Sleep characteristics and risk factors in patients of insomnia comorbid with anxiety

BACKGROUND: Insomnia is a common sleep disturbance that often co-occurs with anxiety. While insomnia is a known risk factor for anxiety, anxiety can also exacerbate insomnia. This hypothesis-driven study aimed to: (1) test whether patients of insomnia comorbid with anxiety exhibit greater subjective-objective sleep discrepancy than those with insomnia alone; and (2) identify risk factors for comorbidity and predictors of anxiety severity. METHODS: A total of 221 patients with insomnia were recruited and divided into an anxiety group (n = 163) and a non-anxiety group (n = 58) according to the Beck Anxiety Inventory (BAI) ≥ 10 points or < 10 points. Subjective sleep and emotional states were assessed using the Pittsburgh Sleep Quality Index (PSQI), Insomnia Severity Index (ISI), Beck Depression Inventory (BDI), and BAI. Objective sleep parameters (e.g., total sleep time, sleep efficiency) were monitored via 7-day actigraphy (ACT). Differences and correlations among subjective scales and objective indicators were analyzed using Logistic regression and Spearman correlation analyses to identify sleep characteristics and risk factors for insomnia comorbid with anxiety. RESULTS: The anxiety group showed significantly higher PSQI, ISI, BDI, and BAI scores compared to the non-anxiety group (all P < 0.001). Act data indicated that the anxiety group spent more time in bed (P = 0.017), but no significant differences were observed in total sleep time, sleep efficiency, sleep onset latency, or wake time after sleep onset (P > 0.05). The anxiety group also exhibited more severe sleep disturbances and daytime dysfunction in PSQI subitems (P < 0.05). Spearman correlation analysis revealed positive correlations between BAI grade and PSQI, ISI, and BDI grades (all P < 0.05). Both univariate and multivariate binomial logistic regression analyses indicated that a disease duration of 36–120 months significantly increased the risk of insomnia comorbid with anxiety (P < 0.05). Ordinal logistic regression further identified sleep efficiency (P = 0.008), sleep disturbances (P