Heart rate stability in a clinical setting and after a short exercise in healthy male volunteers

Limited data exist on heart rate stabilization in the domiciled nature of phase I clinical studies, particularly when frequent measurements of QT intervals are involved. The present analysis aimed to evaluate heart rate stability in the domiciled nature of, and stabilization after a short exercise.


| INTRODUCTION
Twelve-lead surface electrocardiograms (ECGs) are frequently performed in phase I human pharmacology trials to evaluate a novel compound's effect on cardiac conductivity and as a screening method to detect potentially fatal drug-induced ventricular electrical instability (Kinter & Valentin, 2002).Of particular interest is the QT interval, which reflects the duration of ventricular depolarization and subsequent repolarization on an ECG, and QT interval prolongation which is linked to drug-induced ventricular arrhythmia.However, the QT interval varies over time and is inversely related to a subject's RR interval (time elapsed between two successive R-waves of the QRS signal on an ECG) with shorter QT intervals with increasing heart rates.And although the RR interval is in turn a function of intrinsic properties of the sinus node as well as autonomic influences, the subsequent QT interval changes encompasses both a swift and a gradual process (Baumert et al., 2016;Malliani et al., 1991;Surawicz & Knilans, 2008;Tsuji et al., 1996).
The process of how swiftly the QT interval adapts to RR interval fluctuations over time is referred to as QT/RR hysteresis (Malik et al., 2016).Previous research has shown that where heart rhythm exhibits evident variations, it is necessary to take into account the hysteresis lag present in the QT interval adaptation to RR changes (Alessandrini et al., 1997).Ample research about heart rate recovery following exercises with different degrees of intensity report recovery times of RR interval stabilization to pre-exercise levels of up to 45-60 min (Imai et al., 1994;Javorka et al., 2002;Kannankeril et al., 2004;Pelchovitz et al., 2012;Wang et al., 2011).Hence a stable heart rate is preferable for accurate QT interval measurements, and thus the Society for Cardiological Science and Technology states that time should be taken for any given individual to be relaxed and comfortable before ECG measurements without further detailing a certain time window (Campbell et al., 2017).Alternatively, other groups advocate a minimum period of 10 min or longer in a supine position before the QT measurement (Batchvarov et al., 2002).
However, in a clinical research unit, subjects are typically domiciled during phase I human pharmacology studies, making the previously cited studies not necessarily applicable.Hypothetically, the time period for the heart rate to return to baseline in domiciled healthy volunteers is shorter (Batchvarov et al., 2002;Javorka et al., 2002;Kligfield et al., 1996).It is desirable to maintain a resting window that is on one hand long enough to allow stabilization of the RR interval, but as short as possible short to minimize interference in operational study conduct.Therefore, the aim of the present analysis was to evaluate heart rate stability in the domiciled nature of, and stabilization after a short exercise in male subjects.

| METHODS
In the present analysis, data of two clinical studies were included.
The first clinical study covered the effects of supine resting in a clinical trial setting in male subjects, the second clinical study covered the effects of supine resting after a limited exercise bout in male subjects.All data were collected at the Centre for Human Drug Research in Leiden, the Netherlands, a clinical research organization specialized in early phase drug development studies.
Before any study-related activities, an informed consent according to Declarations of Helsinki recommendations was signed by all subjects.Both clinical studies were performed in compliance with Good Clinical Practice (GCP).The clinical trial setting data were collected as part of a large study, registered under toetsingonline number NL68390.056.19 and approved by an independent ethics committee.The exercise bout setting data were collected as part of a study that was not submitted to the ethics committee as per Dutch legislation.A written informed consent was obtained before any study activities.All activities were performed in accordance with applicable standard operating procedures.

| Clinical trial setting study
Included subjects were screened before study participation and were either young (18-25 years) or elderly (>60 years) males.Key exclusion criteria were evidence of any active or chronic disease or condition that could interfere with the study objectives.Furthermore, the use of any medication within less than five half-life times before study participation was prohibited.Subjects were instructed to roam about in the clinical research unit between 45 and 60 min before the first measurement.In accordance with the standard operating procedures use of staircases, any form of physical exercises or activities that may induce excessive heart rate fluctuations were prohibited.They were then instructed to assume a supine position during which a holter-ECG was performed for at least 10 min.This setting was chosen to mimic a typical clinical research setting.

