cardiovascular risk factors, part 11.serum urate, serum
TRANSCRIPT
Dietary Intake andCardiovascularRisk Factors,Part 11.SerumUrate, SerumCholesterol, andCorrelatesThis report presents analyses of relationship:
among serum urate, serum cholesterol, and
nutritional variables including dietary intake
and selected biochemistries among U.S.
adults ages 18-74 and 25-74 years by age,
sex, race, body mass, and selected
behavioral patterns or attributes. These
estimates are based on standardized
examination findings from the national
probability samples of the civilian
noninstitutional ized population examined in
the first National Health and Nutrition
Examination Survey of 1971-75.
Data From the National HealthSurveySeries 11, No. 227
DHHS Publication No. (PHS) 83-1677
U S. Department of Health and Human Servlc~
Publlc Health Service
National Center for Health Statistics
Hyattsville, Md.
March 1983
I
COPYRIGHT INFORMATION
All material appearing in this report is in the public domain and may be
reproduced or copied without permission; citation as to source, however,
is appreciated.
SUGGESTED CITATION
National Center forliealth Statistics, W. R, Harlan, M, D., A, L, Hill, Ph, D.,R. P. Schmouder, and others: Dietary intake and cardiovascular riskfactors, pertll. Serum urate, serum cholesterol, andcorrelates: UnitedStates, 1971-75. Vital arrd Health Statistics. Series 11, No. 227, DHHSPub. No. (PHS) 83-1677, Public Health Service, Washington, U.S.Government Printing Office, March 1983
Library of Congreaa Cataloging in Publication Data
Main entry under title:
Dietary intake and cardiovascular risk factors.
(Data from the national health survey; ser. 11, no. 227) (DHHS
publication ; no. (PHS) 83-1 677)
Supt. of Dots. no.: HE 20.6209:11 /227
Contents: pt. l, Blood pressure correlates,
1.Hypertension-Nutritional aspects. 2. Hypertension-United
States—Statistics. l. Harlan, William R. 11. National Center for Health
Statistics (U.S.) Ill. Series Vital and health statistics. Series 11, Data from
the national health survey; no. 227. IV. Series DHHS publication no,
(PHS) 83-1677. [DNLM: 1. Nutrition surveys-United States, 2.
Cardiovascular diseases—Etiology. W2AN148vk no. 226-227]
RA407.3.A347 no. 227 [RE48] 312’.0973s 82-6(XM 98
[RC685.H8] [61 6.1 ‘32071]
National Center for Health Statistics
MANNING FEINLEIB, M. D., Dr.P. H., Director
ROBERT A. ISRAEL, Deputy Director
JACOB J. FELDMAN, Ph. D., Associate Directorfor Analysis
and Epidemiology
GARRIE J. LOSEE, Associate Director for Data Processingand Services
ALVAN O. ZARATE, Ph. D., Assistant Director forInternational Statistics
E. EARL BRYANT, Associate Director for Interview andExamination Statistics
ROBERT L. QIJAVE, A ctirrg Associate Director forManagement
GAIL F. FISHER, Ph.D., Acting Associate Director forfiogram Planning, Evaluation, and Coordination
MONROE G. SIRKEN, Ph. D., Associate Director forResearch and Methodology
PETER L. FIURLEY, Associate Director for Vital and HealthCare Statistics
ALICE HAY WOOD, Information Ojjjcer
Interview and Examination Statistics Program
E. EARL BRYANT, Associate Director
MARY GRACE KOVAR, Dr. P. H., Special Assistant for Data
Policy and Analysis
Division of Health Examination Statistics
ROBERT S. MURPHY, Director
JEAN ROBERTS, Chiej Medical Statistics Branch
SIDNEY ABRAHAM, ChieJ Nutrition Statistics Branch
KURT R. MAURER, C’hieJ Survey Planning andDevelopment Branch
RITA WEINBERGER, ChieJ fiogramming Staff
Dlata Processing and Services Program
GARRIE J. LOSEE, Associate Director
ALAN K. KREGER, ChieJ Computer Users Stafl
Division of Data Services
JAMES C. JACKS, Ph. D., Director
PHILLIP R. BEATTIE, Deputy Director
DAVID L. LARSON, ChieJ Health Examination FieldOperations Branch
Cooperationof the U.S. Bureauof the Census
Underthe legislationestabliifsingthe National Health Survey, the PublicHealth SeMcc is authorized to use, insofaras possible.the servieesor facilitiesof (otherFederal, State, or privateagencies.In accordancewith speeiticationsestablishedby the National Center for Health Statistics, the U.S. Bureauof theCensus participated in the designand seketion of the sample and carried outthe household interviewstage of the data coketion and certain parts of thestatiitieal processing.—
Foreword andacknowledgments
The National Health and Nutrition ExaminationSurvey (NHANES) is the only source of general U.S.population data that provides a direct link betweenindicators of health and nutritional status and reporteddietary intake information. The Congress providedresources in the Departments of Labor and Health,Education, and Welfare, and Related Agencies Appro-priation Bill, 1980 to the National Center for HealthStatistics (NCHS) to fund an initiative to undertakemore detailed analyses of nutrition-related healthproblems as measured in the first NHANES. As partof this initiative, the Division of Health ExaminationStatistics funded a contract (No. 223-79-2090) with theSchool of Public Health at the University of Michiganto examine relationships among dietary intake andcardiovascular risk factors.
The approach and depth of analysis presented inthis report differ from most reports from the Divisionof Health Examination Statistics. This report is basedon a statistical rather than a descriptive presentation ofthe data. The tables and text present the results of aregression analysis that incorporates the fall designeffect of the complex survey.
Cognizant that the underlying assumptions oftraditional statistical analyses are violated to someextent, the degree of which is unknown, the authorsand NCHS staff jointly determined that the assump-tions made in the analyses presented in this report arereasonable in light of present knowledge. In addition,the authors have presented throughout the text andtechnical appendix material concerning appropriatequaWications that the reader should consider ininterpreting the results and conclusions presented.
Jean Roberts, the NCHS Project OfEcer, wasinstrumental in bringing the project to a successfulcompletion. Her continuing interaction with the au-thors and their cooperation throughout the projectaided the Center in dealing with diflicult and highlytechnical analytic issues not faced previously byNCHS.
Robert S. MurphyDirectorDivision of Health Examination Statistics
...Ill
Contents
Foreword and acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Highlights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Dietary intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Medical and laboratory examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Serum cholesterol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Body mass index and skinfold thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Blood pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Dietary intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Behavioral and demographic variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Clinical hematology and biochemistries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Multivariate analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Serum urate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Age, sex, race, body mass index, and geogriiphic region, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Dietary intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Demographic and behavioral variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Clinical hematology and biochemistries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Multivariate analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...!!.
List ofdetailed tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . !>.
Appendixes
1. Statistical notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Il. Definitions ofselected terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
List of text figures
1.Mean serum cholesterol levels in quintile strata of body mass index for adults”United States, 1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
iv
8-74 years of age by race and sex:. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
...Ill
1
2
3344
666
1010111214151517202021
22
27
30
7891
7
2.
3.
4.
5.
6.
7.
8.
9.
10.
Mean serum cholesterol levels in selected strata of body mass index for white and black males and females byage: United States, 1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean serum cholesterol levels in quintile strata of total skinfolds for adults 18-74 years of age by race and sex:United States, 1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean serum cholesterol levels in percentile strata of serum calcium levels within BMI quartile strata for males andfemales 25-74 years of age: United States, 1971-75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean serum cholesterol levels in percentile strata of serum magnesium levels within BMI quartile strata for malesand females 25-74 years of age: United States, 1971-75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean serum urate levels in selected strata of body mass index for white and black males and females by age:United States, 1971-75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean serum urate levels in quintile strata of body mass index for adults 25-74 years of age by race and sex:United States, 1971-75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean serum urate levels in quintile strata of total skinfolds for adults 25-74 years of age by race and sex: UnitedStates, 1971-75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean serum urate levels in selected strata of weekly ethanol consumption within BMI quartile strata for males andfemales 25-74 years of age: United States, 1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean serum urate levels in percentile strata of serum glutamic oxalacetic transaminase (SGOT) within BM I quar-
8
9
12
13
16
17
18
19
tile strata for males and females 25-74 years of age: United States, 1971-75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
List of text tables
A.
B.
c.
Standardized beta coefficients and standard errors for regression of serum cholesterol on selected variables, bysex foradults ages 25-74 years: United States, 1971-75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Bivariate correlation coefficients, standardized beta coefficients, and standard errors for regression of serumcholesterol on computed diet scores, age, and body mass index, by sex for adults ages 25-74 years: UnitedStates, 1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Standardized beta coefficients and standard errors for regression of serum urate on selected variables, by sex foradults ages 25-74 years: United States, 1971-75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Symbols
--- Data not available
. . . Categov not applicable
Quantity zero
0.0 Quantity more than zero but lessthan
0.05
z Quantity more than zero but lessthan
500 where numbers are rounded to
thousands* Figure does not meet standards of
reliability or precision (more than
30 percent relative standard error)
# Figure suppressed to comply with
confidentiality requirements
Cardiovascular RiskFactors, Part 11.SerumUrate, SerumCholesterol, andCorrelatesby William R. Harlan, M. D.; Alan L. Hull, Ph.D.; Robert P.Schmouder, M. P.H.; Frances E. Thompson, M. P.H.;Frances A. Larkin, Ph.D.: and J. Richard Landis, Ph.D.,School of Public Health, University of Michigan
IntroductionDiet and nutritional status have been related to the
development of arteriosclerotic disease, primarily is-chemic heart disease, when international comparisonsare made. lJ However, the relationship between dietarypatterns and ischemic heart disease becomes lessrobust or disappears when comparisons are madewithin countries.1-s The existence of a relationship inthe United States is particularly controversial withconflicting dietary advice being offered to the public.However, it is generally accepted that personal attri-butes relate to the risk of developing heart disease.These personal characteristics or risk factors includeserum cholesterol, blood pressure, and cigarette smok-ing, although other variables including body weight,behavioral pattern, and serum urate have been impli-cated as well.l~bWhat remains controversial is theinfluence of diet and nutritional status on these riskfactors and on the development of disease. Thepersonal and public health implications are important.Clear evidence for a linkage would provide a scientificrationale for public health action and a stronger basisfor therapeutic intervention by health providers.s~g
The first National Health and Nutrition Examina-tion Survey assessed the nutritional status and dietaryintake of a representative sample of the U.S. civiliannoninstitutionalized population, and data from thissurvey provide an opportunity to explore relationshipsamong measures of nutrition, cardiovascular riskfactors, and cardiovascular health. Although the pri-mary intent of the survey was to ascertain the
nutritional and health status of the U.S. populationand not to relate specific nutritional variables tocardiovascular risk factors and disease, the data arebroad in scope and reflect “state-of-the-art” reliabilityand validity. The data may be used to test specifichypotheses regarding nutritional and cardiovascularrelationships, but some limitations should be noted.The survey is cross-sectional and accurately reflectsonly current measures of nutritional and health statusand does not disclose prior eating behavior or pre-existent or future disease conditions. Thus, the sur-veyed population can be characterized only in terms ofcurrent associations.
The primary focus of analysis must, therefore, bedirected to exploring relationships between nutritionalpatterns and cardiovascular risk factors rather thanattempting to predict future disease outcome orexplain past episodes. In this report, attention isdirected to associations between nutritional variablesand risk factors associated with development of coro-nary artery disease. The nutritional variables werederived from food frequency questionnaires, 24-hourdietary recall histories, physical examination findings,and serum biochemical measurements on the examin-ees. Additionally, personal habits and attributes, suchas cigarette smoking, use of oral contraceptives,psychological well-being, and socioeconomic statuswere examined in the analysis. Serum cholesterol andserum urate are the dependent variables in this report.Blood pressure and adiposity are also important riskfactors but are addressed as dependent variables in acompanion analysis of the survey data.7
1
Highlights
Relative weight or body mass (weight/helghtz )was found to be an important independent predictor ofserum cholesterol and serum urate levels in U.S.adults. It was also an important predictor of bloodpressure levels of U.S. adults as determined in the firstNational Health and Nutrition Examination Survey of1971-75. This variable, age, and sex accounted for themajority of explained variance. No dietary variablesfrom either the 24-hour dietary recall or the foodfrequency questiomaire had important or consistentassociations with serum cholesterol or serum uratelevels except for reported use of alcohol, which wasrelated directly to serum urate concentration. Other
attributes and behavioral patterns (such as smoking,psychological well-being, and oral contraceptive use)were not significantly or independently related to thedependent variables. Several unsuspected but provoca-tive associations were found with serum biochemis-tries. Serum calcium and magnesium levels ancl serumglutamic oxalacetic transarninase were directly andindependently related to serum cholesterol and serumurate, and these latter variables were interrelated.Cross-sectional dietary data from relatively homoge-neous subgroup populations were not related to serumcholesterol and serum urate, although other nutrition-al biochemical measurements were related.
2
Methods
Data collection in the first National Health andNutrition Examination Survey (NHANES I) wasbegun in April 1971, and the initial survey wascompleted in June 1974. The sample design, plan ofoperation, and details of data collection have beenpublished,s$ and only features of the study pertinent tothe present analysis are described here. Teams of theNational Center for Health Statistics traveled to 65primary sampling units (PSU’S). A PSU is typically acounty or set of contiguous counties. The teamsincluded professional and paraprofessional medicaland dental examiners, along with technicians, inter-viewers, and other staff. The selected sample personsfor whom appointments could be made were broughtinto specially constructed mobile examination centersthat were moved into a central location in each PSUarea.
Of the 28,043 sample persons selected to represent194 million persons ages 1-74 years in the U.S.population, the program examined 20,749 or 74percent of the sample at the 65 locations visitedbetween April 1971 and June 1974. This is aneffective response rate of 75 percent when adjustmentis made for the effect of oversampling among preschoolchildren, women of childbearing age, the poor, and theelderly.
A subsample of approximately 20 percent (3,854)of those ages 25-74 years in the initial sample receiveda more detailed examination. An additional sample of3,059 persons ages 25-74 years was identified toaugment the data collected during the detailed exami-nation in April 197l-June 1974. This augmentationsurvey (NHANES IA) was conducted in 35 additionalPSU’S between July 1974 and September 1975. Theseadditional groups are referred to as the “detailed” and“augmentation” components, respectively. Several ad-ditional measures were collected on persons includedin the augmentation survey. 10
Data presenting breakdowns by race are based onfindings from a sample of 27,730 white and blackpersons, of whom 20,514 were examined. Estimates in
this report are based on weighted observations; that is,the data obtained for the examined persons are inflatedto the level of the U.S. population from which thesample was drawn using the appropriate weights toaccount for both sampling fractions and responseresults. (See appendix I). Analyses included in thisreport utilized the largest number of persons on whomdata were available. Some data were available only onthe general and detailed components (20,749), somewere available on the detailed and augmented compo-nents (6,9 13), others only on the augmented compo-nent (3,054), and some only on the detailed component(3,859). The text and tables indicate the data sourcesused for each analysis.
Dietary intakeDietary intake was a primary data source for the
relationships described in this report. The informationfor dietary intake was determined through 24-hourdietary recall and 3-month food frequency recall. Adietary interview was conducted with each sampleperson to obtain information about his or her totalfood and drink consumption during the 24 hours-midnight to midnight-preceding the interview. Thiswas followed by questions about the frequency of foodintake for the preceding 3 months.’ 1.12The parent orother adult responsible for a child’s feeding providedinformation about preschool children. Usually boththe parent and child were interviewed for subjects ages6-12 years.
The dietary interview lasted approximately 20minutes (with a maximum allowance of 30 minutes)and usually was administered in the mobile examina-tion center. A small percent of the interviews tookplace in the subjects’ homes.
Food portion models were used to assist therespondent in estimating amounts of foods consumedfor the 24-hour recall. Models developed for anothersurvey were used with slight modifications. 13A com-puter program was used to determine nutrient values
3
of foods consumed. The computer program to processfood recall data for nutrient contents was adaptedfrom one developed and used in the Ten-State Nutri-tion Survey and was based on a program developedoriginally at Tulane University. 14The program uses thenutritive values of food items appearing in the U.S.Department of Agriculture Handbook No. 8 (1963),Table 1,15as well as information from other sources.Because of the constantly changing food supply,nutrient composition values for new food productswere added or updated continually according toinformation provided by the U.S. Department ofAgriculture, food processors, and manufacturers.
Dietary intake measurements considered in thisreport include the following:
1. Frequency of consumption of the following foodgroups: butter and margarine, dried beam andpeas, breads and cereals, dairy foods (whole milk,eggs, and cheese and cheese dishes), meat andpoultry, total fruits and vegetables, desserts andsweets, candy, sweetened beverages, coffee andtea, and snack foods.
2. Frequency of consumption of the following spe-cial food groups: complex carbohydrate and fiber-containing foods, high fat foods, meats, sweets,snack foods, coffee and tea, alcohol, proportion ofcalories as total fat, ratio of saturated to unsatu-rated fats, cholesterol intake, cholesterol plussaturated/unsaturated fats (linoleic acid only),and fat intake (calculation of dietary cholesterolscores using Keys et al. lband HegstedlT).
3. Proportion of calories as carbohydrate or protein.
4. Relative dietary purine intake.
Medical and laboratory examination
Complete descriptions of the clinical examination,body measurements, and laboratory assessments areavailable,g,lo,lg,lgand only aspects pertinent to thepresent analysis are described here. A medical historyquestionnaire was completed by participants ages12–74 years. This instrument requested information onhealth habits and general medical status, as well asspecific answers regarding known disease conditionsand medical treatments. The medical history question-naire was reviewed by the examining physician beforethe scheduled examination.
All examinees received a physical examinationwith emphasis on nutritional aspects. Blood pressurewas recorded in the sitting position near the beginningof the examination for persons ages 6-74 years. Therecommendations of the American Heart Associationwere followed, and a complete discussion of themeasurement techniques and sources of variabilityy anddiagnostic error is available.TAfter the examination,the physician used standard diagnostic codeszoto
classify the presence of disease as observed. Bodymeasurements including height, weight, and skinfoldthickness were made by specially trained techniciansusing equipment designed for the study and checkedweekly and before each examination stand com-menced. IS.Z1,ZZ
The detailed sample exarninees and those in theaugmentation survey also completed supplementalquestionnaires related to arthritis, respiratory (disease,and cardiovascular disease. These examinees alsoreceived a more detailed examination related to theseconditions. A detailed history on tobacco use, includ-ing duration and amount of cigarette smoking, wasobtained for this group. These examinees also com-pleted the General Well-being Questionnaire.zs
Laboratory assessments of adults in the generalexamination included hematologic examinations andnutritional biochemistries on serum and urine speci-mens. On the detailed and augmentation samples thefollowing clinical biochemistries were performed onblood samples from nonfasting examinees: totall biliru-bin, serum glutamic oxalacetic transaminase (SGOT),alkaline phosphatase, calcium, phosphorus, and uricacid. Details regarding examinee preparation, samplecollection and standardization, and analytic proce-dures are described in detail elsewhere.lg Serum choles-terol determinations on blood samples from nonfastingexaminees in the general, detailed, and augmentationnational surveys were made by the Lipid Standardiza-tion Laboratory, Centers for Disease Control, PublicHealth Service, Atlanta, Georgia. The method of Abellet al.zdwas modified for a semiautomated productionline.zsSGOT, sodium, and potassium were analyzedindividually, and calcium, phosphate, uric acid, andcreatinine were analyzed on a Techicon SequentialMultiple Analyzer (SMA) 12/60.19
Analysis
Statistical considerations and procedures for statis-tical analysis are detailed in appendix I. Definitions ofselected terms, including those related to the statisticalmethods, are given in appendix II. The weightedsample and sample design factors were considered inall statistical analyses presented in this report. Thegeneral analytical approach was to screen initially forsignificant relationships among nonnutritional vari-ables as a means of identifying interrelationships thatcould confound the further analyses. Relationshipswere then sought between nutritional and nonnutri-tional variables, and, where necessary, potentiallyconfounding variables were controlled. For analysis,the population was divided into age, race, and sexgroups. The category of “other” racial groups was notconsidered in further analyses because of the smallnumbers and the heterogeneityy of the group. Thefollowing age ranges were used: 18-24, 25–34, 3544,
45-54,55-64, and 65-74 years. Each age, race, and sexgroup was examined separately.
In general, the independent variable was dividedinto strata bounded by appropriate percentile cut-offpoints, and means and standard errors were deter-mined for the dependent variable within those strata.Because body mass (weight/height2) was consistentlyrelated to many of the dependent variables, quintilestrata of this variable were used to control for thisconfounding influence. When significant and apparent-ly important interrelationships between nutritional andnonnutritional variables were discovered, potentiallyconfounding relationships were examined by control-ling for the confounding variable and by use ofmultivariate analysis. Apparent relationships were
assessed in three ways before inferring biologic impor-tance to the relationship. Tests of statistical sig-nificance were used to contrast values for the depend-ent variables within strata. A probability of 5 percentor less that the finding was the result of chance wastaken as statistically signiikant. The relationships wereexamined for consistency within age, sex, and racegroups. Finally, the quantitative dilTerences were usedto arrive at inferences regarding the biologic impor-tance of relationships.
Each independent variable found to have a sig-nificant and consistent relationship to the dependentvariable on univariate analysis was entered into amultiple regression analysis against the dependentvariable.
5
Findings
Serum cholesterolBody mass index and skinfold thickness
Serum cholesterol levels for adults have beendescribed by age, sex, and race in an earlierreport.za Serum cholesterol concentration was relatedto body mass index (Quetelet’s index orweightfieightz ). In the present report, national esti-mates for the population were separated into quintilestrata of body mass index (BMI), and the serumcholesterol levels were compared for those in eachquintile strata (tables 1 and 2 and figures 1 and 2). Forthe total group of males and for white and blacksubgroups, successively greater quintile strata of BMIwere associated with higher serum cholesterol levels.The progression of serum cholesterol levels withhigher BMI was less consistent for black males,reflecting in part greater variances and smaller samplesizes. The mean differences in serum cholesterolbetween the highest and lowest quintile strata of BMIwere 31 milligrams/deciliter (mg/dl) for all males, 30mg/dl for white males, and 39 mg/dl for black males.Serum cholesterol is higher with age in men through64 years and then lower at ages 65–74 years.2cTheinfluence of age on cholesterol was also apparentwithin each quintile strata of BMI (table 1). Withineach age range, the relationship of serum cholesterol toBMI was not progressive, although statistically sig-nificant cholesterol differences were found between thelowest and highest quintile strata of BMI, and thesedifferences were relatively consistent, ranging from 19to31 mg/dl.
Similar relationships were found for women (table2 and figure 1). Mean serum cholesterol was higher forsuccessively greater quintile strata of body mass indexfor the total group and for white and black females.The differences in mean serum cholesterol between thelowest and highest quintile strata were 36 mg/dl, 36mg/dl, and 34 mg/dl for all females, white females,and black females, respectively. Within each quintilestrata of body mass index the mean serum cholesterol
tended to be lower in black women than white. Whencross-classified by age and BMI quintile strata, theindependent effect of age and BMI on cholesterolconcentration was apparent (figure2). Within each quintilestrata, serum cholesterol increased with age, althoughthe differences were less consistent in the 55-[64- and65-74-year age groups. The magnitude of BMI effecton cholesterol was slightly less for females than maleswhen age was controlled, although the ag[: effectwithin quintile strata of BMI was similar for bothsexes. Therefore, for both men and women, body massindex has an influence on serum cholesterol concentra-tion that is independent of age, sex, and race.
Skinfold thickness is an indirect measure of subcu-taneous fat tissue that correlates well with bodyadiposity and, to a lesser degree, with body massindex.zT~Z8For analysis, the skinfold thicknesses fromtwo sites, the triceps and subscapular areas, werecombined. This combination of measurements affordsassessment of limb (triceps) and truncal (subscapular)adiposity and provides a more representative sample ofsubcutaneous fat. The relationships between skinfoldthickness and serum cholesterol levels were similar tothose for BMI and serum cholesterol (tables 3 and 4).At progressively greater quintile strata of skinfoldthickness, serum cholesterol concentration was higher.This effect was observed for males and females, whiteand black persons, and generally for each adult agerange (figure 3). The magnitude of serum cholesteroldifferences between the highest and lowest quintilestrata was similar when skinfold thickness was theindependent variable to that found when BMI was theindependent variable. The association between skinfoldthickness and serum cholesterol was independent ofsex, race, and age.
Because body mass has a pervasive influence onserum cholesterol concentration, weight/heightz wasused in subsequent analyses to control for this possibleeffect. This parameter was selected because the (effectisequivalent to that of skinfolds, and it is a readilyavailable clinical assessment.
6
V~Whiteferndes
252
240
228
192
180
0
— T
T
T
T
T
0-19” 20-39 40-59
T
T
0 Black males
I lBlack females
I 95-percentconfidence limit
60-79 80-100
Body mass index quintile strata
—
Figure 1. Mean serum cholesterol levels in quintile strata of body mass index for adults 18-74 yeers of age by race and sex
United States, 1971-74
7
~ 285
I 250
215
180
145
320
285
250
215
180
145
320
White males White females
Body mass, indexpercentiles
— 0-19
--40-59-.-80-100
I I I I I I I I I 1 I 145 ~J
20 25 30 35” 40 45 50 55 60 65 70
Age in years
320
81ack malas
..
Y’~u 145
20 25 30 35 40 45 50 55 60 65 70
Age in years
20 25 30 35 40 45 50 55 60 65 70
Age in years
81ack females
~-l20 25 30 35 40 45 50 55 60 65 70
Age in years
Figure2. Mean wrumcholes@rol levels unselected strata of body rnassindex forwhite and black males and females byagw
United States, 1971-74
8
252
240
228
216
204
192
180
0
TI
T
T
0-19 20-39
T T
m White males
Izz2White females
nBlack males
nBlack females
I
TT
95-percent
confidence limit
TT
40-59 60-79 80-100
Total skinfold thickness quintile strata
Figure 3. Mean serum cholesterol levels in quintile strata of total skinfolds for adults 18-74 years of age by race and sex-
United Statas, 1971-74
9
Blood pressure
Serum cholesterol levels were determined for fourstrata of systolic and diastolic blood pressure for malesand females (tables 5–8). Cutoff points for the stratawere the 15th, 50th, and 85th percentiles. Becausetherapy for hypertension may raise plasma lipids,examinees reporting treatment for blood pressure wereexcluded from analysis.zgFor systolic pressure (tables 5and 6), there were statistically significant differencesbetween mean serum cholesterol concentrations in thelower (0-15) percentile stratum and upper (85-100)percentile stratum of systolic pressure. The higher thepressure, the greater the serum cholesterol. Thisrelationship was observed for both sexes and bothraces, and the mean difference between the lowest andhighest percentile strata was 25 mg/dl for males and46 mg/dl for females. When age was controlled, therelationship persisted but the differences were less andtended to be greater in younger adults and decreasing(and even reversing in females 65-74 years) in thoseover 35 years of age. The relationship persisted whenbody mass index was controlled. Mean differencesbetween low and high strata ranged from 11 to 63mg/dl and were greatest in the lowest quartile stratumof BMI. When sex and age were controlled, similartrends were noted, but the differences were smaller orinconsistent. However, the numbers of examineeswithin the groups were small and the variabilityrelatively large.
For diastolic blood pressure, the relationships weregenerally similar to those for systolic pressure (tables 7and 8). Higher diastolic pressure was associated withhigher mean serum cholesterol, and this relationshippersisted when race, sex, and body mass index werecontrolled singly. However, in those ages 45 years andover, a positive relationship was no longer present. Themagnitudes of mean serum cholesterol differenceswhen stratified by diastolic pressure were similar tothose when systolic pressure was used for strati-fication. However, controlling for age and sex togetherled to a dissipation of the relationship betweendiastolic pressure and serum cholesterol.
Dietary intake
Reported dietary intake was separated into catego-ries, and the relationships between the dietary intakecategories and serum cholesterol concentration weredetermined. Particular attention was directed to di-etary intake of total fat, saturated/unsaturated fats,and cholesterol because of the reported associationbetween these nutrients and serum cholesterol lev-els.1’sFrequency of food consumption and a detailed24-hour food intake were analyzed separately. Ingeneral, there were noships.
The frequency of
10
consistent or strong relation-
fatty food consumption was
determined from the food frequency questionnaire,which describes the number of times a particular foodwas consumed (table 9). For the total group and forboth races and sexes, there was an inverse relationshipbetween reported frequency of fatty food consumptionand serum cholesterol. This trend persisted when bodymass index was controlled. However, controlling forage or age and sex dissipated any consistent trends.Tlhis is not surprising because of conflicting trends fordietary intake and serum cholesterol in progressivelyolder groups. Serum cholesterol is higher in oldergroups, while dietary fat intake is lower. 11,26Therefore,an~inverse relationship would be expected when allages are considered, but when age is controlled, therelationship might change. No consistent or importantrelationships were found between serum cholesterolconcentration and the following reported food frequen-cies: complex carbohydrate (including fiber), fattyfood/complex carbohydrate ratio, coffee and tea con-sumption, alcohol intake, and refined sugar intake.
Data from the 24-hour dietary recall were used toexplore relationships between specific proportions ofnutrient intake and serum cholesterol levels. The 24-hour dietary recall provides a quantitative assessment,but this is representative of only one brief pmiod.However, one would expect a direct relationshipbetween food frequency reports and 24-hour recall atthe extremes of ingestion of particular foods if thle 24-hour recall is representative. The energy consumed bymale and female respondents for one 24-hour periodwas divided at the 15th, 50th, and 85th percentiles(tables 10 and 11). For males, there was an inverseassociation between total dietary calories and serumcholesterol values when the total group and whitemades were considered. The same direction of associa-tion was present for black males and for quartile strataof BMI, but not all associations were statisticallysignificant.
When age groups were contrasted, no clear pat-terns were present, and the mean cholesteroldifferences between high and low energy strata werenot significant except for the youngest age group(18-24 years). Females (table 11) had a fairly consist-ent pattern between total caloric intake and meanserum cholesterol values. As noted for males, thepattern was inverse and persisted when sex, race, andage were controlled, although the level of associationand the magnitude of cholesterol differences betweenthe highest and lowest energy groups varied considera-ble.
Total dietary fat intake, as obtained from the. 24-hour dietary recall, tended to be inversely related tocholesterol concentration except when age was con-trolled (table 12). The pattern resembled that of totalcalories, which would be expected. The findings weresimilar for females (table 13), except that a weak butpersistently inverse relationship was found when agewas controlled.
The proportion of total calories contributed by fatwas determined from the 24-hour recall data, whichprovided total caloric and total fat intake (table 14).There were no statistically significant or consistentdifferences in serum cholesterol levels when proportionof fat calories was stratified.
Total polyunsaturated fatty acids were not avail-able in the food composition tables, but linoleic acid,which constitutes approximately 90 percent of thetotal polyunsaturates in the U.S. diet, was availablefrom the 24-hour recall. A ratio of linoleic to totalsaturated fatty acids was developed and used as asurrogate measure of polyunsaturated/saturated fattyacids (table 15). When stratified by this measure, onlysmall, inconsistent, and statistically insignificantdifferences were found for serum cholesterol values.
Dietary cholesterol represents a fraction of thetotal fat consumed, but intake does not necessarilyparallel total fat intake. Cholesterol intake differed formales and females. No clear pattern of association wasfound for either males or females, and there were nostatistically significant differences between serum cho-lesterol values at different levels of dietary cholesterol(tables 16 and 17).
Dietary sodium and potassium were calculatedfrom the 24-hour recall. The ratio of dietary sodium topotassium was found to be related to bloodpressureT and was, therefore, applied to serum choles-terol. An inverse pattern of association was found forserum cholesterol (table 18). When mean serumcholesterol values were compared in the low and highsodium/potassium strata, a statistically significant (p< 0.001) difference was found for the total group andfor females and white respondents. The trend was inthe same direction but less strong for males and blackrespondents. When age was controlled, the associationwas no longer statistically significant but remainedinverse.
Other nutrients from the 24-hour dietary recallwere analyzed, but no statistically significant, consist-ent associations were found.
Behavioral and demographic variables
The recent use of oral contraceptive agents wasassociated with significantly higher serum cholesterollevels for some females (table 19). White femalesreporting current use of oral contraceptive agents hadsignificantly higher serum cholesterol concentrationsthan those reporting no use within the past 6 months.The differences were most prominent for white femalesages 18-24 years (amounting to 20 mg/dl) and weresmall for those ages 25-34 and 354-4 years (5 mg/dl).Information on the use of oral contraceptive agentswas limited to women ages 18-44 years.
