- Email: [email protected]
Arm and leg blood pressures — are they really so different in newborns?
Early Human Development, 26 (1991) 203-211 Elsevier Scientific Publishers Ireland Ltd.
203
EHD 01177
Arm and leg blood pressures - are they really so different in newborns? Frances
Cowana,
Marianne
Thoresen”
and Lars Wall@eb
“Department of Paediatrics. Ullevaal Hospital, Oslo and hDeparment of Physiology. University of Oslo (Norway) (Received
6 September
1990; revision
received
28 August
1991; accepted
2 September
1991)
Summary Controversy still exists over differences between upper and lower limb blood pressures (BP) in neonates. We measured upper arm and calf systolic (S) and diastolic (D) BP and heart rate (HR) using 2 Dinamap 847 instruments simultaneously every half minute for several measurement periods of 5-10 min. Nine healthy term newborns were studied in active (AS) and quiet (QS) sleep on post-natal days 1 and 5. The results were examined using unbalanced analyses of variance. Arm SBP was 62.3 f 1.6 mmHg and DBP 35.5 + 1.0 mmHg on day 1 in AS and QS. Calf values were not significantly different but were slightly higher (by 2-3 mmHg) in AS. Arm SBP and DBP rose by 8.5 and 5 mmHg, respectively, between days 1 and 5 but calf pressures rose less. Calf SBP and DBP were significantly lower (by 4.6 and 3.4 mmHg, respectively) than the arm values in QS on day 5. Arm SBP and DBP were dependent on post-natal age but not on sleep state while calf SBP and DBP and HR were dependent on both. Mean HR rose with age from 1!4 to 117.6 bpm in QS and from 118.6 to 122.4 bpm in AS. Our non-invasive BP measurements were similar to available invasive data. We postulate that differences in arm and calf vasoreactivity account for the different dependence on sleep state and for the unequal changes in arm and calf BE from days 1 to 5.
blood
pressure;
oscillometric
Correspondence to: Dr. Frances Hospital. DuCane Rd.. London
method;
arm; calf; sleep; newborn
Cowan, Dept. of Paediatrics WI2 OHS. U.K.
0378-3782/91/$3.50 0 1991 Elsevier Scientific Published and Printed in Ireland
Publishers
and Neonatal
Ireland
Ltd.
Medicine,
Hammersmith
204
Introduction A rise in upper limb systolic BP in the first week of life has been well documented [3,13,17,19,25]. However, there is considerable variation in reported absolute values for upper limb BP [3,10,11,13,17,19,24,25], little information on upper limb diastolic BP [ 19,241 and little agreement on lower limb BP [ 10, I 1,13,19,24]. Reported differences between arm BP and leg BP vary from -17 mmHg to +20 mmHg [ 10,13,19,24]. Scant attention has been given to the influence of awake/sleep state during BP measurements [3]. We have tried to overcome some of the problems of other studies by simultaneously measuring upper arm BP and calf BP and HR in a quiet, undisturbed environment on the first and fifth day of life in healthy term infants in active (AS) and quiet sleep
(Qs). Subjects and Methods The study group comprised 9 healthy term newborn infants (mean weight 3501 g) vaginally delivered after uncomplicated pregnancies without any maternal medication. Systolic (SBP) and diastolic (DBP) upper arm BP and calf BP were measured every half minute using an oscillometric method (two Dinamap 847 instruments, Critikon, Tampa, FL) which has been shown to give values corresponding well to invasively determined arterial pressures in the neonate [7,12]. The largest neonatal cuff size possible was used. The infants lay prone and the arm was positioned away from the chest to prevent chest wall movement impinging on the cuff and influencing the oscillometric detection device. The elbow was straightened at least to 90”. The ECG from three chest electrodes was stored on computer for later analysis. Respiratory movements were recorded from a pressure sensor (Graseby Medical, Watford, U.K.) taped on the upper abdomen to help with defining sleep state. Sleep state was determined according to Prechtl [21] using body and eye movements, breathing pattern and HR variability. For each baby, arm and calf BP measurements were obtained during a median of 4 (range 2-7) 5-min periods in AS and QS sleep on the first and fifth days of life. The median value of the 10 BP measurements obtained during each 5-min period was used for the analysis. Recordings were made once the baby had been in a steady state for 5 min and those associated with body movement or change in sleep state were not used. We have previously determined that the 95% CI for one BP reading is f 7.2 mmHg while the 95% CI for 10 readings is only ~2.3 mmHg [26]. A variable number of 5-min measurement periods on each day and in each sleep state was obtained per child. To examine the significance of our observed results unbalanced analyses of variance (BMDP 3V) [2] were performed with the infant as a random variable and with day and sleep state and different interaction terms as possible fixed effect variables. The model gave a good tit to observed data and made it possible to estimate the relative influence of day and sleep state on upper arm and calf BP and HR.
