Different Blood Pressure in Both Arms: When Is the Difference Worrying?

Blood reaches the arms through the subclavian arteries. The right one arises from the brachiocephalic trunk, the left directly from the aortic arch, so their course and angle of origin are not symmetrical. If a narrowing, most often atherosclerotic, develops in one of these arteries, the pressure in the arm lying 'beyond' it falls, because part of the flow energy is lost at the site of the narrowing. The effect is greater the more severe the narrowing. Besides atherosclerosis, less common causes include inflammation of large vessels (such as Takayasu disease), compression of vessels at the thoracic outlet, and congenital aortic defects. A serious but rare cause of a sudden pressure difference is aortic dissection - which is why we describe the warning signs below.

The best source of numbers is the meta-analysis published in The Lancet in 2012, which pooled studies on the link between the blood pressure difference between arms and vascular disease and mortality.

Note two numbers: sensitivity and specificity. At the 15 mmHg threshold, sensitivity for peripheral artery disease is only 15%, so most people with that disease do not have such a large difference. Specificity, on the other hand, is high (96%), so a large difference is a signal worth attention, but the absence of a difference rules nothing out. It is a test that 'catches' some of the sick, not a screening tool that replaces vascular examination.

The other side of the issue is error. Blood pressure changes from heartbeat to heartbeat, so if you measure the right arm first and the left a moment later, part of the difference simply reflects the passage of time. Technique also matters greatly: a cuff that is too small or too large on one arm, a different position of the hand relative to the heart, talking during measurement, and even a difference in arm circumference in people with a dominant, muscular arm. A systematic review examined how the measurement method affects how often a difference is detected.

Similar conclusions come from a review on the prevalence of the difference (Clark et al., 2016): a systolic difference of at least 10 mmHg affected 3.6% of adults in the general population, 7.4% of people with diabetes and 11.2% of people with hypertension, and sequential measurement overestimated its frequency almost threefold compared with simultaneous measurement (RR 2.9). In other words: the worse the technique, the more 'false' differences.

The stability of the difference is also limited. In the MESA study (6,725 people aged 45-84, Doppler pressure measured on both arms three times over 9.5 years), a high inter-arm difference often did not persist at subsequent examinations. The practical conclusion: a single large result needs to be repeated, preferably on another day.

The size of the effect is moderate: a hazard ratio of 1.05 per 5 mmHg is much less than the influence of age, smoking or diabetes. The inter-arm difference is therefore more of an 'additional clue' in risk assessment than a diagnosis in itself. A more detailed picture comes from the MESA study, which evaluates not only the size of the difference but also its persistence.

Guidelines recommend measuring blood pressure on both arms at the first assessment and using the arm with the higher reading for later measurements. The justification was provided by the INTERPRESS-IPD analysis of 2022 (23 cohorts, 53,172 participants).

In home practice this means: at the start measure both arms alternately, establish on which arm the result is higher, and measure on that arm in the following weeks. This rule protects against underestimating blood pressure - 12% of people changed category depending on which arm the result was taken from. If the difference is large or changes from measurement to measurement, tell your doctor and do not switch arms without a consultation.

The most accurate assessment is simultaneous, that is measuring both arms at the same moment with two devices, but at home it is rarely possible. The sequential version is an acceptable compromise if you take several alternating measurements and average them - this limits the effect of natural variability shown by the review of Verberk et al.

With a reproducible difference of at least 10-15 mmHg, a doctor will usually start by examining the pulse in both arms and listening over the arteries, and then order vascular imaging. A first step is often the ankle-brachial index (ABI), which assesses the general state of the limb arteries, and Doppler ultrasound of the subclavian and carotid arteries. If significant narrowing is suspected, further diagnostics are CT or MR angiography. You can read more about the tests themselves in our knowledge-base entries on vascular Doppler ultrasound and on angiography and CT angiography.

Diagnosis of subclavian stenosis also changes the approach to measurements: in the arm beyond the narrowing, pressure is falsely low, so the doctor decides which arm to regard as reliable. That is why you should not switch arms 'on your own' or base your treatment on one arm if the other gives significantly different results. Subclavian stenosis does not always cause symptoms; when they appear, they may be pain or fatigue of the arm during exertion, a cold, pale hand, or dizziness during intense arm work.

Equally mistaken is the belief that 'the left arm is more important because it is closer to the heart'. In the MESA analysis risk did not depend on the side on which pressure was higher, and the advice to measure the higher arm comes from the need to avoid underestimating pressure, not from a preference for one side of the body.

Limitations: the results we rely on are mainly observational data from cohorts and reviews; at the 15 mmHg threshold sensitivity for peripheral artery disease is only 15%, so the absence of a difference rules nothing out. Estimates differ depending on the population (different in the general population, in diabetes and in hypertension), on the measurement method, and on whether the difference was confirmed several times. There is also no single, universally accepted threshold for diagnosing disease: 10 mmHg is a proposal for the upper limit of normal, and 15 mmHg indicates a stronger link with risk. Finally, most data concern systolic pressure - the significance of a diastolic difference is less well known. Any diagnostic work-up is always decided by a doctor.

Measure both arms once properly, not ten times nervously. If after several alternating measurements the systolic difference is less than 10 mmHg, forget about it and keep measuring on the arm with the higher result. If it is at least 10-15 mmHg and repeats on subsequent days, treat it as a valuable diagnostic signal: it is not proof of disease, but it is a good reason for a doctor to examine the arteries. That one extra minute at the first measurement is one of the cheapest tests you can do.