Skip main navigation

Cumulative Psychosocial Stress and Ideal Cardiovascular Health in Older Women

Data by Race/Ethnicity
Originally published 2019;139:2012–2021



Research implicates acute and chronic stressors in racial/ethnic health disparities, but the joint impact of multiple stressors on racial/ethnic disparities in cardiovascular health is unknown.


In 25 062 women (24 053 white; 256 Hispanic; 440 black; 313 Asian) articipating in the Women’s Health Study follow-up cohort, we examined the relationship between cumulative psychosocial stress (CPS) and ideal cardiovascular health (ICH), as defined by the American Heart Association’s 2020 strategic Impact Goals. This health metric includes smoking, body mass index, physical activity, diet, blood pressure, total cholesterol, and glucose, with higher levels indicating more ICH and less cardiovascular risk (score range, 0–7). We created a CPS score that summarized acute stressors (eg, negative life events) and chronic stressors (eg, work, work-family spillover, financial, discrimination, relationship, and neighborhood) and traumatic life event stress reported on a stress questionnaire administered in 2012 to 2013 (score range, 16–385, with higher scores indicating higher levels of stress).


White women had the lowest mean CPS scores (white: 161.7±50.4; Hispanic: 171.2±51.7; black: 172.5±54.9; Asian: 170.8±50.6; Poverall<0.01). Mean CPS scores remained higher in Hispanic, black, and Asian women than in white women after adjustment for age, socioeconomic status (income and education), and psychological status (depression and anxiety) (P<0.01 for each). Mean ICH scores varied by race/ethnicity (P<0.01) and were significantly lower in black women and higher in Asian women compared with white women (β-coefficient [95% CI]: Hispanics, −0.02 [−0.13 to −0.09]; blacks, −0.34 [−0.43 to −0.25]; Asians, 0.34 [0.24 to 0.45]); control for socioeconomic status and CPS did not change these results. Interactions between CPS and race/ethnicity in ICH models were not significant.


Both CPS and ICH varied by race/ethnicity. ICH remained worse in blacks and better in Asians compared with whites, despite taking into account socioeconomic factors and CPS.

Clinical Perspective

What Is New?

  • Cumulative psychological stress and socioeconomic status were evaluated as potential explanations for racial/ethnic differences in ideal cardiovascular health among older women.

  • Our results show that these factors did not explain the racial/ethnic differences noted in ideal cardiovascular health among older female health professionals.

  • Stressors were jointly evaluated and included both acute (eg, life events) and chronic (eg, job, financial, work-family conflict, neighborhood) sources of stress.

What Are the Clinical Applications?

  • Psychosocial stressors are social determinants of health that likely have different prevalence according to race/ethnicity.

  • Clinicians should incorporate psychosocial stressors in their assessment of patients because they might help to better target cardiovascular care, particularly since different patients will have wide variability in socioeconomic status, a potential source of stress.

  • This analysis used a psychosocial stress questionnaire with potential clinical utility in part because of its succinct but comprehensive components.


Cardiovascular disease (CVD) is a leading cause of death for every racial/ethnic group in the United States and accounts for much of the excess premature mortality among blacks compared with whites.1 Several decades of research have sought to unpack the mechanisms through which race/ethnicity impacts traditional modifiable risk factors for CVD, such as hypertension, elevated cholesterol, poor diet, and low physical activity, with the aim of targeting cardiovascular prevention programs to reduce differences in CVD outcomes.2 Although it is a known risk factor for CVD, including myocardial infarction,3,4 much of the work related to psychosocial stress in different racial/ethnic groups is limited to single domains of stress, such as discrimination, or to general perceived stress. Thus, almost no data exist pertaining to the relationship of a combination of acute and chronic stressors (cumulative psychosocial stress [CPS]) to race/ethnicity or to CVD health, resulting in significant gaps in the literature in these regards. Indeed, individuals typically experience a combination of stressors, and the examination of single domains of stress likely does not accurately capture the impact of cumulative stress over time. Moreover, whether the association between CPS and CVD varies based on race/ethnicity is unknown.

Psychosocial stress can result from multiple sources, including trauma, interpersonal relationships, employment, and neighborhood environment, potentially resulting in physiological consequences, including dysregulation of the inflammatory, neurohormonal, and autonomic nervous systems.5,6 Psychosocial stress associated with social inequalities is arguably disproportionately experienced by racial/ethnic minorities living in the United States.7,8 For example, job instability and financial strain are associated with increased serum C-reactive protein concentration in Mexicans living in the United States.9 Additionally, everyday discrimination is associated with increased serum biomarkers of endothelial dysfunction and blood pressure.6,10 These aforementioned forms of psychosocial stress in turn may be potential mechanisms by which racial/ethnic disparities in CVD risk occur.7–11

Ideal cardiovascular health (ICH) is a metric supported by the American Heart Association 2020 Impact Goals to improve the cardiovascular health of all Americans by 20%; in this metric, a higher score is associated with a more favorable cardiovascular profile.12 More than a mere tally of the absence of 7 lifestyle-related risk factors for CVD (tobacco use, blood pressure, cholesterol, fasting glucose, weight, physical activity, and diet), ICH was created to take into account both behavioral and clinical risk factors to be used both in research and in clinical practice as a pragmatic tool to empower patients to reduce their CVD risk. To date, emerging data indicate an inverse association of ICH with incident myocardial infarction, heart failure, stroke, and cardiovascular mortality in men and women.13,14 Although women live longer than men, black women in particular do not achieve as much of the cardiovascular benefit associated with sex as women of other races/ethnicities.15,16 Although lower socioeconomic status (SES) regardless of race/ethnicity is associated with poor cardiovascular health, differences in cardiovascular health outcomes by race/ethnicity are not fully explained by SES differences.17 Given that psychosocial stress is as important a risk factor for CVD as traditional risk factors such as smoking and dyslipidemia, and that psychosocial stressors might be differentially associated with heightened cardiovascular risk based on race/ethnicity and culture,3,18 we sought to assess the relationship between CPS and race/ethnicity and to evaluate the impact of CPS on any observed ICH-race/ethnicity relationship in women participating in the WHS (Women’s Health Study).13,19,20


Data Availability

The data will not be made available to other researchers for purposes of reproducing results; however, we highly encourage collaboration and contacting the corresponding author regarding sharing of methods and other information.