205
Results The observed values (and S.D.) for HR and arm and calf BP in AS and QS on day 1 and 5 are summarised in Table 1. The differences between sleep states for each day and the increase between days for each sleep state are also given. There was a large individual variation in absolute HR values. HR increased from day 1 to 5 in both AS and QS and was higher in AS than in QS on both days. The difference between sleep states was greater on day 1 than day 5 and the increase from day 1 to 5 was greater during QS than AS. BP was much less variable between individuals than HR. Between days I and 5 observed arm SBP rose 9.5 and 8.1 mmHg while arm DBP rose only 5.5 and 5 mmHg in AS and QS, respectively. In the calf the increase in both SBP and DBP between days 1 and 5 was less than in the arm. There was little difference in arm SBP or DBP between sleep states on either day but leg SBP and DBP differed more especially on day 1. Since the number of 5-min recording periods per baby was different and there were individual differences in resting HR and BP the increases observed could be
TABLE
I
Observed heart rate and blood pressure results with SD. states and the rise from day AS. active
sleep; QS,
I
showing showing the differences between sleep
to day 5.
quiet sleep, S.D.,
standard
deviation.
Number
of observations: day I AS. II = IX:
day I Q’S, n = 23; day 5 AS. n = 40; day 5 QS n = 19.
AS
Heart
rate
tbpm)
(mmHg)
(mmHg)
(SD.)
diff
118.7
f
9.2
112.2
f
II.6
-6.5
day 5
122.5
f
8.3
118.2
f
13.9
-4.3
3.8
6.0
day I
62. I
f
2.8
62.4
f
3.3
-0.3
day 5
71.6
f
7.0
70.5
zt 6.9
-1.1
rise Calf SBP
QS
day I rise
Arm SBP
(S.D.)
9.5
8.1
day I
66. I
??
5.3
61.8
f
4.5
-4.3
day 5
68.9
f
6.3
67.8
zt 6.5
-1.1
rise
2.8
6.0
Arm DBP
day I
35.8
f
I.7
35.3
+z 2.6
-0.5
tmmHg)
day 5
41.3
f
4.8
40.3
?? 4.5
-1.0
rise
5.5
5.0
Calf DBP
day I
38.0
f
3.5
35.2
f
4.2
-2.x
0mnHg)
day 5
39.0
*
3.7
37.8
f
4.6
-1.2
rise
I .o
2.6
206
due to a skew distribution of the number of measurement periods between the individuals. To examine the significance of our observed results unbalanced analyses of variance were performed (see Subjects and Methods). In the first analysis both sleep stage, day and an interactive term between sleep state and day were included as possible fixed effects in the model. The result given in Table II, panel A showed
TABLE
II
Results of the unbalanced pressure.
analyses
of variance
for heart
rate and arm and calf systolic
SW’. systolic blood pressure; DBP, diastolic blood pressure: err, var., error variance; Variable
Parameter
Estimate
A. Heart rate
err. var. constant
47.18 118.14
7.00 2.85
day sleep day and sleep infant
-1.79 -2.32 0.34 68.03
0.77 0.76 0.77 34.26
err. var. constant day sleep
47.34 118.20 -1.82 -2.38
7.02 2.83 0.76 0.75
infant
67.20
33.88
err. var. constant
Il.65 66.59
I .73 1.51
day sleep infant
-4.27 -0.01 19.15
0.38 0.37 9.57
err. var. constant
2.70
day sleep
18.23 65.72 -I .98 -1.47
0.47 0.46
infant
16.81
8.77
err. var. constant
6.51 38.20
0.96 0.94
day sleep infant
-2.64 -0.03 7.18
0.28 0.28 3.69
err. var. constant day sleep
II.05 37.42 -0.84 -0.90
1.64 0.82 0.37 0.36
infant
4.83
2.78
B. Heart
rate
C. Arm SBP
D. Calf SBP
E. Arm DBP
F. Calf DBP
SD.
