Low-dose propranolol and exercise capacity in postural tachycardia syndrome

Abstract

Objective:

To determine the effect of low-dose propranolol on maximal exercise capacity in patients with postural tachycardia syndrome (POTS).

Methods:

We compared the effect of placebo vs a single low dose of propranolol (20 mg) on peak oxygen consumption (VO₂max), an established measure of exercise capacity, in 11 patients with POTS and 7 healthy subjects in a randomized, double-blind study. Subjects exercised on a semirecumbent bicycle, with increasing intervals of resistance to maximal effort.

Results:

Maximal exercise capacity was similar between groups following placebo. Low-dose propranolol improved VO₂max in patients with POTS (24.5 ± 0.7 placebo vs 27.6 ± 1.0 mL/min/kg propranolol; p = 0.024), but not healthy subjects. The increase in VO₂max in POTS was associated with attenuated peak heart rate responses (142 ± 8 propranolol vs 165 ± 4 bpm placebo; p = 0.005) and improved stroke volume (81 ± 4 propranolol vs 67 ± 3 mL placebo; p = 0.013). In a separate cohort of POTS patients, neither high-dose propranolol (80 mg) nor metoprolol (100 mg) improved VO₂max, despite similar lowering of heart rate.

Conclusions:

These findings suggest that nonselective β-blockade with propranolol, when used at the low doses frequently used for treatment of POTS, may provide a modest beneficial effect to improve heart rate control and exercise capacity.

Classification of evidence:

This study provides Class II evidence that a single low dose of propranolol (20 mg) as compared with placebo is useful in increasing maximum exercise capacity measured 1 hour after medication.

Postural tachycardia syndrome (POTS) is a heterogeneous disorder characterized by sustained tachycardia (≥30 bpm) upon standing, unrelated to medications or other medical conditions.¹ Although the underlying cause of POTS remains unclear, several mechanisms have been proposed, including sympathetic activation, hypovolemia, and, more recently, deconditioning.^2–4 Indeed, endurance exercise training decreases standing heart rate (HR), restores blood volume, and improves orthostatic tolerance and associated symptoms in POTS.^5–7

Another frequently used treatment for POTS is nonselective β-adrenergic blockade with propranolol.^8,9 At low doses, propranolol reduces standing HR and improves orthostatic symptoms.^8,10,11 In contrast, higher doses can more effectively control tachycardia, but do not improve quality of life or orthostatic symptoms.^5,8 Because β-blockers impair exercise tolerance in healthy subjects,¹² this raises the concern that these drugs would counteract the beneficial effects of exercise training in POTS. We hypothesized, however, that by controlling tachycardia, β-blockers could improve exercise capacity in these patients.

Thus, we compared the effects of placebo vs low-dose propranolol on peak oxygen consumption (VO₂max), an established measure of exercise capacity, in patients with POTS and healthy subjects during a single session of semirecumbent bicycling exercise to maximal effort. As a secondary objective, we assessed whether high-dose propranolol or cardioselective β₁-blockade with metoprolol would provide additional benefit for exercise capacity in POTS.

METHODS

Standard protocol approvals, registrations, and patient consents.

This study was approved by the Vanderbilt University Institutional Review Board. Written informed consent was obtained from all participants. This study was registered at ClinicalTrials.gov (NCT00770484).

Study participants.

We studied 16 female patients with POTS admitted to the Vanderbilt Autonomic Dysfunction Center between November 2008 and October 2011 (figure 1). For comparison, we studied 7 matched female healthy volunteers. The diagnosis of POTS was according to current guidelines.¹ Patients with medical conditions (i.e., prolonged bed rest, dehydration) or medications predisposing to tachycardia were excluded. All participants were nonsmokers, nonpregnant, and none were endurance-trained athletes. Screening included medical history, physical examination, 12-lead ECG, and routine laboratory tests. Subjects with systemic illness, hematologic disease, or abnormalities in liver or renal function were excluded.

Figure 1. Enrollment, treatment allocation, and follow-up of study participants.

POTS = postural tachycardia syndrome.

General protocol.

The patients with POTS received a low-monoamine, methylxanthine-free, and fixed sodium (150-mEq) and potassium (70-mEq) diet. Medications affecting the autonomic nervous system, blood pressure (BP), or volume were withheld ≥5 half-lives before testing. Healthy volunteers were studied on an outpatient basis and withheld from methylxanthine-containing products or medications for ≥3 days before the study.

Autonomic function and orthostatic stress testing.

