Integration of lung function in allostatic load scoring and its impact on mortality prediction

Allostatic load refers to the physiological "wear and tear" that results from adaptation to stressors over the lifespan. In this study, we integrated lung function (LF) parameters into the calculation of the allostatic load score (ALS) to evaluate changes in its performance for predicting all-cause mortality. Data from 8,775 participants (aged 25-82 years; 52% female) who participated in the first wave of the Austrian LEAD cohort were used. "ALS without LF" was calculated using 12 parameters, including cardiovascular, metabolic, body composition, and bone mineral density measures. Z-scores of forced expiratory volume in 1 s (FEV₁) and forced vital capacity (FVC) were integrated with the aforementioned parameters for the calculation of "ALS with LF". Participants scored 1 point for each at-risk marker and 1 point for medication use for hypertension, diabetes, and dyslipidaemia. The total points constituted the ALS. Participants were followed-up for death for an average of 7.7 years. The association of ALS, with and without LF, with mortality was investigated using Cox proportional hazards models. ALS increased with age and was higher in males compared to females across all age categories. ALS (as a continuous variable) with LF [hazard ratio (HR): 1.19 (95% confidence interval (CI) 1.14-1.24)] and without LF [HR: 1.16 (95% CI 1.11-1.21)] showed a significant association with mortality in sex- and age-adjusted models. The adjusted models incorporating ALS as a categorical variable showed that individuals with high ALS, with [HR: 2.44 (95% CI 1.41-4.20)] and without [HR: 2.47 (95% CI 1.46-4.80)] LF parameters, had a higher risk of mortality compared to those in the low ALS group. Cox models incorporating "ALS with LF" parameters exhibited higher concordance index and R² values, along with a lower Akaike's information criterion indicating superior predictive power compared to models that included "ALS without LF". ALS is strongly associated with mortality, with higher ALS linked to an increased risk of mortality across both continuous and categorical analyses. Models that incorporate ALS with LF parameters demonstrated superior predictive performance and greater robustness, underscoring the added value of including LF in the models.