| Exercise bout study
Recruited male subjects were included in case they claimed to be otherwise healthy and were not taking any cardiac-related medication within less than five half-life times before study participation.
After subjects were fitted with a holter-ECG, they were instructed to remain in a resting supine position for 5 min, after which they had to walk up and down three stories (100 steps).Immediately upon completion, they were instructed to assume a supine resting position for 15 min.A holter-ECG measurement was performed continuously.

| Holter ECG
Data were acquired with Mortara H12+ tm monitors (Mortara Instrument), with 10 disposable electrodes placed in the standard anatomical position.Data from the holter monitors were automatically analysed with validated SuperECG Software (Mortara Instrument).Validation consisted of automated and manual WILDENBEEST ET AL.
| 37 calculations of various ECG measurements in 30 patients in normal sinus rhythm with continuous 12-lead Holter recordings for a 24-h period, which illustrated a mean difference of 5 ms in QT interval analyses (Thomas & Carey, 2007).Also, this software has previously been used in other research to analyse ECG measurements (Tinoco et al., 2017).
The ECG parameters that were included in the present analysis were PR, RR, QT and corrected QT (QTcF) interval.The QTcF interval values were calculated with the Fridericia formula: Intervals of 30 s were used to calculate means of each parameter.
In the clinical setting study, the final 120 s of recording were considered a stable heart rate and were used as baseline for analysis.
In the exercise bout study, the final 300 s of the first 10 min in supine position were used to create a baseline.

| Statistical analysis
Statistical analyses were performed using IBM SPSS version 20 (IBM Corporation).Data are reported as mean ± standard deviation (SD), or with a percentage where appropriate.The collected data were visually assessed for normal distribution, indicating that it followed a normal distribution pattern.Paired t-tests were used to compare the means of the 30 s intervals with baseline means.Two criteria were set to determine a significant difference with the baseline.The first criterion was a statistically significant p-value between the evaluated epoch and the baseline measurement.The second criterion was a change from baseline greater than 25% for the PR interval or above the threshold of >240 ms, 20 ms for the RR interval, 1.35 ms for the QT interval and 1.9 ms for the QTcF interval (Nada et al., 2013;Thomas & Carey, 2007;Tinoco et al., 2017).These values were based on reported literature for PR interval, calculated using half the intrasubject SD reported in literature for RR and QT interval or based the data set from our own exercise bout study for QTcF, using the last 5 min from the baseline holter recordings.

| RESULTS
Data of two studies, one in a typical domiciled clinical setting and one with a limited exercise bout, were retrospectively analysed.Baseline characteristics of each group are shown in Table 1.

| Clinical setting ECG parameter measurements
The

| Exercise study ECG parameter measurements
The male subjects of the exercise challenge (n = 10) had a mean age of 26.9 ± 4.7 years which differed significantly from both clinical trial