Study Population

We used participants from the follow-up cohort of the WHS, a completed randomized clinical trial of the effect of low-dose aspirin, vitamin E, and beta carotene in the primary prevention of cancer and CVD.21 Female health professionals ≥45 years of age in the United States were invited to participate in the trial (N=39 876). Study randomization began in April 1993, and follow-up for the clinical trial ended in March 2004. After trial conclusion, consenting participants from the trial were recruited into an observational cohort starting in 2005 (N=33 796). Follow-up questionnaires were conducted every 6 months during the first year and every year starting in year 2 to assess self-reported sociodemographic information and health outcomes. In 2012 to 2013, participants in the WHS observational cohort with no history of CVD were invited to participate in this stress cohort. A total of 25 335 participants were willing and eligible and provided informed consent for the stress study. Women with complete data on race/ethnicity, CPS, and ICH metrics were included in this analysis (N=25 062); 193 women without complete race/ethnicity, stress, and ICH data were excluded. Because of the small sample size of American Indian women (n=49) and women who reported their race/ethnicity as other (n=31), we do not present these data. Therefore, 273 women were excluded, representing 1.1% of this stress cohort. The distributions of available baseline characteristics were similar in the excluded group compared with women in this analysis (data not shown). This study was approved by the institutional review boards of Brigham and Women’s Hospital and the University of California, San Francisco.

Assessment of CPS

Details of the WHS stress follow-up study have been detailed previously.22 A mailed written questionnaire to evaluate acute and chronic psychosocial stressors was completed by each participant. Acute stress domains (reported as yes or no) included items regarding negative events in the past 5 years (eg, fired from a job, moving to a worse neighborhood, unemployment) and traumatic life events (eg, life-threatening illness or death of child/spouse, physical attack/assault victim). Chronic stress domains included work stress (eg, unable to express work creativity, lack of decision authority, monotony or excessive work, conflicting demands, job insecurity), work-family conflict (eg, too stressed out to participate in activities with family/others), financial stress (eg, inadequate funds, difficulty paying bills), intimate partner stress (eg, relationship happiness, partner demands, conflict resolution), neighborhood stress (eg, safety, community member trust/support), and everyday discrimination (eg, less respect, poor service, treated as unintelligent/dangerous). These chronic stress domains were assessed using 4 or 5 option Likert-style responses depending on the domain.22 Weights were assigned to each of the 8 domains that constituted the cumulative stress score based on the reciprocal of the SD of the scores for questions in each domain. The 8 domain-specific weighted scores were then added to create the cumulative stress score (CPS; range, 16–385), with higher values representing higher stress.

Assessment of ICH

We created an ICH score using the American Heart Association Strategic 2020 Impact Goals, which captures an index of ICH consisting of 7 health behaviors and factors. ICH is defined as body mass index (BMI) <25 kg/m2; ≥150 min/wk of moderate physical activity; a healthy diet pattern that includes sufficient amounts of fruits and vegetables; optimal blood pressure (<120/<80 mm Hg); fasting glucose <100 mg/dL; total cholesterol <200 mg/dL; and never smoking or smoking cessation for >12 months.12 ICH was determined for each participant in WHS using self-reported data collected within 1 year of stress study baseline in 2011 to 2012 because of the WHS follow-up data collection schedule. The accuracy of self-reported health conditions in WHS, including blood pressure, diabetes mellitus, and weight, is >90%.23–26 Diet information was collected in a self-administered food frequency questionnaire administered in 2004. Missing values for total cholesterol were imputed using the value from the data collection period immediately preceding the 2011 to 2012 follow-up assessment (ie, 2010–2011).

The definition of each component of ICH is consistent with the American Heart Association definition, with the following modifications to accommodate the methodology of the WHS (Table 1):

Table 1. Definition of Poor, Intermediate, and Ideal Cardiovascular Health in the Women’s Health Study

Ideal Cardiovascular Health
Health behaviors
 Weight/BMI≥30 kg/m225 to <30 kg/m2<25 kg/m2
 Physical activityModerate physical activity <1 min/wkModerate physical activity between 1 and 149 min/wkModerate physical activity ≥150 min/wk
Diet0–1 Components2–3 Components4–5Components
 Fruits and vegetables (>75th percentile)
Fish (>75th percentile)
Sodium (<25th percentile)
Sugar-sweetened beverages (<25th percentile)
Fiber-rich whole grains: fiber (>75th percentile); whole grains (>75th percentile)
Health factors
 Blood pressureSBP ≥140 mm Hg or DBP ≥90 mm HgSBP 120–139 mm Hg or DBP 80–89 mm Hg<120/80 mm Hg
 Total cholesterol≥240 mg/dL200–239 mg/dL<200 mg/dL
 Glycemia/fasting plasma glucosePositive T2DM diagnosisN/ANo history of T2DM

BMI indicates body mass index; DBP, diastolic blood pressure; N/A, not applicable; SBP, systolic blood pressure; and T2DM, type 2 diabetes mellitus.

  1. Physical activity: Ideal, moderate physical activity ≥150 min/wk; intermediate, 1 to 150 min/wk; poor, <1 min/wk.

  2. Diet: The diet metric is composed of 5 components consistent with a DASH (Dietary Approaches to Stop Hypertension) diet: fruits and vegetables, fish, fiber-rich whole grains, sodium, and sugar-sweetened beverages. In this analysis, each component of the DASH diet was assessed by 25th or 75th percentiles for the WHS cohort. Participants were classified as ideal if they reported fruit and vegetable consumption above the 75th percentile, fish intake above the 75th percentile, sodium intake below the 25th percentile, and sugar-sweetened beverage intake below the 25th percentile. Fiber-rich whole grains consumption was divided into whole grains intake and fiber intake, and each received a score of 0.5 above the 75th percentile. Diet was rated according to the number of dietary components that met the criteria for ideal (ideal, 4 to 5; intermediate, 2 to 3; or poor, 0 to 1) and presence or absence of glycemia (ideal, no self-report of type 2 diabetes mellitus; or poor, self-report of diabetes mellitus), consistent with the definitions used in the American Heart Association 2020 goals.12

For each component of ICH, participants were given a score of 1 for ideal classification, 0.5 for intermediate, and 0 for poor. The scores for the 7 metrics were summed to give a total ICH score ranging between 0 and 7, with 7 indicating an ideal risk factor category for all components. This continuous ICH score was categorized into 3 groups: low (scores 0–3), intermediate (scores 3.5–5.5), and high (scores 6–7).