For black females, higher serum cholesterol valueswere observed to be higher among users only for the
18-24-year age group, and the pattern was reversed inthose ‘ages 25~ years, but the numbers of oralcontraceptive users were relatively small in these oldergroups. When body mass index was controlled, thepositive association between oral contraceptive use andserum cholesterol was strengthened for the youngergroup (ages 18-24 years) and tended to shift theobserved association from negative to positive forfemales ages 25-44 years.
Socioeconomic status was scored using educationalattainment and income levels.30This scale uses meas-ures of income and educational attainment rather thanoccupation and was created so that the five categoriesfit a Gaussian distribution with approximately 15percent of the sample in the two extreme categories.The lowest category had income of $4,000 or less andeducation of grade school or less, while the highestcontained people with college education and income of$10,000 or more. Serum cholesterol levels were sig-nificantly higher in the lowest socioeconomic class forthe total group, for females, and for white persons(table 20), and the mean differences between the lowestand highest strata ranged from 15 mg\dl for the totalgroup to 27 mg/dl for females and 17 mg\dl for whiterespondents. This inverse pattern persisted when bodymass index was controlled. When age and sex werecontrolled, a significant inverse trend was found forfemales only, while males tended to have a directrelationship. However, the number of respondentscategorized in the low category by age was small andthe variability of serum cholesterol values was great.
When serum cholesterol was contrasted across thefour regions of the contiguous United States, small butstatistically significant differences were found amongthe regions (table 21). For males, cholesterol levelswere higher in the Northeast; and for females, serumcholesterol levels were observed to be higher in theNortheast and lowest in the West. Levels were highestfor white respondents living in the Northeast and wereobserved to be highest for black respondents living inthe Midwest. The differences by region were small andrarely exceeded 5 mg/dl. Controlling for age and BMIdid not lead to a consistent pattern for males orfemales.
Respondents in the detailed and augmentationsurveys completed a survey of General Well-Being(GWB).’S The higher the score, the greater the self-rated physical and psychological status. For the totalgroup of males and for all white males the higherGWB scores were observed to be associated withhigher serum cholesterol levels (table 22). The reversepattern was found for black males, but the numberswere small. None of these differences was statisticallysignificant. No consistent trends were found when ageor body mass index were controlled, and no consistenttrends were found for females (table 23).
Cigarette smoking was analyzed, but no relation-ship to serum cholesterol concentration was found.
11
Clinical hematology and biochemistries
Hemoglobin concentration was related
function, although the main application is determina-tion of major acute injury to tissues such as liver and
to serum cardiac muscle. For the total group and for both sexescholesterol (tables 24 and 25). Significant meandifferences in serum cholesterol levels were foundbetween the lowest and highest strata for males and forwhite males, but the differences for black males werenot significant. Within all age ranges except 55–64years, a similar pattern was found (table 24). However,when body mass index was controlled the differencesdiminished. Similar relationships were found for fe-males but the consistency and magnitude of differenceswere greater (and in general significant). The meandifferences between the lowest and highest hemoglobingroups averaged 20 mgidl for all females and 10mg/dl for all males.
Serum glutamic oxalacetic transaminase (SGOT)is often used as a test for subtle impairment of liver
and both races, higher strata of SGOT were associatedwith significantly higher mean serum cholestercd levels(table 26). The differences in cholesterol between thehighest and lowest strata of SGOT in these groupsranged from 6 to 24 units/milliliter. The relationshipwas more prominent for females. When age wascontrolled, a clear positive relationship was apparentonly for ages 25–54 years. Controlling for body massindex did not change the positive relationship forfemales, but it was less consistent for males after bodymass index was taken into account.
Serum calcium level distribution in the populationwas divided into strata at the 15th, 50th, and 85thpercentiles (table 27 and figure 4). A consistent andlarge difference in mean serum cholesterol was found
Figure 4. Mean sarum cholesterol levels in percentile strata of serum calcium levels within BMI quartile strata for males and females 25-74 yearsof age: United States, 1971-75
12
when the population was stratified by serum calciumlevels. This relationship was positive with higherserum cholesterol being associated with higher serumcalcium and was generally progressive through allstrata of serum calcium concentration. The relation-ship was not altered when sex, race, age, body massindex, or age and sex were controlled. The magnitudeof mean differences in serum cholesterol between thelowest and highest strata of serum calcium wasconsiderable, averaging 29 mg/dl for the group andranging from –2 (females ages 65-74 years) to 47mg/dl (males ages 55-64 years). Serum inorganicphosphate and the serum calcium\phosphate ratiowere analyzed, but no consistent relationships werefound.
Serum magnesium levels were stratified using
cutoff points of 1.555, 1.685, and 1.825 mg/dl whichare at the 15th, 50th, and 85th percentiles (table 28).There was a consistent, statistically significant positiverelationship between mean serum cholesterol concen-trations in the lowest and highest strata of serummagnesium. For the entire group the mean differencebetween lowest and highest strata amounted to 19mg\dl of serum cholesterol. This positive associationwas found for both sexes, both races, and aftercontrolling for age, body mass index, and age and sex(figure 5). The relationship was generally progressivethrough the strata and ranged from 5 to 8 mg/dl aftercontrolling for these factors.
Serum urate levels were stratified separately formales and females because of the considerably differentlevels for each sex. The cutoff points for adult males
1- T
T
Lessthan 1.555 1.555-1.684
Male
TI
T-
1.685-1.824
Female
Body massindex
m First quatiile
n Second quartile
~ Third quartile
m Fourth quartile
~ 95-percentconfidence limit
T
more
Serum magnesium percentile strata (milligrams/deciliter)
1.825 or Lessthan 1.5551.555-1.684 1.685-1.824 1.825 or
more
Serum magnesium percentile strata (milligrams/deciliter)
Figure 5. Mean serum cholesterol levels in percentile strata of sarum magnesium levels within BMI quartile strata for males and females 25-74 yearsof age: United States, 1971-75
13
were 4.95, 6.15, and 7.55 mg/dl; for females, 3.65,4.65, and 5.95 mg/dl. These are at the 15th, 50th,and 85th percentiles.
There was a positive association between serumurate and serum cholesterol (tables 29 and 30). Thedifferences in mean serum cholesterol were 16 and 22mg/dl (males and females, respectively) between thehighest and lowest strata of serum uric acid. Thisrelationship persisted after controlling for race, age,and body mass index. However, for two subgroups,males in the highest quartile strata of body mass indexand females ages 55–64 years, the mean differenceswere small or reversed, but for other groups thedifferences ranged from 4 to 45 mg/dl .
Multivariate analyses
Variables that were found to have consistent andimportant relationships to serum cholesterol were usedas independent variables for regression on the depend-ent variable, serum cholesterol (table A). Males andfemales were analyzed separately, but race was enteredas a dichotomous variable (1 = white, 2 = black).Because serum urate and calcium were available onlyon examinees in the detailed and augmentation sur-veys, the regression analysis was confined to thisgroup, ages 25–74 years, and comprising 3,134 malesand 3,707 females. Relatively little variance of serumcholesterol was explained; for males the R2 was 0.12,and for females, 0.20. The standard beta weightsprovide an indication of the relative influence of theindependent variables. For males and females, age,body mass index, and serum calcium clearly had themost important effects. Lesser effects were found forserum urate and serum magnesium. There were somebody mass index and socioeconomic status being lessdifferential influences between males and females withimportant in females, although serum urate wasslightly more influential in females.
Despite the lack of clear relationships betweendlietary fat variables and serum cholesterol, when thedlietary variables were assessed individually, thesevariables were used in predictive models of serumcholesterol that were developed by Keys et al. 16andHegsted. 17
The formulas for the diet scores were as follows:
Keys 1.35 (2S – P) + 1.52( C/1,000E)VzHegsted 2.16 S– 1.65P+ 0.0677C
where S = proportion of dietary calories fromsaturated fatP = proportion of dietary calories
from polyunsaturated fatC = cholesterol intake per day (milli-
grams)E = energy intake per day (calories)
For polyunsaturated fatty acids, only linoleic acidintake was available in the nutrient data of NHANESI. However, this fatty acid accounted for over 90percent of all polyunsaturated fatty acids in the U.S.diet at the time of the survey. The value for li,noleicacid was used in the analysis without correction. Ascore for each examinee was calculated using the twoformulas and a correlation determined between thesescores and the serum cholesterol values measuredduring the survey. For each model, only a small,statistically insignificant correlation of less than 0.05w~as found between lipid intake scores and serumcholesterol. To compare the relative contribution ofdietary lipid and physiologic variables to serum choles-terol concentration, the dietary score, body massindex, and age were regressed against the dependentvariable, serum cholesterol (table B). When the betaweights for the independent variables were compared,it was apparent that the dietary score made small and
Table A. Standardized beta coefficients and standard errors for regression of serum cholesterol on selected variables, by sex for adultsages 25-74 years United States, 1971-75
Male Female
(N = 3, 134) (N = 3,707)Variable Multiple R = 0.34 Multiple R = 0.45
Beta Standard errorof beta Beta
Standard errorof beta
Body mass index .........................................................................Age~................................................................................................Race .............................................................................................Socioeconomic status .................................................................Systolic blood pressure ..............................................................General well. being .......................................................................Serum urate ..................................................................................Serum calcium .............................................................................Serum glutamic oxalacetic transaminase.., .............................Serum magnesium .......................................................................
0.120.280.040.05
-0.040.030.030.24
-0.050.08
0.030.030.020.030.030.020.030.020.020.02
0.050.37
-0.01-0.01-0.010.000.060.140.050.05
0.030.030.020.020.030.020.020.020.020.02
NOTE: For race, 1 = white, 2 = black, other racial groups excluded,N = number of examinees.R = multiple correlation coefficient.
14
Table B. Bivariate correlation coefficients, standardized beta coefficients, and standard errors for regression of serum cholesterol oncomputed diet scores, age, and body mass index, by sex for adults ages 25-74 year.% United States, 1971-74
Male (N = 1,753) Female (N = 3,424)
Variable
Keys et al. diet score ............................................Hegsted diet score .................................................Age .............................................................................Body mass index .....................................................
Regression of serum cholesterol on computeddiet scores, age, and body mass index
Keys et al. diet score ............................................Age ........................................................................Body mass index ................................................
Hegsted diet score .................................................Age ........................................................................Body mass index ................................................
50.0 0.58 0.02 0.0357.8 0.82 -0.03 0.0339.4 0.27 0.26 0.0325.9 0.11 0.14 0.03
Multiple R = 0.29
Beta
0.010.250.13
-0.020.250.13
Standard errorof beta
0.020.020.020.020.020.02
48.7 0.47 0.02 0.030.66 0.00 0.02
% 0.21 0.37 0.0224.9 0.12 0.16 0.02
Multiple R = 0.39
BetaStandard error
of beta
0.02 0.020.36 0.020.11 0.020.00 0.020.36 0.020.11 0.02
Bivariate correlation between Keys, et al. Correlation Standard error Correlation Standard errorand Hegsted diet scores coefficient (r) of r coefficient (r) of r
0.83 0.01 0.88 0.01
NOTE N = number of examinees.R = multiple correlation coefficient.r = simple correlation coefficient.
not statistically significant contributions to explanationof variance. Compared with the major predictors ofserum cholesterol levels, age and body mass index, thecontribution of dietary lipid was not important.
These findings can be contrasted with those report-ed by Shekelle and otherss in a recent longitudinalstudy of men employed at the Western ElectricCompany, who were ages 40-55 years on entry intothe study. The computed dietary scores were consist-ently higher in the study of Shekelle et al. SSome of thedifference may be attributable to the different nutrientdata sources used to compute polyunsaturated fatcontent. In the Western Electric study, archadomicacid was estimated but not in NHANES I. Thecorrelation between the initial dietary score and serumcholesterol was significant and slightly but not sig-nificantly higher in the Western Electric study than thenational estimates from NHANES I shown in table B.When dietary scores and other relevant variables wereregressed against serum cholesterol concentration, thebeta weights for body mass index and age were higherin NHANES I than in the Western Electric study, andthe beta weights for dietary score were lower. Thedifferences may be attributed to a number ofdtierences in the study, including selected population,age range, variability of dietary reporting, failure toinclude all polyunsaturated fatty acids in NHANES I,and possibly lack of comparability in the nutrient databank used for the two studies.
Serum urate
Age, sex, race, body mass index, and geographicregion
Serum urate concentration was measured in exam-inees of the detailed and augmentation surveys. Thelevels of serum urate by sex and race are presented intable 31 together with selected percentiles in thedistribution of these levels. Levels of serum uric acidwere higher in males than in females at all ages (table31 and figure 6). Mean serum urate concentration wasnot higher in older age groups in males, but there wasan age-related difference in urate levels among females.For both males and females, uric acid concentrationwas slightly but significantly higher in black persons.This racial difference in mean urate concentrationreflected higher levels in the upper distribution of uratefor all ages.
Uric acid concentration was related to body massindex (BMI) in adult males (table 32 and figure 7). Forthe total male group, the serum urate difference was1.2 mg/dl between the lowest and highest quintilestrata of BMI. DiiTerences of similar magnitude werefound for white and black males. There was agraduated increase in serum urate for progressivelygreater quintile strata of BMI for males. A similarrelationship was observed for adult females with amean difference of 1.3 mg/dl in mean serum uratebetween the lowest and highest quintile strata of BMI
15
Body massindexpercentile groups
— 0-19
9.0
r
4.2
t
White males
3 ~J25 30 35 40 45 50 55 60 66 70
Age in years
9.0 r Black males
4.2 F
f
,i
f
!’
3 ~J
25 30 35 40 45 50 55 60 65 70
Age in years
9.0
r
I
---40-59m- m 80.1oo
White females
3 i I I I I I I I I I I2.5 30 35 40 45 50 55 60 65 70
Age in years
9.0
[
81ack females
,/””/’ ●\, #//“
i~
--
3~
25 30 35 40 45 50 55 60 65 70Age in years
Figure 6. Mean serum urate levels in selected strata of body mass index for white and black males and females by age: United States, 1971-75
16
(table 33 and figure 7). This observed relationship wasnot altered by race and age.
Triceps and subscapular skinfold thicknesses weresummed to provide a representative estimate of subcu-taneous adiposity. When these combined skinfoldmeasurements were divided into quintile strata, agraded positive relationship was found for skinfoldthickness and serum urate (tables 34 and 35 and figure8). Controlling for race and age did not impair thisrelationship. The differences in mean serum uratebetween the lowest and highest skinfold thicknessstrata were less than the differences observed withbody mass index, but the differences were statisticallysignificant. Because of the more pervasive influence ofBMI in serum urate concentration, this confoundingvariable was controlled in subsequent analyses.
Mean serum urate concentration was determinedfor the four geographic regions into which the countrywas divided for this national study (table 36). Therewere no significant differences, although serum uratetended to be highest in the Midwest for the total groupand for female examinees and highest in the Northeastfor male examinees. No prominent confounding of thisregional trend was present for race, age, or BMI.
Dietary intake
The food frequency questionnaire provided infor-mation on customary dietary intake patterns. Noconsistent or significant associations were found be-tween serum urate concentrations and the followingfood and beverage groups: fatty foods, complex carbo-
- I
I
I
c1 Black males
lsss White females
n Black females
I 95-percentconfidence limit
T
0-19 20-39 40-59
Body massindex percentile groups
60-79 BO-1OO
Figure 7. Mean serum urate levels in quintile strata of body mass index for adults 25-74 years of age by race and sex: United States, 1971-75
17
hydrateifiber, sugar-containing foods, and coffee-teaconsumption,
Reported alcohol consumption was directly andstrongly related to serum urate levels, The averageweekly consumption of alcoholic beverages was con-verted to ounces of ethanol per week,’ and the follow-ing four groups were developed according to alcoholconsumption: zero ounces /week (OZ/ wk), 0.001-0.999oz / wk, 1.000–6.999 oz / wk, and 7.000 oz / wk or more(table 37 and appendix II). For the entire populationand for each sex, race, age, and BMI-sex subgroup,abstainers had lower mean serum urate levels thanrespondents reporting ethanol intake of 7.000 oz / wkor more (figure 9). The differences were significant forall groups except when divided by age and sex wherethe numbers within each cell were small and the vari-
ability relatively great. Abstainers tended to have lowerserum urate levels than those with the lowest level ofalcohol consumption, but the differences were smalland not always statistically significant. On the otherhand, the differences in mean serum urate betweenabstainers and the heaviest consumers of alcohol wereconsistent, statistically significant, and sizable. Tlhedif-ference for the total group averaged 1.34 mg/dl with theheaviest ethanol consumers having serum urate levels2’7percent higher than abstainers. This magnitude ofdifference was observed for each sex, race, and agegroup, and after controlling for body mass index.
There are two sources of urate in body fluids,exogenous and endogenous. The exogenous source,diet, may contain sufficient purines to provide 3CIto 60millimoles (0.5 to 1.0 gram) of urate per day.31About
8
7
I
T
IL00-19
T
20-39 40-59
Total skinfadd thickness quintile strata
T
White males
D
,.,;, :...,, .,, Black males
IzzzJWhite females
I I Black females
I 95-percent
confidence limit
60-79
T
80-100
Figure 8. Mean serum urate levels in quintile strata of total skinfolds for adults 25-74 years of age by race end sex United States, 1971-75
18
20 percent of dietary purines are destroyed in diges-tion, but the remainder form urate. Foods particularlyrich in purine include glandular meats and meatextracts, and, to a lesser extent, meat, poultry, fish, andlegumes. To test for an association between diet andserum urate, the following three food groups werecombined: meats, fish, and beans from the foodfrequency record. It was assumed that individuals whofrequently consume purine-rich foods, will ingestincreased quantities, but no quantitative measurementwas available. Serum urate levels were not significantlydifferent in strata with high purine consumption thanin those with lower levels of such consumption (table38). Higher consumption tended to be associated withhigher serum urate levels, although the differenceswere slight except in the 35-44-year age group.Moreover, there was no consistent pattern when bodymass index was controlled. Therefore, no consistentassociation was found between purine intake andserum urate levels.
The 24-hour dietary recall was used to explorerelationships between serum urate and the following
variables: total caloric intakq proportion of caloriesfrom alcoho~ proportion of calories from fat, carbohy-drate, and protein; saturated to polyunsaturated fattyacid ratiq and sodium content and sodium to potassi-um ratio. There were no consistent, statistically sig-nificant relationships between consumption of thesenutrients and serum urate levels.
Systolic and diastolic blood pressure levels ofexaminees were stratified at the 15th, 50th, and 85thpercentiles, and mean serum urate was determined forthose within each strata (tables 39-42). For progres-sively higher systolic pressure strata, the mean serumurate was greater (tables 39 and 40). The differences inserum urate averaged 0.8 mg/dl between the lowestand highest strata of systolic pressure. The significantdifferences between extremes of systolic blood pressurepersisted for each sex, race, and age group. When bodymass index was controlled, the relationship decreasedand became inconsistent for males although generallypersisted at the same level for females. A similarrelationship was observed when diastolic blood pres-sure was used as the independent variable (tables 41
-r-
0
8.0
Male
I
T“
TI
8ody massindex
- First q.artile
m Second quartile
~Third quartile
D Fourth quartile
I 95-percentconfidence Iimit
Female
0.001-0.999 1.000-6.999 7.000 or o
more
[-r-
Ekl
0.001-0.999 1.000-6.999 7.000 or
Ethanol per week (ounces) Ethanol per week (ounces)
Figure 9. Mean serum urate Iavels in selected strata of weekly ethanol consumption within BMI quartile strata for males and females 25-74 yearsof age: United States, 1971-74
19
and 42). The magnitude of differences between thelowest and highest strata was less, averaging 0.5 mgidlof serum urate. The differences in serum urate betweenthe lowest and highest strata of disatolic pressuregenerally persisted across age, sex, and race groups.However, when body mass index was controlled therewere no consistent differences among men or amongwomen. The positive trends persisted but with consid-erably less quantitative differences. Therefore, serumurate is directly related to blood pressure, but therelationship is related partially to the confoundingeffect of body mass, particularly in males.
Demographic and behavioral variables
Socioeconomic status, general well-being scores,and the reported use of oral contraceptive agents andtobacco were examined, but no relationship was found
Clinical hematology and biochemistries
No significant or consistent relationship was foundbetween serum urate and hemoglobin concentration.
For progressively higher levels of serum glutamicoxalacetic transaminase (SGOT), the serum urateconcentrations were significantly higher (table 43,figure 10). The mean difference in serum urate was 1.4mg/dl between groups with the lowest and lhighestlevel of SGOT. This significant difference persistedtlhrough all sex, race, and age groups except femalesages 65-74 years. When body mass index wals con-trolled, the differences in serum urate level between thehighest and lowest SGOT strata were slightly less,averaging 0.76 mg/dl or half the difference whenEIMI was not controlled. Therefore, the relationshipbetween SGOT and serum uric acid is consistent,although it is partially attributable to a confounding
with serum urate levels. relationship to body mass index.
8.0
rI
Male
Body mass index
m First quartile
~ Second quanlile
~ Third quartile
n Fourth quartile
8.0 r I 95-percentFemale confidence limit
T I
Less than 16.2 -21.7 21.8-29.6 29.7 or Less than 16.2-21.7 21.8-29.6 29.7 o[r16.2 more 16.2 more
SGOT percentile strata (units/milliliter) SGOT percentile strata (units/milliliter)
Figure 10. Mean serum urate levels in percentile strata of serum glutamic oxalacatic transaminase (SGOT) within BM I quartile strata for males and
females 25-74 years of age: United States, 1971-75
20
Significant differences in mean serum urate levelswere found between those in the lowest and higheststrata of serum calcium concentration (table 44). Themean difference was 0.77 mg/dl for the total group.This relationship was present for both sexes and races,but the magnitude of difference was greater for femalesand black respondents. When age was controlled, asignificant, positive relationship was observed exceptfor those ages 55-64 years. When both age and sexwere controlled, significant differences were presentonly for males ages 35-44 and 65-74 years and forfemales ages 25-34 and 65–74 years. The relationshipin females persisted after controlling for BMI but wasnot significant in males when BMI was considered. Norelationship was found with serum inorganic phos-phate.
Multivariate analyses
Variables found to have consistent and importantrelationships to serum urate were entered as inde-
pendent variables in a regression on the dependentvariable, serum urate (table C). Males and femaleswere analyzed separately. Serum urate levels wereavailable only on those examined in the detailed andaugmentation surveys and, therefore, this analysis
comprised data from 2,731 male and 3,038 femalerespondents. Relatively little variance was explained asindicated byl?z = 0.14 for males and 0.20 for females.The independent variables with the greatest betaweights for males and females were body mass index,ethanol consumption, and SGOT. For females, ageand serum calcium concentration also had relativelyhigh beta weights, while serum calcium had a lesser,but significant beta weight in males. For both malesand females, systolic blood pressure levels made aminor but statistically significant contribution, whilesocioeconomic status, serum cholesterol, and serummagnesium gave minor, statistically insignificant pre-dictions for serum urate concentrations.
Table C. Standardized beta coefficients and standard errors for regression of serum urate on selected variables, by sex for adults ages25-74 years: United States, 1971-75
Male Female
(N = 2,731) (N = 3,038)Variable Multiple R = 0.37 Multiple R = 0.45
BetaStandard error
BetaStandard error
of beta of beta
Body mass index......................................................................... 0.32 0.02 0.32 0.02Age................................................................................................. 0.00 0.03 0.11 0.03Race ............................................................................................. 0.02 0.02 0.00 0.02Socioecomonic status................................................................. 0.00 0.02 -0.01 0.02Systofic blood pressure.............................................................. 0.03 0.03 0.02 0.03Ethanol consumption.................................................................. 0.09 0.02 0.10 0.02Setum cholesterol....................................................................... 0.02 0.02 0.06 0.02Serum calcium............................................................................. 0.07 0.02 0.11 0.02Serum glutamic oxalacetic transaminase................................ 0.09 0.02 0.13 0.02Serum magnesium....................................................................... -0.05 0.02 0.00 0.02
NOTE For race, 1 = white, 2 = black, other racial groups excluded.N = number of examinees.R = multiple correlation coefficient.
21
Discussion
Nutritional variables have important relationshipsto serum cholesterol and serum urate in a populationrepresentative of U.S. adults. Among these assess-ments of nutritional status and dietary intake, bodymass and adiposity have the most pervasive and clearlydemonstrable relationships.
Body mass index (weight/heightz ) and skinfoldthickness, which indirectly measures subcutan~eousadiposity, are directly related to serum cholesterol andserum urate. This association is present at all ages inadults, for both sexes, and for white and black persons.The magnitude of the association indicates that therelationship has biologic importance. The meandifference in serum cholesterol between the lowest andhighest quintile strata of body mass index (BMI) was31.2 milligrams/deciliter which represents values 16percent higher in the highest quintile stratum than inthe lowest quintile stratum. Similarly, the difference inmean serum urate between the lowest and highest BMIquintile strata was 1.2 mg/dl or 22 percent higher.When combined skinfold thickness (skinfolds at thetriceps and subscapular sites) was used as theindependent variable, the relationship was similar inmagnitude for serum cholesterol but somewhat less forserum urate.
The BMI reflects several tissues (e.g., bone, mus-cle, and adipose), while skinfold thickness representsprimarily adipose tissue. Body mass index and skinfoldthickness are intercorrelated, and both have similarrelationships to serum cholesterol. Therefore, it isreasonable to infer that their common relationship tocholesterol levels is predicated primarily on a relation-ship to adiposity. On the other hand, skinfoldthickness has a somewhat weaker relationship toserum urate than does BMI, and one might infer thatanother component measured by BMI—perhaps mus-cle mass or skeletal frame size-has a relationship toserum urate levels. This inference regarding musclemass is compatible with the consistently higher levelsin males, who have proportionately greater muscle andskeletal mass and less adiposity than females, but sex
differences in testosterone also provide a reasonableexplanation.32
Dietary correlates with serum cholesterol levelswere generally inconsistent and of relatively smallmagnitude. Dietary intake as determined by either thefood frequency questionnaire or 24-hour dietary recalldid not have statistically significant or consistentrelationships with serum cholesterol. Based oninterpopulational or intercultural studies and clinicalinvestigations carried out in controlled settings, arelationship would be expected between serum choles-terol and dietary fat, both the amount and type oflipid.sJ-J5This anticipated relationship is formalized inthe Keys and Hegsted equations, which relate dietarycholesterol and saturated fat directly to serum choles-terol and polyunsaturated fat inversely. IGJ7However,in this cross-sectional survey, no important relation-ships were found between serum cholesterol and theseformulas or their component parameters, saturated fat-ty acids, polyunsaturated fatty acids, and cholesterol. Asimilar lack of relationship has been reported in othercross-sectional studies of homogeneous populations.s~
On the other hand, longitudinal studies of popula-tions and studies on metabolic units tend to corrobo-rate a relationship between dietary lipids and serumcholesterol. s,ld,17The reasons for the disparate findingsare unknown but deserve comment. A major problemin population studies is the relatively large within-ir~dividual variability of reported dietary intake. Thisir~traindividual variability often exceeds the variabilitybetween individuals, particularly in a population withrelatively homogeneous eating patterns and foodsources. When a single assessment is made in a cross-sectional survey, this problem is magnified and theopportunity to find a relationship diminished. The useof multiple assessments of 24-hour recall or diet diariesdecreases the intraindividual variability and can im-prove the correlation between dietary intake andserum cholesterol.QQ1It is also apparent from thissurvey that use of a food frequency questionnaire thatassesses customary eating patterns in the preceding 3
22
months affords no better relationship with serumcholesterol than does a single 24-hour recall. More-over, in developing predictive models, it should beappreciated that dietary variables will have less predic-tive power than other nutritional measurements (bodymass index, serum biochemistries) that have lessvariability. In this survey and in a similar study byShekelle et al.,s dietary lipid intake explained far lessvariance of serum cholesterol than did body massindex. Although the results from this survey do notsupport the hypothesis that dietary lipid relates toserum cholesterol, they also do not refute an associa-tion. Rather, it seems fair to conclude that a largecross-sectional survey of the U.S. population is unlike-ly to display a clear relationship between dietary intakeand serum cholesterol unless the variability of dietarymeasurement can be decreased.
Several physiological measurements and clinicalbiochemistry parameters were related to serum choles-terol. Systolic and diastolic blood pressure weredirectly related to cholesterol levels, and this associa-tion obtained for each race, sex, and age group.Because the use of diuretic agents by hypertensivepatients may spuriously elevate serum cholesterol,patients receiving antihypertensive medication wereremoved from analysis.zgMoreover, controlling forbody mass index did not alter this relationship. This isimportant because body mass is a potentially con-founding variable as it is related to both blood pressureand serum cholesterol. However, in multiple regres-sion analysis, systolic blood pressure added littleindependent explanation of variance. A similar associ-ation between higher blood pressure levels and greaterserum cholesterol has been found in children andadolescents.42q
The health implications of this association areimportant. The clustering of two major risk factors forischemic heart disease in the same individual wouldhave a synergistic effect on atherogenesis and makeidentification and intervention in these individualsparticularly important. Further, it is important torecognize the additive risk association because neitherrisk factor, blood pressure or serum cholesterol maybesufficiently elevated to attract notice, although the riskof ischemic heart disease would be increased when thesynergistic effects are considered.45’46
In addition to diet, several other life patterns wereexplored for an association with serum cholesterol.Current use of oral contraceptives was associated withhigher levels of total serum cholesterol, but this effectwas noted only in younger women ages 18–24 years,although found for both white and black women, andthe effect was independent of body mass index.Cigarette smoking was not related to total serumcholesterol levels, although other studies have indicat-ed that it may be associated with lower levels of highdensity lipoproteins without producing a major changein total serum cholesterol.qTLipoprotein fractions were
not measured in NHANES I; therefore, this possibilitycould not be tested.
Demographic factors were associated with serumcholesterol levels. Survey respondents classified in thelower socioeconomic group by income and educationalattainment had signillcantly higher mean cholesterolvalues than those who were in the upper middlesocioeconomic group. This relationship was observedfor both sexes and white and black respondents, and itwas independent of BMI but was confounded by age.The inverse relationship between socioeconomic classand serum cholesterol was observed in respondentsages 1844 years, but a direct relationship was foundfor persons ages 55–74 years. This reversal of patternsuggests a cohort effect. The cohort of persons age 55years and over and in the upper middle socioeconomicclass may have experienced environmental influencesthat differ from younger individuals in the samesocioeconomic class. For example, the older cohortmay have had eating patterns and life styles linked tosocioeconomic status that lead to higher serum choles-terol levels, while the reverse was true for the youngercohort. This speculation is compatible with the ob-served secular changes in eating patterns and mortalityfrom ischemic heart disease that have characterizedthe past 20 years.48
Perceived health status, which reflects self-evalu-ated psychological well-being, was not related to serumcholesterol levels. This lack of association contrastssomewhat with the inverse relationship between bloodpressure and self-assessed well-being that was found inNHANES 1.7Moreover, the lack of association pro-vides no support for the concept that emotional orpsychological tone affects serum cholesterol.
Several hematologic and clinical biochemistrymeasurements were related to serum cholesterol.Hemoglobin concentration was positively associatedwith serum cholesterol levels. This relationship wasparticularly prominent for females, in whom a meancholesterol difference of 20 mg/dl was found betweenthe lowest and highest 15-percent strata of hemoglo-bin. The comparable difference in males was 10mg/dl when the same strata were used. Moreover, apositive association persisted in females after control-ling for age and body mass, while the relationshipbecame inconsistent in males with these factors con-trolled. There is no obvious explanation for theassociation. One might speculate that a dietary constit-uent, such as meat, might be related to hemoglobinand cholesterol levels, but in univariate analyses noassociation was found between serum cholesterol andprotein intake.
Strong associations were found between serumcholesterol and levels of serum calcium, and, thoughunanticipated, the relationship was consistent andpotentially important. Mean serum cholesteroldifferences between the highest (85–100) percentile andlowest (O-15) percentile strata of serum calcium and
23
magnesium were 29 mg/dl and 19 mg\dl, respec-tively. Therefore, those in the highest strata for theseelements had cholesterol values 9 to 17 percent higherthan those in the lowest strata. The pattern waspresent for both sexes, white and black races, and allages. When body mass index was controlled, thedifferences in serum cholesterol were increased. Inmultiple regression analysis, serum calcium concentra-tion had a beta weight that was greater than that forother independent variables except age. This multivari-ate analysis corroborates the strong, independentrelationship to serum cholesterol.
