P-wave axis abnormalities in the UK Biobank: prevalence, associated outcomes, and genetic component

Abstract Background The P-wave axis represents the dominant conduction vector of atrial depolarization on the electrocardiogram (ECG). Previous research has suggested P-wave axis abnormalities as a predictor of atrial fibrillation (AF). The genetic underpinnings of the P-wave axis and associations with atrial remodelling remain unclear. Purpose The study aimed to determine the prevalence of P-wave axis deviations in a large general population cohort and study the associations between abnormal P-wave axis and incident AF and atrial remodelling, and to elucidate a potential genetic component of the P-wave axis. Methods European ancestry individuals with P-wave axis measures on their ECGs were included. Individuals with a history of AF, heart failure or cardiac conduction disorders prior to ECG were excluded. Abnormal P-wave axis was defined as <0 degrees (left axis deviation), and >75 degrees (right axis deviation). Hazard ratios (HRs) for AF were calculated using multivariable Cox-regression adjusted for sex, age, body-mass index (BMI), diagnosed hypertension, coronary artery disease, and diabetes. Cubic spline curves were drawn using a penalized smoothing model, centered at a P-wave axis of 35 degrees. Measures of indexed left atrial (LA) volume and LA emptying fraction were derived from cardiac magnetic resonance imaging (cMRI), and associations were examined using linear regressions adjusted for sex, age and BMI. Genome-wide associations were examined for degrees of P-wave axis deviation from normal range, using regenie. Results Among 36,708 included individuals, 1,284 (3.5%) had left P-wave axis deviation, while 3,210 (8.7%) had right P-wave axis deviation. Median age at ECG recording was 65.3 years (interquartile range: 58.8-70.7 years), and 19,246 (52.4%) were female. During a median follow-up period of 4.5 years, 725 individuals (2.0%) were diagnosed with incident AF. An abnormal P-wave was associated with rates of AF (HR: 1.37, 95% confidence interval [CI]: 1.13-1.66, P=0.002). Considered seperately, left-axis deviation was associated with a 1.57 HR for AF(95% CI: 1.13-2.18, P=0.007), while right-axis deviation was associated with a HR of 1.29 for AF (95% CI: 1.03-1.63, P=0.028). In a sub-cohort with cMRI measures (n=27,450), abnormal P-wave axis was associated with a higher LA volume (β=0.54 ml/m2, P=0.007), and lower LA emptying fraction (β=0.92%, P=7.87x10-9). Genome-wide association analyses revealed four significant genetic loci associated with degrees of axis deviation from the normal range, including one near the CACNB2 gene, encoding an L-type calcium channel subunit. Conclusions In a large biobank, abnormal P-wave axis on ECG was associated with higher rates of AF, increased LA volume and decreased LA function. These findings support P-wave axis abnormalities as a risk marker for future AF and indicate that deviations from a normal P-wave axis may serve as an indirect marker of atrial remodelling and myopathy.Associations with AF and LA measures Manhattan plot of genetic associations