| DISCUSSION
In the present analysis, we observed a stable heart rate in male subjects that were in a domiciled clinical setting after they assumed a supine position, whereas heart rate stabilization took up to 2 min after attaining a supine position following a short exercise.In line with this, PR and QTcF intervals remained stable in the clinical setting, whereas this took up to 2 min after the short exercise.These Heart rate is influenced by several mechanisms including autonomic function (Campbell et al., 2017;Malliani et al., 1991).
However, given that in virtually all patients and healthy subjects the RR interval and in turn QT interval varies on a beat-to-beat basis even during resting, controlled conditions, there is no clear consensus on the resting period before an ECG measurement (Batchvarov et al., 2002;Electrophysiology, 1996).Previous studies focused on QT/RR hysteresis, the process of how swiftly the QT interval adapts to RR interval fluctuations over time, reported stabilization recovery times up to 45-60 min varying based on the level of intensity following exercises (Imai et al., 1994;Javorka et al., 2002;Kannankeril et al., 2004;Pelchovitz et al., 2012;Wang et al., 2011).
Therefore, at minimum a resting period of up to 5-10 minutes is used in a clinical research setting.However, our data suggest that these heart rate stabilization studies are not applicable and do not translate well to clinical trials.In our analysis, both RR and PR interval recovered to baseline levels after a short exercise up to 2 min after assuming a supine position and reflects the normal parasympathetic reactivation following an exercise (Imai et al., 1994).And although a relatively low level of intensity exercise was chosen with a mean heart rate elevation of less than 35 beats per minute, QTcF interval still took longer to recover to baseline levels.This is in line with previous studies where QT adaptation in response to RR changes were reported with an initially rapid QT interval reaction during the first 50 s followed by a slower adjustment that takes up to 2 min to complete (Pueyo et al., 2003).Therefore, the time for the ECG parameters to normalize in heart rate stabilization studies and our clinical study differ a lot from each other.These differences make it inappropriate to compare these two types of studies and applying these recovery times in a clinical setting.Technology states that time should be taken for any given individual to be relaxed and comfortable before ECG measurements a specific time window is not provided (Campbell et al., 2017).In this study, the exercise test was intentionally not designed to be maximal.Another point of concern that may also induce autonomic imbalance and thus lead to irregular RR intervals and inaccurate QT interval analyses are positional change (Yeragani et al., 2000).
Previous studies mainly focused on changing body positions from a supine to standing position, which resulted in short-term (<1 min) increased heart rates and elevated blood pressures to compensate for the initial fall in cardiac output (Borst et al., 1982;Patel et al., 2016;Watanabe et al., 2007;Yeragani et al., 2000).However, in our study, male subjects were instructed to assume a supine position from a standing position after which we observed no significant ECG measurement changes.These findings suggest a limited roll of autonomic imbalances caused by assuming a supine position on cardiac conductivity within a clinical setting.After this study, the question still arises if the ECG parameters have the same values after 3 min of rest and after 10 min of rest.To further investigate this, the ECG parameters should be measured again in a clinical trial and the values of the parameters after 3 min should be compared with timepoints 10 and 15 min to see if it is similar or that there are any changes from the first timepoint.

| LIMITATIONS
In this analysis, automated ECG analyses were used.These analyses are reliable, but analyses done manually by medical experts can be even more accurate (Upasani & Kharadkar, 2012).In automated analyses outliers in the data can be overlooked, and even cause an excessive value of a parameter that can influence the data.In this study, this potential impact was limited by evaluating the automatically generated data through visual inspection of histograms and QQ plots of data.Though automated have been validated, it is nevertheless not considered the gold standard (ICH, 2016).Moreover, it is important to note that the exercise group was primarily used to illustrate the relatively quick stabilization of RR intervals following a short exercise bout.This specific finding may not directly extrapolate to elderly volunteers.Exercise-induced changes in RR intervals can vary among different age groups, and caution should be exercised when generalizing these specific findings to elderly populations.Finally, a limitation of this study is that it only included male subjects, therefore restricting the generalizability of the findings to female individuals.

| CONCLUSION
These results indicate that in a clinical setting with male volunteers any waiting period for ECG parameters normalization is not needed, because in our clinical study the first minute of recording the resting time was already in line with baseline measurement.Furthermore, after an exercise challenge, in which male subjects performed an exercise of walking 100 steps, a maximum of up to 3 min is needed to normalize the heart rate parameters.
clinical study was performed in young (n = 29) and elderly (n = 17) male subjects with a mean age of respectively 21.1 ± 2.0 years and 71.6 ± 6.2 years.In the younger domiciled clinical setting subject group, PR interval was 0.69 ± 4.83 ms (p = 0.47) longer while RR interval was 2.65 ± 58.90 ms (p = 0.82) longer at the 30-s timepoint compared to the first baseline measurement immediately after assuming a supine position following the roaming period (Figure1a,b).QT interval was 1.03 ± 5.34 ms (p = 0.31) longer, while QTcF interval was 0.88 ± 4.42 ms (p = 0.31) longer at the 30-second timepoint compared to the baseline measurement (Figure1c,d).In the elderly male domiciled clinical setting subject group, PR interval was −1.20 ± 5.69 ms (p = 0.38) shorter while RR interval was 7.45 ± 24.21 ms (p = 0.22) longer at the 30-s timepoint compared to the first measurement immediately after assuming a supine position following the roaming period (Figure2a,b).QT interval was 1.14 ± 3.51 ms (p = 0.20) longer, while QTcF interval was 0.21 ± 3.15 ms (p = 0.79) longer at the 30-s timepoint compared to the baseline measurement (Figure2c,d).Each of the subsequently measured ECG parameter values over time (mean per 30 s) of both the younger and older male clinical setting subject groups showed no significant difference from the first baseline measurement immediately after assuming a supine position following the roaming period.Both the p-value of 0.05 from the paired samples t-test and the critical values were not exceeded as shown in Figures1 and 2.
study groups.Other than mean age, no relevant differences between the groups were observed.Different mean postexercise ECG parameters to baseline measurements were found in the exercise group as illustrated in Figure3.True values for each measurement are shown in Supporting Information: TableS1.Baseline heart rate was 74.0 ± 9.6 beats per minute.Mean duration of the short exercise was 121 s with a maximum mean heart rate of 107.3 ± 15.5 beats per minute.PR interval was significantly (p < 0.05) shorter for up to 120 s after assuming a supine position following the exercise with a mean value of −9.8 ± 7.2 ms, but never crossed the threshold of 25% change from baseline.RR interval was significantly (p < 0.05) shorter for up to 30 s after assuming a supine position following the exercise, while the final change from baseline larger than 20 ms is after 60 s with a mean value of −56.5 ± 73.5 ms.Finally, for QT and QTcF interval, the last statistically significant shortening compared to the baseline values were at the 90 (p = 0.04) and 120 (p < 0.1) second time points respectively.Based on the critical clinical values, both QT interval and QTcF interval duration were initially shorter but stabilized after 2 min with between measurement variability less than the critical value (1.35 and 1.9 ms respectively).However, QT interval duration remained prolonged compared to baseline values with a mean prolongation of +5.8 ± 6.3 ms while QTcF interval duration did normalize and returned to baseline values after the initial 2 min.

