Intervention Adherence in Longevity Trials

Key Takeaways

Adherence describes how closely participants follow the assigned intervention protocol, and it affects how much biological exposure the trial actually tests. [1] [2]
Low adherence can dilute an intention-to-treat effect estimate, while adherence-based analyses can introduce bias because adherent participants may differ systematically from less adherent participants. [3] [4] [5]
Longevity trials are especially exposed to adherence problems because behavior, diet, exercise, sleep, medication timing, and repeated biomarker visits may need to be sustained for months or years. [6] [7] [8]
Adherence is not just a compliance detail; it helps define what question the trial answered, whether the intervention was feasible, and how cautiously results should be generalized. [2] [9] [10]

Who This Is Useful For

This page is useful for readers evaluating intervention studies in ageing, healthspan, and longevity science. It is especially relevant when a trial tests a demanding lifestyle protocol, a calorie restriction target, a supplement schedule, a digital monitoring routine, or a multi-visit biomarker program where the assigned intervention may not match what participants actually did. [1] [2] [6] [8]

In a longevity trial, adherence is the difference between assigning an intervention and delivering a sustained biological exposure. A randomized design can allocate participants cleanly, but the contrast between groups may weaken if people miss doses, reduce exercise sessions, abandon diet targets, stop wearing sensors, or skip follow-up visits. [1] [2] [7]

This matters because many longevity outcomes are slow, indirect, or measured through intermediate biomarkers. If adherence is poor, a null result may mean the intervention biology was weak, the endpoint was insensitive, or the participants did not receive enough exposure to test the hypothesis clearly. [6] [8] [9]

Adherence at a Glance

Issue	What It Means	Why It Matters in Longevity Trials
Assigned intervention	The protocol participants were randomized or instructed to follow	Randomization tests assignment first, not automatically the dose actually received
Received exposure	The amount, timing, and consistency of the intervention participants actually completed	Ageing-related mechanisms may depend on cumulative exposure over time
Measurement method	Pill counts, logs, biomarkers, device data, attendance records, or direct observation	Different adherence measures capture different behaviors and can be incomplete or reactive
Analysis choice	Intention-to-treat, per-protocol, as-treated, or complier-focused analysis	Each analysis answers a different question about assignment, feasibility, or biological effect
Generalizability	Whether typical people could sustain the protocol outside the trial setting	A biologically plausible intervention may still be hard to translate if adherence depends on intensive support

1. Assignment Is Not the Same as Exposure

Randomization helps balance measured and unmeasured prognostic factors at baseline, but it does not guarantee that participants receive the planned exposure throughout follow-up. CONSORT reporting guidance treats adherence and protocol deviations as core trial information because they affect how the assigned intervention should be interpreted. [1] [2]

In longevity research, this gap can be large because interventions often involve repeated behavior rather than a single procedure. Exercise frequency, dietary targets, sleep timing, supplement use, and monitoring schedules can all drift over time, changing the biological contrast between intervention and control groups. [6] [7] [8]

2. Why Intention-to-Treat Can Look Smaller

Intention-to-treat analysis compares participants according to their assigned groups, regardless of how closely they followed the protocol. This preserves the protection of randomization and is often the clearest estimate of the effect of offering or assigning an intervention under trial conditions. [1] [3]

When adherence is low, the intention-to-treat estimate may be diluted because the intervention group includes people who received little exposure and the control group may include people who adopted similar behaviors outside the protocol. That smaller estimate is still informative, but it may describe the practical effect of assignment more than the maximum biological effect under high adherence. [3] [4] [9]

3. Why Per-Protocol Analyses Are Not Simple Fixes

Per-protocol and as-treated analyses try to estimate what happened among participants who followed, or appeared to follow, the intervention. These analyses can be useful, but they no longer have the same simple randomization protection if adherence is related to health status, motivation, socioeconomic resources, adverse effects, or early response. [3] [4] [5]

This is a central reading issue in longevity trials because healthier or more organized participants may be more able to sustain demanding protocols. Apparent benefits among adherers can therefore reflect both the intervention and the characteristics that made adherence possible. [4] [5] [11]

4. Measuring Adherence Is Itself a Study Design Choice

Adherence can be measured through self-report, visit attendance, pill counts, food records, wearable data, biochemical markers, or combinations of these methods. Reporting frameworks for interventions emphasize that enough detail is needed to understand what was delivered, how it was modified, and how much participants actually received. [2] [10]

Each method has limits. Self-report can overestimate adherence, device data can miss context, biomarkers may be indirect, and clinic attendance can measure engagement with the study more than completion of the intervention itself. These limits affect how confidently a trial can connect the protocol to the biological or functional outcome. [2] [7] [10]

5. Why Longevity Trials Make Adherence Harder

Trials aimed at ageing biology often need sustained exposure and repeated measurement because lifespan, multimorbidity, frailty, function, and biological-age markers change gradually. Geroscience trial frameworks therefore emphasize feasibility, endpoint selection, and proof-of-concept designs rather than assuming that a short biomarker shift is equivalent to a durable ageing effect. [6] [8] [9]

