{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,26]],"date-time":"2026-03-26T12:42:04Z","timestamp":1774528924572,"version":"3.50.1"},"reference-count":45,"publisher":"Cambridge University Press (CUP)","issue":"1","license":[{"start":{"date-parts":[[2020,11,16]],"date-time":"2020-11-16T00:00:00Z","timestamp":1605484800000},"content-version":"unspecified","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["J. Clin. Trans. Sci."],"published-print":{"date-parts":[[2021]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec id=\"S2059866120005567_as1\"><jats:title>Introduction:<\/jats:title><jats:p>Identifying predictors of patient outcomes evaluated over time may require modeling interactions among variables while addressing within-subject correlation. Generalized linear mixed models (GLMMs) and generalized estimating equations (GEEs) address within-subject correlation, but identifying interactions can be difficult if not hypothesized<jats:italic>a priori<\/jats:italic>. We evaluate the performance of several variable selection approaches for clustered binary outcomes to provide guidance for choosing between the methods.<\/jats:p><\/jats:sec><jats:sec id=\"S2059866120005567_as2\"><jats:title>Methods:<\/jats:title><jats:p>We conducted simulations comparing stepwise selection, penalized GLMM, boosted GLMM, and boosted GEE for variable selection considering main effects and two-way interactions in data with repeatedly measured binary outcomes and evaluate a two-stage approach to reduce bias and error in parameter estimates. We compared these approaches in real data applications: hypothermia during surgery and treatment response in lupus nephritis.<\/jats:p><\/jats:sec><jats:sec id=\"S2059866120005567_as3\"><jats:title>Results:<\/jats:title><jats:p>Penalized and boosted approaches recovered correct predictors and interactions more frequently than stepwise selection. Penalized GLMM recovered correct predictors more often than boosting, but included many spurious predictors. Boosted GLMM yielded parsimonious models and identified correct predictors well at large sample and effect sizes, but required excessive computation time. Boosted GEE was computationally efficient and selected relatively parsimonious models, offering a compromise between computation and parsimony. The two-stage approach reduced the bias and error in regression parameters in all approaches.<\/jats:p><\/jats:sec><jats:sec id=\"S2059866120005567_as4\"><jats:title>Conclusion:<\/jats:title><jats:p>Penalized and boosted approaches are effective for variable selection in data with clustered binary outcomes. The two-stage approach reduces bias and error and should be applied regardless of method. We provide guidance for choosing the most appropriate method in real applications.<\/jats:p><\/jats:sec>","DOI":"10.1017\/cts.2020.556","type":"journal-article","created":{"date-parts":[[2020,11,16]],"date-time":"2020-11-16T08:04:50Z","timestamp":1605513890000},"source":"Crossref","is-referenced-by-count":3,"title":["Variable selection methods for identifying predictor interactions in data with repeatedly measured binary outcomes"],"prefix":"10.1017","volume":"5","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7124-5158","authenticated-orcid":false,"given":"Bethany J.","family":"Wolf","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yunyun","family":"Jiang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Sylvia H.","family":"Wilson","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jim C.","family":"Oates","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"56","published-online":{"date-parts":[[2020,11,16]]},"reference":[{"key":"S2059866120005567_ref12","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9469.2011.00740.x"},{"key":"S2059866120005567_ref2","doi-asserted-by":"publisher","DOI":"10.1534\/genetics.106.067348"},{"key":"S2059866120005567_ref16","doi-asserted-by":"crossref","first-page":"168","DOI":"10.3414\/ME11-02-0021","article-title":"Regularization for generalized additive mixture models by likelihood-based boosting","volume":"51","author":"Groll","year":"2012","journal-title":"Methods of Information in Medicine"},{"key":"S2059866120005567_ref24","doi-asserted-by":"publisher","DOI":"10.2307\/2983440"},{"key":"S2059866120005567_ref19","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9868.2011.00771.x"},{"key":"S2059866120005567_ref30","doi-asserted-by":"crossref","first-page":"275","DOI":"10.1002\/gepi.20459","article-title":"Screen and clean: A tool for identifying interactions in genome-wide association studies","volume":"34","author":"Wu","year":"2010","journal-title":"Genetic Epidemiology"},{"key":"S2059866120005567_ref42","doi-asserted-by":"publisher","DOI":"10.1007\/s00264-018-3967-y"},{"key":"S2059866120005567_ref21","first-page":"1822","volume-title":"A Stepwise Algorithm for Generalized Linear Mixed Models","author":"Neerchael","year":"2014"},{"key":"S2059866120005567_ref18","doi-asserted-by":"publisher","DOI":"10.1080\/10618600.2016.1247005"},{"key":"S2059866120005567_ref36","first-page":"1","article-title":"Fitting linear mixed-effects models using lme4","volume":"67","author":"Bates","year":"2015","journal-title":"Journal of Statistical Software, Articles"},{"key":"S2059866120005567_ref35","first-page":"1","article-title":"A complete classification of epistatic two-locus models","volume":"9","author":"Hallgrmsdttir","year":"2008","journal-title":"BMC Genetetics"},{"key":"S2059866120005567_ref23","unstructured":"23. Flom, P , Cassell, D. Stopping stepwise: Why stepwise and similar selection methods are bad, and what you should use. In: NESUG 2007 Proceedings, NorthEast SAS User Group, Baltimore, MD."