RD-Ri22 187 SALIVARY AND PLASMA TESTOSTERONE AND CORTISOL DURING i/iMODERATELY HERVY EXERCISE(U) NAVAL HEALTH RESEARCHCENTER SAN DIEGO CA R R VICKERS ET AL. RPR 82

UNCLASSIFIED NRVHLTHRSCHC-82-5 F/G 6/5 NLInnnnIIII ~

71 S.BUEUO1:SADll616-

10 118 EM5

L-:':

MICROCOPY RESOLUTION TEST CHARTNATIONA SEM OF STAMIOMCS-1963-A #

.%

AD A122 107

I-.m

SALIVARY AND PLAS TESIUTMER1 AND OMTISCL

mmN NDIERAMLY HEAW MOCISE

Ross R. Vickers, Jr., Ph.D. 1

James A. Hodgdon, Ph.D. 1

Brad L. Bennett, M.S. 2 "

Russel E. Poland, Ph.D. 3

Robert T. Rubin, M.D., Ph.D.3

,.'

Report ND. 82-5, supported by Naval Medical Research md Development Commnd, Bethesda, Nryland, Department of theNavy, under research Work Uhit N0O96.NO01-103S5, under Office of Naval Research Omtract N0014-81-K-0561; and byResearch Scientist evelopment hard WH47363 (to R.T.R.). The views presented in this paper are those of the authors. "No endrsement by the Ipartment of the Navy has been given or should be inferred. r. v

IBavirommental Physiology I-partinot, Naval Health Research Center, P.O. Box 85122, San Diego, CA 921382 Navy Submarine Nedical Research Laboratory, New London, CT.

3leparament of Psychiatry, Herbor-U.C.L.A. Ndical Center, Torrance, CA.

The authors thank Ceryl Spinweber and Earl A. Eklards for their helpful suggestions and comments on an earlier draftof this paper.

ALI!&I:

' ~.........

my -gg

*iWvi ~'~

W.O7 0-. 4-'00

irkV

, 0*"

to' !, pu -

:f.i~ 4

IQ

'IFt

Introduction

The study of steroid horames is of interest to exercise physiologists because these hormones influence metabolic

substrate utilization and fluid and electrolyte balance. However, changes in levels of circulating steroid hormones

during exercise have been inconsistent across studies (1-4). Resolution of these inconsistencies might be facilitated

by more frequent sampling of the hormones to accurately quantify the temporal profile of hormone changes during exer-

cise. Techniques that facilitate frequent sampling therefore my contribute to improved understanding of the physio-

logical processes involved in exercise.

Standard blood collection procedures can be difficult to employ when frequently repeated measures are needed.

Drawing blood for these measures can be troublesome due to problems arising from the need for repeated veniponctures.

Catheterization provides an alternative, but is impractical in many settings. Venipncture by either method may dis-

turb the hornmes to be measured (5).

Salivary measures provide a potentially sensitive alternative means of monitoring steroid hormones. Close paral-

lels between salivary concentrations and plasm or serum concentrations have been dmonstrated for cortisol(6-9),

testosterone (10-13), 17-hydroxyprogesterone (14-16), and aldosterone (17). Salivary measures have the advantages of

low risk, miniml stress, and ease of repetition during short time periods. However, this work has been done in clinical

settings uider resting conditions, mnd the findings my not generalize to conditions of interest to exercise physiolo-

gists. 1hs, the present study was undertaken to examine the Iipact of a 30-minute period of moderately heavy exercise

on the association between plasm and salivary measures of cortisol and testosterone.

Mbthod

* Smep-.

Fifteen male military and civilian persomel (12 white, three black) working for the U.S. Navy in San Diego,

California, participated after giving informed consent nd passing a medical screening exemination. Additional descrip-

tive data are given in Table 1. (See page 4.)

Exercise Protocol

The research design required participants to cmplete a 30-minute run at 751 of maxima oxygen consuqption capacity

CVO2 max). Participants reported to the laboratory betwen 1300 and 1330 for n max assessment, which as measured in

a single graded exercise session on a treadmill. Participants walked at 3.0 miles per hor (mph) at 0 grade for four

minutes to warm up. The speed then was increased 0.5 mph each minute until a comfortable pace between 7 and 8 mph was

reached. The grade then was increased 31 per minute until the participant voluntarily stopped running. Spirometry

values during the run were determined by an online, open-circuit spirometry system' which recorded values on a mag-etic tape and printed out 15-second averages. n 2 max was reached when V02 did not increase in the one minute aver-

age during the test. The measured N02 max indicated that participants were above average in fitness (Table 1).

