Multivariate probit linear mixed models for multivariate longitudinal binary data.
correlation matrix
generalized linear mixed models
heterogeneity
hypersphere decomposition
positive definiteness
Journal
Statistics in medicine
ISSN: 1097-0258
Titre abrégé: Stat Med
Pays: England
ID NLM: 8215016
Informations de publication
Date de publication:
15 Apr 2024
15 Apr 2024
Historique:
revised:
15
07
2023
received:
16
02
2023
accepted:
12
01
2024
medline:
18
3
2024
pubmed:
15
3
2024
entrez:
15
3
2024
Statut:
ppublish
Résumé
When analyzing multivariate longitudinal binary data, we estimate the effects on the responses of the covariates while accounting for three types of complex correlations present in the data. These include the correlations within separate responses over time, cross-correlations between different responses at different times, and correlations between different responses at each time point. The number of parameters thus increases quadratically with the dimension of the correlation matrix, making parameter estimation difficult; the estimated correlation matrix must also meet the positive definiteness constraint. The correlation matrix may additionally be heteroscedastic; however, the matrix structure is commonly considered to be homoscedastic and constrained, such as exchangeable or autoregressive with order one. These assumptions are overly strong, resulting in skewed estimates of the covariate effects on the responses. Hence, we propose probit linear mixed models for multivariate longitudinal binary data, where the correlation matrix is estimated using hypersphere decomposition instead of the strong assumptions noted above. Simulations and real examples are used to demonstrate the proposed methods. An open source R package, BayesMGLM, is made available on GitHub at https://github.com/kuojunglee/BayesMGLM/ with full documentation to produce the results.
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
1527-1548Subventions
Organisme : National Science Council
ID : MOST 111-2118-M-006-001-MY2
Organisme : National Research Foundation of Korea
ID : NRF-2022R1A2C1002752
Organisme : National Research Foundation of Korea
ID : NRF-2022R1F1A1062904
Organisme : National Research Foundation of Korea
ID : RS-2023-00217022
Organisme : National Research Foundation of Korea
ID : RS-2023-00240564
Informations de copyright
© 2024 John Wiley & Sons Ltd.
Références
Xu J, MacKenzie G. Modelling covariance structure in bivariate marginal models for longitudinal data. Biometrika. 2012;99:649-662.
Pourahmadi M. Joint mean-covariance models with applications to longitudinal data: unconstrained parameterisation. Biometrika. 1999;86(3):677-690.
Chiu T, Leonard T, Tsui K-W. The matrix-logarithmic covariance model. J Am Stat Assoc. 1996;91(433):198-210.
Feng S, Lian H, Xue L. A new nested Cholesky decomposition and estimation for the covariance matrix of bivariate longitudinal data. Comput Stat Data Anal. 2016;102:98-109.
Lee K, Cho H, Kwak M-n, Jang E. Estimation of covariance matrix of multivariate longitudinal data using modified Choleksky and hypersphere decompositions. Biometrics. 2020;76:75-86.
Rhee A, Kwak M-S, Lee K. Robust modeling of multivariate longitudinal data using modified Cholesky and hypersphere decompositions. Comput Stat Data Anal. 2022;170:107439.
Agresti A. A model for repeated measurements of a multivariate binary response. J Am Stat Assoc. 1997;92:315-321.
Gueorguieva RA. A multivariate generalized linear mixed model for joint modelling of clustered outcomes in the exponential family. Stat Model. 2001;1:177-193.
Ilk O, Daniels MJ. Marginalized transtion random effects models for multivariate longitudinal binary data. Can J Stat. 2007;35:105-123.
Ribaudo H, Thompson SG. The analysis of repeated multivariate binary quality of life data: a hierarchical model approach. Stat Methods Med Res. 2002;11:69-83.
Yang W, Chaganty N. A contrasting study of likelihood methods for the analysis of longitudinal binary data. Commun Stat Theory Methods. 2014;43(14):3027-3046.
Lin H, Chaganty N. Multivariate distributions of correlated binary variables generated by pair-copulas. J Stat Distrib Appl. 2021;8:1-14.
Lee K, Joo Y, Yoo JK, Lee J. Marginalized random effects models for multivariate longitudinal binary data. Stat Med. 2009;28:1284-1300.
Lee K, Daniels MJ, Joo Y. Flexible marginalized models for bivariate longitudinal ordinal data. Biostatistics. 2013;29:1-15.
Anderson JA, Pemberton JD. The grouped continuous model for multivariate ordered categorical variables and covariate adjustment. Biometrics. 1985;41(4):875-885.
Chib S, Greenberg E. Analysis of multivariate probit models. Biometrika. 1998;85:347-361.
Chib S, Carlin BP. On MCMC sampling in hierarchical longitudinal models. Stat Comput. 1999;9:17-26.
Czado C. Multivariate regression analysis of panel data with binary outcomes applied to unemployment data. Stat Pap. 2000;41:281-304.
Lee K-J, Kim C, Chen R-B, Lee K. Robust probit linear mixed models for longitudinal binary data. Biom J. 2022;64:1307-1324.
Galecki AT. General class of covariance structures for two or more repeated factors in longitudinal data analysis. Commun Stat Theory Method. 1994;23:3105-3119.
Ghaganty NT, Naik DN. Analysis of multivariate longitudinal data using quasi-least squares. J Stat Plan Inference. 2002;103:421-436.
Zhang W, Leng C, Tang CY. A joint modelling approach for longitudinal studies. J R Stat Soc Ser B. 2015;77:219-238.
Pan J, MacKenzie G. Regression models for covariance structures in longitudinal studies. Stat Model. 2006;6(1):43-57.
Daniels MJ, Zhao YD. Modelling the random effects covariance matrix in longitudinal data. Stat Med. 2003;22:1631-1647.
Herron MC. Postestimation uncertainty in limited dependent variable models. Politic Anal. 1999;8(1):83-98.
Saefken B, Kneib T, Waveren C-S, Greven S. A unifying approach to the estimation of the conditional Akaike information in generalized linear mixed models. Electron J Stat. 2014;8(1):201-225.
Vaida F, Blanchard S. Conditional Akaike information for mixed-effects models. Biometrika. 2005;92:351-370.
Geisser S, Eddy WF. A predictive approach to model selection. J Am Stat Assoc. 1979;74(365):153-160.
Piironen J, Vehtari A. Comparison of Bayesian predictive methods for model selection. Stat Comput. 2017;27(3):711-735.
Alexander CS, Markowitz R. Maternal employment and use of pediatric clinic services. Med Care. 1986;24:134-147.
Kim J, Kim E, Yi H, et al. Short-term incidence rate of hypertension in Korea middle-aged adults. J Hypertens. 2006;24:2177-2182.
Ji H, Kim A, Ebinger JE, et al. Sex differences in blood pressure trajectories over the life course. JAMA Cardiol. 2020;5(3):255-262.
Park K-Y, Park H-K, Hwang H-S. Relationship between abdominal obesity and alcohol drinking pattern in normal-weight, middle-aged adults: the Korea National Health and nutrition examination survey 2008-2013. Public Health Nutr. 2017;8(20):2192-2200.
Huang S, Li J, Shearer G, et al. Longitudinal study of alcohol consumption and HDL concentrations: a community-based study. Am J Clin Nutr. 2017;105(4):905-912.
Kim SK, Kim HC, Shim J-S, Kim DJ. Effects of cigarette smoking on blood lipids in Korean men: cardiovascular and metabolic diseases etiology research center cohort. Korean J Intern Med. 2020;35(2):369-382.