Geographically clustered dietary patterns associated with body mass index among nationally represented adults in...

Despite the rising prevalence of obesity, evidence on geographically specific dietary patterns and the potential impact of healthier dietary shifts remains limited. This study therefore aims to identify spatial clustering of dietary patterns and use substitution analysis to estimate the effect of a dietary pattern shift on body mass index (BMI).

Information from the Malaysian Adult Nutrition Survey (MANS) 2014 (n = 2574) were utilised. Dietary patterns were identified using principal component analysis of 17 food groups, and Bartlett scores were derived. Global and Local Moran's I with an egocentric distance of 8 km or five nearest neighbours assessed general and local clustering of the dietary pattern scores, respectively. Direct and substitutive associations between dietary patterns and BMI accounted for hierarchical data structure and socio-demography factors.

Three dietary patterns explaining 36.0% of the variance were identified and labelled as Western, Healthier, and Traditional diets. Significant spatial clustering was observed for all dietary patterns, with the Western diet showing the weakest clustering. Healthier diets predominated in higher-income urban areas, whereas the Traditional diet clustered in lower-income areas. Energy-adjusted Traditional diet score was positively associated with BMI (β = 0.40 kg/m²; p < 0.01), while substituting it with Healthier diet was inversely associated (β = -0.8 kg/m²; p < 0.05).

This study presents a novel application of substitution analysis on the association between dietary pattern and BMI across a multi-ethnic population. The findings reveal geographically distinct dietary patterns linked to BMI, underscoring the need for localised nutrition and obesity prevention strategies.