Objectives: We aimed to characterise the prevalence and burden of HAND and assess associated factors in the global population with HIV.

Methods: We searched PubMed and Embase for cross-sectional or cohort studies reporting the prevalence of HAND or its subtypes in HIV-infected adult populations from Jan 1, 1996, to May 15, 2020, without language restrictions. Two reviewers independently undertook the study selection, data extraction, and quality assessment. We estimated pooled prevalence of HAND by a random effects model and evaluated its overall burden worldwide.

Results: Of 5,588 records identified, we included 123 studies involving 35,513 participants from 32 countries. The overall prevalence of HAND was 42·6% (95% CI: 39·7-45·5), and did not differ with respect to diagnostic criteria used. The prevalence of asymptomatic neurocognitive impairment (ANI), mild neurocognitive disorder (MND) and HIV-associated dementia (HAD) were 23·5% (20·3-26·8), 13·3% (10·6-16·3) and 5·0% (3·5-6·8) according to the Frascati criteria, respectively. The prevalence of HAND was significantly associated with the level of CD4 nadir, with a prevalence of HAND higher in low CD4 nadir groups (mean/median CD4 nadir <200: 45·2%, 40·5-49·9) versus high CD4 nadir group (mean/median CD4 nadir ≥200: 37·1%, 32·7-41·7). Worldwide, we estimated that there were roughly 16145400 (95% CI 15046300-17244500) cases of HAND in HIV-infected adults, with 72% in sub-Saharan Africa (11571200 cases, 95% CI 9600000-13568000).

Conclusions: Our findings suggest that people living with HIV have a high burden of HAND in the ART era, especially in sub-Saharan Africa and Latin America. Earlier initiation of ART and sustained adherence to maintain a high-level CD4 cell count and prevent severe immunosuppression is likely to reduce the prevalence and severity of HAND.

The source data are available from individual articles reviewed.

Supplementary Table 1: Search strategy for systematic review and meta-analysis on the prevalence of HAND in adults with HIV
Supplementary Table 2: Tool used for risk of bias
Supplementary Table 3: Characteristics of included studies
Supplementary Table 4: Univariable meta-regression analyses for moderators on prevalence estimate of HAND and its subtype in adults with HIV
Supplementary Table 5: Multivariable meta-regression analyses for moderators on prevalence estimate of HAND and its subtype in adults with HIV
Supplementary Table 6: Sensitivity analysis for study quality on prevalence estimate of HAND and its subtype in adults with HIV

Supplementary Figure 1: Random-effects meta-analysis results for prevalence of HAND in HIV-infected adults, by diagnostic criteria
Supplementary Figure 2: Random-effects meta-analysis results for prevalence of ANI according to Frascati criteria in HIV-infected adults
Supplementary Figure 3: Random-effects meta-analysis results for prevalence of MND according to Frascati criteria in HIV-infected adults
Supplementary Figure 4: Random-effects meta-analysis results for prevalence of HAD according to Frascati criteria in HIV-infected adults
Supplementary Figure 5: Funnel plot and asymmetry tests of the pooled estimates of HAND, ANI, MND and HAD
Supplementary Figure 6: Proportion on ART and current CD4 count by age group

Reference list of included studies