Contemporary work suggests that consumption of polyunsaturated fatty acids (PUFAs), in place of saturated fats, decreases the risk of cardiovascular disease (CVD) [
MESA is a prospective cohort study that began in July 2000 to investigate the prevalence, correlates, and progression of subclinical CVD in individuals without known CVD at baseline [
Fasting blood samples were collected, and fatty acid analyses were performed at the Collaborative Studies Clinical Laboratory at Fairview-University Medical Center (Minneapolis, Minn, USA) as previously described [
CMR imaging was performed with 1.5-T whole-body MRI systems, Signa CV/I or Signa LX (General Electric Medical Systems), at the first examination. Determination of LV mass and volumes have been previously described and were adjusted for body size [
Evaluation of aortic distensibility has also been previously described [
Means and standard deviations or proportions were calculated for selected variables according to quartiles of plasma fatty acids (Table
Relation between baseline characteristics and plasma phospholipid DHA, LA.‡
LA quartile ( | DHA quartile ( | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Q1 | Q2 | Q3 | Q4 | Q1 | Q2 | Q3 | Q4 | |||
trend | trend | |||||||||
Age in years, mean ± SE | 65.4 ± 0.7 | 63.7 ± 0.7 | 63.7 ± 0.7 | 62.96 ± 0.8 | 0.002 | 64.0 ± 0.9 | 62.7 ± 1.0 | 62.5 ± 1.0 | 62.1 ± 0.7 | 0.01 |
Gender, %female | 65.5 | 61.1 | 61.3 | 51.8 | 0.0006 | 54.2 | 61.6 | 60 | 69.1 | 0.0008 |
Race/ethnicity, % | ||||||||||
White | 20.2 | 12.9 | 14.6 | 11.1 | <.0001 | 24.3 | 15.9 | 4.9 | 5.1 | 0.56 |
Black | 22.9 | 16.5 | 11.3 | 4.7 | 8.2 | 16.8 | 19.2 | 15.9 | ||
Chinese | 12.7 | 19.4 | 24.2 | 34.7 | 4.9 | 18 | 37.2 | 45.8 | ||
Hispanic | 44.2 | 51.1 | 49.8 | 49.4 | 62.7 | 49.4 | 38.7 | 33.1 | ||
BMI, mean ± SE | 28.1 ± 0.3 | 28.4 ± 0.3 | 27.7 ± 5 | 27.7 ± 0.3 | 0.04 | 28.4 ± 0.3 | 28.3 ± 0.3 | 27.7 ± 0.3 | 27.5 ± 0.3 | 0.005 |
Education ≥ high school, % | 82 | 79.5 | 79.7 | 83.9 | 0.51 | 79.9 | 81.1 | 82.8 | 81.1 | 0.64 |
Cigarette smoking, % | ||||||||||
Current | 15.1 | 16.5 | 20.3 | 18.7 | 0.06 | 21.8 | 17.6 | 17.4 | 11.9 | 0.0005 |
Phys activity, mod-heavy | ||||||||||
Mean ± SE MET-h/wk | 5143 ± 374 | 5343 ± 362 | 5547 ± 366 | 5446 ± 389 | 0.35 | 6193 ± 377 | 4991 ± 365 | 4673 ± 378 | 5386 ± 387 | 0.08 |
SBP, mean ± SE mmHg | 128 ± 1.4 | 126 ± 1.3 | 125 ± 1.3 | 125 ± 1.4 | 0.02 | 127 ± 1.4 | 124 ± 1.3 | 127 ± 1.4 | 126 ± 1.4 | 0.83 |
Total : HDL cholesterol | 4.3 ± 0.08 | 4.3 ± 0.08 | 4.2 ± 0.08 | 4.2 ± 0.09 | 0.16 | 4.3 ± 0.09 | 4.3 ± 0.08 | 4.2 ± 0.09 | 4.2 ± 0.09 | 0.31 |
Cholesterol medication, % | 25.7 | 15 | 9.6 | 5.9 | <.0001 | 9.1 | 14.4 | 18.9 | 18.2 | 0.001 |
Diabetic, % | 15.5 | 17 | 15.6 | 15.5 | 0.43 | 15.5 | 17 | 15.6 | 15.6 | 0.86 |
Total energy, mean | ||||||||||
kcal/day | 1591 ± 53 | 1616 ± 52 | 1640 ± 52 | 1592 ± 56 | 0.87 | 1797 ± 54 | 1539±52 | 1546 ± 54 | 1512 ± 55 | <.0001 |
Alcohol, avg drinks/wk | 4.6 | 3.7 | 3.3 | 2.8 | 0.02 | 3.1 | 3.7 | 4 | 3.9 | 0.35 |
Saturated fat, %kcal | 18.9 | 20.3 | 20.5 | 20.3 | 0.09 | 11.2 | 10.8 | 10.4 | 9.8 | <.0001 |
Fruits, avg srv/day | 1.7 | 1.5 | 1.5 | 1.6 | 0.24 | 1.6 | 1.5 | 1.6 | 1.8 | 0.05 |
Cruciferous vegetables, savg srv/day | 0.35 | 0.37 | 0.4 | 0.37 | 0.35 | 0.33 | 0.37 | 0.38 | 0.44 | 0.002 |
‡Adjusted for age, gender, race/ethnicity, education, and site.
