Household Income as a Predictor of Body Mass Index Among Adults in Poland: A Multilevel Analysis

Abstract

The relationship between body mass index (BMI) and different predictors is very important aspect from the point of view of the whole society because the costs of preventing and fighting the overweight and obesity are borne by everyone. This study investigates the relationship important from the point of view of economists, i.e., between BMI and income household. This relationship was examined against other relationships. Linear multilevel models were fitted, with 18,534 adult Polish individuals nested in 9786 households and 66 subregions (Nomenclature of Territorial Units for Statistics 3—NUTS 3). We used data from the Social Diagnosis project (individual and household level) and from Statistics Poland (subregion level). We found that BMI was significantly associated with individual and household characteristics. Individuals in households from higher income categories (middle- and high-income) had higher values of BMI compared to low-income households. Two variables from subregion level (green areas and length of bicycle tracks) were related statistically insignificant with BMI. Just less than 0.5% of the unexplained variance was located on the subregion level. The study highlighted the importance of the household environment (household characteristics explain 14–18% of variance) which confirms that treating individuals as a part of households in which attitudes and behavior are shaped is a correct approach.