Sports-Based Cognitive Game Program and Attention in Athletes

Design: Single-center, two-arm, parallel-group randomized controlled trial with pre-test and post-test measurements, conducted at Düzce University, Faculty of Sport Sciences (Türkiye), between December 2025 and March 2026. The trial was registered retrospectively after data collection had been completed.

Participants and allocation: One hundred physically active university students aged 18-23 years without a neurological or psychiatric diagnosis in the previous six months were enrolled after giving written informed consent. After baseline measurements, participants were randomly assigned by one member of the research team using computer-generated random allocation (1:1) to the experimental group (n = 50) or the control group (n = 50). Allocation concealment by a third party was not used, and participants and assessors were not blinded to group allocation.

Intervention: The experimental group took part in a basketball-based sports cognitive game development program delivered twice weekly for 12 weeks (24 sessions of 45 minutes) by the same researcher. The program had five phases with progressively increasing cognitive load: (1) adaptation and basic cognitive infrastructure (weeks 1-2), (2) attention and reaction development (weeks 3-4), (3) decision making and strategy (weeks 5-8), (4) concentration and multitask performance (weeks 9-10), and (5) reinforcement and transfer to sport (weeks 11-12). Each session included greeting and orientation (3 min), warm-up (5 min), sport-based cognitive games (15 min), table-based cognitive games (15 min), and evaluation and closing (7 min). Difficulty was increased for the whole group according to a predefined plan (number of stimulus-response rules, rule reversals and no-go signals, working-memory load, time pressure and dual tasking). Delivery was documented with structured session control forms, and attendance was recorded at every session. Participants who were absent from more than two of the 24 sessions were excluded from the analysis (n = 5), so 95 participants (experimental 45, control 50) were analyzed.

Control condition: The control group received no intervention and continued their daily lives without contact with the research team until the post-test.

Outcome measures: (1) d2 Attention Test (Turkish adaptation), TN-E score = total number of items processed minus total errors (omission + commission); higher scores indicate better performance. (2) Resting EEG theta/beta ratio at Fz (linked-earlobe reference), recorded with a NeXus-10 Mk II amplifier and BioTrace+ software during a 10-min eyes-closed rest (256 samples/s; first and last 30 s discarded; automatic artifact rejection at ±100 µV; at least 8 min of artifact-free data; FFT with 2-s Hanning-windowed epochs and 75% overlap; theta 4-8 Hz, beta 13-21 Hz; ratio of absolute powers). The theta/beta ratio is treated as an exploratory neurophysiological measure. Both outcomes were assessed at baseline and after 12 weeks; the d2 test and EEG were performed on the same day, d2 first, separated by at least 30 min.

Statistical analysis: Each outcome was analyzed with a 2 (group) x 2 (time) mixed-design ANOVA, with the Group x Time interaction as the primary test. Simple effects were examined with paired and independent/Welch t-tests with Bonferroni correction (three comparisons per outcome, alpha = .0167). Effect sizes are reported with 95% confidence intervals (partial eta squared, Cohen's d, and the pre-post-control effect size dppc2 with bootstrap confidence intervals). As a supplementary analysis, ANCOVA was performed on post-test scores with pre-test score as covariate and gender as a fixed factor. The a priori sample size (128 participants) was not reached; a sensitivity analysis indicated a minimum detectable between-group effect of d = 0.58 for the analyzed sample.

Ethics: Approved by the Düzce University Scientific Research and Publication Ethics Committee (No. E-78187535-050.04-426788).