A similar though less robust relationship wasfound between serum magnesium and cholesterolconcentration. Serum inorganic phosphate concentra-tion, which varies reciprocally with serum calciumlevels, did not have a relationship to serum cholesterol.The similarity of calcium and magnesium relationshipsto serum cholesterol is not surprising. Both ions haveimportant roles in neuromuscular transmission, andthey share many metabolic characteristics with respectto absorption, storage, and excretion.qg$oAdditionally,nutrition deprivation and disease states associated withabnormalities in one are often accompanied by parallelchanges in the other, and both are influenced byparathyroid hormones and renal function. However,no metabolic concept for these ions affords a biologicexplanation for their association with levels of circulat-ing cholesterol, and no disease states marked byexcesses or deficiencies of calcium or magnesium areaccompanied by striking changes in serum cholesterol.A possible explanation is the binding or chelation ofcalcium of phospholipid, the concentration of which isdirectly related to cholesterol levels. This provocativerelationship deserves further investigation.
Among the other biochemical parameters, serumglutamic oxalacetic transaminase (SGOT) and serumurate had consistent and quantitatively importantrelationships with serum cholesterol concentration.Higher strata of SGOT were associated with highervalues for serum cholesterol. There was a graduatedincrease in mean cholesterol across the strata ofSGOT, with a mean difference of 13 mg\dl betweenthe highest and lowest groups. This trend persistedafter accounting for race and sex, but the differences inserum cholesterol were diminished or reversed inrespondents ages 55–74 years and in males when bodymass index was controlled. SGOT is frequently used toscreen for mild liver inflammation or impairment, andin a well population the most common liver toxin isalcohol. It seems likely, therefore, that this relation-ship reflects liver inflammation secondary to alcoholand a resultant modest increase in serum cholesterol.One might speculate further that the mechanism is anincrease in very low density lipoproteins (VLDL),which can accompany altered hepatic metabolism. TheVLDL fraction transports a portion of cholesterol, andelevation of this fraction would be associated with an
increase in total serum cholesterol. In NHANES I,serum cholesterol was the only lipid measureld, and,therefore, this hypothesis cannot be explored further.
Serum cholesterol and serum urate levels weredirectly related, and adjustment for sex, race, and agedid not change this relationship. When body mass wascontrolled, the same pattern obtained, but the quanti-tative differences were somewhat less, particularly inmales, indicating that adiposity influences the relation-ship. Studies in other groups and populations havefound a relationship between levels of serum urate andserum triglyceride or serum cholesterol. sl,szIn somestudies, these relationships dissipated when weight wascontrolled.sl The postulated mechanism is similar tothat linking SGOT and cholesterol levels. Obesity andperhaps the metabolic pathways linked to pro[iuctionof uric acid are associated with increased productionand secretion of VLDL, whose lipid fraction comprisesprimarily triglyceride and cholesterol. Thus, severalmetabolic situations including diabetes mellitus andalcohol ingestion would be accompanied by elevatedVLDL, serum triglyceride, and, to a lesser degree,increased serum cholesterol. Obesity may coexist inthese situations and further confound the relationship.
Serum urate measurements were available on thedetailed and augmentation samples of respcmdentsages 25–74 years. Serum urate level means w{ere 1.5mg/dl higher in males than in females, and a meandifference of 0.4 mg\dl was found between white andblack persons. In successively older groups of men,there were slightly higher levels of serum urate throughage 64 years for black men, but little difference forwhite men. However, in women, mean serum uratewas considerably higher in successively older groupsfor both racial groups through age 64 years and thedifferences between ages 25 and 74 years averaged 0.8to 1.0 mgl dl. These sex differences in serum urate andthe male and female trends after age 25 are similar tothose reported in a population survey of Tecurnseh,Michiganss and in studies of representative populationsin Japansq and Israel.ss In NHANES I and in theTecumseh study, the higher mean values with increas-ing age reflect skewing of the distribution to highervalues. The finding of higher values in males persists instudies of even diverse populations as does the rise inserum urate levels in progressively older femalegroups, while male levels are not related to age after 25years. This does not result from age-related increasesin weight in females because the age trends in uratepersist after adjusting for body mass index. Menopausemay have an influence on this age-related increase inwomen, although the mechanism is unknown.sq
The small but consistent difference between blackand white examinees persists at all ages and after bodymass is controlled. Ethnic variations have been noted incross-cultural and interpopulational studies,32includi-ng a small (0.5 mgidl) but consistently higher serumurate level in white adolescents.sq’sGThe explanation for
24
racial differences in the United States is not apparent,but it would not seem to be related to differences indietary intake, particularly of purine-containing foods,as none of the estimates of food intake, except alcohol,were related to serum urate in this survey. Only smalland inconsistent differences in serum urate were notedby geographic region, with levels being slightly but notsignificantly higher in the Northeast and Midwest, butthe differences were 0.1 mg/dl or less.
Reported alcohol consumption was the only di-etary variable having an important relationship toserum urate levels. Abstainers, who comprised 27percent of the adults, had consistently lower levels ofuric acid than those who consumed alcohol. Atprogressively greater levels of reported alcohol intake,serum urate was higher. This relationship held foreach sex and race group and at all ages from 25 to 74years. Interestingly, the differences between black andwhite persons with respect to serum urate levels werediminished when respondents were categorized byalcohol consumption. Controlling for body mass didnot change the pattern of relationship to alcoholingestion in any subgroup. Historically, excessivealcohol consumption has been related to gout, al-though the relationship was partially due to associatedlead ingestion, which produces Saturnine gout.57Ex-cessive acute alcohol ingestion to the point of inebria-tion commonly promotes elevated uric acid levels,which lead to suppressed renal excretion of uric acidand result in hyperuricemia. On the other hand, only arelatively few reports indicate habitual alcohol intakein noninebriating amounts leads to higher mean serumurate Ievels.ssjsgThe postulated mechanisms for theeffect of chronic intake of ethanol include changes inpurine synthesis, extracellular fluid volume changes,and the increased ingestion of purines that are presentin considerable quantities in beer.sz
A direct relationship was also found betweenserum urate and SGOT. This association could berelated to the confounding influence of chronic alcoholingestion, which can produce an elevation of bothserum urate and SGOT. However, in the multivariateanalysis, both reported alcohol ingestion and SGOTlevel were independently related to serum urate levels.This suggests that each of these independent variablesmakes a unique contribution of the variance of serumurate. A possible explanation is that both alcoholingestion and subtle hepatic inflammation relate touric acid metabolism, perhaps through different mech-anisms. For example, alcohol ingestion may tiectrenal excretion of urate and thereby lead to elevatedserum levels. Alcohol, particularly with large orprolonged ingestion, may also result in subtle changesin hepatic metabolism that affect the rate of synthesisof uric acid. This speculation cannot be resolved byfurther population data but requires investigation ofmetabolic alterations.
Among the other clinical biochemistries, serum
calcium had a significant relationship to serum uratewhen the effects of age and body mass index weretaken into account. At progressively greater serumcalcium concentrations, serum urate levels were high-er, and the difference was 0.77 mg/dl of serum uratebetween the lowest and highest 15 percentile strata ofserum calcium. The relationship was consistent in bothsexes and in white and black respondents, although itwas not consistent over all ages. The association wassignificant in the multiple regression analysis, and thebeta weight was similar to those for SGOT and ethanolconsumption. Interestingly, serum magnesium had aninconsistent negative relationship with uric acid in theunivariate analysis but a relatively robust negative betaweight in the multiple regression analysis for males.No biologic mechanism or explanation is apparent, butit is of interest that serum calcium and serummagnesium have contrasting rather than similar rela-tionships, a situation not found for their associationswith blood pressureT or serum cholesterol relation-ships.
Two measures of social and psychological statuswere available, but neither had a clear or consistentrelationship to serum urate concentration. The generalwell-being questionnaire, which assesses self-reportedpsychological and physical well-being, had no associa-tion with urate levels. Socioeconomic status, summa-rized from income level and educational attainment ofhead of household, also had no association. Theseobservations contrast with reports that uric acid levelsare higher in those with higher “social class” oreducational attainment.@,61Observations suggesting arelationship have been made in groups specificallyselected to provide sharp contrasts or dichotomies inachievement. In a general population survey, theextremes of achievement are not so sharply drawn. Forexample, in a survey of Israeli men, educationalattainment had a relatively weak though significantassociation.ss However, this does not detract from theinteresting finding that economic and educationalachievement are not related to serum urate in thegeneral US. population.
Although the association between uric acid andblood pressure observed in this survey is discussed in acompanion report,T the relationship deserves comment.Serum urate was related to systolic and diastolicpressure in each sex, race, and age group, andadjusting for body mass index decreased but did notremove this association. Other population studiessS@have found a similar relationship that persists afteradjustment for weight or body mass. The reason forthis association and the potential biologic mechanismsthat might be responsible are not clear, but mostspeculation has been directed to the kidney and its rolein excretion of uric acid and in control of bloodpressure.
In multiple regression analysis, only four variablescontributed independently to explanation of serum
25
urate levels. Body mass index offered the greatestexplanation of variance, but alcohol intake, serumcalcium, and SGOT also made significant andindependent contributions. Other variables with rela-tionships in univariate analysis but no independentrelationship in the multivariate analysis include race,blood pressure, serum cholesterol, and serum magne-sium. Therefore, the association for these variables wasaccounted for by the other significant variables.
The general caveats regarding inferences that maybe drawn from a large survey were stated at the outsetof this report and deserve comment here. It is apparentthat a survey because of its cross-sectional naturecannot effectively describe variables that are associatedwith subsequent development of disease, that is, riskfactors. Physiologic variables and behavior patternsmeasured at the time of the survey may be altered byprior disease manifestations or treatment, and this maybias or confound valid relationships. In analysis, it ispossible to minimize this effect by removing individu-als with a prior diagnosis of disease or who receivetreatment. However, these individuals are usuallyimportant to a relationship, and their removal inanalysis may weaken or blur a valid association.
A second reservation relates to the representa-tiveness of variables measured on a single occasion andthe variability of the measurement. The two problemsare related and are particularly important consider-ations with respect to nutritional variables. The cross-sectional data from this survey indicate that there arerather marked changes in dietary patterns with age.There may even be a cohort effect with some patternschanging within younger groups who may have re-sponded to public advice regarding diet. Similarly,there are age-related changes in weight, blood pres-sure, and serum cholesterol that can enhance or negaterelationships with other variables.
In addition to the problem of long-term represent-ativeness of a single assessment of dietary patterns,there is considerable variability of that single measure-ment with respect to characterization of currentdietary intake. This problem has been discussedelsewhere, and the conclusion from both theoretical
explorations and practical experiences, including thatgained in NHANES I, would indicate that currentmethodology of dietary recall and food frequencyquestionnaires are unlikely to disclose strong relation-ships with other variables. In a population withrelatively homogeneous dietary patterns, the variabili-ty of the measurement, both in terms of technologicmeasurement and day-to-day variability, exceeds thedifferences among individuals. The food frequencyquestionnaire probably affords a more representativeassessment of customary dietary intake than does asingle 24-hour dietary recall, but the precision ofmeasurement is worse and is semiquantitative at best.Food frequency questionnaires can supplement orconfirm eating patterns determined by 24-hour recall,but there is no feasible means of combining quantita-tive and nonquantitative assessments. When the vari-abilityy of assessed dietary intake is relatively great, itfollows that parameters having less variability (e.g.,body mass index and serum biochemistries) willdisplay stronger relationships than dietary variables.Therefore, body mass index and serum biochemistries,which have little variability and can be well-ccmtrolledmeasurements, will appear to have more robust rela-tionships than dietary variables even if a relativelystrong relationship were to exist with a dietaryvariable.
Data from NHANES I can be used as an epidemi-ologic tool to define interrelationships among cardio-vascular variables and nutritional characteristics. Theanalyses presented here confirm and extend somerecognized relationships and suggest others that couldbe profitably explored. Body mass and, more spe-cifically, adiposity were related to serum cholesteroland urate. From the NHANES I data, it is clear thatthis relationship extended across all adult age groupsand to both sexes and races. Among the unanticipatedassociations was a consistent, direct relationship be-tween serum cholesterol and urate and serum calciumand magnesium. While the relationships may be anartifact related to binding of calcium by lipoproteincomponents, this provocative finding deserves furtherinvestigation.
26
References
lBlackbum, H.: Progress in the epidemiology and prevention ofcoronary heart disease, in P. Yu and J. Goodwin, eds., Progress inCardioiogv. Philadelphia. Lea and Feblger, 1974. p. 1.
ZAhems, E. H., Jr.: The management of hyperlipidemia Whether,rather than how. Ann. Intern. Med. 85:87-93, July 1976.
sWerko, L.: Risk factors and coronary heart dke*Facts orfancy? Am. Heart 1 91(1):87-93, July 1976.
4Turpeinen, O.: Effect of cholesterol-lowering diet on mortalityfrom coronary heart disease and other causes. Circulation. 59:1-7,1979.
%hekelle, R. B., Shryock, A. M., Paul, O., and others Diet, serumcholesterol and death from coronary heart disease The WesternElectric Study. N EngL L Mea! 304:65-70.1981.
Wditorial: Serum uric acid and coronary heart disease. Lancet.1:358, 1969.
TNational Center for Health Statistics, W. R. Harlan and others:Dietzmy intake and cardiovascular risk factors, part I. Bloodpressure correlates, United States, 1971–75, Vital and HealthStatistics. Series 1l-No. 226. DHHS Pub. No. (PHS) 82-1676.Public Health Service, Hyattsville, Md. In preparation.
sNational Center for Health Statistics, H. Miller: Plan andoperation of the Health and Nutrition Examination Survey, UnitedStates, 1971–1973, Vital and Health Statistics. Series 1-Nos. 10aand 10b. DHEW Pub. No. (HS~ 73-1310. Health Services andMental Health Administration. Washington. U.S. GovernmentPrinting Office, Feb. 1973.
gNational Center for Health Statistics, J. R. Landisj J. M.Lepkowski, S. A. Eldund, and S. A. Stehouwe~ A generalstatistical methodology for the analysis of data from a complexsurvey. Fnt National Health and Nutrition Examination Survey.Vital and Health Statistics. Series 2–No. 92, DHHS Pub. No.(PHS) 82-1366. Public Health Service, Hyattsville, Md. hpreparation.
10Nationd Gnter for Health Statistics: Plan and operation of theHANES I Augmentation Survey of Adults, 25-74 years, UnitedStates, 1974-1975, by A. Engel, R. Murphy, K. Maurer and E.Collins. Vital and Health Statistics. Series l-No. 14. DHEW Pub.No. (PHS) 78-1314. Public Health Service. Washington. U.S.Government Printing Office, June 1978.
llNation~ ~nter for Health Statistic= Food consumption prOfik%of white and black persons aged 1-74 y- United States,1971-74, by C. M. Dresser, M. D. Carroll, and S. Abraham. Vitaland Health Statistics. Series 1l–No. 210. DHEW Pub. No. (PHS)79–1658. Public Health Service. Washington. U.S. GovernmentPrinting Office, May 1979.
lZNational Center for Health Statistic Dietary intake findings:United States, 1971-74, by S. Abraham, M. D. Carroll, C. M.Dresser, and C. L. Johnson. Vital and Health Statistics. Series1l-No. 202. DHEW Pub. No. (HRA) 77-1647. Public HerdthService. Washington. U.S. Government Printing G&e, July 1977.
ISyoulmd, D. M., and Engle, A.: Dietmy data methodology inHANES. Z Am. Diet. Assoc. 68(l): Jan. 1976.
14Tuhme Dietant Listing of Dietant Identification Data and FoodValues per 100 Gram Edible Portion. Unpublished computerprintout. New Orleans. Nutrition Section, BloMedicd ComputingSystem, Tulane University, 1969.
IsWatt, B. K., and Merrill, A. L.: Composition of Foods-Raw,Processed, Prepared. Agriculture Handbook, No. 8 (rev.). Wash-ington. U.S. Department of Agriculture, 1963.
lbKeys, A., Anderson, J. T., Grande, F.: Serum cholesterolresponse to changes in the diet. 1-5. Metabolim 14747–787,1965.17H@~, D. M., Mc&dy, R. B., Myers, M. L., S*e, F. J.:Quantitative effects of dietary fat on serum cholesterol in man. Am.J. Cli~ Nutr. 1%281–295, 1965.
WHANES: Examination statT procedures manual for the Healthand Nutrition Examination Survey, part 15a. Public HealthService, June 1972.
WHANES I: Hematology and clinical chemistry proceduresdeveloped or utilized by the Centers for Disease Control, Bureau ofLaboratories, 1971-75, part 16. Public Health Service, Aug. 1979.
2oNational Center for Health Statistic% Eighth Revision Interna-tional Class@ation of Dkeases, Adapted for Use in the UnitedStates PHS Pub. No. 1693. Public Health Service. Washington.U.S. Govemment Printing Of@ 1967.ZINation~ ~nter for HealtlI Statistics Weight ad Height ofAdults 18–74 Years of Age United States, 1971–74 by S.Abraham, C. L. Johnson, and M. Najjar. Vital and HealthStatistics. Series 11-No. 211. DHEW Pub. No. (PHS) 79-1659.Hyattsville, Md., 1979.
~National Center for Health Statistics: NCHS Growth Curves forChildren Bti-18 Y- United States by P. V. V. Hamill, T.Drizd, C. L. Johnson, R. Reed, A. Roche. Vital and HealthStatistics. Series 1l-No. 165. DHEW Pub. No. (PHS) 78-1650,Public Health Service. Washington. U.S. Government PrintingOffice, May 1979.23Nation~ ~nter for H~th s~tisti= A ~nc~ent v~dationstudy of the NCHS General Well-Being Schedule, by A. F. Fazio.Vital and Health Statistics. Series 2-No. 73. DHEW Pub. No.(EIRA) 78-1347. Health Resources Administration. Washington.U.S. Government Printing O&ce, Sept. 1977.
27
z4Abell, L. L., Levy, B. B., Brodie, B. B., and Kendall, F. E.: Asimplified method for the estimation of total cholesterol in serumand demonstration of its specificity. J. Biol. Chem. 195:357-366,1952.25Eavenson, E), @er, c). T., Cisson, J. G., ~d Witter, ~. F.; A
semiautomated procedure for the determination of serum choles-terol using the Abell-Kendall method. J. Am. Oil Chem. Sot.43:652-656, 1966.26Nation~ center for Health Statistics: Total semm cholesterol
levels for adults 18-74 years, by S. Abraham, C. L. Johnson, M. D.Carroll. Vital and Heahh Statisdcs. Series 1l-No. 205. DHEWPub. No. (PHS) 78-1652. Public Health Service. Washington. U.S.Government Printing GfIice, Apr. 1978.
zTKidleen,J., Vanderburg, D., Harlan, W. R.: Application of weightheight ratios and body indices to a juvenile population: TheNational Health Examination Survey data. 1 Chronic Dis.31:529-537, 1978.
Z8Womers1y, J., and Dumin, J. V. G. A.: A comparison of theskinfold method with extent of overweight and various weight-height relationships in the assessment of obesity. Br. 1 Nutr.38:271-284, 1977.
zgGrimm, R. I-L,Leon, A. S., Hunninghake, D. B., et al.: Effects oftheazide diuretics on plasma lipids and lipoproteins in mildlyhypertensive patients. Arm. Intern. Med. 947-11, 1981.
3’JBerkman,L. F.: Social Networks, Host Resistance and Mortality:A Follow-up Study of Alameda County Residents. Doctoraldissertation, University of California, Berkely University Mi-crofilms International, 1977.31Davi&on, St, Passmore, R., Brock, AtD., Tmswell, A.; ~~~a~
Nutrition and Dietetics, 6th ed. London. Churchill Livingston,1975.
32wYnge~den,J. B,, and Kelley, W. N.: Gout and Hyperurecemia.New York. Grume and Stratton, 1976.
33Keys, A. B.: seven Countries: A Mtdtivariate Analysis of Deathand Coronary Heart Disease. Cambridge. Harvard UniversityPress, 1980.
34sacks, F, Mm, Castefli, w, p., Dormer, & Kass, E. H.: P]asma
lipids and lipoproteins in vegetarians and controls. N. Engl. J.Med. 292:1148-1151, 1975.
JSStmler, J.: Lifestyles: Major risk factors, proof and public policy.Circulation. 58:3-19, 1978.
JGMcGi~, H, C. Jr.: The relationship of dietary cholesterol toserum cholesterol concentration and to atherosclerosis in man. Am.J C[in. Nutr. 32:2664-2702, 1979.
3TGordon, T.: The Framinghm Diet Study: Diet and the regula-tion of serum cholesterol, in W. B. Kannel and T. Gordon, eds.,The Framingham Study: An Epidemiological Investigation ofCardiovascular DLrease (Section 24). Washington. U.S. Gover-nmentPrinting Office, 1970.
JBNichols, A. B., Ravenscroft, C., Lamphiear, D. E., Ostrander, L.D. Jr.: Independence of serum lipid levels and dietary habits TheTecumseh Study. J.A.M.A. 236:1948-1953, 1976.39Kahn, H. & Me~e, J. H., Ne~eld, H. N., et ala: Semm
cholesterol: Its distribution and association with dietary and othervariables in a survey of 10,000 men. Isr. 1 Med. Sci. 5:1117-1127,1969.
4’3Liu,K., Stmler, J., Dyer, A., et al,: Statistical methods to assessand minimize the role of intra-individual variability in obscuringthe relationship between dietary lipids and serum cholesterol. XChronic Dis. 31:399418, 1978.
41Jawbs, D. R., Jr., &derson, J. T., Blackbuq H.: Diet and
serum cholesterol: Do zero correlations negate the relationship?Am. J Epidemiol. 110:77-87, 1979.
42Fra&. G. c.. Berensen, G. s. and Webber, L- &: Diem studies-,––———and the relationship of diet to cardiovascular disease risk factorvariablea in 10-year-old children—The Bogalusa Heart Study. Am.J CIin. Nutr. 31:328-340, Apr. 1978.
4JHarlan, W. R., Comoni-Huntley, J., and Leaverton, P. E.: Bloodpressure in childhood: The National Health Examination Survey.Hypertension. 1:559–566, 1979.
44Comoni-Huntley, J., Harlan, W. R., and Leaverton, P. E.: Bloodpressure in adolescence The United States Health ExaminationSurvey. Hypertension. 1:566-571, 1979.45Kame1, J?/. B., Castelli, w. p., Gordon, T., and McNiamara. pm
M.: Serum cholesterol, lipoproteins, and the risk of coronary heartdisease The Framingham Study. Ann. Intern. Med. 74:1--12, 1971.
46Chapman, J. M., and Massey, J. T.: The interrelationship ofserum cholesterol, hypertension, body weight and risk of coronarydisease. L Chronic Dis. 17:933-949, 1964.
4TGarrison, R. J,, Kannel, W. B., Feinleib, M., et a].: Cigarette\smoklng and HDL cholesterol: The Framingham Offspring Study.Atherosclerosis. 30:17-25, 1978.
‘~proceedings of the Conference on the Decline in Corona~ HeartDisease Mortality, R. J. Havlik and M. Feinleib, eds. NIH Pub. No.‘79–161O.National Institutes of Health. Public Health Service, May‘1979.
~lgLinkswiler,H.; Calciw, in Present Knowledge in Nuti~ition. D.M. Hegsted, et al., eds. New York, Washington, NutritionIFoundation Inc., 1976, pp. 232–246.
%hils, M. E.: Magnesium, in Present Knowledge in Nutrition. D.M. Hegsted, et al., eds. New York, Washington. NutritionFoundation Inc., 1976, pp. 247-258.
sl~lbson, T., and Grahame, R.: Gout and hyperlipidaemia. Ann.ltheum. Dis. 33:298-303, 1974.
SzBerkowitz, D.: Blood lipid and uric acid interrelationships.~[/4.M./f. 19&85&858, 1964.
sq~lkkelsen, W. M., Dodge, H. J., Valkenburg, H.: The distribu-tion of serum uric acid values in a population unselected as to goutc}rhyperuricemia Tecumseh, Michigan, 1959–1960. Am. J Med.~9:242_251, 1965C
5Okada, M., Takesheta, M., Ued~ K., Omae, T. and Hirota, Y.:Factors influencing the serum uric acid level: A study based on apopulation survey in Hisayam Town, Kyushu, Japan. Z ChronicI)is. 33:607-612, 1980.
~SGoldbourt, U., Medalie, J. H., Herman, J. B.j and Neufeld, H. N.:$emm u~c acid: correlation with biochemical, anthropometric,clinical and behavioral parameters in 10,000 Israeli men. JChronic Dis. 33:435=143, 1980.sGHulan, W. R., Comoni-Huntley, J., and Leaverton, P. E.:Physiological determinants of serum urate levels in adolescence.Pediatrics. 63:569-575, 1979.3’?Gamod,A. B.: The Nature and Treatment of Gout and Rheumat-ic Gout. London. Walton and Maberly, 1863. p.251.5BEvans, J. G-, Pfior, 1. A. M., ~d ~rvey, H. P. B.: Relation of
serum uric acid to body bulk, hemoglobin, and alcohol intake intwo South Pacific Polynesian populations. Ann. Rheum Dir.27:319-325, 1968.59MacJ_,achlan, M. J., ~d Rodnm, G. p.: Effects of food fast adalcohol on serum uric acid and acute attacks of gout. Am. J Med.42:38-56, 1967.
~Dunn, J. P., Brooks, G. W., and Mausner, J. et al.: Social classf9;#ent of serum uric acid levels in man. J.A.M.A. 185:431436,
GICobb, S.: Hyperuricemia in executives, in The Epidemiology ofChronic Rheumatism, vol. 1. Oxford, Blackwell Scientific Pub.1963. p. 182.
28
62Be@so~, c., ~d Tibbh, E.: Serum uric acid Ievek in women
— An epidemiological study with special reference to women withhigh serum uric acid vrdues. Acts. Med. Stand. 19693-102, 1974.
cJNational Center for Health Statistics A study of the effect ofremuneration upon response in the Health and Nutrition Examina-tion Survey, United States, by E. E. Bryant, et al. Vital and HealthStatistics Series 2-No, 67. DHEW Pub. No. (HIL%) 76-1341.Health Resources Administration. Washington. U.S. GovernmentPrinting Ottice, Oct. 1975.64Nation~ Center for Health Statistic B1OOCI pressure levels Of
persons 6-74 years of age Untied States, 1971–1974, by J. Robertsand K. Maurer, Vitai and Health Statistics. Series 11 - No. 203.DHEW Pub. No. (HRA) 78-1648. Health Resources Administra-tion. Washington. US. Government Printing Office, Sept. 1977.65National &nter for Health stat~ti~. Hypefiension in adults
25-74 years of age United States, 1971-1975, by J. Roberts and M.
Rowland. Vital and Health Sfatktim series 11-No. 221. DHHSPub. No. (PHS) 81-1671. Public Health Service. Washington. U.S.Government Printing Ofiice, Feb. 1981.
%epkowski, J. M.: Design Effects for Multivariate CategoricalInteractions. Doctoral thesis, University of Michigan, 1980.
Wuwey Research Center Computer Support Group: OSIRIS IVUser’s Manurd. University of Michigan, Institute for SocialResearch, 1979.
Wiiter, S. T.: Survey sampling errors with OSIRIS IV. Paperpresented at the COMPSTAT conference, Aug. 1980.69~en@ J.: Elemen~ of fionome~cs. New York, Macm~
Publishing Co., Inc., 1971.
TOHolt,D., Smith, T. M. F., and White, P. D.: Regression andYSk
of data from complex surveys. Z R. Stat. Sac. A. 143(4):474-487,1980.
29
List of detailed tables
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11<
12.
13.
30
Serum cholesterol levels of adult males ages 18-74 yearswithin body mass index strata showing means and standarderrors of means by race and age United States, 1971-74 ........Serum cholesterol levels of adult females ages 18-74 yearswithin body mass index strata showing means and standarderrors of means by race and age United States, 1971-74 ........Serum cholesterol levels of adult males ages 18-74 yearswithin total skinfold (triceps and subscapular) thickness stratashowing means and standard errors of means by race andage United States, 1971.74 .........................................................Serum cholesterol levels of adult females ages 18-74 yearswithin total skinfold (triceps and subscapular) thickness stratashowing means and standard errors of means by race andage: United States, 1971.74 .........................................................Serum cholesterol levels of adult males ages 18-74 yearswithin systolic blood pressure level strata showing means andstandard errors of means by race, age and body mass indexUnited States, 1971-74 .................................................................Serum cholesterol levels of adult females ages 18-74 yearswithin systolic blood pressure level strata showing means andstandard errors of means by race, age and body mass indexUnited States, 1971-74 .................................................................Serum cholesterol levels of adult males ages 18-74 yearswithin diastolic blood pressure level strata showing meansand standard errors of means by race, age and body massindex United States, 1971-74 ......................................................Serum cholesterol levels of adult females ages 18-74 yearswithin diastolic blood pressure level strata showing meansand standard errors of means by race, age and body massindex United States, 1971-74 ......................................................Serum cholesterol levels of adults ages 18-74 years withinfrequency of fatty food consumption strata showing meansand standard errors of means by sex, race, age and bodymass index United States, 1971-74 ............................................Serum cholesterol levels of adult males ages 18-74 yearswithin total dietaty calories strata showing means and stan-dard errors of means by race, age and body mass indexUnited States, 1971-74 .................................................................Serum cholesterol Ievelsof adult females ages 18-74 yearswithin total dietary calories strata showing means and stan-dard errors of means by race, age and body mass indexUnited States, 1971-74 .................................................................Serum cholesterol levels of adult males ages 18-74 yearswithin total dietary fat strata showing means and standarderrors of means by race, age and body mass index UnitedStates, 1971-74 .............................................................................Serum cholesterol levels of adult females ages 18-74 yearswithin total dietaty fat strata showing means and standarderrors of means by race, age and body mass index UnitedStates, 1971-74 .............................................................................
32
33
34
35
36
37
38
39
40
41
42
43
44
14. Serum cholesterol Ievelsof aduksages 18-74 yeara withinpercent of caloric intake from fat strata showing means andstandard errors of means by sex, race, age and body massindex United States, 1971-74 .......................................................
15. Serum cholesterol Ievelsof adults ages 18-74 years withinIinoleic to saturated fatty acid intake ratio strata showingmeans and standard errors of means by sex, race, age andbody mass index United States, 1971-74 ....................................
16. Serum cholesterol levels of adult males ages 18-74 yearswithin dietary cholesterol strata showing means and standarderrors of means by race, age and body mass index UnitedStates, 1971-74 ..............................................................................
‘17. Serum cholesterol levels of adult females ages 18-74 yearswithin dietary cholesterol strata showing means and standarderrors of means by race, age and body mass index UnitedStates, 1971-74 ..............................................................................
18. Serum cholesterol levels of adults ages 18-74 years withindietary sodium/potassium ratio strata showing means andstandard errors of means by sex, race, age and body massindex United States, 1971-74 .......................................................
19. Serum cholesterol levels of adult females ages 18-74 ‘yearsby extent of use of oral contraceptive agents showing meansand standard errors of means by race, age and body massindex United States, 1971-74 ......................................................
;20. Serum cholesterol levels of adults ages 18-74 years within
:21.
:22.
23.
24.
25<
socioeconomic class strata showing means and standarderrors of means by race, age and body mass index UnitedStates, 1971-74 ..............................................................................
Serum cholesterol levels of adults ages 18-74 years withingeographic region showing means and standard errors ofmeans by race, age and body mass index United States,1971-74 ...........................................................................................
Serum cholesterol levels of adult males ages 25-74 yearswithin total general well-being strata showing means andstandard errors of means by race, age and body mass indexUnited States, 1971-75 .................................................................
Serum cholesterol levels of adult females ages 25-74 yearswithin total general well-being strata showing meana andstandard errors of means by race, age and body mass indexUnited States, 1971-75 .................................................................Serum cholesterol levels of adult males ages 18-74 yearswithin hemoglobin strata showing means and standard errorsof means by race, age and body mass index United States,1971-74 ..........................................................................................Serum cholesterol levels of adult females ages 18-74 yearswithin hemoglobin strata showing means and standard errorsof means by race, age and body mass index United States,1971-74 ..........................................................................................
45
46
47
48
49
50
51
52
53
54
55
56
26. Serum cholesterol levels of adults ages 25-74 years withinSGOT strata showing means and standard errors of means bysex, race, age and body mass index United States, 1971 -75...
27. Serum cholesterol levels of adults ages 25-74 years withinserum calcium strata show-rig means and standard errora ofmeans by sex, race, age and body mass index United States,1971-75 ..........................................................................................
