Journal of Human Hypertension (2003) 17, 459462 & 2003 Nature Publishing Group All rights reserved 0950-9240/03

$25.00
www.nature.com/jhh

ORIGINAL ARTICLE

Both body and arm position significantly influence blood pressure measurement
RT Netea, JWM Lenders, P Smits and Th Thien
Division of General Internal Medicine, Department of Medicine, University Hospital Nijmegen, Nijmegen, The Netherlands

The position of both the body and the arm during indirect blood pressure (BP) measurement is often neglected. The aim of the present study was to test the influence of the position of the patient on BP readings: (1) sitting with the arms supported precisely at the right atrium level and (2) supine: (a) with the arms precisely at the right atrium level and (b) with the arms on the examination bed. In a first group of 57 hypertensive patients, two sessions of BP and heart rate (HR) measurements were performed in two positions: sitting and supine with the arms supported precisely at right atrium level in both positions. BP was measured simultaneously at both arms, with a Hawksley Random Zero sphygmomanometer at the right arm, and with an automated oscillometric device (Bosomat) at the left arm. BP and HR readings obtained in the two positions were then compared. In a second group of 25 normoand hypertensive persons, two sessions of BP and HR readings were performed in supine with the arms in two different arm positions: (a) the arm placed precisely at right atrium level and (b) the other arm on the examination bed. The measurements were performed at both arms with two automated devices (Bosomat). The read-

ings taken in the two positions were compared. Both systolic BP (SBP; by 9.5 7 9.0 (standard deviation, s.d.); right arm) and diastolic BP (DBP; by 4.8 7 6.0 mmHg; right arm) were significantly higher in the supine than in the sitting position. When the two different arm positions (body continously supine) were compared in the second part of the study, significantly higher SBP (by 4.6 7 6.1 mmHg) and DBP (by 3.9 7 2.8 mmHg) were obtained when the arm of the patient was placed on the bed (below the right atrium level), than when the arm was placed at the level of the right atrium. BP readings in sitting and supine positions are not the same. When according to guidelines the arm of the patient is meticulously placed at the right atrium level in both positions, the difference is even greater than when the arm rests on the desk or on the arm support of the chair. Moreover, in the supine position small but significant differences in BP are measured between arm on a 5 cm-high pillow and arm on the bed. In every study reporting BP values, the position of both the body and especially the arm should be precisely mentioned. Journal of Human Hypertension (2003) 17, 459462. doi:10.1038/sj.jhh.1001573

Keywords: blood pressure measurements; hypertension; body and arm position

Introduction
The indirect blood pressure (BP) measurement with a mercury or aneroid sphygmomanometer and a stethoscope by a trained observer is widely recommended as the cheapest and the most accurate manner of measuring the BP in the daily routine. The position of the patient during the measurement is often neglected. The reference point for the measurement of BP is the right atrium, the so-called heart level.1 The guidelines of the World Health Organisation/International Society of Hypertension (WHO/ISH) recommend that the BP should be

Correspondence: Dr Th Thien, Department of Medicine (541), University Medical Center St Radboud, PO Box 9101, 6500 HB Nijmegen, The Netherlands. E-mail: t.thien@aig.umcn.nl Received 28 September 2002; revised 18 January 2003; accepted 18 March 2003

routinely measured with the patient comfortably seated with the arms supported at heart level.2 To detect orthostatic hypotension, BP should also be measured with the patient first supine and subsequently in the standing position.2 It has been suggested that BP readings taken in sitting and supine positions can be considered equivalent if the patients arm is placed at heart level (right atrium) in both positions.3 As a practical approximation of the heart level when sitting or standing, it has been proposed to take the level of the fourth intercostal space2 or the level of the midsternum.3 We could find only scarce information about the approximation of the heart level in the supine position. Using computerised tomography, it has previously been shown that the level of the right atrium in the supine position is situated approximately half way between bed surface and sternum.4 A special pillow has been developed to be placed under the arm, in order to support it at the right atrium level in the supine

Influence of body and arm position on BP measurement RT Netea et al 460

position.4 This pillow is however not available in most hospitals. For practical reasons, positioning the arm of the patient on the bed has been recommended as acceptable during BP measurement in the supine position.5 The first aim of the present study was to test the influence of body posture on the indirectly measured BP values when the arm was placed at the right atrium level. The second aim of our study was to test the effect of the level of the arm on the indirectly measured BP values in the supine position: arm on the bed surface and arm at half distance between the sternum and the bed.