I .44
prob.,
and diastolic
probability,
Two-tail
prob.
0.000 0.020 0.002 0.657
0.000 0.017 0.001
0.000 0.000 0.982
0.000 0.000 0.002
0.000 0.000 0.903
0.000 0.022 0.012
207
that the interaction between sleep and day did not have significant effect at the 5% level on HR. Similarly there was no interactive effect between sleep and day on any of the BP results (not given in the Table). When the analysis was run without this interactive term, we obtained the results shown in Table II, panels B to F and Figs. 1 and 2. Day and sleep had significant effects on HR (Table II, panel B, and Fig. 1). Arm SBP and DBP were significantly affected by day but not by sleep (Table II, panels C and E). However calf SBP and DBP were significantly affected both by day and sleep (panels D and F). The BP results are summarised in Fig. 2. Table III shows a comparison of the observed and predicted results (with S.D.) from the best fitting models.
w 2
0 A active 0
70
68
arm
sleep P,/I’
a quiet sleep
I
1
leg
0 1
122
120 i
arm
c 118e
/’
c % 116I
4’ I’ /’ __-I_--
114-
112 p
Fig.
leg __--
_A i
/
I
1
1
2
I
3
Day after I. The rise in heart 0 ) as estimated
( 0 __
/’
,0’
I
4
I
5
I
I
I
I
I
1
2
3
4
5
Day after
birth
rate from day from the model.
I to day 5 in active sleep ( 0 -
Fig. 2. The upper half shows the rise in arm and leg systolic
pressure
birth 0 ) and quiet sleep
from day I to day 5 in active and
quiet sleep. The lower half shows the values obtained for diastolic pressure. The values plotted estimated from the model. Note the breaks in the blood pressure scale on the .r axis.
are those
(0) heart rate and blood
111 pressure
Blood Arm Calf Arm Calf
pressure SBP SBP DBP DBP
Heart rate (bpm)
62.3 62.3 35.5 35.7
114.0
112.2
(mmHg) 62.4 61.8 35.3 35.2
P
62.1 66. I 35.8 38.0
118.7
?? 3.l
+ 1.6 f 1.6 f I.0 ??I.0
0
SD.
AS
QS
0
day I n = I8
pressure;
day I n = 23
blood
results compared
AS, active sleep; QS, quiet sleep; SBP, systolic
Observed
TABLE
62.3 65.2 35.6 37.5
118.8
P
DBP. diastolic
zt zt * f
I.6 I.6 1.0 I.0
70.5 67.8 40.3 37.8
118.2
3.l ??
70.8 66.2 40.8 37.4
117.6
f zt f f
1.6 I.6 I.0 I.0
*3.l
71.6 68.9 41.3 39.0
122.5
0
AS
day 5
of the predicted
QS SD.
deviation
n = 40
P
S.D.. standard
day 5 n = I9
pressure;
0
blood
(P) from the model.
S.D.
to those predicted
70.9 69.2 40.9 39.2
122.4
P
results
??I.5 f I.5 ??I.0 zto.9
jz2.9
SD.
