Assessing the Relationship Between Depression and Obesity Using Structural Equation Modeling

Abstract

<p> In this project we used structural equation modeling to analyze the data collected for the Canadian Community Health Survey (CCHS) Cycle 1.2 - Mental Health and Well-Being conducted by Statistics Canada. The data are cross-sectional.</p> <p> We looked at the relation between depression and obesity adjusting for gender, socioeconomic status, gene-environment interactions, eating and physical activity and stress.</p> <p> We used the AMOS and Mplus softwares to analyze our data. The first one used continuous variables for depression ("persistence of depression", in years) and obesity ("body mass index"-BMI), while the second used categorical variables: lifetime depression, 12 month depression and obesity (normal weight, overweight and obese). We also used two variables to measure different aspects of stress: self-perceived ability to handle an unexpected problem and work stress-social support.</p> <p> We fitted the models across the entire data, but also across different groups: males versus females and groups based on gender and BMI.</p> <p> The results indicated that the relationship between depression and obesity is different across gender.</p> <p> The limitations of the study are also discussed. </p>