Internet-based CBT reduces sexual risk behavior in MSM: exploratory RCT (n=50)
CBT-based online intervention reduced unprotected anal intercourse in the intervention group vs. control (p=0.016), but absence of absolute numbers and 95% CI limits clinical interpretability.
| Population | MSM (men who have sex with men), n=50, China |
|---|---|
| Intervention | Online intervention model based on Cognitive Behavioral Therapy (CBT) |
| Comparator | Traditional intervention methods (control group; insufficiently described in available text) |
| Outcome | Unprotected anal intercourse; Gut microbiota changes (Bifidobacterium longum, Prevotella); Pre-intervention cognitive level (risk behavior despite adequate cognition) |
Summary of findings
| Outcome | Effect | 95% CI | Certainty | Clinical relevance | Notes |
|---|---|---|---|---|---|
| Unprotected anal intercourse | p=0.016; in the RR, OR, SMD or 95% CI reported | — | Low | — | 1 studies |
| Gut microbiota changes (Bifidobacterium longum, Prevotella) | exploratory; in the statistical values reported | — | Very low | — | 1 studies |
| Pre-intervention cognitive level (risk behavior despite adequate cognition) | 44% prevalence pre-intervention; in the comparative statistic reported | — | Low | — | 1 studies |
Context
MSM bear a disproportionate burden of STIs, including HIV, syphilis, and gonorrhea. Digital behavioral interventions are a low-cost strategy for populations with limited healthcare access. This study additionally explores associations between behavioral change and gut microbiota, a dimension still lacking causal support.
What the study showed
Incidence of unprotected anal intercourse was significantly lower in the CBT group vs. control (p=0.016), but absolute pre- and post-values, effect size, and 95% CI were not reported in the available text. Pre-intervention, 44% of participants engaged in risk behaviors despite adequate cognitive levels. Changes in Bifidobacterium longum and Prevotella were observed but declared by the authors themselves as exploratory and hypothesis-generating, with no confirmatory value.
How it was done
RCT with n=50 MSM, conducted in China, with online CBT-based intervention. Duration, inclusion/exclusion criteria, randomization method, blinding, and outcome protocol are insufficiently detailed in the available text. Microbiota assessed by 16S rRNA.
Effect magnitude
Only p=0.016 was reported for the primary outcome. No RR, OR, NNT, SMD, or 95% CI was provided, precluding effect size assessment.
Risk of bias
Sample of 50 is critically insufficient for an RCT with behavioral and microbiological outcomes; statistical power not reported. Absence of 95% CI, effect size, and absolute data compromises GRADE assessment. Control arm description is inadequate, raising risk of performance and detection bias (RoB 2 applicable but not used by authors). Follow-up and dropout rate not declared.
What this study does NOT prove
This study does not prove causality between CBT and sustained reduction in risk behavior, nor causal association between gut microbiota and behavioral change. Results are not generalizable beyond convenience samples of MSM in urban Chinese contexts.
In clinical practice
Results do not support immediate adoption of clinical protocols based on this model. Clinicians may consider digital CBT as an adjuvant approach in STI prevention programs for MSM, acknowledging that current evidence is preliminary. There is no basis for microbiota intervention from this study.
Limitations
Sample of 50 is critically insufficient for an RCT with behavioral and microbiological outcomes; statistical power not reported. Absence of 95% CI, effect size, and absolute data compromises GRADE assessment. Control arm description is inadequate, raising risk of performance and detection bias (RoB 2 applicable but not used by authors). Follow-up and dropout rate not declared.
What is still missing
Adequately powered RCT (n≥200), with complete control description, minimum 12-month follow-up, and full reporting of 95% CI and effect sizes. Separate mechanistic studies for the microbiota-behavior hypothesis.
Technical appendix
Version history
- 1.0 · 2026-09-20 — Auto-generated under Evidence Standard v1.0