Virtual reality training gives older brains a small cognitive lift

Keeping the mind sharp in later life is one of the biggest open problems in ageing, and cognitive training is a crowded field with mixed results. Virtual reality offers something conventional training does not: immersive, engaging tasks that can combine movement, navigation and problem solving in one session. The question is whether that novelty translates into measurable cognitive benefit for older adults, whatever their starting cognitive level. This review set out to answer it by comparing virtual reality-based training against conventional training or information-based treatment across the randomised trial literature.

The researchers ran a systematic search across four databases and identified 31 randomised controlled trials of virtual reality-based training in older adults. Trials were included regardless of participants' cognitive level. The team pooled effect sizes for two outcomes, global cognition and executive function, assessed each trial's risk of bias, and graded the overall strength of the evidence. They also ran meta-regression, a technique for exploring which study features might explain differences in results between trials.

Virtual reality-based training beat control conditions on global cognition, with a small but statistically significant pooled effect (Hedges' g 0.42) across 876 participants in 20 trials, graded as low evidence. Executive function also improved, with a slightly smaller effect (Hedges' g 0.35) across 810 participants in 16 trials, graded as very low evidence. The trials disagreed with each other considerably on both outcomes. Meta-regression, intended to explain those differences, produced inconclusive results, so it remains unclear which features of the training drive the benefit.

This synthesis suggests virtual reality-based training may be more effective than conventional training or information-based treatment for cognition in older adults, which extends the broader cognitive training literature into newer technology. The certainty is limited, though: the evidence was graded low for global cognition and very low for executive function, and the studies varied widely in design. That leaves open whether particular VR formats, session lengths or populations account for the gains. The authors' own conclusion is measured: VR may help, and more high-quality studies are needed.

For anyone thinking about cognitive health in later life, or helping a parent do so, this is a reason to treat VR-based training as a credible option rather than a gimmick. The measured gains were small, and the low quality of the underlying trials means future research could shrink them further. It is worth noting the comparison was against conventional training and information-based treatment, so VR appeared to add something beyond simply doing an activity. Nothing here settles which kind of VR training matters most or how much of it is enough.