Globally, the number of older persons—aged 60 and over—is growing faster than any other age group [
1]. Therefore, the prevalence of later-life cognitive disorders can be expected to increase. Dementia, or major neurocognitive disorder (DSM-5), is the significant impairment of cognitive performance in one or more cognitive domains (e.g. complex attention, learning and memory, executive function, language), ultimately resulting in functional incapacity and death [
2]. The most common cause of dementia is Alzheimer’s disease, where it affects 6–9% of adults aged over 60 worldwide [
3]. Mild cognitive impairment (MCI) involves cognitive decline that is greater than that which occurs in normal ageing, with some limitations to daily function (
Statistical Manual of Mental Disorders (fifth ed.;
DSM-5). It is estimated that MCI affects up to 42% of older adults and precedes the onset of dementia [
4].
Studies have indicated that computer-based interventions may be beneficial for improving or maintaining cognitive function to slow the trajectory of cognitive decline, for both those with MCI and dementia [
5‐
10]. In recent years, advances in computer technology have allowed such interventions to be administered using personal computers, laptops, tablets and other mobile devices, in an increasingly accessible, individualised and cost-effective manner. The usefulness of computer-based interventions has also been demonstrated in meta-analyses [
11,
12]. However, these meta-analyses have been associated with methodological limitations. For example, Garcia-Casal et al. [
11] used a fixed-effects statistical model to calculate summary effect sizes. Such a model assumes that all differences between studies can be accounted for by sampling error [
13]. Hence, this study may have unintentionally inflated effect sizes. To correct for such inflation, Hill et al. [
12] used a random-effects statistical model [
13] but failed to discriminate between active and passive control groups in their statistical analysis. Similarly, Sitzer et al. [
14] included mixed treatment types (e.g. computerised training with exercise); therefore, were not able to distinguish the effects of computerised training from those associated with adjunct interventions.
Aim
We aim to conduct a systematic review and meta-analysis that uses a random effects model and that compares the effects of computerised training that is not mixed with other interventions, with three types of control groups—placebo, active and passive. The review will consider the effects of such training for individuals diagnosed with dementia and MCI. The purpose of this systematic review protocol is to transparently present the method we will undertake in order to conduct the systematic review, such that it could adequately be replicated. This method includes information regarding eligibility criteria, information sources and search strategy to be used, and the process of data extraction, synthesis and analysis.