Sleep and circadian disorders as risk factors for autoimmune disease: A population-based study

Abstract

Background: Sleep and circadian disruption have been increasingly linked to immune dysregulation, yet population-level associations with autoimmune disease remain underexplored. We examined whether delayed sleep phase disorder (DSPD), obstructive sleep apnea (OSA), primary insomnia, and hypersomnia were associated with autoimmune conditions in a large, diverse U.S. cohort.

Methods: Data were drawn from the All of Us Research Program Registered Tier Dataset v8. Participants were categorized into sleep disorder groups based on clinical diagnoses, with regular sleepers serving as controls. Autoimmune disease was defined using SNOMED-coded records. DSPD and primary insomnia were analyzed using rare disease logistic regression; OSA and hypersomnia were analyzed using 1:5 propensity score matching. Adjusted logistic regression models included age, sex at birth, race, ethnicity, income, BMI, and chronic inflammatory diagnosis. E-values assessed robustness to unmeasured confounding.

Results: All four sleep disorder groups showed significantly higher odds of autoimmune diagnosis relative to regular sleepers (p < 2.2 × 10^-16). Adjusted odds ratios were: DSPD (OR = 0.26; 95 % CI: 0.15-0.45), OSA (OR = 0.46; 95 % CI: 0.41-0.52), primary insomnia (OR = 0.46; 95 % CI: 0.41-0.52), and hypersomnia (OR = 0.48; 95 % CI: 0.46-0.50). Older age, female sex, and chronic inflammation were associated with higher autoimmune prevalence. Asian race and BMI were inversely associated with autoimmune risk; higher income was unexpectedly associated with greater autoimmune diagnosis.

Conclusions: Distinct sleep phenotypes were associated with autoimmune conditions. These associations may reflect shared or bidirectional links between sleep disruption and immune dysregulation.

Keywords: Autoimmune disease; Chronotype; Delayed sleep phase disorder; Hypersomnia; Obstructive sleep apnea; Primary insomnia.

Graphical abstract

Fig. 1

Adjusted Odds Ratios for Autoimmune Diagnosis Following Propensity Score Matching (Delayed Sleep Phase Disorder) Forest plot displaying adjusted odds ratios (ORs) and 95 % confidence intervals (CIs) for the association between delayed sleep phase disorder (DSPD) and autoimmune diagnosis, based on a rare disease case-control design with 1:5 matching of controls to cases. Logistic regression with robust standard errors (clustered by participant) was used to estimate associations, adjusting for demographic and clinical covariates. Odds ratios are plotted on a logarithmic scale. Blue points indicate statistically significant associations (FDR-adjusted p < 0.05); gray points indicate non-significant associations. Reference groups: Sleep = DSPD, Sex = Female, Race = White, Ethnicity = Hispanic or Latino, Income = <10k.

Fig. 2

Adjusted Odds Ratios for Autoimmune Diagnosis Following Propensity Score Matching (Obstructive Sleep Apnea) Forest plot showing adjusted odds ratios (ORs) and 95 % confidence intervals (CIs) for the association between OSA status and autoimmune disease diagnosis, as well as demographic covariates, following 1:5 case-control propensity score matching. Odds ratios are plotted on a log scale. Robust standard errors were used to account for potential clustering and heteroskedasticity in the matched sample. Blue points indicate statistically significant associations (FDR-adjusted p < 0.05), and gray points indicate non-significant associations. Reference groups: Sleep = OSA, Sex = Female, Race = White, and Ethnicity = Hispanic.

Fig. 3

Adjusted Odds Ratios for Autoimmune Diagnosis (Primary Insomnia) Forest plot showing adjusted odds ratios (ORs) and 95 % confidence intervals (CIs) for the association between sleep group (Regular vs Primary Insomnia) and autoimmune disease diagnosis, as well as demographic covariates, using a rare disease logistic regression model. Odds ratios are plotted on a log scale. Blue points indicate statistically significant associations (FDR-adjusted p < 0.05), while gray points indicate non-significant associations. Reference groups: Sleep = Primary Insomnia, Sex = Female, Race = White, Ethnicity = Hispanic, and Income = <10k.

Fig. 4

Adjusted Odds Ratios for Autoimmune Diagnosis Following Propensity Score Matching (Hypersomnia) Forest plot showing adjusted odds ratios (ORs) and 95 % confidence intervals (CIs) for predictors of autoimmune disease diagnosis in a propensity score–matched sample (1:5 matching). Odds ratios are displayed on a logarithmic scale. Blue points indicate statistically significant associations (FDR-adjusted p < 0.05); gray points indicate non-significant associations. Reference groups: Sleep = Hypersomnia, Sex = Female, Race = White, Ethnicity = Hispanic, and Income < $10,000.