PR Interval and the Risks of Recurrent Syncope and All‐Cause Mortality in Patients Presenting With Syncope

Authors: Asger Knudsen, Johannes Jan Struijk, Sam Riahi, Mikkel Porsborg Andersen, Helle Collatz Christensen, Christian Torp‐Pedersen, Kristian Kragholm, Christoffer Polcwiartek, Jørgen K. Kanters, Claus Graff

Abstract

Background

Risk assessment of patients with syncope does not consider a short PR interval despite its association with increased risk of atrial fibrillation and all‐cause mortality. This study aimed to explore the association between the PR interval and all‐cause mortality and recurrent syncope in patients admitted to the hospital with syncope.

Methods

We included patients with a diagnosis of syncope and an ECG recorded within 24 hours of hospital admission from the Danish Nationwide Electrocardiogram Cohort and divided patients into short (<120 ms), normal (120–200 ms), or long PR interval (>200 ms). Patients with ECG abnormalities or comorbidities influencing the PR interval or outcomes were excluded.

Results

A total of 52 038 patients were included. Adjusting for age, sex, and relevant covariates the highest hazard ratio (HR) was observed in patients with short PR interval with an HR of 1.50 (95% CI, 1.24–1.80, P<0.001). A long PR interval did not show an association with all‐cause mortality (HR, 1.02 [95% CI, 0.97–1.08], P=0.3566). Adjusted 5‐year cumulative incidence of all‐cause mortality was 18% for short PR interval, 14% for normal PR interval, and 13% for long PR interval. Regarding recurrent syncope, a HR of 1.14 (95% CI, 1.09–1.20. P<0.001) was seen for long PR interval. Adjusted 5‐year cumulative incidence of recurrent syncope was 23% in patients with a long PR interval.

Conclusion

In patients with syncope, a short PR interval was associated with higher risk of all‐cause mortality; however, a long PR interval was associated with increased rate of recurrent syncope.

Research PerspectiveWhat is New? This study demonstrates that the entire spectrum of the PR interval holds prognostic value when assessing patients with syncope.Short PR intervals were independently associated with an increased risk of all‐cause mortality, and prolonged PR intervals were associated with increased risk of recurrent syncope, underscoring the importance of considering both short and prolonged PR intervals in risk assessment. What Question Should Be Addressed Next? What pathophysiological mechanisms explain the association between short PR intervals and increased all‐cause mortality?

Syncope is a common clinical presentation accounting for 0.8% to 2.4% of all attendances to the emergency department and ∼6% of all hospital admissions. ^1^ , ^2^ Syncope is associated with an increased risk of all‐cause mortality, recurrent syncope, and cardiovascular events. ^3^ However, it is evident that the poor outcome of syncope is related to the underlying disease rather than syncope itself. Particularly, syncope due to cardiac diseases is associated with a higher mortality compared with patients with noncardiac syncope. ^3^ , ^4^

In the initial evaluation of patients with syncope, a resting 12‐lead ECG can provide information about the potential and specific cause of the syncope. ^1^ , ^5^ A PR interval lower than the normal range of 120 to 200ms has shown to be associated with an increased risk of all‐cause mortality. ^6^ , ^7^ In addition, both short and prolonged PR intervals are also associated with an increased risk of atrial fibrillation. ^8^ Nonetheless, the significance of a prolonged PR interval and the risk of all‐cause mortality is still widely debated with some studies showing no association of all‐cause mortality and other studies showing an increased risk of all‐cause mortality ^7^ , ^9^ , ^10^ , ^11^ In patients with syncope, the presence of atrial fibrillation, intraventricular conduction disorders, left ventricular hypertrophy, or ventricular pacing is associated with increased risk of 1‐year all‐cause mortality. ^12^

Current syncope risk stratification scores rely on clinical interpretation of the ECG and not on continuous ECG parameters. For example, some current risk scores include interpretations of the PR interval, that is, the presence of advanced atrioventricular blocks but not the PR interval as a continuous parameter. ^5^ , ^13^ , ^14^ , ^15^ , ^16^ , ^17^ , ^18^ However, the risk stratification scores have not performed better than clinical judgment ^5^ , ^13^ , ^14^ , ^15^ , ^16^ , ^17^ , ^18^ and better parameters for risk stratification of patients with syncope are needed. Therefore, the aim of this study was to determine the association between the PR interval and recurrent syncope and all‐cause mortality in patients admitted with a first‐time syncope.