Patients and methods

Influence of body position on BP readings

A total of 57 hypertensive patients (29 males, mean age 55 7 12 (s.d.) years, BMI 25.7 7 4.4 (s.d.) kg/m2, arm circumference right 29.2 7 3.3 (s.d.) cm and left 28.9 7 3.3 (s.d.) cm) participated in this part of the study. Of these, 50 patients were on antihypertensive drug treatment. Two sessions of BP measurements, separated by a 10 min pause, were performed in two positions: (1) lying on a bed with the arms supported with the palms upwards and both the cubital fossa and the upper arm at halfway between the bed and the sternum4 and (2) sitting on a chair with the arms supported horizontally with the palms upwards and again both the cubital fossa and the upper arm at the level of the midsternum. The arm was placed at the approximated right atrium level using a levelling instrument. Each session (at random first sitting or supine) started with 5 min rest in the respective position. Then BP recordings were taken in triplicate 1 min apart, simultaneously at both arms, using a Hawksley random-zero sphygmomanometer (Hawksley and Sons, Lancing, UK; bladder size 36 13 cm) and a stethoscope at the right arm and an automatic oscillometric device (Bosomat; Boso oscillomat, Bosch, Jungingen, Germany; bladder size 28 12.5 cm)6 at the left arm. In between, a 20 s heart rate (HR) was counted at the right radial artery.
Influence of arm position on supine BP readings

neously at both arms, with the subject comfortably supine on an examination bed with one small cushion under the head. To test for BP differences between the two arms, one BP reading was first performed with the patients in the described supine position and both arms placed on the bed, with the palms upwards. One arm (at random right or left) was then supported with a specially designed pillow with the cubital fossa and upper arm at half-distance between the bed and the sternum (approximation of the right atrium level), whereas the other arm remained on the bed (Figure 1). After 5 min of rest, three BP and HR were measured, 1 min apart. The arm position was switched thereafter, with the arm first supported on the pillow now being placed on the bed and vice versa. Again after 5 min of rest, the three BP and HR readings were repeated. In both studies, all BP measurements were performed by the same trained observer (RTN), in a quiet room, with comfortable temperature. Every patient that agreed to participate in our study received detailed explanation before the procedure commenced and talking was avoided during the measurements.

Statistical analysis

The mean of three readings of BP and HR obtained, respectively, in each position was considered representative for that position. Paired t-test was used to compare the differences between the BP and HR, respectively, in different positions. The test was considered statistically significant when Po0.05 (two sided). The results are expressed as mean 7 s.d. unless otherwise stated.

In all, 25 subjects (16 hypertensive patients, 11 males, mean age 52.9 7 16. 5 (s.d.) years, BMI 26.1 7 4.9 (s.d.) kg/m2, arm circumference right 29.5 7 0.7 (s.d.) cm and left 29.3 7 1.2 (s.d.) cm) participated in the study. Pregnant women, patients with BP differences between the two arms larger than 10 mmHg and patients with arm circumferences larger than 35 cm were excluded. The BP measurements were performed using two automatic oscillometric devices (Bosomat; Boso oscillomat, Bosch, Jungingen, Germany; bladder size 28 12.5 cm). All BP readings were taken simultaJournal of Human Hypertension

Figure 1 Computerised tomography illustrating the difference between the arm levels used in the present study (right arm supported at the right atrium level and left arm supported on the examination bed). R: right arm; L: left arm; RA: right atrium.

Influence of body and arm position on BP measurement RT Netea et al

Table 1 Systolic blood pressure (SBP), diastolic blood pressures

(DBP) in mmHg, and heart rate (HR) in bpm, in the 57 patients in the sitting and supine positions. Results are given as means 7 s.d. Sitting Hawksley Random Zero SBP 135.5 7 24.0 DBP 79.2 7 9.7 HR 63.3 7 9.0 Automatic oscillometric SBP 135.7 7 24.8 DBP 79.5 7 9.7 HR 63.3 7 9.7 Supine Difference P-value

low than arm low/arm high (7.0 7 4.3 mmHg vs 2.3 7 6.7 mmHg, P 0.048), no significant effect of the sequence of the two positions was detected on the BP differences. With respect to the HR, after an initial small decrease in the first 5 min of rest it remained unchanged during the procedure.

