Spousal Influence on Body Weight: A Life Course Network Analysis

• In Event: Medical Sociology Section Roundtables
  In Refereed Roundtable (90 minute): Table 7. The Life Course and Health I

Abstract

Obesity is a socially embedded health condition that unfolds within intimate and family relationships across the life course. While prior research documents spousal resemblance in body weight, less is known about how much observed similarity reflects direct marital influence versus broader shared family contexts. Drawing on a life-course and linked-lives perspective, this study examines whether spousal overweight and obesity status predict individuals’ subsequent weight transitions after accounting for pre-marital family-network environments. Using longitudinal data from the Panel Study of Income Dynamics (1999–2021), we construct a family-network influence measure based on Graph Convolutional Networks (GCNs) to model structured kinship contexts prior to marriage. Logistic regression models with clustered standard errors show that having an overweight or obese spouse increases individuals’ subsequent risk of corresponding weight transitions by approximately 3–5 percentage points, net of demographic characteristics, prior weight status, and family-network influences. Although point estimates are slightly larger for male spouses, interaction tests indicate no statistically significant gender differences. These findings suggest that spousal influence on obesity persists beyond shared family-of-origin contexts and operates largely symmetrically across genders. By integrating family-network modeling with dyadic analysis, this study advances a more relational understanding of obesity and health inequality within marriage.

Authors

• Yujia Shen, The Pennsylvania State University

• Bing Han, Penn State

• Yubai Yuan, The Pennsylvania State University

• Tianxi Li, University of Minnesota

• Liying Luo, Pennsylvania State University