Computer Cognitive Training May Improve Brain Functions for Individuals With HIV

Results of a meta-analysis of randomized clinical trials published in JAMA shows improvement in 6 of 8 domains.

Computer cognitive training (CCT) programs were associated with improvements in cognitive and daily functions for individuals with HIV, according to the results of a meta-analysis of randomized clinical trials published in JAMA.

Investigators included 12 different randomized clinical trials to be included in the meta-analysis. When combined, there were 596 total individuals included, with 320 in the CCT arms and 276 in the control arm.

The average HIV inhibition ratio, which is the proportion of individuals who achieved virological suppression, had a range of approximately 30% to 100%. The CD4+T-cell count ranged from 471 to 833 cells/μL, and the time since training ranged anywhere between 3 and 24 weeks. They used standardized mean differences (SMD) to calculate the mean change between pre- and post-training.

Investigators found that after an individual received CCT, function significantly improved in 6 of 8 domains: abstraction and executive function (0.58 SMD), attention and working memory (0.62 SMD), daily function (0.44 SMD), memory (0.59 SMD), motor skills (0.50 SMD), and speed of information processing (0.65 SMD).

However, language and verbal and perceptual and sensory skills skills did not significantly improve afterward, and individuals who underwent CCT had SMD of 0.49 and 0.06, respectively.

Investigators also conducted a sensitivity analysis among all the factors that were used from the randomized clinical trials for each domain. Just 5 of factors had a substantial change in the outcome, which included age, CD4+ T-cell count, HIV inhibition ratio, session hours, and time since training.

Eleven of the studies had moderate-quality evidence that supported them, and just 1 had low-quality evidence. Additionally, the 6 domains that significantly improved had moderate-quality evidence that supported the findings, while the 2 domains that did not improve had very low-quality evidence.

Investigators also found that 11 studies had a low risk of bias for random sequencing, 10 for selective reporting, 8 for incomplete outcome data, and 10 for other types of bias. Additionally, 8 studies did not blind who assessed or implemented the CCT and/or those who reported relevant information.

Investigators included the studies if they included individuals aged 18 years or older; reported changes between baseline and post-training; used CCT as the primary intervention or combined CCT with other types of interventions; used passive control conditions, a placebo, single training tasks, or traditional cognitive training, as control conditions; and were randomized clinical trials.

The studies were excluded if they did not report findings on the domains that were used in the meta-analysis, were case reports, were feedback reports or research protocols, or were not associated with HIV.

Data was used from the Cochrane Library, PsycINFO, PubMed, and Web of Sciences that were searched from database inception to December 15, 2020. Investigators also used supplementary searches to identify studies that were missing in other databases but conducted in Google Scholar from inception of the database to November 18, 2021.


Wei J, Hou J, Mu T, et al. Evaluation of computerized cognitive training and cognitive and daily function in patients living with HIV: a Meta-analysis. JAMA Netw Open. 2022;5(3):e220970. doi:10.1001/jamanetworkopen.2022.0970