Digital motor intervention effects on physical activity performance of individuals with developmental disabilities: a systematic review.

Only 16 studies met the inclusion criteria, and they varied widely in who was included, which game was used, how long the programme ran and what was measured, so the results could not be pooled into a single estimate. Only 3 of the 9 controlled studies were at low risk of bias. No study met the criteria for blinding participants or trainers, only one concealed who was assigned to which group, and only one reported checking that the intervention was delivered as intended. Among the before-and-after studies, only one justified its sample size, none blinded the person doing the assessment, and none measured outcomes more than once after the intervention. The authors state the literature is too limited to develop clinical practice guidelines, and that their recommendations come from studies that reported improvements rather than from a formal synthesis. The number of participants is reported inconsistently: one part of the results says 609 people received digital interventions, another says 359, and the conclusion says 400. The review excluded people with significant mobility impairments, naming cerebral palsy, fragile X syndrome, muscular dystrophy, Rett syndrome and paediatric stroke, so it says nothing about non-ambulatory patients. Only 2 of the 16 studies reported socioeconomic status and only 2 reported race or ethnicity, so it is impossible to judge how representative the samples were. Fewer than half of the studies measured heart rate, which the authors say is needed to know whether the activity was intense enough to meet physical activity guidelines.

The authors declare that they have no conflicts of interest to report. The paper does not state who funded the review.

Do not read this as proof that exercise games work for people with developmental disabilities. It rests on 16 studies, only 3 of the 9 controlled studies were at low risk of bias, none blinded participants or trainers, and only one study checked that the intervention was actually delivered as intended. The authors themselves say the evidence is too limited to write clinical practice guidelines.