Autonomic function testing included orthostatic stress, sinus arrhythmia, Valsalva maneuver, hyperventilation, cold pressor, and isometric handgrip.¹³ BP was measured intermittently with an arm cuff by automated oscillometric method (VitalGuard 450C; Ivy Biomedical, Branford, CT) and continuously with the finger volume clamp method (Nexfin; BMEYE, Amsterdam, the Netherlands). HR was measured by continuous ECG (VitalGuard 450C). Orthostatic stress testing was routinely performed at 8:00 am to minimize circadian differences. Subjects remained supine after overnight rest and then stood for 30 minutes, or as long as tolerated. Supine and standing BP and HR were measured by automated sphygmomanometer, and fasting blood samples were collected via an antecubital vein catheter, placed ≥30 minutes before testing. Plasma catecholamines were measured by high-performance liquid chromatography with electrochemical detection.¹⁴

Spectral analysis.

Data segments of 300 seconds were recorded in supine subjects using the WINDAQ acquisition system (DI720; DATAQ, Akron, OH). Data were processed offline using PhysioWave software (Visual Numerics Inc., Boulder, CO) with power spectral densities estimated in low-frequency (0.04–0.15 Hz) and high-frequency (0.15–0.40 Hz) ranges using a fast Fourier transform–based Welch algorithm.¹⁵

Study design.

We performed a randomized, double-blind, crossover study assessing the effects of oral low-dose propranolol (Inderal, 20 mg; Nicholas Piramal) vs placebo on exercise capacity in 11 patients with POTS and 7 healthy subjects. The primary outcome was selected a priori as VO₂max. As a secondary objective, we compared effects of high-dose propranolol (80 mg), equipotent metoprolol (Lopressor, 100 mg; Novartis Pharmaceuticals Corp.), and placebo on VO₂max in a separate cohort of 5 patients with POTS. The order of interventions was randomized within each protocol using computer-generated random numbers, with ≥2 washout days between drugs. Medications were blinded to patients and investigators through distribution of identical pills by the Health and Wellness Pharmacy (Nashville, TN).

Acute recumbent exercise protocol.

Exercise testing commenced 1 hour after receiving medication, consistent with peak plasma levels of these β-blockers.¹⁶ VO₂ was measured at rest and during graded exercise to maximal effort, with VO₂max determined during the final 1 minute of exercise. Exercise testing was valid if 2 of the following were met: a) predicted maximal HR was achieved, b) respiratory exchange rate exceeded 1.15, or c) VO₂ reached a plateau. Subjects were placed on a semirecumbent stationary bicycle (Ergometrics 800; Ergoline, Bitz, Germany), rested for 10 minutes, and were instrumented with a low-resistance mouthpiece connected to a bag to capture expired air. After a 5-minute warmup at 0-W resistance, the workload was increased by 25 W every 2 minutes until maximum effort was achieved. Subjects maintained a speed of 60 rpm and received verbal encouragement to reach maximum effort. The overall testing period lasted approximately 30 minutes. Expired ventilation was processed by an Innocor metabolic system (Innovision, Odense, Denmark) for beat-to-beat VO₂ measurements, with data averaged over the last 30 seconds of each resistance cycle. BP was measured at baseline and at the end of each resistance cycle using an automated sphygmomanometer. HR was measured by continuous ECG and arterial oxygen saturation by pulse oximetry. Cardiac output (CO) was measured at 0 W and 75 W, a submaximal resistance reached by all participants, using the Innocor inert gas rebreathing technique. Stroke volume (SV) was calculated as CO divided by HR, and systemic vascular resistance (SVR) as mean arterial pressure (MAP) divided by CO. Predicted VO₂ was based on established thresholds for sedentary females and predicted HR determined from the formula 210 − (age × 0.65).¹⁷ Predicted VO₂ and HR were expressed as percent achieved during exercise testing.

Statistical analysis.

Data are presented as mean ± SEM. Analyses were performed using SPSS version 19.0 software (IBM Corp., Armonk, NY). A 2-tailed α level <0.05 was defined as statistical significance. The primary objective was to test the null hypothesis that VO₂max will not be different following placebo vs low-dose propranolol in POTS. As a secondary objective, we assessed whether high-dose propranolol or metoprolol would further improve VO₂max. Comparisons of outcomes by intervention within subjects were analyzed using Friedman analysis of variance by ranks. Comparisons of outcomes between groups were analyzed using Mann-Whitney U or Kruskal-Wallis test. We did not correct for order of intervention within protocols.

Sample size.

