ESTIMATING LIFE EXPECTANCY IN THE CANADIAN ELDERLY POPULATION WITH DEMENTIA USING PREVALENT COHORT SURVIVAL DATA

• 成果类型: Article
• 署名作者: Shariati, Ali; Asgharian, Masoud; Fakoor, Vahid
• 署名单位: Macquarie University; University of New South Wales Sydney; McGill University; Ferdowsi University Mashhad
• 刊物名称: ANNALS OF APPLIED STATISTICS
• ISSN/ISSBN: 1932-6157; 1941-7330
• DOI: 10.1214/25-AOAS2039
• 发表日期: 2025-09
• 页码: 2129-2154
• 关键词: Cross-sectional sampling; dementia; epidemiology; follow-up study; geriatrics; informative censoring; life expectancy; loss to follow-up; prevalent cohort; selection bias; survival analysis; alzheimers-disease; risk-factors; nonparametric-estimation; Empirical Likelihood; UNITED-STATES; education; HEALTH; mortality; diagnosis; BIAS

摘要

Dementia is globally one of the leading causes of death and the primarycause of dependency and disability in senior citizens. Life expectancy withdementia, defined as the average remaining lifespan for cases with dementia,is a key epidemiological concept in geriatrics. In spite of its significance formedical research and policy-making, this measure has not been studied forpeople with dementia in the Canadian population. We employ data from theCanadian Study of Health and Aging, a nationwide cross-sectional study ongeriatrics with follow-up for survival, to study life expectancy among elderlyCanadians with dementia. Even though practically more feasible, the col-lected survival data using such sampling mechanism suffer from two formsof bias: selection bias due to left truncation, also known as survivor bias,and bias owing to loss to follow-up. While the latter is often inevitable inlongitudinal studies when the subjects under study may drop out before theterminating event occurs, the former is a structural cross-sectional samplingbias occurring because long-term survivors are favoured by such a samplingmechanism. To the best of our knowledge, life expectancy and margins oferror under these two types of bias have not hitherto been studied in the lit-erature. Taking these complexities into account, we study the nonparametricmaximum likelihood estimator of age-specific life expectancy and its uni-form margins of error. Based on this estimator, we devise the first two-samplemethod for constructing uniform margins of error for the difference in lifeexpectancy between two groups of patients, which is then applied to scru-tinise the effects of various covariates. Our methodology enjoys robustnessand high efficiency while avoiding restrictive constraints. Simulation stud-ies are conducted to validate the performance of the proposed procedures.Our analysis provides novel information on the progression of the disease inCanada, revealing the pronounced effects of sex and type of dementia on lifeexpectancy. A comprehensive body of theoretical results, essential for pavingthe way for methodological development and beyond, is documented in theSupplementary Material.