Quantifying the Association Between Disease Onset and Lifetime Under Cross-Sectional Sampling

• Activity Number: 292
• Type: Topic Contributed
• Date/Time: Tuesday, August 2, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Marco Carone*+ and Daniel O. Scharfstein and Masoud Asgharian
• Companies: University of California at Berkeley and The Johns Hopkins University and McGill University
• Keywords: cross-sectional sampling ; disease onset ; measures of association ; semiparametric model ; successive durations ; survival analysis

Abstract

Common measures of association usually quantify the departure of random variables from independence. Such measures often do not provide a scientifically meaningful interpretation when the variables are successive durations with a constrained sum, such as age at disease onset and time from onset to death. We propose a novel semiparametric model to quantify in a sensible manner the impact of disease on an individual's lifetime. A class of estimating equations for the parameters of this model are constructed and inference using data from a cross-sectional survey with longitudinal follow-up is discussed. Such data often arise in the study of the natural history of a disease. Because these data are subject to both systematic biases and loss to follow-up, their analysis is challenging. The asymptotic properties of the proposed inferential procedures are derived, and data from the Canadian Study of Health and Aging are analyzed to learn about the impact of dementia on the lifetime of elderly Canadians.