Answers are grounded in this paper and cite their sources.
Inter-arm systolic blood pressure differences of ≥10 mmHg correlate with cardiovascular risk factors but do not typically indicate obstructive arterial disease, supporting the guideline recommendation to measure BP in both arms and use the higher reading.
It has long been known that even apparently normal human beings can show differences in both systolic and diastolic blood pressure readings between the two arms. As a result, those who determine clinical blood pressure are taught (or are supposed to be taught) to determine blood pressure in both arms and to note that the arm with the higher reading should be used for future determinations. Indeed, this is embodied in national and international guidelines on hypertension such as those of the European Society of Hypertension–European Society of Cardiology and the Joint National Committee on Prevention, Detection, Evaluation and Treatment of Hypertension (JNC-7) [1,2]. Those of us who have the opportunity to observe medical students, house officers and others in the clinical setting determining blood pressure appreciate that this procedure is not generally followed. Measurement of blood pressure of subjects in controlled clinical trials is usually performed by personnel who are carefully trained in the correct technique. There is no such control over regular office and clinical personnel. If the arm used to determine blood pressure is selected randomly or for convenience, rather than it being the one with the higher pressure, then the clinician could be misled into making treatment errors. In this issue of the journal, Kimura et al. [3] reinforce this concept by reporting a carefully controlled study of 1090 people living in the community of Ohasama, Japan. Ohasama is a small inland town in northern Japan. Its population has been the subject of other studies on hypertension. The authors specifically excluded those subjects who were known to have diseases associated with aortic obstructive pathology. By contrast to previous studies in which blood pressure determinations were made sequentially, they used an automated device that determines blood pressure simultaneously in all four extremities. Therefore, they were able to obtain simultaneous readings in both arms that were free from observer error. In addition, the device used measures pulse wave velocity and the ankle-brachial index. The findings they obtained are quite interesting. If the systolic and diastolic blood pressure reading in the left arm is simply subtracted from the simultaneous reading in the right arm for these 1090 subjects, the result is a bell-shaped normal distribution. If the total of all right arm readings is subtracted from the total of all left arm readings, the result is a small and clinically unimportant difference of 0.6 mmHg systolic lower in the right arm and 1.1 mmHg diastolic higher in the right arm. That does not address the important clinical issue. If the absolute pressure difference between the arms for each of the paired readings is considered (i.e. the difference between the two arms irrespective of which arm has the higher pressure), then the distribution is quite different and clinically meaningful. More than half of the subjects had less than a 5 mmHg difference and 90.9% had less than a 10 mmHg systolic difference. This provides a convenient cut-off point for identifying those individuals with a 10 mmHg or more difference in inter-arm systolic blood pressure as being ‘different’ from most other people. Is this difference important? Kimura et al. [3] attempted to determine this by very careful analysis of the data with regard to laboratory and clinical data. Not surprisingly, they found a correlation with hypertension, hypercholesterolemia, body mass index ≥ 25 kg/m2, HbA1c ≥ 5.8% and ankle-brachial index ≤ 1.0. On the other hand, neither they nor other investigators have been able to attribute differences of even more than 20 mmHg to obstructive disease [4]. Singer and Hollander [4] noted that 19% of their 610 emergency department patients had an inter-arm difference of more than 20 mmHg and yet were unable to find any evidence of substantial intrinsic arterial disease. However, they did appear to have more coronary artery disease. A practical concern is that a device for measuring blood pressure in both arms simultaneously is unlikely to be widely distributed to most clinics or other sites where clinical blood pressure is routinely determined. Nevertheless, blood pressure readings taken sequentially within a few minutes of each other are likely to be adequate for clinical purposes. The arm with the higher systolic reading (which may not be the one with the higher diastolic reading) should be identified and future determinations made from that arm. If the patient is younger and has predominantly diastolic hypertension, then the arm with the higher diastolic reading should be similarly identified and selected. What should the clinician do with the minority of people who have an inter-arm systolic pressure difference of ≥ 10 mmHg? Given the huge absolute number of people that will fall into this category, proposing invasive testing or costly non-invasive investigation as a first step is not warranted. The first step should be to ensure that the subject does not have a known cause of obstructive disease, such as syphilitic aortitis, obstructive vasculitis (Takayasu's Disease), evidence of extensive arteriosclerosis or, especially in young hypertensive patients, coarctation of the aorta. Clinical history and simple physical examination should establish or refute most of these entities. If this is negative and the patient is otherwise not symptomatic, it is probably best to do nothing other than to repeat the measurements periodically. If the index of suspicion for an obstructive lesion is high, non-invasive testing such as comparison of pulse waves from both arms might be of value as a next step. In summary, these inter-arm blood pressure differences are clinically important in terms of the diagnosis and treatment of hypertension. They do not appear to indicate obstructive lesions in most instances and routine invasive procedures to find such lesions are not warranted.
No takes yet. Share an insight, caveat, or question.
Barry J. Materson (2004) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: