Andrzej T Galecki, MD, PhD
- Research Professor, Internal Medicine, Geriatric and Palliative Medicine
- Research Professor, Biostatistics
Andrzej Galecki's research interests include application of modern statistical methods to studies in geriatrics and gerontology. He is interested in nonlinear mixed effects models, population pharmacokinetics and pharmacodynamics analysis, modeling of covariance structure in longitudinal data analysis, generalized linear models and categorical data analysis. Dr. Galecki serves as a Director of Design, Data and Biostatistics Core of the Older Americans Independence Center.
- PhD, Institue of Mother and Child Care, Warsaw, Poland, 1987
- MD, Medical Academy of Warsaw, Warsaw, Poland, 1981
- MS, Department of Applied Mathematics of Technical University of Warsaw, Warsaw, Poland, 1977
My primary interests lie in developing computational methods for analyzing correlated and over dispersed data, which are frequently encountered in many fields of application, such as pharmacokinetic and pharmacodynamic (PK/PD studies, longitudinal studies, survey sampling and gene mapping in genetics studies. A class of models often considered in this context are hierarchical or mixed-effects models. These models are an extension of the regression models whereby random effects are introduced to describe between-subject variation.
My interests related to mixed effects models lie in:
1. An extension of mixed models which allows between-subject variation to be modeled as a mixture of underlying distributions.
2. Computational methods for PK/PD population studies. Here models are often expressed as a solution of a system of ordinary differential equations. To address these advanced problems, I developed a SAS/IML NLMEM and successfully used to analyze existing data. Specific application of these models occurs in population studies when intravenous glucose tolerance test (IVGTT) studies are used to evaluate glucose metabolism in patients.
3. Modeling covariance structure in the presence of two or more repeated factors. A class of models proposed in Galecki, 1994 has been implemented in PROC MIXED, which is part of commercial statistical software SAS, starting with version 6.12. The proposed class of covariance structures is especially useful for the joint modeling of several outcomes measured longitudinally.
My other research interests involve a wide range of methodological and practical aspects of research on elderly including study design and conducting study itself. I am involved in several large NIH funded projects including Design, Data and Biostatistics Core at the Claude D. Pepper Older Americans Independence Center.
Galecki, A. and Burzykowski, T., 2013. Linear Mixed-Effects Models Using R. A Step-by-Step Approach. Springer, New York, NY. https://link.springer.com/book/10.1007/978-1-4614-3900-4
West, B.T., Welch, K.B. and Galecki, A.T., 2022. Linear Mixed Models: A Practical Guide Using Statistical Software (Third Edition). Chapman and Hall/CRC. https://www.taylorfrancis.com/books/mono/10.1201/9781003181064/linear-mixed-models-brady-west-kathleen-welch-andrzej-galecki
Doria, A., Galecki, A.T., Spino, C., Pop-Busui, R., Cherney, D.Z., Lingvay, I., Parsa, A., Rossing, P., Sigal, R.J., Afkarian, M., Aronson, R. et al, 2020. Serum urate lowering with allopurinol and kidney function in type 1 diabetes. New England Journal of Medicine, 382(26), pp.2493-2503. https://www.nejm.org/doi/full/10.1056/NEJMoa1916624
Preisser, J.S., Galecki, A.T., Lohman, K.K. and Wagenknecht, L.E., 2000. Analysis of smoking trends with incomplete longitudinal binary responses. Journal of the American Statistical Association, 95(452), pp.1021-1031. https://www.tandfonline.com/doi/abs/10.1080/01621459.2000.10474299
Galecki, A.T., Ten Have, T.R. and Molenberghs, G., 2001. A simple and fast alternative to the EM algorithm for incomplete categorical data and latent class models. Computational statistics and data analysis, 35(3), pp.265-281. https://www.sciencedirect.com/science/article/abs/pii/S0167947300000153
Galecki, A.T., 1994. General class of covariance structures for two or more repeated factors in longitudinal data analysis. Communications in Statistics-Theory and Methods, 23(11), pp.3105-3119. https://www.tandfonline.com/doi/abs/10.1080/03610929408831436