Age, race/ethnicity, education level, and annual household income (income) were self-reported. Other conditions included in the analyses were obesity/overweight (normal, <25.0 kg/m2; overweight, 25.0 to <30.0 kg/m2; and obese, ≥30.0 kg/m2; for Asians: normal, <23.0 kg/m2; overweight, 23.0 to <27.0 kg/m2; and obese, ≥27.0 kg/m2), hypertension history, hypercholesterolemia, and alcohol use (<1 glass/d versus ≥1 glass/d).27,28

Physical activity was evaluated by self-report at initiation of the WHS and was updated every 2 to 3 years with questions based on the College Alumni Health Study, for which reliability and validity have been investigated extensively.29,30 Ascertained data include information about usual pattern of stair climbing and walking and average time spent weekly doing leisure-time activities (eg, walking/jogging, aerobic exercise, dance, swimming). Blood pressure was self-reported using the following criteria from the annual WHS follow-up questionnaire: (1) physician diagnosis of hypertension; (2) initiation of blood pressure–lowering medication; and (3) systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg. Validation of self-report of hypertension in WHS is 96%.24

Participants completed a validated, reproducible 131-item semiquantitative food frequency questionnaire before WHS randomization and in 2004 to assess dietary intake.31 Women responded to questions that inquired about frequency of consumption of the following foods: fruits and vegetables (includes fruit juices and excludes potatoes), fish (eg, canned tuna, shrimp/lobster, dark meat fish, breaded fish), fiber (eg, dark bread, brown rice, oatmeal, grains), sugar-sweetened beverages (eg, carbonated beverage with caffeine and sugar, punch, lemonade), and sodium. Responses were on a 9-point scale ranging from never/less than once monthly to ≥6 times/d. Because of skewedness related to food consumption distribution, food intake was treated as both continuous and dichotomous variables (ie, cut points were for fruits and vegetables ≥4.5 cups/d, fish ≥2 3.5-oz servings/wk, fiber-rich whole grains ≥3 1-oz equivalent servings/d, sugar-sweetened beverages ≤450 kcal (36 oz)/wk, sodium <1500 mg/d). Total caloric intake (kcal/d) was a continuous variable.32

Lipid levels were initially obtained via blood from WHS participants before randomization and by self-report. Thereafter, history of hypercholesterolemia was self-reported on annual follow-up questionnaires. The reliability of self-report is extremely high and correlates well with CVD risk.33 Self-reported type 2 diabetes mellitus is also collected on the annual WHS follow-up questionnaire and validated using American Diabetes Association criteria (positive predictive value for validation is 91%).23 Finally, self-reported weight correlates strongly (r=0.96) with directly measured weight.26 Psychological status including depressed mood (eg, feeling downhearted/blue/in the dumps) and anxiety (eg, nervousness, feeling calm/peaceful) was assessed by the stress questionnaire.

Statistical Analysis

Baseline characteristics are reported by race/ethnicity as means and medians; significance testing was performed with ANOVA or Kruskal-Wallis and χ2 tests. CPS is reported as median and associated 25th and 75th interquartile range based on race/ethnicity. To compare CPS by race/ethnicity, we used linear regression analyses and report the β-coefficients (reference group, whites). Linear regression models assessed the relationship between race/ethnicity as follows: model 1, unadjusted; model 2, age-adjusted; model 3, model 2 + SES (income and education); model 4, model 3 + psychological status (depression and anxiety symptoms); and model 5, model 4 + hypertension, BMI, smoking, alcohol use, and physical activity.

The relationship between race/ethnicity and ICH (continuous variable) was examined by linear regression: model 1, unadjusted; model 2, age-adjusted; model 3, age + CPS; model 4, model 3 + SES (education, income); and model 5, model 4 + psychological status. We also performed mediation analyses using the CAUSALMED procedure in SAS (SAS Institute Inc, Cary, NC) to understand the effect of cumulative stress on racial/ethnic differences in ICH. In addition, we examined the individual components of the ICH and CPS scores by race/ethnicity. Pearson correlation coefficients between CPS and ICH were also calculated. The proportional odds assumption and models of unequal slopes for race/ethnicity were performed. A 2-tailed P value <0.05 indicated statistical significance. Analyses used SAS 9.4 (SAS Institute Inc).


Among the 25 062 participating women (24 053 white; 256 Hispanic; 440 black; 313 Asian), overall mean age at the time of the stress questionnaire was 72.2±6.0 years. Compared with white women, Hispanic women had lower educational attainment and income, higher prevalent diabetes mellitus and hypercholesterolemia, higher BMI, and less current smoking; black women had lower income, more diabetes mellitus and hypertension, higher BMI, and more current smoking (Table 2). Asian women had higher educational attainment and income, more diabetes mellitus, and lower BMI and current smoking. White women had more prevalent alcohol use than women of other racial/ethnic groups.

Table 2. Sociodemographic Characteristics and Cardiometabolic Risk Factors of Participants in the Women’s Health Study by Race/Ethnicity

White (n=24 053)Hispanic (n=256)Black (n=440)Asian (n=313)P Overall
Age, y70.9 (67.3–75.8)70.4 (67.2–75.1)71.5 (68.0–75.4)71.1 (67.8–74.6)0.56
Less than college (<bachelor’s degree)50.855.550.620.9
College or more (≥bachelor’s degree)45.144.549.479.1
Household income<0.001
 Less than $50 00040.346.445.615.6
 $50 000 or more57.253.654.484.4
Type 2 diabetes mellitus9.819.523.019.5<0.001
Current hypertension69.870.388.470.9<0.001
Current hypercholesterolemia (>240 mg/dL)73.679.373.379.20.02
Current BMI, kg/m225.8 (23.0–29.7)26.5 (23.4–30.1)28.3 (25.7–32.1)23.0 (21.1–25.6)<0.001
Current BMI<0.001
 <25.0 kg/m242.738.019.371.3
 25.0 to <30.0 kg/m234.035.743.421.7
 ≥30.0 kg/m223.326.337.37.0
 <1 glass/d  ≥1glass/d84.6
Current METs (h/wk)17.5±16.516.3±18.116.1±17.521.7±17.5<0.001
Depression score5.5±2.15.7±2.65.3±2.05.7±2.00.01
Anxiety score4.4±1.64.4±1.84.0±1.54.4±1.6<0.001

Data are mean±SD, median (interquartile range), or %. BMI indicates body mass index; and MET, metabolic equivalent.