28. Serum cholesterol levels of adulta ages 25-74 years withinserum magnesium strata showing means and standard errorsof means by sex, race, age and body mass index UnitedStates, 1971-75 .............................................................................
29. Serum cholesterol levels of adult males ages 25-74 yearswithin serum urate strata showing means and standard errorsof means by race, age and body mass index United States,
31,
32.
33.
34.
35.
Serum cholesterol levels of adult females ages 25-74 yearswithin serum urate strata showing means and standard errorsof means by race, age and body mass index United States,1971-75 ..........................................................................................Serum urate levels of adults ages 25-74 years showingmeans, standard erros of means and selected percentiles bysex and race: United States, 1971 -75 ..........................................Serum urate levels of adult males ages 25-74 years withinbody mass index strata showing means and standard errorsof means by race and age United States, 1971 -75 ...................Serum urate levels of adult females ages 25-74 years withinbody mass index strata showing means and standard errorsof means by race and age United States, 1971 -75 ...................Serum urate levels of adult males ages 25-74 years withintotal skinfold (triceps and subscapular) thickness stratashowing means and standard errors of means by race andage United States, 1971-75 .........................................................Serum urate levels of adult females ages 25-74 years withintotal skinfold (triceps and subscapular) thickness stratashowing means and standard errors of means by race andage: United States, 1971.75 .........................................................
57
58
59
60
61
62
63
64
65
66
36.
37.
38.
39.
40.
Serum urate levels of adults ages 25-74 years withingeographic region strata showing means and standard errorsof means by race, age and body mass index United States,1971-75 ..........................................................................................
Serum urate levels of adults ages 25-74 yeara within strata ofweekly ethanol consumption showing means and standarderrors of means by sex, race, age and body mass indexUnited States, 1971-74 ..................................................................
Serum urate levels of adults ages 25-74 yeara within strata ofdietary purine showing means and standard errors of meansby sex, race, age and body mass index United States,1971.74 ...........................................................................................
Serum urate levels of adult males ages 25-74 years withinsystolic blood pressure level strata showing means andstandard errors of means by race, age and body mass indexUnited States, 1971-75 ..................................................................
Serum urate levels of adult female ages 25-74 years withinsystolic blood pressure level strata showing means andstandard errors of means by race, age and body mass indexUnited States, 1971-75 ......................=...........................................
41. Serum urate levels of adult males ages 25-74 years withindiastolic blood pressure level strata showing means andstandard errors of means by race, age and body mass indexUnited States, 1971-75 ..................................................................
42. Serum urate levels of adult females ages 25-74 years withindiastolic blood pressure level strata showing means andstandard errors of means by race, age and body mass indexUnited States, 1971-75 ..................................................................
43. Serum urate levels of adults ages 25-74 years within SGOTstrata showing means and standard errors of means by sex,race, age and body mass index United States, 1971 -75 ............
44. Serum urate levels of adults ages 25-74 years within serumcalcium strata showing means and standard errors of meansby sex, race, age and body mass index United States,1971-75 ...........................................................................................
67
68
69
70
71
72
73
74
75
31
cdhr Table 1. Serum cholesterol levels of adult males ages 18-74 years within body mass index strata showing means and standard errors of means by race and age United States,
f 971-74
Body mass index (kilograms/metersZ)
Less than 21.9095 21.9095-24.0785 24.078&26.1385 26.1388-28.4735 28.4736 or moreRace and age
Standard Number Standard Number Standard NumberMean
Standard Number Standard Numbererror of of Mean error of of Mean error of of Mean error of ofmean
Meanexaminees
error ofmean
ofexaminees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall .......................... 193.1 1.44 1,036 201.7 2.12 1,036 216.7 1.89 1,036 221.1 1.93 1,036 224.3 2.02 1,035
RaceWhite ........................... 192.8 1.62 804 202.3 2.25 867 217.0 2.02 884Black ........................... 195.1 3.79 232 195.9
221.74.83 169
1.79 924213.3
223.25.08
2.17152
865211.6 7.40 112 233.9 7.88 180
Age
16-24 years ............... 170.9 2.54 258 167.4 2.74 203 186.5 3.93 12625-34 years ............... 190.3 2.57 146 196.0
197.13.78
4.04162
82203.0
193.13.57
6.05160
87
35-44 years ............... 205.9 6.81 87 216.6202.7
5.172.52
120146
223.3 4.97212.8 4.59
133150
45-54 years ............... 205.5 5.92 123 228.7223.4
5.273.65
131163
228.4225.8
3.155.02
161154
55-84 years ............... 216.7 5.99 92 223.0235.8
3.84 924.46 169
236.1 6.00236.8 4.34
124176
65-74 years ............... 213.4 3.23 328 228.1226.4
2.654.36
308143
223.2237.5
4.324.98
332133
232.1 2.98 333 231.9 5.12 335
I Excludes“other”racialgroups.
Table 2. Serum cholesterol levels of adult females ages 18-74 years within body mass index strata showing means and standard errors of means by race and age United States,1971-74
Body mass index (kilograms/meters9
Less than 20.6325 20.6325-22.8135 22.813&25.3195 25.3196-29.3295 29.3296 or moreRace and age
Standard Number Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examirrees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall .......................... 194.3 1.57 1,661 205.9 1.70 1,661 217.2 1.45 1,656 229.3 2.25 1,660 230.1 2.32 1,660
RaceWhite ........................... 194.3 1.69 1,413 206.5 1.65 1,468 217.9 1.38 1,396 230.0 2.53 1,302 230.7 2.51 1,171Black ........................... 193.3 3.05 246 197.5 4.76 193 209.6 4.90 262 224.3 3.59 358 227.2 3.91 469
Age16-24 years ............... 175.0 2.50 530 182.4 2.24 376 187.0 2.88 282 192.4 5.09 175 200.0 4.77 13325-34 years ............... 187.3 1.85 502 189.5 1.93 447 198.0 2.91 360 203.9 2.32 268 202.8 3.89 29635-44 years ............... 198.4 2.99 265 201.5 2.42 349 204.5 2.41 329 215.2 3.46 334 214.7 2.95 3574G54 years ............... 212,7 4.84 100 223.0 3.28 159 236.5 4.09 171 240.8 4.23 204 237.6 7.45 19355-64 years ............... 238.6 8.03 250.3 5.48 80 238.9 4.85 144 247.7 4.13 171 249.4 3.80 19965-74 years ............... 235.1 5.19 1: 246.3 3.02 250 250.5 3.48 372 253.7 3.43 508 253.6 2.93 482
t Excludes“other” racialgroups.
w.P Table 3. Serum cholesterol levels of adult males ages 16-74 years within total skinfold (triceps and subscapular) thickness strata showing means and standard errors of means by
race and age United States, 1971-74
Total skinfold thickness (millimeters)
Less than 76.5 16.5-22.9 23.0-28.9Race and age
29.0-36.4 36.5 or more
Standard Number Standard Number Standard Number Standard NumberMean
Standard Numbererror of of Mean error of of Mean error of of Meanmean
error of of Meanexaminees mean
error of ofexaminees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall .......................... 191.6 1.59 1,030 205.9 1.57 1,149 217.9 2.10 1,013 220.6 2.06 972 221.76 2.39 1,015
Race
White ........................... 191.2 1.90 748 205.1 1.64 977 218.2Black ........................... 193.4 4.00 282
2.22 903203.8 4.85
220.4172
2.19 846214.1 7.06
221.2 2.26110
860223.1 3.91 126 226.6 8.80 155
Age
16-24 years ............... 169.1 3.10 260 170.0 2.79 180 183.6 4.3825-34 years ............... 189.5 3.37 147
116194.2 3.37
194.3163
4.14 88203.2 3.93
191.9 4.82 112
35-44 years ............... 207.8 5.76 100151 210.6 3.74 135
223.0 3.58 133 223.4 5.06207.5 4.07 170
45-54 years ............... 205.1 5.86 114129
218.5 3.95220.0
1653.82 150
234.6 4.41224.3 4.55 145
55-64 years ............... 211.7 4.97 92148
227.4 3.10237.8
1314.54 152
235.2 !3.77237.4 4.80 181
65-74 years ............... 212.6 3.07 317124
224.1229.6
2.874.08
357123 236.7 6.94 114
228.3 3.84 345 229.0 3.68 324 233.7 4.85 293
*Excludes “other” racial groups.
Table 4. Serum cholesterol levels of adult females ages 16-74 yeare within total .sIdnfold (trioeps and subscapular) thickness strata showing means and standard errors of means byrace and age United States, 1971-74
Tolal skinfold thickness (millimeter)
Less than 26.5 26.5-34.9 35.&44.4 44.5-%9 57.0 or moleRace and age Starrdarti Number Standard Number S&snokni Number Standard Number Standard Number
Mean error of of Meerr errw of of Msarr errvr of of Mean errvr of of Mean en-or of ofmean examinees mean examiness mean examiiwes mean examinees mean examinees
fvlNigrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Total .......................... 196.1 2.00 1,652 208.1 2.05 1,720 216.7 1.76 1,617 227.0 1.81 1,674 225.9 1.83 1,637
RaceWhite ........................... 196.5 2.21 1,364 207.6 2.04 1,496 217.7 1.75 1,404 227.4 1.92 1,337 226.3 2.21 1,149Black ........................... 192.9 3.00 266 213.0 5.93 224 203.8 4.09 213 223.6 4.08 337 224.2 3.33 486
Age16-24 years ............... 174.1 2.42 506 163.7 3.02 379 184.4 2.57 262 189.1 4.33 197 202.2 4.20 15025-34 years ............... 167.6 2.08 464 192.0 2.37 434 196.1 2.07 336 199.4 2.61 319 201.2 2.68 3203544 years ............... 197.9 3.21 268 202.5 3.41 307 208.8 2.72 316 213.0 2.14 342 210.0 2.53 40145-54 years ............... 217.5 4.66 100 224.6 4.36 128 230.8 4.95 169 239.7 4.52 204 236.9 5.02 22655-64 years ............... 238.6 6.15 243.2 5.16 113 245.6 5.08 107 252.0 4.34 173 243.4 4.17 19185-74 years, .............. 236.2 4.66 22 246.0 2.90 359 250.9 3.03 427 254.9 2.98 439 256.5 3.45 349
I Excludes“other”racialgroups.
Table 5. Serum cholesterol levels of adult males ages 18-74 years within systolic blood pressure level strata showing means and standard errors of means by race, age and bodymass index United States, 1971-74
Systolic blood pressure (millimeters ofmercwy)
Race, age, and body massLess than 114 114-129 130-153 154 or more
index quartile strata Standard Number StandardMean
Number Standard Number Standarderror of of Mean
Numbererror of of Mean
meanerror of of Mean
examineeserror of
meanof
examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 197.5 2.48 686 208.2 1.47 1,842 216.9 1.52 1,458
Race
222.9 2.14
White..., ........................................ 198.4 2.55 584 207.9 1.36Black ............................................ 187.6 5.73
1,601102
217.9211.0
1.845.20
1,249241
224.1206.9
2.565.62 209 216.0
Age
5.48
18-24 years ................................ 170.8 4.57 178 180.3 1.9825-34 years ................................ 195.4 3.10
374154
180.4201.8
3.662.78
181 198.1 5.14
35-44 years ................................ 216.7 4.79421
109201.5
217.93.45
4.22181 220.0 5.78
45-54 years ................................ 220.4 6.29305
85227.0
225.64.23
3.61186 214.8 9.73
55-64 years ................................ 213.4 5.36275
51233.7
235.43.87
4.70227 228.1 4.17
65-74 years ................................ 207.8 6.08155
109229.9
231.44.23
3.04190
312221.9
226.94.48
4.40 491 225.4 2.38
Body mass indexIst quartile .................................. 181.1 5.11 174 187.1 1.942d quartile ................................... 192.6 4.73
290286
190.6203.0
3.682.48
150 211.4 8.29
3d quartile ................................... 215.3 4.29697
187208.7
218.02.15
2.25465 222.3 3.95
4th quartile ................................. 210.2 6.69661
37222.6
221.32.17
5.41553
194227.9
229.03.52
3.79 290 220.9 4.98
I Excludes “other” racial groups.
682
513169
152131
104104407
80178241183
Table 6. Serum cholesterol levels of adult females ages 16-74 years within systolic blood pressure level strata showing means and standard errors of means by race, age and bodymass index United States, 1971-74
Systolic blood pressure (millimeters of mercury)
Race, age, and body mass Less than 100 106-119 120-149 150 or more
index quartiie strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean er70r of of
mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliier Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 192.2 1.63 1,070 203.6 1.31 2,650 219.3 1.73 2,678 237.7 2.80 924
RaceWhite ............................................ 192.4 1.95 883 203.7 1.37 2,211 219.4 1.86 2,251 239.7 3.11 692Black ............................................ 189.9 5.41 187 202.9 3.31 439 217.6 3.40 427 226.9 3.59 232
Age16-24 years ................................ 175.0 2.16 363 185.4 2.04 741 185.0 3.31 367 177.9 ,, 9.0225-34 yeara ................................ 188.3 2.08 381 194.8 1.90 860 195.6 1.49 556 213.4 6.82 ;35-44 years ................................ 201.5 2.74 209 202.5 1.88 619 209.5 2.39 589 218.2 4.38 10645-54 years ................................ 226.0 8.22 62 231.1 3.70 222 227.5 3.47 328 237.1 5.99 11855-64 years ................................ 228.2 9,04 26 242.4 5.96 85 248.7 3.93 259 241.9 6.24 14365-74 years ................................ 255.3 16.97 29 246.3 5.61 123 252.3 2.83 579 248.2 3.08 501
Body mass indexIst quartile .................................. 178.9 3.15 303 191.7 2.43 480 202.5 4.13 262 241.5 15.18 552d quartile ................................... 193.9 2.69 491 197.9 1.84 1,063 210.5 2.38 811 234.3 4.59 1793d quartile ................................... 205.8 2.73 248 213.7 2.53 862 227.7 2.73 1,070 240.2 3.62 3924th quartile ................................. 197.4 12.86 28 221.2 6.49 245 225.0 3.04 535 236.1 4.90 298
I Excludes“other”racialgroups.
Table 7. Serum cholesterol levels of adult males ages 16-74 years within diastolic blood pressure level strata showing means and standard errors of means by race, age and bodymass index United States, 1971-74
Diastolic blood pressure (millimeters of mercury)
Race, age, and body massLess than 70 70-82 83-94 95 or more
index quartile strata Standard Number Standard NumberMean error of
Standard Number Standardof Mean
Numbererror of of Mean
meanerror of of Mean
examinees meanerror of of
examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Total’ ........................................... 194.4 2.68 401 205.5 1.66 2,069 215.8 1.92 1,514
Race
225.4 2.01 684
White ............................................ 194.4 2.85 344 206.1Black ............................................ 194.4
1.546.27
1,798 215.857 198.2
2.055.95
1,288271
226.8215.5
1.988.09
517226 217.8 6.24 167
Age
18-24 years ................................ 170.6 3.55 131 178.8 2.5825-34 years ................................ 199.1 4.95
44877
181.6199.9
3.582.21
143 194.3 7.00 26
35-44 years ................................ 208.3 7.13396
34201.6
216.83.93
3.79249 209,9 4.84 55
45-54 years ................................ 223.8 13.2259 217.7
27 225.53.93
3.43240 232.9 4.75 100
55-64 years ................................ 231.9 9.72238
24230.3
227.33.80
4.07275 229.0 3.66 151
65-74 years ................................ 219.7 5.34196
108229.5
221.44.20
3.41173
532228.9
233.66.17
3.87107
434 227.2 4.30 245
Body mass index
1st quartile .................................. 178.1 4.83 113 186.6 2.492d quartile ................................... 191.3 4.46
384 190.4180 201.4
3.802.62
143 211.2 8.99 54
3d quartile ................................... 222.7 6.02823 208.8
89 215.22.90
1.95460 221.3 4.61 165
4th quartile ................................. 195.4 5.92684 221.7
19 221.72.71
5.21608
178230.2
225.43.00
3.43261
303 225.2 3.86 204
1Excludes“’other”racialgroups.
Table 8. Serum cholesterol levels of adult females ages 18+4 years within diastolic blood pressure level strata showing means and standard errors of means by race, age and bodymass index United States, 1971-74
Diastolic blood pressure (millimeters ofmercuw)
Race, age, and body mass Less than 68 68-77 78-89 90 or more
index quartile strata Standard Number Standard Number Standard NumberMean
Standard Numbererror of of Mean error of of Mean error of of Mean error of ofmean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 192.0 1.46 1,101 205.5 1.79 2,576 218.9 1.57 2,718 231.4 2.86 927
RaceWhite., .......................................... 191.9 1.58 930 205.7 1.86 2,204 219.4 1.69 2,238 233.2 2.97 665Black ............................................ 192.4 4.69 171 203.3 4.16 372 213.8 4.41 480 223.0 4.97 262
Age18-24 years ................................25-34 years................................35-44 years ................................45-54 years ................................55-64 years................................65-74 years ................................
Body mass index1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
177.1188.7195.3228.1248.5253.7
177.1192.4208.9209.7
2.452.312.167.13
12.906.17
2.782.563.298.92
440355161542269
314489260
38
183.9193.8201.8226.2243.3250.1
189.1200.9217.0223.0
2.241.681.883.175.044.09
2.422,302.776.77
639764527219129298
4471,071
874184
185.4196.7210.9231.4245.8251.5
208.7208.0226.9227.0
2.681.752.483.604.312.98
5.352,092.343.31
365621649305221557
266820
1,086526
209.5200.5212.3233.6244.5246.9
239.4230.5233.0229.2
16.843.902.985.994.964.30
10.275.324.143.90
4199186152141306
53164352358
t Excludes“other” racialgroups.
.Po Table 9. Serum cholesterol levels of adults ages 16-74 years within frequency of fatty food consumption strata showing means and standard errors of means by sex, race, age and
body mass index United Statea, 1971-74
Frequency of fa~ food consumption (times/week)
Sex, race, age, and bodyLess than 17.0 17.0-27.9 28.042.4 42.5 or more
mass index quartile strata Standard Number StandardMean
Number Standarderror of
Number Standard Numberof Mean error of of Mean
meanerror of of
examineesMean
meanerror of of
examinees mean examinees mean examinees
Totall ...........................................
SexMale.. ...........................................Female .........................................
Race
White ............................................Black ............................................
Age—
Total
16-24 years ................................25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Malej=ml .,-...-T , “-, . ................................25-34 years ................................36-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Female
16-24 years ................................25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Body masa index
Male
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
216.4
219.2214.7
217.0213.6
182.7201.3217.9232.1236.0240.8
$~~~
212:9222.5235.0233.3227.1
185.9194.0215.0230.3237.8253.6
197.1216.4227.1228.4
197.8207=1222.7228.1
1.76
3.782.06
1.784.76
3.604.804.204.624.403.48
. . .4.(0
10.287.636.717.634.01
4.253.314.286.476.556.09
5.555.204.927.92
3.803.625.494.90
1,603
5561,047
1,132471
287303280192153388
.-
;668362
218
222241214109
91170
145136121154
231239
274303
210.2
210.1210.3
210.2210.8
176.6196.2208.3229.5235.8242.0
177.7202.5215.6224.2220.9226.9
179.2191.6202.8234.2248.7253.5
190.4205.4220.7224.2
192.22!)4,8221.7227.1
1.11
1.551.59
1.143.06
1.971.702.082.933.692.44
3.513.083.593.455.922.36
2.312.021.884.494.513.72
2.644.023.503.45
2.84~.f ~
2.24
3,631
1,2642,367
2,954677
631830739444288699
168226189198140343
463604550246148356
314301318331
594636589548
210.0
209.7210.3
210.5202.7
181.4197.5214.5228.6240.6239.1
179.2200.5220.6233.3234.8225.8
184.2194.0206.1224.2247.4253.7
189.8209.3216.9224.5
193.5208.4224.5
3.73 227.9
1.24
1.811.47
1.333.57
1.861.812.702.514.013.87
2.672.875.323.734.504.82
2.661.422.343.685.154.06
2.483.393.164.15
2.213.282.393.49
3,611
1,5252,086
3,144467
716886633436263677
279307195221139384
437579438215124293
412407371335
679574
204.9
205.3204.3
205.0203.9
180.3198.4213.4222.7238.7235.7
177.6199.1217.4224.9234.2231.0
184.4197.2204.8218.1246.5242.6
190.0202.3218.8214.8
193.7204.8
469 218.1343 222.3
1.43
1.912.63
1.546.81
2.212.593.213.235.153.91
2.843.724.334.195.635.99
3.343.063.256.259.546.28
2.835.113.223.16
3.924.656.025.21
1,531
781750
1,395136
425337238170115246
19913711410972
150
226200124614396
233216193139
312201135102
tExcludes“other” racial groups.
Table 10. Serum cholesterol levels of adult males ages 18-74 years within total dietary calories strata showing means and standard errors of means by race, age and body massindex United States, 1971-74
Average fofal caloric infake per day (catories)
Race, age, and body mass Less than 1,425.6 1,425.6-2,201.5 2,201.6-3,402.3 3,402.4 or more
index quartile strafa Sfandard Numbar Sfandard Number Sfandard Number Sfandard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totalj ........................................... 219.1 2.35 517 216.8 1.60 1,205 209.8 2.35 1,205 197.5 1.87 516
Race
White ............................................ 221.4 2.58 387 218.0 1.86 1,021 210.3 2.36 1,058 198.1 2.11 461Black ............................................ 206.0 5.72 130 207.5 4.56 184 204.3 6.52 147 189.0 8.46 55
Age
18-24 years ................................ 190.1 6.71 28 183.3 4.36 93 181.4 2.77 234 174.5 3.18 17125-34 years ................................ 193.4 6.27 34 199.9 4.35 140 206.2 2.92 258 196.9 4.27 13335-44 years ................................ 216.3 8,61 38 220.5 3.61 131 218.5 4.16 207 219.9 7.60 7545-54 years ................................ 230.3 6.27 69 227.8 3.79 208 227.1 3.75 188 219.4 4.79 6255-84 years ................................ 229.3 5.71 66 226.8 3.74 154 225.4 5.62 107 223.5 10.14 3865-74 years ................................ 223.9 3.77 280 230.6 4.17 479 220.7 3.04 211 237.0 10.43 37
Body mass index
1st quartile .................................. 205.8 5.63 127 200.3 3.71 277 193.1 3.62 298 178.6 2.83 1582d quartile ................................... 217.5 4.94 117 215.5 4.55 263 207.6 3.12 336 192.4 3.79 1453d quartile ................................... 225.8 5.28 113 223.4 2.67 332 219.0 3.36 305 209.7 3.66 1114th quartile ................................. 224.6 4.73 159 222.3 2.70 333 220.6 4.03 266 219.8 4.68 102
~Excludes “other” racial groups.
h Table 11. Serum cholesterol levels of adult females ages 18-74 years within total dietary calories strata showing means and standard errors of means by race, age and body massindex United States, 1971-74
Average total caloric intake per day (calories)
Race, age, and body massLess than 989.68 989.68-7,509.5 1,509. G2,232.9 2,233.0 or more
index quartile strata Standard Number Standard NumberMean
Standard Number Standarderror of of Mean
Numbererror of of Mean error of
meanof Mean
examinees meanerror of of
examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
TotaP ........................................... 224.1 2.51 760 216.8 1.54 1,771 208.8 1.65 1,772 199.2
Race
1.78 759
White ............................................ 224.9 2.64 577 216.7 1.69Black ............................................ 219.6 5.66
1,495183
206.9217.0
1.804.85
1,521 199.8276
1.97207.8 4.33
625251 193.7
Age
3.74 134
18-24 years ................................ 205.9 6.38 90 178.3 2.99 30625-34 years ................................ 198.1 3.25 155
182.5194.1
3.002.05
383 179.9 3.04 227
35-44 years ................................ 212.7406
3.43 155192.5
211.02.11 471
1.97191.2
3712.87 231
45–54 years ................................ 240.2 4.95 100206.1
239.52.31
4.81413 203.7 2.97 147
55-64 years ................................ 241.2182
7.41 63223.7
244.14.14
3.93170 217.2 6.22 79
65-74 years ................................ 250.5 6.82149
197251.4
257.15.83
2.95107 235.9
357 247.96.93
5.3529
228 247.8 5.92 46
Body mass index
Ist quartile.... . .. ........................... 218.1 8.87 112 1955 3.332d quartile ................................... 213.4
3665.03 145
~91.2209.3
2.892.30
510 185.2 2.99 277
3d quartile ................................... 224.3 5.07457
208203.7
223.73.16
2.29475 200.8 3.52 190
4th quartile ................................. 233.3494
2.88 294222.1
234.63.51
3.78399 210.2
454 227.64,30
2.66163
388 212.0 4.32 129
1Excludes “other” racial groups.
Table 12. Serum cholesterol levels of adult males ages 16-74 years within total dietary fat strata showing means and standard errors of means by race, age and body mass indexUnited States, 1971-74
Total fat intake per day (grams)
Race, age, and body massLess than 57.12 51.12-90.04 90.05-150.01 150.02 or more
index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Total~ ............................................ 218.1 1.92 517 213.9 1.71 1,205 211.0 2.19 1,205 199.8 1.77 516
RaceWhite ............................................ 220.2 2.15 389 215.6 1.92 1,032 210.5 2.21 1,050 200.8 1.95 456Black ............................................ 204.4 5.05 128 198.6 4.22 173 216.7 6.31 155 186.8 8.41 60
Age16-24 years ................................ 183.8 5.13 30 180.0 3.49 126 181.2 3.05 211 177.1 3.64 15925-34 years ................................ 197.8 5.56 37 200.2 3.90 148 203.7 3.75 255 200.3 4.37 12535-44 years ................................ 217.3 6.62 46 219.5 3.91 133 219.9 3.62 199 217.5 7.94 7345-54 years ................................ 232.2 5.64 71 227.4 4.00 195 227.7 4.74 196 217.1 5.35 6555-64 years ................................ 231.8 6.88 65 224.1 4.17 145 226.9 4.56 110 226.2 6.57 4765-74 years ................................ 221.3 3.97 268 228.2 2.02 458 230.4 8.02 234 226.4 9.83 47
Body mass index1st quartile .................................. 200.5 5.40 125 196.6 3.10 292 192.6 3.25 285 185.1 2.93 1582d quartile ................................... 215.5 5.26 115 212.8 2.95 283 208.0 3.65 327 194.2 4.59 1363d quartile ................................... 233.0 3.73 116 221.9 3.63 321 217.9 3.39 313 210.5 3.07 1114th quartile ................................. 219.2 3.69 161 220.8 3.09 308 224.6 4.05 280 218.7 4.98 111
t Excludes“other”racialgroups.
*.P Table 13. Serum cholesterol levels of adult females ages 16-74 years within total dietary fat strata showing means and standard errors of means by race, age and body mass index
United States, 1971_74
Total fat intake per day (grams)
Race, age, and body massLess than 35.00 35.00-60.01 60.02-96.25 96.26 or more
index quartile strata Standard Number Standard NumberMean
Standarderror of of Mean
Number Standarderror of of Mean
Number
meanerror of of
examineesMean error of
meanof
examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 220.5 2.49 761 215.5 1.63 1,770 209.9 1.52
Race
1,772 201.8 2.17 759
White ............................................ 220.3 2.58 567 215.8Black ............................................ 221.1
1.716.25
1,499194
210.0212.1
1.584.28
1,521 202.2271 208.8
2.363.91
631251 196.8 3.35 128
Age
18–24 years ................................ 193.6 5.43 110 180.8 2.6125-34 years ................................ 197.2 3.07
311155
182.3192.6
2.502.34
380 181.1 3.88 205
35-44 years ................................ 213.1 4.02411
144192.8
207.51.95
2.40463 193.6 2.71 234
45-54 years ................................ 232.4 3.54380 206.9
89 238.82.21
2.60399 209.4 3,14 163
55-84 years ................................ 241.8 7.59191 226.8
67 244.05.27
4.12177 217.5 5.88 74
65-74 years ................................ 248.9 5.45148
196250.4
256.35.09
4.52104
329240.3
249.57.10
5.2929
249 243.6 5.62 54
Body mass indexIst quartile .................................. 206.5 5.54 126 i 94.6 2.702d quartile ................................... 214.0 5.54
383155
191.5207.0
3.052.83
482 188.9 3.28 273
3d quartile ................................... 222.6 4.11438
207203.8
225.73.08
2.81475 204.1 3.70 199
4th quartile ................................. 230.4 2.88478
271218.7
232.02.72
2.70420
471213.9
231.33.64
4.50159
395 212.3 5.17 128
1Excludes“other”racialgroups.
Table 14. Serum cholesterol levels of adults ages 16-74 years within percent of caloric intake from fat strata showing means and standard errors of means by sex, race, age andbody mass index: United States, 1971-74
Calories from fat per day (percent)
Sex, race, age, and body Less than 28.15 28.15-36.73 36.74-45.26 45.27 or more
mass index quatiile strata Standard Number Standard Number Standard NumberMean error of of Mean
Standard Numbererror of of Mean error of of Mean
examineeserror of of
mean mean examinees mean examinees mean examinees
Totalj ...........................................
Sex
Male .............................................Female .........................................
Race
White ............................................Black ............................................
18-2425-3435-4445-5455-8465-74
18-2425-3435-4445-5455-6465-74
18-2425-3435-4445-5455-6465-74
Age—
Total
years ................................years ................................years ................................years ................................years ................................years ................................
Male
years ................................years ................................years ................................years ................................years ................................years ................................
Female
years ................................years ................................years ................................years ................................years ................................years ................................
Body mass index
Male
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Milligrams/deciliter
214.2
212.6215.6
214.6211.1
179.0195.5216.3236.0238.9239.6
175.4197.2224.8234.5231.3220.5
181.4193.7208.2237.2245.5255.0
182.2208.9223.4222.5
192.0207.2227.5232.2
1.50
2.302.19
1.624.49
2.603.433.382.836.013.61
4.445.176.393.918.133.69
3.773.553.954.437.825.63
5.015.393.983,68
5.134.523.422.90
1,276
489787
1,018258
224236210160116330
62726071
1$
1621641508961
161
103124133129
163197204223
Milligrams/deciliter
209.8
208.9210.5
210.3203.8
180.7198.0209.7226.2234.0241.0
176.0204.2216.3223.3226.2227.7
183.2192.6204.0229.0239.9252.5
193.9206.0220.4214.7
192.1207.0219.4226.6
1.54
2.261.58
1.603.52
1.962.692.643.103.282.83
2.434.085.194.794.552.70
3.312.422.063.864.504.45
3.203.023.943.87
3.332.532.723.34
2,976
1,1371,841
2,533445
551662551345244625
191195155171108319
360467396174138306
294273287283
468453472448
Milligrams/deciliter
211.3
211.2211.3
211.7206.7
182.3197.0215.5226.9237.3236.0
183.3201.0220.4227.4223.1224.9
181.3193.3210.7226.4254.8249.2
194.2208.9218.0224.8
193.3203.3220.8230.9
1.36
1.751.93
1.424.16
2.431.712.223.684.332.01
3.793.293.484.554.582.43
2.542.022.805.185.073.30
2.403.673.142.92
2.472.993.264.03
2,976
1,2311,745
2,551425
554645538386249604
192194165186140354
362451373200109250
316326299289
460443418424
Milligrams/deciIiter
1.87212.5
211.2214.0
212.4212.7
182.9198.7214.8230.8233.0246.7
179.3201.5217.1226.5229.8238.2
187.5195.2212.7236.2239.0256.5
192.3204.4218.3227.7
197.2210.5221.4229.4
2.342.19
2.005.20
2.053.223.413.865.337.06
4.274.985.924.516.20
11.52
4.012.883.366.097.155.14
4.604.264.405.42
3.464.654.524.03
1,275
586689
1,043232
203285238167106276
61104
7199
IE
122181167
6840
111
147138142159
175174170170
.Pul I Excludes “other” racial groupa.
Table 15. Serum cholesterol levels of adults ages 18-74 years within Iinoleic to saturated fatty acid intake ratio strata showing means and standard errors of means by sex, race, ageand body mass index: United States, 1971–74
Lino/eic/saturated fatty acid ratio (alai& intake)
Sex, race, age, and bodyLess than 0.1163 0. 1163-Q2693 0.2694+551 1 0.5512 or more
mass index quartile strata Standard Number Standard Number Standard Number StandardMean error of of Mean
Numbererror of of Mean error of of
meanMean
examineeserror of of
mean examinees mean examinees mean examinees
Totall ...........................................
SexMale .............................................Female .........................................
Race
White ............................................Black ............................................
Age
Total
18-24 years ................................25-34 years ................................35-44 years ................................45–54 years ................................55-84 years ................................85–74 years ................................