F
I G U R E 1 Electrocardiographic parameters (PR interval in [a], RR interval in [b], QT interval in [c] and QTcF interval in [d]) over time as measured in the younger (18-25 years) male domiciled clinical setting subject group, where each timepoint indicates the change from baseline (time point 0) means with standard deviations over intervals of 30 s.Each of the electrocardiographic parameters showed no significant difference from baseline.ms, milliseconds.data suggest that the current practice of an extended duration of supine positioning to normalize heart rate and QTcF interval could be reduced without affecting the ECG data quality.
However, in contrast to these exercise-focused heart rate recovery studies, in clinical trials these markedly increased heart rates will not be reached because subjects are prohibited to any form of physical exercise or activity that may induce excessive heart rate fluctuations.And although the Society for Cardiological Science & F I G U R E 2 Electrocardiographic parameters (PR interval in [a], RR interval in [b], QT interval in [c] and QTcF interval in [d]) over time as measured in the elderly (>60 years) male domiciled clinical setting subject group where each timepoint indicates the change from baseline (time point 0) means with standard deviation over intervals of 30 s.Each of the electrocardiographic parameters showed no significant difference from baseline.ms, milliseconds.
Instead, we aimed to create a more natural and representative daily setting during a clinical trial where individuals may roam freely within a facility.Our intention was to simulate a moderate level of physical activity that individuals might typically engage in during their daily routines.The present analysis illustrates that in such a typically domiciled clinical setting any waiting period for ECG parameters normalization is not needed and thus guidelines of resting time (5-10 min) may be eliminated.An interval of only 5 min may translate into operational/logistical benefits for clinical study conduct.In the vast majority of clinical trials there is a tight schedule for measurements.Shortening the resting time in supine position of subjects can leave time for extra non-ECG measurements or it can lead to a better performance of the existing schedule.

F
I G U R E 3 Electrocardiographic parameters (PR interval in [a], RR interval in [b], QT interval in [c] and QTcF interval in [d]) as measured in the male exercise bout group where each timepoint indicates the change from baseline (first 10 min) means with standard deviation over intervals of 30 s. Dotted square indicates the exercise period.(a) PR interval significantly (p < 0.05) shorter for up to 120 s but never crossed the threshold of 25% change from baseline.(b) RR interval significantly (p < 0.05) shorter for up to 30 s and change from baseline larger than 20 ms for up to 60 s.(c) QT interval significantly (p < 0.05) shorter for up to 90 s and stabilized after 120 s but remained prolonged compared to baseline.(d) QTcF interval significantly (p < 0.05) shorter for up to 120 s and remained stable afterwards within baseline values.

1
Baseline characteristics of young and elderly male subjects in the domiciled clinical setting and the exercise bout group were compared.
Note: Values are shown as mean ± standard deviation.ms, milliseconds.