How being in a study changes your biology before the experiment even starts
Imagine a trial on a dietary intervention finding that participants become biologically younger.
The Hawthorne effect is a classic phenomenon from social science: people who know they are being observed change their behavior. They sleep better, exercise more, eat healthier, drink less alcohol. That sounds harmless, but in clinical research it is a serious methodological problem. If those behavioral changes influence biological biomarkers — and they do — researchers can no longer tell whether it was the intervention that worked, or simply the fact of being in a study.
In aging research, the problem is especially acute, argues a commentary published in Nature Aging. Geroscience studies measure biomarkers of biological age: blood values, epigenetic clocks, inflammation markers. These are more sensitive to lifestyle changes than many other outcomes. A participant who eats slightly better because she’s enrolled in a study can produce a measurable shift in her biological age score — a shift that has nothing to do with whatever pill or protocol is being tested.
How large is the problem, exactly?
The authors demonstrate that observation-induced behavioral changes can produce biomarker shifts comparable to — or even larger than — the effects of the interventions being tested. That is a striking claim. It means that some of the positive results reported in aging trials may not be attributable to the intervention at all, but to study participation itself.
The conventional answer is double-blind placebo control: neither participants nor researchers know who receives the real treatment and who receives the placebo. But in many aging studies, this is difficult to implement. A dietary intervention or exercise programme cannot easily be ‘blinded.’ And even in drug trials with a genuine placebo, participants may make behavioral changes that blinding does not address.
A methodological framework as a partial fix
The commentary goes beyond identifying the problem: it proposes a methodological framework for isolating and quantifying the Hawthorne effect. This includes adding an ‘observation arm’ — a group that is monitored without any intervention, to map the baseline effects of study participation. The authors also suggest that studies incorporate longer acclimatization periods, so that initial behavioral changes have subsided before measurement begins.
For readers following aging research, this offers a useful lens. When a study reports dramatic improvements in biological age biomarkers without explicitly controlling for the Hawthorne effect, some skepticism is warranted. The question is not whether the effect is real — it may be very real — but whether it is attributable to the intervention. That distinction is precisely what science tries to make, and in this field it has not always succeeded.