Inter‐Arm Blood Pressure Difference

Tyger, tyger, burning bright

In the forests of the night,

What immortal hand or eye

Could frame thy fearful symmetry?

‐‐William Blake 1757–1827

Patients frequently complain of lack of expected symmetry: “Doctor, I have different blood pressure (BP) in my left and right arm,” an examining physician may sometimes hear especially since home BP measurement has become more widespread. Such differences may also be documented in the clinical and research settings. According to different studies (and quite different patient populations), the prevalence of a ≥10 mm Hg inter‐arm BP difference (IABPD) may range between 24% in primary care,1 7%2 and 20%3 in research settings, or even 53% in an emergency department setting.4

Hypertension evaluation and treatment guidelines,5, 6 BP measurement guidelines,7, 8 and a widely recognized hypertension textbook9 recommend assessing BP in both arms (at least during initial examination). Nevertheless, they do not specify explicitly how to do it. A relatively recent meta‐analysis of simultaneous vs sequential measurements suggests that a simultaneous measurement is superior,10 as does another important meta‐analysis11 and the study reported in this issue of the Journal by van der Hoeven and colleagues.12 In their carefully designed and executed study, van der Hoeven and colleagues evade most of the methodological problems of previous studies by simultaneously and sequentially measuring BP in both arms with the same BP instrument and cuff in patients stratified by age and BP, as previous studies have suggested these factors may be associated with IABPD. Other factors may be those predisposing for atherosclerosis,2, 13 and indeed peripheral arterial disease (PAD) is frequently associated with IABPD.2, 11, 14 It is also believed that the adverse outcome associated with large IABPD is related to generalized vascular disease. This is supported only in part in a recent study of stroke survivors,14 where an adverse outcome is demonstrated and is associated with PAD, as was found in other studies.2, 11, 15 Nevertheless, somewhat surprisingly, no significant difference in aortic, cerebral, or coronary artery disease was found between patients with large IABPD and those without it.14 This leaves the question to be answered by future research of what causes the excess mortality in patients with large IABPD. Nevertheless, the cumulative recent evidence about the adverse outcome of patients with large IABPD is impressive.2, 11, 14, 15, 16

So far, very few studies have demonstrated the vascular lesion in IABPD, and then in only a handful of patients. Studies have found that a significant subclavian artery stenosis is associated with large IABPD, as summarized nicely in the meta‐analysis of Clark and colleagues.11 That does not mean that IABPD is necessarily a sign of subclavian stenosis, as was shown in a study that used large IABPD17 as a synonym to subclavian stenosis.

A careful study by Kim and coworkers, for instance, did not find such lesions in stroke survivors with large IABPD, where an extensive vascular workup was systematically applied.14 A recent study from China demonstrated an additional benefit of measuring BP simultaneously in all limbs in elderly patients,15 which again suggests an association with PAD.

In addition, the prognostic significance of IABPD is only one motivation for measuring BP in both arms: A usual reason for doing this in the first visit is to determine in which arm BP is higher, so that subsequent BP measurement will be measured in the arm with the higher BP to avoid underdiagnosis and undertreatment of high BP.7, 8, 9 Another motivation is to identify the much less common patient with pronounced obstructive arterial disease, be it for etiologies such as atherosclerotic subclavian stenosis, coarctation of the aorta, or even rarer entities such as Takayasu disease and other vascular diseases in the clinic and acute aortic dissection in the emergency department.

van der Hoeven and associates12 achieved this, as did other researchers, measuring BP with an instrument specifically dedicated for simultaneous measurements such as those used for measuring ankle brachial ratio.2, 14 Yet, other researchers used 2 standard BP instruments attempting near‐simultaneous measurements.3, 18 Although simultaneous measurement is attractive, it should be remembered that there are also nonvascular reasons for IABPD, such as patients with a paretic arm, and the direction of the difference may depend on whether it is a flaccid paralysis (usually lower BP) or a spastic one (frequently a higher BP).19

It makes sense that an instrument designed for simultaneous measurements should be superior. Nevertheless, such instruments are not widely available. Given a web‐promoted price of £1295.00 before VAT, it is not likely to be widely available even in high‐resource countries. Therefore if inter‐arm BP assessment is important, as guidelines suggest and this author also believes, how are most of us to measure it?

First of all, I think we ought to realize that in many of the patients who declare high IABPD, as well as in many with such a difference identified in the doctor's office, it is not a true difference. As Eguchi and colleagues18 have shown, it is to a great extent a simple product of BP variability and measurement, order effect, and regression to the mean, that these effects can be reduced by multiple measurements. This is also suggested in the study by van der Hoeven and collegues,12 where first measurement was higher, be it in simultaneous or sequential measurement. Indeed, it is BP reactivity that is responsible for the first BP measurement to be higher than a subsequent one, thus “IABPD” is the difference between subsequent measurements rather than between the arms in many instances. As this type of reactivity (the white coat‐effect) is more common in older and hypertensive people,7 it should not be surprising that in the sequential 2‐arm measurement studies, age and hypertension are a strong predictor of a large IABPD, as is left ventricular hypertrophy and aortic stiffness.2 With that in mind, it should be remembered that age, hypertension, and vascular and cardiac hypertrophy are intimately biologically entangled, and probably cannot be easily fully separated by either sophistication of epidemiologic design or statistical analysis.

I have suggested previously20 that it is feasible to have repeated measurements in an arm until these yield a stable value, than switch to the other arm. If the value is significantly different (say ≥10 mmHg), switching back to the first arm will reveal in many if not most cases that this IABPD disappears.

Until BP measurement instruments that measure BP simultaneously can be widely available, we can still reasonably assess IABPD with standard BP measurement instruments, as described above. This way we can identify the arm with the higher BP with universally available equipment, employ effective preventive measures, and select the rare patient with a suspicion of significant arterial disease to undergo the appropriate imaging to best identify the culprit lesion.