Drug-Free Sleep Therapies Show Real Promise for People Living with HIV

Sleep disturbances are among the most common yet underappreciated burdens of living with HIV. Even as antiretroviral therapy has transformed the infection into a manageable chronic condition, many people on successful treatment still struggle to fall asleep, stay asleep, or wake feeling rested. A new systematic review and meta-analysis published in BMC Public Health now offers the first quantitative synthesis of whether non-drug approaches—ranging from cognitive behavioural therapy for insomnia to mindfulness training and sleep hygiene education—can genuinely improve sleep quality in this population. The findings point to a statistically significant benefit, but the researchers are careful to stress that the evidence base remains small and preliminary.

The review, led by Hening Pujasari of Universitas Indonesia together with colleagues in Indonesia and Malawi, followed the PRISMA 2020 reporting guidelines and was prospectively registered with PROSPERO. The team searched ten electronic databases from their inception through October 2025, looking for both randomized and non-randomized studies that assessed sleep outcomes in people living with HIV using validated instruments. Those instruments included the Pittsburgh Sleep Quality Index, the Insomnia Severity Index, actigraphy, and polysomnography—the gold-standard laboratory measurement of sleep architecture. Two reviewers independently screened the records and extracted the data, a standard safeguard against individual judgment errors.

Methodological rigor was a defining feature of the analysis. Risk of bias was assessed with design-appropriate tools: the Cochrane RoB 2.0 instrument for randomized trials and Joanna Briggs Institute tools for other study designs. Importantly, the authors judged bias domain by domain rather than collapsing everything into a single score, which preserves the nuance that a trial might be well-conducted overall yet vulnerable in one specific area such as blinding or outcome measurement. Eleven studies ultimately met the eligibility criteria, spanning a diverse array of interventions and comparison conditions.

That diversity posed a genuine analytical challenge. Because the eleven studies differed substantially in how their comparator groups were structured and which outcome instruments they used, the researchers restricted quantitative pooling a priori to studies with comparable controlled post-intervention Pittsburgh Sleep Quality Index data. Only three randomized trials—by Molavi and colleagues in 2020, Qi and colleagues in 2024, and Wen and colleagues in 2025, together enrolling 187 participants—met that comparability threshold. These were combined using a random-effects model, the standard approach when true effect sizes are assumed to vary across studies rather than being identical.

The pooled result was striking. The standardized mean difference, expressed as Hedges g, came out at −1.38 with a 95 percent confidence interval of −1.78 to −0.99, and heterogeneity was modest at I² = 27.5 percent. In plain terms, the interventions improved subjective sleep quality by an amount conventionally considered large, and the variation between studies was relatively low. Because only three trials were pooled, the team also ran a Hartung-Knapp sensitivity analysis, which is better suited to small numbers of studies than the default methods. That check produced a wider but still statistically significant interval of −2.31 to −0.48, reinforcing confidence in the direction of the effect even under conservative assumptions.

The authors went further with robustness testing. A leave-one-out analysis, in which the pooled estimate is recalculated after omitting each study in turn, retained both the direction and the statistical significance of the effect no matter which single trial was excluded. This is a meaningful safeguard in small meta-analyses, where a single influential study can sometimes drive the entire result. With only three pooled studies, however, publication bias could not be reliably assessed—a limitation the researchers acknowledge explicitly, since small meta-analyses are precisely the setting in which unpublished negative trials are hardest to detect.

Not all the evidence could be pooled. One study used a mixed pharmacological comparator, and several others reported outcomes with instruments other than the Pittsburgh Sleep Quality Index, so these were synthesized narratively rather than merged into the quantitative estimate. This decision reflects a growing consensus in evidence synthesis: forcing heterogeneous studies into a single pooled number can produce a deceptively precise figure that obscures real clinical differences. By pre-specifying their pooling criteria, the authors traded breadth of quantification for integrity of the estimate they did report.

Why does sleep matter so much in HIV care? The review’s background notes that sleep disturbances among people living with HIV are associated with poorer quality of life, suboptimal adherence to antiretroviral therapy, and adverse clinical outcomes. Adherence is particularly critical: consistent daily medication is what keeps viral load suppressed and prevents drug resistance. A patient who sleeps badly may struggle with the energy, mood, and routine organization that adherence demands, creating a pathway from disturbed sleep to worse virological control. Interventions that improve sleep without adding drug interactions or side effects therefore carry substantial public health potential, especially since they are described as scalable and low-risk.

The interventions represented in the eligible studies spanned several categories, including cognitive behavioural therapy for insomnia, brief behavioural treatment for insomnia, mindfulness-based cognitive therapy, emotional freedom techniques, and sleep hygiene training. The authors are careful not to crown a winner: with only three studies in the pooled analysis and considerable heterogeneity in intervention type, no single category can be established as superior. This is an honest reading of the data. Clinicians hoping for a definitive prescription—say, eight sessions of CBT-I over sleep hygiene pamphlets—will not find one here, and the authors explicitly say the findings should be regarded as preliminary pending larger, standardized trials.

What the review does establish is a foundation. It demonstrates that the question is answerable with rigorous methods, that subjective sleep quality as measured by the Pittsburgh Sleep Quality Index responds to non-pharmacological intervention in this population, and that future trials should standardize their outcome instruments and comparator structures so that their results can be combined. For the roughly thirty-eight million people living with HIV worldwide, many of whom report poor sleep, the message is cautiously encouraging: drug-free approaches appear to work, the effect may be large, and the research community now has a clear template for confirming it at scale.

Subject of Research: Non-pharmacological interventions for improving sleep quality in people living with HIV

Article Title: Non-pharmacological interventions to improve sleep quality among people living with HIV: a systematic review and meta-analysis

Article References: Pujasari, H., Putri, Y. S. E., Umar, E., Nurhayati, N., Friska, F., & Mulyana, A. S. (2026). Non-pharmacological interventions to improve sleep quality among people living with HIV: a systematic review and meta-analysis. BMC Public Health. https://doi.org/10.1186/s12889-026-29661-7

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29661-7

Keywords: HIV, sleep quality, non-pharmacological interventions, systematic review, meta-analysis, PSQI, CBT-I, mindfulness, sleep hygiene, public health, antiretroviral therapy, insomnia