Skip to content

Exercise trials suggest lower drug-craving scores, with low confidence in the evidence

Randomized trials reported favorable craving results for several exercise approaches, while the best type and amount remained uncertain.

By 100HP editorialAbstract-based explanation checked

Based on the published abstract. The full paper may contain additional methods, results and limitations.

The 30-second takeaway

Aerobic exercise, high-intensity interval exercise and combined aerobic and resistance exercise were associated with lower craving scores than control conditions in the analysis. The evidence was rated low quality. Because craving was the measured outcome, these results do not establish effects on abstinence, relapse or long-term recovery from substance use disorders.

Keep in mind

Evidence quality was rated low. The abstract omits control activities and intervention lengths, limiting assessment of how comparable the exercise programs and control conditions were.

THE NUMBERS, WITH CONTEXT

What researchers found

SMD -0.23; 95% confidence interval -0.41 to -0.04

Drug-craving scores after acute aerobic exercise

Acute aerobic exercise versus control: average scores differed by -0.23 standard deviations, meaning lower craving with exercise. These units reflect score variability; exact assessment timing was unspecified.

Addict Sci Clin Pract, 2025 · Original source ↓

Understanding the score

The acute aerobic-exercise finding expresses a difference relative to variation in craving scores. Standard-deviation units allow different craving scales to be compared, but the result cannot be read as a percentage reduction in drug use.

Interpreting the dose model

The authors modeled how exercise amount related to craving and proposed a preferred amount. The abstract gives a dose estimate alongside a session schedule, but does not explain the conversion between them.

CHECK THE ORIGINAL

The original publication

Effects of different exercise types on craving in substance use disorder patients with drug dependence -network meta-analysis and dose-response relationships based on frequentist and Bayesian models.

Wang C, Yang Y, Wang K et al.
Addict Sci Clin Pract · 2025

PubMed ID
41413589
Record checked

AI-assisted research and writing. This explains one selected publication; it is not a complete review of everything known. Our approach.