Appropriate assessment of neighborhood effects on individual health: integrating random and fixed effects in multilevel logistic regression.

By K. Larsen, J. Merlo. American Journal of Epidemiology, 2005

977 citations recorded by the source corpus.

The logistic regression model is frequently used in epidemiologic studies, yielding odds ratio or relative risk interpretations. Inspired by the theory of linear normal models, the logistic regression model has been extended to allow for correlated responses by introducing random effects. However, the model does not inherit the interpretational features of the normal model. In this paper, the authors argue that the existing measures are unsatisfactory (and some of them are even improper) when quantifying results from multilevel logistic regression analyses. The authors suggest a measure of heterogeneity, the median odds ratio, that quantifies cluster heterogeneity and facilitates a direct comparison between covariate effects and the magnitude of heterogeneity in terms of well-known odds ratios. Quantifying cluster-level covariates in a meaningful way is a challenge in multilevel logistic regression. For this purpose, the authors propose an odds ratio measure, the interval odds ratio, that takes these difficulties into account. The authors demonstrate the two measures by investigating heterogeneity between neighborhoods and effects of neighborhood-level covariates in two examples--public physician visits and ischemic heart disease hospitalizations--using 1999 data on 11,312 men aged 45-85 years in Malmo, Sweden.

Subfields: Sociology and Political Science, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/9c28da59fbdbf625e3a7e8566aba708384b94304

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Sociology and Political Science

1. Peter C. Austin
2. M. C. S. Wong
3. M. Wang
4. J. Merlo
5. B. Chaix

In Statistics and Probability

1. Peter C. Austin
2. M. C. S. Wong
3. M. Wang
4. J. Merlo
5. B. Chaix
6. Elizabeth A. Stuart
7. N. Cuong
8. Uri Shalit

Citations in the corpus

• Happiness and Leisure Across Countries: Evidence from International Survey Data (2013, methodology): “…research has been widely employed in various fields such as education, health care and medicine, sociology, and geography (Paterson 1991; Hox and Kreft 1994; Jones and Duncan 1996; Langford et al. 1998; Hill et al. 2005; Larsen and Merlo 2005; Leyland and Goldstein 2001; Magnus et al. 2001).”
• A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena (2006, methodology): “However, for binary outcomes, the partition of variance between different levels does not have the intuitive interpretation of the linear model. Despite these difficulties several methods have been developed in logistic regression to obtain suitable epidemiological information on area level variance”
• Intermediate and advanced topics in multilevel logistic regression analysis (2017, methodology): “Merlo proposed the ‘sorting out index’ [15] or, using a more informative term, the POOR [10] as an alternative to the IOR as a measure of the magnitude of the effect of cluster variables.”