Problematic Social Media Use and Insomnia Severity Among Saudi University Students: Cross-Sectional Indirect-Association Analyses Involving Fatigue and Anxiety

Purpose: This study sought to examine cross-sectional associations of problematic social media use with insomnia severity and to estimate indirect associations involving fatigue and anxiety under a hypothesized statistical ordering among Saudi university students. Methods: A cross-sectional online survey of 430 Saudi undergraduate social media users used the Bergen Social Media Addiction Scale, Fatigue Severity Scale, Depression Anxiety Stress Scales anxiety subscale, and Athens Insomnia Scale. Gender and daily social media-use duration were included as covariates in the primary observed-score regression path models; age was examined as a sensitivity covariate. Bootstrapped indirect associations, a paired-bootstrap contrast between two alternative concurrent-mediator specifications, and ordinal confirmatory-factor-analysis sensitivity analyses were conducted. These analyses do not establish temporal order. Results: Problematic social media use was associated with fatigue (B = 0.161, p < 0.001), anxiety (B = 0.388, p < 0.001), and insomnia severity (B = 0.161, p < 0.001) after adjustment. In the hypothesized fatigue-first decomposition, the fatigue-to-anxiety indirect association was B = 0.072, 95% CI [0.045, 0.105]. Its paired numerical difference from the reverse-order decomposition was B = 0.045, 95% CI [0.013, 0.083]. The AIS two-factor measurement sensitivity model fit better than the one-factor model. Daily social media-use duration was not robust across sensitivity analyses. Conclusions: Problematic social media use, fatigue, anxiety, and insomnia severity were concurrently associated in this sample. The indirect-effect decompositions are statistical descriptions of cross-sectional covariation, not evidence of a temporal, causal, or clinical sequence; longitudinal and experimental studies are needed.