Impact of obesity on ECG waveforms across racial/ethnic groups in the general population

Obesity is associated with distinct ECG manifestations due to altered body fat distribution, which may vary by race/ethnicity. However, few studies have evaluated these differences. We analyzed 7315 participants (mean age 59.6 ± 13.5 years; 52.4% female; 51.5% non-Hispanic White, 24.5% non-Hispanic Black, 24.0% Mexican American) from NHANES III, excluding those with prior cardiovascular disease or missing ECGs. Obesity was defined as BMI ≥30 kg/m 2 . Standard 12‑lead ECGs were recorded and processed centrally. Wave amplitudes, durations, PR, QRS, QT intervals, P, QRS, and T axes were measured automatically with visual inspection. Multivariable linear regression assessed associations between obesity and ECG parameters overall and stratified by race/ethnicity. Obesity was associated with prolonged QRS, QT, and PR durations (differences [msec] obese vs. non-obese [95% CI]: 2.17 [1.43–2.90], 1.90 [0.87–3.01], and 4.82 [3.40–6.24], respectively; all p < 0.001), and more leftward P, QRS, and T axes (differences [degrees]: −7.36 [−8.54, −6.18], −9.53 [−11.52, −7.53], −7.02 [−8.60, −5.43]; all p < 0.001). Obesity was further associated with lower R and S amplitudes and prolonged wave durations in nearly all leads, with exceptions in III and aVF. Most associations were consistent across races, though effects on S amplitude in V3/V4, R duration in aVL, and S duration in II were weaker in Mexican Americans and non-Hispanic Whites (interaction p = 0.01–0.03). Obesity influences ECG waveform amplitudes and durations, largely independent of race/ethnicity. These findings support non–race-specific ECG interpretation while highlighting potential limitations of amplitude- and duration-based criteria.