209
Discussion Many studies on HR and BP in newborn have been performed and the values given differ considerably. We feel that this is largely due to varying awake/sleep states and the use of methods that disturb the infant. We have measured BP in relation to post-natal age and sleep state in a situation where disturbance to the infant was minima1 and have discarded measurement periods associated with any movement or change of state. The effect of ductal closure on BP has not been well documented but most of our day 1 measurements were made after 12 h of age when for many full term infants the ductus arteriosus is functionally closed. We used the median of 10 consecutive measurements made at half-minute intervals as an estimate of each measurement period. We have previously found [26] the S.D. around 1, 3 and 10 measurements to be *7.2, +4.1 and h2.3 mmHg respectively. We took great care over the choice of cuff size and position as the use of inappropriately small cuffs overestimates the true BP [ 12,151. In addition, because it was not possible to obtain exactly the same number of measurement periods in each state for each baby on both post-natal days the significance of our observed results was examined using unbalanced analyses of variance. Our values for arm SBP and DPB on day 1 are almost identical to those of Versmold [27] obtained using invasive methods and slightly lower than those of Kitterman [l4] and Moss [16]. There is no invasive data for day 5. Values reported for newborn term babies using the oscillometric or Doppler methods are often higher condithan ours [ 17,24,25]. We feel this is most likely due to less stable measurement tions than ours in awake rather than asleep infants. Calf BP is not often measured invasively in newborn infants. However, Butt and Whyte [l] found the same good correlation between posterior tibia1 and dorsalis pedis and umbilical artery measurements as between the radial and the umbilical artery. In older children Park and Guntheroth [18] found simultaneous invasive measurements of BP in the brachial and femoral artery to be the same. Using noninvasive methods the calf BP is reported to vary from 20 mmHg higher to 17 mmHg lower than the arm BP [ 10,13,19,24]. We could not confirm these reported differences. We found calf and arm BP to be the same on day 1 as did Park in a recent study [19] but calf BP was slightly lower on day 5. This difference was statistically significant in QS (see Table III, predicted results). The rise in BP from day 1 to day 5 is well documented and is due to increased cardiac output in response to the large rise in metabolic demands in the perinatal period [23] and rise in peripheral resistance and possibly to ductal closure. Calf vascular resistance increases by 20-30% in the first week of life [ 11,281. We have not been able to find equivalent data for arm vascular resistance. We found that calf BP rose between days 1 and 5 by about half the rise in the arm. This suggests a difference in the vasoreactivity of the upper and lower limb major vessels. In adults the resistance vessels of the calf and forearm have been shown to behave differently in many physiological situations, calf vasculature being less reactive than that of the forearm [4,5,22]. Similar work has not been published in infants. The lower limbs of newborns are underdeveloped and have had a poorer, lower oxygen supply than the upper limbs during fetal life. In the light of adult studies it could well be that the
210
forearm responses are more marked accounting for the greater increase in arm BP compared to the calf. Arm BP was not dependent on sleep state but the calf BP was. Similar adult values for comparison are not known. However, this finding again suggests a difference in nervous regulating mechanism between the upper and lower limb of a more central origin [22]. We did not find any interactive effect of day and sleep state. HR is known to increase during the first week of life and our findings confirm this. Our values are similar to those of Finlay [6] and Picton-Warlow [20] though lower than many reported [8,9] reflecting the sleeping state and stable conditions of the infants. HR was dependent on sleep state being IO”% lower during QS than AS. In summary, we found little difference between our Dinamap BP measurements and available invasive data. Arm and calf measurements were very similar in this group of healthy term babies. Arm and calf BP rose differentially between days 1 and 5 and were not equally dependent on sleep state. The most likely explanation for this is that the newborn infant can preferentially increase peripheral resistance in the upper extremities as does the mature adult. Acknowledgements F.C. and M.T. were supported by the Norwegian Research Council for Science and the Humanities. We thank the staff of the post-natal ward (Barsel 2, Ullevaal Hospital) and the mothers for their cooperation with the study. References I
Butt, W.W. and Whyte. H. (1984): Blood pressure monitoring in neonates: and peripheral artery catheter measurements. J. Pediatr., 105, 630-632.
