Shang · BMC public health 2026 · cross-sectional study · n=10,778

Prevalence, subtypes, and comorbidity of DSM-5 insomnia disorder among adults in Beijing, China: a large-scale cross-sectional study.

Cited 0 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional survey with standardized diagnostic interviews

PubMed 41639643 · doi:10.1186/s12889-026-26453-x · record verified 2026-08-30

What was done

A community-based cross-sectional study evaluated the prevalence, subtypes, sociodemographic correlates, and psychiatric comorbidities of insomnia among adults in Beijing, China, between October and December 2021. Using multistage stratified random sampling, 10,778 adults were recruited. Trained psychiatrists administered standardized diagnostic interviews based on DSM-5 criteria to assess insomnia disorder and comorbid mental disorders. Data were analyzed using weighted descriptive statistics and Rao-Scott chi-square tests.

What was found

The weighted prevalence was 17.7% for subjective sleep problems, 10.7% for clinically assessed insomnia symptoms, and 3.0% for DSM-5 insomnia disorder. Initial insomnia was the most common subtype for symptoms (8.1%) and disorder (2.6%). Insomnia disorder prevalence was higher in females, adults aged 60 years or older, and those with lower education. Among individuals with insomnia disorder, 31.3% had comorbid mental disorders (13.4% alcohol-related disorders). Among individuals with other mental disorders, 14.4% had insomnia disorder, peaking in depressive (21.4%) and anxiety disorders (19.2%).

Why it matters

This study highlights the substantial divergence between self-reported sleep complaints and clinician-diagnosed DSM-5 insomnia disorder in a large urban population. It emphasizes that while formal insomnia disorder is relatively low in prevalence, it carries a heavy psychiatric comorbidity burden.

Limits

The cross-sectional design cannot establish causality or the temporal order between insomnia and psychiatric comorbidities. Findings from an urban Beijing sample may not generalize to rural or other regional populations. Exact confidence intervals and multivariable-adjusted odds ratios are not reported in the abstract.

Cited by