Arterial pressure and arm circumference.

Abstract

Listen

Translate article

The relation between sphygmomanometer reading and arm circumference is of more than academic interest. It has obvious implications in the field of population studies in that arm circumference varies with sex, age, weight, and physical demands of occupation, while, at the clinical level, whether or not a man is accepted for life insurance or for an occupation such as crane or lorry driving may turn on an estimate of his arterial pressure made with a sphygmomanometer. Yet evidence about the relation ship remains inconclusive and confusing. Ragan and Bordley (1941) compared, for 51 subjects, the direct measurement of arterial pressure taken in one arm by the intra-arterial method of Hamilton with the indirect measurement of pressure taken simultaneously in the other arm by a standard mercury sphygmomanometer. They found that there was a tendency for indirect readings to be lower than intra-arterial readings for subjects with thin arms and higher for subjects with thick arms. But their data which did not include body weight, were derived from a small, mixed (male and female), and very unrepresentative sample. (Some of the subjects were suffering from hypertension, others had syphilitic aortitis, and their ages ranged from 19 to 60 years). Excluding thirteen subjects with aortic regurgitation or aortitis, Pickering, Roberts, and So wry (1954) used the same data to draw up a Table of Corrections for arm circumference to be applied to sphygmomanometer readings. They admitted that the corrections might be misleading when used on individual readings, but claimed that they were well worth making for the comparison of large groups and then proceeded to apply them to their own data. Others have since followed their example (Miall and Oldham, 1955). More recently, however, Holland and Humerfelt (1964) have failed to demonstrate any relationship between arm circumference and the difference between simultaneously taken direct and indirect measurements of arterial pressure, although they did find a positive correlation between arm circum ference and the direct intra-arterial readings. But their report is incomplete, in view of the fact that it do s not take body weight into account; for arm circumference is highly correlated with weight, and weight itself is known to be related to indirect readings of arterial pressure (Bee, Humerfelt, and W dervang, 1957; Society of Actuaries, 1959). In an analysis of data on physique, occupation, and indirect readings of arterial pressure in a population of more than 5,000 employed men between the ages of 15 and 69 years, Lowe (1964) has confirm d that, for a given age, sphygmomanometer re dings increase with body weight and that weight and arm circumference are highly correlated. When he corrected for weight, he found that the effect of arm circumference upon pressure reading was negligible and he concluded that the greater part, perhaps all, of the observed relationship between rm circumference and sphygmomanometer read ing was attributable to the close association between arm circumference and weight. His con clusion can be criticized, however, on the ground that it is based on the combined readings of twelve different observers, for he had already shown (Lowe and McKeown, 1962) that the observers differed widely in the level at which they recorded systolic and diastolic pressure measurements in comparable populations. In what follows some of the same data are subjected to a more rigorous analysis to meet this objection.