Hostname: page-component-8448b6f56d-sxzjt Total loading time: 0 Render date: 2024-04-24T14:00:04.663Z Has data issue: false hasContentIssue false
  • English
  • Français

Minimizing attrition bias: a longitudinal study of depressive symptoms in an elderly cohort

Published online by Cambridge University Press:  17 March 2009

Chung-Chou H. Chang ,
Hsiao-Ching Yang ,
Gong Tang  and
Mary Ganguli
Chung-Chou H. Chang*
Affiliation:
Department of Medicine, School of Medicine, University of Pittsburgh, Pittsburgh, U.S.A. Department of Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, U.S.A.
Hsiao-Ching Yang
Affiliation:
Department of Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, U.S.A.
Gong Tang
Affiliation:
Department of Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, U.S.A.
Mary Ganguli
Affiliation:
Department of Psychiatry, School of Medicine, University of Pittsburgh, Pittsburgh, U.S.A. Department of Neurology, School of Medicine, University of Pittsburgh, Pittsburgh, U.S.A. Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, U.S.A.
*
Correspondence should be addressed to Chung-Chou H. Chang, 200 Meyran Ave., Suite 200, Department of Medicine, University of Pittsburgh, Pittsburgh, PA 15213, U.S.A. Phone: +1 412 692 4868; Fax: +1 412 246 6954. Email: changj@pitt.edu.
Get access Rights & Permissions [Opens in a new window]

Abstract

Background: Attrition from mortality is common in longitudinal studies of the elderly. Ignoring the resulting non-response or missing data can bias study results.

Methods: 1260 elderly participants underwent biennial follow-up assessments over 10 years. Many missed one or more assessments over this period. We compared three statistical models to evaluate the impact of missing data on an analysis of depressive symptoms over time. The first analytic model (generalized mixed model) treated non-response as data missing at random. The other two models used shared parameter methods; each had different specifications for dropout but both jointly modeled both outcome and dropout through a common random effect.

Results: The presence of depressive symptoms was associated with being female, having less education, functional impairment, using more prescription drugs, and taking antidepressant drugs. In all three models, the same variables were significantly associated with depression and in the same direction. However, the strength of the associations differed widely between the generalized mixed model and the shared parameter models. Although the two shared parameter models had different assumptions about the dropout process, they yielded similar estimates for the outcome. One model fitted the data better, and the other was computationally faster.

Conclusions: Dropout does not occur randomly in longitudinal studies of the elderly. Thus, simply ignoring it can yield biased results. Shared parameter models are a powerful, flexible, and easily implemented tool for analyzing longitudinal data while minimizing bias due to nonrandom attrition.

Keywords

discrete failure time modeldropoutnon-ignorable nonresponseshared parameter modelWeibull model
Type
Research Article
Information
International Psychogeriatrics , Volume 21 , Issue 5 , October 2009 , pp. 869 - 878
Copyright
Copyright © International Psychogeriatric Association 2009