Methods

Data Availability Statement

Due to Danish data protection regulations, the data are not publicly available and cannot be shared by the authors. Access to the data requires approval from Statistics Denmark and is available only to authorized researchers affiliated with a Danish research institution. Applications for data access can be made directly to Statistics Denmark.

Ethics Approval

The Capital Region of Denmark granted permission for data access in compliance with the General Data Protection Regulation. By Danish law, registry‐based studies do not require approval from an ethics committee nor informed consent if the study is conducted for the purpose of statistics and scientific research. ^19^

Setting and Data Sources

The DNEC (Danish Nationwide Electrocardiogram Cohort) comprises all ECGs recorded in Danish ambulances and at Danish hospitals from January 1, 2000, to December 31, 2021, making it the largest population‐based ECG data set that can be linked to comprehensive registry‐based data on comorbidities, medication use, operation procedures, and clinical outcomes including mortality data. ^20^ DNEC contains more than 11.9 million ECGs from more than 2.4 million patients. ^20^ Using a unique civil registration number assigned to all residents in Denmark upon birth or immigration, it was possible to link all patients in DNEC to multiple national health registries. Information on all hospital, ambulatory, and emergency department diagnoses was obtained from the Danish National Patient Register. ^21^ All diagnoses were classified according to the International Classification of Diseases, Tenth Revision (ICD‐10). The Danish Register of Pharmaceutical Sales ^22^ was used to obtain information on dispensed drug prescriptions from all Danish pharmacies, according to international Anatomical Therapeutic Classification codes. Demographic information, such as sex and age, was obtained from the Danish Civil Registry. ^23^ Information regarding mortality was obtained from the Danish Register of Causes of Death. ^24^ All data were available through Statistics Denmark.

Study Population

All patients ≥18 years of age with a diagnosis of syncope (ICD‐10 R55.9) and an ECG recorded at a hospital within 1 day of admission with first‐time syncope were identified. The ICD‐10 code for syncope has previously been validated with a positive predictive value of 95%. ^25^ For patients where multiple ECGs were available within the time window, we selected the ECG recorded on the same day as the diagnosis. If no ECGs were available on the same day, we selected the ECG recorded the day after and lastly the day before. If multiple ECGs were available the same day, we selected the ECG recorded first that day.

The date of the ECG recording associated with the first syncope diagnosis was used as the ECG index date. Patients who had pacemaker implantation before the index date were identified through procedure codes from the Danish National Patient Register and excluded from the analyses. Similarly, patients who immigrated <5 years before the index date were excluded due to limited information on baseline comorbidities and treatments. We also excluded patients with possible confounding comorbidities defined as cancer, chronic obstructive pulmonary disease, or anemia. Further, with the help of the clinically validated Marquette 12SL™ algorithm we excluded ECGs with atrial fibrillation, atrial flutter, junctional rhythms, sinus arrest or sinoatrial block, nonsinus rhythms, and advanced atrioventricular blocks (second‐ and‐ third degree atrioventricular block). ECGs with premature atrial or ventricular complexes, fusion complexes, escape complexes, ventricular preexcitation, or aberrant conduction of supraventricular beats were also excluded. The diagnostic statements have previously been validated with high accuracy and validity. ^26^ Also, ECGs with bradycardia (ventricular rate ≤40) or tachycardia (ventricular rate ≥110) were excluded to ensure a homogeneous population with stable sinus rhythm, minimizing potential confounding effects of arrhythmias on PR interval measurement and prognosis. We excluded the 0.25% and 99.75% percentiles of the PR interval distribution to exclude outliers, Finally, cases where the P wave duration was greater than the PR interval were excluded.

