Sodium and Health: Evidence, Policy, Reality

Cheryl A. M. Anderson, PhD, MPH, MS University of California San Diego

Department of Family Medicine and Public Health

Email: c1anderson@ucsd.edu

Sodium Intake in Populations: Assessment of Evidence

Statement of Task The CDC asked the IOM to examine the studies on sodium intake and direct health outcomes in the general U.S. population and among individuals with hypertension; pre-hypertension; those 51 years of age and older; African Americans; and those with diabetes, chronic kidney disease, and congestive heart failure.

Source: Institute of Medicine Consensus Report, Released May 14, 2013

Outline  

Evidence  related  to  sodium  and  health  outcomes  •  Sodium:”  an  old  enemy  or  a  new  friend?”    Recent  policy  making  on  sodium  •  Individual  and  public  health      PopulaAon  sodium  reducAon    •  RealisAc?  Feasible?  Sustainable?  

Points to consider

•! Sodium has been evaluated in relation to cardiovascular, kidney, and cancer outcomes

•! Policies mostly based on evidence that reduced sodium intake lowers blood pressure

•! Lower levels of blood pressure should reduce the risk of cardiovascular disease, the leading cause of death in the US and worldwide

•! The estimated benefits of sodium reduction are substantial and warrant major public health efforts to reduce intake

Outline  

Evidence  related  to  sodium  and  health  outcomes  •  Sodium:”  an  old  enemy  or  a  new  friend?”    Recent  policy  making  on  sodium  •  Individual  and  public  health      PopulaAon  sodium  reducAon    •  RealisAc?  Feasible?  Sustainable?  

Survey approaches to Sodium Assessment:

Approach Characteristics

Behavioral Strumylaite et al, Medicina 2006

•  “Do you put additional salt on your meal?”

•  Categorized as yes or no Qualitative food pattern Sjodahl et al, Cancer  Epi  Biomarkers  Prev  2008

•  “How often do you sprinkle extra salt on your hot food?”

•  “How often do you eat salted meat?” •  Categorized as high or low based on

frequency data Semi-quantitative usual intake Peleteiro et al, Br J Cancer 2011

•  Food list, frequency and amount used to determine intake

•  Categorized as high or low

Food  Frequency  QuesAonnaire  

• Derives  typical  intake  from  quesAons  on  amount  and  frequency  

 • PotenAal  for  misreporAng  

Incomplete  list  of  foods  and  aggregaAon  Poor  porAon  size  esAmaAon    

• PotenAal  for  recall  bias    • PotenAal  for  underesAmaAon  

DiscreAonary  sodium  intake  not  assessed  

24-hour recall interview

•  Requires detailed probing •  Databases must have specific sodium

content of processed foods that contribute most to daily intake

•  Multiple recalls needed to account for day-

to-day variability

9

D%/1.-

!! !"#$%&'%()*+,-*.',!! /..01',&'*+%&",23,"$,1*45,6&$,1*7',89991+,0:(4',(4,8,;*"

9,

39,

899,

839,

A99,

A39,

J99,

!"#$%

&'(&)*

*+,-./&

0#*%&)1+2"3/&

D-'.

9,

39,

899,

839,

A99,

A39,

J99,

!"#$%

&'(&)*

*+,-./&

0#*%&)1+2"3/&

D-'.

Measurement of Na Intake Optimal §  Multiple, high quality 24 hour urine collections Suboptimal §  24 hour urine collected with limited or no

attention to quality control §  Spot, overnight or timed urines §  24 hour dietary recalls §  Food frequency questionnaire

Adverse Effects of Excess Sodium Intake

•! Established relationship –! Increased blood pressure

•! Probable relationship –!Gastric cancer

•! Suggestive relationship –! Increased risk of osteoporosis –! Increased left ventricular mass

•! Hypothesized relationship –!Overweight/obesity

! CVD and Stroke

Types of Evidence Linking Sodium Intake to Blood Pressure

Epidemiology Over 50 population studies Migration Several, e.g. Kenya Genetic All defects identified so far impair the ability

of the kidney to excrete salt. Animal All forms of hypertension are caused or

aggravated by salt No study has documented increased CVD risk from reduced sodium

Trials Children: ~10 trials, one trial in infants Adults: > 50 trials, 10 dose-response

Population Interventions

Northern Japan Finland Portuguese villages

Does modification of whole dietary

patterns affect blood pressure?

