Drugs That Distort Heart Rate Variability (HRV): What Skews Your Results

What HRV Actually Measures and Why Medications Matter

Heart rate variability quantifies the fluctuation in time intervals between consecutive heartbeats, reported as millisecond-level variation in R-R intervals on an electrocardiogram. A healthy heart does not beat like a metronome. It speeds up during inhalation and slows during exhalation, a pattern called respiratory sinus arrhythmia governed by the vagus nerve 1.

The two branches of the autonomic nervous system (the sympathetic "fight-or-flight" arm and the parasympathetic "rest-and-digest" arm) constantly modulate cardiac pacing through the sinoatrial node. HRV metrics capture this tug-of-war. Time-domain measures like SDNN (standard deviation of normal-to-normal intervals) and RMSSD (root mean square of successive differences) reflect overall variability and vagal tone, respectively. Frequency-domain analysis separates low-frequency (LF, 0.04-0.15 Hz) power, often linked to mixed sympathetic and parasympathetic input, from high-frequency (HF, 0.15-0.4 Hz) power, which tracks parasympathetic activity almost exclusively 2.

Any drug that alters sinoatrial node firing, vagal transmission, or sympathetic outflow will change these numbers. The clinical problem is straightforward: a patient on metoprolol may show an SDNN of 160 ms that looks reassuringly normal, but the reading reflects pharmacology, not genuine autonomic resilience. The 1996 Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology stated that "pharmacological interventions that modify autonomic tone will predictably alter HRV indices, and this effect must be considered in interpretation" 2.

Beta-Blockers: The Most Common HRV Inflator

Beta-adrenergic blockers are the single most prescribed drug class that distorts HRV readings. They work. By blocking sympathetic input to the heart, they lengthen R-R intervals and widen beat-to-beat variability, producing higher SDNN and RMSSD values that can mimic improved vagal tone.

A 2003 study published in the American Journal of Cardiology found that metoprolol 100 mg/day increased SDNN by a mean of 34% and RMSSD by 28% in post-myocardial infarction patients over 8 weeks, independent of any change in underlying autonomic function 3. Propranolol, a non-selective beta-blocker, showed even larger HF power increases in a crossover trial of healthy volunteers because it also blocks beta-2 receptors on bronchial smooth muscle, deepening respiratory sinus arrhythmia 4.

Cardioselective agents (atenolol, bisoprolol, metoprolol) inflate HRV less dramatically than non-selective agents (propranolol, carvedilol, nadolol). The key point for patients tracking wearable HRV data is this: if you start a beta-blocker and your Oura ring or Apple Watch shows a 30-point RMSSD jump overnight, that is not recovery. That is the drug.

Dr. George Billman, a cardiac electrophysiologist at Ohio State University, has noted: "Beta-blocker-induced increases in HRV do not carry the same prognostic significance as intrinsic increases in vagal tone. The underlying autonomic substrate remains unchanged" 5.

Anticholinergic Medications: Stripping Away Parasympathetic Signal

Anticholinergic drugs block muscarinic acetylcholine receptors at the sinoatrial node, directly suppressing the vagal brake that generates HRV's high-frequency component. The result is a flattened HRV profile with low RMSSD and collapsed HF power.

This class includes medications patients may not think of as cardiac drugs. Oxybutynin (for overactive bladder), diphenhydramine (Benadryl), tricyclic antidepressants like amitriptyline, and first-generation antipsychotics like chlorpromazine all carry anticholinergic burden 6. A 2008 analysis in the Journal of the American Geriatrics Society examined 544 older adults and found that those with a high anticholinergic burden score (ACB score ≥3) had SDNN values 22% lower and HF power 38% lower than matched controls taking no anticholinergics 6.

Atropine, the prototypical muscarinic antagonist, eliminates respiratory sinus arrhythmia almost entirely at doses above 0.04 mg/kg 7. This makes atropine a useful pharmacological probe in research settings (confirming that HF power is vagally mediated) but a confound in clinical HRV interpretation. Patients on combination regimens that include an anticholinergic inhaler (ipratropium, tiotropium) plus an overactive bladder agent plus a sedating antihistamine may show profoundly suppressed HRV that mimics severe autonomic neuropathy. The drug stack, not the nervous system, is the problem.

SSRIs and SNRIs: Direction Depends on the Agent

Selective serotonin reuptake inhibitors produce inconsistent HRV effects that vary by drug, dose, and treatment duration. This inconsistency itself is clinically important because depression lowers HRV 8, and SSRIs treat depression, creating competing pharmacological and therapeutic signals.

