Intakes of fish and polyunsaturated fatty acids and mild-to-severe cognitive impairment risks: a dose-response meta-analysis of 21 cohort studies.

BACKGROUND: The intake of fish and polyunsaturated fatty acids (PUFAs) may benefit cognitive function. However, optimal intake recommendations for protection are unknown.

OBJECTIVE: We systematically investigated associations between fish and PUFA intake and mild-to-severe cognitive impairment risk.

DESIGN: Studies that reported risk estimates for mild cognitive impairment (MCI), cognitive decline, dementia, Alzheimer disease (AD), or Parkinson disease (PD) from fish, total PUFAs, total n-3 (ω-3) PUFAs, or at least one n-3 PUFA were included. Study characteristics and outcomes were extracted. The pooled RR was estimated with the use of a random-effects model meta-analysis. A dose-response analysis was conducted with the use of the 2-stage generalized least-squares trend program.

RESULTS: We included 21 studies (181,580 participants) with 4438 cases identified during follow-up periods (2.1-21 y). A 1-serving/wk increment of dietary fish was associated with lower risks of dementia (RR: 0.95; 95% CI: 0.90, 0.99; P = 0.042, I(2) = 63.4%) and AD (RR: 0.93; 95% CI: 0.90, 0.95; P = 0.003, I(2) = 74.8%). Pooled RRs of MCI and PD were 0.71 (95% CI: 0.59, 0.82; P = 0.733, I(2) = 0%) and 0.90 (95% CI: 0.80, 0.99; P = 0.221, I(2) = 33.7%), respectively, for an 8-g/d increment of PUFA intake. As an important source of marine n-3 PUFAs, a 0.1-g/d increment of dietary docosahexaenoic acid (DHA) intake was associated with lower risks of dementia (RR: 0.86; 95% CI: 0.76, 0.96; P < 0.001, I(2) = 92.7%) and AD (RR: 0.63; 95% CI: 0.51, 0.76; P < 0.001, I(2) = 94.5%). Significant curvilinear relations between fish consumption and risk of AD and between total PUFAs and risk of MCI (both P-nonlinearity < 0.001) were observed.

CONCLUSIONS: Fishery products are recommended as dietary sources and are associated with lower risk of cognitive impairment. Marine-derived DHA was associated with lower risk of dementia and AD but without a linear dose-response relation.