Individual Health Trajectories: Population-Based Analysis of the Effects of Incidence and Recovery

Igor Akushevich, Duke University
Julia Kravchenko, Duke University

Lack of population-based analyses representative at a national level impedes better addressing health demands in the US elderly population. To understand age patterns of incidence and remission/recovery rates of aging-related diseases and their time trends, we analyzed individual histories of medical service use reconstructed from Medicare-linked datasets: the National Long Term Care Survey (NLTCS-M, 34,077 individuals followed-up for 5 years) and the Surveillance, Epidemiology and End Results (SEER-M, 2,154,598 individuals). Age, disability, and comorbidity patterns of incidence rates and time-after-diagnosis patterns of long-term remission/recovery rates were evaluated for cardio- and cerebrovascular diseases, most prevalent cancers, Parkinson's and Alzheimer's diseases, diabetes, and asthma. Decline in age patterns with age was detected for majority of diseases. Recovered individuals had higher survival and time trend in recovery rates were positive for all diseases except cancers. Estimates were validated using two Medicare datasets. Sensitivity analysis proved stability of evaluated rates.