Abstract: To investigate the effects of late-life foreign language learning on executive function (EF), a three-level meta-analysis was conducted on 20 studies (63 effect sizes) identified through rigorous systematic search criteria. The results showed that late-life foreign language learning produced a small but significant improvement in EF (g = 0.37, p < 0.001). However, when isolating the specific effects of language learning through high-quality randomized controlled trials with active controls, the effect size iminished to a non-significant level (g = 0.11, p = 0.119). This disparity implies that language‑specific cognitive gains are limited, and that previously observed gains may stem largely from general cognitive stimulation rather than language learning per se. Moderation analyses showed that the improvement effect on EF was significantly moderated by linguistic distance, with greater distance associated with larger cognitive gains. Exploratory analyses suggested a greater benefit for older adults with cognitive impairment than for healthy peers, although this finding is based on only three studies and awaits further confirmation. Conversely, the moderating effects of learning duration, learning intensity, control group type, and specific EF sub-components were non-significant. Furthermore, the findings provide stronger empirical support for the processing complexity hypothesis over the interference inhibition hypothesis, aligning more closely with the Bilingual Executive Processing Advantage (BEPA) hypothesis rather than the Bilingual Inhibitory Control Advantage (BICA) hypothesis. This study provides a basis for developing low-cost, non-pharmacological cognitive interventions and establishes a foundation for exploring language learning and its cognitive effects across the lifespan.

Key words: older adults, foreign language learning, executive functions, cognitive enhancement, meta-analysis