Secular temporal shifts in alexithymia among Chinese college students (2003-2024): a cross-temporal meta-analysis

Alexithymia constitutes a key indicator for assessing mental-health status among college students. Employing cross-temporal meta-analytic methods, the present study synthesizes repeated cross-sectional empirical data spanning 2003 to 2024 to map secular patterns of alexithymia in Chinese undergraduate samples and characterize observational temporal covariation between lagged aggregated macrosocial indicators and alexithymia-related traits. In total, 140 eligible studies involving 141723 undergraduate participants were incorporated; analyses were restricted strictly to datasets collected using the fully validated Chinese adaptation of the TAS-20 scale. Extensive between-study heterogeneity was observed for all alexithymia-related outcome measures. Univariate meta-regression detected significant positive temporal covariation of total alexithymia, Difficulty Identifying Feelings (DIF), and Difficulty Describing Feelings (DDF) with data-collection years, whereas Externally Oriented Thinking (EOT) exhibited no discernible temporal covariation. Multiple lagged macrosocial indicators showed bivariate observational covariation with elevated alexithymia scores. Multivariate adjusted meta-regression further revealed differentiated correlational profiles linking economic predictors to distinct alexithymia subdimensions. Supplementary sensitivity tests and publication-bias assessments were implemented to evaluate the robustness of these observational patterns, alongside explicit acknowledgment of methodological limitations intrinsic to observational meta-analytic research. Aggregated findings document secular increasing patterns in emotion-identification and emotion-description deficits across pooled college-student samples, as well as observational temporal covariation with multifaceted macrosocial transformations. The present results reflect population-level correlational covariation and provide no evidence for causal connections or intra-individual longitudinal shifts.

This cross‑temporal meta‑analysis is observational social‑science evidence‑synthesis and falls outside the eligibility scope of the PROSPERO registry, which mainly accepts health‑intervention‑focused systematic reviews. The study protocol, full extracted dataset and reproducible analytical R‑scripts have been publicly archived on the Open Science Framework (OSF). Public accessible project URL: https://osf.io/4d5pt/.