A Two-Part Model to Assess the Impact of Coping Strategies on the Duration of Depression in Young Adults

• Activity Number: 424
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Statistics in Mental Health Research
• Author(s): Jun Yan*+ and Steven Chiou and Robert Aseltine and Elizabeth Schilling
• Companies: University of Connecticut and University of Connecticut and University of Connecticut Health Center and University of Connecticut Health Center
• Keywords: induced smoothing ; Buckley-James estimator

Abstract

As part of a longitudinal study of young adults in the greater Boston metropolitan area, participants were screened for major depression. A large portion of the subjects did not experience depression, resulting in a point mass of duration at zero. For those who did experience depression, the duration was right censored. Researchers were interested in how coping strategies on stress affect the duration of depression. We propose a two-part model to analyze the depression data. The probability of experiencing depression is modeled with logistic regression. The duration, if positive, is modeled with an accelerated failure time model. Both rank-based and least-squares based inferences are used for the accelerated failure time model with recently available computation methods. The effect of coping strategies on stress is assessed in both parts of the model.