Paroxetine, the SSRI with the strongest anticholinergic side-effect profile, tends to reduce HRV. A 2000 study in Biological Psychiatry showed that paroxetine 20 mg/day decreased HF power by 12% over 16 weeks compared to placebo 9. Sertraline, by contrast, showed a neutral-to-positive effect on HRV in the SADHART trial (N=369), where sertraline-treated post-ACS depression patients had slightly higher SDNN at 16 weeks versus placebo, although the difference did not reach statistical significance (P=0.07) 10.

SNRIs (venlafaxine, duloxetine) add norepinephrine reuptake inhibition to the serotonergic effect, which can increase resting heart rate by 2-7 bpm and lower overall HRV. Venlafaxine at doses above 150 mg/day has been associated with dose-dependent reductions in RMSSD of 15-20% 11. For clinicians ordering HRV analysis in depressed patients, the practical question is not whether to stop the antidepressant (usually, no) but whether to adjust the reference range and document the agent.

Opioids: Suppressing Variability Across All Domains

Opioid medications reduce HRV through centrally mediated suppression of both sympathetic and parasympathetic outflow, with a disproportionate effect on vagal tone. Acute morphine administration decreases HF power by 30-50% within 60 minutes of intravenous dosing 12.

Chronic opioid therapy presents a larger interpretive challenge. A cross-sectional study of 75 chronic non-cancer pain patients on long-term opioid therapy (mean morphine equivalent dose 90 mg/day) found that mean SDNN was 89 ms, roughly 25% below age-matched controls, and RMSSD was 18 ms, well below the 20 ms threshold often used to flag autonomic dysfunction 13. Methadone carries additional risk because it blocks the hERG potassium channel, prolonging QTc and introducing repolarization abnormalities that can confound automated HRV algorithms 14.

Buprenorphine appears to have a milder HRV-suppressive effect than full mu-agonists, likely because of its partial agonist pharmacology 13. Patients transitioning from methadone to buprenorphine may see HRV metrics rise without any actual change in autonomic health.

Stimulants, Sympathomimetics, and Thyroid Hormones

Amphetamine-based medications (Adderall, Vyvanse), methylphenidate (Ritalin, Concerta), and pseudoephedrine all increase sympathetic outflow and shift the LF/HF ratio toward sympathetic dominance. A controlled study of 30 adults with ADHD found that methylphenidate 36 mg/day reduced RMSSD by 19% and increased the LF/HF ratio by 0.8 units after 4 weeks of treatment 15.

Exogenous thyroid hormone (levothyroxine) in supraphysiologic doses produces similar effects. Subclinical hyperthyroidism, whether iatrogenic or endogenous, increases resting heart rate and reduces HRV. A study of 42 patients on TSH-suppressive levothyroxine therapy for differentiated thyroid cancer found that SDNN was 18% lower and HF power was 29% lower compared to euthyroid controls 16. Even modest over-replacement (TSH 0.1-0.4 mIU/L) may affect HRV readings in patients tracking autonomic markers for fitness or recovery purposes.

Caffeine, though not a prescription drug, deserves mention. Doses above 300 mg (roughly three cups of coffee) acutely reduce HF power and increase the LF/HF ratio for 2-4 hours post-ingestion 17. Standardizing caffeine intake before HRV recordings is a basic but often neglected step.

GLP-1 Receptor Agonists and Other Newer Agents

GLP-1 receptor agonists (semaglutide, tirzepatide, liraglutide) increase resting heart rate by a mean of 2-4 beats per minute, an effect documented across multiple phase 3 trials. In SUSTAIN-6 (N=3,297), semaglutide was associated with a 2.5 bpm increase in resting heart rate versus placebo at 104 weeks 18. This baseline shift lowers time-domain HRV metrics even if autonomic function is unchanged.

The mechanism appears to involve direct GLP-1 receptor activation at the sinoatrial node rather than sympathetic stimulation, which means the usual frequency-domain interpretation (higher LF/HF = more sympathetic drive) may not apply 19. For the growing population of patients on semaglutide or tirzepatide who track wearable HRV, this is a practical concern: their overnight RMSSD will trend 5-15% lower than pre-medication baseline without indicating worse autonomic health.

Ivabradine, a funny-channel (If) blocker, presents the opposite pattern. By selectively slowing sinoatrial firing without affecting contractility or autonomic signaling, ivabradine increases time-domain HRV metrics. A 2014 trial in patients with stable coronary artery disease found that ivabradine 7.5 mg twice daily increased SDNN by 17 ms (from 112 to 129 ms) and RMSSD by 9 ms after 8 weeks 20. This increase is pharmacological, not a sign of improved vagal fitness.

