Dietary pattern derived by reduced-rank regression and cardiovascular disease: A cross-sectional study

Abstract

Background and aims: Diet can affect cardiovascular health by changing lipid profiles or obesity levels. However, the association of dietary patterns reflecting lipid metabolism and adiposity measures with cardiovascular disease (CVD) is unclear. This study aimed to derive dietary patterns that explained variation in blood lipids and adiposity and investigate their associations with prevalent CVD.

Methods and results: A cross-sectional study was constructed in Beijing MJ Health Screening Center from 2008 to 2018. A dietary pattern was derived using reduced-rank regression among 75,159 participants without CVD. The dietary pattern explained the largest in predicting lipid profiles and adiposity measures. The dietary pattern was associated with a higher level of LDL-cholesterol and triglyceride, and high body mass index and waist circumference, but lower HDL-cholesterol. The dietary pattern was characterized by high intakes of staple food, red meat, processed food, fried food, edible offal, and less intakes of jam or honey, fruits, milk, and dairy products. Among 89,633 participants, we evaluated its association with prevalent CVD using multivariate logistic regression with adjustment for age, sex, annual income, education attainment, marital status, family history of CVD, smoking status, alcohol use, physical activity, and daily energy intake. Individuals with the highest quintile of dietary pattern score were 1%-38% more likely to have prevalent CVD than the lowest quintile (OR = 1.18, 95% CI = 1.01-1.38).

Conclusion: A diet pattern reflecting lipid profiles and obesity level was positively related to prevalent CVD, which could provide new insights in optimizing blood lipids and body shape for the prevention of CVD through dietary approaches among the Chinese population.

Keywords: Adiposity measures; Cardiovascular disease; Dietary pattern; Lipid metabolism; Reduced-rank regression.