Estimating the correlation between rates of change in longitudinal studies in presence of dropout
Abstract
Estimating the correlation between subject-specific rates of change is a common objective in longitudinal studies. A natural approach is to fit ordinary least squares regressions per subject and correlate the resulting slope estimates directly. Under complete data, this slopes-first estimator is algebraically equivalent to a contrast-based plug-in estimator derived from the joint variance-covariance matrix of the two signals across time. Under independent dropout, however, the slopes-first approach implicitly applies heterogeneous weights across missing-data patterns, introducing bias whenever the cross-covariance between the two signals is not constant over time. We formalise this result and propose an unbiased alternative that estimates the joint covariance matrix before applying the contrast. We further extend our results to include covariates as well. Simulations confirm that the slopes-first estimator is biased under non-constant cross-covariance while the proposed alternative remains unbiased. Further, we show in a real-world example that the two estimators can lead to different scientific conclusions and recommend that researchers working with longitudinal data inspect the missing data pattern to make an informed decision about the estimator of choice.
Related
- Spatially varying effect estimates of mediation in the presence of spatial auto-correlation
- Estimating duration in sex work from cross-sectional studies: a Bayesian global synthesis suggests shorter durations
- Longitudinal Changes in Testicular Volume and the Additional Value of Body Weight for Estimating Histology-Linked Testicular Development in Male Cynomolgus Macaques (Macaca fascicularis).
- Associations of digital exclusion with joint trajectories of depression and self-rated health: evidence from two prospective longitudinal studies
- Presence of Paradoxical Mid-Systolic Longitudinal Strain (Stretching) with Speckle Tracking in the Basal Inferior-to-Lateral Segments in Mitral Valve Prolapse.
BibTeX
@misc{dorfel2026estimating,
title = {{Estimating the correlation between rates of change in longitudinal studies in presence of dropout}},
author = {Ruben P. Dörfel and Brice Ozenne and the Alzheimer’s Disease Neuroimaging Initiative (ADNI)},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.16.26365737},
url = {https://www.medrxiv.org/content/10.64898/2026.10.16.26365737v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Estimating the correlation between rates of change in longitudinal studies in presence of dropout
AU - Ruben P. Dörfel
AU - Brice Ozenne
AU - the Alzheimer’s Disease Neuroimaging Initiative (ADNI)
PY - 2026
DA - 2026/10/16
PB - medRxiv
DO - 10.64898/2026.10.16.26365737
UR - https://www.medrxiv.org/content/10.64898/2026.10.16.26365737v1
AB - Estimating the correlation between subject-specific rates of change is a common objective in longitudinal studies. A natural approach is to fit ordinary least squares regressions per subject and correlate the resulting slope estimates directly. Under complete data, this slopes-first estimator is algebraically equivalent to a contrast-based plug-in estimator derived from the joint variance-covariance matrix of the two signals across time. Under independent dropout, however, the slopes-first approach implicitly applies heterogeneous weights across missing-data patterns, introducing bias whenever the cross-covariance between the two signals is not constant over time. We formalise this result and propose an unbiased alternative that estimates the joint covariance matrix before applying the contrast. We further extend our results to include covariates as well. Simulations confirm that the slopes-first estimator is biased under non-constant cross-covariance while the proposed alternative remains unbiased. Further, we show in a real-world example that the two estimators can lead to different scientific conclusions and recommend that researchers working with longitudinal data inspect the missing data pattern to make an informed decision about the estimator of choice.
ER -