Game-based Cognitive Training for School-age Children Living With or Exposed to HIV: A Randomized Controlled Trial in Uganda and Malawi

ABSTRACT Objective: Children with perinatally acquired HIV (PHIV) and those exposed to HIV but uninfected (HEU) are at increased risk for adverse cognitive and behavioral outcomes, particularly in sub-Saharan Africa. This study evaluated the efficacy of brain powered games (BPG), a mobile computerized cognitive rehabilitation intervention, on neurocognitive outcomes among children with PHIV, HEU, and HIV unexposed uninfected children (HUU) in Uganda and Malawi. Methods: We conducted a randomized controlled trial in Uganda and Malawi enrolling 599 children aged 5 to 12 years (PHIV = 120, HEU = 239, HUU = 240). Participants were randomized to a 12-week BPG intervention or a waitlist. Neuropsychological outcomes were assessed at baseline, postintervention, and 6-month follow-up using the Kaufman Assessment Battery for Children Second Edition (KABC-II), test of variables of attention, and Cogstate. Linear mixed-effects models estimated intervention effects, separately by country. Results: In Uganda, children receiving BPG demonstrated higher Learning composite scores (3.34; 95% CI, 1.03–5.65), Mental Processing Index scores (2.39; 95% CI, 0.78–4.01), and Nonverbal Index scores in KABC-II (2.28; 95% CI, 0.27–4.28) immediately postintervention compared with waitlist. Cogstate Maze Learning was higher postintervention (0.02; 95% CI, 0.004–0.043) and at 6 months (0.02; 95% CI, 0.003–0.040). In Malawi, Planning scores (−2.22; 95% CI, −4.39 to −0.05) and Delayed Recall scores (−2.87; 95% CI, −5.22 to −0.51) in KABC-II were lower in the intervention arm postintervention, whereas Cogstate Identification scores were higher (−0.03; 95% CI, −0.05 to −0.004). Conclusion: Mobile cognitive training produced modest, domain-specific cognitive benefits, with effects varying by context and HIV exposure status.