Choosing marginal or random-effects models for longitudinal binaryresponses: Application to self-reported disability among older persons

被引：47
作者：
Carrière I. [1 ]
Bouyer J. [2 ]
机构：
[1] INSERM Unité 500, 34093 Montpellier
[2] INSERM Unité 569-IFR69, 94276 Le Kremlin Bicetre
来源：
BMC Medical Research Methodology | / 2卷 / 1期
关键词：
Generalize Estimate Equation; Random Model; Marginal Model; Disable Woman; Work Correlation Matrix;
D O I：
10.1186/1471-2288-2-15
中图分类号：
学科分类号：
摘要：
Background: Longitudinal studies with binary repeated outcomes are now widespread in epidemiology. The statistical analysis of these studies presents difficulties and standard methods are inadequate. Methods: We consider strategies for modelling binary repeated responses and focus on two specific issues: the choice between marginal and random-effects models, and the choice of the time point origin. These issues are addressed using the example of self-reported disability in older women assessed annually for 6 years. The indicator of disability "needing help to go outdoors or home-confined" is used. Results: In view of the observed associations between the responses for consecutive years, the baseline response was considered as a covariate. We compared the marginal and random-effects models first when only the influence of time and age is analysed and second when individual risk factors are studied in an aetiological perspective. There were substantial differences between the parameter estimates. They were due to differences between specific concepts related to the two models and the large between-individual heterogeneity revealed by the analysis. Conclusions: A random-effects model appears to be most suitable for the analysis of self-reported disability in older women.
引用
收藏
页码：1 / 10
页数：9
相关论文
共 22 条
  • [1] Beckett L.A., Brock D.B., Lemke J.H., Analysis of change in self-reported physical function among older persons in four population studies, Am J Epidemiol, 143, pp. 766-778, (1996)
  • [2] Hebert R., Brayne C., Spiegelhalter D., Factors associated with functional decline and improvement in a very elderly community-dwelling population, Am J Epidemiol, 150, pp. 501-510, (1999)
  • [3] Dargent-Molina P., Favier F., Grandjean H., Fall-related factors and risk of hip fracture: The EPIDOS prospective study, Lancet, 348, pp. 145-149, (1996)
  • [4] WHO/EUROPE, NCBS Third Consultation to Develop Common Methods and Instruments for Health Interviews Surveys, (1992)
  • [5] Diggle P.J., Liang K.Y., Zeger S.L., Analysis of Longitudinal Data, (1994)
  • [6] McCullagh P., Nelder J.A., Generalized Linear Models - Second Edition, (1989)
  • [7] Zeger S.L., Liang K.Y., Albert P.S., Models for longitudinal data: A generalized estimating equation approach, Biometrics, 44, pp. 1049-1060, (1988)
  • [8] Pendergast J.F., Gange S.J., Newton M.A., A survey of methods for analyzing clustered binary response data, Int Stat Rev, 64, pp. 89-118, (1996)
  • [9] Neuhaus J.M., Kalbfleisch J.D., Hauck W.W., A comparison of cluster-specific and population-averaged approaches for analysing correlated binary date, Int Stat Rev, 59, pp. 25-36, (1991)
  • [10] Liang K.Y., Zeger S.L., Longitudinal data analysis using generalized linear models, Biometrika, 73, pp. 13-22, (1986)
123