The relationship between aging and depression is enigmatic. Epidemiologic studies using standardized diagnostic criteria often report a decreasing prevalence and incidence of depressive disorders with age, even within very old adult age groups. Recent interpretations of available research suggests that age differences in depressive symptoms suggests a lower burden among older adults, perhaps via risk reduction and improvements in emotional resiliency accompanying old age. However, other research has suggested that the burden of depressive symptoms increases with aging. An informed inquiry of how depression relates to age and aging requires longitudinal data collected in a representative sample using established instruments. A resource for furthering our understanding of the relationship between aging and depression is the Health and Retirement Survey (HRS) conducted by the University of Michigan, a large, representative, and longitudinal study that includes a depressive symptom inventory modeled after the Center for Epidemiologic Studies - Depression scale (CES-D). However, longitudinal research using the HRS/CES-D has been hampered because the response scale and symptom coverage varies across study wave.
The specific aims of this research project are: (1) to validate a scaling system, grounded in Item Response Theory, that places CES-D responses obtained using different response formats across HRS waves on an equivalent metric, (2) to examine CES-D responses for evidence of differential item functioning (DIF) and determine if patterns of DIF vary by response format, (3) to analyze the course of depressive symptoms in among aging adults, and (4) to archive depression values for HRS respondents for use by other investigators and publish the scoring algorithm for use by investigators in other studies using the CES-D. Successful completion of this research project will provide new data and understanding of the longitudinal trajectory of depressive symptoms with age among older Americans, and will lead to larger cross-study integration projects linking other studies using the CES-D -- even those different response scales -- or possibly other depression measures that provide some overlap of symptom coverage.
| Falba, Tracy A; Sindelar, Jody L; Gallo, William T (2009) Work expectations, realizations, and depression in older workers. J Ment Health Policy Econ 12:175-86 |
| Yang, Frances M; Jones, Richard N (2007) Center for Epidemiologic Studies-Depression Scale (CES-D) item response bias found with Mantel-Haenszel method was successfully replicated using latent variable modeling. J Clin Epidemiol 60:1195-200 |
| Jones, Richard N (2006) Identification of measurement differences between English and Spanish language versions of the Mini-Mental State Examination. Detecting differential item functioning using MIMIC modeling. Med Care 44:S124-33 |
| Jones, Richard N; Fonda, Stephanie J (2004) Use of an IRT-based latent variable model to link different forms of the CES-D from the Health and Retirement Study. Soc Psychiatry Psychiatr Epidemiol 39:828-35 |