How to Interpret HRV When You Cannot Stop Medications

Most patients taking cardiovascular, psychiatric, or pain medications cannot discontinue them for the sake of a clean HRV reading. The practical approach involves documentation, within-person trending, and appropriate reference adjustments.

First, every HRV report (clinical or consumer wearable export) should list concurrent medications with doses. The European Heart Rhythm Association's 2023 consensus statement on HRV assessment recommends that "a complete pharmacological inventory accompany all HRV analyses, as interpretation without medication context is inherently unreliable" 21.

Second, within-person longitudinal trends are more useful than single-timepoint comparisons to population norms. A patient on stable metoprolol 50 mg/day whose RMSSD drops from 55 ms to 32 ms over 6 months is showing a meaningful decline in autonomic function, even though both values fall within "normal" range. The drug effect is constant; the change is the signal.

Third, frequency-domain analysis can sometimes separate drug effects from autonomic effects. Beta-blockers primarily increase HF power (parasympathetic-associated), while stimulants primarily increase LF power (sympathetically-associated). If a patient on a beta-blocker shows rising LF and falling HF, the underlying sympathetic shift is real despite the drug's HF-boosting effect 2.

For research-grade HRV assessment, the gold standard remains a minimum 48-72 hour washout of the offending medication, performed under physician supervision 22. This is feasible for some agents (caffeine, short-acting stimulants) and infeasible for others (beta-blockers in heart failure, opioids in chronic pain). When washout is not possible, drug class and dose should be treated as covariates in any statistical analysis.

Common Medication Classes and Their HRV Effects: A Quick-Reference

Beta-blockers (metoprolol, propranolol, atenolol): Increase SDNN 20-40%, increase RMSSD and HF power. Effect begins within 24-48 hours of first dose.

Anticholinergics (oxybutynin, diphenhydramine, amitriptyline): Decrease RMSSD, decrease HF power by 25-40%. Dose-dependent suppression of vagal markers.

SSRIs (sertraline, paroxetine, fluoxetine): Variable. Paroxetine tends to decrease HRV; sertraline is neutral. Effects emerge over 4-8 weeks.

SNRIs (venlafaxine, duloxetine): Decrease overall HRV at higher doses, increase resting heart rate 2-7 bpm.

Opioids (morphine, oxycodone, methadone): Decrease all HRV domains. HF power drops 30-50% acutely. Chronic use lowers SDNN by approximately 25%.

Stimulants (amphetamine, methylphenidate): Shift LF/HF ratio toward sympathetic dominance, decrease RMSSD 15-20%.

GLP-1 RAs (semaglutide, tirzepatide): Increase resting heart rate 2-4 bpm, reduce time-domain metrics 5-15%.

Ivabradine: Increases SDNN and RMSSD by slowing sinoatrial rate without autonomic modulation.

Thyroid hormone (levothyroxine, excess): Decreases SDNN and HF power when doses suppress TSH below 0.4 mIU/L.

Normal HRV Ranges and How Drugs Shift Them

Healthy adult SDNN on 24-hour Holter monitoring typically ranges from 100 to 180 ms, with values below 50 ms considered a marker of increased cardiovascular risk in the post-MI population 2. Short-term (5-minute) RMSSD in resting healthy adults averages 20-75 ms, with age being the strongest non-pharmacological determinant: RMSSD declines approximately 3-4 ms per decade after age 30 23.

Consumer wearables report overnight RMSSD or proprietary "HRV scores" that are not directly comparable to clinical 5-minute or 24-hour recordings. A patient whose Apple Watch HRV reads 45 ms while on atenolol 25 mg/day might have an unmedicated baseline closer to 30-35 ms. Without the drug context, the wearable data suggests better autonomic fitness than actually exists.

The direction of distortion matters clinically. Artificially elevated HRV (from beta-blockers or ivabradine) may mask autonomic deterioration in conditions like diabetic cardiac autonomic neuropathy, where falling HRV is an early warning sign 24. Artificially suppressed HRV (from anticholinergics or opioids) may trigger unnecessary autonomic testing or cardiology referrals. Both errors waste resources and mislead patients.

The American Diabetes Association's 2024 Standards of Care recommend cardiovascular autonomic reflex testing (CARTs), not passive HRV monitoring, as the diagnostic standard for cardiac autonomic neuropathy, precisely because HRV is too susceptible to pharmacological confounding 25.