WOME: Theory-Based Working Memory Training — A Placebo-Controlled, Double-Blind Evaluation in Older Adults

Abstract

Background: Scientifically evaluated cognitive intervention programs are essential to meet the demands of our increasingly aging society. Currently, one of the “hottest” topics in the field is the improvement of working memory function and its potential impact on overall cognition. The present study evaluated the efficacy of WOME (WOrking MEmory), a theory-based working memory training program, in a double-blind, placebo-controlled, and randomized controlled trial (www.drks.de, DRKS00013162).Methods: N = 60 healthy older adults were allocated to (1) the WOME intervention, (2) an active low-level intervention, or (3) a passive control group. Overall, the intervention groups practiced twelve sessions of 45 min within 4 weeks of their respective training. Transfer effects were measured via an extensive battery of neuropsychological tests and questionnaires both pre-/post-training and at a 3-month follow-up.Results: WOME led to a significant improvement in working memory function, demonstrated on a non-trained near transfer task and on two different composite scores with moderate to large effect sizes. In addition, we found some indication of relevant impact on everyday life. The effects were short-term rather than stable, being substantially diminished at follow-up with only little evidence suggesting long-term maintenance. No transfer effects on other cognitive functions were observed.Conclusion: WOME is an appropriate and efficient intervention specifically targeting the working memory system in healthy older adults.Trial Registration: German Clinical Trials Register (DRKS), Identifier: DRKS00013162.

A Systematic Review on Predictors of Working Memory Training Responsiveness in Healthy Older Adults: Methodological Challenges and Future Directions

“…Only one study conducted a multi-domain WMT, as next to working memory tasks one executive control task was included within the training regime (Zinke et al, 2014 ). All training regimes were conceptualized as adaptive, except for those studies in which adaptivity was investigated as a prognostic factor for WMT responsiveness (Brehmer et al, 2011 ; Tusch et al, 2016 ; Simon et al, 2018 ; Weicker et al, 2018 ).…”

“…In total, nine studies applied digital WMT: four studies used commercially available digital WMT programs (Cogmed and WOME/ RehaCom®) (Brehmer et al, 2011 ; Tusch et al, 2016 ; Simon et al, 2018 ; Weicker et al, 2018 ), three studies used a digital n-back training (Heinzel et al, 2014a , b ; Matysiak et al, 2019 ), and two used a study–individual composition of digital WMT tasks (McAvinue et al, 2013 ; Borella et al, 2014 ). Five studies used a WMT with the Categorization Working Memory Span (CWMS) Task based on audio recordings (Borella et al, 2013 , 2017a , b , c ; Brum et al, 2018 ); however, all of these studies were conducted by the same group of researchers.…”

“…One study used Bayesian modeling approach (Borella et al, 2017b ) and one used linear mixed effect modeling (Simon et al, 2018 ), both with time as one predictor, therefore abandoning the use of change scores as dependent variable. Ten studies were factor finding studies, including a total of 18 factor finding analysis approaches: seven used a generalized linear model approach (e.g., ANOVA) (Brehmer et al, 2011 ; Zinke et al, 2012 ; Borella et al, 2014 ; Heinzel et al, 2014b ; Tusch et al, 2016 ; Brum et al, 2018 ; Weicker et al, 2018 ), one used ANCOVA (Borella et al, 2017a ), five used Pearson correlations (Brehmer et al, 2011 ; Zinke et al, 2012 ; McAvinue et al, 2013 ; Heinzel et al, 2014a ; Tusch et al, 2016 ), and one used linear regressions (Weicker et al, 2018 ) and one Linear Mixed Models (Matysiak et al, 2019 ). Three studies used a (descriptive) comparison of effect sizes (Borella et al, 2013 , 2014 ; Brum et al, 2018 ).…”

“…We defined verbal and visuospatial working memory, i.e., direct training and near-transfer effects, as primary outcomes. Most of the included studies distinguished between these two domains; however, four studies did not (Brehmer et al, 2011 ; Zinke et al, 2012 ; Simon et al, 2018 ; Weicker et al, 2018 ), and four studies addressed verbal working memory only (Heinzel et al, 2014a , b ; Tusch et al, 2016 ; Matysiak et al, 2019 ). Three of the 16 included studies (18.8%) investigated direct training effects (i.e., effects in trained tasks) only (Heinzel et al, 2014a , b ; Matysiak et al, 2019 ).…”

“…For both near- and far-transfer outcomes, baseline working memory and/or baseline performance in outcome measure was consistently found to be a negative predictor for WMT responsiveness (Zinke et al, 2012 ; Borella et al, 2017b , c ), i.e., individuals with lower performance at baseline improved more from WMT than individuals with higher baseline performance. However, for analyses on direct training effects, heterogeneous results appear, with some analyses indicating baseline working memory and/or baseline performance in outcome measure to be a positive predictor (Brehmer et al, 2011 ; Heinzel et al, 2014a ; Weicker et al, 2018 ; Matysiak et al, 2019 ), i.e., individuals with higher baseline performance in training tasks achieving higher WMT task gains than individuals with lower baseline performance. Baseline performance, as a continuous variable, was dichotomized into high vs .…”