The study was approved by the Ethics Committee of Henan Provincial People's Hospital (2018 Ethics Review No. 24) prior to conducting performance and clinical investigations, adhering to the principles outlined in the Declaration of Helsinki.

We included 94 consecutive patients [51.0 (38.0–60.0) years; 48 males] who were admitted to the Henan Provincial People's Hospital with suspected VVS. All patients underwent a tilt table test (TTT) as part of their diagnostic assessment. VVS was diagnosed based on clinical characteristics suggestive of a reflex mechanism and after excluding other differential diagnoses (5, 15). Exclusion criteria comprised individuals with a history of overt heart failure, myocardial infarction, left ventricular ejection fraction less than 50%, significant valvular heart disease, dilated cardiomyopathy, hypertrophic cardiomyopathy, sinus node dysfunction, arrhythmogenic right ventricular cardiomyopathy, Brugada syndrome, Long QT syndrome, and symptomatic orthostatic hypotension. Seventy-six healthy subjects [53.0 (44.3–62.8) years; 46 males] were enrolled as controls. They had undergone normal routine physical examinations and had never reported syncopal episodes in the past. Additionally, all subjects had normal 24-h Holter monitoring and echocardiograms.

A Holter monitor test was performed on each patient using a portable electrocardiogram device (CONTEC Medical System LTD, Qinhuangdao, China). The test measured various indices included average, fastest and slowest HR, deceleration capacity, acceleration capacity and HRV. HRV included standard deviation of normal-to-normal (NN) intervals (SDNN), standard deviation average of NN intervals (SDANN), root mean square successive difference of normal R-R intervals (RMSSD), and the percent of the number of times that the difference between adjacent normal RR intervals >50 ms in the total number of NN intervals (PNN50). The SDNN and SDANN were considered as measured of vagal and sympathetic influences, while RMSSD and PNN50 were regarded as indicators of parasympathetic nerve activity (16).

The heart rate deceleration and acceleration capacities were measured by the Holter system (5). Calculation methods as previous described: Firstly, heart beat intervals were selected as decelerating anchors when >1.00 but ≤1.05 of the preceding heartbeat interval; heart beat intervals were selected as accelerating anchors when <1.00 but ≥0.95 of the preceding heartbeat interval. Secondly, the segments of heartbeat intervals around decelerating and accelerating anchors were collected. Thirdly, the above segments were aligned at the decelerating and accelerating and the signals of segments were averaged to obtain the phase-rectified signal averaging signal X(i).

In the end, the following formula was used to quantify deceleration capacity (DC) and acceleration capacity (AC): DC/AC=[X(0)+X(1)−X(−1)−X(−2)]/4.

The values of DC are over 0, quantifying vagal nerve activity and the values of AC are less than 0, quantifying sympathetic nerve activity (17).

In our study, all patients underwent transthoracic echocardiography (TTE) using a Sonos 5500 Ultrasound machine (Philips). The following parameters measured by the M-mode technique: right atrial diameter (RAd) and left atrial diameter (Lad), left ventricular end-diastolic diameter (LVEDd), and left ventricular end-systolic diameter (LVESd). Simpson's biplane method was used to measure the left ventricular ejection fraction (LVEF).

Continuous data are presented as mean ± SD or median (25th–75th percentile). Continuous variables were compared using a Student t-test for normally distributed data or a Wilcoxon ranksum test otherwise. Categorical variables were presented as sample size (percentage) and were compared using a Pearson χ² test. multivariate logistic regression were employed to assess associations among various variables of interest with syncope. The receiver operator characteristics curve was performed to test the best cutoff value and the area under curve (AUC) of DC and AC to differentiate VVS and controls. All statistical analyses were two-sided, and a P-value < 0.05 was considered statistically significant.

There were no significant differences in gender, body mass index (BMI), echocardiogram, biochemical examinations and HR (all p > .05). DC was significantly higher (9.3 ± 2.1 vs. 7.4 ± 1.4 ms, p < .001) and AC was significantly lower (−9.3 ± 2.1 vs. −7.3 ± 1.3 ms, p < .001) in patients with VVS compared with controls. The HRV parameters including SDNN, RMSSD and PNN50 were increased in VVS group compared with controls (p < .05). Clinical characteristics of both groups are summarized in Table 1.

There are 114 participants <60 years of age (45 in the control group and 69 in the VVS group) and 56 participants ≥60 years of age (31 in the control group and 25 in the VVS group). Table 2 shows that in patients with VVS, the DC and absolute value of AC were significantly higher compared to controls, regardless of age.

However, SDNN, RMSSD, and PNN50 differed between VVS and control groups only in participants <60 years of age. The mean HR and maximum HR were higher in the VVS group compared to controls in participants <60 years of age, lower in participants ≥60 years of age. Interestingly, the DC and absolute value of AC were lower in control groups across different age groups, while there was no significant difference in DC and absolute value of AC in VVS groups across different age groups

The univariate regression analysis (Supplementary Table 1) showed that DC, AC, SDNN, RMSSD, and PNN50 were associated with VVS. However, the results of the multivariable regression analyses indicated that only DC [odds ratio = 1.746 (95% CI, 1.389–2.195); p < .001] and AC [odds ratio = 0.553 (95% CI, 0.435–0.702); p < .001] were significantly correlated with syncope (Table 3).

In participants under 60 years of age, the univariate regression analysis indicated that DC, AC, SDNN, RMSSD, PNN50, mean HR, and maximum HR were correlated with VVS (Supplementary Table 2). Furthermore, the multivariable regression analyses demonstrated that not only DC and AC, but also mean HR (odds ratio = 1.111, 95% CI 1.018–1.213; p = .000), were significantly linked to syncope (Table 4).

In participants aged 60 years and above, the univariate regression analysis showed correlations between VVS and DC, AC, mean HR, and maximum HR (Supplementary Table 3). Further multivariable regression analyses revealed that only DC and AC were significantly associated with syncope (Table 5).

ROC analysis was utilized to assess the predictive value of DC and AC in determining VVS. The AUC values for DC/AC in predicting syncope were 0.755/0.765, with sensitivities and specificities of 56.4%/60.6% and 93.4%/88.2% respectively at a cutoff of 9.4 ms/−9.0 ms.

In participants under 60 years of age, the area under the curve (AUC) for DC/AC predicting syncope was 0.743/0.758, with a sensitivity and specificity of 66.7%/50.7% and 82.2%/97.8% at the cutoff of 8.8 ms/−9.8 ms, respectively. For participants aged 60 years or above, the AUC for DC/AC predicting syncope was 0.759/0.767, with a sensitivity and specificity of 52.0%/52.0% and 100%/96.8% at the cutoff of 9.5 ms/−9.0 ms (Figures 1, 2). The AUC was similar for DC, AC, DC combined with HRV index, and AC combined with HRV index (Supplementary Table 4). In Supplementary Figure 1, the ROC curves almost completely overlap.

This study has several limitations. Firstly, there were only 94 patients with VVS included in our study, and more prospective studies are needed to investigate the association between deceleration and acceleration capacities with VVS patients. Secondly, we did not specifically exclude conditions such as diabetes, hypertension, obesity, and chronic lung disease, which might have an impact on autonomic activity. For instance, cardiovascular autonomic neuropathy (CAN) is an under-recognized yet highly prevalent microvascular complication of diabetes, affecting approximately 20% of those with the condition. However, the enrolled patients were relatively healthy, and few demonstrated the above conditions. Third, further detailed, larger sample, multicenter and longitudinal studies may be required in the future.