Dietary Approaches to Stop Hypertension (DASH) Trial DASH diet emphasizes: •  Fruits •  Vegetables •  Low-fat dairy products •  Whole grains •  Poultry •  Fish •  Nuts •  Reduced red meat •  Reduced sweets and sugared-beverages

D.A.S.H. Diets and Systolic BP Change

122

124

126

128

130

132

0 1 2 3 4 5 6 7&8

ControlFruits/VegsCombination

Sys

tolic

BP

(mm

Hg)

Intervention Week

Source: Appel LJ et al. N Eng J Med 1997;336:1117-1124

Control Diet

Randomization

Run-in: (11-14 days)

Intervention (Three 30-day periods, random order)

Intermediate Sodium

Higher Sodium

Lower Sodium

Higher Sodium

Intermediate Sodium

Lower Sodium

Study Design

Control Diet, N = 204

DASH Diet, N = 208 N = 412

Source: Sacks FM et al, N Engl J Med. 2001 Jan 4;344(1):3-10

D.A.S.H.-SODIUM Diets and Systolic BP Change

Estimated BP Reductions from Lowering Sodium Intake

Children1 Non-HTN2 HTN2 Resistant

HTN3

SBP (mmHg)

-1.2 -2.0 -5.0 -22.7

DBP (mmHg)

-1.3 -1.0 -2.7 -9.1

Na (mg/d) Reduction

n/a 1,700 1,800 2,300

1He,HTN 2006;48:861 2Cochrane Review, 2006; 3Pimenta, HTN 2009;54:475

Trials of Sodium Reduction in Patients with Diabetes

•  13 trials with 254 individuals – 75 individuals with type 1 diabetes – 158 individuals with type 2 diabetes

•  Duration – median: 1 week, range: 5 days to 12 wks

*Source: Suckling RJ, Cochrane Review, 2010 Dec 8;(12):CD006763

BP Reductions from Lowering Sodium Intake: Diabetes

Type 1 DM Type 2 DM

SBP (mmHg) -7.1 -6.9

DBP (mmHg) -3.1 -2.9

Median Na (mg/d) Reduction

4,700 2,900

*Source: Suckling RJ, Cochrane Review, 2010 Dec 8;(12):CD006763

Population-Based Strategy SBP Distributions

Stamler R. Hypertension 1991;17:I-16–I-20.

Hypertension

Reduction in SBP mmHg

2 3 5

% Reduction in Mortality

Before Intervention

Reduction in BP

After Intervention

Stroke CHD Total -6 -4 -3 -8 -5 -4 -14 -9 -7

Projected Effects of Dietary Salt Reduction

Source: Bibbins-Domingo, K et al, N. Engl. J Med; 2010; 362:590-599

Sodium Reduction Lowers CVD Risk: Meta-Analysis of Trials

Source: He FJ, MacGregor GA. Lancet. 2011;378:380–382

Events Na/Cntl

Trial

TOHP I 17 / 32

TOHP II 71 / 80

Morgan 6 / 5

TONE 36 / 46

Total 130 / 163

Trials of Sodium Reduction in Heart Failure

•  Multiple, randomized controlled studies by one investigative team (Paterna et al, 2009, 2011)

•  Low sodium intake (to 1840 mg/day)

increases risk for adverse events

Trials of Sodium Reduction in Heart Failure

Participants – Hospitalized NYHA Class III Heart Failure –  Intensive medication regimens

