Multilevel modelling for measuring interaction of effects between multiple categorical variables: An illustrative application using risk factors for preeclampsia
Paediatric and Perinatal Epidemiology · 2023-02 · 37 (2) · 154-164
DOI: 10.1111/ppe.12932 ↗Kurz gefasst
Fachartikel in der Zeitschrift Paediatric and Perinatal Epidemiology, online erschienen am 10.11.2022. Verfasst von Merida Rodriguez-Lopez und 3 weiteren Autor:innen. Beteiligte Einrichtung: Unit for Social Epidemiology, Faculty of Medicine, Lund University, Malmö, Sweden. Hauptthemen laut MEDLINE: Pre-Eclampsia. Die Arbeit wurde bisher 25-mal zitiert (Crossref) und nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Vetenskapsrådet.
Abstract (Originalsprache)
Background
Measuring multiple and higher‐order interaction effects between multiple categorical variables proves challenging.
Objectives
To illustrate a multilevel modelling approach to studying complex interactions.
Methods
We apply a two‐level random‐intercept linear regression to a binary outcome for individuals (level‐1) nested within strata (level‐2) defined by all observed combinations of multiple categorical exposure variables. As a pedagogic application, we analyse 36 strata defined by five risk factors of preeclampsia (parity, previous preeclampsia, chronic hypertension, multiple pregnancies, body mass index category) among 652,603 women in the Swedish Medical Birth Registry between 2002 and 2010.
Results
The absolute risk of preeclampsia was 4% but was predicted to vary from 1% to 44% across strata. The stratum discriminatory accuracy was 30% according to the variance partition coefficient (VPC) and 0.73 according to the area under the receiver operating characteristic curve (AUC). While the risk heterogeneity across strata was primarily due to the main effects of the categories defining the strata, 5% of the variation was attributable to their two‐ and higher‐way interaction effects. One stratum presented a positive interaction, and two strata presented negative interaction.
Conclusions
Multilevel modelling is an innovative tool for identifying and analysing higher‐order interaction effects. Further work is needed to explore how this approach can best be applied to making causal inferences.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Unit for Social Epidemiology, Faculty of Medicine, Lund University, Malmö, Sweden.
- Faculty of Health Sciences, Universidad Icesi, Cali, Colombia.
- Centre for Multilevel Modelling, University of Bristol, Bristol, UK.
- Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, Quebec, Canada.
- Center for Primary Health Care Research, Region Skåne, Malmö, Sweden.
ORCID-Profile: Merida Rodriguez-Lopez
Förderung
- Vetenskapsrådet
Zitieren
APA
Rodriguez-Lopez, M., Leckie, G., Kaufman, J. S., & Merlo, J. (2023). Multilevel modelling for measuring interaction of effects between multiple categorical variables: An illustrative application using risk factors for preeclampsia. Paediatric and Perinatal Epidemiology, 37(2), 154-164. https://doi.org/10.1111/ppe.12932
Vancouver
Rodriguez-Lopez M, Leckie G, Kaufman JS, Merlo J. Multilevel modelling for measuring interaction of effects between multiple categorical variables: An illustrative application using risk factors for preeclampsia. Paediatric and Perinatal Epidemiology. 2023;37(2):154-164. doi:10.1111/ppe.12932
BibTeX
@article{RodriguezLopez2023,
author = {Rodriguez-Lopez, Merida and Leckie, George and Kaufman, Jay S and Merlo, Juan},
title = {Multilevel modelling for measuring interaction of effects between multiple categorical variables: An illustrative application using risk factors for preeclampsia},
journal = {Paediatric and Perinatal Epidemiology},
year = {2023},
volume = {37},
number = {2},
pages = {154--164},
doi = {10.1111/ppe.12932}
}
EUVIMED