Identification of Dietary Patterns Related to Metabolic Diseases and Their Association with Cardiovascular Disease: From the Korean Genome and Epidemiology Study

Abstract

Using data from the community-based cohort of the Korean Genome and Epidemiology Study (KoGES), we evaluated the dietary patterns (DPs) related to metabolic diseases and their associations with the incidence of non-fatal cardiovascular disease (CVD). After excluding those with a history of CVD or cancer, we analyzed the data of 8352 subjects aged 40-69 years. Based on their daily intake of 26 food groups at baseline, the DPs of the subjects with metabolic diseases (n = 1679, 20.1%) were analyzed using principal component analysis. Due to regional differences in the effect of DPs on CVD, we performed analyses stratified by region. The association between DPs and the incidence of non-fatal CVD was evaluated by calculating the hazard ratio (HR) and 95% confidence interval (CI) using the Cox proportional hazards model. During the 12-year follow-up, the incidence of non-fatal CVD was 5.4 per 1000 person-years (n = 431). An animal-based DP made the greatest contribution to the total variance and was characterized by a high intake of pork, beef, chicken, fish, and shellfish. The effect of DP on CVD differed by region (industrial/rural regions, p < 0.05) and was dominant in industrial regions, irrespective of metabolic disease status. In industrial regions, subjects in the top quintile of DP had a 0.42-fold (95% CI = 0.24-0.74) lower risk of incident CVD than those in the bottom quintile, even after adjusting for various covariates. In addition, the risk of CVD was high in individuals with a history of metabolic disease in both regions (HR = 1.74, 95% CI = 1.24-2.43 in industrial regions; HR = 1.88, 95% CI = 1.42-2.48 in rural regions). DP and a history of metabolic diseases, but not their interaction, were independently associated with incident CVD. In our study, an animal-based DP related to metabolic disease was independently associated with incident CVD, and this effect was noticeable only in industrial regions.