1 The spirometry system cmisted of an Applied Electrochemistry 02 analyzer O4del S3-A), A Beci CO2 analyzer 4bdel

(LB-2), a I Digineering pmnmscan spirometer (Model S-300), a Yellow Springs air teqerature thermeter (odel 43T-

A) md an Nmld Ibepl one-way valve. All interfaced with a Hewlett-Packard ?4bdel 982SA programmble desk-top calcu-

lator.

3

_,,, ,/ - ,.. ..-.. , ~.. ..... ...... .-. ..- ....... .. .. ..............• x ', , , ' . . ,. , , ¢ . , . ., ,' .. . . . '. . .. .. ¢ . , .. .. - .. .. -. ' . . . _ • . , . . .. . . . .- , . .- . .,. ., .. , - . . ,' -. .. . .' . .. , .. ., . r.

TABLE 1PARTICIPANT DESCRIPTION

MEAN IIMfl3

Am (waws 26.1 4.5hIt (an) 177.2 8.5SI (I) 76.6 10.4mSut of dunfokIb 292 10L6

% VO 2 mm 4I/mini 4.48 0.8V0 2 (mi/g ) mi8.4 4.4

PHYSIOLOGICAL RESPONSES TOTHE 30-MINUTE TREADMILL RUN

Hlan ratm (bpn) 165.5 11.3Vemtlation (I/min STPS) 92. 1"V0 2 (1/n STPD) 3.40 0.50V02 (mi/kg mn STPO) 45. 2AV0 2 (% V0 2 mM) 77.5 4.2

ON - 15.

bComputed frm amesures of bkcps, tceps, ,wprIlec, end suhscaput snfold.

The day following V02 max determination, the participants returned to the laboratory at the same time of day for

- the 30-minute run. Each participant ran at a speed calculated to represent 751 of his VD2 max. Om the average, actual

V02 during the run was quite close to 751 (Table 1). This work level ws expected to produce testosterone and corti-

sol changes in moderately conditioned individuals (1,2). Prior to beginning the rim, the participants provided blood

and saliva samples (see below) and then warmed up at six mph for five minuts. They thn ran for 15 minutes, gave

another blood and saliva sample, and completed the last IS minutes of the run. At the end of the run, final blood and

saliva samples were collected.

Blood and Saliva Collection and Assay Procedures

Venous blood was dram for a superficial vein of the mtecubital fossa into a IS-al vacutainer collection tube.

Samples were centrifuged imediately awd the plams separated frm the red blood cells and fTom. Three-minute samples

I of whole saliva were collected by having the participants expectorate directly into 3S-mi containers. Salivation was

stimulated by chewing a 1.0 pa bolus of paraffin. Saples were from immdiately after collection. The time re-

quired to collect both samples was three to four minutes, as blood was damn during the saliva collection. In addi-

tion to the three samples taken during the 30-minute run, a baseline mame was taln the day before the run during

the participants' initial laboratory visit prior to the V02 max detemination.

To determine flow rate, saliva volume wms estimted as follow: The vacutainer was first weighed empty. Emty

*weight was subtracted from the combined weight of the sale and the vacutainer to determine net saple weight, uhich

then ws translated to a volume estimate by assmnig that the specific grity of the saliva was 1.000. This assimp-

tion should have introduced an error of I percent at most into the vmlume estimates (18, pp. 22-23). Such an error

is very sall cmpared to the observed differences in estimated sople volum. Plow rate was finally determined as

milliliters per minute by dividing the estimated volume by three.

4

".-]

The samples were stored frozen until assayed. All samples were measured in a single assay. Cortisol radioimno-

assay (RIA) employed the technique of Foster and Ian (19) with ntra-assay coefficients of variation of 17.61, 5.21,

and 14.0 for low, imditm, and high concentration pools, respectively. The testosterone RIA used the tedmique of

Odell, et al. (20) with intra-assay coefficients of variation of 17.2%, 8.8%, and 2.0% for low, mdium, and high con-

centration control pools, respectively. The RIAs measured total blood and total salivary hormone concentrations.

Analysis Procedures

Pearson product-ament correlations were used to describe the association between plasma and salivary concentra-

tions for cortisol and testosterone. Additional analyses ensured that these product-moment correlations accurately

summarized the data. The Kolzogroff-Smirnov test (21) indicated that the score distributions met the asstmption of

normality. Nonprmetric correlation (21) and jackknife statistics (22) showed that the parametric statistics yielded

reasonable estimates of magnitude of association and significance, even if the normality assumptions had been relaxed.

