Assessing Dietary Intake Patterns Through Cluster Analysis Among Adolescents in Selected Districts of Bihar and Assam From India : A Cross-Sectional Survey

Background: In the recent decade, dietary pattern assessment has evolved as a promising tool to describe the whole diet and represent inter-correlations between different dietary components. We aimed to derive the dietary patterns of adolescents (10–19 years) using cluster analysis on food groups and evaluate these patterns according to their socio-demographic profile. Methods: This community-based cross-sectional study was conducted in two districts, each from Bihar and Assam in India. Adolescents (10–19 years) were enrolled from both rural and urban areas. The dietary intake was assessed through a pre-validated single food frequency questionnaire. Cluster analysis was performed by a 2-step procedure to explore dietary patterns, pre-fixed at 2 clusters. Clusters were analyzed with respect to socio-demographic characteristics using binomial logistic regression. Results: A total of 826 girls and 811 boys were enrolled in the study. We found two major dietary patterns, namely a low- and high-mixed diet. The low-mixed diet (76.5% prevalence) had daily consumption of green vegetables, including leafy vegetables, with less frequent consumption of other foods. The high-mixed diet (23.5% prevalence) had more frequent consumption of chicken, meat, egg, and milk/curd apart from green vegetables. Adolescent boys had 3.6 times higher odds of consuming a low-mixed diet compared to girls. Similarly, adolescents with lower education grades and from marginalized social classes had two times higher odds of taking a low-mixed diet than their respective counterparts. Conclusions: The high consumption of a low-mixed diet and relatively less milk consumption limit the comprehensive growth of adolescents. Improvement in dietary intake of adolescents from marginalized sections of society can prove to be an important deterrent in mitigating India's nutritional challenges