Association between social determinants of health and asthma: a cross-sectional analysis based on NHANES data

This study examines the link between social determinants of health (SDoH) and asthma, utilizing data from the National Health and Nutrition Examination Survey (NHANES).

We analyzed 39,340 participants, including 5,645 with asthma and 33,695 without. Multivariable logistic regression assessed the SDoH-asthma relationship, with subgroup analyses for effect modifiers. Machine learning models combining demographic, clinical, and SDoH variables were evaluated using area under the curve (AUC).

Asthma was more prevalent among younger individuals (mean age: 45.11 years), females (58%), and those with significant differences in BMI, education, marital status, and race (all p < 0.001). Unadjusted models showed a 5.5% increased asthma risk per SDoH index unit (OR = 1.055, p < 0.001), remaining significant after demographic and clinical adjustments (Model 2: OR = 1.045, Model 3: OR = 1.039). Stronger SDoH-asthma associations were found among smokers (OR = 1.08) and diabetics (OR = 1.12), but not in participants with higher education (OR = 1.00). Model 3, including SDoH variables, demonstrated superior predictive performance (training AUC = 0.779) with minimal generalizability loss (ΔAUC = 0.206).

SDoH is an independent risk factor for asthma, particularly among smokers, diabetics, and individuals with less education. Incorporating SDoH into predictive models enhances performance and offers insights for clinical and policy interventions.