q 1999, British Geriatrics Society
Age and Ageing 1999; 28: 537–542
The normal range for inter-arm
differences in blood pressure
SUSAN ORME, SUSAN G. RALPH1, ANDREW BIRCHALL, PETER LAWSON-MATTHEW2, KATHERINE MCLEAN3,
KEVIN S. CHANNER
Department of Cardiology, Royal Hallamshire Hospital, Glossop Road, Sheffield S10 2JF, UK
1
Department of Genitourinary Medicine, The General Infirmary at Leeds, Leeds, UK
2
Department of Health Care for Elderly People, Northern General Hospital, Sheffield, UK
3
Department of Medicine for the Elderly, Derbyshire Royal Infirmary, Derby, UK
Address correspondence to K. S. Channer. Fax: (+44) 114 271 2042
Abstract
Objective: to establish the mean and normal range for the difference in simultaneous systolic and diastolic blood
pressure measurements between the right and left arm.
Subjects: 462 subjects, age range 20–89 years, in sinus rhythm and with no history of stroke, 98 of whom had a
history of cardiovascular disease or were taking vasoactive medication.
Methods: four simultaneous recordings of blood pressure in both arms were made using two automated
sphygmomanometers with the subject supine after resting for 10 min.
Results: inter-arm systolic and diastolic blood pressure differences show a near normal distribution of values. Some
individuals had clinically important differences in systolic and diastolic blood pressure between their arms. The
magnitude of these differences was not related to the mean baseline blood pressure. Linear regression analysis did
not demonstrate any significant relationship between inter-arm systolic or diastolic blood pressure difference and
age in patients of either sex. For systolic blood pressure the mean difference between the right and left arm was 1.1
mmHg and the normal range was ¹9 mmHg to 11 mmHg. For diastolic blood pressure the mean difference was 0
mmHg and the normal range ¹10 mmHg to 10 mmHg.
Conclusion: the frequency of significant inter-arm systolic and diastolic blood pressure differences suggests that
the blood pressure should be taken in both arms at the initial consultation. At subsequent visits, the arm in which
measurements are taken should be recorded in the case notes. The higher of the two readings should be used to
guide further management decisions.
Keywords: aortic dissection, blood pressure, blood pressure measurement, hypertension
Introduction
The diagnosis of hypertension requires that several
measurements of blood pressure are made [1]. If there
is a wide normal range of systolic and diastolic blood
pressure difference between the arms, then recordings
made in different arms on different occasions may
influence whether the diagnosis of hypertension is
made. In addition, a difference in systolic blood
pressure between the arms is an important clinical
sign in the diagnosis of acute aortic dissection.
However, there are patients who have clinically
important differences in indirectly recorded blood
pressure between their arms in the absence of aortic
dissection. For example, indirect measurements in
patients with a hemiparetic stroke often show higher
blood pressure in the hypertonic paretic arm and
lower blood pressure recordings in a flaccid paretic
arm [2, 3].
An earlier study indicated that some individuals
have clinically important inter-arm blood pressure
differences, but showed no significant relationship to
age [4]. This was also the conclusion of another group,
who studied the inter-arm blood pressure difference
of patients attending an emergency department and
found no relationship with gender or age. They
demonstrated that 53% of subjects presenting acutely
to an emergency department had an inter-arm blood
pressure difference of >10 mmHg [5].
We aimed to determine the influence of mean
blood pressure, age and gender on inter-arm systolic
and diastolic blood pressure differences. Secondly,
537
S. Orme et al.
we aimed to establish the mean and normal range
for inter-arm systolic and diastolic blood pressure
differences.
automated blood pressure monitors to eliminate
observer bias and digit preference associated with
the manual mercury sphygmomanometer. The Dinamap 8100 measures the diastolic pressure at Korotkov
phase V. Since automated sphygmomanometers have
limited accuracy in atrial fibrillation, subjects in atrial
fibrillation were excluded [7]. We used blood pressure
cuffs appropriate to the size of the subject. One
observer triggered both automatic sphygmomanometers simultaneously to record synchronous measurements of the blood pressure in the right and left arm.
Measurements were repeated and then the cuffs were
immediately interchanged and two further simultaneous
recordings of blood pressure made.
The data were entered into a computer spreadsheet. We determined the mean of the four readings
obtained from each arm and calculated the difference
in systolic blood pressure between the right and left
arm for all subjects by subtracting the mean value in
the left arm from the mean value in the right arm. We
calculated summary statistics for each patient group.
Regression analysis was used to determine the relationship of inter-arm blood pressure differences and mean
baseline blood pressure. Regression analysis was also
used to determine the correlation of inter-arm blood
pressure differences with age for each gender and
patient group.
Methods
We studied 462 subjects in sinus rhythm. They were
ward visitors to the Royal Hallamshire Hospital,
Sheffield, patients attending for non-cardiovascular
day-case surgery and volunteers from the workforce
of the Manpower Services Commission, based in
Sheffield.
