Ivabradine Lowered Heart Rate in Long COVID POTS Without Improving Symptoms
The largest randomized drug trial ever run in POTS gave patients a medication that reliably lowers heart rate, then measured whether they felt any better. Their heart rates came down. Their symptom scores stayed where they were.
Those are the headline results of RECOVER-AUTONOMIC, the NIH-funded trial of ivabradine in Long COVID POTS, presented on March 28, 2026 in a Late-Breaking Clinical Trials session at the American College of Cardiology Annual Scientific Session. A full journal publication is still pending, so what follows comes from the trial registration, the published design paper, and the investigators' own presentation of the findings.
What the RECOVER-AUTONOMIC ivabradine trial tested
RECOVER-AUTONOMIC is a platform trial, meaning it runs more than one treatment question under a single master protocol. The platform enrolled 381 participants across two drug arms, one testing ivabradine and one testing intravenous immunoglobulin. The ivabradine study enrolled 181 adults with POTS that developed after COVID-19, ran from March 2024 to October 2025, and was registered as a phase 2 trial. Pam Taub, a cardiologist and professor of medicine at UC San Diego who presented the results, described it as the largest trial ever conducted in POTS patients.
Ivabradine was a deliberate choice for this question. It lowers heart rate by selectively inhibiting the sinoatrial node, and unlike beta blockers and calcium channel blockers it does not lower blood pressure or reduce the force of cardiac contraction. That selectivity is what makes the trial interesting. It comes close to isolating heart rate as a single variable and asking what happens to a patient when you move it.
Participants were randomized twice. The first randomization assigned them to ivabradine or placebo for 3 months. The second assigned them to coordinated care or usual care over the same period. Coordinated care meant volume expansion through a high salt diet and water intake, an abdominal binder, exercise and rehabilitation, education, and assisted care through a care coordinator. Usual care meant a one-page handout of instructions.
One detail in the registration is worth pulling out. The primary endpoint was the change in the Orthostatic Hypotension Questionnaire composite score, a 6-item symptom scale paired with a 4-item daily activity scale, each item scored from 0 to 10. Change in heart rate was listed as a secondary outcome. The trial was built from the start to be judged on how patients felt, with heart rate as a supporting measurement.
Heart rate fell and symptom scores stayed flat
Ivabradine produced a significant reduction in heart rate compared with placebo. On the registered symptom endpoint, it produced no significant improvement.
The drug did exactly what it is designed to do at the level of the sinus node, and the people taking it did not report feeling better for it. Because ivabradine leaves blood pressure and contractility alone, this is about as clean a test as the field has of whether lowering the heart rate in POTS is itself therapeutic. Over 3 months, in 181 patients, it was not.
Why lowering heart rate may not relieve POTS symptoms
The result is less surprising than it first sounds, because the tachycardia in POTS has always been the body's response to something upstream rather than the thing that makes people sick. When a person stands and venous return drops, raising heart rate is one of the ways the circulation defends output. A drug that blocks that response removes the compensation while leaving whatever provoked it untouched.
This is a thread that runs through much of the existing literature. Heart rate elevation does not predict how severe a patient's symptoms are. POTS affects far more systems than heart rate and blood pressure. Cerebral blood flow can fall before the heart rate rises at all, and orthostatic intolerance tracks brain perfusion more closely than it tracks the numbers on a monitor.
RECOVER-AUTONOMIC did not measure cerebral blood flow, so the trial itself cannot confirm that explanation. What it does establish is the gap: a measurable, drug-responsive number moved, and the patients attached to it did not. Taub's own summary for clinicians was that heart rate lowering alone will not translate into improvement in patient-reported outcomes, and that POTS after COVID is a complex multi-system disease rather than a cardiac one.
The coordinated care arm produced the improvement
The second randomization is where something did work. In a prespecified analysis, participants who received ivabradine plus coordinated care improved on the symptom questionnaire compared with those who received ivabradine plus usual care, with a p value for the interaction of 0.0042.
The component that moved symptoms was the non-drug program. Salt, fluids, an abdominal binder, graded exercise and rehabilitation, education, and a coordinator staying in contact with the patient. Taub drew the comparison to cardiac care after a heart attack, where a beta blocker and a statin are standard but the functional gains come from rehab, and argued that the same synergy applies here.
One caveat belongs alongside that finding, and the investigators put it in writing themselves. Section 6.3 of the RECOVER-AUTONOMIC protocol states that participants and investigators "will be blinded throughout the study for the study intervention, but not for the non-pharmacologic intervention," because "due to the nature of non-pharmacologic intervention, blinding is not possible." A program that supplies equipment, education and regular human contact carries attention and expectation effects that a one-page handout does not. The result is real and it is prespecified, but it is a different grade of evidence than the blinded drug result sitting next to it.
That design choice is worth noticing on its own terms. Faced with an intervention that cannot be placebo-controlled, an NIH-funded trial did not set it aside as unmeasurable. The protocol's scientific rationale section says the factorial design was used "where the non-pharmacologic care is not blinded due to its practical implementation," and the trial randomized it anyway. Rigor came from randomization and prespecification rather than from masking, and the arm that could not be blinded is the one that moved the primary endpoint.
How this fits the rest of the POTS literature
POTS is a syndrome with multiple upstream causes rather than a single disease, which is why treatment has to match the driver in the individual patient. A trial that applies one heart rate drug to everyone who meets the heart rate criterion is testing the syndrome definition as much as it is testing the drug.
It also sits alongside the repeated finding that the standard measurements and the patient's experience come apart. How a patient feels does not predict what their autonomic testing shows, and normal vitals do not establish normal brain blood flow. RECOVER-AUTONOMIC adds the treatment-side version of that same dissociation. Peter Novak, whose work on treating central rather than peripheral problems we have covered, is a co-author on the trial's design and rationale paper.
What the trial leaves unanswered
These results were presented at a conference and have not yet been published in full, so effect sizes, confidence intervals and subgroup analyses are still to come. The drug exposure was 3 months in a phase 2 trial of 181 patients, and a null result on a symptom questionnaire over that window does not rule out benefit in a narrower group or over a longer period.
Taub was direct about that limitation. POTS has multiple phenotypes, patients do not all behave the same way, and identifying subtypes through biomarkers such as norepinephrine levels could reveal a hyperadrenergic group that does respond to heart rate control. The larger mechanistic questions are still open, including immune overactivation, autoantibodies, persistent viral reservoirs, endothelial dysfunction and mitochondrial function, and she named all of them as targets for the next round of research. The immunoglobulin arm of the same platform is a separate trial with its own results.
For a patient whose tachycardia is itself intolerable, ivabradine may still be worth trying, and this trial does not say otherwise. What it says is that bringing the number down should not be mistaken for treating the illness, and that the part of the protocol that actually moved how people felt was the part aimed at the patient rather than at the reading.