Objective: To evaluate the application of Multinomial Logistic Regression to assess the factors affecting the women to be underweight and overweight.
Methods: This was a cross-sectional study. The socio-demographic detail with height and weight of the women was recorded. The simple random sampling was adopted in the selection of women. The pregnant women were excluded from the study. Women's weight status, indicated by their BMI category, was used as the outcome variable in the analyses. A total of 435 women were interviewed. To assess the net effect of exposure variables on the outcome measures, multinomial logistic regression analysis was contemplated to be suitable as the outcome measure is polychotomous by nature.
Results: Most of the women were between 20-30 years of age (44.4%). More than one third of the women had family income between Rs. 5001-10, 000 (57.7%) per month and were illiterates (68.5%). The results of multinomial logistic regression showed that overweight increased with age and education with higher prevalence among urban women.
Conclusion: The growing demand which now appears before the Government or urban health planners is to address this rising urban epidemic with equal importance as given to other issues in the past.
Multinomial regression, Body mass index, underweight, Overweight.