ECG and Cardiovascular Disease

All ECGs in DNEC were standard 12‐lead ECGs and were digitally stored in the MUSE NX ECG Management System (GE Healthcare, Milwaukee, WI, USA). All ECGs were processed using version 23 of the Marquette 12SL™ algorithm, ^27^ ensuring standardized and uniform ECG diagnostic statements as well as global and lead‐specific ECG measurements ^26^ , ^27^ The PR interval was derived from the global measurements of the Marquette 12SL algorithm. Previously, we validated the PR interval measurements from the computer‐based Marquette 12SL algorithm with good agreement. ^8^

We defined ST‐T deviation according to the fourth universal definition of myocardial infarction. ^28^ When defining QTc prolongation, the QT interval was corrected for heart rate using the Fridericia formula. A QTc interval >450 ms for men and >470 ms for women were defined as prolonged. ^29^ Left ventricular hypertrophy was defined according to the Sokolow–Lyon ECG criteria. ^30^ We defined baseline cardiovascular disease as an ICD‐10 diagnosis of coronary artery disease (including myocardial infarction), heart failure, valvular heart disease, cardiomyopathy, cerebrovascular disease, peripheral vascular disease, diabetes, renal failure, cardiac conduction disorders, atrial fibrillation, or other cardiac arrhythmias given before the index date. See Tables S1 and S2 for a full list of the ICD‐10 and Anatomical Therapeutic Classification codes used in this study. We included I44.0 (first‐degree atrioventricular block), as a separate covariate because it represents a clinically recognized baseline diagnosis, meaning the patient had been previously evaluated and diagnosed, typically during a prior hospital contact. This distinction is important, as patients known to have atrioventricular block may differ in prognosis and management from those with incidental PR prolongation identified on the index ECG.

Outcome Measures

The study included 2 primary outcomes, the association between PR interval and all‐cause mortality, and the association between PR interval and recurrent syncope. Recurrent syncope was defined as a diagnosis of syncope more than 24 hours after discharge from the initial hospitalization. Patients were followed from the index date to the first event of either end of study (May 26, 2023), emigration, or death. Median follow‐up time was calculated using the reversed Kaplan–Meier estimator.

Statistical Analysis

Categorical variables are presented as numbers and percentages and continuous variables are presented as means±SD.

A Cox regression model was used to investigate the associations between PR interval and all‐cause mortality, and a cause‐specific Cox regression model was used to estimate the association between PR interval and recurrent syncope with death and pacemaker implantation as competing events. The PR interval was entered into the model as a categorical variable with a normal PR interval (120–200ms) as reference. Further, the PR interval was categorized as either short (<120 ms) or long (>200 ms). Both models were adjusted for age, sex, and baseline cardiovascular disease. Each cardiovascular disease, described previously, was entered into the model as individual parameters. Age was entered as a categorical variable with age groups of 18 to 42, 43 to 60, 61 to 74, and >74 years old. Age cutoffs were determined by the 25th, 50th, and 75th percentiles of the age distribution. Sex and cardiovascular disease were included in the model as binary variables. The associations were described as hazard ratios (HRs) with 95% CIs.

Using the Cox‐ and cause‐specific Cox regression models the 5‐year absolute risk of both outcomes was estimated in the overall population, stratified by PR interval groups to assess differences between groups.

To account for confounding, we standardized the predicted risks using marginal standardization (g‐computation), averaging individual predicted risks over the covariate distribution of the overall study population. This yielded marginal model‐based risk estimates stratified by PR interval groups and ensured that the estimated risk reflected differences in the PR interval groups independently of the covariates included in the models. To generate smooth model‐based risk estimate curves we predicted the risk at daily time points. As a final step, we created unadjusted cumulative incidence curves for both all‐cause mortality and recurrent syncope, both on the entire population and divided by age groups. For recurrent syncope the cumulative incidence of the first recurrent event was estimated using the Aalen–Johansen method, treating pacemaker implantation and death as competing risks.

