The Annals of Statistics

A nonparametric dynamic additive regression model for longitudinal data

Torben Martinussen and Thomas H. Scheike

Full-text: Open access


In this work we study additive dynamic regression models for longitudinal data. These models provide a flexible and nonparametric method for investigating the time-dynamics of longitudinal data. The methodology is aimed at data where measurements are recorded at random time points. We model the conditional mean of responses given the full internal history and possibly time-varying covariates. We derive the asymptotic distribution for a new nonparametric least squares estimator of the cumulative time-varying regression functions. Based on the asymptotic results, confidence bands may be computed and inference about time-varying coefficients may be drawn. We propose two estimators of the cumulative regression function. One estimator that involves smoothing and one that does not. The latter, however, has twice the variance as the smoothing based estimator. Goodness of fit of the model is considered using martingale residuals. Finally, we also discuss how partly-conditional mean models in which the mean of the response is regressed onto selected time-varying covariates may be analysed in the same framework. We apply the methods to longitudinal data on height development for cystic fibrosis patients.

Article information

Ann. Statist., Volume 28, Number 4 (2000), 1000-1025.

First available in Project Euclid: 12 March 2002

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Primary: 62M10: Time series, auto-correlation, regression, etc. [See also 91B84]
Secondary: 62G07: Density estimation 62G10: Hypothesis testing 62G20: Asymptotic properties

Dynamic linear models estimating equations least squares longitudinal data nonparametric methods partly conditional mean models time-varying-coefficient models


Martinussen, Torben; Scheike, Thomas H. A nonparametric dynamic additive regression model for longitudinal data. Ann. Statist. 28 (2000), no. 4, 1000--1025. doi:10.1214/aos/1015956705.

Export citation


  • Aalen, O. O. (1975). Statistical inference for a family of counting processes. Ph.D. dissertation, Dept. Statistics, Univ. California, Berkeley.
  • Aalen, O. O. (1978). Nonparametric inference for a family of counting processes. Ann. Statist. 6 534-545.
  • Aalen, O. O. (1980). A model for non-parametric regression analysis of counting processes. Mathematical Statistics and Probability. Lecture Notes in Statist. 2 1-25. Springer, New York.
  • Aalen, O. O. (1989). A linear regression model for the analysis of life times. Statistics in Medicine 8 907-925.
  • Aalen, O. O. (1993). Further results on the non-parametric linear regression model in survival analysis. Statistics in Medicine 12 1569-1588.
  • Andersen, P. K., Borgan, Ø., Gill, R. and Keiding, N. (1993). Statistical Models Based on Counting Processes. Springer, New York.
  • Br´emaud, P. (1981). Point Processes and Queues, Martingale Dynamics. Springer, New York.
  • Diggle, P., Liang, K. and Zeger, S. (1994). Analysis of Longitudinal Data. Cambridge Univ. Press.
  • Fan, J. and Zhang, W. (1999). Statistical estimation in varying-coefficient models. Ann. Statistics 27 1491-1518.
  • Hall, W. and Wellner, J. (1980). Confidence bands for a survival curve from censored data. Biometrika 67 133-143.
  • Hastie, T. and Tibshirani, R. (1993). Varying-coefficient models. J. Roy. Statist. Soc. Ser. B 55 757-796.
  • Hoover, P., Rice, J., Wu, C. and Yang, L. P. (1998). Nonparametric smoothing estimates of time-varying coefficient models with longitudinal data. Biometrika 85 809-822.
  • Huffer, F. W. and McKeague, I. W. (1991). Weighted least squares estimation for Aalen's additive risk model. J. Amer. Statist. Assoc. 86 114-129.
  • Laursen, E., Koch, C., Petersen, J. and Muller, J. (1999). Secular changes in anthropometric data in cystic fibrosis patients. Acta Pædiatrica 88 171-179.
  • Liang, K. Y. and Zeger, S. L. (1986). Longitudinal data analysis using generalized linear models. Biometrika 73 13-22.
  • Lin, D. Y. and Ying, Z. (1994). Semiparametric analysis ofthe additive risk model. Biometrika 81 61-71.
  • Martinussen, T. and Scheike, T. H. (1998). A non-parametric dynamic additive regression model for longitudinal data. Research report, Dept. Biostatistics, Univ. Copenhagen.
  • Martinussen, T. and Scheike, T. H. (1999). A semi-parametric additive regression model for longitudinal data. Biometrika 86 691-702.
  • Martinussen, T. and SØrensen, T. I. A. (1998). Age-dependent U-shaped risk functions and Aalen's additive risk model. Biometrics 54 989-1001.
  • McKeague, I. W. (1988). Asymptotic theory for weighted least squares estimators in Aalen's additive risk model. Contemp. Math. 80 139-152.
  • McKeague, I. W. and Sasieni, P. D. (1994). A partly parametric additive risk model. Biometrika 81 501-514.
  • McKeague, I. W. and Zhang, M. J. (1994). Identification ofnonlinear time series from first order cumulative characteristics. Ann. Statist. 22 495-514.
  • Murphy, S. A. and Li, B. (1995). Projected partial likelihood and its application to longitudinal data. Biometrika 82 399-406.
  • Murphy, S. A. (1995). A central limit theorem for local martingales with applications to the analysis oflongitudinal data. Scandinavian J. Statist. 22 279-294.
  • Pepe, M. and Couper, D. (1997). Modeling partly conditional means with longitudinal data. J. Amer. Statist. Assoc. 92 991-998.
  • Ramlau-Hansen, H. (1983). Smoothing counting process intensities by means ofkernel functions. Ann. Statist. 11 453-466.
  • Scheike, T. H. (1994). Parametric regression for longitudinal data with counting process measurement times. Scandinavian J. Statist. 21 245-263.
  • Scheike, T. H. and Zhang, M. (1998). Cumulative regression function tests for longitudinal data. Ann. Statist. 26 1328-1355.
  • Scheike, T. H. (1997). A general framework for longitudinal data through conditional distributions. Biometrical J. 39 57-67.
  • Scheike, T. H. (2000). Comparison ofnon-parametric regression functions through their cumulatives. Statist. Probab. Lett. 46 21-32.