Exploring the association between heart rate variability and self-efficacy: A secondary analysis of a cross-sectional pilot study

Background General self-efficacy (SE) may be associated with autonomic functioning, but evidence from healthy community samples remains limited. This exploratory study examined the cross-sectional associations between general SE and resting heart rate variability (HRV) and assessed whether SE explained HRV variance after adjustment for demographic characteristics. Methods This secondary analysis used data from 97 healthy Chinese adults who participated in a cross-sectional pilot study. General SE was assessed using the Chinese General Self-Efficacy Scale, and resting HRV was recorded for five minutes using an ear-lobe photoplethysmography device. Spearman rank correlations were used to examine unadjusted associations between SE and 10 HRV indices, with false discovery rate correction for multiple comparisons. Principal component analysis was subsequently used to reduce redundancy among the HRV indices. Two-block hierarchical linear regression models examined whether SE explained incremental variance in the resulting HRV components after adjustment for age group and sex. Results In the unadjusted analyses, SE showed small inverse correlations with low-frequency power, SDNN, RMSSD, and mean inter-beat interval and a weak positive correlation with mean heart rate. Principal component analysis identified three components that collectively explained 84.9% of the variance in the HRV indices. After adjustment for age group and sex, SE did not explain a statistically significant increase in variance in any HRV component, contributing 0.3% to 2.4% of additional variance. Age group was significantly associated with two of the three components, whereas sex was not significantly associated with any component. Conclusions Weak associations between SE and several resting HRV indices were observed in the unadjusted analyses, but these associations were attenuated in the covariate-adjusted component models. Demographic differences, particularly age, may partly account for the observed bivariate associations. Given the small convenience sample, cross-sectional design, and potential residual confounding, these findings should be considered exploratory and require confirmation in larger studies using physiological measurements.