Figure 1 displays CPS scores (mean, median, and interquartile range) by race/ethnicity. Overall, nonwhites had higher CPS scores than whites. Specifically, mean CPS scores were lowest for white women (161.7±50.4). Black women had the highest mean CPS score (172.5±54.9), followed by Hispanic women (171.2±51.7) and then Asian women (170.8±50.6). Table 3 shows the results of linear regression analyses of CPS by race/ethnicity (white women=referent). After adjustment for potential confounders, black (β=0.08; 95% CI, 0.06–0.10), Hispanic (β=0.05; 95% CI, 0.03–0.08), and Asian (β=0.04; 95% CI, 0.02–0.07) women reported higher CPS than white women (all P<0.001). Table 4 shows the relationship between the individual stressors and ICH by race/ethnicity. These results revealed that the impact of each stress domain on ICH was relatively similar in magnitude in blacks and Asians. The effect of individual stressor domains in Hispanics was negligible and not statistically significant.

Table 3. Comparison of Cumulative Psychosocial Stress (CPS) Scores by Race/Ethnicity in the Women’s Health Study

ModelHispanic (n=256)Black (n=440)Asian (n=313)
Model 1: unadjusted6.7 (3.9 to 9.6)
7.7 (5.6 to 9.9)
2.4 (−0.03 to 4.9)
Model 26.4 (3.7 to 9.0)
7.4 (5.3 to 9.4)
2.5 (0.007 to 4.8)
Model 36.1 (3.4 to 8.7)
7.4 (5.3 to 9.5)
4.8 (2.4 to 7.3)
Model 45.5 (3.1 to 7.8)
9.5 (7.6 to 11.3)
4.0 (1.8 to 6.2)
Model 5: Fully adjusted5.4 (3.1 to 7.8)
8.2 (6.1 to 10.1)
4.3 (2.1 to 6.5)

Data are percent differences from white women (reference) (95% CI).

Model 1: unadjusted.

Model 2: age.

Model 3: model 2 + income + education.

Model 4: model 3 + depression/anxiety;

Model 5: model 4 + hypertension + smoking + alcohol use + metabolic equivalents + body mass index.

Table 4. Ideal Cardiovascular Health in Relation to Individual Psychosocial Stress Domains by Race/Ethnicity

Individual StressorWhiteHispanicBlackAsian
β (SE)β (SE)P Valueβ (SE)P Valueβ (SE)P Value
Work stress1.00−0.08 (0.06)0.16−0.41 (0.04)<0.010.41 (0.05)<0.01
Work-family stress1.00−0.08 (0.06)0.18−0.43 (0.04)<0.010.43 (0.05)<0.01
Financial stress1.00−0.05 (0.06)0.38−0.35 (0.04)<0.010.38 (0.05)<0.01
Traumatic events1.00−0.07 (0.06)0.24−0.42 (0.04)<0.010.41 (0.05)<0.01
Negative life events (past 5 y)1.00−0.07 (0.06)0.24−0.41 (0.04)<0.010.42 (0.05)<0.01
Discrimination stress1.00−0.07 (0.06)0.23−0.39 (0.04)<0.010.44 (0.05)<0.01
Relationship stress1.00−0.06 (0.06)0.31−0.36 (0.04)<0.010.44 (0.05)<0.01
Neighborhood stress1.00−0.04 (0.06)0.47−0.38 (0.04)<0.010.43 (0.05)<0.01

White women=reference. β indicates β-coefficient; and SE, standard error.

Figure 1.

Figure 1. Cumulative psychosocial stress score by race/ethnicity in the Women’s Health Study stress study. Mean (◊) and median cumulative stress scores and associated interquartile range according to race/ethnic group among participants in the Women’s Health Study stress study. P value across race/ethnicity groups <0.001.

As shown in Figure 2, mean ICH score was the highest in Asian women (5.0±0.9), followed by white (4.6±0.9), Hispanic (4.5±1.0), and black (4.1±1.0) women. Black women had the highest proportion of participants with low ICH scores (19.1%), that is, lowest cardiovascular health, and the lowest percentage of participants with high ICH scores (3.9%) (Table I in the online-only Data Supplement). Asian women had the lowest percentage of participants with low ICH scores (4.8%) and the highest percentage of participants with high ICH scores (19.2%), that is, with ICH. The odds of having ideal ICH classification (ICH score=6–7) versus poor ICH classification (ICH score=0–3) were significantly decreased for black women compared with white women (odds ratio [OR], 0.46; 95% CI, 0.28–0.75) after adjustment for age, CPS, and SES (Table I in the online-only Data Supplement). Asian women were 2 times more likely than white women to have ideal versus poor ICH (OR, 2.21; 95% CI, 1.65–2.98).

Figure 2.

Figure 2. Ideal cardiovascular health score by race/ethnicity in the Women’s Health Study. Mean (◊) and median ideal cardiovascular health scores and associated interquartile range according to race/ethnic group among participants in the Women’s Health Study stress study. P<0.001.

We also observed that in all race/ethnicity groups, physical activity was the ICH metric with the largest proportion of women categorized as ideal, with a range from 73.8% to 88.6% (Table II in the online-only Data Supplement). The BMI component of the ICH score had the widest range of percentage of participants meeting the criteria for ideal classification, ranging from 19.3% in black women to 71.3% in Asian women. Asian women had the highest percentage of participants meeting the criteria for an ideal rating for the greatest number of individual components of the ICH score. When the individual components of ICH were examined (ie, odds for ideal versus poor component), the odds of absence of diabetes mellitus (ideal) were lower in Hispanics (OR, 0.49; 95% CI, 0.36–0.67), blacks (OR, 0.42; 95% CI, 0.34–0.53), and Asians (OR, 0.45; 95% CI, 0.34–0.60) compared with whites. Additionally in Asians, the odds of other ideal ICH components were significantly higher (eg, never smokers: OR, 3.92; 95% CI, 1.72–8.86) compared with whites, besides total cholesterol. The odds of ideal versus poor blood pressure were significantly lower in blacks (OR, 0.61; 95% CI, 0.44–0.86) compared with whites.

