Ordinal regression model for analyzing factors affecting obesity levels in middle-aged adults in Indonesia

Abstract

Obesity is one of the major issues in many countries, including in Indonesia. In this study, the obesity category based on body mass index (BMI) is categorized into four levels, namely non-obese with BMI below 30 kg/m2, obesity I with BMI 30 – 34.9 kg/m2, obesity II with BMI 35 – 39.9 kg/m2, and obesity III with BMI above 40 kg/m2. The objective of this study was to assess factors affecting Indonesian obesity levels using ordinal logistic proportional odds models for each gender aged 40-65 years based on the fifth Indonesian Family Life Survey (IFLS 5) data. In totals of 4,658 male participants and 4,772 female participants were included in the analyses. The prevalence of severe obesity was associated with age, education, working status, and waist-hip circumference ratio for both men and women. With addition that the frequency of eating affects the prevalence of severe obesity for women. Men and women who had smaller waist-hip circumference ratio occurred to have higher probability of non-obese.