Predicting of Changes in Obesity Indices Regarding to Dietary Patterns in Longitudinal Tehran Lipid and Glucose Study

This study aimed at investigating the association between obesity indices (Body Mass Index (BMI), Waist Circumference (WC), Waist to Hip Ratio (WHR)) and dietary patterns derived from Reduced Rank Regression (RRR) analysis among Tehranian adults in a cohort study.

Over a duration of 6 years, 141 adults were examined before and after 6 years for obesity indices. Dietary intakes were recorded at baseline by two 24 hour dietary recalls. RRR analysis derived five patterns with respect to total fat, polyunsaturated-to-saturated fat ratio, cholesterol, fiber and calcium intakes. The calculated factor scores were categorized in quintiles and changes in obesity indices were scrutinized within these quintiles.

The first pattern (traditional) showed high loadings on refined carbohydrates, whole grain, vegetables, red and processed meat, saturated and trans fat, egg, showing significant high positive correlation with fat, cholesterol and calcium intakes (r=0.478, 0.626, and 0.486, respectively p<0.001). All obesity indices had an increasing trend across quintiles of pattern score (p<0.05 for BMI and WHR, p<0.001 for WC). The fifth pattern (egg) had high loading for eggs, salty snacks, and fruits, and negative loadings for red and processed meat, saturated and trans fat, plant oil, and dairy products, indicating increasing trends for WC (p<0.001) and WHR (p<0.05) after adjustment for potential confounders. Other patterns showed non-significant trends for obesity indices.

RRR analyses were indicative of two dominant dietary patterns correlated with increase in obesity indices in Tehran.