Fish contains essential polyunsaturated fatty acids (n-3 PUFA) which increase n-3 PUFA concentrations in the cardiac membrane and influence cardio electrophysiology, which might have antiarrhythmic effects and so lower risk of fatal CHD.⁽ Reference London ^{1 ,} Reference Mozaffarian ^{2 )} Clinical trials of n-3 PUFA supplements conducted in high risk populations show no significant benefit,⁽ Reference Chowdhury ^{3 )} results from observational studies on fish intake show heterogeneous results.⁽ Reference Zheng ^{4 )} N-3 PUFA containing supplements, mainly cod liver oil, are widely used in the UK and by 24 % in the Norfolk-based European Prospective Investigation into Cancer (EPIC-Norfolk).⁽ Reference Lentjes ^{5 ,} Reference Lentjes ^{6 )} We studied the association between n-3 PUFA Total Nutrient Intake (TNI, i.e. intake from foods and supplements), n-3 PUFA supplement use and fatal CHD in a general population-based cohort.

EPIC-Norfolk recruited men and women, between 39–79 y (N = 25,639) between 1993–1997. Anthropometry was measured. Participants completed a 7-day diet diary, from which n-3 PUFA TNI, energy intake and disaggregated food consumption were determined. Participants were classified into three groups: non-supplement users (NSU), supplement users without n-3 PUFA supplements (SU-n3) and supplement users with n-3 PUFA supplements (SU+n3). General questionnaires ascertained social class, education, smoking, physical activity, alcohol consumption and prevalent diseases. Analyses were based on n = 22,137 with complete data. After a median follow-up of 18 years, 1393 participants died from CHD (ICD 410-414/I20-25). Cox proportional hazards regression was used to analyse differences between supplement groups as well as quintiles (Q5 v Q1) of TNI intake.

SU + n3 (compared to NSU) were more likely to be women, >60 years, and to be non-smokers and alcohol consumers. They reported fewer higher educational qualifications and less physical activity. SU + n3 and SU-n3 had lower self-reported history of myocardial infarction, diabetes or stroke. Differences in median (Med) n-3 sourced intake are shown in the top half of the table. SU + n3 did not have lower risk of fatal CHD; however higher n-3 PUFA intake was associated with a 22 % lower risk of fatal CHD, after adjusting for fish consumption, indicating that other sources than fish are associated with fatal CHD.

^* sex, age, smoking, body mass index, alcohol, social class, education, season, physical activity, energy intake, fruit, vegetables, red meat, processed meat, white meat, prevalent diabetes/stroke/myocardial infarction.

Non-fish n-3 PUFA was negatively associated with fatal CHD. The negative confounding observed from fish might be explained by preparation methods⁽ Reference Mozaffarian ^{7 )} or UK dietary patterns (fish ‘n chips); alternatively, contamination of fish with methylmercury might play a role.⁽ Reference Stern ^{8 )}