Reviewed: 26 October 2026
A cannabinoid study reports a measurable biological change. The result sounds promising, but what does it tell us about people's lives? Before interpreting the finding, ask what the researchers actually measured. A laboratory result, a symptom rating and the ability to manage everyday activities answer different questions.
Understanding those differences helps readers recognise what a study supports, what remains uncertain and why a convincing headline may go further than the evidence.
Start with the outcome, not the headline
An endpoint is a defined outcome used to evaluate a study. Clinical outcomes directly concern how people feel, function or survive. A biomarker is a measured biological characteristic. Researchers may use biomarkers to investigate biological processes without claiming that they demonstrate a benefit people can experience.
A surrogate endpoint goes further: it substitutes for a clinical outcome, with the intention of predicting that outcome. Such endpoints can make some studies shorter or more practical. However, showing that a marker changes is different from establishing that the change reliably predicts a meaningful clinical benefit. The FDA's explanation of surrogate endpoints describes why substantial supporting evidence matters.
When reading cannabinoid research, therefore, separate a finding about biological activity from a finding about people's health or daily experience. The distinction does not make the biological result worthless; it defines the question it can answer.
A fictional example: the marker and the morning routine
Imagine an entirely fictional study of an unspecified cannabinoid intervention. Researchers measure an unnamed blood marker and ask participants how easily they complete their usual morning activities. These details illustrate study interpretation; they do not describe a real trial or an established cannabinoid effect.
Suppose the marker changes more in the intervention group than in the comparison group, but morning activities do not become easier. The marker result may justify further investigation. It does not establish improved daily functioning.
Now imagine that the study measures only the marker. Readers cannot fill the gap by assuming that daily life improved. The relevant outcome was not measured. Equally, if daily functioning improves but the marker does not change, the marker may fail to capture what matters in this setting.
These possibilities are reasons to inspect the outcome list before accepting a broad conclusion.
Why correlation does not validate a surrogate
A marker may be associated with better health without being a reliable substitute for measuring health outcomes. People with one marker level might differ in many other ways. Even changing a marker through an intervention may not reproduce the relationship seen in observational data.
The FDA–NIH BEST resource describes validation as requiring a scientific rationale and clinical evidence that the surrogate predicts a specific clinical benefit in a defined context. Evidence usually comes from multiple randomised trials; observational association alone is insufficient.
Validation is therefore context-specific, not a permanent badge that transfers automatically between populations, interventions or uses. This article does not identify any cannabinoid biomarker as a validated surrogate. Our guide to correlation and causation in cannabis health reporting explains another reason associations need careful interpretation.
Patient-reported does not mean second-rate
Some important outcomes are best described by the person experiencing them. A patient-reported outcome comes directly from the patient, without someone else interpreting the response. Symptoms such as pain or nausea may require this kind of measurement. Other assessments can examine performance on specified tasks, as explained in the FDA's patient-focused glossary.
A laboratory number is not inherently better evidence than a patient report. The appropriate measure depends on the research question. For either type, readers should consider whether the assessment is suitable, how consistently it was collected and whether missing responses could affect interpretation.
Keep benefits and harms visible
A favourable result on one measure does not provide a complete account of an intervention. Patient-important benefits and adverse outcomes need separate attention. An improved surrogate can coexist with unchanged or worse outcomes that matter to people, a concern highlighted by the Cochrane Handbook.
Ask what harms were measured, over what period and in how many participants. An encouraging biological signal does not cancel an adverse finding, and a short study cannot answer every longer-term question.
A practical reading checklist
- Name the outcome: Is it a biological marker, a patient report, a performance measure or another clinical outcome?
- Identify the claim: Does the conclusion describe what was measured, or imply benefits beyond it?
- Check surrogate evidence: If a marker substitutes for a clinical outcome, what supports that use in this particular context?
- Look for the original plan: Were outcomes specified before results were known? Our guide to clinical trial registration explains where to start.
- Read beyond benefits: Look for adverse outcomes, missing data and follow-up length.
- Consider the wider evidence: A single result needs context. Learn how systematic reviews assess research certainty.
The central question is simple: did the study measure something that matters to people, or something proposed to predict it? Both can contribute to research, but they support different conclusions.
This article provides methodological education, not treatment advice. The US sources explain research terminology; they do not establish European approval status.