Data management and statistical analysis were conducted with the use of R version 4.4.1 (R Foundation for Statistical Computing, Vienna, Austria). A P value <0.05 was considered statistically significant.

Supplementary Analyses

To assess the robustness of our primary findings and minimize potential confounding, we performed a 1:1 nearest‐neighbor propensity score matching without replacement based on age, sex, and comorbidities to balance individuals with short, normal, and long PR intervals. Postmatching balance was evaluated using standardized mean differences. Then we conducted a Cox regression analysis on the matched cohort to evaluate the association between PR interval and all‐cause mortality.

Second, to explore potential nonlinear associations in the long PR interval group we first divided the individuals with long PR interval (>200ms) into 2 groups based on the median (218 ms) resulting in 2 200 to 218 ms and >218 ms. In a separate analysis we modeled the PR interval as a continuous variable using restricted cubic splines in a Cox regression model, adjusting for age, sex, and comorbidities to assess the association with all‐cause mortality across the full PR interval spectrum. To assess whether the cubic spline model improved fit relative to the linear model, a likelihood ratio test was performed. The 95% CIs were derived from the model.

Third, to investigate the physiological basis of the PR interval abnormalities we calculated the P‐wave duration/PR interval (P/PR) ratio, which provided insight into the relative contribution of atrial conduction time versus atrioventricular nodal conduction time. We stratified the distribution of the P/PR ratio by PR interval group and created boxplots to assess whether the ratio differed across PR interval groups. Further, we examined the association between the P/PR ratio as a continuous variable and all‐cause mortality using a Cox regression model with restricted cubic splines. We conducted separate similar analyses on the PR segment duration and P‐wave duration to further explore conduction abnormalities and their prognostic relevance.

Fourthly, as a supplementary analysis we adjusted both multivariable models for baseline use of atrioventricular‐nodal and antiarrhythmic medications, including digoxin, betablockers, calcium channel blockers, loop diuretics, and class I, III, and IV antiarrhythmic agents.

RESULTS

Baseline Characteristics

A total of 113 198 patients had an ECG recorded at a Danish hospital within 24 hours of a diagnosis of syncope. Of these 50 816 patients were eligible for inclusion (Figure 1). The final study population had their ECG recorded between October 30, 2001, and January 31, 2021, with an average number of 1.35 ECGs per index event. Baseline characteristics of the study population are shown in Table 1, and ECG characteristics are shown in Table 2. Median follow‐up time was 6.37 years (interquartile range, 3.82–8.97) for all‐cause mortality and 6.55 years (interquartile range, 4.03–9.12) for recurrent syncope.

Risk of All‐Cause Mortality

The 5‐year unadjusted cumulative incidence of all‐cause mortality was 8% for short PR interval, 12% for normal PR interval, and 23% for long PR interval (Figure 2). The age‐divided cumulative incidence showed that in the older population (>61 years old), patients with a short PR interval had the highest 5‐year event for patients aged >74 years old with a short PR interval the cumulative incidence was 46.3% whereas it was 38.6% for patients with a long PR interval in the same age group.

The 5‐year model‐based estimated risk of all‐cause mortality was 18% for short PR interval, 14% for normal PR interval, and 13% for long PR interval (Figure 3). The 1‐year model‐based estimated risk of all‐cause mortality was also higher for short PR interval (4%) compared with both normal (3%) and long (3%) PR interval.

