Dietary questions as determinants of mortality: the OXCHECK experience

Objective: To determine whether responses to simple dietary questions are associated with specific causes of death. Design: Self-reported frequency intakes of various classes of foods and data on confounding factors were collected at the baseline survey. Death notifications up to 31 December 1997 were ascertained from the Office for National Statistics. Relative risk (RR) of death and 95% confidence intervals (CI) associated with baseline dietary factors were calculated by Cox regression. Setting: Prospective follow-up study based on five UK general practices. Subjects: Data were used from 11 090 men and women aged 35-64 years (81% of the eligible patient population) who responded to a postal questionnaire in 1989. Results: After 9 years of follow-up, 598 deaths were recorded, 514 of these among the 10 522 subjects with no previous history of angina. All-cause mortality was positively associated with age, smoking and low social class, as expected. Among the dietary variables, all-cause mortality was significantly reduced in participants who reported relatively high consumption of vegetables, puddings, cakes, biscuits and sweets, fresh or frozen red meat (but not processed meat), among those who reported using polyunsaturated spreads and among moderate alcohol drinkers. These associations were broadly similar for deaths from ischaemic heart disease (IHD), cancer and all other causes combined, and were not greatly attenuated by adjusting for potential confounding factors including social class. Conclusions: Responses to simple questions about nutrition were associated with mortality. These findings must be interpreted with caution since residual confounding by dietary and lifestyle factors may underlie the associations.