For protocol 1, we performed a blinded sample size calculation from preliminary data in 3 patients with POTS showing a 2.7 mL/min/kg difference in VO₂max means, an approximate 11% difference, with SD of difference of 2.9 mL/min/kg. Based on these data, we estimated 11 patients would have 80% power to detect a significant difference in VO₂max between treatments. For protocol 2, we expected that high-dose propranolol or metoprolol could conservatively increase VO₂max by 20%, representing a 4.9 mL/min/kg difference in means. Assuming similar variability as preliminary studies, we estimated 5 patients would have 80% power to detect a difference in VO₂max among treatments. These calculations assumed an α level of 0.05 and were performed with paired t test analysis (PS Dupont software, version 3.0.34).

RESULTS

Clinical characteristics.

Subjects were female with no differences in age or body mass index among groups (table 1). Supine low-frequency variability of systolic BP, an indirect measure of sympathetic modulation,¹⁸ was higher in POTS. High-frequency variability of HR (HF_RRI), a measure of cardiac parasympathetic tone,^19,20 was lower in POTS, and low-frequency variability of HR (LF_RRI) was similar among groups. Therefore, the ratio of LF_RRI:HF_RRI, an index of sympathovagal balance,^15,19 was higher in POTS, suggesting increased sympathetic tone. Autonomic function tests revealed a higher HR ratio and increased systolic BP overshoot during phase IV of the Valsalva maneuver in POTS. BP responses to phase II of the Valsalva, hyperventilation, isometric handgrip, and cold pressor tests were similar between groups. Patients with POTS had higher supine and standing HR, with no differences in BP (table 1). Postural norepinephrine increases were higher in POTS, consistent with increased sympathetic tone. While clinical characteristics of patients with POTS in the 2 protocols were similar, sinus arrhythmia ratio was lower in group 2, but was within normal limits.

Table 1.

Clinical characteristics of study participants^a

Baseline hemodynamic measures.

Resting hemodynamic function was not significantly different between study days in either group of subjects, thus results from only the first day are reported. Data for patients with POTS were combined because there were no differences in hemodynamic measures between the 2 protocols. Resting HR was higher in patients with POTS compared with healthy subjects (87 ± 3 vs 66 ± 4 bpm, respectively; p = 0.001), with no differences in MAP (81 ± 2 POTS vs 75 ± 2 mm Hg healthy; p = 0.165). Resting SV was significantly lower in POTS (69 ± 3 POTS vs 85 ± 7 mL healthy; p = 0.047). There were no differences in resting CO (6.0 ± 0.3 POTS vs 5.6 ± 0.5 L/min healthy; p = 0.341), SVR (14.2 ± 0.8 POTS vs 13.1 ± 1.0 mm Hg/L/min healthy; p = 0.535), VO₂ (27.4 ± 1.6 POTS vs 27.2 ± 1.2 mL/min/kg healthy; p = 0.452), or the respiratory quotient (0.77 ± 0.03 POTS vs 0.73 ± 0.05 healthy; p = 0.535) between groups.

Protocol 1: Low-dose propranolol and exercise capacity.

All exercise measures, including VO₂max and peak HR, were similar between groups following placebo, suggesting exercise capacity was not impaired in POTS. Furthermore, all healthy subjects and 9 of 11 patients with POTS achieved >80% predicted VO₂, suggesting most participants were not deconditioned. Peak respiratory quotient and percent predicted HR and VO₂ were not different between groups on either study day (table 2), suggesting a similar level of exertion was achieved.

Table 2.

Effect of low-dose propranolol (20 mg) on maximal exercise capacity and hemodynamic measures^a

In healthy subjects, low-dose propranolol reduced peak HR compared with placebo, with no effect on VO₂max or other measures of exercise capacity (figure 2, table 2). Low-dose propranolol also reduced peak HR in POTS, but, in contrast to healthy subjects, it increased VO₂max (figure 2, table 2; 95% confidence interval [CI] = 25.6–29.6 mL/min/kg; 53.24% absolute risk reduction; 1.88 number needed to treat), lowered anaerobic threshold for HR, and increased peak workload (table 2). Thus, this study provides Class II evidence that a single low dose of propranolol as compared with placebo is useful in increasing maximum exercise capacity in POTS.

Figure 2. Low-dose propranolol improves exercise capacity in patients with POTS.

The effect of placebo vs low-dose propranolol on maximal exercise capacity was determined in 7 healthy subjects and 11 patients with postural tachycardia syndrome (POTS). There was no effect of low-dose propranolol on peak oxygen consumption (VO₂max) in healthy subjects (A), despite a significant lowering of peak heart rate responses (C). In contrast, propranolol significantly improved VO₂max in the patients with POTS (B), and reduced peak heart rate to a similar level as in controls (D).