As our previous work has suggested a gender difference in the association between omega-3 fatty acids and endothelial function [
Characteristics of participants according to plasma phospholipid omega-6 fatty acid LA and omega-3 fatty acid DHA quartiles are presented in Table
After adjustment for age, gender, race/ethnicity, and site, plasma phospholipid AA showed no association with LV mass, LV mass/volume ratio, EF, stroke volume, or aortic distensibility as assessed by CMR. These results were unchanged after adjustment for additional covariates (Table
Relation between plasma phospholipid omega-6 fatty acid levels and measures of cardiac structure and function.‡,*
Measure | Plasma AA ( | Plasma LA ( | ||||||||
Q1 | Q2 | Q3 | Q4 | Q1 | Q2 | Q3 | Q4 | |||
trend | trend | |||||||||
Aortic distensibility, 10(−3) mmHg(−1) | 1.25 | 1.25 | 1.22 | 1.23 | 0.4 | 1.23 | 1.25 | 1.24 | 1.23 | 0.98 |
LV mass, grams | 143.3 | 144.4 | 145.8 | 143.8 | 0.62 | 144.2 | 144.4 | 145.7 | 142.8 | 0.69 |
LV mass : volume | 0.12 | 0.11 | 0.11 | 0.14 | 0.64 | 0.13 | 0.13 | 0.12 | 0.12 | 0.70 |
Ejection fraction, % | 68.32 | 68.05 | 68.01 | 68.61 | 0.67 | 68.99 | 68.66 | 67.52 | 68.01 | 0.02 |
Stroke volume, mL | 84.92 | 86.10 | 86.82 | 83.86 | 0.64 | 86.30 | 85.32 | 85.57 | 84.27 | 0.20 |
‡Multivariate linear regression analysis with adjustment for age, gender, race/ethnicity, BMI, current smoking status, education level, site, and total : HDL cholesterol, systolic blood pressure, and total energy intake.
*AA, arachidonic acid; LA, linoleic acid; LV, left ventricular.
§Sample sizes are for each plasma phospholipid quartile associated with LV mass, LV mass/volume ratio, ejection fraction, and stroke volume.
Sample sizes for each plasma phospholipid quartile associated with aortic distensibility varied slightly due to fewer available aortic distensibility measures and were as follows for AA: Q1 = 204, Q2 = 232, Q3 = 140, Q4 = 238 and for LA: Q1 = 255, Q2 = 227, Q3 = 223, Q4 = 209.