461

145.0 7 23.2 84.1 7 9.0 62.2 7 9.0 141.3 7 25.5 84.6 7 10.5 62.5 7 9.7

9.5 7 9.0 4.8 7 6.0 1.1 7 3.8 5.6 7 8.3 5.1 7 5.3 0.8 7 3.8

o0.001 o0.001 0.037 o0.001 o0.001 0.128

Discussion
In our group of hypertensive patients, in an outpatient setting, we found significantly lower BP readings in the sitting than in the supine position when the patients arms were supported at the right atrium level as exactly as possible in both positions. This confirms previous observations in normotensive subjects.4,7 The BP measurement in sitting position is most often used in the general physicians office, whereas in some hospitals BP is often measured in supine position. Both diagnostic and therapeutic decisions are based on BP readings taken sometimes in sitting and sometimes in supine positions, since they are often regarded as equivalent. The differences we found using the Hawksley Random Zero sphygmomanometer are sustained by those found with the automatic instrument. The explanation of the slight differences in results between the two instruments could be the different measurement principle. One should realise that with the oscillometric method the mean arterial pressure is determined by the maximal amplitude of the oscillations, and SBP and DBP are subsequently computed from this value. With the mercury sphygmomanometer, however, both SBP and DBP are separately measured. Another factor that is often neglected during BP measurement is the position of the arm of the patient. We and others79 have shown that this induces a consistent difference between BP readings taken with the arm in different positions. In patients with a length of 190 cm, we measured differences as large as 25 cm between the level of the arm support of a common chair and the mid-sternum level. Such differences may result in large errors in BP readings that could obscure or even inverse the difference in

Results
Influence of body position on BP readings

The main results are given in Table 1. Both the systolic (SBP) and the diastolic (DBP) blood pressures were significantly higher in the supine position regardless of the instrument used (Po0.001 for both). The HR was slightly, but significantly, higher in the sitting than supine position (63.3 7 9.0 vs 62.2 7 9.0 bpm respectively; P 0.037).
Influence of arm position on supine BP readings

Table 2 presents the BP and HR readings taken in the supine position with the arm on the examination bed and with the arm supported at the right atrium level. As seen in Table 2, no difference was found in the initial BP readings between the two arms. However, significantly higher BP readings were recorded when the arm was placed on the bed than when the arm was supported at right atrium level. Statistical significance was not reached for the SBP in the right arm but the trend of the results was similar to that in the left arm. Irrespective of the arm position, the average BP in the first attended position (after the first 5 min of rest) was significantly lower than the initial BP (start BP). Except for a slightly higher SBP difference at the right arm when the sequence of positions was arm high/arm

Table 2 Supine systolic (SBP), diastolic (DBP) blood pressure in mmHg, and heart rate (HR) readings in bpm, in 25 subjects with the
arms supported at the heart level and with the arms on the bed; the sequence of the two positions was randomized. Results are given as mean 7 s.d. Starta (arm on the bed) Right arm SBP DBP HR Left arm SBP DBP HR
a

Arm high (at heart level)

Arm low (arm on the bed)

Difference

P-value

153.0 7 33.6 85.0 7 15.6 69.7 7 19.2 151.0 7 31.6 86.7 7 14.2 70.0 7 19.4

138.3 7 29.2 77.8 7 13.7 67.3 7 16.0 137.4 7 29.0 78.2 7 14.4 67.4 7 16.7

140.2 7 29.4 80.6 7 14.1 67.5 7 17.0 142.1 7 28.0 82.1 7 13.4 67.1 7 16.0

1.9 7 6.3 2.8 7 2.7 0.2 7 2.6 4.6 7 6.1 3.9 7 2.8 0.2 7 3.5

0.105 o0.0001 0.57 0.0009 o0.0001 0.638

The start readings represent the initial control readings to exclude leftright differences. Journal of Human Hypertension

