LAUSR

BACKGROUND Depression is associated with cognitive deficits across multiple domains, including working memory. The n-back task, a convenient psychometric tool capable of computerised delivery and concurrent use with neuroimaging, can provide enhanced insight into working memory dysfunction in depression. This meta-analysis sought to investigate the n-back task under varying cognitive load conditions (i.e. different levels of 'n') to clarify the pattern of working memory deficits in depression. METHODS We conducted a systematic review and meta-analysis of studies involving unipolar depressed participants and matched controls utilising the n-back task. Meta-analyses were performed for accuracy and response times at four levels of cognitive load (0-, 1-, 2-, and 3-back). RESULTS 31 studies (total 1,666 participants) met inclusion criteria and were included for quantitative analyses. Depressed individuals had significantly reduced accuracy compared to controls for 1-, 2-, and 3-back tasks, but not the attentional 0-back task. Likewise, response latencies were prolonged for all task levels (0-, 1-, 2-, and 3-back). Additional meta-regression analyses indicated that participant age and clinical status (i.e. inpatient/outpatient) may exacerbate working memory deficits associated with depression. LIMITATIONS Our results indicate high levels of heterogeneity between studies, particularly for response times. CONCLUSIONS Accuracy impairments were worse at higher levels of n, with the largest effect size obtained on the 2-back task, suggesting deficits to higher executive functions. Response times were consistently prolonged at all cognitive loads in agreement with a pattern of generalised psychomotor retardation.