After adjustment for age, gender, race/ethnicity, and site, plasma phospholipid DHA showed no significant association with LV mass, LV mass/volume ratio, EF, stroke volume, or aortic distensibility in either minimally or fully adjusted models (Table
Measure | Plasma EPA ( | Plasma DHA ( | ||||||||
Q1 | Q2 | Q3 | Q4 | Q1 | Q2 | Q3 | Q4 | |||
trend | trend | |||||||||
Aortic distensibility, 10(−3) mmHg(−1) | 1.22 | 1.26 | 1.23 | 1.26 | 0.26 | 1.23 | 1.22 | 1.26 | 1.26 | 0.17 |
LV mass, grams | 143.14 | 143.61 | 146.48 | 145.38 | 0.14 | 144.14 | 142.19 | 145.45 | 147.19 | 0.07 |
LV mass : volume | 0.12 | 0.12 | 0.14 | 0.12 | 0.31 | 0.12 | 0.12 | 0.13 | 0.14 | 0.21 |
Ejection fraction, % | 67.94 | 68.50 | 68.22 | 68.68 | 0.26 | 68.13 | 68.82 | 68.35 | 68.53 | 0.29 |
Stroke volume, mL | 84.52 | 85.93 | 85.15 | 87.04 | 0.12 | 84.67 | 85.56 | 85.47 | 86.55 | 0.27 |
‡Multivariate linear regression analysis with adjustment for age, gender, race/ethnicity, BMI, current smoking status, education level, site, and total:HDL cholesterol, systolic blood pressure, and total energy intake.
*EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; ALA, alpha-linolenic acid; LV, left ventricular.
§Sample sizes are for each plasma phospholipid quartile associated with LV mass, LV mass/volume ratio, ejection fraction, and stroke volume.
Sample sizes for each plasma phospholipid quartile associated with aortic distensibility varied slightly due to fewer available aortic distensibility
measures and were as follows for EPA, Q1 = 225, Q2 = 212, Q3 = 245, Q4 = 232 and for DHA, QQ = 213, Q2 = 239, Q3 = 220, Q4 = 242.
Measure | Plasma ALA ( | ||||
Q1 | Q2 | Q3 | Q4 | ||
Trend | |||||
Aortic distensibility, 10(−3) mmHg(−1) | 1.23 | 1.26 | 1.21 | 1.24 | 0.59 |
LV mass, grams | 143.41 | 144.87 | 145.19 | 144.67 | 0.48 |
LV mass : volume | 0.13 | 0.13 | 0.11 | 0.12 | 0.40 |
Ejection fraction, % | 68.18 | 68.53 | 68.20 | 68.07 | 0.73 |
Stroke volume, mL | 84.20 | 85.40 | 87.46 | 85.37 | 0.18 |
‡Multivariate linear regression analysis with adjustment for age, gender, race/ethnicity, BMI, current smoking status, education level, site, and total:HDL cholesterol, systolicblood pressure, and total energy intake.
*ALA, alpha-linolenic acid; LV, left ventricular.
§Sample sizes are for each plasma phospholipid quartile associated with LV mass, LV mass/volume ratio, ejection fraction, and stroke volume. Sample sizes for each ALA quartile associated with aortic distensibility varied slightly due to fewer available aortic distensibility measures and were as follows: Q1 = 258, Q2 = 226, Q3 = 199, Q4 = 231.
When analyses were stratified by gender, a trend for a positive association was noted between plasma phospholipid DHA and LV mass in women (
Mean plasma phospholipid fatty acid quartile values, by gender. ‡
Plasma fatty acid | Men | Women | ||||||
Q1 | Q2 | Q3 | Q4 | Q1 | Q2 | Q3 | Q4 | |
AA | 8.95 ± 0.18 | 10.70 ± 0.07 | 12.39 ± 0.07 | 15.21 ± 0.20 | 8.99 ± 0.17 | 10.61 ± 0.07 | 12.28 ± 0.07 | 15.18 ± 0.20 |
LA | 17.46 ± 0.24 | 20.65 ± 0.10 | 22.74 ± 0.09 | 25.59 ± 0.31 | 17.37 ± 0.24 | 20.63 ± 0.09 | 22.70 ± 0.09 | 25.21 ± 0.29 |
EPA | 0.4 ± 0.01 | 0.62 ± 0.01 | 0.95 ± 0.02 | 2.02 ± 0.25 | 0.41 ± 0.01 | 0.63 ± 0.01 | 0.93 ± 0.02 | 2.03 ± 0.25 |
DHA | 2.54 ± 0.06 | 3.69 ± 0.04 | 4.76 ± 0.05 | 6.06 ± 0.22 | 2.64 ± 0.06 | 3.67 ± 0.04 | 4.77 ± 0.05 | 6.4 ± 0.22 |
ALA | 0.10 ± 0.01 | 0.15 ± 0.01 | 0.19 ± 0.01 | 0.29 ± 0.02 | 0.10 ± 0.01 | 0.15 ± 0.01 | 0.19 ± 0.01 | 0.27 ± 0.02 |
‡Mean ± SE, expressed as %total plasma phospholipid fatty acids and adjusted for age, race/ethnicity, BMI, current smoking status, education level, site, total:HDL cholesterol, systolic blood pressure, and total energy intake.