A multivariate analysis of variance for repeated measures was used to determine whether mean hormone levels changed

significantly during the 30-minute run. All analyses except the jackknife statistics were performed with the Sta-

tistical Package for the Social Sciences (SPSS) (23). The jackknife statistics were computed on a hand calculator

from correlations produced by the SPSS program.

results

Cortisol levels did not change significantly during the study, but testosterone levels did (Table 2). Examina-

tion of values for individual participants showed that the nonsignificant cortisol trends represented the effect of

averaging over people with very different response patterns. In the case of testosterone, there was a general cor-

respondence in the trends for mean plasm and man saliva hormone concentrations.

TABLE 2PLASMA AND SALIVA CONCENTRATIONS OF CORTISOL AND TESTOSTERONE:

MEAN LEVELS AND INTERCORRELATIONS

eINK PF3-1IX 15Miu MI3UNUT AVERAGE F VALUE SURICANCE

CORTISOL Plaume Muna 84.40 94.3 91.87 05.27 91.62 2.17 .145S. D. 22.72 32.44 29.77 33.81 29.5

Saliva Mine 1.89 2.31 2.12 2.47 2.20 1.84 .194S.D. .88 1.64 1.08 1.60 1.30

Corra eon .o .87"** .8"** .96"** .88"**

TESTOSTERONE Plame Muna 5.37 5.54 6.8t 6.73t 6.06 5.40 .014S. D. 1.74 1.58 2.60 2.21 2.11

SaivM Mov e .079 .085 .105 .11st .08 4.19 .030S. D. .033 .019 .030 .035 .033

Plume/SeIvaCaoruetlon .66* .24 .48* .37 .81*

All concentralon repor'td in nlmlp <.05, otte~elled

"op <.01, ot-tlled%*p < .001 one.tledtSlflcenity diff n t (p < .05) from baeine and preexerclse iles

... ... .. .. . ... ... ....... .. .. .

Paired plasma and saliva cortisol concentrations were highly correlated (r!..82, Table 2). However, plasm:salivw

testosterone correlations were consistnly lower (r_.65, Table 2).

Salivary cortisol concentrations ranged between 2.14 and 2.36 percent of the corresponding plasma concentrations.

' Corresponding values for testosterene were 1.48 to 1.79 percent. In both cases, there was a nonsignificant trend

toward lower ratios during the later sampe collections. Salivary horme concentrations were not correlated with

saliva flow rate at any point.

Discussion

}Nbderately heavy exercise did not affect the relationship between plasma and salivary cortisol. The correlations

were consistently high. Man levels of cortisol did not vary, possibly due to the only moderate exercise intensity (3).

Mether this will bold true during heavier or more prolonged exercise is yet to be determined.

The testosterone findings were equivocal. Paired saliva and plasm values were not highly correlated. Oanges

in mean testosterone levels during exercise were significant and similar in pattern for both plasm and saliva. The

low correlation between the paired samples is at odds with previous findings in resting subjects (13), but even the

baseline correlation in this study was only r-.65. One aspect of our data that might account for this discrepancy wasa relatively restricted range of testosterone values compared to those in other studies. Wang, et al. (13) reported

a high correlation (r..94) and presented a scatterplot for paired saliva and plasma testosterone values. The scatter-

plot suggests that elimination of a few extreme samples would have produced results quite similar to ours. Within a

restricted range of values, the well-dcumented individual differences in the ratio of free to total testosterone

(24-26) could have had ore influence on the saliva-plasm associations. This issue may be important for salivary

steroid research in general because similar distributional factors appear to affect some of the reported associations

for cortisol (16), 17-hydroxyprogresterone (16), and aldosterone (17). Previous reports of high salivary-plasm or

salivary-sen. correlations therefore my have depended on the inclusion of saples from persons with endocrine prob-

lem or who received drugs that altered steroid levels.

The present findings do not rule out the possibility that salivary testosterone might have been highly correlated

with the free plasm testosterone fraction in our subjects. It is cmmnly held that salivary steroids reflect free

plasma steroids more thain they do total plasm steroids (6-8,10-13). Ulforfmnately, this belief is based largely on

indirect evidence (e.g., absence of steroid binding globulins in saliva, low salivary steroid concentrations, and

similar half-lives of free plasm and salivary steroids). Katz and Shamon (9) have shown that salivary cortico-

steroids were more highly corrlated with free plasm corticosteroids (r - +. 83) than with total plasm corticosteroids

(r - +.7S), but this difference is too sml for final conclusions, even with circumstantial support from other lines

of investigation. The fact that total plasm values and salivary values can be relatively independent at times, as

in the case of our testosteome data, jmkes it critical to examine this issue in more detail. Ihfortnately, the

importance of the free vs. total blood hormne distinction was not anticipated at the outset of this project because

of the consistently high correlations reported in other research; therefore, free hormne measures are not available in

or data. Despite the expense and difficulty of obtaining both free and total hormone values, further investigation of

the differences between the two is essential. This is particularly true if salivary measurs are to be employed in

studies involving norml populations.

