Dietary and lifestyle patterns identified through reduced rank regression and their association with insulin-related disorders: a prospective analysis from the Tehran Lipid and Glucose Study.

Abstract

Background: Since foods are consumed in combinations that also interact with other lifestyle variables such as body mass index(BMI) and physical activity, it is difficult to separate the role of single foods or a lifestyle variable alone in predicting the risk of chronic diseases such as metabolic disorders. Therefore, a suitable way to examine the combined effect of food consumption and its interaction with other lifestyle variables is to derive dietary patterns and lifestyle patterns using appropriate statistical methods. This study aimed to derive two dietary and lifestyle patterns related to hyperinsulinemia and insulin resistance(IR) using reduced rank regression(RRR) analysis.

Methods: The current study was conducted on 1063 individuals aged ≥ 25 years old of the Tehran Lipid and Glucose Study who have complete data on fasting blood sugar, plasma insulin, anthropometric variables, and nutritional intakes. Dietary intakes were collected using a food frequency questionnaire. Dietary and lifestyle patterns were identified via RRR analysis, using 34 food groups, BMI, smoking, and physical activity as predictor variables, and fasting serum insulin and HOMA-IR as response biomarkers.

Results: RRR derived a dietary pattern with a higher intake of processed meat, doogh, pickles, lemon juices, fish, and a lower intake of starchy vegetables, garlic and onion, dried fruits, nuts, red meat, dairy products, and coffee as predictive variables for IR and hyperinsulinemia. Also, RRR derived a lifestyle pattern based on the above-mentioned dietary pattern and high BMI as response variables. In the final adjusted model of cross-sectional analysis, the odds of hyperinsulinemia(OR:1.23,95%CI:1.08-1.41,P_trend=0.002) and IR(OR:1.52,95%CI:1.25-1.86,P_trend<0.001) were elevated with increasing each quartile of RRR-derived dietary pattern score. Also, a higher adherence to RRR-derived lifestyle pattern was associated with higher odds of hyperinsulinemia(OR:2.49,95%CI:2.14-2.88,P_trend<0.001) and IR(OR:3.20,95%CI:2.50-4.10,P_trend<0.001). Moreover, after three years of follow-up, the risk of hyperinsulinemia(OR:1.30,95%CI:1.08-1.56,P_trend=0.006) and IR(OR:1.26,95%CI:1.01-1.58,P_trend=0.037) incidence were increased per each quartile increase of the RRR-derived lifestyle pattern.

Conclusions: Our findings suggested that a dietary pattern and lifestyle with elevated BMI level, higher consumption of processed meat, doogh, pickles, lemon juices, and fish, and lower consumption of starchy vegetables, garlic and onion, dried fruits, nuts, red meat, dairy products, coffee may be associated with a higher risk of hyperinsulinemia and IR. It is suggested that further studies with a larger sample size and more extended follow-up duration, especially in other populations with different lifestyles and food habits be performed to confirm the findings of the current study.