Results from the Cox regression model on the adjusted association between all‐cause mortality and PR interval are summarized in Table 3. Regarding the PR interval, the highest HR was seen in patients with short PR interval with an HR of 1.46 (95% CI, 1.21–1.77; P<0.001). A long PR interval did not show an association with all‐cause mortality (HR, 1.03 [95% CI, 0.97–1.08]; P=0.330). Further, increasing age was associated with an increased risk of all‐cause mortality with HRs of 5.60 (95% CI, 4.72–6.64, P<0.001), 13.63 (95% CI, 11.57–16.07, P<0.001), and 49.90 (95% CI, 42.43–58.67, P<0.001) for patients aged 43to 60, 61 to 74, and >74 years, respectively, compared with the reference group aged 18to 42 years. Men had an HR of 1.27 (95% CI, 1.22–1.33; P<0.001) compared with women. Notably, baseline heart failure, valvular heart disease, and atrial fibrillation were associated with increased risk of all‐cause mortality with HRs of 1.51 (95% CI, 1.39–1.64; P<0.001), 1.24 (95% CI, 1.14–1.34; P<0.001), and 1.09 (95% CI, 1.01–1.17; P=0.030), respectively.

Risk of Recurrent Syncope

The 5‐year unadjusted cumulative incidence of recurrent syncope showed that a long PR interval was associated with the highest event rates with 27% (Figure 4). The cumulative incidence of recurrent syncope for short PR interval was 17% and for normal PR interval the cumulative incidence was 20%. The age‐divided cumulative incidence did not show a change in this pattern for any of the age groups; however, for the patients aged >74 years old, a short PR interval had the highest 1‐year event rates.

For the 5‐year model‐based estimated risk of recurrent syncope the highest risk was in patients with a long PR interval (23%), whereas it was 21% and 20% for normal and short PR interval, respectively (Figure 3). Notably, the estimated risk of recurrent syncope was greatest within the first day of the first syncopal event. The Cox regression model on the association between recurrent syncope and PR interval are summarized in Table 3. All results from the cause‐specific Cox model on recurrent syncope are presented in the (Table S3). Long PR interval showed an HR of 1.16 (95% CI, 1.10–1.22. P<0.001), whereas short PR interval showed no association with recurrent syncope with a HR of 1.00 (95% CI, 0.86–1.16; P=0.9905). Increasing age was associated with higher HRs for recurrent syncope with HRs of 1.12 (95% CI, 1.06–1.19; P<0.001), 1.44 (95% CI, 1.36–1.52; P<0.001), and 1.99 (95% CI, 1.88–2.11; P<0.001) for patients aged 43 to 60, 61 to 74, and >74 years, respectively, compared with the reference group aged 18 to 42 years. Men had a slightly higher rate of recurrent syncope with a HR of 1.08 (95% CI, 1.04–1.13, P<0.001) compared with women. Baseline cardiac conduction diseases, coronary artery disease, heart failure, valvular heart disease, and atrial fibrillation were associated with increased rate of recurrent syncope with HRs of 1.55 (95% CI, 1.39–1.73; P<0.001), 1.11 (95% CI, 1.05–1.18; P<0.001), 1.16 (95% CI, 1.05–1.27; P=0.003), 1.16 (95% CI, 1.05–1.27; P=0.002), and 1.15 (95% CI, 1.06–1.25; P<0.001), respectively.

Supplementary Analyses

We matched the individuals with a short PR interval 1:1 with individuals with a normal or long PR interval based on age, sex, and comorbidities. After matching, the mean standardized mean difference was 0.002 between short and normal PR interval, and 0.050 between short and long PR interval, indicating a good overall between the matched groups. The association between short PR interval and increased all‐cause mortality remained significant after matching with an HR of 1.69 (95% CI, 1.24–2.31; P=0.001), supporting the robustness of our findings.

Dividing the long PR interval group into two different groups based on the median (200–218 ms and >218 ms), showed that the PR >218 ms group had an HR of 1.13 (95% CI, 1.05–1.20; P<0.001), and the 200 to 218 ms group had an HR of 0.93 (95% CI, 0.87–1.00; P=0.043).