The fact that testosterone levels increased with exercise adds additional data to support the position that tes-

tosterom concentrations increase during exercise; previous evidence an this issue has been mixed (2-4).

The sin cmclusion from our study is that salivary cortisol cam be substituted for plasm cortisol in research

ibolving mdrately heavy exercise. The equivocal testosterone findings point to a need for direct investigation of

6

q .14

the relationship between salivary steroids and both total and free plasm steroids; this need applies to other steroid

W. hornmes as well.

11eferences

subjects. In J.H. iiMoore 6 J.F. Keogh (Eds.). Exercise and Sport Science Re-views 1975, Acadmic Press, pp. 1-30.

cise InPootmn, J 4 ise, G.(Ed.),Biohemstr ofExercise IV-B, 1981, thiversity Park Press, Baltinre,

S. RseLM.4 Hrst M.. Pasmcoriso an grwthhoraneresponses to intravenous catheterization. Journal of

Hman~~~Sao Ttes17,.,2-6

7. Sten .. Vraini aiaylvl f17-hydroxycorticosteroids as an index of adrenocortical activity andcirclatig coticsterdsTheNow ork tat Denal JXM_ 1960, 2.6 368-370.

8. WleL. idRbyD. Red .. Arnlsauasesdby direct radioisumoassay of cortisol in wholesaliva or parotid saliva. Clinical Chemistry, 1978, 24, 1460-1463.

*9. Katz, F.H. 4 Shannon, I.L. Parotid fluid cortisol. and cortisone. Journal of Clinical Investigatli 1969, 48,838-8S5.

10. Landam, A.D., Sanford, L.M. Howland, B.E., Dames, C., 4 Pritchard, E.T. Testosterone in hum saliva.Experientia 1976, 32, 940-941.

11. Luisi, M., Bernini, G.P., Dal Genovese, A., Birindelli, L., Barletta, D., Gesperi, )L, 4 Pranchi, F. Radtioimmom-assay for "free" testosterone in hmn saliva. The Journal of Steroid Biochemistry, 1980, 1 lZ S1-16.

12. Smith, R.G., Bosch, P.K., Dill, G., 4 Buttran, Jr., V.C. Saliva as a metrix for measuring free androgens: Com- S

parison with serum androgens in polycystic ovarian disease. Fertility and Sterility, 1979, 31 51-517.13. Wang, C., Plymete, S., Nieschlag, E., 6 Paulsen, A. Salivary testosterone in men: Further evidence of a direct

correlation with free serim testosterone. Journal of Clinical Bmdocrinolom, mid Nbtabolim, 1981, S3 1021-1024.

14. Walker, LF., RAW, G.F., Itigher, I.A., 6 Riad-Fabie, D. Radiolummoassay of M7-hydroxyproreterone In saliva,parotid fluid, and plasm of Congenital adrenal hyperplasia patients. Clinical Chemistry, 1979, 2S, 542-545.

15. Walker, LF., Read, G.F., 6 Riad-Raboy, D. Pladiolumoassay of perogesterone in saliva: A~plication to the assess-mat of ovarian ftnction. Clinical amimistry, 1979, 25, 2030-2033.

16. Price, D.A., Astin, M.P., Oiad, C.R., 6 Addison, G.M. Assay of hydroxyprogesterone in saliva. Lancet, 18 Aug

17. WVie, L, Levine L S, 4 New, I. The biologic sigaificance of aldosterone concentration in saliva. Pediatric

Reserch 197, liF~i-7'1

7

18. Kerr, A. C. The ftiolojicai Iuation of S~aivay Secreths in Nan. New York: Pergmn Press, 1961.

19. Poster, L.B. 4 Dn, LT. Single antibody tedmique for radioinioassay of cortisol in umextracted seru orplasma. .Clinical Chemistry, 1 74, 0 365-368.

20. Odell, .D., Swerdloff, LS., Bain, J., Vollesen, F., & Grover, P.K. The effect of sexual maturation on testicu-lar response to LH stimulation of testosterone secretion in the intact rat. Endecrinology 1974, 1580-1384.

