How to estimate the effect of proximity to fast food establishments on individuals’ BMI while accounting for possibly correlated neighborhood effects of residence and workplace?

James O’Malley of The Dartmouth Institute photographed in Hanover, New Hampshire on Tuesday, April 11, 2017. Copyright 2017 Robert C Strong II

The paper “Modeling a Bivariate Residential-Workplace Neighborhood Effect when Estimating the Effect of Proximity to Fast-Food Establishments on Body Mass Index” by James O’Malley and colleagues was published online in Statistics in Medicine on November 20, 2018 (https://onlinelibrary.wiley.com/doi/epdf/10.1002/sim.8039). This paper makes an important advance in the statistical methodology for hierarchical models by allowing the latent or random effects of a neighborhood to have a bivariate impact through both residential and workplace exposure to fast-food establishments on an individuals’ Body Mass Index (BMI). The paper solves a general gap at the intersection of statistical methodology and statistical computing that occurs when a clustering variable impacts outcomes through multiple possibly correlated forms of exposure. The research team was led by Dr. James O’Malley and included colleagues at Dartmouth, Harvard School of Public Health, Harvard Medical School and Harvard Pilgrim Health Care Institute, all co-authors on this paper.