Exploration of model misspecification in latent class methods for longitudinal data: Correlation structure matters
Statistics in Medicine · 2023
DOI: 10.1002/sim.9730 ↗At a glance
Journal article in Statistics in Medicine, 2023. Written by Megan L Neely and 4 co-authors. Institution: Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, North Carolina, USA. Major MEDLINE topics: Software, Biomedical Research. Cited 3 times so far (Crossref); 56 references. The full text is freely available in PubMed Central. Funded by: NIA NIH HHS.
Abstract (original language)
Modeling longitudinal trajectories and identifying latent classes of trajectories is of great interest in biomedical research, and software to identify latent classes of such is readily available for latent class trajectory analysis (LCTA), growth mixture modeling (GMM) and covariance pattern mixture models (CPMM). In biomedical applications, the level of within‐person correlation is often non‐negligible, which can impact the model choice and interpretation. LCTA does not incorporate this correlation. GMM does so through random effects, while CPMM specifies a model for within‐class marginal covariance matrix. Previous work has investigated the impact of constraining covariance structures, both within and across classes, in GMMs—an approach often used to solve convergence problems. Using simulation, we focused specifically on how misspecification of the temporal correlation structure and strength, but correct variances, impacts class enumeration and parameter estimation under LCTA and CPMM. We found (1) even in the presence of weak correlation, LCTA often does not reproduce original classes, (2) CPMM performs well in class enumeration when the correct correlation structure is selected, and (3) regardless of misspecification of the correlation structure, both LCTA and CPMM give unbiased estimates of the class trajectory parameters when the within‐individual correlation is weak and the number of classes is correctly specified. However, the bias increases markedly when the correlation is moderate for LCTA and when the incorrect correlation structure is used for CPMM. This work highlights the importance of correlation alone in obtaining appropriate model interpretations and provides insight into model choice.
Subjects
MeSH terms (MEDLINE)
Author keywords
Authors and institutions
- Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, North Carolina, USA.
- Center on Aging and Human Development, Duke University Medical Center, Durham, North Carolina, USA.
- Department of Quantitative and Computational Biology, Dana and David Dornsife College of Letters, Arts and Sciences, University Southern California, Los Angeles, California, USA.
- Department of Psychological Sciences, Northern Arizona University, Flagstaff, Arizona, USA.
ORCID profiles: Megan L Neely
Funding
- NIA NIH HHS
Cite
APA
Neely, M. L., Pieper, C. F., Gu, B., Dmitrieva, N. O., & Pendergast, J. F. (2023). Exploration of model misspecification in latent class methods for longitudinal data: Correlation structure matters. Statistics in Medicine. https://doi.org/10.1002/sim.9730
Vancouver
Neely ML, Pieper CF, Gu B, Dmitrieva NO, Pendergast JF. Exploration of model misspecification in latent class methods for longitudinal data: Correlation structure matters. Statistics in Medicine. 2023. doi:10.1002/sim.9730
BibTeX
@article{Neely2023,
author = {Neely, Megan L and Pieper, Carl F and Gu, Bida and Dmitrieva, Natalia O and Pendergast, Jane F},
title = {Exploration of model misspecification in latent class methods for longitudinal data: Correlation structure matters},
journal = {Statistics in Medicine},
year = {2023},
doi = {10.1002/sim.9730}
}
EUVIMED