The CALERIE trial illustrates this problem in a relatively direct way: participants were assigned to long-term calorie restriction, but achieved restriction was lower than the original target. That does not make the trial uninformative; it shows why actual exposure and protocol feasibility are part of the evidence, not background logistics. [7] [12]

6. How to Read Adherence Claims

A useful trial report should clarify the assigned intervention, how adherence was measured, how much exposure participants received, how missing adherence data were handled, and whether adherence differed by age, baseline health, side effects, or trial burden. These details help separate intervention failure, implementation failure, and endpoint mismatch. [1] [2] [10]

Interpretation is strongest when the paper distinguishes the effect of assignment from the effect of sustained exposure and explains the assumptions behind any adherence-adjusted analysis. Without that distinction, readers can mistake a feasibility problem for a biological result, or mistake an adherence-selected subgroup for a randomized comparison. [3] [4] [5]

What This Does Not Mean

It does not mean low-adherence trials are useless; they can reveal feasibility, tolerability, and real-world implementation limits. [1] [2] [9]
It does not mean high adherence proves an intervention works; endpoint validity, sample size, bias control, and follow-up still matter. [6] [8]
It does not mean per-protocol analyses should be ignored; it means they should be read as assumption-dependent analyses rather than automatic causal answers. [3] [4]
It does not mean participants are at fault when adherence is low; protocol burden, adverse effects, social context, and study design can all shape adherence. [2] [10] [11]

Practical Interpretation Examples

If a trial reports no biomarker effect but adherence was low: the result may reflect weak biological activity, insufficient exposure, endpoint insensitivity, or some combination of these factors. [6] [8] [9]
If only highly adherent participants improved: check whether the analysis accounts for differences between adherent and less adherent participants. [3] [4] [5]
If a protocol required intensive coaching or frequent clinic visits: strong adherence inside the trial may not imply easy adoption in routine life. [2] [9] [10]
If adherence is measured only by self-report: the exposure estimate may be less certain than if several adherence measures point in the same direction. [2] [10]

Summary

Adherence determines how much of an intervention a longevity trial actually tests. It affects effect size, bias risk, feasibility, and generalizability, especially when protocols require sustained behavior change or repeated measurement. The strongest interpretation separates the effect of assignment from the effect of exposure and treats adherence as evidence about the intervention, not as a minor procedural footnote. [1] [3] [8] [9]

References

Schulz, K. F., Altman, D. G., and Moher, D. (2010). CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. BMJ. https://www.bmj.com/content/340/bmj.c332
Hoffmann, T. C., Glasziou, P. P., Boutron, I., et al. (2014). Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ. https://www.bmj.com/content/348/bmj.g1687
Hernan, M. A., and Robins, J. M. (2017). Per-protocol analyses of pragmatic trials. New England Journal of Medicine. https://pubmed.ncbi.nlm.nih.gov/28564548/
Hernan, M. A., and Robins, J. M. (2006). Estimating causal effects from epidemiological data. Journal of Epidemiology and Community Health. https://pmc.ncbi.nlm.nih.gov/articles/PMC2652882/
Shrank, W. H., Patrick, A. R., and Brookhart, M. A. (2011). Healthy user and related biases in observational studies of preventive interventions: a primer for physicians. Journal of General Internal Medicine. https://pmc.ncbi.nlm.nih.gov/articles/PMC3077477/
Justice, J. N., Kritchevsky, S. B., and Ferrucci, L. (2018). Frameworks for proof-of-concept clinical trials of interventions that target fundamental aging processes. The Journals of Gerontology: Series A. https://pmc.ncbi.nlm.nih.gov/articles/PMC6523054/
Ravussin, E., Redman, L. M., Rochon, J., et al. (2015). A 2-year randomized controlled trial of human caloric restriction: feasibility and effects on predictors of health span and longevity. The Journals of Gerontology: Series A. https://pmc.ncbi.nlm.nih.gov/articles/PMC5315691/
Cummings, S. R., and Kritchevsky, S. B. (2022). Endpoints for geroscience clinical trials: health outcomes, biomarkers, and biological age. GeroScience. https://pmc.ncbi.nlm.nih.gov/articles/PMC9768060/
Loudon, K., Treweek, S., Sullivan, F., Donnan, P., Thorpe, K. E., and Zwarenstein, M. (2015). The PRECIS-2 tool: designing trials that are fit for purpose. BMJ. https://www.bmj.com/content/350/bmj.h2147
Borrelli, B. (2011). The assessment, monitoring, and enhancement of treatment fidelity in public health clinical trials. Journal of Public Health Dentistry. https://pmc.ncbi.nlm.nih.gov/articles/PMC3194367/
Chewning, B. (2006). The healthy adherer and the placebo effect. BMJ. https://pmc.ncbi.nlm.nih.gov/articles/PMC1488753/
Kraus, W. E., Bhapkar, M., Huffman, K. M., et al. (2019). 2 years of calorie restriction and cardiometabolic risk: exploratory outcomes of a multicentre, phase 2, randomised controlled trial. The Lancet Diabetes & Endocrinology. https://pubmed.ncbi.nlm.nih.gov/31402175/

Educational Disclaimer

This content is provided for educational purposes only and does not constitute medical advice.