At Odds: Concerns Raised by Using Odds Ratios for Continuous or Common Dichotomous Outcomes in Research on Physical Activity and Obesity

Abstract

Purpose Research on obesity and the built environment has often featured logistic regression and the corresponding parameter, the odds ratio. Use of odds ratios for common outcomes such obesity may unnecessarily hinder the validity, interpretation, and communication of research findings. Methods We identified three key issues raised by the use of odds ratios, illustrating them with data on walkability and body mass index from a study of 13,102 New York City residents. Results First, dichotomization of continuous measures such as body mass index discards theoretically relevant information, reduces statistical power, and amplifies measurement error. Second, odds ratios are systematically higher (further from the null) than prevalence ratios; this inflation is trivial for rare outcomes, but substantial for common outcomes like obesity. Third, odds ratios can lead to incorrect conclusions during tests of interactions. The odds ratio in a particular subgroup might higher simply because the outcome is more common (and the odds ratio inflated) compared with other subgroups. Conclusion Our recommendations are to take full advantage of continuous outcome data when feasible and to use prevalence ratios in place of odds ratios for common dichotomous outcomes. When odds ratios must be used, authors should document outcome prevalence across exposure groups.

Dietary Variety Is Inversely Associated with Body Adiposity among US Adults Using a Novel Food Diversity Index

“…However, concordance between BMI and excess adiposity is greater once BMI reaches a threshold of 24 in women and 28 in men (46), suggesting that using existing cut-points may bias associations by incorrectly misclassifying more adults as normal weight. As a result, it may be more valid to examine BMI continuously because the associations with both overweight and obesity are likely attenuated (47).…”

A Prospective Examination of the Mechanisms Linking Childhood Physical Abuse to Body Mass Index in Adulthood

“…Thus, the risk of higher BMI for poorer health outcomes among obese individuals may be missed with BMI categorization (Papalia et al, 2010). Furthermore, categorization contributes to loss of power in the analyses and increases type II error (Jolliffe, 2004; Lovasi et al, 2012). …”

Detectable Blood Lead Level and Body Size in Early Childhood

“…Log-binomial regression was used instead of logistic regression as the prevalence of overweight/obesity was not rare (i.e. >10%) [17]. Models were fit (1) unadjusted and (2) adjusted for race, sex and birth weight Z-score.…”