The adjusted association between PR interval as a continuous variable and the risk of all‐cause mortality showed a U‐shaped relationship (Figure 5). Likelihood ratio testing confirmed that the cubic spline model provided a significantly better fit than a linear model (P<0.001), supporting the use of splines to flexibly model this association. Notably, the HRs increased more rapidly moving towards short PR intervals compared with moving towards longer PR intervals. For example, PR interval of 125 ms had an HR of 1.4, (39 ms decrease from the median) whereas a PR interval of 280 ms (116 ms increase from the median) had a similar HR. Further, PR intervals inside the normal PR interval group showed increased HRs compared with the median PR interval, and both for short and long PR intervals the extremes showed the highest risk.

The P/PR ratio was highest in the short PR interval group and lowest in the long PR interval group (Figure 6) indicating high P‐wave contributions or short PR segments for individuals with short PR interval. The short PR interval group had the shortest PR segments, and the long PR interval group had the longest PR segments. The P‐wave duration did not differ notably between the 3 PR interval groups. Both a long PR segment and a long P‐wave duration were associated with increased risk of all‐cause mortality, however, with wide 95% CIs in the extremes. Short PR segment and short P‐wave duration were also associated with increased risk of all‐cause mortality. The P/PR ratio showed a J‐shaped relationship with all‐cause mortality with a low P/PR ratio being associated with the highest risk.

Adjustment for baseline medication use resulted in HRs that were slightly attenuated compared with those in the main analysis with HRs of 1.46 (95% CI, 1.20–1.76; P<0.001) and 1.00 (95% CI, 0.95–1.06; P=0.894) for short and long PR interval for all‐cause mortality, respectively. Regarding recurrent syncope the HRs were 1.00 (95% CI, 0.86–1.16; P=0.952) and 1.15 (95% CI, 1.09–1.22; P<0.001) for short and long PR interval, respectively.

DISCUSSION

In patients with syncope a short PR interval was associated with higher risk of all‐cause mortality when adjusting for confounders. This is consistent with previous findings suggesting an association between short PR interval and increased risk of all‐cause mortality in patients with coronary artery disease ^6^ and the general population. ^7^ On the other hand, a long PR interval was associated with increased rate of recurrent syncope, which supports existing findings, indicating that syncope with cardiac cause may be associated with greater rate of recurrent syncope. ^31^

The incidence of death during follow‐up in the current study was 18%, which is comparable to previous findings reported in a Danish registry‐based study by Ruwald et al. ^32^ who reported an incidence of all‐cause mortality of 20% in patients with syncope. The lower incidence observed in this study might be explained by the fact that the current study, in contrast to Ruwald et al., excluded patients with cancer and chronic obstructive pulmonary disease. Further, only 31% had cardiovascular disease in the current study compared with 38% in the study by Ruwald et al., possibly explaining the observed differences.

Previous studies on the association between the PR interval and all‐cause mortality are inconsistent, with some studies showing lack of association and other studies showing an increased risk of mortality for both short and long PR interval. Rasmussen et al. ^7^ demonstrated a nonlinear association between the PR interval and cardiovascular mortality, with both short and long PR intervals showing an increased risk. ^7^ However, for all‐cause mortality only short PR interval was associated with an increased risk supporting the findings of the current study. This was also the case in the propensity score matching, which showed that in a homogeneous population a short PR interval was still associated with increased risk of all‐cause mortality. Similarly, Holmqvist et al. ^6^ showed that a short PR interval but not a long PR interval was associated with an increased risk of all‐cause mortality in patients with known coronary artery disease. However, the explanation of the increased risk of all‐cause mortality in patients with short PR interval is still debated. A short PR interval could be the result of accessory pathways from the atria to the ventricles, as observed in patients with Wolff‐Parkinson‐White syndrome. ^33^ Wolff‐Parkinson‐White is associated with an increased risk of ventricular tachyarrhythmias, ^33^ potentially explaining the increased risk of all‐cause mortality for short PR interval. However, we excluded patients with ventricular preexcitation in the current study opposing that hypothesis. Previously, an association between both short PR interval and increased risk of atrial fibrillation has been shown. ^8^ Further, atrial fibrillation is associated with an increased risk of ventricular tachycardia and ventricular fibrillation, ^34^ , ^35^ possibly explaining the increased risk of all‐cause mortality. The supplementary analyses investigating the physiological basis of the PR interval as the P/PR ratio, P‐wave duration, and PR segment demonstrated that patients with short PR interval had a significantly short PR segment. This finding supports the hypothesis that a dysfunctional atrioventricular node may excessively stimulate the ventricles, particularly in patients with atrial fibrillation, potentially causing ventricular fibrillation. Further, consistent with our findings, Soliman et al. ^9^ showed that a high P/PR ratio was associated with increased risk of mortality in patients with short and long PR interval. Also, both a short P‐wave duration and a short PR segment were associated with increased risk of all‐cause mortality. However, patients with a short PR interval exhibited both shorter P‐wave durations and shorter PR segments, making it difficult to determine which component drives the increased risk. Although Soliman et al. ^9^ suggested that the prognostic value of the PR interval is primarily due to P‐wave duration, our findings indicate that, in the context of a short PR interval, both P‐wave duration and PR segment contribute to the observed risk.

