Age-related modification of the association between arterial blood pH and intensive care unit mortality: a nationwide retrospective observational study using the Japanese Intensive Care Patient Database (JIPAD)

The lowest arterial blood pH within the first 24 h of intensive care unit (ICU) admission is a widely used prognostic indicator in critically ill patients and is incorporated into multiple established prognostic models. Although older adults represent a growing proportion of ICU admissions and have reduced physiological reserves, it remains unclear whether age modifies the prognostic association between arterial blood pH and ICU mortality. We aimed to evaluate this association and to examine whether it differed by patient age. We conducted a nationwide retrospective cohort study using data from the Japanese Intensive Care Patient Database (2018–2023), including adult patients (≥ 18 years) admitted to participating ICUs. Primary exposure was defined as the lowest arterial blood pH recorded within the first 24 h of ICU admission, and the primary outcome was ICU mortality. Associations were evaluated using multivariate Cox proportional hazard models with restricted cubic splines adjusted for clinically relevant covariates and stratified by institution. Effect modification by age (18–64 vs ≥ 65 years) was assessed using an interaction term, and an additional model with continuous spline-transformed age was used to examine dose–response relationships. Among 324,806 ICU admissions, 215,880 (66.5%) were aged ≥ 65 years; ICU mortality was 4.4%. Lower arterial pH was nonlinearly associated with higher ICU mortality, with a stronger association in patients aged ≥ 65 years than in those aged 18–64 years (p for interaction < 0.001). At pH 7.0 versus 7.35–7.45, the aHR for ICU mortality was 3.40 (95% CI, 3.07–3.76) in patients aged 18–64 years and 4.73 (95% CI, 4.33–5.16) in those aged ≥ 65 years. Continuous-age modeling supported this age-related modification across the age spectrum. A lower arterial blood pH was associated with higher ICU mortality, and this association was modified by age. The prognostic impact of acidemia was more pronounced in older patients, suggesting that the interpretation of the lowest arterial blood pH may benefit from consideration of patient age. Further studies are needed to evaluate whether age-specific approaches to acid–base assessment improve prognostic accuracy in patients with critical illness.