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

Andreescu, C., Chang, C. H., Mulsant, B. and Ganguli, M. (2008). Twelve-year depressive symptom trajectories and their predictors in a community sample of older adults. International Psychogeriatrics, 20, 221–236.CrossRefGoogle Scholar
Beunckens, C., Molenberghs, G. and Kenward, M. G. (2005). Direct likelihood analysis versus simple forms of imputation for missing data in randomized clinical trials. Clinical Trials, 2, 379–386.CrossRefGoogle ScholarPubMed
Casey, E., Hughes, J. W., Waechter, D., Josephson, R. and Rosneck, J. (2008). Depression predicts failure to complete phase-II cardiac rehabilitation. Journal of Behavioral Medicine, 31, 421–431.CrossRefGoogle ScholarPubMed
Eerola, M., Huurre, T. and Aro, H. (2005). The problem of attrition in a Finnish longitudinal survey on depression. European Journal of Epidemiology, 20, 113–120.CrossRefGoogle Scholar
Folstein, M. F., Folstein, S. E. and McHugh, P. R. (1975). “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189–198.CrossRefGoogle ScholarPubMed
Funatogawa, T., Funatogawa, I. and Takeuchi, M. (2008). An autoregressive linear mixed effects model for the analysis of longitudinal data which include dropouts and show profiles approaching asymptotes. Statistics in Medicine, 27, 6351–6366.CrossRefGoogle ScholarPubMed
Ganguli, M. et al. (1993). Sensitivity and specificity for dementia of population-based criteria for cognitive impairment: the MoVIES project. Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 48, M152–M161.Google ScholarPubMed
Ganguli, M. J., Gilby, J., Seaberg, E. and Belle, S. (1995). Depressive symptoms and associated factors in a rural elderly population: the MoVIES project. American Journal of Geriatric Psychiatry, 3, 144–160.CrossRefGoogle Scholar
Ganguli, M., Lytle, M. E., Reynolds, M. D. and Dodge, H. H. (1998). Random versus volunteer selection for a community-based study. Journal of Gerontology Series A: Biological Sciences and Medical Sciences, 53A, M39–M46.CrossRefGoogle Scholar
Ganguli, M., Dodge, H. H. and Mulsant, B. H. (2002). Rates and predictors of mortality in an aging, rural, community-based cohort: the role of depression. Archives of General Psychiatry, 59, 1046–1052.CrossRefGoogle Scholar
Gomeni, R., Lavergne, A. and Merlo-Pich, E. (2009). Modelling placebo response in depression trials using a longitudinal model with informative dropout. European Journal of Pharmaceutical Sciences, 36, 4–10. doi:10.1016/j.ejps.2008.10.025CrossRefGoogle ScholarPubMed
Jiang, J. (1998). Consistent estimators in generalized linear mixed models. Journal of the American Statistical Association, 93, 720–729.CrossRefGoogle Scholar
Jorm, A. F., Henderson, A. S., Kay, D. W. K. and Jacomb, P. A. (1991). Mortality in relation to dementia, depression and social integration in an elderly community sample. International Journal of Geriatric Psychiatry, 6, 5–11.CrossRefGoogle Scholar
Katona, C. L. E. and Shankar, K. K. (2004). Depression in old age. Reviews in Clinical Gerontology, 14, 283–306.CrossRefGoogle Scholar
Kristman, V. L., Manno, M. and Côté, P. (2005). Methods to account for attrition in longitudinal data: do they work? A simulation study. European Journal of Epidemiology, 20, 657–662.CrossRefGoogle ScholarPubMed
Lavery, L. L., Dodge, H. H., Snitz, B. and Ganguli, M. (2008). Cognitive decline and mortality in a community-based cohort: the MoVIES Study. Journal of the American Geriatrics Society. E-published ahead of print, doi: 10.1111/j.1532-5415.2008.02052.CrossRefGoogle Scholar
Lin, H., McCulloch, C. E. and Mayne, S. T. (2002). Maximum likelihood estimation in the joint analysis of time-to-event and multiple longitudinal variables. Statistics in Medicine, 21, 2369–2382.CrossRefGoogle ScholarPubMed
Little, R. J. A. (1995). Modeling the drop-out mechanism in longitudinal studies. Journal of the American Statistical Association, 90, 1112–1121.CrossRefGoogle Scholar
Little, R. J. A. and Rubin, D. B. (2002). Statistical Analysis with Missing Data, 2nd edn. New York: Wiley.CrossRefGoogle Scholar
Matthews, F. E., Chatfield, M., Freeman, C., McCracken, C., Brayne, C. and MRC-CFAS (2004). Attrition and bias in the MRC cognitive function and ageing study: an epidemiological investigation. BMC Public Health, 4, 12. doi:10.1186/1471-2458-4-12.CrossRefGoogle ScholarPubMed
Matthews, F. E., Chatfield, M., Brayne, C. and MRC-CFAS (2006). An investigation of whether factors associated with short-term attrition change or persist over ten years: data from the Medical Research Council Cognitive Function and Ageing Study (MRC-CFAS). BMC Public Health, 6, 185. doi:10.1186/1471-2458-6-185.CrossRefGoogle ScholarPubMed
Ong, J. C., Kuo, T. F. and Manber, R. (2008). Who is at risk for dropout from group cognitive-behavior therapy for insomnia? Journal of Psychosomatic Research, 64, 419–425.CrossRefGoogle ScholarPubMed
Pawitan, Y. (2004). In All Likelihood: Statistical Modelling and Inference Using Likelihood. Oxford: Oxford University Press.Google Scholar
Roy, J. (2003). Modeling longitudinal data with nonignorable dropouts using a latent dropout class model. Biometrics, 59, 829–836.CrossRefGoogle ScholarPubMed
Rubin, D. B. (1976). Inference and missing data. Biometrika, 63, 581–592.CrossRefGoogle Scholar
Touloumi, G., Babiker, A. G., Pocock, S. J. and Darbyshire, J. H. (2001). Impact of missing data due to drop-outs on estimators for rates of change in longitudinal studies: a simulation study. Statistics in Medicine, 20, 3715–3728.CrossRefGoogle ScholarPubMed
Tsiatis, A. A. and Davidian, M. A. (2004). Joint modeling of longitudinal and time-to-event data: an overview. Statistica Sinica, 14, 809–834.Google Scholar
Troxell, A. B., Ma, G. and Heitjan, D. F. (2004). An index of local sensitivity to nonignorability. Statistica Sinica, 14, 1221–1237.Google Scholar
van Beijsterveldt, C. E. M., van Boxtel, M. P. J., Bosma, H., Houx, P. J., Buntinx, F. and Jolles, J. (2002). Predictors of attrition in a longitudinal cognitive aging study: the Maastricht Aging Study (MAAS). Journal of Clinical Epidemiology, 55, 216–223.CrossRefGoogle Scholar
Vonesh, E. F., Greene, T., Mark, D. and Schluchter, M. D. (2006). Shared parameter models for the joint analysis of longitudinal data and event times. Statistics in Medicine, 25, 143–163.CrossRefGoogle ScholarPubMed
Wilson, R. S., Beck, T. K., Bienias, J. L. and Bennett, D. A. (2007). Terminal cognitive decline: accelerated loss of cognition in the last years of life. Psychosomatic Medicine, 69, 131–137.CrossRefGoogle ScholarPubMed
Wu, M. and Carroll, R. (1988). Estimation and comparison of changes in the presence of informative right censoring by modeling the censoring process. Biometrics, 44, 175–188.CrossRefGoogle Scholar
Wulfsohn, M. S. and Tsiatis, A. A. (1997). A joint model for survival and longitudinal data measured with error. Biometrics, 53, 330–339.CrossRefGoogle ScholarPubMed
Supplementary material: File

Chang Appendices

Chang Appendices

Download Chang Appendices(File)
File 153.1 KB