*AA, arachidonic acid; LA, linoleic acid; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; ALA, alpha-linolenic acid.
Relations between plasma phospholipid docosahexaenoic acid (DHA) levels, left ventricular (LV) mass, and ejection fraction, by gender. Results shown are least squares means ± SD after adjustment for age, race/ethnicity, BMI, current smoking status, education level, site, total : HDL cholesterol, systolic blood pressure, and total energy intake. Concentrations of plasma phospholipid fatty acid levels (expressed as percentages of total fatty acids) including DHA were summed and ranked into quartiles from lowest to highest based on sample range. Significant gender × DHA interactions were noted for both LV mass (
Plasma phospholipid polyunsaturated fatty acids are a marker of dietary patterns, with higher proportions generally reflecting greater consumption of polyunsaturated fats and less cardiovascular disease. These data from 1,274 healthy, multiethnic volunteers corroborate this concept by showing that men within higher plasma phospholipid DHA quartiles (with quartiles two through four corresponding to average intake of one or more servings of nonfried fish per week, as compared with food frequency questionnaire data and as supported by others [
The beneficial effects of dietary EPA and DHA in prevention of sudden cardiac death, acute coronary syndrome, and heart failure are well documented [
In contrast to the results in men, the observed positive association between plasma phospholipid DHA and increased LV mass among women within in this cohort initially appears surprising. However, there remains considerable discrepancy in the literature regarding gender differences in the association between PUFAs and cardiovascular outcomes. Among population-based cohorts, the nurses’ health study demonstrated a significantly negative association between fish consumption and incidence of CHD, and the ARIC study showed an inverse relationship in plasma phospholipid omega-3 fatty acids (particularly DHA) and incident heart failure in women [
Endothelial dysfunction is an independent predictor of cardiovascular events [
The strengths of this study include the ethnically diverse population, the inclusion of both men and women, and the availability of objective plasma phospholipid fatty acid measures as well as CMR measures of cardiovascular structure and function. The limitations include the observational cross-sectional study design with its known inability to infer causation and potential for temporal bias, and like any observational study, there remains the possibility of residual confounding for factors (such as health status) that could have meaningful impact on the observed results. It is also recognized that plasma phospholipid fatty acid measures are reflective of dietary intake over weeks, hence may not be representative of long-term intake [
In conclusion, this study is the first to examine the relationship between plasma phospholipid fatty acid levels and CMR measures of cardiovascular structure and function. Within this cohort of mixed ethnic/racial groups without known coronary artery disease, a gender difference was suggested in the association between plasma phospholipid DHA and ejection fraction: a positive association between DHA and ejection fraction was observed in men but not women, and a trend for a gender × DHA interaction was noted. In addition, a positive association between plasma phospholipid DHA and LV mass was found among women but not men, and a statistically significant gender × DHA interaction was observed. Additional research is warranted to clarify the effects of omega-3 fatty acids on cardiac structure and function in women versus men.
This research was supported by the National Institutes of Health, National Heart, Lung, and Blood Institute T32 training grant (no. 2 T32 HL 076132-06 A1, to J. S. Anderson), National Institute of Diabetes and Digestive and Kidney Diseases K01 (no. 5K01DK082729-02, to J. A. Nettleton) and Contracts nos.N01-HC-95159 through N01-HC-95169 from the National Heart, Lung, and Blood Institute. The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at
The authors declare that there is no conflict of interests.
Timothy Morgan provided statistical guidance for this paper.