Ninety-eight subjects (48 women and 50 men), age
range 29–89 years, mean age 68.4, were taking
vasoactive medication or had a history of cardiovascular disease (hypertension, ischaemic heart disease or
peripheral vascular disease). Subjects with a history of
stroke were excluded because of the difficulties in
measuring indirect blood pressure. The remaining
364 subjects (155 women and 209 men), age range of
20–89 years, mean age 49.1, had no history of
cardiovascular disease and were not taking medication
that affected the blood pressure.
Each subject rested in the supine position for
10 min before the blood pressure measurement. Blood
pressure recordings were made using two Critikon
Dinamap 8100 monitors that had been calibrated in
accordance with the manufacturer’s instructions and
the supplied calibration equipment. This model has
previously been validated in accordance with the
British Hypertension Society protocol [6]. We used
Results
Some individuals with and without a history of
Table 1. Mean values, with ranges and standard deviations (SDs), for blood pressure measurements in the left and right arms of
subjects with and without cardiovascular disease
Mean value (and 95% confidence interval), mmHg
..............................................................................................................................................................................................
Systolic pressure
Diastolic pressure
.................................................................................................
Population
Right arm
Left arm
a
Difference
...................................................................................
Right arm
Left arm
Differencea
........................................................................................................................................................................................................................
Without history of cardiovascular disease (n = 364)
Mean
134 (132–136)
132 (130–134)
Range
84 to 215
77 to 215
SD
18.7
18.9
1.1 (0.6 –1.6)
74 (73–75)
74 (73–75)
43 to 138
46 to 135
4.6
12.6
12.3
5.0
0.9 (–0.1–1.9)
75 (72–78)
74 (71–77)
0.4 (¹0.6–1.4)
47 to 125
47 to 129
6.1
15
15.6
5.6
1.1 (0.6–1.6)
74 (73–75)
74 (73–75)
0 (¹0.5–0.5)
44 to 141
46 to 135
13.2
13.1
¹15 to 19
0 (¹0.5–0.5)
¹20 to 17
With history of cardiovascular disease (n = 98)
Mean
Range
SD
138 (133–143)
7 to 212
25.6
137 (132–142)
79 to 205
25.7
¹16 to 28
¹15 to 20
Total study population (n = 462)
Mean
134 (132–136)
133 (131–135)
Range
78 to 215
77 to 215
SD
20.4
20.6
a
Right–left.
538
¹16 to 28
4.9
¹20 to 20
5.1
Inter-arm differences in blood pressure
cardiovascular disease have clinically important
differences in systolic and diastolic blood pressure
between their right and left arms. The range of systolic
and diastolic blood pressures in the group without a
history of cardiovascular disease indicates that there
were some subjects in this population who had
previously undiagnosed hypertension. The results are
summarized in Table 1.
Figure 1. Inter-arm difference in a systolic and b diastolic blood pressure (mmHg) versus frequency in subjects without
cardiovascular disease (n = 364).
539
S. Orme et al.
We performed a linear regression analysis on the
right arm–left arm difference against the mean [(right +
left blood pressure)/2] blood pressure. This showed no
statistically significant relationship between the mean
blood pressure and the magnitude of the inter-arm
blood pressure difference [systolic, r = 0.06 (¹0.03 to
0.01), P = 0.2; diastolic, r = 0.01 (¹0.03 to 0.04),
P = 0.8].
Figure 1 shows the frequency distribution of the
systolic and diastolic blood pressure difference in
patients without a history of cardiovascular disease,
while Figure 2 shows the frequency distribution in the
Figure 2. Inter-arm difference in a systolic and b diastolic blood pressure (mmHg) versus frequency in all subjects (n = 462).
540
Inter-arm differences in blood pressure
Table 2. Regression analysis of the relationship between inter-arm blood pressure
difference and age, gender and history of cardiovascular disease
Inter-arm blood pressure difference
.......................................................................................................................
Systolic
Population
Diastolic
................................................
.....................................................
r (and 95% CI)
r (and 95% CI)
P
P
........................................................................................................................................................
Without history of cardiovascular disease
Women
0.01 (¹0.03 to 0.03)
0.9
0 (¹0.04 to 0.03)
0.9
Men
0.08 (¹0.06 to 0.01)
0.2
0.03 (¹0.02 to 0.04)
0.7
All
0.03 (¹0.03 to 0.01)
0.5
0.02 (¹0.02 to 0.03)
0.7
With history of cardiovascular disease
Women
0.16 (¹0.3 to 0.1)
0.3
0.04 (¹0.2 to 0.2)
0.8
Men
0.16 (¹0.2 to 0.02)
0.1
0.3 (¹0.2 to 0.01)
0.8
All
0.1 (¹0.2 to 0.03)
0.1
0.05 (¹0.13 to 0.07)
0.6
CI, confidence interval.
study population as a whole. All the histograms
approach a normal distribution.
