Adults whose resting heart rate falls below 50 beats per minute or climbs to 90 and above face a measurably higher risk of stroke than those whose pulse sits in the 60 to 69 bpm range, according to a large-scale UK Biobank analysis presented at the European Stroke Organisation Conference in 2026. The finding holds even after accounting for atrial fibrillation, the irregular rhythm most commonly linked to stroke. The results trace a U-shaped curve: risk rises at both ends of the heart-rate spectrum, challenging the assumption that a slow resting pulse is always a sign of cardiovascular fitness.
Why a U-shaped stroke curve demands attention now
Most stroke-prevention strategies focus on atrial fibrillation screening, blood-pressure control, and cholesterol management. Resting heart rate rarely features in that conversation. The UK Biobank analysis, cataloged as the conference abstract ESOC2026A395, shifts that calculus by showing that the association between resting heart rate and incident stroke is independent of atrial fibrillation. In practical terms, a patient cleared of AF could still carry elevated stroke risk if their pulse consistently lands outside the 60 to 69 bpm window.
The hypothesis worth testing is whether individuals whose resting heart rate drifts outside that window between two separate assessment-center visits face a steeper rise in subsequent stroke incidence than those whose rate stays stable, regardless of a single baseline reading. UK Biobank collects pulse rate during automated blood-pressure measurement at its assessment centers, and repeat visits create the longitudinal data needed to track that drift. If confirmed, heart-rate trajectory rather than a single snapshot could become a practical screening signal.
That prospect matters now because low-cost wearables and home monitors have made pulse tracking ubiquitous, but clinical risk tools have not caught up. Current stroke calculators emphasize age, blood pressure, smoking status, diabetes, and prior cardiovascular events. A reproducible U-shaped curve, if validated in full peer-reviewed form, would argue for adding resting heart rate as a routinely assessed vital sign with clear interpretive thresholds rather than a number that clinicians glance at and ignore.
Three cohorts, one consistent signal linking pulse to stroke
The UK Biobank result does not stand alone. In the United States, the REGARDS study, a large prospective cohort, found that a higher resting heart rate was linked to increased ischemic stroke risk. That work focused on the upper end of the heart-rate distribution, establishing that faster resting pulses carry danger even after adjusting for traditional cardiovascular risk factors such as hypertension, smoking, and diabetes. The association persisted when participants with known atrial fibrillation were excluded, suggesting that heart rate itself, or the physiology it reflects, contributes additional risk.
Separately, the Framingham Study tied elevated resting heart rate and autonomic imbalance to incident stroke, offering a mechanistic explanation. A chronically fast pulse may reflect sympathetic overdrive and reduced vagal tone, both of which can stiffen blood vessels, impair endothelial function, and promote pro-thrombotic states. Over years, that autonomic imbalance could accelerate atherosclerosis in cerebral arteries or increase the likelihood that small clots form and travel to the brain.
What the UK Biobank analysis adds is the left side of the curve. While REGARDS and Framingham concentrated on high heart rates, the new data show that very low rates, below 50 bpm, also correlate with higher stroke incidence. That pattern produces the U shape. Athletes and highly conditioned individuals often have resting rates in the 40s, so the finding raises a question about whether bradycardia in the general population signals something different from athletic conditioning, perhaps subclinical conduction disease, autonomic dysfunction, or medication effects that impair cerebral perfusion.
All three datasets rely on resting heart rate captured through standardized clinical protocols. In UK Biobank, the exposure variable is recorded in Data-Field 102, which logs pulse rate in beats per minute during blood-pressure measurement. Stroke outcomes in UK Biobank are identified through linked coded healthcare data, and a validation study reported positive predictive values that support the reliability of incident stroke capture. That validation matters because misclassified outcomes would be expected to blur the U-shaped signal rather than create it, meaning the true association could be even sharper than reported if outcome coding is imperfect.
Open questions about heart-rate drift and stroke prediction
Several gaps remain. The conference abstract does not publish full adjusted hazard ratios with confidence intervals for each heart-rate category, so the exact magnitude of risk at the extremes is not yet available for independent scrutiny. Peer review of the complete manuscript will determine whether the U shape survives adjustment for medications such as beta-blockers and calcium-channel blockers, which lower heart rate pharmacologically and carry their own cardiovascular profiles. Without those details, clinicians should treat the current signal as hypothesis-generating rather than practice-changing.
Exact counts of incident stroke cases stratified by both resting-heart-rate band and AF status have not been released. Without those numbers, it is difficult to judge whether the low-heart-rate arm of the curve is driven by a small subgroup with conduction abnormalities, pacemaker indications, or structural heart disease, or whether it reflects a broader population pattern. The distinction matters for clinical translation: a signal concentrated in a narrow subgroup would call for targeted screening and follow-up, while a population-wide gradient would argue for integrating heart rate into general risk calculators used in primary care.
The repeat-measurement question is equally unresolved. UK Biobank’s physical-measurement protocol includes repeat assessment visits, but the abstract and press materials do not specify whether heart-rate change over time was analyzed or whether the findings rest on a single visit. If future analyses show that individuals whose pulse migrates from the 60 to 69 bpm range into either extreme between visits face disproportionately higher stroke rates, that would strengthen the case for serial monitoring rather than one-time measurement. It would also align with the growing recognition that dynamic changes in risk markers, such as rising blood pressure or new-onset atrial fibrillation, often herald vascular events more strongly than static baseline values.
Another unanswered issue is causality. A U-shaped association does not by itself prove that modifying resting heart rate will alter stroke risk. Low or high pulse may be a marker of underlying pathology rather than a modifiable driver. Randomized trials of heart-rate–lowering or –raising interventions with stroke outcomes are unlikely to be conducted at scale, so researchers will need to rely on sophisticated observational methods, including Mendelian randomization and time-updated analyses, to separate signal from confounding.
What patients and clinicians can do now
For anyone who tracks resting heart rate through a wearable device or clinical visits, the immediate implication is not panic but context. A single low or high reading, especially during illness, stress, or after caffeine, is less informative than a stable pattern over weeks to months. Individuals whose usual resting pulse sits well below 50 bpm or persistently above 90 bpm should mention this to their clinician, particularly if they are older, have other vascular risk factors, or experience symptoms such as dizziness, palpitations, shortness of breath, or reduced exercise tolerance.
Clinicians, in turn, can treat heart rate as a low-cost prompt for deeper evaluation rather than an isolated target. For a very slow pulse, that might include reviewing medications, checking for conduction abnormalities on an electrocardiogram, and asking about syncope or near-syncope. For a fast pulse, it might mean screening for anemia, thyroid disease, sleep apnea, deconditioning, or occult arrhythmias. In both directions, the goal is to uncover and address the underlying process that may be increasing stroke risk, rather than simply chasing a specific number.
Until the full UK Biobank analysis is published, guideline-level recommendations are unlikely to change. Nonetheless, the emerging U-shaped curve offers a practical takeaway: resting heart rate, long relegated to the margins of the chart, deserves a more deliberate look. As larger datasets and repeat measurements clarify how pulse and stroke risk move together over time, heart rate may evolve from a background vital sign into a meaningful component of stroke prevention strategies.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
