A Binary Logistic Regression Model for Assessing Predisposing Factors of Obesity among Urban Traders in Benue State

Abstract

The study investigated the predisposing factors to obesity among adult urban traders in four selected towns in Benue State-Nigeria using binary logistic regression modeling procedure. The study utilized a cross -sectional design and a stratified random sampling technique using 
structured questionnaire to elicit responses from 400 selected traders considering their lifestyle, habits and socio-economic factors, anthropometric measurements and physical exercise. Sensitivity 
and specificity tests were also performed. Sensitivity and specificity tests results showed 99.2% and 97.9% respectively, indicating the extent to which the obese and non -obese cases were correctly classified. Results showed that increase in health challenges ( = 0.000), household income ( =
0.000), body weight ( = 0.018), parental history ( = 0.027), alcohol intake ( = 0.001), educational level ( = 0.034) were associated with the increased likelihood o f traders being obese, whereas increase in physical exercise ( = 0.000) was associated with the decreased likelihood of traders being obese. The prevalence of obesity was found to be 61.75% among the traders in the study area. The study recommends that promoting healthy dietary and weight management practices as well as encouraging regular physical exercise and other healthy life style changes like not drinking alcohol might be of great importance to the adult urban traders in the study area.