2 3
Dixon, W.J. (Ed.) (1981): BMDP Statistical Software. University of California Press, Berkeley. Earley, A., Fayers, P., Ng, S., Shinebourne. E.A. and de Sweit. M. (1980): Blood pressure in the first 6 weeks of life. Arch. Dis. Child.. 55, 755-757. Essandoh. L.K., Houston. D.S.. Vanhoutte, P.M. and Shepherd. J.T. (1986): Differential effects of lower body negative pressure on forearm and calf blood flow. J. Appl. Physiol.. 61, 994-998. Essandoh, L.K., Duprez, D.A. and Shepherd, J.T. (1987): Postural cardiovascular reflexes: comparison of responses of forearm and calf resistance vessels. J. Appl. Physiol., 63. 1801-1805. Finlay. J.P.. Hamilton, R. and MacKenzie. M.G. (1984): Heart rate response to tilting in newborns
4 5 6
Comparison
of umbilical
in quiet and active sleep. Biol. Neonate, 45. l-10. Friesen, R.H. and Lichtor. J.L. (1981): Indirect measurement of blood pressure in neonates and infants utilizing an automatic noninvasive oscillometric monitor. Anesth. Analg.. 60, 742-745. Haddad. G.G.. Epstein. R.A.. Epstein, M.A.F.. Leistner. H.L. and Mellins. R.B (1980): The R-R interval and R-R variability in normal infants during sleep. Pediatr. Res., 14, 809-81 I. Harper, R.M.. Hoppenbrouwers. T.. Sterman, M.B., McGinty. D.J. and Hodgman. J. (1976): Polygraphic studies of normal infants during the first six months of life. 1. Heart rate and variability as a function of state. Pediatr. Res.. 10. 945-951. Ho, T.F.. Yip. W.C.L., Tay. J.S.H. and Wong. H.B. (1985): Ultrasonic systolic blood pressure gradient between upper and lower limbs in normal term Chinese neonates. Clin. Pediatr., 24. 146-148. Kidd. L.. Levison. H., Gemmel. P., Aharon. A. and Swyer. P.R. (1966): Limb blood flow in normal and sick newborn. Am. J. Dis. Child., 112. 402-407. Kimble, K.J.. Darnall, R.A., Yelderman, M.. Ariagno. R.. and Ream. A.K. (1981): An automated oscillometric technique for estimating mean arterial pressure in critically ill newborns. Anesthesiology. 54, 423-425.
211
13 14 I5 16 17 I8 19 20 21 22
23 24 25 26 27
28
Kirkland,
R.T. and Kirkland,
J.L. (1972): Systolic blood pressure
measurement
in the newborn
in-
fant with the transcutaneous Doppler technique. J. Pediatr., 80. 52-56. Kitterman, J.A., Phibbs. R.H. and Tooley, W.H. (1969): Aortic blood pressure in normal newborn infants during the first I2 hours of life. Pediatrics, 44. 959-968. Lum, L.C. and Douglas Jones, M., Jr. (1977): The effect of cuff width on systolic blood pressure measurements in neonates. J. Pediatr.. 91. 963-966. Moss. A.J., Duffe, E.R. and Emmanouilides, G. (1963): Blood pressure and vasomotor reflexes in the newborn infant. Pediatrics, 32. 175-179. Oh, W.. Lind, J. and Gessner, I.H. (1966): The circulatory and respiratory adaptation to early and late cord clamping in newborn infants. Acta Paediatr. &and., 55, 17-25. Park, M.K. and Gunteroth, W.G. (1970): Direct blood pressure measurements in brachial and femoral arteries in children. Circulation, XLI, 231-237. Park, M.K. and Lee, D.-H. (1989): Normative arm and calf blood pressure values in the newborn. Pediatrics, 83, 240-243. Picton-Warlow, C.G. and Mayer, F.E. (1970): Cardiovascular responses to postural changes in the neonate. Arch. Dis. Child., 45. 354-59. Prechtl, H.F.R. (1974): The behavioural states in the newborn infant. Brain Res., 76. 185-212. Rusch. N.J., Shepherd, J.T., Clinton Webb, R. and Vanhoutte, P.M. (1981): Different behavior of the resistance vessels of the human calf and forearm during contralateral isometric exercise. mental stress, and abnormal respiratory movements. Circ. Res., Suppl. I. 48. 118-130. Stopfkuchen, H. (1987): Changes of the cardiovascular system during the perinatal period. Eur. J. Pediatr., 146, 545-549. Swiet. M. de, Peto, J. and Shinebourne, E.A. (1974): Differences between upper and lower limb blood pressure in normal neonates using Doppler technique. Arch. Dis. Child.. 49. 734-735. Tan, K.L. (1987): Blood pressure in full-term healthy neonates. Clin. Pediatr., 26, 21-24. Thoresen. M. and Cowan F. (1991): Dinamap blood pressure measurements in the newborn: how many - what effects? Acta Paediatr. Stand., in press. Versmold. H.T.. Kitterman. J.A.. Phibbs. R.H.. Gregory. G.A.. and Tooley. W.H. (1981): Aortic blood pressure during the first I2 hours of life in infants with birth weight 610 to 4,220 grams. Pediatrics, 67. 607-613. Wu, P.Y.K., Wong. W.H., Guerra. G.. Miranda. R.. Godoy, R.R.. Preston, B., Schoentgen, S. and Levan, N.E. (1980): Peripheral blood flow in the neonate. I. Changes in total: skin, and muscle blood flow with gestational and postnatal age. Pediatr. Res., 14. 1374-78.