Table 5 shows results of linear regression analyses of ICH according to race/ethnicity. As previously noted, compared with white women, black women and Hispanic women had lower ICH scores and Asian women had higher ICH scores. These racial/ethnic differences persisted and remained statistically significant despite adjustment for CPS (model 3) and additionally for SES (model 4), except among Hispanics. In age-adjusted models, among blacks compared with whites, ≈12.7% of ICH was mediated by CPS (P<0.001). No significant associations were noted in Hispanics (P=0.27) and Asians (P=0.06). The overall correlation between ICH and the CPS score was small (r=−0.06, P<0.001).

Table 5. β-Coefficients of Ideal Cardiovascular Health Score by Race/Ethnicity

Model 1−0.07
(−0.18 to −0.05)
(−0.45 to −0.28)
(0.32 to 0.52)
Model 2−0.06
(−0.18 to −0.05)
(−0.45 to −0.28)
(0.31 to 0.52)
Model 3−0.03
(−0.14 to 0.09)
(−0.40 to −0.23)
(0.33 to 0.56)
Model 4−0.02
(−0.13 to 0.09)
(−0.42 to −0.24)
(0.24 to 0.45)
Model 5−0.02
(−0.13 to 0.09)
(−0.43 to −0.25)
(0.24 to 0.45)

Data are β-coefficient (95% CI). Reference=white women. Negative values indicate lower ideal cardiovascular health.

Model 1: unadjusted.

Model 2: age.

Model 3: model 2 + cumulative psychosocial stress.

Model 4: model 3 + education + income.

Model 5: model 4 + depression/anxiety.


In this cross-sectional study of older women participating in the WHS, both CPS and ICH varied by race/ethnicity. White women had lower CPS scores than other women. Black women had the highest CPS and the lowest ICH scores. Asian women had the highest ICH scores and were more likely categorized as having ICH than white women. The observed racial/ethnic differences in ICH persisted after adjustment for CPS or SES. These data suggest that the mechanisms by which racial/ethnic differences in cardiometabolic risk in this cohort likely occur are through multiple pathways beyond SES and measured acute and chronic stressors. Given that racial/ethnic heterogeneity exists in all major groups discussed in this article and is primarily a social construct that serves as a surrogate for other individual and societal variables that might affect health behaviors and health outcomes, our analysis significantly adds to existing literature by additionally examining the impact of cumulative stress on ICH. In future research, it will be important to evaluate whether alternative modes to assess stress can be of additional value for understanding the association between race/ethnicity and cardiovascular health.

Although the prevalence of ICH was higher in our study than in other cohorts, the racial/ethnic differences in ICH were similar to those observed in other cohorts.19,20,34 NHANES (National Health and Nutrition Examination Survey) 2003 to 2008 data demonstrated a prevalence of ICH in women age ≥65 years of <2%, which is considerably less than what was reported for each of the race/ethnicity groups in our study.34 However, the participants of WHS were health professionals and thus had a relatively higher educational attainment than other cohorts and likely higher exposure to information on healthy behaviors, which might account for the higher prevalence of ICH. Still, despite the overall increased prevalence of ICH in the WHS, racial/ethnic differences in ICH remain. In other cohorts, despite the considerable heterogeneity in the manner in which the components of ICH have been defined and measured, particularly in diet, the pattern of worse ICH in blacks is a recurrent observation. When the NHANES (2003–2008) data were examined by race/ethnicity, black women aged ≥65 years had the lowest prevalence of ideal classification in the areas of tobacco, BMI, diet, and blood pressure, whereas Mexican American women of the same age group had the lowest prevalence of ideal physical activity and fasting glucose.34 Of note, Asian American women were not included in many of the studies of ICH with the exception of a study from the BRFSS (Behavioral Risk Factor Surveillance System). In an analysis of BRFSS 2009 data, patterning in overall ICH scores by race/ethnicity was similar to our study, with Asians having the highest prevalence of ICH, followed by whites, Hispanics, and blacks.19 Importantly, in this analysis, we used recommended cut points for Asian women, who tend to have higher CVD risk at lower BMI.27,28 ICH in Hispanics differs by national background, as observed in the Northern Manhattan Study, in which Caribbean Hispanics had the lowest prevalence of ICH and the lowest prevalence of ideal classification in the areas of BMI and blood pressure.13

In the present data, ICH varied by race/ethnicity despite taking into account CPS and SES. This is surprising, because the CPS scores were statistically different by race/ethnicity, and previous work from this cohort indicates that robust socioeconomic diversity is linked to CVD outcomes even though the WHS participants were female health professionals.35 Despite this, the overlapping CIs between CPS by race/ethnicity and the likelihood that racial/ethnic differences in ICH in women occur through multiple mechanisms (including genetic but more likely environmental, behavioral, and social factors) may have impacted the findings in this cohort of health professionals.7,36 For example, characteristics of neighborhood physical and social environment, such as walkability and safety, are associated with hypertension and obesity prevalence.37 Neighborhood characteristics also influence cardiovascular health behaviors.36–42 Indeed, the ICH metric comprises both cardiometabolic risk factors and health behaviors, each of which might have different upstream social determinants, including cultural practices that influence food choices and leisure-time physical activity. Although the CPS measurement herein included neighborhood stressors, it is possible that the magnitude of the range in the neighborhood stress score for these female health professionals might be different from that noted in the general population. Despite this, the ability to capture multiple domains of potential stressors in this well-phenotyped cohort is a significant addition to the current literature. Certainly, WHS data about race/ethnicity and inflammation was one of the first published studies about racial/ethnic differences in high-sensitivity C-reactive protein and have remained generalizable over time to other cohorts and to the general population.43

Despite the preceding, limitations of this study merit discussion. First, these data are cross-sectional. Stress caused by medical problems was omitted from the CPS assessment for this analysis to help mitigate the potential for reverse causality. Second, participants in WHS are older female healthcare professionals, and thus, similar analyses in other populations, including in men and younger individuals, are needed. Health behaviors of female health professionals might differ from those of other populations. For example, as noted, physical activity levels tend to be greater among participants in the WHS, albeit this would tend to underestimate any observed differences noted in this analysis. Third, the stressors measured in the stress questionnaire are self-reported; however, the perceived experience of psychosocial stress likely represents the actual stress burden of the participant, because it is an individual’s perception of a stressor and response to it that influences their physiological response or health outcome via the inflammatory, neurohormonal, and autonomic regulatory systems.5 Additionally, although the WHS consists of a mailed survey, and other variables such as hypertension and diabetes mellitus are self-reported, the validity of these self-reported variables within the WHS has had >90% accuracy with physician diagnosis.23–26 Weight as a self-reported variable has been validated in other cohorts.44 Fourth, ≈42% and 35% of variables related to work and relationship were missing, likely because many of the women were older retirees. However, sensitivity analyses revealed that our results without these missing data were similar. Fifth, the dietary questionnaire (2004) was conducted 7 years before the stress questionnaire (2012–2013), and thus, dietary patterns of these women could have changed over time. We are unable to correct for this limitation but note that there have been several published reports from the WHS using these same diet data that accurately predict disease risk.45–47 Finally, we cannot account for unmeasured confounders.

