A prediction-interval framework for identifying outlier control cohorts in lifespan studies

Authors

Maeyens, L. T., Luciano, A., Nelson, J. F., Churchill, G. A.

Abstract

Control cohorts anchor the interpretation of lifespan-extending interventions, yet their lifespans vary even under identical experimental conditions. Without quantifying this expected variation, it is difficult to distinguish unusually short- or long-lived cohorts arising by chance from those reflecting potential experimental anomalies, with important consequences for interpreting treatment effects. The National Institute on Aging Interventions Testing Program (ITP), which tests interventions in mice across three independent sites using standardized protocols, provides a unique opportunity to address this problem. Previously, unusually short-lived males at the University of Texas in 2017 and unusually long-lived females at The Jackson Laboratory in 2022 were identified as potentially anomalous, but no standardized framework existed to determine whether such deviations exceeded expected variation. Here, we analyzed 10,757 control mice spanning 18 years of the ITP to quantify historical variation within and between sites and develop a prediction-interval framework for prospectively identifying unusual cohorts. We found that year-to-year variation, particularly in males, contributed substantially more to variation among cohort means than expected from random sampling alone. Accounting for this variation, our predictive framework flagged the 2022 Jackson Laboratory females but not 2017 University of Texas males, while flagging a previously unrecognized male cohort. Excluding this newly flagged cohort from the original lifespan analysis changed the statistical significance of two rapamycin regimens. These findings highlight the importance of a standardized framework for outlier detection and demonstrate the importance of accounting for natural variability in aging studies.

Preprint server: bioRxiv
The authors list and abstract were imported from bioRxiv on 23 Oct 2026.

Stats

1-terrible, 9-excellent. How would you rate this preprint? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.