Effects of Semaglutide on Cognitive Function in People With HIV: A Randomized, Controlled Trial

In this randomized, placebo-controlled trial, 32 weeks of semaglutide improved visuospatial cognitive function in people with HIV, with smaller gains in naming/language and delayed recall that did not all persist after adjustment. The visuospatial benefit remained statistically significant after adjusting for sex and CD4 count and appeared related to changes in inflammation.

Randomized controlled trial — Randomized, double-blind, placebo-controlled phase 2b trial. Population: Adults with HIV on antiretroviral therapy. Sample size: 108 participants (54 per arm). Follow-up: 32 weeks. Interventions: subcutaneous semaglutide. Cited by 2 other papers.

The trial included 108 participants (54 per arm); 65% were non-White, 40% were female, and median age was 53 years. Compared with placebo, semaglutide significantly increased visuospatial, naming/language, and delayed recall scores at 32 weeks (P = .01, .05, and .04, respectively). After adjusting for sex and absolute CD4 count, only the visuospatial score remained statistically significant (P = .05). Semaglutide's total natural direct effect maintained a positive effect on the visuospatial score while accounting for potential changes in high-sensitivity C-reactive protein and soluble CD163 levels (P = .04). The authors concluded that semaglutide may have a beneficial impact on visuospatial cognitive function in people with HIV through its effect on inflammation.

This paper provides randomized, placebo-controlled evidence that semaglutide may selectively improve visuospatial cognition in people with HIV, potentially through anti-inflammatory mechanisms. It does not establish a general cognitive benefit, a specific dose recommendation, or a fully confirmed mechanism, and it does not compare semaglutide with other GLP-1 receptor agonists.

Key findings

Compared with placebo, semaglutide significantly increased visuospatial, naming/language, and delayed recall scores at 32 weeks (P = .01, .05, and .04, respectively).
After adjusting for sex and absolute CD4 count, only the visuospatial score remained statistically significant (P = .05).
Semaglutide's total natural direct effect maintained a positive effect on the visuospatial score while accounting for changes in high-sensitivity C-reactive protein and soluble CD163 levels (P = .04).
The trial included 108 participants (54 per arm); 65% were non-White, 40% were female, and median age was 53 years.

Limitations

Authors do not state limitations in the abstract.
The 32-week follow-up in a phase 2b trial cannot establish long-term cognitive effects.
After adjustment, only one of the three initially significant cognitive domain scores remained significant.
The proposed mediation by inflammation is inferred from statistical modeling rather than directly demonstrated.

The record

Journal: Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 82(5): 753-761
Published: 2026-05-20
PMID: 41098140 — DOI: 10.1093/cid/ciaf577
MeSH terms: Humans, HIV Infections, Inflammation, Double-Blind Method, Cognition, Adult, Middle Aged, Female, Male, Adiposity, Glucagon-Like Peptides
Funding: NIDDK NIH HHS; NIH HHS; National Institutes of Health
Author countries: Lebanon, USA

Peptide profiles: Semaglutide.

All indexed evidence: Semaglutide trials & papers.