Defining dietary guidelines requires a quantitative assessment of the influence of diet on the development of diseases. The aim of the study was to investigate how dietary patterns were associated with mortality in a general population sample of Switzerland. We included 15,936 participants from two population-based studies (National Research Program 1A (NRP1A) and Monitoring of Trends and Determinants in Cardiovascular Disease (MONICA)—1977 to 1993) who fully answered a simplified 24-h dietary recall. Mortality data were available through anonymous record linkage with the Swiss National Cohort (follow-up of up to 37.9 years). Multiple correspondence analysis and hierarchical clustering were used to define data-driven qualitative dietary patterns. Mortality hazard ratios were calculated for all-cause, cancer and cardiovascular mortality using Cox regression. Two patterns were characterized by a low dietary variety (“Sausage and Vegetables”, “Meat and Salad”), two by a higher variety (“Traditional”, “High-fiber foods”) and one by a high fish intake (“Fish”). Males with unhealthy lifestyle (smokers, low physical activity and high alcohol intake) were overrepresented in the low-variety patterns and underrepresented in the high-variety and “Fish” patterns. In multivariable-adjusted models, the “Fish” (hazard ratio = 0.82, 95% CI (0.68–0.99)) and “High-fiber foods” (0.85 (0.72–1.00)) patterns were associated with lower cancer mortality. In men, the “Fish” (0.73 (0.55–0.97)) and “Traditional” (0.76 (0.59–0.98)) patterns were associated with lower cardiovascular mortality. In summary, our results support the notion that dietary patterns affect mortality and that these patterns strongly cluster with other health determinants.
Read full abstract