Ejaculatory latency is associated with autonomic balance as assessed by heart rate variability (HRV). To evaluate the association between ejaculatory latency and heart rate variability parameters in young men.

Cross-sectional study conducted between July and December 2025, including men aged 18–40 years. Age, body mass index (BMI), MELT, and HRV (SDNN, RMSSD, pNN50, LF, HF, and LF/HF) were assessed. A Polar heart rate monitor was used to obtain HRV data. Effect size was 0.24. Analyses included mean and standard deviation, Shapiro–Wilk and Levene tests, independent samples t-test, and Mann–Whitney test. The association between variables was evaluated using Pearson correlation. Binary logistic regression was performed, with variable selection based on physiological relevance and control of multicollinearity (RMSSD as a parasympathetic marker and LF/HF as an indicator of autonomic balance), and model fit assessed using the Hosmer–Lemeshow test. Predictive accuracy was analyzed using ROC curves.

A total of 40 volunteers were evaluated, of whom 26 (65%) had MELT <10 minutes and 14 (35%) >10 minutes. Only BMI and pNN50 showed normal distribution. The groups were homogeneous. Mean age was 24.54±3.14 years in the <10 min group and 24.29±2.70 years in the >10 min group. BMI was 24.45±2.96 kg/m² and 22.60±2.75 kg/m², respectively. Mean masturbatory time was 18.00±4.86 min in group 1 compared to group 2 (7.04±2.94 min). For HRV parameters, SDNN was 51.43±30.11 ms vs 56.51±19.56 ms, RMSSD 52.48±37.68 ms vs 52.94±19.71 ms, pNN50 24.45±19.55% vs 23.34±13.50%, LF 1573.84±1963.77 ms² vs 1712.78±1166.48 ms², HF 1636.53±2460.13 ms² vs 1008.50±662.81 ms², and LF/HF 1.72±2.06 vs 1.82±0.74, for the <10 and >10 min groups, respectively. Between-group comparisons using independent samples t-test (homogeneity of variances by Levene’s test) showed no significant differences for BMI (p=0.60; Δ=1.85; 95% CI: −0.86–3.80) and pNN50 (p=0.85; Δ=1.10; 95% CI: −10.77–12.99). Variables analyzed using the Mann–Whitney test also showed no statistically significant differences (p>0.05). Pearson correlation analysis showed no significant association between MELT and HRV: SDNN (r=0.092; p=0.57), RMSSD (r=0.007; p=0.96), pNN50 (r=−0.031; p=0.85), LF (r=0.039; p=0.81), HF (r=−0.149; p=0.35), and LF/HF (r=0.031; p=0.85). Binary logistic regression identified no independent predictors of ejaculatory latency. The model showed good fit according to the Hosmer–Lemeshow test (χ²=6.792; p=0.559). Age (OR=0.986; 95% CI: 0.771–1.261; p=0.911), RMSSD (OR=1.000; 95% CI: 0.978–1.023; p=0.986), and LF/HF (OR=1.184; 95% CI: 0.765–1.833; p=0.448) were not significantly associated. BMI showed a trend toward association (OR=0.779; 95% CI: 0.603–1.007; p=0.057). ROC analysis showed low discriminative ability for RMSSD (AUC=0.584), with no adequate cutoff identified. LF/HF showed modest discriminative ability (AUC=0.646), with an approximate cutoff of 1.56, sensitivity of 64.3%, and specificity of 57.7%. Similarly, SDNN showed an AUC of 0.637, with an approximate cutoff of 45.1, sensitivity of 64.3%, and specificity of 53.8%. Comparatively, LF/HF showed the best discriminative performance, followed by SDNN, while RMSSD showed the lowest accuracy.

Resting HRV was not associated with ejaculatory latency, indicating that ejaculatory control is not determined by baseline autonomic tone. LF/HF showed better discriminative performance, although modest, suggesting that autonomic balance is more relevant than isolated or global measures. These findings reinforce the predominant role of dynamic, context-dependent autonomic responses and central mechanisms.

Ejaculatory latency is not determined by resting autonomic modulation, and autonomic balance appears to be more relevant than isolated measures. These findings indicate that ejaculatory control predominantly depends on dynamic and central mechanisms, reinforcing the need for approaches based on functional assessment of autonomic response.

Continence 19S (2026) 102789
DOI: 10.1016/j.cont.2026.102789