},{"key":"S2059866120005567_ref26","doi-asserted-by":"publisher","DOI":"10.1111\/j.2044-8317.1992.tb00992.x"},{"key":"S2059866120005567_ref13","first-page":"771","article-title":"A short introduction to boosting","volume":"14","author":"Freund","year":"1999","journal-title":"Journal of Japanese Society for Artificial Intelligence"},{"key":"S2059866120005567_ref31","doi-asserted-by":"publisher","DOI":"10.1080\/00224065.1992.11979383"},{"key":"S2059866120005567_ref38","volume-title":"glmmLasso: Variable Selection for Generalized Linear Mixed Models by L1-Penalized Estimation","author":"Groll","year":"2016"},{"key":"S2059866120005567_ref41","doi-asserted-by":"publisher","DOI":"10.1002\/art.39623"},{"key":"S2059866120005567_ref39","volume-title":"threeboost: Thresholded Variable Selection and prediction Based on Estimating Equations","author":"Wolfson","year":"2014"},{"key":"S2059866120005567_ref8","doi-asserted-by":"publisher","DOI":"10.1007\/978-1-4757-3462-1"},{"key":"S2059866120005567_ref22","doi-asserted-by":"publisher","DOI":"10.1198\/0003130043277"},{"key":"S2059866120005567_ref3","doi-asserted-by":"publisher","DOI":"10.1038\/ejhg.2008.106"},{"key":"S2059866120005567_ref10","doi-asserted-by":"publisher","DOI":"10.1080\/00401706.1995.10484371"},{"key":"S2059866120005567_ref33","doi-asserted-by":"publisher","DOI":"10.1214\/13-AOS1096"},{"key":"S2059866120005567_ref40","doi-asserted-by":"publisher","DOI":"10.1002\/art.37940"},{"key":"S2059866120005567_ref7","volume-title":"Generalized Linear Mixed Models: Modern Concepts, Methods and Applications.","author":"Stroup","year":"2012"},{"key":"S2059866120005567_ref29","doi-asserted-by":"publisher","DOI":"10.1007\/s00180-012-0382-5"},{"key":"S2059866120005567_ref44","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9868.2005.00532.x"},{"key":"S2059866120005567_ref4","doi-asserted-by":"publisher","DOI":"10.3390\/ijerph5010004"},{"key":"S2059866120005567_ref17","doi-asserted-by":"publisher","DOI":"10.1198\/jasa.2011.tm10098"},{"key":"S2059866120005567_ref5","doi-asserted-by":"publisher","DOI":"10.1038\/nrg1407"},{"key":"S2059866120005567_ref25","first-page":"51","article-title":"Estimating the residual variance in orthogonal regression with variable selection","volume":"40","author":"Copas","year":"1991","journal-title":"Journal of the Royal Statistical Society Series D (The Statistician)"},{"key":"S2059866120005567_ref14","doi-asserted-by":"publisher","DOI":"10.1214\/aos\/1013203451"},{"key":"S2059866120005567_ref27","doi-asserted-by":"publisher","DOI":"10.1146\/annurev.es.21.110190.001021"},{"key":"S2059866120005567_ref9","doi-asserted-by":"publisher","DOI":"10.5705\/ss.2011.030a"},{"key":"S2059866120005567_ref43","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9868.2005.00503.x"},{"key":"S2059866120005567_ref15","doi-asserted-by":"publisher","DOI":"10.1198\/016214503000125"},{"key":"S2059866120005567_ref6","first-page":"119","article-title":"Strategy for detecting susceptibility genes with weak or no marginal effect","volume":"62","author":"Kotti","year":"2008","journal-title":"Human Heredity"},{"key":"S2059866120005567_ref20","unstructured":"20. Breiman, L. Bias, Variance, and Arcing Classifiers. Technical report, Statistics Department, University of California, Berkeley, 1996, pp. 1\u201322."},{"key":"S2059866120005567_ref32","doi-asserted-by":"publisher","DOI":"10.1214\/09-AOAS254"},{"key":"S2059866120005567_ref28","first-page":"477","article-title":"Boosting algorithms: Regularization, prediction and model fitting","volume":"22","author":"B\u00fchlmann","year":"2007","journal-title":"Sci"},{"key":"S2059866120005567_ref34","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.2014.881741"},{"key":"S2059866120005567_ref1","doi-asserted-by":"publisher","DOI":"10.1038\/nrg1578"},{"key":"S2059866120005567_ref37","volume-title":"GMMBoost: Likelihood-Based Boosting for Generalized Mixed Models","author":"Groll","year":"2013"},{"key":"S2059866120005567_ref11","doi-asserted-by":"publisher","DOI":"10.1016\/j.ygeno.2013.08.006"},{"key":"S2059866120005567_ref45","doi-asserted-by":"publisher","DOI":"10.1214\/07-AOS584"}],"container-title":["Journal of Clinical and Translational Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.cambridge.org\/core\/services\/aop-cambridge-core\/content\/view\/S2059866120005567","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,8,17]],"date-time":"2024-08-17T12:44:49Z","timestamp":1723898689000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.cambridge.org\/core\/product\/identifier\/S2059866120005567\/type\/journal_article"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,11,16]]},"references-count":45,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2021]]}},"alternative-id":["S2059866120005567"],"URL":"https:\/\/doi.org\/10.1017\/cts.2020.556","relation":{},"ISSN":["2059-8661"],"issn-type":[{"value":"2059-8661","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,11,16]]},"article-number":"e59"}}