We performed linear regression analysis to determine the influence of age and sex on inter-arm systolic
and diastolic blood pressure difference (Table 2). There
was no statistically significant relationship between
inter-arm difference in either systolic or diastolic blood
pressure with age in subjects of either gender regardless
of whether they have a history of cardiovascular disease.
As we have established that there is no statistically
important relationship between inter-arm blood pressure differences with age and sex, we can define a
normal range for inter-arm blood pressure differences
for the study population. The normal range in which
95% of the population would be expected to lie is
defined as the mean 6 2 standard deviations. The
calculated normal ranges are shown in Table 3, which
translates these into differences which could be
measured in clinical practice.
Conclusion
We have demonstrated that some subjects of all ages
have clinically important differences in indirect systolic
and diastolic blood pressure measurements between
their arms. The wide range of inter-arm diastolic and
systolic blood pressure differences demonstrated leads
us to agree with the recommendations of previous
groups that at the initial consultation blood pressure
should be taken in both arms. At subsequent visits the
arm in which the blood pressure was taken should be
recorded. The higher of the two values should be used
to inform clinical decisions about the treatment of
hypertension [4, 5].
The normal range for systolic and diastolic blood
pressure differences we have defined should be of
use in clinical practice. As we have discovered that
the group of subjects without any previous history
of cardiovascular disease included individuals with
previously undiagnosed hypertension, these subjects
may be amongst those who present for diagnosis and
management of their hypertension.
To improve the validity of our results, we calculated
the mean of four blood pressure measurements in both
arms. In clinical practice, when fewer readings of
blood pressure are taken on a single occasion, the
range of inter-arm blood pressure may be greater than
we have demonstrated. Nevertheless, the size of the
normal ranges which we have established indicate
that inter-arm blood pressure differences can be an
important clinical problem. The normal range for the
difference in systolic blood pressure between the right
and left arm could be used as a working reference range
to aid in the clinical diagnosis of possible aortic
Table 3. The normal ranges of inter-arm blood pressure
difference for those without a history of cardiovascular
disease and for the whole study population, according to
experimental results and translated for use in clinical practice
Normal range of inter-arm difference, mmHg
....................................................................
Experimental
In clinical
practice
........................................................................................................
Systolic
Without history
¹8.1 to 10.3 (¹8.6 to 10.8)a
¹8 to 11
All
¹8.7 to 10.9 (¹9.2 to 11.4)
¹9 to 11
Diastolic
Without history
¹10 to 10 (¹10.5 to 10.5)
¹10 to 10
All
¹10.2 to 10.2 (¹10.7 to 10.7)
¹10 to 10
a
95% confidence limits.
541
S. Orme et al.
dissection. However, in order to use the normal ranges
defined above for this purpose four measurements of
blood pressure would need to be taken in each arm.
References
Key points
2. Yagi S, Ichikawa S, Sakamaki T et al. Blood pressure in the paretic
arms of patients with stroke. N Engl J Med 1986; 315: 836.
• Some subjects of all ages have clinically important
differences in systolic and diastolic blood pressure
between their arms.
• Age has no influence on inter-arm systolic or
diastolic blood pressure differences in patients of
either gender and there is no correlation between
mean baseline blood pressure and inter-arm blood
pressure difference.
• For systolic blood pressure, the normal range for
clinically measurable difference between the right
and left arm is ¹9 mmHg to 11 mmHg.
• For diastolic blood pressure, the normal range for
clinically measurable difference between the right
and left arm is ¹10 mmHg to 10 mmHg.
1. Anon. The sixth report of the Joint National Committee on
prevention, detection, evaluation and treatment of high blood
pressure. Arch Intern Med 1997; 157: 2413–46.
3. Dewar R, Sykes D, Mulkerrin E et al. The effect of hemiplegia on
blood pressure measurement in the elderly. Postgraduate Med J 1992;
68: 888–91.
4. Fotherby MD, Panayiotou B, Potter JF. Age-related differences in
simultaneous interarm blood pressure measurements. Postgraduate
Med J 1993; 69: 194–6.
5. Singer AJ, Hollander JE. Blood pressure, assessment of interarm
differences. Arch Intern Med 1996; 156: 2005–8.
6. O’Brien E, Mee F, Atkins N et al. Short report: accuracy of
the Dinamap portable monitor, model 8100 determined by the
British Hypertension Society protocol. J Hypertens 1993; 11:
761–3.
7. Stewart MJ, Gough K, Padfield PL. The accuracy of automated
blood pressure measuring devices in patients with controlled atrial
fibrillation. J Hypertens 1995; 13: 297–300.
Acknowledgements
We would like to thank the Sheffield Statistical Unit for
advice on statistical methods.
542
Received 19 January 1998; accepted in revised form 3 February
1999