Influence of body and arm position on BP measurement RT Netea et al 462

BP readings between the sitting and supine position. The effect of the arm position on BP readings in the supine position was especially observed during nocturnal BP registration in ambulatory BP measurement (ABPM) studies1013 and in pregnant women,1416 when the BP was measured in the lateral recumbent position. The BP values recorded when the arm was placed above the heart level were significantly lower than those recorded when the arm was situated below the heart level. When the patient is lying on the examination bed, one assumes that the arms of the patient rest at the level of the right atrium.5 However, Ljungvall et al4 demonstrate that even in this case the arms of the patient are situated clearly below the right atrium level. Our results in a group of patients with a wide range of BP values, show that even such a small variation in the arm position can result in significant BP differences. To obtain a double control of the results and to avoid observer bias, we used automated devices, and we performed simultaneous readings in both arms and for each arm in both sequences of the two positions. The results showed similar trends in both arms and were independent of the sequence of the two positions and of the BP level. The magnitude of differences we obtained was slightly smaller (ranging from 2.8 to 4.6 mmHg) than in the study of Ljungvall et al (5.5 mmHg). A possible explanation for the difference between these two studies could be the fact that oscillometric instruments were used in the present study, whereas standard mercury sphygmomanometers were used in the study of Ljungvall et al.4 However, the differences we obtained are closer to those theoretically expected, when one considers the effects of hydrostatic forces as a major determinant of the differences in BP between the two positions.17 Differences of 34 mmHg may be small as absolute value, but are still important in epidemiological terms. In conclusion, the present study shows that the assumption that blood pressure in sitting and supine position can be considered similar is incorrect even when the arm of the patient is placed at the correct right atrium level in both positions, as officially recommended. Supporting the arm of the patient on the arm support of a common chair in the sitting position could partially correct and potentially reverse the differences between sitting and supine positions. However, this introduces a new source of variation in BP measurement: the various distances between the heart level and the arm support of the chair, which can be as large as 25 cm in tall patients. In

consequence, such practice is incorrect and should be discouraged.

References
1 Guyton A. Textbook of Medical Physiology. WB Saunders: Philadelphia. 1986. 2 1993 Guidelines for the management of mild hypertension: memorandum from a WHO/ISH meeting. Hypertension 1993; 22: 392403. 3 Ramsay L et al. Guidelines for the management of hypertension: report of the third working party of the British Hypertension Society. J Hum Hypertens 1999; 13: 569592. 4 Ljungvall P, Thorvinger B, Thulin T. The influence of a heart level pillow on the result of blood pressure measurement. J Hum Hypertens 1989; 3: 471474. 5 OBrien E, OMalley K. ABC of hypertension. The patient. BMJ 1987. 6 Sloan PJ et al. Standardised methods for comparison of sphygmomanometers. J Hypertens 1984; 2: 547551. 7 Terent A, Breig-Asberg E. Epidemiological perspective of body position and arm level in blood pressure measurement. Blood Press 1994; 3: 156163. 8 Netea RT, Smits P, Lenders JWM, Thien T. Does it matter whether blood pressure measurements are taken with patients sitting or supine? J Hypertens 1998; 16: 263268. 9 Zachariah PK, Sheps GS, Moore AG. Office blood pressures in supine, sitting and standing positions: correlation with ambulatory blood pressures. Int J Cardiol 1990; 28: 353360. 10 Van der Steen MS, Pleijers AMLJ, Lenders JWM, Thien T. Influence of different supine body positions on blood pressure: consequences for night blood pressure/dipper-status. J Hypertens 2000; 18: 17311736. 11 Cavelaars M et al. Assessment of body position to quantify its effect on nocturnal blood pressure under ambulatory conditions. J Hypertens 2000; 18: 17371743. 12 Parati G. Blood pressure reduction at night: sleep and beyond. J Hypertens 2000; 18: 17251729. 13 Newton KM. Comparison of aortic and brachial cuff pressure in flat supine and lateral recumbent positions. Heart Lung 1981; 10: 821825. 14 Hovinga G, Aarnoudse JG, Huisjes HJ. The effect of supine and lateral position on intra-arterial pressure in hypertensive pregnancies. Am J Obstet Gynecol 1978; 131: 233238. 15 Villar J et al. The measuring of blood pressure during pregnancy. Am J Obstet Gynecol 1989; 161: 10191024. 16 Wichman K, Ryden G, Wichman G. The influence of different positions and Korotkoff sounds on the blood pressure measurements in pregnancy. Acta Obstet Gynaecol Scand 1984; 18: 2528. 17 Netea RT et al. Influence of the arm position on intraarterial blood pressure measurement. J Hum Hypertens 1998; 12: 157160.

Journal of Human Hypertension