Evaluating functional covariate‐environment interactions in the Cox regression model

• Autores: Ling Zhou, Haoqi Li, Huazhen Lin, Peter X.-K. Song
• Localización: Canadian Journal of Statistics = Revue Canadienne de Statistique, ISSN 0319-5724, Vol. 47, Nº. 2, 2019, págs. 204-221
• Idioma: inglés
• DOI: 10.1002/cjs.11486
• Enlaces
  • Texto completo
• Resumen

  • AbstractENTHIS LINK GOES TO A ENGLISH SECTIONFRTHIS LINK GOES TO A FRENCH SECTION Children exposed to mixtures of endocrine disrupting compounds such as phthalates are at high risk of experiencing significant friction in their growth and sexual maturation. This article is primarily motivated by a study that aims to assess the toxicants‐modified effects of risk factors related to the hazards of early or delayed onset of puberty among children living in Mexico City. To address the hypothesis of potential nonlinear modification of covariate effects, we propose a new Cox regression model with multiple functional covariate‐environment interactions, which allows covariate effects to be altered nonlinearly by mixtures of exposed toxicants. This new class of models is rather flexible and includes many existing semiparametric Cox models as special cases. To achieve efficient estimation, we develop the global partial likelihood method of inference, in which we establish key large‐sample results, including estimation consistency, asymptotic normality, semiparametric efficiency and the generalized likelihood ratio test for both parameters and nonparametric functions. The proposed methodology is examined via simulation studies and applied to the analysis of the motivating data, where maternal exposures to phthalates during the third trimester of pregnancy are found to be important risk modifiers for the age of attaining the first stage of puberty. The Canadian Journal of Statistics 47: 204–221; 2019 © 2019 Statistical Society of Canada