Development of Food Group Tree-Based Analysis and Its Association with Non-Alcoholic Fatty Liver Disease (NAFLD) and Co-Morbidities in a South Indian Population: A Large Case-Control Study

Abstract

Background: Non-alcoholic fatty liver disease (NAFLD) is a global problem growing in parallel to the epidemics of obesity and diabetes, with South Asians being particularly susceptible. Nutrition and behaviour are important modifiers of the disease; however, studies to date have only described dietary patterns and nutrients associated with susceptibility to NAFLD.

Methods: This cross-sectional case-control study included 993 NAFLD patients and 973 healthy controls from Trivandrum (India). Dietary data was collected using a locally validated food frequency questionnaire. A tree-based classification categorised 2165 ingredients into three levels (food groups, sub-types, and cooking methods) and intakes were associated with clinical outcomes.

Results: NAFLD patients had significantly higher consumption of refined rice, animal fat, red meat, refined sugar, and fried foods, and had lower consumption of vegetables, pulses, nuts, seeds, and milk compared to controls. The consumption of red meat, animal fat, nuts, and refined rice was positively associated with NAFLD diagnosis and the presence of fibrosis, whereas consumption of leafy vegetables, fruits, and dried pulses was negatively associated. Fried food consumption was positively associated with NAFLD, whilst boiled food consumption had a negative association. Increased consumption of animal fats was associated with diabetes, hypertension, and cardiovascular outcomes among those with NAFLD, whereas consumption of wholegrain rice was negatively associated with these clinical-related outcomes.

Conclusions: The tree-based approach provides the first comprehensive method of classifying food intakes to enable the identification of specific dietary factors associated with NAFLD and related clinical outcomes. This could inform culturally sensitive dietary guidelines to reduce risk of NAFLD development and/or its progression.

Keywords: NAFLD; South Asians; case-control; co-morbidities; dietary factors; food groups.

Figure 1

Representation of the food group levels based on the proposed food tree structure. Level 1 of the tree (inner circle) represents the 10 main food groups. Level 2 of the tree (middle circle) represents the 23 sub classifications related to each main food group in level 1. Level 3 (outer circle) represents the 8 cooking/processing methods related to sub classifications in level 2.

Figure 2

CIRCOS plot showing the association of individual food group levels (A-Level 1; B-Level 2; C-Level 3) with NAFLD status. Bars represent Odds ratio’s (ranging from 0.5–2) and 95% CIs from logistic regressions of unadjusted intakes, adjusted for age, gender and weight. (* p < 0.05; ** p < 0.001).

Figure 3

Heatmap showing positive and negative associations of the intakes of significant food groups with clinical outcomes amongst individuals with NAFLD. The axis represents the range of positive (red) and negative (blue) associations. p values adjusted for age, gender, and BMI (* p < 0.05; ** p < 0.001).