Table 1. Prospective studies that investigated the association of dietary and circulating EPA and/or DHA with risk of all-cause mortality.
| First author, year (Country/region) | Source of populations, duration | Participants | No. of deaths | Comparison | RR (95% CI) | Diet assessment | Adjustment for potential confounders | Quality scores |
|---|---|---|---|---|---|---|---|---|
| Dietary intake | ||||||||
| Folsom, 2004 (United States) | IWHS, 14yr | 41 836 F aged 55–69 yr | 4653 | Q5 vs. Q1 | 0.96 (0.86–1.06) | Self-reported FFQ | Age, BMI, WHR, education, physical activity, smoking, age at first live birth, estrogen use, vitamin use, diabetes, hypertension, and intake of energy, alcohol, whole grains, fruit and vegetables, red meat, cholesterol, and saturated fat. | 7 |
| Nagata, 2012 (Japan) | Takayama Study, 16 yr | 28 356 M/F aged ≥35 yr | 2499 M; 2117 F | Q5 vs. Q1 | 1.03 (0.86–1.23) (M) 1.04 (0.85–1.26) (F) | Self-reported FFQ, validated | Age, height, BMI, physical activity, smoking, education, marital status, histories of diabetes and hypertension, and intakes of energy, alcohol, protein, SFA, MUFA, non-long-chain n-3 PUFA, fruit, vegetables, fiber, and percent energy from carbohydrate in foods other than rice. | 9 |
| Takata, 2013 (China) | SMHS, 5.6 yr; SWHS, 11.2 yr | 61 137 M aged 40–74 yr; 73 159 F aged 40–70 yr | 2170 M; 3666 F | Q5 vs. Q1 | 0.79 (0.72–0.87) | Self-reported FFQ, validated | Age, income, occupation, education, comorbidity index, physical activity, smoking, and intakes of energy, alcohol (for M), red meat, poultry, fruit, and vegetable. | 8 |
| Miyagawa, 2014 (Japan) | NIPPON DATA80, 24 yr | 9190 M/F aged ≥30 yr | 2551 | Q5 vs. Q1 | 0.80 (0.66–0.96)a 0.97 (0.84–1.12)b | Food records | Age, sex, BMI, smoking, SBP, blood glucose, serum total cholesterol, antihypertensive medication status, residential area, intakes of alcohol SFA, n-6 PUFA, vegetable protein, total dietary fiber, and sodium. | 8 |
| Bell, 2014 (United States) | VITAL Study, 5.0 yr | 70 495 M/F aged 50–76 yr | 3051 | Q4 vs. Q1 | 0.84 (0.76–0.93) | Self-reported FFQ | Age, sex, BMI, smoking, race/ethnicity, marital status, education, physical activity, self-rated health, mammogram in, prostate- specific antigen test, sigmoidoscopy, uses of cholesterol-lowering medication, aspirin, non-aspirin NSAIDs, estrogen, and estrogen+progestin, morbidity score, age at menopause, age at death of father or mother, and intakes of total energy and energy from trans fat and SFA, alcohol, fruit, and vegetables, | 7 |
| Villegas, 2015 (United States) | SCCS, 5.5 yr | 77 100 M/F aged 40–79 yr | 6917 | Q5 vs. Q1 | 0.94 (0.86–1.03) | Self-reported FFQ, validated | Age, sex, BMI, smoking, physical activity, income, education, insurance coverage, race, and intakes of energy, alcohol, and total meat. | 8 |
| Circulating levels | ||||||||
| Warensjö, 2008 (Sweden) | ULSAM, 30.7yr | 1885 M aged 50 yr | 1012 | Per SDc increase | EPA: 1.00 (0.94–1.08) DHA: 0.95 (0.89–1.02) | Gas chromatography | Total cholesterol, BMI, smoking, physical activity, and hypertension. | 8 |
| Chien, 2013 (Taiwan) | Residents living in Chin-Shan Township, 9.6 yr | 1833 M/F aged >35 yr | 568 | Q4 vs. Q1 | EPA: 0.77 (0.59–1.00) DHA: 0.89 (0.68–1.18) | Gas chromatography | Age, sex, BMI, smoking, alcohol drinking, marital status, education, occupation, sports activity, hypertension, diabetes, LDL-C and HDL-C levels. | 8 |
| Mozaffarian, 2013 (United States) | CHS | 2692 M/F aged ≥65 yr | 1625 | Q5 vs. Q1 | EPA: 0.83 (0.71–0.98) DHA: 0.80 (0.67–0.94) | Gas chromatography | Age, sex, BMI, WC, physical activity, race, education, enrollment site, fatty acids measurement batch, smoking, prevalent diabetes, AF, and drug- treated hypertension, and intakes of alcohol, tuna or other broiled or baked fish, fried fish, red meat, fruit, vegetables, and dietary fiber. | 7 |
| Marklund, 2015 (Sweden) | Residents living in Stockholm County, 14.5 yr | 4232M/F aged 60 yr | 356 | Q4 vs. Q1d | EPA: 0.81 (0.72–0.91) DHA: 0.75 (0.68–0.84) | Gas chromatography | Sex, BMI, smoking, physical activity, education, alcohol intake, diabetes mellitus, drug-treated hypertension, and drug-treated hypercholesterolemia. | 8 |
AF, atrial fibrillation; BMI, body mass index; CHS, Cardiovascular Health Study; CVD, cardiovascular disease; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; F, females; FFQ, Food frequency questionnaire; HDL-C, high-density lipoprotein cholesterol; IWHS, Iowa Women’s Health Study; LDL-C, low-density lipoprotein cholesterol; M, males; MUFA, monounsaturated fatty acid; NSAIDs, non-aspirin nonsteroidal anti-inflammatory drugs; PUFA, polyunsaturated fatty acid; Q, quartile/quintile; SCCS, Southern Community Cohort Study; SBP, systolic blood pressure; SFA, saturated fatty acids; SMHS, Shanghai Men’s Health Study; SWHS, Shanghai Women’s Health Study; ULSAM, The Uppsala Longitudinal Study of Adult Men; WC, waist circumference; WHR, waist/hip ratio; yr, years.
aRisk estimates for total CVD mortality.
bRisk estimates for total non-CVD mortality.
cThe estimated SD was 0.52% for EPA and 0.19% for DHA, respectively.
dThis study also reported risk estimates for per SD (1% for EPA and 0.2% for DHA, respectively) increase in circulating long-chain n-3 fatty acids, and presented sex-specific results.