AI Insight
This study used ecological momentary assessment over 14 days with 60 Chinese university students in sport-related majors to examine how physical, academic, and mental health states relate to each other in daily life. Researchers found that physical strain (soreness and fatigue) predicted subsequent changes in other states, while tension, anxiety, and irritability were most strongly connected to other mental states in the moment. Poor sleep quality predicted elevated levels of all eight measured states (including fatigue, academic stress, depressed mood, and concentration difficulty) the following morning, even after accounting for individual differences and previous levels.
Why it matters
The findings suggest that monitoring physical recovery and sleep quality may be important for supporting the mental health and academic performance of student-athletes. Understanding these day-to-day connections could inform interventions targeting specific states like rumination or tension that appear central to students' momentary well-being.
Understand the Science
University students enrolled in sport-related majors may simultaneously manage academic requirements and regular training or competition, yet the short-term organization of their physical, academic/performance-related, affective, and cognitive states remains insufficiently understood. This study used a 14-day ecological momentary assessment (EMA) design with three smartphone assessments per day and multilevel vector autoregressive modeling. The original recruitment pool included 72 students from one university in Chongqing, China. Five participants were excluded for repeated responses outside the designated EMA windows, and seven were excluded for broader protocol or response-quality failures, including discontinuation of EMA completion and/or repeated invariant, non-informative response patterns. The retained analytic sample comprised 60 students and a balanced panel of 2,520 observations. Eight brief, study-specific momentary indicators were modeled: fatigue, soreness/bodily pain, academic stress, performance worry, depressed mood, tension/anxiety/irritability, rumination, and concentration difficulty. The revised primary temporal model used population-level fixed lagged slopes and showed no singularity or convergence problems. Soreness and fatigue showed the largest temporal out-strength, whereas the tension/anxiety/irritability indicator showed the largest temporal in-strength. In the contemporaneous network, the tension/anxiety/irritability indicator and rumination showed the highest expected-influence and bridge-centrality estimates. Participant-level bootstrap and case-dropping analyses indicated high centrality stability within the retained sample, although between-person edge estimates were substantially less precise. Sensitivity analyses showed that the broad outgoing and contemporaneous structure was comparatively robust to measured training, academic, and time-of-day context, whereas several incoming temporal relations, particularly those involving academic stress, were context-sensitive. Poorer-than-usual sleep remained prospectively associated with higher next-morning levels of all eight states after adjustment for participants’ mean sleep, the previous-evening value of the same outcome, study day, and participant-level clustering. These findings characterize conditional associations among the specific momentary indicators measured in this single-university sample; they do not establish causal pathways, comprehensive dual-career functioning, or intervention targets.