Male
18-24 years ................................25-34 years ................................35-44 years ................................45-54 years ................................55-84 years ................................65-74 years ................................
Female
18-24 years ................................25-34 years ................................35-44 years ................................45–54 years ................................55-84 years ................................65–74 years ................................
Body mass index
Male
Ist quartile, .................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th auartile .................................
Milligrams/deciliter
212.6
210.8214.5
212.7210.1
184.0193.8213.2228.0245.9239.2
183.9196.4216.0226.0232.1230.6
184.2191.1210.6230.5260.1249.3
190.2203.8222.0225.8
192.9211.5219.7234.9
2.11
2.312.75
2.324.10
3.132.313.185.726.325.92
4.773.394.817.778.677.97
4.473.952.956.267.507.50
5.383.794.425.23
3.634.384.916.39
1,270
564706
1,123147
230232z~~
141115332
8079688060
197
1501531526155
135
145133132154
170178179179
Milligrams/deciliter Milligrams/deciliter
210.9
210.6211.3
211.4206.5
177.5197.9214.5226.7232.1237.5
174.3198.6222.5226.9223.2226.1
180.4197.3205.4226.4244.2250.8
193.4207.2222.2220.3
192.0208.1221.3227.2
1.16
1.451.63
1.272.75
1.802.012.282.913.931.69
2’173.343.633.024.662.43
2.781.872.234.754.883.78
2.183.563.222.61
2.742.103.003.80
2,963
1,2791,684
2,447516
483607550
389251663
j&j
197184195139398
317410366194112285
330306334308
434414395441
211.3
211.0211.7
211.4210.4
181.2197.7,214.1232.0233.9240.9
180.8203.0220.2230.9226.2222.5
181.6192.8209.4233.0240.0256.2
191.9209.3216.5224.2
193.7203.4221.2230.8
1.04
1.821.46
1.133.96
2.212.142.eo2.453.673.98
3.26
3.245.414.023.743.48
2.642.112.284.525.085.03
3.122.953.632.95
3.332.982.952.75
2,962
1,1421,820
2,483479
567656540362249586
195201144180118303
371457396182131283
287310276269
463449466442
Milligrams/deciliter
211.0
209.6212.2
212.0203.1
186.8196.5210.0226.9236.0245.4
1e4.9205.0210.7218.9230.1236.2
188.4190.1209.6233.0245.7251.4
197.9203.6216.7215.5
195.3203.9223.0226.2
2.02
2.622.52
2.183.85
4.172.883.564.795.563.60
5.015,816.956.477.744.86
4.642.443.376.207.485.11
5.726.234.864.44
5.025:315.064.12
1,269
428841
1,060209
24332021716197
231
7%81506746
106
167239167
9449
125
89101113125
195222223201
j Excludes “other” racial groups.
Table 16. Serum cholesterol levels of adult males ages 18-74 years within dietary cholesterol strata showing means and standard errors of means by race, age and body mass indexUnited States, 1971-74
Average daily ctrolesterol intake (milligrams)
Race, age, and body mass Less than 177.945 177.945-412.874 412.875-435.584 835.585 or more
index quartile strata Standard Number Standard Number Standard NumberMean
Standard Numbererror of of Mean error of of Mean error of of Mean error of ofmean examirreas mean examirraes mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Total! ........................................... 212.3 2.34 512 211.4 1.75 1,195 211.8 1.70 1,195 205.4 2.81 541
RaceWhite ............................................ 213.7 2.42 425 211.7 1.81 1,036 212.4 1.87 1,017 206.5 2.90 449Black ............................................ 199.5 6.79 67 207.5 4.62 159 206.5 4.79 178 197.3 5.51 92
Age13-24 years ................................ 177.5 4.42 65 183.0 3.14 191 178.3 2.84 163 177.7 4,35 10725-34 years ................................ 198.1 5.56 62 204.0 3.03 200 202.8 4.18 190 197.6 5.19 1133544 years ................................ 223.0 8.14 34 219.5 4.29 145 220.5 4.53 166 215.1 6.16 10645-54 years ................................ 235.1 5.03 70 224.5 3.98 191 227.0 4.76 182 225.0 5.18 8455-64 years ................................ 227.2 7.39 56 224.4 5.11 113 225.2 3.53 147 235.5 9.59 5165-74 years ................................ 220.0 3.30 225 231.4 4.37 355 227.9 2.91 347 222.7 5.80 80
Body mass index1st quartile .................................. 192.7 5.05 136 190.0 3.15 266 197.7 3.08 301 186.2 4.84 1572d quartile ................................... 214.9 5.62 110 208.7 2.80 326 207.6 3.49 301 199.1 4.83 1243d quartile ................................... 227.9 5.28 126 220.3 2,80 305 217,1 4.05 297 218.0 4.20 1334th quartile ................................. 213.1 3.70 140 222.4 3.73 296 225.2 2.34 295 220.4 5.49 127
*Excludes“other”racialgroups.
& Table 17. Serum cholesterol levels of adult females ages 18-74 years within dietary cholesterol strata showing means and standard errors of means by race, age and body massindex United States, 1971-74
Average daiiy cholesterol intake (milligrams)
Race, age, and body massLess than 106.52 106.52-244.78 244.79-570.78 570.79 or more
index quartile strata Standard Number StandardMean
Number Standard Number Standarderror of of Mean error of of Mean
Numbererror of of Mean error of of
mean examinees mean examinees.
mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 213.6 2.18 758 212.5 2.12 1,768 211.0 1.20 1,768 212.2 2.79 768
Race
White ............................................ 212.7 2.25 619Black ............................................
212.8219.4
2.226.97
1,500 211.1139 208.1
1.24 1,4924.23 268
212.5209.4
3.034.11
607276 210.3 3.65 161
Age
18-24 years ................................ 185.6 4.22 154 181.9 2.7125-34 years ................................ 192.8 3.33
371 183.8165
3.06193.6 2.34
325 180.1 3.45 156
3544 years ................................ 209.0 3.87451 193.4
146 205.61.99
2.06426 193.2 3.18 221
45-54 years ................................ 231.8 4.21374 207.1
74 236.32.08
3.60397 217.6 3.41 169
55-64 years ................................ 237.9 6.53167 225.5
63 251.53.61
5.89191 229.0 7.80 99
65-74 years ................................ 255.5118
7.66240.4
156 256.15.08
4.17115
287252.6
246.16.30
3.8452
314 261.3 7.73 71
Body mass index1st quartile .................................. j 99.F3 4.81 156 189.3 2.83
----
2d quartile ...................................432 1Y4.U
204.5 4.86 182 206.53.16
2.92490 195.5 3.99 188
3d quartile ................................... 217.7468
4.09206.5
202 228.92.28 431
2.95206.9 3.55 186
4th quartile ................................. 229.0 2.69459 217.2
218 228.43.64
3.63416
409215.6
228.94.76
3.23187
431 233.3 7.06 207
1Excludes “othed’ racial groups.
Table 18. Serum cholesterol levels of adults ages 18-74 years within dietary sodium/potassium ratio strata showing means and standard errors of means by sex, race, age and bodymass index United States, 1971–74
Dietaty sodium/potassium ratio
Sexj race, age, and bodyLess than 0.5335 0.5335-0.9526 0.9527-1.6360 1.6361 or more
mass index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Total~ ...........................................
SexMale .............................................Female .........................................
Race
White ............................................Black ............................................
Age
Total
18-24 years ................................25-34 years ................................35-44 years ................................45-54 years ................................55-84 years ................................65-74 years ................................
Male
18-24 years ................................25-34 years ................................35-44 years ................................45-54 years ................................55-84 years ................................65-74 years ................................
Female
18-24 years ................................25-34 years ................................35-44 years ................................45-54 years ................................55-84 years ................................65-74 years ................................
Body mass index
Male
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Milligrams/deciliter
216.7
215.3217.6
216.9214.2
184.2198.2214.5236.0239.9245.6
181.4203.1228.0235.6226.5227.9
186.0195.2207.3236.2248.8257.5
204.2204.5222.8227.1
193.9214.8226.4234.3
1.37
2.751.92
1.454.60
3.033.283.633.035.353.52
5.485.766.605.026.694.77
3.752.544.123.297.565.28
6.595.794.454.68
3.193.445.073.64
1,276
408668
1,097179
208256235169120288
5864535550
128
15019218211470
160
1::105113
198219222229
Milligrams/deciliter
211.9
211.2212.7
212.3208.0
180.5196.1212.6232.8233.8243.6
180.6200.3216.6234.4220.4230.8
180.4191.0209.1231.1245.8256.0
192.9210.7219.6220.7
194.6207.8219.4233.2
1.22
1.491.64
1.363.10
2.042.012.022.433.643.91
2.913.153.343.233.884.94
2.472.371.663.854.274.75
3.152.583.312.93
2.702.632.873.60
2,977
1,2121,765
2,584393
492666525382238874
178227150190113354
314439375192125320
293314304300
425460447413
210.1
210.2210.0
210.1209.7
180.7199.2213.5221.8234.4234.8
179.6205.4217.4219.6229.0222.6
181.6193.8209.6224.5240.4248.3
192.5207.0219.5220.6
192.8202.0220.2225.5
1.33
1.701.60
1.404.19
1.832.412.852.413.592.99
2.023.514.403.603.992.55
3.032.552.593.775.014.95
2.383.183.462.64
2.323.122.753.09
2,977
1,2751,702
2,507470
532625572353270625
193189179198146372
339436393157124253
343301314317
448401434419
Milligrams/deciliter
207.4
207.1207.6
208.6200.7
181.8196.2214.7223.9240.4233.3
177.3196.7222.3218.2230.2226.8
185.9195.6203.4230.9259.4242.1
186.2203.1217.6221.6
189.7198.1221.8224.3
2.37
3.463.02
2.403.65
4.033.664.125.698.395.19
6.035.506.428.47
10.205.45
4.723.773.255.12
12.498.15
5.455.144.345.87
6.993.714.094.53
1,275
548727
957316
300281205154
87248
9785698658
153
203196136
682995
139141138130
195167161204
.z t Excludes“other”racialgroups.
(no Table 19. Serum cholesterol levels of adult females ages 16-74 years by extent of use of oral contraceptive agents showing means and standard errors of means by race, age and
body mass index United States, 1971–74
Oral contraceptive use
Race,l age, and body mass Not usedin past 6 months (.Aedin past 6 months but not now Use now
index quartile strata Standard Number Standard NumberMean
Standarderror of
Numberof Mean error of of Mean
meanerror of of
examinees mean examinees mean examinees
Milligrame/deciliter Milligrams/deciliter
White
Ail ages 16-44 years ........................................................18-24 years ................ .........................................................25-34 years ..........................................................................35-44 years .................................................... .....................
Black
All ages 16-44 years ........................................................
16-24 years ....... ..................................................................25-34 years ..........................................................................35-44 years ..........................................................................
Body mass index
All ages: 16-44 years1St quafitle ............................................................................2d quafiile .............................................................................3d quatile .............................................................................4th quartile ...........................................................................
18-24 years
1st tetile ...............................................................................2d tetile ................................................................................3d tertle ................................................................................
24-34 years
Ist t*le ...............................................................................2d tertile ................................................................................3d tetile ................................................................................
35-44 yeara
1st teflile ................. ..............................................................2d tertile ................................................................................3d tertile ................................................................................
191.8 0.98175.6 2.18192.3 1.74205.1 1.42
2,829
7141,0381,077
194.6
166.8195.8209.7
3.33
4.535.247.68
213
7610730
198.0 1.67195.3 3.20197.0 2.04210.1 4.18
864
373379132
197.7 1.91181.5 3.32199.3 4.29211.1 3.15
669
182231256
192.0186.5194.2208.0
6.64
12.536.05
18.79
55
2724
4
191.8 4.03194.3 5.73189.9 7.72185.7 8.55
1941126022
.—-1 f&3189.3198.2204.7
1.501.971.331.85
788869910988
183.8192.4198.6215.1
4.455.315.097.01
86696650
196.3193.1197.9208.4
3.312.342.044.24
355300254184
.-. .10%0
177.2188.3
x%2.643.38
---Zfz
337309
1/1.8192.2203.6
4.966.34”8.47
194.1191.4201.8
6.592.713.79
146185160
182.0194.7200.1
2.333.012.32
330478474
190.1188.7212.2
6.264.979.06
425635
194.5193.2204.6
3.433.474.59
188149107
197.0 1.70 368 202.5205.6
10.511.86
16505
212.4235.2
7.58 60
214.821.39 6
2.43 461207.3
207.85.64
9.4559
12 200.1 8.12 39
! Excludes “other” racial groups.
Table 20. Serum cholesterol levels of adults ages 16-74 years within socioeconomic class strata showing means and standard errors of means by race, age and body mass indexUnited States, 1971-74
Socioeconomic status
Sex, race, age, and Low Low middle Midde U~r middle Upperbody mass index
quartile strata Standard Number Standard Number Standard Numlwr Skurdard Number Standard NumLrarMean error of of Mean envr of of Mean error of of Mean error of of Mean emor of of
mean examinaes mean examineas mean examirwes mean axaminaes mean examinees
Milligrams/deciliter
Totall .......................... 222.5 1.73
2.532.10
2.252.66
3.623.175.755.574.082.19
6.487.496.009.655.513.32
6.713.376.896.735.083.71
2.033.753.163.15
4.724.312.503.10
1,861
6181,043
1,278583
169157176191217951
60294984
105491
109128127107112460
496394335636
304368542647
213.5
209.5216.7
214.8207.1
180.6199.6212.9225.7236.7239,8
179.1203.7218.2220.1220.6228.4
161.9197.1206.4229.7248.1250.3
198.2211.1224.5225.7
195.6202.6211.5225.8
1.43
2.351.66
1.723.13
2.542.783.372.514.563.44
3.655.716.244.186.464.78
3.601.762.044.134.833.63
1.982.942.792.16
2.613.242.402.04
3,026
1,0911,935
2,286740
557583483333287783
170134125143128391
387449358190159392
945660561860
608661878879
MNigrams/deoiliter
sexMale ........................._ 215.8Female ........................ 228.1
RaceWhtie ........................... 224.1Black ........................... 217.3
Age—
Total18-24 years ............... 181.125-34 years ............... 198.635-44 years ............... 206.945-54 years ............... 221.055-64 years ............... 239.565-74 years ............... 240.7
Male16-24 years ............... 180.625-34 years ............... 189.835-44 years ............... 203.045-54 years ............... 217.155-64 years ............... 230.765-74 years ............... 230.3
Female
16-24 years ............... 181.625-34 years ............... 203.03544 years ............... 210.645-54 years...,,., ........ 223.855-64 years ............... 247.765-74 years ............... 249.6
Body mass index
Male
Ist quartile ................. 206.32d quartile .................. 222.83d quartile .................. 231.04th quartile ................ 230.9
Female
Ist quartile ................. 204.82d quartile .................. 214.73d quartile .................. 226.34th quartile ................ 230.6
mI Excludes “other” racial groups.
205.4
204.8205.9
206.0196.2
179.8195.4214.3226.8235.3234.0
176.5199,4223.2227.6224.8216.2
182.7191.3207.7229.6246.9248.4
190.7204.5214.7219.2
185.2200.4211.2218.7
1.27
1.701.79
1.293.56
2.062.252.573.083.373.93
2.973.724.974.434.723.61
2.842.472.174.146.105.65
1.982.753.231.98
2.542.092.182.12
2,667
1,0391,626
2,343324
752573469336i98319
278192127176102164
474381362160
96155
893640507627
526667840634
Milligrams/deciliter
210.1
212.0208.2
210.3205.7
162.0197.6212.6234.3239.7238.0
179.6202.4220.6232.0237.3219.6
184.2192.6204.5236.9242.6254.9
195.2206.8221.2226.5
191.5200.6217.4226.5
1,16
1.501.51
1.225.26
2.14*.682.053.194.053.14
3.182.473.343.444.783.54
2.341.832.254.745.774.46
1.682.072.242.26
2.151.971.762.31
3,440
1,3062,134
3,206234
6351,042
799435216313
206336258224117165
429706541211
Iz
1,322797643676
762927
1,039692
Milligrams/deciliter
208.0
214.3201.2
207.4237.5
173.1197.6207.6231.7217.8231.9
182.0198.3216.9243.2221.6225.4
165.5196.9196.4218.3214.7241.9
192.4210.0222.9219.6
189.7199.2219.5221.5
2.45
4.082.53
2.4713.38
7,054.033.494.147.007.89
10.276.415.766.298.83
10.01
5.674.513.916.739.009.84
3.214.685.235.45
4.823.244.514.30
500
199301
42185126
673644
12614434
E
3012482331715
2061438269
126152150
72
mM Table 21. Serum cholesterol levels of adults ages 18-74 years within geographic region showing means and standard errors of means by race, age and body mass index United
States, 1971-74
Geographic region,4, – –.. .
Sex, race, age, and bodyNortheast
mass index quartile strata Standard NumberMean error of of
mean examinees
Total} ...........................................
SexMale .............................................Female .........................................
Race
White ............................................Black ............................................
Age—
Total
18–24 years ................................25-34 years ................................35-zt4 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Malej&~A ~g~rs ................................25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Female
18-24 years ................................25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Body mass index
Male
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Milligrams/deciliter
216.4
215.8216.9
216.9210.3
180.2200.1215.8
235.0239.0240.4
~~o.~205.4228.5234.5230.7221.7
180.2195.0205.4235.4245.8254.8
194.9210.3228.7230.2
198.2207.1228.7238.1
1.28
2.351.68
1.421.97
1.362.082.654.253.642.64
~.23
3.885.446.894.923.80
2.872.342.315.163.692.64
3.463.324.243.49
2.223.243.564.51
2,901
1,1111,790
2,480421
4536075+9
359272691
~~~
180156163124332
297427
196148359
273
239290
476432460422
MIUwest South wesr
Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees
212.1
209.2214.9
211.9214.4
179.8195.8213.4231.3236.2235.7
175.5
196.4217.7226.7229.6221.3
183.2195.2209.5236.5243.0246.2
191.2209.8215.1219.4
196.7209.0227.4230.7
1.18
1.671.53
1.314.17
3.281.742.232.194.992.39
&$~
2.093.923.436.101.61
3.742.482.302.884.602.78
2.924.312.803.73
3.232.081.963.25
3,215
1,2451,970
2,728487
532698574396283732
?g7
213167199133346
345465407197150386
317279319330
546490448486
Milligrams/deciliter
212.1
209.6214.2
212.2211.7
181.1199.4209.7227.5237.6237.0
477C,,, .”
202.5215.2226.7228.5223.1
184.1196.7205.2228.1246.0247.1
198.3213.6213.6219.3
199.2213.4223.5227.6
1.99
2.312.34
2.122.94
3.153.893.692.975.581.49
~.~o
6.298.125.416.032.27
3.702.612.004.965.893.26
2.504.993.893.56
3.422.283.923.64
3,759
1,4202,339
2,7061,053
658657570412350
1,112
218196143191161511
440481427221189601
498351271300
678534555572
Milligrams/deciliter
212.2 3,604
212.2212.2
212.3210.8
181.2195.32i 3.3228.1238.1245.1
179.0200.9220.4228.1227.5237.0
163.0190.5206.3228.1246.8252.0
192.6209.1223.9227.2
1.74
1.622.26
1.655.65
2.462.473.622.114.373.36
~ ~~
3:395.352.527.034.88
3.351.973.204.234.284.45
1.943.203.533.01
1.983.SQ5.093.52
1,4032,201
3,170434
609697628420346904
. .-.rI za197191207166447
414500437213180457
308355356312
195.121 4=0225.7225.4
1Excludes “other” racial groups.
Table 22. Serum cholesterol levels of adult males ages 25-74 yeare within total general well-being strata showing means and standard errors of means by race, age and body massindex United States, 1971-75
Total general well-being score
Race, age, and body mass Less than 67 67-66 87-99 100 or more
index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examirrees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 217.9 2.56 437 218.3 1.89 1,114 217.9 1.83 1,105 221.7 3.30 421
RaceWhite ............................................ 215.4 2.76 350 217.5 1.71 984 217.8 1.93 1,000 222.3 3.60 361Black ............................................ 233.1 9.40 87 226.0 9.89 130 219.4 4.63 105 215.7 6.57 60
Age25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Body mass index1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
194.9219.5221.5233.6224.6
203.1219.7225.3226.5
5.025.175.864.897.29
3.935.165.648.03
7767
11910173
13384
102117
204.7215.3227.5229.1226.6
207.6216.6221.6226.9
4.403.003.503.444.13
3.453.583.164.32
264189260211190
274285275280
200.0217.6234.5225.5218.1
202.1219.7219.0230.1
3.042.963.533.343.23
4.082.982.543.30
237195265202206
253299290262
206.5222.3233.6220.5228.7
203.4228.1229.7222.5
4.718.036.325.374.48
4.095.396.544.59
795989
1%
10410399
114
~Excludes “other” racial groups.
Table 23. Serum cholesterol levels of adult females ages 25-74 years within total general well-being strata showing means and standard errors of means by race, age and body massindex United States, 1971-75
Total general well-being score
Race, aae, and bodv massLess than 60 60-79 80-95 96 or more
iride; quartile st;ata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
“Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 220.5 2.65 541 224.1 1.43 1,275 220.2 1.74 1,263 225.8 2.62 531
RaceWhite ...................................... ...... 221.8 2.92 417 225.2 1.65 1,093 220.0 1.78 1,143 225.0 2.84 484Black ............................................ 215.4 5.33 ?24 216.2 4.48 182 222.9 4.55 120 235.8 8.39 47
Age
25-34 years ........... .................... 194.6 4.70 124 201.9 2.36 313 192.5 2.72 31635-44 years ................................
190.3 4.36 lot211.5 4.54 107 208.6 3.07 234 205.0 3.21 243 203.4 4.29 76
45-54 years ................................ 227.0 6.06 127 233.1 3.91 287 232.7 3.34 295 235.2 6.19 13955-84 years ................................ 251.1 6.26 93 242.9 4.06 234 245.2 3.99 201 247.8 5.26 11265-74 years ................................ 242.2 5.05 90 254.3 3.18 207 250.4 3.64 208 251.4 4.93 103
Body mass index ‘
1st quartile .................................. 213.1 4.49 112 211.7 2.71 315 205.5 3.03 3312d quartile ...................................
203.6 4.04 137215.1 4.33 t 22 222.0 3.17 2S8 216.5 2.88 353
3d qutiile ................................... 226.3 4.88 125222.0 4.45 141
232.1 3.56 306 227.8 4.48 318 234.9 5.16 1494th quartile ................................. 227.0 6.31 181 231.6 2.51 353 236.4 3.58 261 246.7 6.31 104
~Excludes“othsr”rscislgroups.
Table 24. Serum cholesterol levels of adult males ages 18-74 years within hemoglobin strata showing means and standard errors of means by race, age and body mass indexUnited States, 1971-74
Hemoglobin level (grams/deciliter)
Race, age, and body mass Less than 14.35 14.35-15.54 15.55-16.74 16.75 or more
index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
maan examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
TotaP ........................................... 208.6 2.60 729 208.2 1.29 1,647 210.9 1.80 1,443 218.4 2.76 634
RaceWhite ............................................ 208.4 2.57 522 209.1 1.29 1,355 210.6 1.79 1,322 217.3 2.80 589Black ............................................ 209.8 7.61 207 201.6 3.82 292 215.9 10.22 121 244.9 20.02 45
Age18-24 years ................................ 173.6 4.75 60 173.4 2.06 255 180.9 3.05 282 185.0 4.98 11225-34 years ................................ 194.8 4.84 88 197.8 3.08 256 203.4 3.18 278 207.4 4.90 11835-44 years................................ 213.3 6.18 66 216.6 3.47 254 224.0 3.55 195 232.8 5.36 8445-54 years ................................ 214.7 5,92 101 223.9 2.93 237 232.3 3.36 219 240.2 4.82 11055-64 years................................ 229.6 6.64 89 226.8 3.60 186 230.8 5.33 136 229.7 10.55 6365-74 years................................ 215.1 2.75 325 231.0 4.19 459 226.2 3.84 333 233.6 4.69 147
Body mass index1st quartile .................................. 191.1 3.45 271 186.6 2.78 398 195.8 3.73 320 193.8 4,37 1132d quartile ................................... 208.3 5.14 179 205.7 3.20 421 206.7 3.77 359 214.3 5.22 1623d quartile ................................... 212.0 3.00 145 219.2 2.62 440 220.3 3.04 357 226.0 5.22 1734th quartile ................................. 232.7 6.05 134 218.5 3.08 390 219.4 3.13 407 230.7 6.09 186
1Excludes“other” racialgroups.
mm Table 25. Serum cholesterol levels of adult females ages 18–74 years within hemoglobin strata showing means and standard errors of means by race, age and body mass index
United States, 1971-74
Hemoglobin level (grams/deciliter)
Race, age, and body massLess than 12.65 12.65-13.74 13.75-14.94 14.95 or more
index quartile strata Standard Number Standard Number Standard NumberMean
Standard Numbererror of of Mean error of of Mean error of of Mean error of ofmean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totalj ........................................... 204.7 1.51 1,216 206.9 1.56 2,327 213.4 1.65 2,462 224.8 2.34 929
Race
White ............................................ 205.5 2.08 797 206.5 1.72 1,936 213.1 1.70B[ack ............................................ 202.3
2,1753.75 419
224.9209.7 2.30
2.37 863391 217.3 4.44 267 223.9 7.23 66
Age
18–24 years ................................ 180.9 2.94 282 180.3 2.46 537 186.8 2.94 463 189.4 5.9425–34 years ................................ 193.8 2.95 315 190.8 1.70 604
115195.1 1.67
35-44 years ................................617 201.8
206.0 3.99 2712.29
201.1 2.15203
503 207.7 1.8745–54 years ................................ 220.0
4934.56
215.7110
3.44229.1 4.20 205
163227.7 4.64
55-64 years ................................ 247.2256
10.11236.8
556.84
245.7 4.23 140116
242.6 3.6465-74 years ................................
191244.6 7.49
250.6183 246.4
5.104.52 338
107254.5 3.46 442 254.4 3.38 205
Body mass index
1st quartile .................................. 185.1 3.73 307 190.8 2.37 626 197.7 3.402d quartile ................................... 200.6 3.44
585 209.2325 202.3
3.802.77
211615 208.8 2.26
3d quartile ................................... 218.2 4.49633
318 216.7216.3 5.33
1.96 566180
221.5 1.984th quartile ................................. 219.4 5.05
625266
229.2224.0
5.273.10 520
243225.5 4.29 619 239.9 4.57 295
I Excludes “other” racial groups.
Table 26. Serum cholesterol levels of adults ages 25-74 years within SGOT strata showing means and standard errors of means by sex, race, age and body mass index UnitedStates, 1971-75
SGOT level (units/millimeter)
Sex, race, age, and body Less than 16.25 16.25-21.74 21.75-29.74 29.75 or more
mass index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Total~ ...........................................
SexMale .............................................Female .........................................
RaceWhite ............................................Black ............................................
Age—
Total25-34 years ................................35-44 years ................................45-54 years ................................55-84 years ................................65-74 years ................................
Male25-34 years ................................3!544 years ................................45-54 years ................................55-84 years ................................65-74 years ................................
Female25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Body mass index
Male
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th cruartile .................................
Milligrams/deciliter
210.0
213.5208.8
210.5204.5
189.9210.6222.6238.3232.0
193.7211.2223.6235.9225.5
189.0210.4222.2239.3236,5
205.5211.0215.8223.6
197.9204.2218.0221.6
1.42
3.631.78
1.553.80
2.943.474.454.385.05
6.436.266.886.495.92
3.024.135.745.838.00
5.356.367.567.52
3.723.044.354.57
812
211601
710102
265192158102
94
4841423!545
217151117
6749
65544943
154179142126
Milligrams/deciliter
216.6
214.2216.7
217.2211.5
196.8209.8226.7235.7244.4
196.4214.8222.7223.5227.7
195.6206.2229.0247.1258.0
202.6217.0218.9220.8
204.8216.2225.7229.2
1.40
1.991.59
1.474.72
2.002.562.063.183.09
3.363.663.034.254.25
2.373.612.673.463.59
3.403.633.344.45
3.343.253.083.60
Milligrams/deciliter
1,850
7151,135
1,660170
548362391302247
193121137143121
355241254159126
192199179144
299272286275
221.5
219.2224.5
220.9227.4
203.0213.8229.6234.9236.1
202.1220.7228.9226,2224.7
204.4203.1230.7241.6250.5
199.5221.2223.0230.2
216.5224.6226.6230.0
1.53
2.071.97
1.327.42
3.733.243.243.322.96
4.813.664.194.463.73
4.423.723.844.444.46
2.824.012.664.12
3.773.414.824.49
1,636
1,012826
1,647191
378309481350320
215181257161176
163128224169142
245267236262
193210204216
Milligrams/deciliter
223.1
219.2232.4
223.2222.5
206.1212.9235.1239.6226.3
206.5213.4226.8234.2220.2
206.3211.5249.4246.9236.6
201.3220.4221.7227.7
224.1217.0245.1235.1
2.20
2.583.70
2.495.86
3.673.444.504.764.50
4.574.434.236.005.16
11.806.818.195.868.96
6.118.214.104.12
7.518.667.426.63
764
535229
655109
173132209138112
13194
1459273
4238644639
113105135182
46466077
t Excludes“other”racialgroups.
(X4
Uim Table 27. Serum cholesterol levels of adults ages 25-74 years within serum calcium strata showing means and standard errors of means by sex, race, age and body mass index
United States, 1971-75
Serum calcium level (milligrams/deciliterj
Sex, race, age, and bodyLess than 9.25 9.25-9.74 9.75-10.14 10.75 or more
mass index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter MNigrams/deciliter
Totall ...........................................
Sex
Male .............................................Female .........................................
Race
White ............................................Black ............................................
25-3435-4445-5455-8465-74
25-343544
45-5455-8465-74
25-3435-4445-5455-8465-74
Age
Total
years ................................years ................................years ................................years ................................years ................................
Male
years ................................years ................................years ................................years ................................years ................................
Femaleyears ................................years ................................years ................................years ................................years ................................
Body mass index
Male
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
204.4
205.0204.0
204.9199.3
186.9195.6214.0212.1225.8
185.2205.4212.2204.6211.2
187.4192.0215.6221.1239.5
190.9208.6210.3212.5
192.2203.2214.4207.6
1.34
1.881.69
1.445.99
4.142.323.124.153.99
4.074.714.015.125.98
5.152.753.855.774.95
4.194.862.406.55
4.223.834.843.80
Milligrams/deciliter
812
323489
73082
180162i 63137150
3743667582
143119
976268
86729174
112153106118
214.9
213.2216.4
215.4208.6
192.3209.7224.3234.9232.3
193.1212.2221.3226.2221.6
191.7207.3227.1243.3243.7
201.3212.3219.0219.4
205.0211.9222.2226.9
1.04
1.631.84
1.173.06
2.112.701.972.593.15
2.833.’4s23.012.684.52
2.983.603.593.834.37
3.463.333.322.74
3.363.083.283.96
2,277
1,0521,225
2,042235
551416555404351
205176257217197
346240298187154
270248265268
303288322311
223.3
221.7225.0
223.3223.2
202.6214.6239.0242.6251.0
203.3218.2243.1239.7231.4
201.6209.9235.3245.1265.4
204.8222.5222.9233.3
210.8219.3235.5239.1
1.42
1.941.68
1.654.69
2.253.422.723.853.82
3.52. --4.!X+3.015.444.78
2.373.793.766.204.68
4.484.143.014.01
2.934.424.114.19
1,553
773780
1,353200
434301376244198
220181181117
94
214140195127104
165221189196
204194187192
MiHiarams/decilitel
233.2
232.1234.8
232.7236.9
216.1228.1244.9256.3244.3
214.4236.4250.3251.1251.0
220.4219.1238.8260.0237.9
217.8235.9235.1239.9
223.0224.8242.4250.3
2.63
3.383.66
2.836.31
4.444.975.253.729.95
4.886.776.664.86
12.17
8.026.597.465.25
12.35
5.135.996.686.90
5.596.388.727.32
613
341272
52687
18511913910961
12763674636
5656726125
95917283
74646173
I Excludes “other” racial groups.
Table 28, Serum cholesterol levels of adults ages 25-74 years within serum magnesium strata showing means and standard errors of means by sex, race, age and body mass index:United States, 1971-75
Serum macmesium level (milligrams/deciliter).-
Sex, race, age, and bodyLess than 1.555 1.555-1.684 1.685-1.824 1.825 or more
mass index auartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Totalt ...........................................
