Computerised cognitive training for preventing dementia in people with mild cognitive impairment

Background The number of people living with dementia is increasing rapidly. Clinical dementia does not develop suddenly, but rather is preceded by a period of cognitive decline beyond normal age‐related change. People at this intermediate stage between normal cognitive function and clinical dementia are often described as having mild cognitive impairment (MCI). Considerable research and clinical efforts have been directed toward finding disease‐modifying interventions that may prevent or delay progression from MCI to clinical dementia. Objectives To evaluate the effects of at least 12 weeks of computerised cognitive training (CCT) on maintaining or improving cognitive function and preventing dementia in people with mild cognitive impairment. Search methods We searched to 31 May 2018 in ALOIS (www.medicine.ox.ac.uk/alois) and ran additional searches in MEDLINE, Embase, PsycINFO, CINAHL, ClinicalTrials.gov, and the WHO portal/ICTRP (www.apps.who.int/trialsearch) to identify published, unpublished, and ongoing trials. Selection criteria We included randomised controlled trials (RCTs) and quasi‐RCTs in which cognitive training via interactive computerised technology was compared with an active or inactive control intervention. Experimental computerised cognitive training (CCT) interventions had to adhere to the following criteria: minimum intervention duration of 12 weeks; any form of interactive computerised cognitive training, including computer exercises, computer games, mobile devices, gaming console, and virtual reality. Participants were adults with a diagnosis of mild cognitive impairment (MCI) or mild neurocognitive disorder (MND), or otherwise at high risk of cognitive decline. Data collection and analysis Two review authors independently extracted data and assessed risk of bias of the included RCTs. We expressed treatment effects as mean differences (MDs) or standardised mean differences (SMDs) for continuous outcomes and as risk ratios (RRs) for dichotomous outcomes. We used the GRADE approach to describe the overall quality of evidence for each outcome. Main results Eight RCTs with a total of 660 participants met review inclusion criteria. Duration of the included trials varied from 12 weeks to 18 months. Only one trial used an inactive control. Most studies were at unclear or high risk of bias in several domains. Overall, our ability to draw conclusions was hampered by very low‐quality evidence. Almost all results were very imprecise; there were also problems related to risk of bias, inconsistency between trials, and indirectness of the evidence. No trial provided data on incident dementia. For comparisons of CCT with both active and inactive controls, the quality of evidence on our other primary outcome of global cognitive function immediately after the intervention period was very low. Therefore, we were unable to draw any conclusions about this outcome. Due to very low quality of evidence, we were also unable to determine whether there was any effect of CCT compared to active control on our secondary outcomes of episodic memory, working memory, executive function, depression, functional performance, and mortality. We found low‐quality evidence suggesting that there is probably no effect on speed of processing (SMD 0.20, 95% confidence interval (CI) ‐0.16 to 0.56; 2 studies; 119 participants), verbal fluency (SMD ‐0.16, 95% CI ‐0.76 to 0.44; 3 studies; 150 participants), or quality of life (mean difference (MD) 0.40, 95% CI ‐1.85 to 2.65; 1 study; 19 participants). When CCT was compared with inactive control, we obtained data on five secondary outcomes, including episodic memory, executive function, verbal fluency, depression, and functional performance. We found very low‐quality evidence; therefore, we were unable to draw any conclusions about these outcomes. Authors' conclusions Currently available evidence does not allow us to determine whether or not computerised cognitive training will prevent clinical dementia or improve or maintain cognitive function in those who already have evidence of cognitive impairment. Small numbers of trials, small samples, risk of bias, inconsistency between trials, and highly imprecise results mean that it is not possible to derive any implications for clinical practice, despite some observed large effect sizes from individual studies. Direct adverse events are unlikely to occur, although the time and sometimes the money involved in computerised cognitive training programmes may represent significant burdens. Further research is necessary and should concentrate on improving methodological rigour, selecting suitable outcomes measures, and assessing generalisability and persistence of any effects. Trials with long‐term follow‐up are needed to determine the potential of this intervention to reduce the risk of dementia.