Rethinking Representation: BMI, GLP-1s and the Evolution of Clinical Trials

Clinical trials are only as useful as the populations they study. That sounds obvious. The history of clinical research suggests it has not always been treated that way.

For decades, trial populations skewed toward narrower demographic profiles, often favouring individuals within tighter BMI ranges, predominantly male, and drawn from specific geographic regions. The scientific rationale was straightforward: reduce variability, increase statistical clarity, produce cleaner results. What this approach also produced, quietly and consistently, was evidence that did not reflect the real-world populations it was meant to serve.

That gap is now harder to defend.

Global obesity rates have changed the picture significantly. Individuals with a BMI of 27 and above present distinct metabolic profiles that influence drug absorption, pharmacokinetics, and clinical outcomes. Excluding or underrepresenting this population does not reduce noise. It produces findings of limited applicability and, in some cases, genuine risk when treatments developed on unrepresentative samples are applied more broadly.

The response requires more than amended inclusion criteria. Adaptive trial design, stratification by BMI and metabolic status, and more sophisticated monitoring of physiological response are all part of it. But so is patient recruitment strategy. Reaching underrepresented populations requires more than broader advertising. It requires trust-building, culturally sensitive communication, and honest acknowledgement that certain communities have historically had good reasons to be cautious about participating in research.

Obesity in particular remains a sensitive area. Communication around eligibility, body composition, and metabolic health requires care at every stage. Stigma does not disappear because a study is scientifically rigorous. It has to be actively designed out of the participant experience.

The evolution underway in clinical trial design is ultimately about relevance. Science that reflects the full diversity of the populations it aims to help will produce better evidence, build broader trust, and deliver outcomes that are meaningful beyond the controlled environment in which they were generated.

References:
Kushner RF, et al. J Clin Endocrinol Metab. 2020.
Nauck MA, et al. Diabetes Care. 2021.