The Effect of Cognitive Training on Domains of Attention in Older Adults With Mild Cognitive Impairment and Mild Dementia: A Meta-Analysis of Randomised Controlled Trials
Journal of Global Health
Sung, C. M., Jen, H. J., et al. (2023).
Journal of Global Health, 13, 04078.
<div>This systematic review and meta-analysis investigates the effects of cognitive training on attention in adults with mild cognitive impairment or mild dementia.</div>
No funding received
From database inception to November 3, 2022
<div>Randomized-controlled trials</div>
17
<div>Cognitive training improved overall attention (pooled Hedge's g = 0.41). Sensitivity analysis showed no significant difference from the primary pooled effect size. Additional findings showed:</div>
<div>
<ul>
<li>significantly improved selective attention (pooled Hedge's g = 0.37), with sensitivity analysis showing no significant difference from the primary value and</li>
<li>significantly improved divided attention (pooled Hedge's g = 0.38), with sensitivity analysis showing the same effect size as the primary pooled effect.</li>
</ul>
<div>Limitations to this review included heterogeneity of the included studies.</div>
</div>
<div>Cognitive training improved global cognitive function (pooled Hedge's g = 0.30). Sensitivity analysis showed no significant difference from the main analysis. Limitations to this review included heterogeneity of the included studies.</div>
<div>Subgroup analysis showed that for cognitive training:</div>
<div>
<ul>
<li>individual intervention was superior to group format (Hedge's g = 0.44);</li>
<li>< 60 minute length of training was superior to ≥ 60 minute length of training (Hedge's g = 0.97);</li>
<li>< 3 weekly training sessions was superior to ≥ 3 sessions (Hedge's g = 0.63);</li>
<li>total training of ≥ 8 weeks was superior to < 8 weeks (Hedge's g = 0.52); and</li>
<li>total sessions of < 24 was superior to ≥ 24 sessions (Hedge's g = 0.30).</li>
</ul>
<div>Limitations to this review included heterogeneity of the included studies.</div>
</div>
