The Impact of Transcranial Direct Current Stimulation Combined with Retrieval Practice on Memory Retention in Healthy Older Adults

We hypothesize that older adults' memory performance will be significantly better with anodal tDCS than with sham or no stimulation. Additionally, RP is expected to outperform repeated learning, with anodal tDCS in the RP condition leading to better recall organization than in the relearning group.This study employed a randomized, single-blind, sham-controlled experimental design.A total of 176 older adults were initially recruited by elderly communities in Xixia District, Yinchuan City, Ningxia Province, the final sample included 158 participants (74 males), with an average age of 67.63±7.62 years. All participants had a Mini-Mental State Examination score of 26 or higher. The design was a 3 (stimulation type: anodal stimulation, sham stimulation, no stimulation) × 2 (learning strategy: retrieval practice, repeated learning) between-subjects factorial design. Both stimulation type and learning strategy were between-subject variables. The dependent variables were the correct recall rate, the number of correctly recalled words, and the adjusted ratio of clustering (ARC) scores at different testing stagesThe data were analyzed using SPSS 27.0.1and R. Free recall performance from the initial and final tests was calculated across three stimulation conditions and two learning methods. For the RP)group, the joint probability of free recall performance during the two initial test stages was computed. Additionally, recall organization was measured using the Adjusted Ratio of Clustering (ARC) score , which ranges from -1 to 1. An ARC score of 0 indicates chance-level clustering, while a score of 1 reflects perfect clustering. Negative ARC scores, indicating atypical recall patterns , were excluded from the analysis.ANOVAs and ANCOVAs were used to analyze the number of correctly recalled words, recall accuracy, and ARC scores for the two learning strategies across the three stimulation conditions. All tests were conducted with an alpha level of 0.05. Where interactions were detected, post-hoc t-tests with Bonferroni corrections were applied. Effect sizes were reported as Cohen’s d for t-tests and partial eta-squared for ANOVAs. In cases of non-significant results, Bayesian ANOVAs were performed to assess the evidence strength for the null hypothesis, with Bayesian Factor (BF01) values greater than 3 supporting the null hypothesis, and values below 0.33 supporting the alternative hypothesis . Additionally, mediation analyses were conducted to explore whether recall organization mediated the relationship between learning method or stimulation type and free recall performance in the final test. Relationships among anodal tDCS, PANAS mood changes, and memory retrieval were also examined. Mediation models were tested using the lavaan package in R, with direct and indirect effects estimated via bootstrap methods using 5,000 bootstrap samples to compute coefficients and 95% confidence intervals.