Heart rate variability associated with cardiometabolic risk factors in Ghanaian adults with hypertension: A quantile regression analysis

Background Obesity modifies cardiovascular risk in hypertension, but associations with cardiac autonomic regulation may be heterogeneous and missed by mean-based analyzes. We assessed whether obesity relates to short-term heart rate variability (HRV) among hypertensive adults in Ghana using quantile regression, a distribution-aware approach.Methods In a cross-sectional study at Ejisu Government Hospital, Ghana, adults (≥18 years) with hypertension were classified as obese (body mass index; BMI ≥ 30 kg/m2) or non-obese. Five-minute resting ECG recordings were used to derive relevant HRV parameters. We compared groups using ordinary least-squares, rank-based regression, and quantile regression.Results Among 223 participants, 79 (35.4%) were obese and 144 (64.6%) were non-obese. Dimensionality-reduction showed substantial overlap between obesity groups without distinct HRV clustering. Mean-based models were largely null, whereas quantile regression revealed marked distributional heterogeneity. Diabetes was consistently associated with lower HRV across multiple indices, with the strongest effects in lower quantiles of SDNN, rMSSD, and HF power distributions. BMI-defined obesity showed minimal association with HRV across quantiles, while high-risk waist circumference showed selective quantile-specific associations including lower rMSSD and HF power and higher LF/HF ratio, particularly in upper-middle portions of the LF/HF distribution. MAP showed little evidence of association with HRV across quantiles.Conclusions In hypertensive adults in Ghana, HRV profiles were heterogeneous and not clearly partitioned by BMI-defined obesity. Diabetes and central adiposity showed distribution-specific associations with HRV that were partly missed by mean-based analyzes, although several adiposity-related findings remained sensitive to alternative adjustments.