Sex
Male .............................................Female .........................................
Race
White ............................................Black ............................................
Age
Total
25–34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Male
25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Female
25-34 years ................................35-44 years ................................45-54 years ................................5544 years ................................65-74 years ................................
Body mass index
Male
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
MiHiqrams/deciliter
209.5
209.3209.5
209.2210.5
194.1199.5226.4228.9226.1
192.8203.8221.6221.4214.7
194.6197.3229.6235.7238.5
196.1214.1210.0218.0
200.8203.2215.1220.7
1.54
3.002.30
1.664.63
2.773.134.266.234.27
4.256.125.557.637.18
3.563.596.927.415.50
4.625.266.105.81
5.483.114.505.37
891
340551
713178
280168207122114
8154816163
1991141266151
98837584
142136128144
Milligrams/deciliter
215.6
215.9215.4
215.0221.6
199.2209.8225.7233.5236.1
204.2214.0225.9225.6223.4
194.7206.7225.5240.6248.8
200.2214.3219.8227.7
204.5212.3220.5225.1
1.54
2.091.70
1.297.22
3.362.122.583.953.56
5.102.694.434.844.44
2.763.042.654.835.54
2.943.433.365.37
3.122.894.433.34
Milligrams/deciliter Milligrams/deciliter
1,997
8891,108
1,771226
553416446322260
231165183160150
322251263162110
220228216224
266296271273
220.5
218.3223.1
220.7217.9
199.5213.8230.3238.6235.2
201.7217.5230.8232.0217.7
196.4208.8229.8244.2251.3
203.3224.6219.6226.5
210.1221.9228.7232.0
1.10
1.791.92
1.183.94
2.092.872.362.893.09
3.163.943.403.583.66
2.803.633,683.663.52
3.713.872.723.36
3.102.923.314.60
2,066
1,0531,013
1,854212
480366535362323
237183286176171
243183249186152
260251274267
249252269243
228.1
226.3230.4
227.9232.7
204.4222.8237.3242.0245.6
205.8230.4234.5229.6241.2
202.5210.7241.2256.3249.3
209.1222.5237.1233.5
214.9216.8247.1242.1
2.17
2.423.59
2.2210.62
3.464.823.993.534.06
3,486.053,874.836.73
6.465.507.595.415.91
3.434.444.384.19
7.705.627.634,20
615
449366
76253
161141188166159
9182
1049280
7059847479
102120117109
947891
102
I Excludes “other” racial groups.
8
Table 29. Serum cholesterol levels of adult males ages 25-74 years within serum urate strata showing means and standard errors of means by race, age and body mass indexUnited States, 1971-75
Serum urate level (milligrams/deciliter)
Race, age, and body massLess than 4.95 4.95-6.14 6.15-7.54 7.55 or more
index quartile strata Standard Number StandardMean
Number Standard Number Standard Numbererror of of Mean error of of Mean error ofmean
of Meanexaminees mean
error of ofexaminees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totail ........................................... 212.4 2.19 403 216.2 1.76 955 216.5 1.93 935 228.8 3.49
Race
438
White ...................................... . . 212.6 2.35 360 216.7 1.86Black ............................................ 209.3
8567.90
216.543
1.99210.6
8374.30 99
224.7 3.23216.2
3676.07 98 254.3 14.51
Age
71
25-34 years ................................ 188.2 4.12 71 202.6 3.91 24335-44 years ................................ 210.2 3.98
202.065
2.87213.9
2343.70
212.2 10.03 92
45-54 years ................................ 223.6177
6.84218.5
1024.01
228.7166
3.19 212228.0 5.54 76
55-64 years ................................ 221.6 4.40228.8
813.68
228.9221
3.50234.3 5.72 119
65-74 years ................................ 224.4172
6.32223.2
644.13
222.5167
3.52 151243.8 6.29
216.669
4.08 147 234.9 6.80 82
Body mass index
1st quartile .................................. 203.0 2.86 167 203.4 3.27 2692d quartile ................................... 219.6
197.0 4.095.71
175iil 217,8 3.35
209.0 6.26 69
3d quartile ................................... 211.8276
4.60218.5
874.65
219.6220
2.82 220231.9 7.24 75
4th quantize ................................. 232.2 7.63223.5
382.56
227.4258
5.28 188230.4 5.50
220.7117
2.55 281 234.2 6.31 177
1Excludes “other” racial groups.
Table 30. Serum cholesterol levels of adult females ages 25-74 years within serum urate strata showing means and standard errors of means by race, age and body mass indexUnited States, 1971-75
Serum urate level (milligrams/deciliter)
Race, age, and body massLess than 3.65 3.65-4.64 4.65-5.94 5.95 or more
index quartile strata Standard Number Standard NumberMean
Standard Number Standarderror of of Mean
Numbererror of of Mean error of
meanof
examineesMean error of
meanof
examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 207.2 2.44 478 214.1 1.74 1,054 224.1
Race
1.83 999 228.9 2.48 507
White..,..., .....................................Black
207.2 2.56 432 214.6206.7
1.73 952 225.2 2.07............................................ 6.43 46 208.2 5.19
889102
228.8211.9
2.474.98
407
Age
110 229.1 7.92 100
25-34 years ................................ 190.7 3.88 168 194.13%44 years ................................ 199.6
2.544.97
321105
198.9204.4
2.80 259 204.5 5.51
45-54 years ................................2.60 235 211.7
88
215.1 4.78 111 224.52.90 192 204.1 6.51
55-64 years ................................3.32 288 239.8
75
246.9 6.48 54 243.74.68 229 236.5 5.80
65-74 years ................................4.53 134 241.1
114
233.7 6.06 42 249.53.78
4.23176 245.9
96 251,25.91
4.47119
143 250.1 5.07 111
Body mass index
1st quartile .................................. 198.4 3.26 189 208.32d quartile ................................... 206.0 4.64
3.19 320135
210.8214.3
3.27 205 221.7 10.48
3d quartile ..................................2.72 319 220.1
37
218.7 4.99 109 216.42.72 225 211.1 7.27
4th quartile .................................3.26 245 231.1
83
220.6 9.09 45 221.73.17
3.88264
168234.2
231.16.61
3.86141
304 232.7 3,30 245
1Excludes “other” racial groupe.
Table 31. Serum urate levels of adults ages 25–74 years showing means, standard erros of means and selected percentiles by sex andrace: United States, 1971-75
StandardPercentile
NumberEstimated
Sex, race, and age Mean error of ofpopulation
the mean 5th 10th 25th 50th 75th 90th 95th examineesin
thousands
Male
Total’ ...................... ...................
White ...........................................
Age
25-34 years ...............................35-44 years ...............................45-54 years ...............................55–64 years ...............................65-74 years ...............................
Black ...........................................
Age
25-34 years ...............................35-44 years ...............................45-54 years ...............................55-64 years ...............................65-74 years ...............................
Female
Total’ ..........................................
White ...........................................
Age
25-34 years ...............................35-44 years ...............................45-54 years ...............................55-64 years ...............................65-74 years ...............................
Black ...........................................
Age
25-34 years ...............................35-44 years ...............................45-54 years ...............................55-64 years ...............................65-74 years ...............................
Milligrams/deciliter
6.3
6.2
6.26.26.36.26.3
6.6
6.36.76.66.96.6
4.8
4.8
4.54.64.85.25.3
5.1
4.64.75.35.85.6
0.02
0.02
0.050.060.050.060.06
0.07
0.150.210.150.170.17
0.02
0.02
0.040.040.040.060.06
0.06
0.120.130.130.170.18
4.3
4.3
4,64,54.13.7.4.5
4.5
4,54,64,74.43.7
2.9
2.9
2.82.93.03.43.3
2.9
2.82.93.52.93.3
4.7
4.7
4.94.74,64.44.7
4.8
4.94.84.94.94.1
3.5
3.5
3.13.33.53.73.7
3.5
3.13.53.73.53.7
5.5
5.4
5.55.45.45.45.4
5.6
5.75.65.65.85.6
4.0
4.0
3.83.83.94.24.4
4.0
3.83.94.34.54.3
6.2
6.2
6.26.26.36.26.3
6.5
6.36.56.57.06.7
4.7
4.7
4.44.54.75.15.3
4.9
4.54.55.25.65.6
7.2
7.1
7.07.07.37.27.2
7.7
7.27.67.88.07.8
5.6
5.5
5.25.35.56.16.3
6.1
5.55.46.37.36.9
8.1
8.1
7.87.98.28.16.2
8.7
6.29.59.18.68.5
6.6
6.5
5.86.26.47.17.2
7.2
6.36.67.38.17.8
8.7
8.5
8.48.56.78.68.9
9.5
8.610.0
9.89.29.6
7.2
7.2
6.36.77.17.67.6
7.9
6.97.27.98.48.4
3)1 71
2,744
592466647536501
390
7253997690
3,742
3,224
778580756572538
483
10497
1078590
48,857
43,903
11,6469,2199,8868,0064,946
44,114
1,2201,005
994724471
53,927
47,705
12,3249,518
10,5888,8766,399
5,733
1,6581,3481,253
845629
1Excludes “other” racial groups.
62
Table 32. Serum urate levels of adult males ages 25-74 years within body mass index strata showing means and standard errors of means by race and age United States, 1971-75
Body mass index (kilograms/meter#)
Less than 22.270 22.270-24.449 24.450-26.353 26.354-28.482 28.483 or moreRace and age
Standard Number Standard Number Standard NumberMean
Standard Number Standard Numbererror of of Mean error of of Mean error of of Mean error of of Mean error of ofmean examinees mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall .......................... 5.7 0.05 523 5.9 0.07 530 6.2 0.06 534 6.3 0.06 532 6.9 0.06 534
RaceWhite ........................... 5.7 0.06 454 5.9 0.07 479 6.2 0.06 476 6.3 0.07 489 6.8 0.08 466Black ........................... 5.7 0.23 69 6.1 0.17 51 6.6 0.19 58 6.3 0.23 43 7.4 0.22 68
Age25-34 years ............... 5.9 0.11 143 5.9 0.10 141 6.2 0.11 116 6.3 0.13 110 7.0 0.18 11435-44 years ............... 5.5 0.15 5.9 0.16 96 6.2 0.15 99 6.3 0.13 102 6.7 0.1645-54 years ............... 5.6 0.17 1% 6.0 0.11 118 6.2 0.12 132 6.4 0.12 138 6.9 0.14 1:55-64 years ............... 5.8 0.20 94 5.5 0.24 79 6.3 0.16 108 6.1 0.14 95 6.9 0.12 10565-74 years ............... 5.7 0.16 96 6.0 0.14 96 6.3 0.16 79 6.3 0.19 87 6.7 0.18 87
I Excludes “other” racial groups.
!3 Table 33. Serum urate levels of adult females ages 25-74 years within body mass index strata showing means and standard errors of means by race and age United States,1971-75
Body mass index (kilograms/metersz)
Less than 20.706 20.708-22.843 22.844-25.203 25.204-28.907 28.908 or moreRace and age
Standard Number Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliier Milligrams/deciliter Milligrams/deciliter MMgrams/deciliter
Totall .......................... 4.2 0.04 592 4.4 0.05 598 4.6 0.05 589 4.9 0.06 590 5.5 0.06 596
Race
White ........................... 4.2 0.05 549 4.4 0.06 565 4.6 0.05 536 4.9 0.07 501 5.5 0.07Black ........................... 4.3 0.15 43
4724.5 0.17 33 4.5 0.22 53 4.9 0.11 69 5.4 0.14 124
Age
25-34 years ............... 4.2 0.08 256 4.4 0.07 205 4.4 0.?1 136 4.5 0.14 101 5.3 0.1335-44 years ............... 4.1 0.08 131
1164.3 0.13 137 4.6 0.09 111 4.7 0.10 115 5.5 0.12
45-54 years ............... 4.2 0.08 108 4.395
0.10 142 4.6 0.10 145 5.0 0.11 166 5.3 0.1155-64 years ............... 4.3 0.17 53 4.7 0.18
15170 4.8 0.13 109 5.0 0.10 111 5.5 0.14
65-74 years ............... 4.5 0.15 42 4.7 0.18 44128
5.0 0.14 88 5.4 0.15 97 5.7 0.15 106
*Excludes“other”racialgroups.
Table 34. Serum urate levels of adult males ages 25-74 years within total skinfold (triceps and subscapular) thickness strata showing means and standard errors of means by raceand age United States, 1971-75
Total skinfold thickness (millimeters)
Less than 17.0 17.(X22.9 23.0-28.9Race and age
29.0-36.5 36.6 or more
Standard Number Standard Number Standard Number Standard NumberMean
Standard Numbererror of of Mean error of of Mean error of of Mean error of of Mean error of ofmean examinees mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Total’ ................ ......... 5.7 0.07 470 6.0 0.07 556 6.2 0.07 509 6.4 0.07 535 6.6 0.06 580
RaceWhite ........................... 5.6 0.06 385 6.0 0.07 497 6.2 0.07 461 6.4 0.06 474 6.6 0.06 523Black ........................... 5.9 0.16 65 6.1 0.22 59 6.8 0.26 28 6.7 0.26 61 7.0 0.28 57
Age25-34 years ............... 5.9 0.10 115 6.0 0.10 143 6.2 0.10 116 6.4 0.19 109 6.7 0.13 13835-44 veals ............... 5.6 0.15 75 6.1 0.17 93 6.2 0.14 89 6.3 0.15 106 6.5 0.15 10645-54 ~ears............... 5.6 0.15 105 6.1 0.12 123 6.4 0.14 120 6.4 0.13 117 6.6 0.10 16555-84 years ............... 5.7 0.25 78 5.8 0.14 110 6.3 0.15 93 6.5 0.15 105 6.6 0.14 9665-74 years ............... 5.6 0.13 97 6.3 0.12 87 6.1 0.18 89 6.5 0.13 98 6.4 0.21 75
1Excludes“other”racialgroups.
Table 35. Serum urate levels of adult females ages 26-74 years within total skinfold (triceps and subscapular) thickness strata showing means and standard errors of means by raceand a!ae United States, 1971-75
Total skinfold thickness (millimeters)
Less than 28.0 28.0-36.4 36.5-45.9 46.0-58.0 58.1 or moreRace and age
Standard Number - Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall .......................... 4.2 0.06 608 4.4 0.05 605 4.7 0.07 601 5.0 0.06 580 5.3 0.07 553
RaceWhite ........................... 4.2 0.06 551 4.4 0.05 566 4.6 0.07 561 5.0 0.06 489 5.4 0.07Black ........................... 4.5 0.14
44057 4.6 0.26 39 5.2 0.30 40 4.8 0.18 91 5.0 0.13 113
Age
25-34 years ............... 4.2 0.07 236 4.3 0.09 190 4.4 0.14 143 4.7 0.12 109 5.1 0.13 12535-44 years ............... 4.1 0.08 129 4.3 0.10 124 4.5 0.14 111 4.9 0.09 116 5.3 0.1045-54 years ............... 4.2 0.09
107113 4.4 0.09 125 4.6 0.09 156 4.9 0.12 150 5.2 0.11
55-64 years ............... 4.4 0.27 63 4.7 0.12 95164
4.8 0.13 99 5.2 0.11 113 5.5 0.16 9965-74 years ............... 4.6 0.14 65 4.7 0.14 71 5.3 0.15 90 5.4 0.19 92 5.8 0.18 58
I Excludes “other” racisl groups.
Table 36. Serum urate levels of adults agee 25-74 years wittin geographic region strata showing means and standard errors of means by race, age and body mass index UnitedStates, 1971-75
Geographic region
Sex, race, age, and body Northeast Midwest South West
mass index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examirrees mean examinees mean examinees
Milligrams/deciliter
5.5 0.05
Milligrams/deciliter
5.5 0.05
Milligrams/deciliter
5.4 0.07
Milligrams/deciliter
Totall ...........................................
SexMale .............................................Female.........................................
RaceWhite............................................Black ............................................
Age25-34 years................................35-44 years ................................45-54 years................................55-64 years..,, ............................65-74 years................................
Body mass index
1,333 1,421 1,352 5.4 0.04 1,518
6.3 0.064.7 0.06
630703
6.3 0.064.7 0.06
647774
6.2 0.094.7 0.07
648704
6.24.7
0.080.06
731787
0.07E 0.26
1,213120
5.4 0.055.9 0.12
1,269152
5.3 0.075.6 0.15
1,104248
5.45.5
0.040.18
1,406112
5.4 0.125.4 0.105.5 0.115.6 0.145.8 0.11
330258337231177
5.4 0.095.3 0.105.6 0.085.6 0.155.6 0.09
379285329236192
5.2 0.105.3 0.115.3 0.085.6 0.165.7 0.16
345226331226224
5.35.45.5
%
0.080.080.100.090.09
388290347261232
MaleIst quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
FemaleIst quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
5.96.26.26.9
0.130.100.090.11
155148157168
5.9 0.126.0 0.10
0.13% 0.16
130152197166
5.7 0.116.0 0.126.5 0.216.6 0.10
197165137148
5.56.06.36.7
0.060.120.100.16
173204171183
4.24.54.65.3
0.050.110.130.07
160155180207
4.2 0.084.5 0.114.6 0.135.6 0.10
196195188193
4.2 0.124.4 0.114.9 0.125.2 0.10
174171163174
4.34.64.85.3
0.090.090.120.16
212225163187
j Excludes“other” racialgroups.
Table 37. Serum urate levels of adults ages 25-74 years within strata of weekly ethanol consumption showing means and standard errors of means by sex, race, age and body massindex United States, 1971-74
Ethanol consumption (ounces/week)
Sex, race, age, and bodyAbstainers (0) Light (0.001-0999) Moderate (1.000-6.999) Heavy (~OOO or more)
mass index quartile strata Standard Number Standard NumberMean
Standard Number Standard Numbererror of of Mean error of of Mean error of of Mean error of ofmean examinees mean examinees mean examinees meen examinees
Totall ...........................................
SexMale .............................................Female .........................................
Race
White ............................................Black ............................................
Age—Total
25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Male
25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Female
25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Body mass index
Male
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
5.01
5.734.64
5.014.96
4.935.034.895.085.15
6.036.665.545.695.42
4.474.484.544.765.00
4.544.825.135.60
4.384.785.055.63
0.070
0.0650.077
0.0810.181
0.1740.1970.1190.1180.090
0.273- ---U.ZW0.1720.1860.135
0.1090.2290.1420.1270.102
0.1160.0860.1050.108
0.1230.1140.0810.120
668
238430
557111
8385
165155180
2027635375
6358
102102105
190153122203
122157208181
5.18
6.044.49
5.155.49
5.014.995.285.505.77
5.905.946.166.216.25
4.344.334.674.665.14
4.575.195.705.76
4.354.995.535.82
0.048
0.0570.056
0.0540.124
0.0970.1010.1050.1820.165
0.1120. i 440.1240.1620.176
0.0810.0830.1160.1810.176
0.0840.0970.1690.090
0.0670.1300.1020.105
786
363423
69195
227135167124113
9749786970
13086
1095543
267187139193
178190252166
5.71
6.284.75
5.695.88
5.755.335.885.616.09
6.445.856.456.266.33
4.464.724.715.175.49
5.015.676.036.53
4.855.395.986.68
0.057
0.0770.094
0.0600.263
0.1370.1200.1140.1780.164
0.1530.1500.1270.2350.186
0.1850.1850.1140.2200.194
0.1080.0950.0960.149
0.1630.1180.0760.166
750
488262
635115
186150211112
91
12284
1387569
6466733722
235187147181
147204245154
6.35
6.495.44
6.356.31
6.056.396.266.886.83
6.186.516.506.936.85
5.115.475.196.496.62
5.806.376.536.77
5.736.166.417.07
0.088
0.1040.201
0.0940.255
0.1280.1690.1380.2950.284
0.1570.1800.1710.3240.265
0.4320.4210.2840.4840.865
0.1170.1890.1870.227
0.1820.1570.143
228
19434
19533
4953623529
4044523127
99
1042
75505152
446285
0.215 37
t Excludes“other”racialgroups.
Table 38. Serum urate levels of adults ages 25-74 years within strata of dieta~ purine showing means and standard errors of means by sex, race, age and body mass index UnitedStates, 1971-74
Purine-riohfood krtake (times/wsak)
Sex, raoe, age, and body Less than 8.00 8.O&l 1.99 12.0&16.99 17.00 or more
mass index quartile strata Standard Numlwr Standard Number Standard Number Standard NumberMean errvr of of Mean error of of Mean error of of Mean error of of. .. .
mean examinees mean examirrees mssrr examinees mean examinees
TotaIl ...........................................
sexMale .............................................Female .........................................
Race
White ............................................Black ............................................
Age
Total
25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Male25-34 years ................................35-44 years ................................45-54 years ................................55-64 years................................65-74 years................................
Female25-34 years ................................35-44 years ................................46-54 yeara ................................55-64 years ................................65-74 years ................................
Body maaa index
Male
Ist quartile ..................................2d quartile.., ................................3d quarhle ...................................4th quartile .................................
Female
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Milligrams/deoiliier
5.3
6.14.8
5.35.4
z5.35.55.6
6.05.56.26.36.0
4.64.44.85.15.3
5.46.05.86.6
4.54.54.85.6
0.08
O.fl0.10
0.080.23
0.170.140.150.170.10
0.190.340.210.220.21
0.160.160.130.210.17
0.330.170.160.16
0.210.110.100.14
377
153224
31067
64347987
113
229
313556
4225485257
50504445
57727684
MNigrams/decUiter
5.3
6.24.6
5.25.7
5.25.15.45.45.7
6.55.86.36.16.3
4.34.54.74.85.1
5.65.96.36.6
4.34.54.65.1
0.04
0.090.06
0.050.28
0.130.130.110.130.15
0.260.200.150.200.18
0.100.140.090.110.20
0.140.150.120.22
0.110.090.130.12
915
392523
781134
161150230171183
655889
Iz
11692
1419381
116108131105
148147157156
5.4
6.24.6
5.45.3
5.35.25.45.75.6
6.16.16.26.46.1
4.54.54.64.95.1
5.76.16,16.7
4.14.35.05.2
0.08
0.050.07
0.080.16
0.110.100.080.170.13
0.100.110.100.160.18
0.130.090.100.160.19
0.100.090.100.09
0,080.090.110.13
MNiarams/deoiliter
1,023
552471
915108
259208271186119
13799
1479673
122109124
7046
130163148173
161141131116
Milligrams/deciliter
5.6
6.14.6
5.65.8
5.55.85.65.45.9
6.06,46.35.75.9
4.44.64.64.95.8
5.86.26.26.9
4.24.34.75.3
0.09
0.090.11
0.100.21
0.120.210.210.340.21
0,100.230.170.380.26
0.160.290.260.430.37
0.130.160.180.15
0.140.220.200.28
415
272143
34570
11785
1214943
7554773432
4231441511
86756167
42403839
I Excludes“other”racialgroups.
ma
-10 Table 39. Serum urate levels of adult males ages 25-74 years within systolic blood pressure level strata showing means and standard errors of means by race, age and body mass
index United States, 1971-75
Systolic blood preasure (millimeters of mercury)
Race, age, and body massLess than 114 116129 130-150 151 or more
index quartile strata Standard Number Standard Number Standard Number StandardMean
Numbererror of of Mean error of of Mean error of of Mean error of ofmean examinees mean examinees mean examirrees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 6.0 0.07 351 6.0 0.04 1,042 6.4 0.06 894 6.7 0.08 369
RaceWhite ............................................ 6.0 0.07 322 6.0 0.05 965 6.3 0.07 787 6.6Black ............................................ 6.0 0.26 29 6.1 0.17
0.0877
2926.6 0.19 107 6.9 0.18 77
Age25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ....... ........................
Body mass index1st quatille ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
6.15.95.86.26.0
5.66.06.26.3
0.110.130.230.240.16
0.090.120.170.19
10962525078
1511006634
6.16.06.25.86.1
5.76.06.16.5
0.070.100.080.160.11
0.090.070.080.10
294171223178176
280297273191
6.56.46.36.36.2
5.86.16.36.9
0.150.120.110.100.14
0.100.100.120.11
178170238171137
174201238280
6.66.76.76.76.8
0.240.190.140.220.22
0.200.210.160.11
4367
1196456
507186
161
I Excludes“other” racialgroups.
Table 40. Serum urate levels of adult female ages 25-74 yeara within systolio blood pressure level strata showing means and standard errors of means by race, age and body massindex United States, 1971-75
Systolio blood pressure (nrillirnetemof mercury)
Race, age, and body mass Less tfran 108 10s121 722-750 151 or more
index quartile strata Stanlfmi Number Standard Number Standard Number Standani NumberMean m of of &lean emvr of of Mean emor of of Mean errvr of of
rrwn examr”nees mean examinees mean examinees mean examinees
Milligrams/deciliter MNiirams/deoiliier Milligrams/deciliter hlilligrams/deoiliter
Total! ........................................... 4.3 0.05 382 4.6 0.04 1,358 4.9 0.05 846 5.2 0.09 382
Racewhite .................................o.......... 4.3 0.08 358 4.6 0.05 1,221 4.9 0.08 757 5.2 0.09 280Black ............................................ 4.4 0.22 24 4.7 0.12 135 5.1 0.15 91 5.1 0.19 92
Age25-34 YSSLS ................................ 4.2 0.08 189 4.4 0.06 425 4.8 0.11 169 4.9 0.29 3535-44 years................................ 4.1 0.10 79 0,07 284 4.9 0.11 170 5.0 0.21 5645-54 yf3SfS . ............................... 4.2 0.15 53 z 0.08 316 4.7 0.06 229 5.2 0,19 11455-64 yeara................................ 4.9 0.25 27 4.8 0.08 200 5.2 0.15 156 0.17 8865-74 years................................ 4.6 0.34 34 5.1 0.15 131 5.2 0.12 124 z 0.18 69
Body mass index1st quartile .................................. 4.2 0.07 161 4.2 0.05 386 4.3 0.10 158 4.8 0.18 392d quartile ................................... 4.3 0.11 116 4.5 0.08 400 4.5 0.09 190 4.7 0.21 403ct quartile ................................... 4.3 0.13 4.7 0.08 333 4.9 0.08 230 5.1 0.22 954th quartile ................................. 4.9 0.31 z 5.2 0.08 236 5.5 0.08 270 5.4 0.10 207
1Excludes“other”racialgroups.
-JN Table 41. Serum urate levels of adult males ages 25-74 years within diastolic blood pressure level strata showing means and standard errors of means by race, age and body mass
index: United States, 1971-75
Diastolic blood pressure (millimeters of mercurv). .
Race, age, and body massLess than 72 72-83 84-98 97 or more
index quartile strata Standard Number Standard Number Standard NumberMean error of
Standard Numberof Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ....... .................................... 6.0 0.08 402 6.2 0.05 1,073 6.4 0.06 805 6.4 0.08 376
RaceWhite ............................................ 6.0 0.08 373 6.1 0.06 976 6.3 0.06Black ............................................ 6.0 0.23
71229 6.4
6.4 0.060.22 97
3056.5 0.17 93 6.9 0.30 71
Age25-34 years ................................ 6.2 0.13 137 6.1 0.08 324 6.4 0.1435-44 years ................................ 5.8 0.13
150 6.692 6.2
0.40 130.09 219 6.4 0.15 132 6.3 0.39
45-54 years ................................ 6.0 0.16 80 6.227
0.10 264 6.4 0.1055-64 years .................. .............. 5.5 0.27
20550 6.0
6.70.16
0.17 63155 6.4 0.11 171 6.3 0.16
65-74 years ................................ 5.9 0.17 43 6.3 0.15 111 6.1107
0.11 147 6.3 0.15 146
Body mass index
1st quartile .................................. 5.6 0.16 ~54 5.7 O.tjg 284 5.7 0.14 146 5.8 0.222d quartile ................................... 6.0 0.08 119 6.0 0.07
72279 0.10
3d quartile ................................... 6.0 0.13192
62 6.26.0
0.090.14
?379
275 0.09 212 6.64th quartile ................................. 6.3 0.23 50 6.8 0.12
0.18 94235 6.8 0.10 251 6.9 0.14 130
1Excludes“other”racialgroups.
Table 42. Serum urate levels of adult females ages 25-74 years within diastolic blood pressure level strata showing means and standard errors of means by race, age and bodymass index United States, 1971-75
Diastolic blood pressure (millimeters of mercury)
Race, age, and body mase Less than 70 70-79 80-91 92 or more
index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter Milligrams/deciliter
Totall ........................................... 4.4 0.05 447 4.6 0.04 1,064 4.8 0.05 1,076 5.1 0.07 379
RaceWhite ............................................ 4.4 0.05 407 4.6 0.04 962 4.9 0.05 969 5.1 0.07 288Black ............................................ 4.7 0.26 40 4.7 0.12 102 4.7 0.14 109 5.5 0.18 91
Age25-34 years ................................ 4.3 0.08 219 4.5 0.06 415 4.5 0.10 173 5.2 0.36 1135-44 years ................................ 4.3 0.07 116 4.5 0.08 269 4.7 0.11 178 5.2 0.36 2645-54 years ................................ 4.5 0.11 78 4.5 0.08 240 4.9 0.07 298 5.0 0.14 9655-64 years ................................ 4.8 0.20 24 4.9 0.13 95 5.0 0.11 249 5.1 0.12 10365-74 years ................................ 5.6 0.66 10 5.0 0.17 45 5.1 0.12 180 5.3 0.16 143
Body mass index
1st quartile .................................. 4.1 0.06 194 4.3 0.06 315 4.3 0.10 193 4.5 0.19 422d quartile ................................... 4.4 0.09 135 4.4 0.09 319 4.5 0.08 241 4.6 0.16 513d quartile ................................... 4.6 0.14 82 4.6 0.11 232 4.9 0.08 316 5.0 0.15 1054th quartile ................................. 5.4 0.31 34 5.2 0.09 198 5.4 0.09 328 5.5 0.10 180
j Excludes“’other”racialgroups.
$ Table 43. Serum urate levels of adults ages 25-74 years within SGOT strata showing means and standard errors of means by sex, race, age and body mass index United States,1971-75
SGOT level (units/milliliterj
Sex, race, age, and bodvLess than 16.2 16.2-21.7 21.6-29.6 29.7or more
ma& index &artile stra~a Standard Number Standard Numbsr Standard Number StandardMean error of of Mean
Numbererror of of Mean error of of Mean error of of
mean examinees mean examinees mean examinees mean examinees
Total’ ...........................................
Sex
Male .............................................Female .........................................
Race
White ............................................Black ............................................
Age—Total
25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Male
25-34 years ................................35-4 j’ea”-, . ................................45-54 years ................................55-64 years ................................65-74 years ................................
Female25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Body mass index
Male
1st quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
1st quertile ..................................2d quartile ...................................3d quartile ...................................4th auartile .................................
Milligrams/deciliter
4.8
5.64.4
4.84.8
4.64.74.74.95.7
5.94.35.84.45.9
4.35.44.66.45.3
5.56.05.76.1
4.14.44.55.0
0.05
0.100.04
0.050.17
0.090.100.1 i0.140.21
0.150.060.250.080.15
0.110.250.170.220.27
0.150.160.220.18
0.080.070.120.15
778
199579
68098
252183153
9991
46~~fj
37146
39
11434654348
58534642
146175136122
Milligrams/deciliter
5.3
6.14.7
5.35.2
5.15.25.25.55.7
6.0~.~
6.24.56.0
4.76.14.96.25.3
5.66.06.36.5
4.24.54.85.3
0.04
0.060.04
0.040.16
0.060.080.070.100.10
0.06(l h.U.(JC)0.150.090.11
0.080.150.110.120.12
0.120.080.130.10
0.060.080.070.09
1,863
7191,144
1,693170
553
%301249
191362124240137
259143156124125
197199179143
306275292271
5.7
6.34.9
5.65.9
5.75.65.75.75.6
6.3..-.
E4.76.3
4.96.35.16.15.0
5.86.26.26.8
4.34.64.95.6
0.04
0.050.06
0.040.15
0.080.100.070.100.09
0.09---U.lz0.110.130.09
0.090.140.090.110.13
0.070.080.100.09
0.080.110.130.10
1,823
997626
1,634189
379304475346317
214---1rm174130254
221179169176141
244262233256
197209203214
6.2
6.65.3
6.26.5
6.46.16.36.26.1
6.74.96.45.26.7
5.46.65.76.45.5
6.16.16.77.1
4.94.75.65.7
0.07
0.080.09
0.080.22
0.170.130.120.170.17
0.170.290.140.150.12
0.180.180.250.230.20
0.150.140.100.16
0.270.210.160.16
Milligrams/deciliter Milligrams/deciliter
792
557235
675117
174143216145114
13341
10241
151
6598477341
119110140188
46466079
I ~xclutjes “other” raciai groups.