In addition, Perez‐Rodon et al. ^5^ investigated multiple ECG predictors of all‐cause mortality in patients with syncope and found that atrial fibrillation had an HR of 6.8 in association with 1‐year all‐cause mortality. Further, the risk of all‐cause mortality may be higher in older patients with short PR interval, since aging is associated with increased fibrosis of the conduction system, possibly leading to reentrant arrhythmias resulting in ventricular fibrillation. Also, older patients carry a higher burden of structural heart disease, which may result in reduced resistance of the rapid conduction from the atria to the ventricles increasing the risk of arrhythmic events. This was also supported by the age‐divided cumulative incidence, which showed that it was in the older population that a short PR interval was associated with increased risk of mortality.

The PR interval is not only dependent on structural or electrophysical properties but also on the parasympathetic and sympathetic nervous system, which actively regulates the atrioventricular conduction time. Increased sympathetic activation shortens the PR interval whereas increased parasympathetic activation prolongs the PR interval, and an autonomic imbalance has previously been shown to be associated with increased risk of cardiovascular outcomes. ^36^ Our findings demonstrate that a short PR interval is associated with higher HRs for all‐cause mortality, which may be explained by increased sympathetic activity. Thus, short PR intervals could be a marker of autonomic imbalance which could promote arrhythmias and other adverse cardiac events.

Recurrent Syncope

Both the unadjusted cumulative incidence and the model‐based estimated risk showed that a long PR interval was associated with an increased rate of recurrent syncope. The incidence of recurrent syncope during follow‐up in the current study was 22%, which is higher than previous reported incidences. ^3^ , ^31^ , ^32^ Ruwald et al. ^32^ reported an incidence of recurrent syncope of 16% in the Danish population in the period 2001 to 2009, and Zimmerman et al. ^31^ reported an incidence of 24‐month recurrent syncope of 20%. In contrast to Ruwald et al., ^32^ we had the unique opportunity to include the ECG in our investigations. Zimmerman et al. ^31^ showed that syncope with cardiac cause was associated with a greater rate of recurrent syncope, which the observations of this study support that, and the increased rate of recurrent syncope in patients with a long PR interval may indicate that cardiac conduction disturbances elevate this risk in patients with a long PR interval. Our findings of increased rate of recurrent syncope in patients with a long PR interval might be explained by the fact that first‐degree atrioventricular block is associated with an increased risk of progressing into more advanced atrioventricular blocks, that is, second‐degree Mobitz type‐II atrioventricular block or third‐degree atrioventricular block, which is associated with increased risk of syncope. ^1^ Hence, the increased rates of recurrent syncope in patients with a long PR interval may have been caused by atrioventricular block, undetected at initial presentation. Also, treating the PR interval as a continuous variable suggests that categorizing it into discrete groups may mask clinically relevant information as the HRs for recurrent syncope increased progressively with longer PR intervals.