In conclusion, these data show that although ICH varied by race/ethnicity among female health professionals participating in the WHS, the association between race/ethnicity and ICH was unattenuated by our joint assessment of acute and chronic stressors or by SES as measured by income and education. Our results are intriguing and suggest that other pathways likely contribute to racial/ethnic differences in ICH within the WHS. Additionally, it is uncertain whether our findings are cohort-specific, and thus, similar work in other populations is required. To the best of our knowledge, data regarding this topic remain sparse. Future work must include prospective analyses about the interplay between race/ethnicity, CPS, and CVD risk, because neither behavior at the individual level nor policy interventions intended to reduce CVD risk currently incorporate psychosocial stressors.


In a cohort of older women, lifetime stress (as defined by acute and chronic stressors) and ICH varied by race/ethnicity. CPS and SES did not account for the racial/ethnic differences in ICH. These data support the need for additional work that addresses the joint impact of multiple social determinants of health, such as psychosocial stress, on CVD in diverse populations.


The authors express much gratitude to the participants of the WHS ongoing cohort for their participation in this stress ancillary study. The authors thank Dr Alan Zaslavsky for his consultation and Manickavasa Vinayagamoorthy, Eva Durazo, and Fumika Matsushita for their contributions.


Sources of Funding, see page 2020

Continuing medical education (CME) credit is available for this article. Go to to take the quiz.