Table 44. Serum urate levels of adults ages 25-74 years within serum calcium strata showing means and standard errors of means by sex, race, age and body mass index UnitedStates, 1971-75
Serum calcium level (millr&ams/deciliter)
Sex race, age, and body Less than 9.30 9.30-9.60 9.61-10.099 10.100 or more
mass index quartile strata Standard Number Standard Number Standard Number Standard NumberMean error of of Mean error of of Mean error of of Mean
examinaeserror of of
mean mean examinaes mean examinees mean examinees
Milligrams/deciliter
Totall ...........................................
sexMale .............................................Female .........................................
Race
White ............................................Black ............................................
Age
T~l
25-34 years ................................35-44 years ................................45-54 years ................................5S64 years.., .............................65-74 years ................................
Male
25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Female
25-34 years ................................35-44 years ................................45-54 years ................................55-64 years ................................65-74 years ................................
Body mass index
Male
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
Female
Ist quartile ..................................2d quartile ...................................3d quartile ...................................4th quartile .................................
5.1
6.14.5
5.15.4
4.64.85.25.75.4
6.04.26.14.46.0
4.56.25.15.95.0
5.65.96.06.5
4.04.34.55.2
0.06
0.090.07
0.060.16
0.150.110.120.140.15
0.160.130.190.120.17
0.130.210.150.160.16
0.200.150.180.18
0.090.110.130.14
609
323486
72980
176160182139150
3714143
11785
9776638268
86
:;74
111151106118
Milligrams/deciliter
5.3
6.24.6
5.35.2
5.25.35.35.45.7
6.24.56.14.66.1
4.66.14.86.35.1
5.76.16.26.6
4.24.54.65.2
0.04
0.060.05
0.050.14
0.090.090.080.080.10
0.110.080.130.100.10
0.080.100.120.120.11
0.060.060.100.11
0.090.080.090.10
1,779
797962
1,593166
428333417326275
152276130203188
169157158117
201163206206
244237255246
Milligrams/deciliter
5.5
6.24.6
5.55.7
5.45.55.65.65.7
6.24.56.14.68.5
4.76.05.16.35.2
5.76.16.26.9
4,34.54.95.5
0.05
0.060.05
0.050.16
0.090.090.100.120.10
0.120.070.110.100.09
0.090.170.130.170.12
0.110.110,100.12
0.070.090.100.11
1,775
889
1,576199
474332451281237
222252174158216
233144137111126
195246204221
230225225223
5.8
6.45.0
5.66.1
5.95.75.95.66.0
6.44.76.54.86.5
5.26.45.16.35.7
5.98,16.77.0
4.44.65.45.7
0.06
0.070.07
0.060.19
0.090.120.100.160.16
0.100.110.140.180.15
0.150.270.150.170.27
0.170.090.150.15
0.100.110.200.11
864
482382
736128
26116519614597
17269927394
10267765740
127126112116
1098785
101
I Excludes“other”racialgroups.
+ol
Appendixes
Contents
1.
II.
Statistical notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78Survey design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78Nonresponse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80Missing data and imputation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Design considerations for examined persons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Analytical strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83Continuous variables: Means . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 84Subgroup comparisons: Means . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...!.. 84
Continuous variables: Multiple regression models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 85Assumptions of themultiple regression moclel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 85Empirical results for regression models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . !0. 87
Definitions of selected terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. !,. 91Demographic and socioeconomic terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91Geographic region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., . ...<!. 91
Family income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ,. 91Population density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...’. 91
Statistical terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92Regression coefficient (B) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ,. 92Sigma (/3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Standardized coefficient (Beta) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Sigma (Beta) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Partial r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92f-statistic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Dietary terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Caloric or total energy intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Ethanol (alcohol) consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Salt andsalty food intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92Sodium intake, combined . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Fatandcomplex carbohydrate intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Coffee and tea consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Linoleic fatty acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Oleic (unsaturated fat) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . !. 93
Dietary cholesterol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Medical and biochemical terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Hypertensive status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Hemoglobin concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Serum cholesterol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Serum urate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Serum glutamic oxalacetic transaminase (SGOT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Serum calcium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94Serum inorganic phosphate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94Serum magnesium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ,. ...0.. 94
List of appendix tables
1.Il.Ill.Ivm
v.
w.
WI.
Vlll.
lx.
x.
xl.
X11.
X111.
NHANES I population estimates for examination locations 1-65, by sex, race, and age at examination, . . . . . . .Sampling rates by age-sex groups for the NHANES I general sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Subsampling rates by age-sex groups for the NHANES I detailed sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Percent distribution of nonresponse adjustment factors: National Health and Nutrition Examination Survey,1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Number of primary sampling units (PSU’S) and number of examined persons for the general, detailed, andaugmentation surveys, by stratum number for the NHANES I design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Comparative analyses of standard errors and design effects for multiple and paired sampling error computingunits (SECU’S) within certainty strata for systolic blood pressure and calories, by age for NHANES I data,1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Number of examined persons, estimated means, standard deviations, standard errors of the means, and designeffects for systolic blood pressure, calories, and age, under analysis options 1-3 for NHANES I data1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Number of examined persons, estimates means, standard deviations, standard errors of the means, and designeffects for systolic blood pressure and calories within age groups, under analysis options 1-3 for NHANES Idata: 1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Number of examined persons in subclasses determined by lowest 15th percentile and highest 15th percentileof skinfold thickness, means, standards errors, test statistics, and design effects for systolic blood pressure:NHANES 1,1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Summary of simple regression models for systolic blood pressure and calories on age under analysis options1-3, by race and sex for NHANES I data, 1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Number of examined persons ages 1-74 years, by race, sex, and stratum number in the NHANES I design,1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Number of examined persons ages 25-74 years, by race, sex, and stratum number in the NHANES I design forthe detailed sample, 1971-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Summary of multiple regression models for systolic blood pressure on age, race, sex, and Quetelet’s index for13,573 examined persons ages 18-74 years, under analysis options 1-3 NHANES 1,1971 -74 . . . . . . . . . . . . . . . .
788080
81
82
84
85
86
87
88
89
90
90
77
Appendix 1. Statisticalnotes
Survey design
The sample design for the first National Healthand Nutrition Examination Survey (NHANES I) isbasically a three-stage, stratified probability sample ofloose clusters of persons in land-based segments. Thesample was designed to be representative of the civiliannoninstitutionalized population within designated ageranges in the coterminous United States, excludingpersons residing on lands set aside for the use ofAmerican Indians. Successive elements dealt with inthe process of sampling were the primary samplingunit (PSU), census enumeration district (ED), segment(a cluster of households), household, eligible person,and finally sample person.
For the period April 1971-June 1974, the “designprovided for selection of a representative sample of thetarget population 1–74 years of age to be given thenutrition-related health interview and examination. Asubsample of adults 25-74 years of age would alsoreceive a more detailed examination focused on otheraspects of health and health care needs. To increasethe size for this subsampling and consequently the
usefulness of the data obtained, the design furtherprovided for the selection of an additional nationallyrepresentative sample of adults 25–74 years of ageb)etween July 1974 and September 1975, to be giventhe more detailed examination. This extension ofNHANES I is referred to as the “augmentationsurvey.”
The estimated civilian noninstitutionalized U.S.population ages 1-74 years is shown in table I by sex,race, and age at the time of examination. The estimatesclosely approximate the U.S. population as estimatedblythe U. S. Bureau of the Census as of the midpoint ofthe survey sample design. The figures in table I maydiffer slightly from the census estimates because thelatter are based on the ages of sample persons at thetime they were examined, whereas the poststrati-ii.cation was based on the ages at interview. Becausecertain analyses must be done on the basis of age atexamination, the population estimates have also beenbased on age at examination for the sake of corlsisten-Cy.
The starting points in the first stage of this designwere the 1960 decennial census lists of addresses and-—
Table 1. NHANES I population estimates for examination locations 1-65, by sex, race, and age at examination
Estimated population
Age at examination Male FemaleTotal
All races White Black All races White Black
Total ..........................................
1 year .......................................2-3 years .................................4-5 years .................................5-7 years .....................+...........8-9 years .................................10-11 years .............................12-14 years .............................15-17 years .............................18-19 years .............................20-24 years .............................25-34 years .............................35-44 years .............................45-54 yeare .............................55-64 years .............................65-74 years .............................
193,976,381
3,313,4586,963,1626,672,3467,193,6637,696,5978,465,793
12,335,32112,318,434
7,352,20017,325,03826,936,00122,268,47723,313,31619,049,00112,773,574
94,239,666
1,693,0743,553,7653,378,5033,652,3223,680,3964,381,7306,312,5916,312,5193,673,3218)1 09,775
13,002,51410,675,73111,150,1109,072,5865,496,351
82,740,899
1,401 ,5C182,997,1 C172,866,3743,060,8833,279,6493,732,5935,397,0615,311,5963,206,4677,094,036
11,594,1159,515,530
10,039,1248,274,9484,969,903
10,413,986
280,212479,362485,672573,867586,419563,823879,377812,321404,045866,201
1,231,7931,004,9531,056,837
702,647486,257
99,736,515
1,620,3843,409,3973,293,8433,541,3413,816,2014,064,0636,022,8026,111,2653,678,8799,215,263
13,933,46711,592,74612,163,206
9,976,4157.277,223
86,867,546
1,327,6572,872,5812,755,0162,951,9273,257,9363,424,0705,122,1895,233,0913,158,9307,972,486
12,160,57810,111,45810,879,167
9,037,1576,603,303
11,999,935
257,289505,442511,134576,578
539,6556’17,793836,252853,294504,417
1,073,3581,646,3371,3’18,0501,237,459
8?1 ,098651,579
_—
78
the nearly 1,900 primary sampling units (PSU’S) intowhich the entire United ‘States”wa; divided. Each PSUis either a standard metropolitan statistical area(SMSA), a county, or two or three contiguous coun-ties. The PSU’S were grouped into 357 strata, as theywere for use in the National Health Interview Surveyduring 1963-72, and subsequently collapsed into 40superstrata for use in NHANES I.
During the April 1971–June 1974 period, 15 of the40 superstrata that contained a single large metropoli-tan area of more than 2 million population werechosen in the sample with certainty. The remaining 25noncertainty strata were classified into 4 broad geo-graphic regions of approximately equal population(when the large metropolitan areas selected withcertainty were included) and cross-classified into 4broad population density groups in each region. Thena modified Goodman-Kish controlled-selection tech-nique was used to select 2 PSU’S from each of the 25noncertainty superstrata, with the probability of selec-tion of a PSU proportionate to its 1960 population, andso that proportionate representation of specified Stategroups and rate of population change classes weremaintained in the sample. In this manner a total first-stage sample of 65 PSU’S was selected. These 65sample PSU’S are the areas within which a clustersample of persons was selected for examination at theparticular examination location designated within eacharea. The mobile examination units were moved fromone location to the next during this 39-month period(1971-74) to permit administering those single-timeexaminations to the cross-sectional sample of thetarget population.
Although the 1970 census data were used as theframe for selecting the sample within the PSU whenthey became available, the calendar of operationsrequired that the 1960 census data be used for the first44 locations in the sample. The 1970 census data werethen used for the final 21 stands of the sample and forthe augmentation survey.
Beginning with the use of the 1970 census data, thesegment size was changed from an expected 6 housingunits selected from compact clusters of 18 housingunits to an expected compact cluster of 8 housingunits. This change was implemented because of opera-tional advantages and results of research by the U.S.Bureau of the Census indicating that precision ofestimates would not be appreciably affected by such amodification. For large enumeration districts thesegments were clusters of addresses from the 1960Census Listing Books (later the corresponding booksfor 1970). For other ED’s area sampling was employedand consequently some variation in the segment sizeoccurred. To make the sample representative of thethen current population of the United States, theaddress or list segments were supplemented by asample of housing units that had been constructedsince 1960.
Within each PSU a systematic sample of segmentswas selected. The enumeration districts selected for thesample were coded into one of two economic classes.The first class, identified as the “poverty stratum,” wascomposed of “current poverty areas” that had beenidentified by the Bureau of the Census in 1970 (pre-1970Census), plus other ED’s in the PSU with a meanincome of less than $3,000 in 1959 (based on 1960Census). The second economic class, the “nonpovertystratum,” included all EDs not designated as belong-ing to the “poverty stratum.” All sample segmentsclassified as being in the poverty stratum were retainedin the sample. For those sample segments in nonppver-ty stratum EDs, the selected segments were dwidedinto eight random subgroups and one of the subgroupswas chosen to remain in the NHANES I sample.Continuing research indicated that efficiency of esti-mates could be increased (sampling variance de-creased) by changing the ratio of poverty to non-poverty segments from 8:1 to 2:1. Therefore, in thelater stands (44-65) the selected segments in thenonpoverty-stratum EDs were divided into two ran-dom subgroups, and one of the subgroups was chosento remain in the sample. This procedure permitsseparate analyses, with adequate reliability of thoseclassified as being below the poverty level and thoseclassified as being above the poverty level.
After identifying the sample segments, a list of allcurrent addresses within the segment boundaries wasmade, and the households were interviewed to deter-mine the age and sex of each household member, aswell as other demographic and socioeconomic infor-mation required for the survey. If no one was at homeafter repeated calls or if the household membersrefused to be interviewed, the interviewer tried todetermine the household composition from question-ing neighbors.
To select the persons in the sample segments to beexamined in NHANES I, all household members ages1-74 years in each segment were listed on a sampleselection worksheet, with each household in thesegment listed serially. The number of householdmembers in each of the six age-sex groups shown intable 11 were listed on the worksheet under theappropriate age-sex group column. The sample selec-tion worksheets were then put in segment numberorder, and a systematic random sample of persons ineach age-sex group was selected to be examined usingthe sampling rates displayed in table H. This samplingstrategy in the 65 stands of the general sample ofNHANES I resulted in the selection of 28,043 samplepersons 1–74 years of age, a sample that can beregarded as representative of the target populationdisplayed in table I.
A subsample of those adults 25–74 years of age inthe total or “nutrition” sample was then selected toalso receive the detailed health examination at the first65 stands of NHANES I. This “detailed” sample was
79
chosen systematically after a random start, using thesampling rates shown in table III. Consequently,adults 45–74 years of age in the first 65 PSU’S weresubsarnpled for the detailed examination at a some-what higher rate than those 2544 years of age.
During the augmentation period, July 1974 toSeptember 1975, the sample of adults 25–74 years ofage selected for examination in locations 66-100constituted a national probability sample of the targetpopulation. Also, when considered jointly with thoseselected for the NHANES I detailed examination inlocations 1–65, the entire 1OO-PSU sample is alsonationally representative of the target population atthat time.
The starting point for the selection of the augmen-tation sample was the 1970 decennial census list ofadddresses and PSU’S. The sampling methods forestablishing the sample frame were generally similar tothose used in the first 65 PSU’S. However, only 5 of the15 superstrata composed of only one very largemetropolitan area of more than 2 million populationwere drawn into the sample for locations 66-100 withcertainty. The remaining 10 of these superstrata werecollapsed into 5 groups of 2 each, only one of whichwas chosen for the augmentation survey with aprobability of selection of 0.5. When these latter 5locations are considered a part of the 1OO-PSUdesign,they are selected with certainty.
In this augmentation survey there was no econom-ic axis of stratification and no oversampling amongspecial groups. One of every two eligible personswithin sample households (using a random startamong those 25–74 years of age) was selected forparticipation in the survey.
Nonresponse
In any health examination survey, after the sampleis identified and the sample persons are requested to
Table Il. Sampling rates by age-sex groups for the NHANES Igeneral sample
Age and sexSampling
rate
1-5 years ............................................................... 1/28-19 years ............................................................. 1/420-44 years (men) ............................................... 1/420-44 years (women) .......................................... 1/245-84 years ........................................................... 1/465-74 years ........................................................... 1
Table Ill. Subsampling rates by age-sex groups for the NHANES [detailed sample
Age and sexSubsampling
rate
25-44 years (men) ............................................... 2/525-44 years (women) .......................................... 1/545-84 years ........................................................... 3/565–74 years ........................................................... 1/4
participate in the examination, the survey meets one ofits more severe problems. Usually a sizable number ofs:ample persons who are willing to complete thehousehold questionnaire and possibly some of themedical history will not participate in the examination.Individual participation is determined by many fac-tors, some of them uncontrollable. Therefore, partici-pation may be treated as a random event with aparticular probability of occurrence.
In this situation, the effect of nonparticipationwould only reduce the sample size, thereby increasingthe sampling variability of the examination findings. Inpractice, however, a potential for bias due to nonre-sponse exists if nonparticipation is not a random eventand if nonparticipants differ from participants. Be-cause of the possibility of bias, intensive efforts weremade in NHANES I to develop and implementprocedures and inducements that would reduce thenumber of nonrespondents and thereby reduce thepotential of bias due to nonresponse. These proceduresare discussed elsewhere.g
Also during the early stages of NHANES 1, whenit became apparent that the response rate for theexaminations was lower than in the preceding healthexamination surveys, a study of the effect of remunera-tion on response in NHANES I was undertaken. Thefindingsbswere considered sufficient to include remun-eration as a routine procedure in NHANES I starting~,ith the 21st and 22d examination locations.
Despite response rates at the household interviewstage of over 98 percent and these intensive efforts ofpersuasion, only 20,749 (74 percent) of the samplepersons from the first 65 stands were examined. Whenadjustments are made for differential sampling forhigh-risk groups, the response rate becomes 75.2percent. Consequently, the potential for a sizable biasdoes exist in the estimates in this publication. How-ever, from what is known about the nonrespondentsand the nature of nonresponse, the likelihood of sizableb[as is believed to be small. For instance, only a smallproportion of sample persons from the first 65 exami-nation locations gave reasons for nonparticipation thatwould lead to the belief that they would never agree toparticipate in examination surveys and that they maydiffer from examined persons with respect to theCharacteristics under examination. Only 15 percent ofmmrespondents gave the following reasons for nonpar-ticipation: personal illness, physical inability, pregnan-cy, antidoctor feelings, or a fear of finding somethingwrong. Typical among the reasons given by the othernonrespondents were the following: inability to taketime off from work, school, or household duties;suspicion or skepticism about the program; uninterest-ed in participating; and considered their privatemedical care sufficient, or they had just visited adoctor.
An analysis of the medical history data obtainedfc~r most nonexaminees as well as examinees also
80
supports the belief that the likelihood of sizable biasdue to nonresponse is small. No large differences werefound between the examined group and the nonex-amined group for the statistics compared. For exam-ple, the percent of persons examined who reportedever being told by a doctor that they had arthritis was20 percen$ the percent for high blood pressure was 18percen~ and for diabetes, 4 percent. The correspond-ing percents for nonexamined persons were arthritis,17 percen~ high blood pressure, 21 percent anddiabetes, 4 percent.
A procedure (similar to that used in previousNational Health Examination Surveys) was used inwhich the reciprocal of the probability of selection ofthe sample persons is multiplied by a factor that bringsestimates based on examined persons up to a level thatwould have been attained if all sample persons hadbeen examined. This factor is the ratio of the sum ofsample weights for all sample persons with a relativelyhomogeneous class defined by age, sex, and five incomegroups (under $3,000; $3,000-$6,999; $7,000-$9,999;$10,000-$14,999 and $15,000 or more) within eachstand, to the sum of sampling weights for all respond-ing sample persons within the same homogeneous class -for the same stand. The poststratified ratio adjustmentmakes the final sample estimates of the populationagree approximately with independent controls pre-pared by the U.S. Bureau of the Census for thenoninstitutionalized population of the United States asof November 1, 1972 (approximately midsurvey point),by race, sex, and age as shown in table I.
To the degree that homogeneous groups can bedefined that are also homogeneous with respect to thecharacteristics under study, this weighting procedurecan be effective in reducing the potential bias fromnonresponse. For the 65-stand sample of NHANES I,the percent distribution of the nonresponse adjustmentfactors used for the 325 cells (determined by the cross-classification of the 5 income groups by the 65 stands)is shown in table IV. Overall, the extent of theadjustment for nonresponse among the detailed exam-inees was 1.45 during the 1971-74 period and 1.40 inthe augmentation survey of 1974-75.
Missing data and imputationExamination surveys are subject to the loss of
information not only through failure to examine allsample persons but also from the failure to obtain andrecord all items of information for examined persons.When data are found to be missing for some of theexaminees, imputation for these values becomes neces-sary in order to minimize the effect on populationestimates.
Among the 13,671 examinees ages 18–74 years ofage in the total or nutrition sample of 1971–74, therewere 76 examinees (O.6 percent) missing the singlemeasurement of systolic or diastolic blood pressure or
Table IV. Percent dktribution of nonresponse adjustment factorsNational Health and Nutrition Examination Survey, 1971-74
Size of rronresponse Number Percentadjustment factor of cells distribution
Total (1 .00-3 .03) ............................ 325 100.0
1.00-1 .24 ......................................... 108 32.6l.2Sl.49 ......................................... 125 38.41.50-1.74 ......................................... 59 18.21.75-1 .99 ......................................... 24 7.42.O&2.49 ......................................... 9 2.82.5G2.99 ......................................... 1 0.33.00-3.03 ......................................... 1 0.3
both. Of the 6,913 exarninees ages 25-74 years in thedetailed and augmentation samples, only 28 (0.4percent) were missing measurements of either systolicor diastolic blood pressure or both in the first sittingposition. For the recumbent position, 59 (0.9 percent)were missing measurements of either systolic ordiastolic blood pressure or both, while for the secondsitting position, 64 (0.9 percent) were missing measure-ments of either or both blood pressures. In no case wasa diastolic measurement present without an accompa-nying systolic measurement.
In the statistical analysis of the blood pressurevariables reported in other Vital and Health Statisticspublications,~@ replacement values for the less than 2percent with missing systolic and diastolic bloodpressure were assigned on the basis of matchedexarninees of the same age, sex, and race, with similararm girth, weight, and height. However, to simplifythe analysis discussed in this report, examinees withsuch missing data were excluded since such exclusionwas found not to seriously alter the findings withrespect to the hypotheses being tested.
Design considerations for examinedpersons
Although the sample design for this survey isdescribed in extensive detail in the previous sectionsand in other documents,8’9the aspects of the designpertaining to data analysis considerations are discussedfurther in this section. All 20,749 examined personsages 1-74 years received a specifically designed nutri-tion-related examination. In addition, approximately a20-percent subsample (3,854 persons) of those ages25-74 years received a more detailed examinationfocused on other aspects of health and health careneeds. An additional 3,059 persons ages 25–74 yearsfrom the augmentation survey were examined toincrease the size of the sample and, hence, thereliability of the estimates from the data collectedduring this detailed survey (including the augmenta-tion portion). The data collection forms for the entire(nutrition) sample, together with the additional formsfor the detailed and augmentation sample, are con-tained elsewhere.gJO
NOTE A fist of references follows the text.
81
Although the sample design for this survey wasfairly complex, the essential feature is the selection ofprimary sampling units (PSU’S) consisting of countiesor groups of counties from each of the defined strata.In particular, the NHANES I design for the 1971–74period involved the selection with certainty of thePSU’S in the 15 large standard metropolitan statisticalareas with more than 2 million population, referred toas “certainty strata” (each PSU consists of a largenumber of enumeration districts), and the selection ofexactly 2 PSU’S from each of the remaining 25 strata.The design was modified for the 1974-75 period bycollapsing 10 of the certainty PSU’S into 5 strata of 2PSU’S each, retaining the remaining strata, and thensampling one PSU per strata. The augmentationsample thus included 10 of the certainty PSU’S fromthe original design and one additional PSU from eachof the 25 noncertainty PSU’S. The data tapes from theNational Center for Health Statistics reflect the index-ing of the certainty strata used in the augmentationsample. The number of PSU’S and the correspondingnumber of examined persons in each of these strata aresummarized in table V. Thus, for analytic purposes,this design can be characterized as having the follow-ing characteristics:
1.
2.
10 (redefined) strata with multiple selection ofPsu’s.
25 strata with paired selection of PSU’S for thegeneral and de~ailed samples and with a singlePSU for the augmentation sample.
Another important aspect of the NHANES Idesign is the need to adjust for the oversampling of thefollowing subgroups thought to be at high risk ofmalnutrition, as outlined in table II:
1. Persons with low income.
2. Preschool children.
3. Women of childbearing age.
4. Elderly persons.
Adjusted sampling weights that reflect the selectionprobabilities and poststratification adjustments werecomputed.
An additional design complication arises becauseat the fist 65 sites of the nutrition survey a subset ofthe sample persons ages 25-74 years received a moredetailed health examination. No particular oversam-pling of subgroups of the population remained in thissubsample; for example, women of childbearing agewere not oversampled as they were for the majornutrition component of NHANES I. However, someslight oversampling remained among the elderly. Thetatal number of persons given this detailed exaunina-tion is 3,854 persons ages 25-74 years, for whichseparate adjusted sampling weights were available.
Moreover, the augmentation survey (fully dis-cussed elsewhere lo) poses additional complications foranalysis. The 3,059 examined persons selected for thissurvey represent a national probability sample of thetarget population when used as a separate 35-stand aswell as when combined with the 65-stand detailedsample to form a 100-stand (PSU) national probabilitysample, in which the combined number of examinedpersons is 6,913. Ten of the PSU’S were included inboth the augmentation and initial surveys. There wasno oversampling of specific groups in either the initialdetailed sample group or the augmentation samplegroup.
Consequently, when computing estimates of ana-lytic statistics and their estimated variance-cova.riancestructure, the appropriate sampling weights need to beutilized in the weighted analyses. Thus, hypothesesinwolving variables from the initial detailed sample ofpersons ages 25–74 years, in stands 1–65 were investi-gated using the adjusted sampling weights associatedwith the detailed sample persons (sampling weight on
—Table V. Number of primary sampling units (PSU’S) and number of examined persons for the general, detailed, and augmentation surveys,
by stratum number for the NHANES I design
Number of PSU’S Number of examined persons
Stratum number—
General andAugmentation
General anddetailed detailed Detailed only Augmentation
Total ............................................................... 1,263 236 20,749 3,854 3,059
l.lo ............................................................... 1,213 211 4,514 853 7011 ..................................................................... 169 21 621 112 55
2 ..................................................................... 106 17 367 80 633 ..................................................................... 125 18 482 87 59
4 ..................................................................... 156 21 737 129 605 ..................................................................... 197 24 741 143 976 .................................................................... 83 22 250 48 827 ..................................................................... 108 23 395 71 72
8 ..................................................................... 61 21 188 42 809 ..................................................................... 89 21 304 57 64
lo ................................................................... 119 23 429 84 69
ll.35 ............................................................. 50 25 16,235 3,001 2,358
82
tape location 170-175). Analyses involving the aug-mentation detailed sample (stands 66-100) used theadjusted sampling weights for this group (tape location182-1 87). When hypotheses were investigated acrossthe combined detailed sample groups (stands 1–100),adjusted sampling weights were used for the combinedgroups (tape location 188-193). Otherwise, hypothesesinvolving variables from the entire initial sample(stands 1-65) utilized the adjusted sampling weightsfor the entire initial sample (tape location 176-181).
Analytical strategiesBecause of the complexities of the sample design,
each analysis could be performed one of three differentways depending on whether the sampling weights wereincluded and/or whether the design structure wasincorporated in the calculations. For simplicity, theseoptions are as follows:
OptionInclusion of sampling
Weights Design
................................. No; .................................
NoYes No
3 ................................. Yes Yes
Most hypotheses initially were investigated underoption 1 to minimize cost and time. Relationshipsfound to be statistically significant at this stage werethen subjected to more definitive analyses under option3 utilizing the sample weights and the survey designeffects. Consequently, the estimated covariance struc-ture for the sample estimators based on the complexi-ties of the survey design was utilized in all final modelsand inferential conclusions.
In survey research, the design effect is commonlydefined to be the ratio of the actual variance for astatistic from a complex sample to the correspondingvariance from a simple random sample. Increasingly,design effects are being used to adjust estimates andstatistics computed under simple random samplingassumptions for the effects of the complexities in thesample design on measures of precision. Given theimportance of these effects to those who design andanalyze surveys, simple but useful models have beensought for design effects. An extensive literaturereview of these design effect considerations and analyt-ical strategies for survey data from complex sampledesigns is presented by Lepkowski.cb A comprehensiveevaluation of the design effects and analytic strategiesspecifically for the NHANES I survey has beenpublished.g
All analyses under option 1 were performed quitesimply and inexpensively using standard statisticalsoftware. In this option sampling weights and designeffects were totally ignored. Thus, the data wereregarded as coming from a simple random sample with
NOTE: A list of references follows the text.
equal representation and probability of selection. Onthe other hand, analyses under option 2 incorporatedthe adjusted sampling weights in estimating the analyt-ic statistics, but simple random sampling computationswere still utilized for the variance estimates. Thesecalculations were performed within the OSIRIS IVsoftware package.c7Finally, analyses under option 3utilized both the adjusted sampling weights and thesampling design in calculating the estimated variance-covariance structure of analytic statistics. In particu-lar, the computer program &PSALMS was used forestimating ratio means and the program &REPERRwas utilized to fit regression models. Both of theseroutines are available within the OSIRIS IV library,and are described in more detail by Vinter.bgBriefly,for relatively simple statistics, such as ratio means,differences of such ratios, and totals, the &PSALMSroutine approximates the complex sample variance ofthese estimators using a linearized Taylor Seriesexpansion. For more complex statistics, such asregression coefficients, several replicated variance esti-mation procedures are available. In particular, thebalanced repeated replication (BRR) option within the&REPERR routine was utilized to fit multiple regres-sion models.
The estimation procedure to implement option 3can be extremely time consuming and expensive,particularly in fitting regression models by the bal-anced half-sample approach, because of the multiplesampling error computing units within the certaintystrata 1-10. To alleviate some of these ditliculties, themultiple sampling error computing unit identificationcodes were randomly allocated into 2 “pseudorepli-cates” for each of these 10 strata. Consequently, thepaired selection computations then could be utilizedfor all 35 strata. The effects of randomly assigning themultiple sampling error computing units to two pairedpseudoreplicates was investigated by the comparativeanalysis of standard errors and design effects forsystolic blood pressure and calories within the selectedage groups shown in table VI. The means and standarderrors were computed both under the multiple sam-pling error computing unit classification as well asunder the paired sampling error computing unitgroupings. At least for these variables, it is apparentthat the random allocation of sampling error comput-ing units in the certainty strata to form a completepaired design has not substantially altered the esti-mates of variances or the corresponding design effects.
As a result of this pairing for the 10 certaintystrata, all variance-covariance computations could beobtained directly as appropriate sums of squares andcross-products of differences across the 35 strata, andthus, 70 sampling error computing units. Thus, all theanalyses under option 3 for the data from the generaland detailed surveys were performed assuming thispaired selection design. On the other hand, theanalyses under option 3 for the combined data from
83
the detailed and augmentation surveys requirecl themultiple selection model because the design could notbe paired for the 25 noncertainty strata.
Continuous variable: MeansThe relative effects of the sampling weights and the
sampling design are displayed in table VII for threevariables of primary interest in these analyses, namely,systolic blood pressure, calories and age. Note that forthe total sample, the unweighed and weighted analy-ses (options 1 and 2) for these variables are quitesimilar, both for the means and variances. However,under option 3, the complex sample design introducesa considerable increase in the estimated variance of themean. In particular, the ratio of the standard error ofthe mean under option 3 to that obtained under option1 in the last column in table VII ranges from 1,498 to2,937. Consequently, the design effects for these threevariables range from 2.24 to 8.63.
In view of the fact that age was a crucial variablein the oversampling aspects of the 1971–74 design, onemight expect the design effects to be less importantwhen stratifying by age. To investigate this possibility,means and standard deviations of these same variableswere computed within age groups as shown in tableVIII. Even though the design effects are somewhatreduced, they are certainly not negligible, ranging from1.39 to 4.99.
Subgroup comparisons: Means
Most of the hypotheses tested in this report involvetlhe comparison of two subgroup means. Because of theclustered design and the sampling weights, thedifference between the mean response for each sub-group was computed as the difference between twoweighted ratio means within the context of the&PSALMS routine described in Vinterbgand otherbasic sampling texts.
In order to assess the effects of the samplingweights and the complex sample design on the magni-tude of the t-statistics associated with the tests forthese differences, a representative analysis was investi-gated in detail under options 1-3. In particular, themean systolic blood pressure was compared for twosubclasses determined by the lowest 15th percentile andhighest 15th percentile of skinfold thickness in selectedage-race subgroups. These results are displayed intable IX under each of the three analysis options. In allsubgroups, the simple random sample estimates ]fortheunweighed and weighted analyses are quite similar,bl~th for the means and variances. However, underoption 3, the complex sample design introduces aconsiderable increase in the estimated variance of thedifference in the means between the two subclasses.Specifically, the ratio of the standard error of the
NOTE: A list of references follows the text.
