Association of P-wave duration with all-cause and cardiovascular mortality in MASHAD cohort study

P wave indices obtained from the electrocardiogram (ECG) serve as indicators of atrial conduction. P-wave duration (PWD) has been associated with various pathological conditions. The relationship between this factor and cardiovascular mortality remains insufficiently explored. To ascertain the association of PWD and cardiovascular and all-cause mortality in the MASHAD cohort study population.

Participants were recruited from the Mashhad Stroke and Heart Atherosclerotic Disorder (MASHAD) cohort study. Participants' demographic information were recorded by checklist. Additionally, a comprehensive medical history was collected to identify cardiovascular risk factors, and history of pre-existing cardiovascular disease (CVD). A 12‑lead electrocardiogram (ECG) was obtained from the participants and they were subsequently categorized into two groups based on their PWD (≥ 120 msec and < 120 msec). All participants were followed for a minimum of 10 years for mortality assessment. We employed Cox regression to evaluate the relationship between PWD and cardiovascular and all-cause mortality. In order to evaluate the incremental predictive value of P wave duration, we calculated the net reclassification improvement (NRI) and integrated discrimination improvement (IDI) at 10-year follow-up timepoint. Subsequently, we developed Kaplan-Meier plots, time-dependent ROC curves, and restricted cubic splines (RCS) to thoroughly evaluate the association between PWD and cardiovascular mortality. Subgroup analysis was conducted to evaluate the impact of PWD across various distinct groups.

A total of 8885 participants were enrolled in the study. Prolonged PWD was associated with higher all-cause, cardiovascular mortality in unadjusted model. After controlling for covariates, PWD remained significantly associated with all-cause and cardiovascular mortality in continuous form of it. However, PWD in categorical form did not show significant association with cardiovascular mortality. The addition of P wave duration to the conventional model did not provide statistically significant improvement in risk reclassification or discrimination for either all-cause or cardiovascular mortality. The RCS models indicated that the PWD displayed a positive linear correlation with the risk of all-cause mortality across all models. The time-dependent ROC curve demonstrated favorable predictive performance for the adjusted models at both the 5-year and 10-year timepoints. The subgroup analysis revealed no interaction among the subgroups in our statistically significant multivariable models (all p for interaction >0.05).

Continuous P-wave duration predicted all-cause and cardiovascular mortality, whereas the conventional 120 ms cut-off showed limited prognostic utility.