Associations of Fish Consumption With Risk of Cardiovascular Disease and Mortality Among Individuals With or Without Vascular Disease From 58 Countries

Abstract

Importance: Cohort studies report inconsistent associations between fish consumption, a major source of long-chain ω-3 fatty acids, and risk of cardiovascular disease (CVD) and mortality. Whether the associations vary between those with and those without vascular disease is unknown.

Objective: To examine whether the associations of fish consumption with risk of CVD or of mortality differ between individuals with and individuals without vascular disease.

Design, setting, and participants: This pooled analysis of individual participant data involved 191 558 individuals from 4 cohort studies-147 645 individuals (139 827 without CVD and 7818 with CVD) from 21 countries in the Prospective Urban Rural Epidemiology (PURE) study and 43 413 patients with vascular disease in 3 prospective studies from 40 countries. Adjusted hazard ratios (HRs) were calculated by multilevel Cox regression separately within each study and then pooled using random-effects meta-analysis. This analysis was conducted from January to June 2020.

Exposures: Fish consumption was recorded using validated food frequency questionnaires. In 1 of the cohorts with vascular disease, a separate qualitative food frequency questionnaire was used to assess intake of individual types of fish.

Main outcomes and measures: Mortality and major CVD events (including myocardial infarction, stroke, congestive heart failure, or sudden death).

Results: Overall, 191 558 participants with a mean (SD) age of 54.1 (8.0) years (91 666 [47.9%] male) were included in the present analysis. During 9.1 years of follow-up in PURE, compared with little or no fish intake (≤50 g/mo), an intake of 350 g/wk or more was not associated with risk of major CVD (HR, 0.95; 95% CI, 0.86-1.04) or total mortality (HR, 0.96; 0.88-1.05). By contrast, in the 3 cohorts of patients with vascular disease, the HR for risk of major CVD (HR, 0.84; 95% CI, 0.73-0.96) and total mortality (HR, 0.82; 95% CI, 0.74-0.91) was lowest with intakes of at least 175 g/wk (or approximately 2 servings/wk) compared with 50 g/mo or lower, with no further apparent decrease in HR with consumption of 350 g/wk or higher. Fish with higher amounts of ω-3 fatty acids were strongly associated with a lower risk of CVD (HR, 0.94; 95% CI, 0.92-0.97 per 5-g increment of intake), whereas other fish were neutral (collected in 1 cohort of patients with vascular disease). The association between fish intake and each outcome varied by CVD status, with a lower risk found among patients with vascular disease but not in general populations (for major CVD, I2 = 82.6 [P = .02]; for death, I2 = 90.8 [P = .001]).

Conclusions and relevance: Findings of this pooled analysis of 4 cohort studies indicated that a minimal fish intake of 175 g (approximately 2 servings) weekly is associated with lower risk of major CVD and mortality among patients with prior CVD but not in general populations. The consumption of fish (especially oily fish) should be evaluated in randomized trials of clinical outcomes among people with vascular disease.

Figure 1.. Fish Intake vs Risk of Composite of Death or Major Cardiovascular Disease (CVD) by Study and by Prior Cardiovascular Disease

^aAdjusted for age, sex, study center (random effect), body mass index, educational level, smoking status, physical activity, alcohol intake, urban vs rural location, history of diabetes, cardiovascular disease, cancer, use of statin or antihypertension medication, and intake of fruit, vegetables, red meat, poultry, dairy, and total energy.

Figure 2.. Associations Between Types of Fish (per 5-g Increment) and Clinical Events in the Outcome Reduction With Initial Glargine Intervention (ORIGIN) Trial (n = 12 422)

Data are adjusted for age, sex, study center (random effect), body mass index, educational level, smoking status, physical activity, alcohol intake, history of diabetes, cardiovascular disease (CVD), cancer, use of statin or antihypertension medication, and intake of fruit, vegetables, red meat, poultry, and dairy. Fish with highest ω-3 levels included herring, mackerel, sable, salmon, tuna (steak), and sardine. Other fish with high ω-3 levels included anchovy, bluefish, oyster, tuna (can), salmon (can), and trout. Fish with lowest ω-3 levels included bass, barramundi, bream, flathead, flounder, perch, snapper, octopus, sword fish, tile fish, and shark. Shellfish included crab, lobster, scallop, and mussel. HR represents hazard ratio.

Figure 3.. Mean Levels of Cardiovascular Risk Markers by Amount of Fish Intake in the Prospective Urban Rural Epidemiology (PURE) Trial (n = 147 541)

Data adjusted for age, sex, study center (random effect), body mass index, educational level, smoking status, physical activity, alcohol intake, urban vs rural location, history of diabetes, cardiovascular disease, cancer, use of statin or antihypertension medication, and intake of fruit, vegetables, red meat, poultry, dairy, and total energy. LDL-C represents low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol.