BackgroundSome studies have already suggested that exergame interventions can be effective to improve physical, cognitive, motor-cognitive, and psychological outcomes in patients with dementia (PwD). However, little is known about the training volume required to induce such positive effects and the inter-individual differences in training response among PwD. The aim of the study was to analyze the time course of changes in motor-cognitive exergame performances during a task-specific training program and to identify predictors of early training response in PwD.MethodsSecondary analyses of data from the intervention group (IG) of a randomized, placebo-controlled trial to improve motor-cognitive performances in PwD. Fifty-six geriatric patients with mild-to-moderate dementia randomized to the IG underwent a 10-week, task-specific training program (2×/week) on an exergame-based balance training system (Physiomat®), combining postural control tasks with cognitive tasks of an established neuropsychological test (Trail Making Test). Main outcome was the time required to complete different Physiomat®-Tasks (PTs) assessed at baseline (T1), training session 7 (TS7) and 14 (TS14), and post-intervention after 20 training sessions (T2). Reliable change indices were used to identify early responders from T1 to TS7. A multivariate logistic regression analysis was performed to determine independent predictors of early training response.ResultsCompletion time significantly improved already from T1 to TS7 in all PTs (p ≤ .001–.006), with moderate to very large effect sizes (r = .38–.52; Cohen’s d = .85–1.45). For most PTs, significant progressive improvements from TS7 to TS14 and TS14 to T2 were not observed. Thirty-one (59.6%) participants were classified as early responders and 21 (40.4%) as non-early responders. Lower baseline exergame performance and lower visuospatial and divided attention abilities were independently associated with early training response.ConclusionsSubstantial task-specific improvements in complex motor-cognitive exergame performances can be obtained within a surprisingly short intervention period in PwD. Our results confirm that not only an excellent training response can be achieved in this patient population, but also that more vulnerable patients with greater deficits in domain-specific cognitive functions associated with fall risk may even reap the most and fastest benefit from motor-cognitive exergame interventions.Trial registrationISRCTN registry, ISRCTN37232817 (retrospectively registered on 04/02/2012).