The online-only Data Supplement is available with this article at

Michelle A. Albert, MD, MPH, 505 Parnasuss Ave, San Francisco, CA 94310. Email


  • 1. Carnethon MR, Pu J, Howard G, Albert MA, Anderson CAM, Bertoni AG, Mujahid MS, Palaniappan L, Taylor HA, Willis M, Yancy CW; American Heart Association Council on Epidemiology and Prevention; Council on Cardiovascular Disease in the Young; Council on Cardiovascular and Stroke Nursing; Council on Clinical Cardiology; Council on Functional Genomics and Translational Biology; and Stroke Council. Cardiovascular health in African Americans: a scientific statement from the American Heart Association.Circulation. 2017; 136:e393–e423. doi: 10.1161/CIR.0000000000000534LinkGoogle Scholar
  • 2. Mensah GA. Eliminating disparities in cardiovascular health: six strategic imperatives and a framework for action.Circulation. 2005; 111:1332–1336. doi: 10.1161/01.CIR.0000158134.24860.91LinkGoogle Scholar
  • 3. Rosengren A, Hawken S, Ounpuu S, Sliwa K, Zubaid M, Almahmeed WA, Blackett KN, Sitthi-amorn C, Sato H, Yusuf S; INTERHEART Investigators. Association of psychosocial risk factors with risk of acute myocardial infarction in 11119 cases and 13648 controls from 52 countries (the INTERHEART study): case-control study.Lancet. 2004; 364:953–962. doi: 10.1016/S0140-6736(04)17019-0CrossrefMedlineGoogle Scholar
  • 4. Cummings DM, Kirian K, Howard G, Howard V, Yuan Y, Muntner P, Kissela B, Redmond N, Judd SE, Safford MM. Consequences of comorbidity of elevated stress and/or depressive symptoms and incident cardiovascular outcomes in diabetes: results from the REasons for Geographic And Racial Differences in Stroke (REGARDS) Study.Diabetes Care. 2016; 39:101–109. doi: 10.2337/dc15-1174CrossrefMedlineGoogle Scholar
  • 5. McEwen BS. Protective and damaging effects of stress mediators.N Engl J Med. 1998; 338:171–179. doi: 10.1056/NEJM199801153380307CrossrefMedlineGoogle Scholar
  • 6. Friedman EM, Williams DR, Singer BH, Ryff CD. Chronic discrimination predicts higher circulating levels of E-selectin in a national sample: the MIDUS study.Brain Behav Immun. 2009; 23:684–692. doi: 10.1016/j.bbi.2009.01.002CrossrefMedlineGoogle Scholar
  • 7. Sternthal MJ, Slopen N, Williams DR. Racial disparities in health: how much does stress really matter?Du Bois Rev. 2011; 8:95–113. doi: 10.1017/S1742058X11000087CrossrefMedlineGoogle Scholar
  • 8. Hatch SL, Dohrenwend BP. Distribution of traumatic and other stressful life events by race/ethnicity, gender, SES and age: a review of the research.Am J Community Psychol. 2007; 40:313–332. doi: 10.1007/s10464-007-9134-zCrossrefMedlineGoogle Scholar
  • 9. Steffen PR, Walker J, Meredith R, Anderson C. The effects of job instability and financial strain on C-reactive protein in a sample of Mexican immigrants.Ethn Dis. 2016; 26:37–44. doi: 10.18865/ed.26.1.37CrossrefMedlineGoogle Scholar
  • 10. Rodriguez CJ, Gwathmey TM, Jin Z, Schwartz J, Beech BM, Sacco RL, Di Tullio MR, Homma S. Perceived discrimination and nocturnal blood pressure dipping among Hispanics: the influence of social support and race.Psychosom Med. 2016; 78:841–850. doi: 10.1097/PSY.0000000000000341CrossrefMedlineGoogle Scholar
  • 11. Williams DR, Mohammed SA. Discrimination and racial disparities in health: evidence and needed research.J Behav Med. 2009; 32:20–47. doi: 10.1007/s10865-008-9185-0CrossrefMedlineGoogle Scholar
  • 12. Lloyd-Jones DM, Hong Y, Labarthe D, Mozaffarian D, Appel LJ, Van Horn L, Greenlund K, Daniels S, Nichol G, Tomaselli GF, Arnett DK, Fonarow GC, Ho PM, Lauer MS, Masoudi FA, Robertson RM, Roger V, Schwamm LH, Sorlie P, Yancy CW, Rosamond WD; American Heart Association Strategic Planning Task Force and Statistics Committee. Defining and setting national goals for cardiovascular health promotion and disease reduction: the American Heart Association’s strategic Impact Goal through 2020 and beyond.Circulation. 2010; 121:586–613. doi: 10.1161/CIRCULATIONAHA.109.192703LinkGoogle Scholar
  • 13. Dong C, Rundek T, Wright CB, Anwar Z, Elkind MS, Sacco RL. Ideal cardiovascular health predicts lower risks of myocardial infarction, stroke, and vascular death across whites, blacks, and Hispanics: the Northern Manhattan Study.Circulation. 2012; 125:2975–2984. doi: 10.1161/CIRCULATIONAHA.111.081083LinkGoogle Scholar
  • 14. Fang N, Jiang M, Fan Y. Ideal cardiovascular health metrics and risk of cardiovascular disease or mortality: a meta-analysis.Int J Cardiol. 2016; 214:279–283. doi: 10.1016/j.ijcard.2016.03.210CrossrefMedlineGoogle Scholar
  • 15. Benjamin EJ, Blaha MJ, Chiuve SE, Cushman M, Das SR, Deo R, de Ferranti SD, Floyd J, Fornage M, Gillespie C, Isasi CR, Jiménez MC, Jordan LC, Judd SE, Lackland D, Lichtman JH, Lisabeth L, Liu S, Longenecker CT, Mackey RH, Matsushita K, Mozaffarian D, Mussolino ME, Nasir K, Neumar RW, Palaniappan L, Pandey DK, Thiagarajan RR, Reeves MJ, Ritchey M, Rodriguez CJ, Roth GA, Rosamond WD, Sasson C, Towfighi A, Tsao CW, Turner MB, Virani SS, Voeks JH, Willey JZ, Wilkins JT, Wu JH, Alger HM, Wong SS, Muntner P; American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics-2017 update: a report from the American Heart Association.Circulation. 2017; 135:e146–e603. doi: 10.1161/CIR.0000000000000485LinkGoogle Scholar
  • 16. Dwyer-Lindgren L, Bertozzi-Villa A, Stubbs RW, Morozoff C, Mackenbach JP, van Lenthe FJ, Mokdad AH, Murray CJL. Inequalities in life expectancy among US counties, 1980 to 2014: temporal trends and key drivers.JAMA Intern Med. 2017; 177:1003–1011. doi: 10.1001/jamainternmed.2017.0918CrossrefMedlineGoogle Scholar
  • 17. Bucholz EM, Ma S, Normand SL, Krumholz HM. Race, socioeconomic status, and life expectancy after acute myocardial infarction.Circulation. 2015; 132:1338–1346. doi: 10.1161/CIRCULATIONAHA.115.017009LinkGoogle Scholar
  • 18. Xu T, Li W, Teo K, Wang XY, Liu LS, Yusuf S; INTER-HEART China Investigators. Association of psychological risk factors and acute myocardial infarction in China: the INTER-HEART China study.Chin Med J (Engl). 2011; 124:2083–2088.MedlineGoogle Scholar
  • 19. Fang J, Yang Q, Hong Y, Loustalot F. Status of cardiovascular health among adult Americans in the 50 States and the District of Columbia, 2009.J Am Heart Assoc. 2012; 1:e005371. doi: 10.1161/JAHA.112.005371LinkGoogle Scholar
  • 20. Yang Q, Cogswell ME, Flanders WD, Hong Y, Zhang Z, Loustalot F, Gillespie C, Merritt R, Hu FB. Trends in cardiovascular health metrics and associations with all-cause and CVD mortality among US adults.JAMA. 2012; 307:1273–1283. doi: 10.1001/jama.2012.339CrossrefMedlineGoogle Scholar
  • 21. Rexrode KM, Lee IM, Cook NR, Hennekens CH, Buring JE. Baseline characteristics of participants in the Women’s Health Study.J Womens Health Gend Based Med. 2000; 9:19–27. doi: 10.1089/152460900318911CrossrefMedlineGoogle Scholar