Table V1. Comparative analyses of standard errors and design effects for multiple and paired sampling error computing units (SEOJ’s)within certainty strata for systolic blood pressure and calories, by age for NHANES I data, 1971-74
Number ofMultiple SECU’S Paired SEW’S
Age examined Mean Standard Square root Standard Square Z
persons error of of design error of of designmean effectl mean effecfl
Systolic blood pressureTotal: 6-74 years ..................................... 17,658 123.95 0.424 2.292 0.409 2.21
6-17 years ................................................. 4,085 108.24 0.492 2.207 0.498 2.23
18-24 years ............................................... 2,290 118.89 0.466 1.573 0.441 1.4$25-34 years ............................................... 2,675 120.93 0.445 1.534 0.440 1.513544 years ............................................... 2,317 125.64 0.580 1.479 0.603 1.5345-54 years ............................................... 1,589 134.14 1.015 1.746 1.037 1.7855-64 years ............................................... 1,255 142.11 0.826 1.214 0.804 1.1865-74 years ............................................... 3,447 150.01 0.793 1.820 0.784 1.79
Calories
Total: 1-74 years ..................................... 20,749 2,000.0 17.80 2.923 17.88 2.93
1-17 years ................................................. 7,104 2,011.0 20.75 2.106 20.03 2.0318-24 years ............................................... 2,297 2,294.8 37.02 1.660 35.32 1.5025-34 years ............................................... 2,694 2,177.5 27.66 1,479 29.44 1.5735-44 years ............................................... 2,327 2,042.9 28.33 1.545 28.94 1.5745-54 years ............................................... 1,599 1,897.3 31.76 1.515 30.41 1.4555-64 years ............................................... 1,262 1,723.2 33.06 1.418 33.45 1.4365-74 years ............................................... 3,466 1,518.9 20.68 1.870 19.99 1.801
I Ratio of standard error of mean from SECU’S to standard error of mean from simple random sampling,
84
Table V1l. Number of examined persons, estimated means, standard deviations, standard errors of the means, and design effects forsystolic blood pressure, calories, and age, under analysis options 1-3 for NHANES I data 1971-74
Inclusion of sampling Number ofOption number Standard Standard Square root
Weights Oesignexamined Mean devktion error of of designoeraons mean effect
Systolic blood pressure1 .................................................. No No 17,658 126.91 24.585 0.185 . . .2 .................................................. Yes No 17,658 123.95 Z.262 0.1683
---.................................................. Yes Yes 17,658 123.95 54.347 0.409 2.211
Calories1 .................................................. No No 20,749 1,827.5 877.00 6.088 ---
.................................................. Yes No 20,749 2,000.0 944.91 6.560; ..................................................
---Yes Yes 20,749 2,000.0 2,575.9 17.883 2.937
Age
1 .................................................. No No 20,749 32.23 22.972 0.1592
---.................................................. Yes No 20,749 30.61 20.120 0.140
3---
.................................................. Yes Yes 20,749 30.61 34.417 0.239 1.498
difference of the mean under option 3 to that obtainedunder option 1 in the last column in table IX rangesfrom 1.1 to 2.0. Thus, the design effects for theset-statistics range from 1.2 to 4.0.
Continuous variables: Multipleregression models
One of the statistical models used for the analysesin this report is the following multiple regression
model:
Yi = BI + BiX2i + BZ~i + . . . + Bdki+ Ei
where Y idenotes the ithe observation of the dependentvariablq Xi denotes the ith observation of each inde-pendent or explanatory variablq and Ei is the randomvariation of the ith observation of Y. The subscripts 1,2. . . , k identify the specific explanatory variables. BI isthe mean of Y~each of the explanatory variables isequal to zercy and Bk is the change in the expectedvalue of Ykcorresponding to a unit change in the kthexplanatory variable, holding all other explanatoryvariables constant. BZ, 13s, . . . . are often referred to asthe regression slopes or (partial) regression coefficients.
Also presented in the regression results tables arebeta coefficients. The beta coefficients are the result oflinear regression in which each variable is “normal-ized” by subtracting its mean and dividing by itsestimated standard deviation. In other words, the betacoefficient adjusts the estimated slope parameter by theratio of the standard deviation of the independentvariable to the standard deviation of the dependentvariable. A beta coefficient of 0.3 maybe interpreted tomean that a standard deviation change of 1.0 in theindependent variable will lead to a 0.3 standard
deviation change in the dependent variable. Betacoefficients are also used to make statements about therelative importance of the X variables in the model.
Assumptions of the multipleregression model
The classical assumptions associated with theregression model are:
1.
2.
3.
4.
5.
The model specification is correct.
The X’s are nonstochastic. In addition, no exactlinear relationship exists among two or more ofthe independent variables.
The random variation has zero expected valueand constant variance for all observations.
Random variations corresponding to dilTerentobservations are uncorrelated.
The random variation term is normally distribut-ed.
Any set of real data is unlikely to meet all theseassumptions, particularly one utilizing complex sampledesign such as that in the NHANES I survey.However, certain violations of these assumptions maynot seriously tiect statistical inferences. For example,under simple random sampling theory, it is straightfor-ward to show that the least squares estimators of theregression coefficients retain their desirable asymptoticproperties (unbiased, consistent, and efficient), provid-ed that the explanatory variables are each distributedindependently of the true errors in the model. See forexample, Kmenta.@ More detailed discussions of theproperties of the regression model estimates fromcomplex sample surveys can be found in Holt et al.70
NOTE A list of references follows the text.
85
Table WI. Number of examined persons, estimated means, standard deviations, standard errors of the means, and design effects for systolic blood pressure and calories within agegroups, under analysis options 1-3 for NHANES I data 1971-74
option 1Number of
Option 2 Option 3
Age examined Sfandard StandardSquare root
Standard StandardStandard
persons Mean error of MeanStandard
denation error of Meanof design
meandeviation deviation
error ofmean
effectmean
Systolic blood pressure
Totak 6-74 years .............. 17,658 126.91 24.585 0.1850 123.95 22.262 0.1675 123.95 54.347 0.40906-17 years .......................... 4,005 108.67
2.21114.245 0.2229 108.24 14.132 0.2211 108.24
16-24 years ........................ 2,29031.829 0.4980
117.96 14.166 0.29602.234
118.89 13.794 0.288325-34 years ........................ 2,675
118.89 21.089 0.4407119.90 15.006 0.2901 120.93 14.710
1.4890.2844 120.93 22.739 0.4397
35-44 years ........................ 2,317 125.76 18.885 0.3923 125.64 17.6651.515
0.367045-54 years ........................ 1,509 135.10
125.64 29.008 0.602623.176 0.5814
1.536134.14 22.782 0.5715
55-64 years ........................ 1,255 143.13134.14 41.317
24.1261.0365 1.783
0.6810 142.11 23.453 0.662065-74 years ........................ 3,447
142.11151.02
28.482 0.8040 1.18125.580 0.4357 150.01 25.056 0.4268 150.01 46.027 0.7840 1.799
CaloriesTotal: 1-74 years .............. 20,749 1,827.5 877.00 6.088 2,000.0 944.91 6.560 2,000.0 2,575.9 17.8831-17 years .......................... 7,104
2.9371,880.4 830.42 9.852 2,011.0 874.24 10.372
16-24 years ........................ 2,2972,011.0 1,688.5
2,084.6 1,068.7020.033 2.033
22.298 2,294.0 1,136.60 23.715 2,294.825-34 year . ....................... 2,694 1,954.5
1,692.6 35.317971.00 18.700
1.5842,177.5 1,050.1 20.232
35-44 years ........................ 2,3272,177.5 1,527.8
1,829.0 884.6529.435 1.573
18.339 2,042.9 986.51 20.036 2,042.945-54 years .......... ............. 1,599 1,040.4
1,395.8 28.935838.33 20.965
1.5781,897.3 816.17 20.411 1,897.3 1,216.0 30.410
55-64 years ........................ i ,262 i ,679.2 828.08 233101.451
i ,723.2 814.0265-74 years ........................
22.9i 43,466
i ,723.2 1,188.51,497.2 651.06 11.059
33.454 1.4351,518.9 649.50 11.032 1,518.9 1,176.9 19.991 1.808
Table IX. Number of examined persons in subclasses determined by lowest 15th percentile and highest 15th percentile of skinfoldthickness, means, standard errors, test statistics, and design effects for systolic blood pressure: NHANES 1, 1971-74
Low $kinfold percentile High skinfold percentile
Number ofSquare root
Option Standard Number of Standard t-statisticexamined Mean examined Mean
of design
persons error persons error effect
All males...................... 1,025 130.6 0.70 1,008 141.7 0.73 11.0
: .............4......... . .
1,025 128.0 0.58 1,008 138.0 0.85 13.6 . . .3 ...................... 1,025 128.0 0.99 1,008 138.0 0.91 9.2 1.3
Black males1 ...................... 280 137.5 1.60 153 146.7 2.40 4.02
. . ....................... 280 131.6 1.32 153 140.1 2.06 3.6 . . .
3 ...................... 280 131.8 2.33 153 140.1 3.25 1.9 1.6
White males1 ...................... 745 127.9 0.73 855 140.4 0.73 12.0 . . .2 ...................... 745 124.8 0.65 855 137.8 0.69 13.53
. . ....................... 745 124.8 1.02 855 137.8 0.88 9.2 1.4
All females...................... 1,644 120.7 0.55 1,621 141.9 0.66 24.7
: ....................... . .
1,644 118.9 0.50 1,621 140.9 0.65 27.1 . . .3 ...................... 1,644 116.9 0.63 1,621 J 40.9 1.05 19.7 1.3
Black females1 ...................... 285 121.2 1.48 482 145.3 1.35 11.52
. . ....................... 285 120.3 1.52 482 146.3 1.41 12.1 . . .
3 ...................... 285 120.3 2.59 482 146.3 3.08 6.2 2.0
White females1 ...................... 1,359 120.6 0.59 1,139 140.5 0.75 21.2 . . .2 ...................... 1,359 118.7 1.16 1,139 139.7 0.74 23.73
. . ....................... 1,359 118.7 0.59 1,139 139.7 1.20 19.8 1.1
Empirical results for regressionmodels
In order to investigate predictive relationshipsamong continuous variables, multiple regression mod-els also can be fitted under either option 1, 2, or 3.Specifically, the affects of the sampling weights andcomplex design on the precision of regressioncoefficients was investigated under options 1-3 forsystolic blood pressure and calories on age as summa-rized in table X. First, it can be observed in thecorresponding entries under options 1 and 2 that theresults are quite similar, particularly for systolic bloodpressure on age, which has a significant linear relation-ship in all the race-sex subclasses. However, forcalories on age, which has extremely small E valuesfor all subgroups, the estimate of the slope is quitedifferent for some subclasses; in fact, for the “othermales” category there is a 12-fold increase in the slopeunder option 2 compared with option 1, and for the“other females” category it differ: by a factor of nearly3. Of course, in both of these subclasses the sample sizeis relatively small.
Otherwise, note ‘h table X that the results underoption 3 are only reported for the white subgroups,even though the number of black persons examined
appears to be reasonably large. This omission is due tothe failure of the balanced half-sample routine in theweighted regression program in OSIRIS IV resultingfrom entire strata with no data for these subclasses asshown in table XI. Modification of this routine or useof another sampling error program could still be usedto obtain these estimates for the other subclasses. Thisproblem of missing sampling error computing units iseven more pronounced within the more restrictivedetailed examination as displayed in table XII. Conse-quently, due to the sparse design across strata, only thewhite and black race data were used in many of theanalyses.
In addition to simple linear regression models,multiple regression models can also be fitted withinthis same framework. Table XIII summarizes theresults of systolic blood pressure regressed jointly onage, race, sex and Quetelet’s index for 13,573 casesages 18-74 years. Here again, the design effects for theregression coefficients range from 2.22 to 4.41.
These empirical results, as expressed in terms ofestimated design effects, demonstrate the critical im-portance of incorporating the sampling weights andthe survey design adjustments into all definitive sub-group comparisons and multiple regression models. .
87
mm Table X. Summary of simple regression models for systolic blood pressure and calories on age under analysis options 1-3, by race and sex for NHANES I data, 1971-74
Unweighed designNumber
Weighted design
Sex, race, and age of(Option 1) (Option 2) (Option 3) Square root
examinees/32 Slope
Standardt-statistic
/72 Slope Standardt-statistic
Standard of design
errort-statistic
error error effect
Total
6-74 years ..............................
White males ...........................Black males ............................Other males ............................
White females ........................Black females ........................Other females ........................
Total
1-74 years ..............................
White males ...........................Black males ............................Other males ............................
White females ........................Black females ........................
17,658
5,8541,326
89
8,2432,037
109
20,749
7,0041,707
109
9,3472,456
Other females ........................ 126
0.40
0.360.460.35
0.410.470.40
0.02
0.010.010.00
0.040.060.00
0.730
0.6050.8150.762
0.7670.9790.920
4.90
-3.39-3.74
1.00
-5.89-6.39-1.23
0.0060
0.01060.02400.1116
0.01020.02300.1066
0.2629
0.48730.9217
3.598
0.30340.6578
3.474
107.45
57.2433.91
6.61
75.5742.55
6.47
-16.64
-6.95-4.05
0.28
-19.41-12.75
-0.35
Systolic blood pressure on age
0.35 0.696 0.0071
0.33 0.610 0.01150.43 0.848 0.02690.14 0.401 0.1064
0.38 0.734 0.01040.44 1.008 0.02520.37 0.818 0.1040
Calories on age0.01 -5.50 0.3236
0.00 -3.52 0.61020.00 -1.08 1.2120.05 12.50 51.01
0.04 -6.44 0.33150.06 -9.45 0.74200.01 -3.35 3.899
98.11
53.1431.53
3.77
70.3940.05
7.87
-16.99
-5.70-0.89
2.45
-19.43-12.74
-0.86
0.0131 53.14
0.0113 54.06--- ------ ---
0.0166 39.03--- ------ ---
0.3171 17.35
0.6314 -5.58--- ------ ---
0.4339 -14.05--- ------ ---
1.93
1.07------
1.65------
1.21
1.30
1.43------
Table Xl. Number of examined persons ages 1-74 years, by race, sex, and stratum number in the NHANES I design, 1971-74
Stratum number TotalNumber of examined persons by race and sex
White males Black males Other males White females Black females Other females
Total ........................
1 ..............................2 ..............................3 ..............................4 ..............................5 ..............................6 ..............................7 ..............................8 ..............................9 ..............................lo ............................11 ............................12 ............................13 ............................14 ............................15 ............................16 ............................17 ............................18 ............................19 ............................20 ............................21 ............................22 ............................23 ............................24 ............................25 ............................26 ............................27 ............................28 ............................29 ............................30 ............................31 ............................32 ............................33 ............................34 ............................35 ............................
20,749
621367482737741580395188304429481517531701486563594505585446790551619449728887664
1,001634868651691619545
1,059
7,004
169146123198232
678567
109138205198232273185178235176237171344114167131225232262259222284221250222236411
1,707
882485
102653590161332
4142
152068
6391213
0107
857373
15623
17451843422
35
74
109
201
1113
2001
130022450241030000101153
2151
9,347
220157171255328
859379
149190267286290396226211346224317246446141249170311305379327292371334367345295479
2,456
13838
102162
8857
127263237
317
4144396
6:1414
0185116122119194
17241
68124523210
193
126
4209
1540
;19
221183121101230020045
121831
89
Table X1l. Number of examined persons ages 25-74 years, by race, sex, and stratum number in the NHANES I design for the detailedsample, 1971-74
Stratum numberNumber of examined persons by race and sex
TotalWhite males Black males Other males White females Black females Other females
3,854
1128087129143
4871425784
1009392129
7810110781
109811628911281158150141182126152113123119100224
1,541
3738234660171619253445404554432952414534722833286745655750634951454699
277
1341815117
18018031121314220171688
226
261014320219
2!1
10014100040000100110010000000012201
1,667
3427294355121718273053494670274154285944902348306765
u58645161695294
335
2711172312112055“4100451807210
20151514181
35811860011
13
000110000311000000000000
:100011300
Table X111. Summaty of multiple regression models for systolic blood pressure on age, race, sex, and Quetelet’s index for 13,573 examinedpersons ages 16-74 years, under analysis options 1-3 NHANES 1, 1971-74
RegressionStandard Square root
Variablecoefficier)t
error of t-statistic of designcoefficient effect
Unweighed SRS design (option 1)Age............................................................................................. 0.677 0.0096 69.44Race ..........................................................................................
---3.896 0.3938 9.89 ---
Sex ............................................................................................. -1.135 0.0335 33.88 ---
Quetelet’s index....................................................................... 1.135 0.0335 33.88 ---
Weighted SRS design (option 2)
Age............................................................................................. 0.584 0.0102 57.49 ---
Race .......................................................................................... 2.908 0.4422 6.58 ---
Sex ............................................................................................. -2.871 0.3162 -9.08Quetelet’s index.......................................................................
---1.177 0.0331 35.56 ---
Weighted complex sampling design (option 3)Age ............................................................................................. 0.584 6.0177 32.92 1.85Race .......................................................................................... 2.908 0.8266 3.52 2.10Sex ............................................................................................. -2.871 0.5206 -5.52 1.49Quetelet’s index....................................................................... 1.177 0.0630 18.69 1.88
90
Appendix IL Definitions ofselected terms
Demographic and socioeconomictermsAge
Two ages were recorded for each examinee age atlast birthday at the time of the examination and age atthe time of the census interview. The age criterion forinclusion in the sample used in this survey was definedas age at the time of the census interview. Theadjustment and weighting procedures used to producenational estimates were based on the age at interview.Data in the detailed tables and text of the report areshown by age at the time of the examination, exceptthat those few who became 75 years of age by the timeof the examination are included in the 65-74-year agegroup.
Race
Race was recorded as “white,” “Negro,” or“other.” “Other” includes Japanese, Chinese, Ameri-can Indian, Korean, Eskimo, and all races other thanwhite or black. Mexicans were included with “white”unless definitely known to be American Indian or of arace other than white. Black persons and those ofmixed black and other parentage were recorded as“Negro.” When a person of mixed racial backgroundwas uncertain about his or her race, the race of thefather was recorded.
Geographic region
The 48 contiguous States and the District ofColumbia, excluding Alaska and Hawaii, were strat-ified into four broad geographic regions, each of aboutthe same population size. With a few exceptions, thecompositions of the regions areas follows:
Region States included
Northeast . . . . . . . . . . . Maine, New Hampshire, Vermont, Massachu-setts, Connecticut Rhode Island, New York,New Jersey. Pennsylvania
Midwest . . . . . . . . . . . . . Ohio, Michigan, Indiana, Illinois, Wisconsin,Minnesota, Iowa, Missouri
South . . . . . . . . . . . . . . . . Delaware, Maryland, Virginia, West Virginia,Kentucky, Arkansas, Tennessee, North Caroli-na, South Carolina, Georgia, Florida, Alabama,Mississippi, Louisiana, District of Columbia
West . . . . . . . . . . . . . . . . . Washington, Oregon, Idaho, Montana, Wyom-ing, Colorado, Utah Nevada, California, Arizo-na, New Mexico, Texas, Oklahoma, Kansas,Nebraska, South Dakota, North Dakota
In a few instances the actual boundaries of theregions did not follow State lines. Some strata in theMidwest and South include primary sampling unitsthat are actually located in the West. Similarly, somestrata in the West contain primary sampling unitslocated in the Midwest and South.
Family income
The income recorded was the total income re-ceived by the head of the household and all otherhousehold members related to the head during the 12months prior to the interview. This income was thegross cash income (excluding pay in kind) except inthe case of a family with its own farm or business. Inthat instance net income was recorded. Also includedwas the income of a member of the Armed Forces wholived at home with the family (even though he or shewas not considered a household member). If theperson was not living at home, allotments and othermoney received by the family from him or her wereincluded in the family income figure.
Population density
The classification of urban-rural areas was thatused in the 1960 census. According to the 1960definition, those areas considered urban are: (1) placesof 2,500 inhabitants or more that are incorporated ascities, boroughs, villages, and towns (except towns inNew England, New York, and Wisconsin); (2) thedensely settled urban fringe, whether incorporated orunincorporated, of urbanized areas; (3) towns in NewEngland and townships in New Jersey and Pennsylva-nia that contain no incorporated municipalities assubdivisions and have either 2,500 inhabitants or
91
more, or a population of 2,500 to 25,000 and a densityof 1,500 persons per square mile; (4) counties in Statesother than the New England States, New Jersey, andPennsylvania that have no incorporated municipalitieswithin their boundaries and have a density of 1,500persons or more per square mile; and (5) unincorporat-ed places of 2,500 inhabitants or more that are notincluded in any urban fringe. The remaining popula-tion is classified as rural.
By means of the first digit of the identification codeon the household questionnaire, the urban and ruralpopulation was divided into the following categoriesaccording to population size: (1) urban, 3,000,000 ormore; (2) urban, 1,000,000-2,999,999; (3) urban,250,000-999,999; (4) urban, under 250,000; (5) urban,not in urbanized area, 25,000 or more; (6) urban, notin urbanized area, 10,000-24,999; (7) urban, not inurbanized area, 2,500-9,999; and (8) rural.
Statistical termsRegression coefficient (B)
The estimated additive effect on the dependentvariable for each unit of change in the independentvariable within the multiple regression model forwhich all the other independent variables are heldconstant.
Sigma (B)
The model-based estimated standard error of theregression coefficient @).
Standardized coefficient (Bet4
The estimated additive effect on the dependentvariable for each unit of change in the independentvariable which has been standardized to have meanzero and variance unity, within the multiple regressionmodel in which all the other independent variableshave been standardized and held constant.
Sigma (Beta)
The model-based estimated standard error of thestandardized coefficient (Beta).
Partial r
The estimated correlation coefficient between thedependent variable and the independent variable with-in the multiple regression model for which all the otherindependent variables are held constant.
t-statistic
The test criterion obtained as the ratio of theregression coefficient (B) to its estimated starldarderror, Sigma (B), to test the hypothesis that B is zero.
Clietary termsCaloric or total energy intake
Total caloric intake computation for food itemslisted in the 24-hour recall.
Ethanol (alcohol) consumption
For each examinee, the average number of ethanolounces per week was calculated in the following way:
(11) Assigning a factor approximating the averagealmount of ethanol in a typical serving (0.48 for beer,0.6 for wine, and 0.45 for liquor), based on the usualtype of alcohol consumed by the individual.(2) Assigning a factor to approximate the averagenumber of drinking occasions per week for eachindividual as follows:
How often do you usually drink?Every day 7.0Almost every day 5.52 to 3 times per week 2.51 to 4 times per month 0.6254 to 12 times per year 0.163Never 0.0
(3I) Multiplying the alcohol content by the weeklyfrequency and then multiplying the result by usualnumber of drinks per drinking occasion.
Thus, ethanol ounces per week = average alcoholcontent of usual alcoholic beverage consumed Xfrequency of individual drinks X number of drinksindividual usually consumes per drinking occasion.
This continuous variable was categorized intoabstainers (less than 0.0001 ethanol ounces per week),light drinkers (0.0001 to 0.9999 ethanol ounces perweek), moderate drinkers, and heavy drinkers (7.000ethanol ounces per week or more).
The second alcohol variable, calories from alcohol-ic beverages, was derived from the 24-hour dietaryrecall by summing calories from all foods coded as“alcoholic beverages food group.” This variable wascategorized into: none (less than 1 calorie),light/moderate (1 to 250 calories), and heavy (250calories or more). The two data sources agreed withrespect to alcohol abstinence for 98.7 percent of theabstainers on the medical history questionnaire whoreported no alcohol intake in the 24-hour recall.
SW and salty food intake
The frequency of use of the table salt shaker andestimated sodium content in food items listed on the24-hour recall, assuming a ratio of one gram of salt to400 milligrams of sodium from grain products, milkand milk products, mixed protein dishes, soups, meats,fruit and vegetables, fats and oils and other foods.15
ilOTE: A list of references follows the text.
92
The sodium content of food is incomplete because thevalues cover only naturally occuring sodium in foodsand sodium added by processors. Table salt used is notincluded in these data.
Sodium intake, combined
A twelve-cell table was constructed from the table-salt-use responses and the dietary sodium content offood reported on the 24-hour recall as follows:
Dietary sodium intake (24 hour recall)
Frequency ofUnder 982.6- 1,884.0- 3,436.8
sa/t shaker use982.6 1,883.9 3,436.7 milligrams
milligrams milligrams milligrams and over
Cell numberRarely .................. al a2 b3 b4Occasionally ........ ~5 b6 b7 c8Frequently ........... b9 bl O c11 d 2
● Low salt users.
bModerate salt users.
cHeavy salt users.
In the analyses, individuals in cells 1,2, and 5 wereclassified as having low salt intak~ those in cells 3, 4,6, 7, and 9 had moderate intake; and those in cells 8,11, and 12 were considered to have heavy salt intake.
Fat and complex carbohydrate intake
For the combined intake of fat and complexcarbohydrates, the distribution of the examinees bytheir frequency of intake of foods high in fats (cheese,milk, eggs, butter/margarine, and meat/poultry) andtheir frequency of intake of complex carbohydratefoods (cereals, grains, fruits, vegetables, beans, andpeas) were each divided into three groups using ascutoff points the 33d and 66th percentiles. The threelevels of each of the two variables were arrayed in athree-by-three table yielding nine cells. The twoextreme cells represented low fat/high complex carbo-hydrate and high fat/low complex carbohydrate in-take, respectively. The remaining seven cells of thethree-by-three table were pooled. The first level of thecombined fat/complex carbohydrate varibable is thehigh complex carbohydrateflow fat extreme cell,which contains approximately 5 percent of the examin-ees. The second level is composed of the sevenintermediate cells, representing 90 percent of the total.The third level, containing the remaining 5 percent ofthe total, is the extreme cell, representing the lowcomplex carbohydrate/high fat intake.
Coffee and tea consumption
Linoleic fatty acid
Estimated content in foods (from the 24-hourrecall) including fats and oils, salty snacks, fruits andvegetables, meats, desserts and sweets, grain products,poultry and other.ls
Oleic (unsaturated fat)
Estimated content in foods (from the 24-hourrecall) including meats, milk and milk products, fatsand oils, desserts and sweets, grain products, mixedprotein dishes, and others.ls
Dietary cholesterol
Estimated content in foods (from the 24-hourrecall) including eggs, meats, milk and milk products,desserts and sweets, fats and oils, and others. IS
Medical and biochemical terms
Hypertensive status
Includes the following four categories:Nonrzotensive.-Systolic pressure less than 140
mm Hg and diastolic pressure less than 90 mm Hg.Borderline.—Systolic pressure less than 160 mm
Hg or diastolic less than 95 mm Hg but not bothsystolic less than 140 mm Hg and diastolic less than 90mm Hg.
Hypertension (deJnite). —Systolic pressure greaterthan or equal to 160 mm Hg and/or diastolic pressuregreater than or equal to 95 mm Hg.
Systolic hypertension.—Systolic pressure greaterthan or equal to 160 mm Hg and diastolic pressure lessthan 90 mm Hg.
Hemoglobin concentration
As determined from the examinees’ blood sampleson the Coulter Hemoglobinometer in the mobileexamination centers. 1$’
Serum cholesterol
As determined from the exarninees’ blood samplesat the Lipid Standardization Laboratory of the Centersfor Disease Control (Atlanta, Ga.) using a modifiedferric-chloride technique.1$’
Serum urate
As determined from the examinees’ blood samplesat the Centers for Disease Control, Bureau of Labora-tories, using the Sobrinho-Simoes method.lg
Serum glutamic oxalacetic transaminase (SGOT)Derived from the food frequency dietary history. As determined from the examinees’ blood sam-
ples at the Centers for Disease Control, Bureau ofNOTE: A list of references follows the text. Laboratories, using the method of Henry et al. 19
93
Serum calcium tories, using an ada~tation of the methods of Hurst
As determined from the examinees’ blood samplesand Kraml.i9
at the Centers for Disease Control, Bureau of Labora-tories, using the method of Kessler and Wolfman.lg
Serum magnesium
Serum inorganic phosphate As determined from the examinees’ blood samplesAs determined from the examinees’ blood samples at the Centers for Disease Control, Bureau of Labora-
at the Centers for Disease Control, Bureau of Labora- tories, using the method of Hansen and Freier.l~
NOTE: A list of references follows the text.
94* U. S. COVERNMENT PR3NT3NG OFFICE : 1953 381-.161/ 109
Vital and Health Statistics series descriptions
SERIES 1.
SERIES 2.
SERIES3.
SERIES4.
SERIES 10.
SERIES 11.
SERIES 12.
SERIES 13.
Programs and Collection Procedures. -Reports describi ng
the general programs of the National Center for Health
Statistics and its offices and divisions and the data col-
lection methods used. They also include definitions and
other material necessa~ for understanding the data.
Data Evaluation and Methods Research. -Studies of new
statistical methodolo~ including experimental tests of
new survey methods, studies of vital statistics collection
methods, new analytical techniques, objective evaluations
of reliability of collected data, and contributions to sta-
tistical theory.
Analytical and Epidemiologicai Studies. -Reports pre-
senting analytical or interpretive studies based on vital
and health statistics, carrying the analysis further than the
expository types of reports in tha other series.
Documents and Committee Reports. -Final reports of
major committees concerned with vital and health sta-
tistics and documents such as recommended model vital
registration laws and revised birth and death certificates.
Data from the National Health Interview Survey .-Statis-
tics on illness, accidental injuries, disability, use of hos-
pital, medical, dental, and other services, and other
health-related topics, all based on data collected in the
continuing national household interview survey.
Date From the National Health Examination Survey and
the National Health and Nutrition Examination Survey.–
Data from direct examination, testing, and measurement
of national samples of the civilian noninstitutional ized
population provide the basis for (1) estimates of the
medically defined prevalence of specific diseases in the
United States and the distributions of the population with
respect to physical, physiological, and psychological
characteristics and (2) analysis of relationships among the
various measurements without reference to an axplicit
finite universe of persons.
Data From the Institutionalized Population Surveys. -Dis-
continued in 1975. Reports from these surveys are in-
cluded in Series 13.
Data on Health Resources Utilization. –Statistics on the
utilization of health manpower and facilities providing
SERIES 14.
SERIES 15.
SERIES 20.
SERIES 21.
SERIES 22.
SERIES 23.
long-term care, ambulatory care, hospital care, and family
planning services.
Data on Health Resources: Manpower and Facilities.–
Statistics on the numbers, geographic distribution, and
characteristics of health resources including physicians,
dentists, nurses, other health occupations, hospitals,
nursing homes, and outpatient facilities.
Data From Spatial Surveys. -Statistics on health and
health-related topics collected in special surveys that are
not a part of the continuing data systems of the National
Center for Health Statistics.
Data on Mortality.–Various statistics on mortality other
than as included in regular annual or monthly reports.
Special analyses by cause of death, age, and other demo-
graphic variables; geographic and time series analyses; and
statistics on characteristics of deaths not available from
the vital records based on sample surveys of those records.
Data on Natality, Marriage, and Divorca.–Various sta-
tistics on natality, marriage, and divorce other than as
included in regular annual or monthly reports. Special
analyses by demographic variables; geographic and time
series analyses; studies of fertility; and statistics on
characteristics of births not available from the vital
records based on sample surveys of those records.
Data From the National Mortality and Natality Surveys.–
Discontinued in 1975. Raports from these sample surveys
based on vital records are included in Series 20 and 21,
respective y.
Data From the National Survey of Family Growth.–
Statistics on fertility, family formation and dissolution,
family planning, and related maternal and infant health
topics derived from a periodic survey of a nationwide
probability sample of ever-married women 1544 years of
age.
For a list of titles of reports published in these series, write to:
Scientific and Technical Information Branch
National Center for Health Statistics
Public Health Service
Hyattsville, Md. 20782
U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICESPublic Health ServiceNational Center for Health Statistics3700 East-West Highway 20782
OFFICIAL BUSINESSPENALTY FOR PRIVATE USE, $300
POSTAGE AND FEES PAIDU.S. DEPARTMENT OF HHSr—HHS 396 A
A& .I -
DHHS Pub, No. (PHS) 83–1 677, Series 11, No 227
For listings of publications in the VITAL AND HEALTH STATISTICS series, call 301 -436 -NCHS