Comparison of blood pressure measurements on the bare arm, below a rolled-up sleeve, or over a sleeve

Ernesto Kahan, John Yaphe, Hadas Knaani-Levinz, Michael A. Weingarten

Research output: Contribution to journalArticlepeer-review

23 Scopus citations


Background. This study examined the effect of measuring blood pressure below subjects' rolled-up sleeves, over the sleeve, or on the bare arm. This is an important day-to-day issue for the busy GP. Methods. The sample consisted of 201 subjects in family practice clinics and residents of a senior citizens' home. A digital device was used in all cases. Each participant underwent three blood pressure measurements in each of the following conditions in random order: cuff on bare arm; cuff over the sleeve; and cuff below the rolled-up sleeve. Differences between measurements were plotted against the mean blood pressure. Confounding factors controlled for were age, sex, clothing pressure and skin-fold thickness. Results. Differences in mean blood pressure readings between the clothed and bare arm were 0.5 mmHg (SD 7.5) for systolic pressure and 1 mmHg (SD 5) for diastolic pressure; neither difference was significant. However, in hypertensive subjects (>140 mmHg systolic), although the mean difference remained small (systolic pressure, 2 mmHg, SD 10), the range of difference for individual subjects was -32 mmHg to +22 mmHg. Conclusion. The degree of clothing under the sphygmomanometer cuff does not have a clinically important effect on the blood pressure measurement. In patients known or found to be hypertensive, measurement on the bare arm is recommended.

Original languageEnglish
Pages (from-to)730-732
Number of pages3
JournalFamily Practice
Issue number6
StatePublished - Dec 2003
Externally publishedYes


  • Blood pressure
  • Blood pressure cuff
  • Family medicine
  • Measurements
  • Screening


Dive into the research topics of 'Comparison of blood pressure measurements on the bare arm, below a rolled-up sleeve, or over a sleeve'. Together they form a unique fingerprint.

Cite this