OSAS.MAP10 and All-Cause Mortality Risk in Hypertensive Patients: A Retrospective Cohort Analysis of NHANES

Background and Research QuestionHypertension and obstructive sleep apnea (OSA) commonly coexist in older adults. OSAS.MAP10 is a symptom-based screening score that estimates OSA risk and is not a polysomnographic measure of disease severity. We examined its association with all-cause mortality, the primary outcome, and explored nonlinear patterns among adults with hypertension.MethodsWe analyzed 6,930 adults with hypertension from NHANES 2005-2008 and 2015-2018. The primary outcome was all-cause mortality; cardiovascular and cancer mortality were secondary exploratory outcomes. We used survey-weighted Cox models and restricted cubic splines adjusting for age, sex, race and ethnicity, BMI, education, family poverty income ratio, smoking status, diabetes, baseline cardiovascular disease, current antihypertensive medication use, and survey period. In adults aged ≥60 years, the age-stratified model without continuous age was primary, and additional continuous-age adjustment was evaluated as a sensitivity analysis.ResultsAmong 6,930 hypertensive adults, 1,199 deaths occurred. In the complete-case analysis of participants aged ≥60 years (n = 3,982; 1,006 deaths), the association between OSAS.MAP10 and all-cause mortality was nonlinear (P for nonlinearity <0.001). In the age-stratified model without continuous age, each 1-point increase at or above 6.19 was associated with higher mortality (HR 1.21; 95% CI 1.02-1.44; P = 0.029). After additional adjustment for continuous age, the corresponding association was attenuated (HR 0.96; 95% CI 0.83-1.12; P = 0.643). Cardiovascular mortality did not show significant nonlinearity.ConclusionsHigher OSAS.MAP10 scores showed a nonlinear association with all-cause mortality among hypertensive adults aged ≥60 years. The increase above approximately 6.2 was observed only in the age-stratified model without continuous age and was attenuated after additional age adjustment. This data-derived inflection point is hypothesis-generating and requires validation in an independent cohort before clinical application.