Crossover CBD Trials: Why the Washout Period Matters

By the Clinical Research Explained Editorial Teamial Team

A washout period is the gap between two treatment periods in a crossover trial, during which participants receive no study treatment so the first one can wear off. It matters because, if the first treatment's effect lingers, the results from the second period cannot be cleanly credited to what was given in that period. When you read a crossover CBD study, check how long the gap was and whether the authors explain why that length was enough.

What a Crossover Trial Does Differently

In a parallel trial, each participant receives one treatment. In a randomized crossover trial, each participant receives several treatments in a random order, so people are randomized to a sequence rather than to a single treatment. The simplest version is the AB/BA design: one group gets treatment A and then B, the other gets B and then A. Each participant acts as their own comparison, according to the CONSORT extension for randomised crossover trials published in the BMJ in 2019.

That guidance says the design suits symptomatic treatment of chronic or relatively stable conditions, where treatment effects are reversible and short-lived. It is a poor fit when a condition can be cured or when participants may die during the trial.

The label can also mislead. The same guidance notes that almost one quarter of records labelled “crossover assignment” in one registry review (17 of 72) did not randomize participants to a sequence. In those studies, people were allowed to change treatment during the trial.

What a Washout Period Does

The BMJ guidance defines washout as the time between treatment periods when no treatment is received, so the treatment can wear off. Its purpose is to prevent carryover, which is when the effect of the first treatment persists into the second period. If carryover occurs, the difference observed between treatments can depend on which one came first.

Carryover has more than one possible cause. The guidance points to a drug remaining in the system, but also to earlier side effects or other reactions to the first treatment that change how a participant responds later.

A washout does not solve every timing problem. A period effect is a change in the outcome over time regardless of treatment, for example because the condition is unstable or varies with the season. Giving equal numbers of participants each sequence means a period effect should not bias the treatment estimate on average, but authors should still report how they handled it.

A Worked Example From a CBD Trial

A published trial of oral cannabidiol in people with hypertension shows how these details appear in a report. According to its PubMed abstract, the study (HYPER-H21-4) was randomized, triple-blind, placebo-controlled and crossover. Seventy patients with mild or moderate primary hypertension received either five weeks of oral CBD or placebo. After a washout of more than two weeks, they switched to the other treatment. The authors reported reductions in average 24-hour blood pressure after 2.5 weeks of CBD, with no serious adverse events or changes in liver enzymes.

This article uses that trial only to illustrate design. It does not assess whether CBD products lower blood pressure. The abstract gives the washout length (more than two weeks) but does not explain why that length was chosen. To judge whether it was long enough, a reader would need the full methods section and the trial's registry record (ClinicalTrials.gov ID NCT05346562, as listed in the abstract). Our guide to matching a study with its registry record explains how to find that record.

Why Washout Length Affects How You Read the Results

If a washout is too short, the first treatment can either improve or suppress the outcome seen with the second, the guidance explains. In either case, the comparison between treatments becomes less reliable.

Statistics cannot reliably repair this. The guidance states that, in the standard two-period design, a carryover effect can neither be identified with enough power nor adjusted for. Authors must therefore assume carryover is negligible and justify that assumption. Testing for carryover and then discarding second-period data if the test is positive is described as flawed and is not recommended.

Reporting is uneven. In one review cited by the guidance, 69% of crossover trials (87 of 127) reported and defended a washout period. In a review of 124 crossover trials of drugs for chronic pain, 28% (35 of 124) reported baseline and post-washout pain levels. Those figures describe crossover trials in general and chronic pain trials specifically, not CBD studies.

The guidance also shows what a justification can look like. One example describes a two-week washout as covering five or more half-lives of either drug. Another trial chose no medicine-free gap for patient-safety reasons and argued that an eight-week treatment period was long enough for the first treatment to wash out before measurement.

This article does not give a standard washout length for CBD. The sources checked for it do not set one, and the appropriate length can depend on the product, the dose, how long it was taken and the outcome measured. The more useful question for a reader is whether the authors state their reasoning.

A Plain-Language Trial-Design Explainer: Four Checks

1. Sequence

Were participants randomized to the order of treatments?
Does the report show baseline characteristics by sequence, so you can see whether the groups started out similar?

2. Washout

Is the length stated, and is a reason given for it?
Does the report show whether measures had returned to their starting level before the second period began?

3. Carryover

Do the authors discuss carryover as a possible limitation?
Are results shown for each treatment in each period, which helps readers see whether order mattered?
Does the report say how many participants dropped out after the first period, and why?

4. Product match

Is the CBD product described in enough detail to replicate, including how and when it was given? The reporting guidance asks for exactly this.
Do the dose, form and placebo match what the study describes? The abstract above ties its safety statement to “the above-noted CBD formulation,” so that statement is about the formulation studied. A different product is a separate question the trial cannot answer. Our guide to comparing a study's dosage with a product label walks through that comparison.

Limits of This Article

It draws on one trial abstract and one reporting guideline. It did not review full-text methods for the trial discussed.
The BMJ guidance centers on the simple two-treatment, two-period design. More complex designs raise additional issues.
Reading a washout correctly does not tell you whether any product is safe or effective for you.

This is general education, not medical advice. If you take blood pressure medicine or other prescriptions, or are considering CBD, talk with a physician or pharmacist first. See our Medical Disclaimer.