Abstract Rationale Obstructive sleep apnea (OSA) is associated with impaired vigilance and higher co-morbid burden, but large-scale data examining its relationship with occupational absenteeism is limited. Understanding whether OSA severity is associated with increased absenteeism may clarify the economic impact of this condition and inform policy. Methods We analyzed 327,894 adults (≥18y) who underwent diagnostic sleep testing within Kaiser Permanente Southern California between 2016-2024. Patients were categorized by apnea-hypopnea index (AHI4%): no OSA (0-4.9), mild (5-14.9), moderate (15-29.9), and severe (≥30 events/h). Descriptive characteristics (age, sex, race/ethnicity, BMI, Charlson/Elixhauser scores, geocoded education/income, office-visit frequency) were compared. Off-work orders and total days off were assessed during the one-year period prior to the sleep study, extracted from electronic health records (Epic Systems). Because most patients had zero off-work days, a two-part (hurdle) model was used: logistic regression for any off-work days and negative binomial regression for the total days off among those with non-zero days. Models included OSA category (reference=no OSA) and covariates (age, sex, BMI, Elixhauser, office visits, education, income). Results Baseline characteristics were compared for the four groups (Table): No OSA (26.3%); Mild (29.3%); Moderate (20.3%); Severe (24.2%). Greater OSA severity was associated with older age, male sex, higher BMI, and greater comorbidity burden. A logistic regression model revealed a higher likelihood of any off-work days for patients with OSA versus no OSA: adjusted odds ratios were 1.17 (95 % CI 1.13-1.22) for mild, 1.11 (1.06-1.17) for moderate, and 1.14 (1.08-1.19) for severe OSA (all p 0.001). Adjusted probabilities of any off-work days were 8.2% for no OSA and 9.1-9.4 % for OSA (without a dose-response relationship with OSA severity). Among the 20,762 patients with non-zero off-work days, OSA versus no OSA was not significantly associated with the number of days absent (Incidence Rate Ratio IRR 0.87 0.66-1.15 mild; 0.75 0.56-1.01 moderate; 0.76 0.56-1.04 severe). Greater comorbidity (Elixhauser per point, IRR 1.08), higher healthcare utilization (office visits, IRR 1.06), male sex (IRR 1.21), and older age (per year, IRR 1.02) were associated with longer absences (all p 0.05). Conclusions The relationship between OSA and work absenteeism was modest. OSA was independently associated with a higher likelihood of any work absence but without a dose-response relationship with the total days of absence. Longer absences were driven primarily by demographic and comorbidity factors rather than OSA severity. This abstract is funded by: National Heart, Lung, and Blood Institute of the National Institutes of Health under award number R01 HL161253-01A1, Data-Driven Sleep Biomarkers of Brain Health, Heart Health, and Mortality.
Hwang et al. (Fri,) studied this question.