In addition, recurrent syncope is associated with a greater risk of all‐cause mortality and major adverse cardiac events, ^31^ , ^32^ hence calling for more focus on first‐degree atrioventricular block and earlier intervention, to reduce future syncopal events. Although we observed long PR interval to be associated with the highest rates of recurrent syncope, this group did not demonstrate the highest risk of all‐cause mortality. This apparent discrepancy may be explained by the fact that recurrent syncope and all‐cause mortality were investigated using separate models. Hence, since recurrent syncope and all‐cause mortality were analyzed using separate models we cannot determine whether or how recurrent syncope might mediate or elevate the risk of all‐cause mortality. The increased rates of recurrent syncope in patients with long PR interval may be driven by non‐lethal mechanisms which may explain the absence of increased risk of all‐cause mortality in this group.

Limitations

Some limitations must be considered for the findings of this study. We did not consider the location of the recorded ECG. Some patients might have had their ECG recorded in the emergency department whereas others had their ECG recorded in the cardiology department. Hence, we did not account for injuries acquired in relation to the syncopal event that is, severe trauma, which might have affected the results. Patients with cardiac arrhythmias were excluded, which might bias the observed results toward lower risk ratios. However, despite excluding those, we were still able to demonstrate an association between PR interval and all‐cause mortality and recurrent syncope. Also, we did not consider the cause of death, which could help explain possible mechanisms causing short PR interval.

An additional limitation of this study is that we could not formally assess atrial fibrillation and atrioventricular block as mediators due to their intermittent presentation, limited event numbers, and the likelihood that high‐grade atrioventricular block often leads to sudden death before diagnosis. Nonetheless, our supplementary analyses support the robustness of the findings and are consistent with this proposed mechanism. We are planning a follow‐up study to investigate these mechanistic relationships in more detail.

Another limitation of the study is that because the study was a register‐based study, we did not know the history leading to the syncope nor the cause of the syncope. Thus, we did not know how many syncope events were deemed as unknown or cardiac, for example. This might have affected the treatment of the patients and hence affected the results observed in this study.

Clinical and Future Implications

A short PR interval might be treated as a benign condition whereas clinical protocols are in place for a prolonged PR interval. ^1^ Our findings suggest that a short PR interval is associated with higher risk of all‐cause mortality in patients with syncope, which clinically might help identify high‐risk individuals who need closer monitoring. However, before integrating these findings into clinical practice additional studies investigating the underlying mechanisms causing the short PR interval are needed. Future studies should focus on explaining the underlying mechanisms of short PR interval and why the association with death is observed, which could be done by including the cause of death in the analysis. Such studies might be appropriate in the general population and not just for syncope patients, as the observed association in the current study is also proven in the general population. ^7^ Studies describing the mechanisms by which the short PR interval is caused, and which consequences triggered by a short PR interval are essential to use the short PR interval in future clinical practice and risk management. Further, this may lead to defining a specific threshold associated with an increased risk for which an action is required. The supplementary analysis treating the PR interval as a continuous variable suggests that categorizing it into discrete groups may mask clinically relevant information. Evaluating the PR interval on a continuous scale may offer greater prognostic accuracy and support a more individualized approach to patient care, consistent with practices applied to other ECG parameters such as QT interval and heart rate variability. ^37^

CONCLUSIONS

In conclusion, in patients with syncope a short PR interval was associated with higher risk of all‐cause mortality. However, no association was found between a long PR interval and all‐cause mortality. Second, a long PR interval was found to be associated with higher rate of recurrent syncope. The observed relationship between PR interval and all‐cause mortality in this study is consistent with previous studies published on the general population. Future studies describing the mechanisms by which short PR interval is associated with increased mortality is still needed for short PR interval to be an actionable parameter in clinical decision making.

Sources of Funding

This research was funded by the Karen Elise Jensen Foundation.

Disclosures

None.

Supporting information