If you work in this category you have seen this study cited, usually as proof that terpenes improve sleep. It is a real trial, it is properly randomized, and it is more interesting than the summaries suggest. It also does not say what most people claim it says.
Worth reading carefully, because it is a good example of how to read any study in this field.
The citation
Wang M, Faust M, Abbott S, Patel V, Chang E, Clark JI, Stella N, Muchowski PJ. “Effects of a cannabidiol/terpene formulation on sleep in individuals with insomnia: a double-blind, placebo-controlled, randomized, crossover study.” Journal of Clinical Sleep Medicine 2025;21(1):69-80. Full text is available on PubMed Central.
Note the year. It is a 2025 paper, not a 2026 one, and it is regularly miscited as new.
What they tested
Oral capsules containing 300 mg CBD plus eight terpenes at 1 mg each: linalool, myrcene, phytol, limonene, alpha-terpinene, alpha-terpineol, alpha-pinene and beta-caryophyllene. No THC in the formulation at all.
Participants took it at least four nights a week for four weeks, then had a one-week washout, then crossed over to the other arm for a second four weeks. Crossover design, so each person acts as their own control, which is a genuine strength.
What they found
The primary outcome was the percentage of nightly time spent in slow-wave plus REM sleep. The formulation produced a mean increase of 1.3 percent, standard error 0.60, 95% confidence interval 0.1 to 2.5, P = .03.
The paper itself describes this as a marginal increase. That is the authors’ word, not a critic’s.
The four things that should shape how you read it
The sample. 125 people enrolled. 56 completed the analysis. That is roughly 50 percent dropout, which the paper flags as reducing statistical power. Anyone citing “a trial of 125 patients” is quoting the enrollment figure, not the analysed one.
The measurement. Sleep stages were measured with a wrist-worn consumer device rather than polysomnography, which is the clinical standard. Wrist actigraphy infers sleep stages, it does not measure them directly. The study was decentralized, which is a reasonable trade-off for scale, but it is a trade-off.
The subjective result. Self-reported sleep measures showed no statistical significance. So the device detected a small change and the people sleeping did not report noticing one. That gap is the most honest thing in the paper.
The funding. The study was funded by Defined Research, Inc., and multiple authors hold equity or stock options in the sponsor. That does not make the findings wrong. It does mean the disclosure belongs in any summary, and it is usually the first thing dropped.
What it does and does not support
It supports the statement that a CBD and terpene formulation produced a small, statistically significant increase in device-measured slow-wave and REM sleep versus placebo in a crossover trial, with no adverse events reported.
It does not support the claim that terpenes improve sleep. The formulation contained 300 mg of CBD alongside 8 mg of total terpenes. The design cannot separate the contribution of the terpenes from the contribution of the CBD, because there was no CBD-only arm. Attributing the effect to the terpenes is not something the data can carry.
It also does not support any therapeutic claim about insomnia. A 1.3 percentage point shift in sleep architecture, detected by a wrist tracker and not noticed by participants, is a signal worth following, not a clinical outcome.
Why it is still worth citing
Because it is one of the better-designed studies in a field with very few of them, and because reporting it accurately is a competitive advantage.
Most content in this category either ignores the research or inflates it. A brand that can describe a trial’s design, its dropout rate, its measurement limitations and its funding, and still find it interesting, is a brand that a formulator or a buyer can trust on the things that are harder to check. That is worth more than a headline claim you would have to walk back.
If you want the wider context on how these compounds are thought to act, the neuroscience behind terpene effects covers the proposed mechanisms, and the entourage effect evidence covers the interaction question that this trial was implicitly testing.
About this article
Written 15 July 2026 and checked against the statute, the Federal Register and the published papers rather than against secondary coverage. Regulatory positions in this area are moving, so check the date before relying on it. Nothing here is legal, tax or medical advice.