•  100% ACEI •  100% Lasix •  85% Spironolactone

Medical management

– Unconventional, leading to untreated long-term volume depletion

Source: Paterna, Am J Card 2009:103:93, Paterna, Am J Med Science 2011, 342: 27

Methodological Challenges in Observational Studies that Relate

Sodium Intake to CVD •  Random error in sodium assessment •  Systematic error in sodium assessment •  Potential for reverse causality •  Major analytic issue, e.g. under-adjustment •  Potential for residual confounding

26 studies with 31 independent samples On average, 2.5 issues/study

Source: Cobb LC et al, 2012 Circulation

Random and Systematic Error in Measuring Na Intake

§  Random error § Reason: high day-to-day variability of Na

within an individual §  Impact: bias to the null

§  Systematic error § Reason: underreporting of intake from 24 hr

recalls OR incomplete urine collection §  Impact: potential for paradoxical relationship

§  Both types of errors – extremely common

Increased CVD Mortality in Persons with Lowest Na (by quartile of Na Intake in mg/d)

1.8 1.94

1.48

1

0

1

2

3

1st 2nd 3rd 4th

Haz

ard

Rat

io (H

R)

Source: Cohen, JGIM 2008;23:1297-302

P=0.03

Q1 vs Q4

Na Quartile of based on mg of Na/d

Case of Systematic Error Leading to Bias:

Increased CVD Mortality in Persons with Lowest Na (by quartile of Na Intake in mg/d)

1.8 1.94

1.48

1

0

1

2

3

1st 2nd 3rd 4th

Haz

ard

Rat

io (H

R)

Source: Cohen, JGIM 2008;23:1297-302

P=0.03

Q1 vs Q4

Na Quartile of based on mg of Na/d

Evidence of Contamination in a Cohort Study1 (NHANESIII) Reporting Increased Mortality in Persons

with Low Sodium Intake on 24Hr Dietary Recall Quartile of Sodium Intake:

1st (Lowest) 2nd 3rd 4th (Highest)

Na (mg/d) 1,501 2,483 3,441 5,497

Energy Intake (kcal) 1,282 1,762 2,152 2,938 BMI (kg/m2) 25.8 26.4 26.3 26.6

1Cohen, JGIM 2008;23:1297-302

Evidence of Massive Underreporting of Calorie Intake Leading to Systematic Error in

Estimate or Sodium Intake

Reduced Systematic Error from Use of

Na/Kcal Ratio as Exposure instead of Na/d

1.25

0.9 0.9 1

0

1

2

1st 2nd 3rd 4th

Haz

ard

Rat

io (H

R)

1Cohen, JGIM 2008;23:1297-302

P=0.26

Q1 vs Q4

Na Quartile of based on mg of Na /Kcal

J-Shaped Relationship of Total Mortality with Urine Sodium Excretion in Patients with Type 1 Diabetes

Thomas, Diabetes Care 2011: 861-6

Extremely low levels are most likely the result of extreme

undercollection

Example of Low Sodium Excretion Related to Under-collection

•  78 year old women, screened for a trial – No special diet – 172 pounds (81 kg), 5’2”, BMI 31 kg/m2

•  Two 24 hour urine collections required – Detailed instructions provided

Urine Lab Range Expected 1st 2nd Sodium (mmol/24hr) 100? 18 21

Volume (ml/24hr) >500 800 725

Creatinine (g/24 hr) .63 to 2.5 1.2 .41 .09

Evidence  Summary  

•  Adverse effects of excess sodium intake – Established adverse effect on blood pressure,

CVD, and stroke – Probable or suggestive relationship for gastric

cancer, osteoporosis, and left ventricular mass

•  Adverse effects of low sodium intake

– Suggestive relationship of increased mortality in those with heart failure or diabetes

Outline  

Evidence  related  to  sodium  and  health  outcomes  •  Sodium:”  an  old  enemy  or  a  new  friend?”    