Exercise increased CO in both groups to a similar extent, with no differences following placebo vs propranolol (figure 3). There was no effect of either treatment on SV in healthy subjects. However, low-dose propranolol increased SV in patients with POTS (figure 3). Exercise increased MAP to a similar extent in both groups, an effect not different following placebo or propranolol (table 2). Similarly, SVR was reduced with exercise in both healthy subjects and patients with POTS, with no differences between placebo and propranolol (table 2).

Figure 3. Hemodynamic changes in response to exercise testing.

Hemodynamic changes in response to semirecumbent exercise testing were assessed at 0-W and at 75-W resistance in healthy subjects and patients with postural tachycardia syndrome (POTS) receiving placebo vs low-dose propranolol. Exercise produced significant increases in cardiac output in healthy subjects (A) and patients with POTS (B), which were not different following placebo vs propranolol. There was no effect of placebo or propranolol on stroke volume in healthy subjects (C). However, low-dose propranolol significantly increased stroke volume during exercise in patients with POTS (D).

Protocol 2: High-dose propranolol and metoprolol on exercise capacity in POTS.

In a separate group of patients with POTS, high-dose propranolol and metoprolol reduced peak HR (143 ± 8 propranolol 80 mg vs 142 ± 10 metoprolol vs 176 ± 7 bpm placebo; p < 0.05) and percent predicted HR achieved (76 ± 5 propranolol 80 mg vs 75 ± 5 metoprolol vs 93 ± 5 placebo; p < 0.05) during exercise. Only 1 of 5 patients with POTS met criteria for deconditioning. However, neither high-dose propranolol (24.6 ± 2.3 mL/min/kg; 95% CI = 20.09–29.11) nor metoprolol (23.9 ± 2.1 mL/min/kg; 95% CI = 19.78–28.02) improved VO₂max compared with placebo (25.8 ± 2.4 mL/min/kg placebo; p = 0.939). Exercise similarly increased CO following placebo (6.7 ± 0.5 baseline vs 8.1 ± 0.6 L/min at 75 W; p = 0.048), high-dose propranolol (6.1 ± 0.7 vs 8.5 ± 0.8 L/min at 75 W; p = 0.028), and metoprolol (5.8 ± 0.6 vs 9.2 ± 1.1 L/min at 75 W; p = 0.043). However, high-dose propranolol did not improve SV (70 ± 6 propranolol vs 75 ± 5 mL placebo at 75 W; p = 0.248) and attenuated MAP increases (99 ± 3 propranolol vs 108 ± 4 mm Hg placebo at 75 W; p = 0.046). In contrast, metoprolol increased SV (82 ± 3 mL; p = 0.028 vs placebo) and maintained MAP (110 ± 10 mm Hg; p = 0.753 vs placebo). All treatments produced similar reductions in SVR with exercise (12.1 ± 1.2 placebo vs 12.0 ± 1.1 propranolol 80 mg vs 11.0 ± 1.7 mm Hg/L/min metoprolol; p = 0.230).

DISCUSSION

The primary findings of this study are that 1) semirecumbent patients with POTS had similar exercise capacity relative to healthy subjects; 2) low-dose propranolol improved VO₂max in POTS, but not healthy subjects; 3) the improvement in VO₂max in POTS was associated with reduced peak HR and increased SV; and 4) despite attenuating peak HR, neither high-dose propranolol nor metoprolol improved exercise capacity in POTS. These data suggest that propranolol can improve HR control and exercise capacity in POTS when given at the low doses frequently used for treatment.

Although there are no established guidelines for treatment, we and others have advocated for use of low-dose propranolol in POTS. Because propranolol has a relatively short duration of action,¹⁶ it can be used as needed before upright activities. At higher doses, however, acute propranolol worsens orthostatic symptoms in POTS.⁸ Chronic treatment with long-acting propranolol (80 mg) does not appear to be a better therapeutic option.⁵ In contrast, endurance exercise training reduces standing HR; increases blood volume, cardiac mass, and exercise capacity; and improves quality of life in POTS.^5–7 Thus, it has been suggested that exercise is more effective than β-blockade for treatment of POTS. However, a direct comparison of these 2 therapies has not been conducted, and the impact of propranolol in combination with exercise has not been determined.

