BACKGROUND:
A protective effect of a small amount of fish on coronary heart disease (CHD) mortality has been established in observational studies in middle-aged people. In the present study this association was investigated in the elderly.

METHODS:
In 1971 CHD risk factors were measured in 272 people born before 1907. They belonged to a general practice in Rotterdam, the Netherlands and were followed for 17 years. The cross-check dietary history method was used to obtain information on fish consumption.

RESULTS:
During the follow-up period 58 people died from CHD, 67 from cancer and 187 from all causes. In 1971 about 60% of the elderly ate fish and 40% did not eat fish. Multivariate Cox proportional hazards analyses, taking the confounding effect of major risk factors into account, showed an inverse relation between fish consumption and 17-year CHD mortality. The risk ratio (RR) for fish eaters compared with no-fish eaters was significantly different from unity (RR = 0.51, 95% CI: 0.29-0.89). Cancer and total mortality were not related to fish consumption.

CONCLUSION:
The results from the present study suggest that the protective effect of a small amount of fish in relation to CHD observed in middle-aged people seems also to be present in the elderly.