Epidemiological studies have shown that eating fish significantly reduces cardiovascular disease (CVD) incidence and mortality. However, more focused meta-analyses based on the most recent results from prospective cohort studies are needed. This systematic review and meta-analysis aims to update the association between fish intake and cardiovascular disease (CVD) risk using recent prospective studies. A systematic review and meta-analysis following the PRISMA guideline was conducted based on a random effects synthesis of multivariable-adjusted relative risks (RRs) of high vs. low categories of fish intake in relation to CVD incidence and mortality. Non-linear meta-regression was applied to investigate the shape of the association between fish intake and CVD risk. Sensitivity analysis and stratifications by type of CVD outcome, type of fish intake and type of cooking were performed. Based on 18 papers reporting 17 independent estimates of CVD risk (1,442,407 participants and 78,805 fatal and non-fatal CVD events), high vs. low intake of fish corresponded to about 8% reduced CVD risk (RR = 0.93 [0.88-0.98]). According to a non-linear dose-response meta-regression, 50 g of fish intake per day corresponded to a statistically significant 9% reduced fatal and non-fatal CVD risk (RR = 0.92 [0.90-0.95]). Similarly, fish intake in the range of a weekly intake of two to three portions of fish with a size of 150 g resulted in 8% fatal and non-fatal CVD risk reduction (RR = 0.93 [0.91-0.96]). The recommended two portions of fish a week reduces the risk of CVD outcomes by approximately 10%. A full portion of fish a day reduces CVD risk by up to 30%.
Keywords: cardiovascular mortality; cardiovascular risk; cohort study; fatty fish; fish; meta-analysis; meta-regression; nutrition; systematic review.