21. Siegel, S. Nonparatric Statistics for the Behavioral Sciences. New York: McGraw-Hill, 19S6.

22. Ducan, G.T. G Layard, .W.J. A Rate-Carlo study of asymptotically robust tests for correlation coefficients.Biometrika, 1973, 60, 551-558.

23. Nie, N.H., Hull, C.H., Jenkins, J.G., Steinbrenmer, K., & Bent, D.H. Statistical Package for the Social Sciences.New York: Mc4raw-Hill Book Company, 1975.

24. Vermeulen, A., Stoica, T., & Verdnck, L. The apparent free testosterone concentration, an index of androgeni-city. Journal of Clinical Endecrinology, 1971, 33, 759-767.

25. )oll, Jr., G.W., Bsenfield, R.L., & Helke, J.H. Estradiol-testosterone binding interactions and free plasma

estradiol under physiological conditions. Journal of Clinical Bndecrinoloy and Metabolism, 1981, 52, 868-874. '1

26. Chopra, l.J. & Tulchinsky, D. Status of estrogen-androgen balance in hyperthyroid men with Graves' disease. iJournal of Clinical Endrocrinology and Wetabolis, 1974, 3 269-277.

,.

ii

C, %

8

. . .

- -- - .. -. - . . ~ n-.-. 77T - - - . . --

Unclassified.SECURITY CLASSIFtCATION OF THIS PAGE (When Data Xntered_

REPORT DOCUMENTATION PAGE 3KF suC'noPIBEFORE COMdPI.,KTG FORM1. RPORT HUMft 2. GOVT ACCMION NO. 3. RECIPIENT'S CATALOG NUMBER82-5 h 7__

4. TITLE (and Subtitle) S. TYPE OF REPORT I PERIOD COVERED

Salivary and Plasma Testosterone and Cortisol InterimDuring Moderately Heavy Exercise 6. PERFORMING ORG. REPORT NUMNER

. AUTHOR(a) S. CONTRACT OR GRANT NUMBER()"

Ross R. Vickers, Jr., James A. Hodgdon,Brad L. Bennett, Russell E. Poland, R. T. Rubin

9. PERFORMING ORGANIZATION NAME AND ADDRESS "0. PREORAM J.I, PROJECT. TASK

Naval Health Research CenterP. 0. Box 85122San Dieao. CA 92138. _ OO96PNO01-1035

II. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE

Naval Medical Research & Development Command April 1982Bethesda, MD 20014 . NMUER OF PAGES

14. MONITORING AGENCY NAME & ADDRESS(If different Item Caontrolin Office) IS. SECURITY CLASS. (o this rone)

Bureau of Medicine & Surgery UnclassifiedDepartment of the NavyWashington, D.C. 20372

i0. DISTRIBUTION STATEMENT (of this Roer,)

Approved for public release; distribution unlimited.

17. DISTRIBUTION STATEMENT (of the abstract entered iBhietk ". It differen te Reeat) .

Approved for public release; distribution unlimited.

IS. SUPPLEMENTARY NOTES

-C *

It. KEY WORDS (CmUtln. an reverse side It news' s, ImEt/ J,& OF iek amber)

Cortisol SalivaExercise TestosteronePlasma

U21IYTSRACT (Contriue a tomvee side It neeseeit a Iesafflp OF Week amb e)

Saliva may provide a useful alternative to blood for measuring steroidhormones. Total plasma and salivary concentrations of cortisol and testos-terone were compared in samples taken twice at rest and twice during exerciseto determine whether physical activity level affects the relationship between rthe two. Correlations were con'lslently high (r>.82) for cortisol, butrelatively low for t, .tostere- (rc.66). Exercise did not affect either cor-relation. Salivary c.,. L. ol s a reasonable alternative to plasma cortisol

DD ,o" 1473 2017100 OF I NOV 4111 OBSOLETEI JAN 7 S/N 0102-LF-414.4601 IISCUIITY CLASSIPICATION OF TNIS PAGE (Ohme DeAta

K ..

UnclassifiedSECURITY CLASSIFICATION OF THIS PAGE (When Data Eatered)

20. Abstract (continued)

even during exercise. The testosterone results were equivocal as salivarytestosterone could be highly correlated with free plasma testosterone despitethe low correlation to total plasma testosterone. Closer examination of thefree/total plasma hormone distinction was not possible in the present study,but should be an important focus for further research on salivary steroids.

CA

I4

IllSECuRiTY CLAIIrICATION OPr"NIIS PA6E(r.(Ia Da se

FIIE

D~IC