Recent  policy  making  on  sodium  •  Individual  and  public  health      

PopulaAon  sodium  reducAon    •  RealisAc?  Feasible?  Sustainable?  

Upper Limit of Sodium Intake •  Set at 2300 mg/day

•  It is not a recommended intake

•  Stated that consuming an intake above the lower limit for sodium provides no benefit

Source: IOM Dietary Reference Intakes for Electrolytes and Water

Adequate Intake for Sodium

•  Set at 1500 mg/day

•  Nutrient Adequacy - a diet that provides 1500 mg/day can also provide an adequate intake of other important nutrients (e.g., DASH Diet)

•  Sufficient to cover sodium sweat losses when exposed to high temperatures or in moderate physical activity

Source: IOM Dietary Reference Intakes for Electrolytes and Water

2010 Dietary Guidelines for Americans recommendeds a gradual reduction in the amount of sodium in the diets of Americans to 2,300 milligrams per person daily, and 1,500 milligrams for some special populations.

2013 AHA/ACC Lifestyle Report •  For blood pressure lowering:

–  lower sodium intake, in general; or – consume no more than 2,400 mg of sodium/

day

•  For even greater reduction in blood pressure: –  lower sodium intake to 1,500 mg per day; or

•  Lower sodium intake by at least 1,000 mg per day even if goals of 2,400 or 1,500 mg per day cannot be met.

Policy Summary

•  Policies in place (national and international) •  NYC DOH efforts have lead to voluntary

initiatives by food manufacturers and restaurants

•  Progress hindered by: –  Confusion resulting from confusing terminology (“salt-

sensitive”, “salt-resistant” hypertension) –  Activities of pro-sodium lobby groups –  Prominent scientists with opposite opinions

Outline  

Evidence  related  to  sodium  and  health  outcomes  •  Sodium:”  an  old  enemy  or  a  new  friend?”    

Recent  policy  making  on  sodium  •  Individual  and  public  health      

PopulaAon  sodium  reducAon    •  RealisAc?  Feasible?  Sustainable?  

Sources of Dietary Sodium

Inherent 12%

Food Processing

77% At the Table

6%

During Cooking 5%

Source: Mattes and Donnelly, 1991, J Am Coll of Nutr, 10(4): 383-93

*Includes breads, cereals, grains, processed meats and dairy, soups, gravies, sauces

Source: Anderson C et al, J Am Diet Assoc, 2010;110:736-745

Source: Johnson CM et al. Arch Intern Med. 2010;170 732-734

!"#$%&'()*+,$(-$.%#)*/&01$2#3*#4#$+"&+$5*#+$&05$3*-#1+,3#$/"&06#1$"&4#$&$7(1*84#$*%7&/+$(0$23((5$7)#119)#$%&0&6#%#0+:$$

>'-&.A/&.$(B//+(>&8/-1.A/&(C/"&,%$(B/"&+.A/&(JUQK(B//+(X(L'.$#2(3"-*'6(KW(

>8(0/1'/&'(7'-'(#-6%&=(#/(-'+",'(#2'%-(?$//+(;-'00"-'(/-(

50

2006

2011

51

2006

2011

Industry Sodium Reduction Initiatives 2010

Company Plans to reduce sodium

Kraft By 10% over next 2 years, in select products

ConAgra

By 20% over next 5 years, in all products

PepsiCo

By 25% over next 5 years, in all products

Campbell’s Soup By 35% over next 4 years; in past 4 years reduced100 products by 25-50%

Reality Summary

•! Current levels of sodium intake exceed physiologic need and guidelines

•! Over 75% of sodium comes from packaged, processed, and restaurant foods

•! There are outspoken critics of population-based recommendations to reduce sodium intake

•! Need rigorous investigations of consequences of low sodium intake

Balancing evidence, policy, and reality

•! New evidence suggestive of adverse effects from low sodium intake must be considered in new policy making for general population and especially population subgroups

•! The estimated benefits of sodium reduction are substantial and warrant major public health efforts to reduce intake