  • 22. Albert MA, Durazo EM, Slopen N, Zaslavsky AM, Buring JE, Silva T, Chasman D, Williams DR. Cumulative psychological stress and cardiovascular disease risk in middle aged and older women: rationale, design, and baseline characteristics.Am Heart J. 2017; 192:1–12. doi: 10.1016/j.ahj.2017.06.012CrossrefMedlineGoogle Scholar
  • 23. Liu S, Lee IM, Song Y, Van Denburgh M, Cook NR, Manson JE, Buring JE. Vitamin E and risk of type 2 diabetes in the Women’s Health Study randomized controlled trial.Diabetes. 2006; 55:2856–2862. doi: 10.2337/db06-0456CrossrefMedlineGoogle Scholar
  • 24. Sesso HD, Cook NR, Buring JE, Manson JE, Gaziano JM. Alcohol consumption and the risk of hypertension in women and men.Hypertension. 2008; 51:1080–1087. doi: 10.1161/HYPERTENSIONAHA.107.104968LinkGoogle Scholar
  • 25. Wang L, Manson JE, Buring JE, Sesso HD. Meat intake and the risk of hypertension in middle-aged and older women.J Hypertens. 2008; 26:215–222. doi: 10.1097/HJH.0b013e3282f283dcCrossrefMedlineGoogle Scholar
  • 26. Willett W, Stampfer MJ, Bain C, Lipnick R, Speizer FE, Rosner B, Cramer D, Hennekens CH. Cigarette smoking, relative weight, and menopause.Am J Epidemiol. 1983; 117:651–658.CrossrefMedlineGoogle Scholar
  • 27. WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies [published correction appears in Lancet. 2004;363:902].Lancet. 2004; 363:157–163.CrossrefMedlineGoogle Scholar
  • 28. Rao G, Powell-Wiley TM, Ancheta I, Hairston K, Kirley K, Lear SA, North KE, Palaniappan L, Rosal MC; American Heart Association Obesity Committee of the Council on Lifestyle and Cardiometabolic Health. Identification of obesity and cardiovascular risk in ethnically and racially diverse populations: a scientific statement from the American Heart Association.Circulation. 2015; 132:457–472. doi: 10.1161/CIR.0000000000000223LinkGoogle Scholar
  • 29. Wolf AM, Hunter DJ, Colditz GA, Manson JE, Stampfer MJ, Corsano KA, Rosner B, Kriska A, Willett WC. Reproducibility and validity of a self-administered physical activity questionnaire.Int J Epidemiol. 1994; 23:991–999.CrossrefMedlineGoogle Scholar
  • 30. Lee I-M, Paffenbarger RSDesign of present-day epidemiologic studies of physical activity and health.Lee I-M, ed. In: Epidemiologic Methods in Physical Activity Studies. New York, NY: Oxford University Press; 2009:100–123.Google Scholar
  • 31. Willett WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi J, Hennekens CH, Speizer FE. Reproducibility and validity of a semiquantitative food frequency questionnaire.Am J Epidemiol. 1985; 122:51–65.CrossrefMedlineGoogle Scholar
  • 32. Samieri C, Grodstein F, Rosner BA, Kang JH, Cook NR, Manson JE, Buring JE, Willett WC, Okereke OI. Mediterranean diet and cognitive function in older age.Epidemiology. 2013; 24:490–499. doi: 10.1097/EDE.0b013e318294a065CrossrefMedlineGoogle Scholar
  • 33. Huang PY, Buring JE, Ridker PM, Glynn RJ. Awareness, accuracy, and predictive validity of self-reported cholesterol in women.J Gen Intern Med. 2007; 22:606–613. doi: 10.1007/s11606-007-0144-1CrossrefMedlineGoogle Scholar
  • 34. Shay CM, Ning H, Allen NB, Carnethon MR, Chiuve SE, Greenlund KJ, Daviglus ML, Lloyd-Jones DM. Status of cardiovascular health in US adults: prevalence estimates from the National Health and Nutrition Examination Surveys (NHANES) 2003-2008.Circulation. 2012; 125:45–56. doi: 10.1161/CIRCULATIONAHA.111.035733LinkGoogle Scholar
  • 35. Albert MA, Glynn RJ, Buring J, Ridker PM. Impact of traditional and novel risk factors on the relationship between socioeconomic status and incident cardiovascular events.Circulation. 2006; 114:2619–2626. doi: 10.1161/CIRCULATIONAHA.106.660043LinkGoogle Scholar
  • 36. Mujahid MS, Diez Roux AV, Cooper RC, Shea S, Williams DR. Neighborhood stressors and race/ethnic differences in hypertension prevalence (the Multi-Ethnic Study of Atherosclerosis).Am J Hypertens. 2011; 24:187–193. doi: 10.1038/ajh.2010.200CrossrefMedlineGoogle Scholar
  • 37. Powell-Wiley TM, Ayers CR, de Lemos JA, Lakoski SG, Vega GL, Grundy S, Das SR, Banks-Richard K, Albert MA. Relationship between perceptions about neighborhood environment and prevalent obesity: data from the Dallas Heart Study.Obesity (Silver Spring). 2013; 21:E14–E21. doi: 10.1002/oby.20012CrossrefMedlineGoogle Scholar
  • 38. Clark CR, Ommerborn MJ, Hickson DA, Grooms KN, Sims M, Taylor HA, Albert MA. Neighborhood disadvantage, neighborhood safety and cardiometabolic risk factors in African Americans: biosocial associations in the Jackson Heart study.PLoS One. 2013; 8:e63254. doi: 10.1371/journal.pone.0063254CrossrefMedlineGoogle Scholar
  • 39. Unger E, Diez-Roux AV, Lloyd-Jones DM, Mujahid MS, Nettleton JA, Bertoni A, Badon SE, Ning H, Allen NB. Association of neighborhood characteristics with cardiovascular health in the Multi-Ethnic Study of Atherosclerosis.Circ Cardiovasc Qual Outcomes. 2014; 7:524–531. doi: 10.1161/CIRCOUTCOMES.113.000698LinkGoogle Scholar
  • 40. Barnes R, Foster SA, Pereira G, Villanueva K, Wood L. Is neighbourhood access to tobacco outlets related to smoking behaviour and tobacco-related health outcomes and hospital admissions?Prev Med. 2016; 88:218–223. doi: 10.1016/j.ypmed.2016.05.003CrossrefMedlineGoogle Scholar
  • 41. Burgoine T, Forouhi NG, Griffin SJ, Brage S, Wareham NJ, Monsivais P. Does neighborhood fast-food outlet exposure amplify inequalities in diet and obesity? A cross-sectional study.Am J Clin Nutr. 2016; 103:1540–1547. doi: 10.3945/ajcn.115.128132CrossrefMedlineGoogle Scholar
  • 42. Creatore MI, Glazier RH, Moineddin R, Fazli GS, Johns A, Gozdyra P, Matheson FI, Kaufman-Shriqui V, Rosella LC, Manuel DG, Booth GL. Association of neighborhood walkability with change in overweight, obesity, and diabetes.JAMA. 2016; 315:2211–2220. doi: 10.1001/jama.2016.5898CrossrefMedlineGoogle Scholar
  • 43. Albert MA, Glynn RJ, Buring J, Ridker PM. C-reactive protein levels among women of various ethnic groups living in the United States (from the Women’s Health Study).Am J Cardiol. 2004; 93:1238–1242. doi: 10.1016/j.amjcard.2004.01.067CrossrefMedlineGoogle Scholar
  • 44. Rimm EB, Stampfer MJ, Colditz GA, Chute CG, Litin LB, Willett WC. Validity of self-reported waist and hip circumferences in men and women.Epidemiology. 1990; 1:466–473.CrossrefMedlineGoogle Scholar
  • 45. Chiuve SE, Sandhu RK, Moorthy MV, Glynn RJ, Albert CM. Dietary fat intake is differentially associated with risk of paroxysmal compared with sustained atrial fibrillation in women.J Nutr. 2015; 145:2092–2101. doi: 10.3945/jn.115.212860CrossrefMedlineGoogle Scholar
  • 46. Rist PM, Buring JE, Kase CS, Kurth T. Healthy lifestyle and functional outcomes from stroke in women.Am J Med. 2016; 129:715–724.e2. doi: 10.1016/j.amjmed.2016.02.002CrossrefMedlineGoogle Scholar
  • 47. Rautiainen S, Wang L, Lee IM, Manson JE, Buring JE, Sesso HD. Dairy consumption in association with weight change and risk of becoming overweight or obese in middle-aged and older women: a prospective cohort study.Am J Clin Nutr. 2016; 103:979–988. doi: 10.3945/ajcn.115.118406CrossrefMedlineGoogle Scholar


eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.