Indeed, low-dose propranolol pretreatment produced an approximate 11% improvement in VO₂max in POTS, an effect similar in magnitude to endurance exercise training.⁶ This improvement was associated with increased SV, perhaps as a compensatory mechanism to maintain CO and BP in the face of reduced HR. The increase in SV could reflect improved venous return or cardiac filling; however, the precise mechanisms were not examined. In contrast, high-dose propranolol did not improve VO₂max in POTS, which may reflect attenuated BP responses during exercise, as well as its known potential to induce fatigue. Although metoprolol increased SV to maintain BP during exercise, there was still no improvement in VO₂max. It is unclear whether lower-dose metoprolol would be beneficial, given our findings for low- vs high-dose propranolol. Importantly, these findings suggest that factors other than SV may be involved in the beneficial effects of low-dose propranolol on exercise capacity.

Of interest, our patients with POTS did not exhibit deconditioning. Because these studies were performed at a tertiary care center for autonomic disorders, these patients may not reflect the broader population. However, the level of VO₂max in this study is similar to observations during upright treadmill exercise in POTS.⁶ Furthermore, previous studies suggest there are no differences in VO₂max between adolescent POTS vs non-POTS patients.²¹ Thus, even nondeconditioned patients could potentially benefit from exercise and low-dose propranolol. In contrast, a recent retrospective study showed deconditioning in >90% of patients with POTS after upright exercise testing.²² This discrepancy may reflect selection bias as well as differences in exercise protocols.

There are potential limitations to the present study. First, we did not examine for sex differences because this condition predominately affects premenopausal women. Second, there were a relatively small number of subjects in this study, especially in the second protocol. Given the small sample size, we were not able to effectively assess for carryover effects. Furthermore, while some variables may have reached significance with additional subjects, we found a trend toward worsening of VO₂max with high-dose propranolol and metoprolol; based on our initial data, we estimated that 59 and 20 subjects, respectively, would be needed to confirm these findings. Given these large numbers, and that these drugs did not improve VO₂max, we decided not to enroll further patients. We also did not compare low- vs high-dose propranolol in the same patients. Third, healthy volunteers were not placed on a controlled diet, which could influence exercise testing. Fourth, it is unclear whether chronic propranolol would provide similar benefit for exercise capacity. Finally, because propranolol is permeable across the blood-brain barrier, we could not differentiate whether there was a contribution of central vs peripheral actions for effects on exercise capacity.

Overall, our results suggest that propranolol, when used at low doses, may provide a modest beneficial effect to improve HR control and exercise capacity in POTS. Given these findings, β-blockers should be tested as an adjuvant to established exercise training programs to provide a more comprehensive treatment strategy in these patients. Furthermore, our findings suggest that despite having a similar cardiovascular phenotype to deconditioned subjects, not all POTS manifests as a consequence of deconditioning. Rather, varying degrees of deconditioning may occur from changes secondary to orthostatic tachycardia. Thus, optimal treatment strategies should be determined on an individual basis and may include both pharmacologic and nonpharmacologic measures.

GLOSSARY

BP

blood pressure

CI

confidence interval

CO

cardiac output

HF_RRI

high-frequency R-R interval

HR

heat rate

LF_RRI

low-frequency R-R interval

MAP

mean arterial pressure

POTS

postural tachycardia syndrome

SV

stroke volume

SVR

systemic vascular resistance

VO₂max

peak oxygen consumption

AUTHOR CONTRIBUTIONS

Amy C. Arnold: drafting/revising the manuscript, analysis or interpretation of data, acquisition of data, statistical analysis. Luis E. Okamoto: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, acquisition of data. André Diedrich: drafting/revising the manuscript, analysis or interpretation of data, acquisition of data, statistical analysis. Sachin Y. Paranjape: drafting/revising the manuscript, acquisition of data, study supervision. Satish R. Raj: drafting/revising the manuscript, study concept or design, analysis or interpretation of data. Italo Biaggioni: drafting/revising the manuscript, study concept or design, study supervision. Alfredo Gamboa: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, acquisition of data, statistical analysis.

STUDY FUNDING

Supported by NIH grants P01 HL056693, 5U54NS065736, and UL1 TR000445.

DISCLOSURE

A. Arnold is funded by American Heart Association grant 11POST7330010. L. Okamoto, A. Diedrich, and S. Paranjape report no disclosures. S. Raj is funded by NIH grant R01 HL102387 and provides expert medical consulting for law firms regarding POTS patients. I. Biaggioni is a consultant for Chelsea Therapeutics and Astra Zeneca, is funded by NIH grants P01 HL056693 and U54 NS065736, and receives research support from Astra Zeneca and Forest Laboratories. A. Gamboa is funded by NIH grant K23 HL95905. Go to Neurology.org for full disclosures.