203
EHD 01177
Arm and leg blood pressures - are they really so different in newborns? Frances
Cowana,
Marianne
Thoresen”
and Lars Wall@eb
“Department of Paediatrics. Ullevaal Hospital, Oslo and hDeparment of Physiology. University of Oslo (Norway) (Received
6 September
1990; revision
received
28 August
1991; accepted
2 September
1991)
Summary Controversy still exists over differences between upper and lower limb blood pressures (BP) in neonates. We measured upper arm and calf systolic (S) and diastolic (D) BP and heart rate (HR) using 2 Dinamap 847 instruments simultaneously every half minute for several measurement periods of 5-10 min. Nine healthy term newborns were studied in active (AS) and quiet (QS) sleep on post-natal days 1 and 5. The results were examined using unbalanced analyses of variance. Arm SBP was 62.3 f 1.6 mmHg and DBP 35.5 + 1.0 mmHg on day 1 in AS and QS. Calf values were not significantly different but were slightly higher (by 2-3 mmHg) in AS. Arm SBP and DBP rose by 8.5 and 5 mmHg, respectively, between days 1 and 5 but calf pressures rose less. Calf SBP and DBP were significantly lower (by 4.6 and 3.4 mmHg, respectively) than the arm values in QS on day 5. Arm SBP and DBP were dependent on post-natal age but not on sleep state while calf SBP and DBP and HR were dependent on both. Mean HR rose with age from 1!4 to 117.6 bpm in QS and from 118.6 to 122.4 bpm in AS. Our non-invasive BP measurements were similar to available invasive data. We postulate that differences in arm and calf vasoreactivity account for the different dependence on sleep state and for the unequal changes in arm and calf BE from days 1 to 5.
blood
pressure;
oscillometric
Correspondence to: Dr. Frances Hospital. DuCane Rd.. London
method;
arm; calf; sleep; newborn
Cowan, Dept. of Paediatrics WI2 OHS. U.K.
0378-3782/91/$3.50 0 1991 Elsevier Scientific Published and Printed in Ireland
Publishers
and Neonatal
Ireland
Ltd.
Medicine,
Hammersmith
204
Introduction A rise in upper limb systolic BP in the first week of life has been well documented [3,13,17,19,25]. However, there is considerable variation in reported absolute values for upper limb BP [3,10,11,13,17,19,24,25], little information on upper limb diastolic BP [ 19,241 and little agreement on lower limb BP [ 10, I 1,13,19,24]. Reported differences between arm BP and leg BP vary from -17 mmHg to +20 mmHg [ 10,13,19,24]. Scant attention has been given to the influence of awake/sleep state during BP measurements [3]. We have tried to overcome some of the problems of other studies by simultaneously measuring upper arm BP and calf BP and HR in a quiet, undisturbed environment on the first and fifth day of life in healthy term infants in active (AS) and quiet sleep
(Qs). Subjects and Methods The study group comprised 9 healthy term newborn infants (mean weight 3501 g) vaginally delivered after uncomplicated pregnancies without any maternal medication. Systolic (SBP) and diastolic (DBP) upper arm BP and calf BP were measured every half minute using an oscillometric method (two Dinamap 847 instruments, Critikon, Tampa, FL) which has been shown to give values corresponding well to invasively determined arterial pressures in the neonate [7,12]. The largest neonatal cuff size possible was used. The infants lay prone and the arm was positioned away from the chest to prevent chest wall movement impinging on the cuff and influencing the oscillometric detection device. The elbow was straightened at least to 90”. The ECG from three chest electrodes was stored on computer for later analysis. Respiratory movements were recorded from a pressure sensor (Graseby Medical, Watford, U.K.) taped on the upper abdomen to help with defining sleep state. Sleep state was determined according to Prechtl [21] using body and eye movements, breathing pattern and HR variability. For each baby, arm and calf BP measurements were obtained during a median of 4 (range 2-7) 5-min periods in AS and QS sleep on the first and fifth days of life. The median value of the 10 BP measurements obtained during each 5-min period was used for the analysis. Recordings were made once the baby had been in a steady state for 5 min and those associated with body movement or change in sleep state were not used. We have previously determined that the 95% CI for one BP reading is f 7.2 mmHg while the 95% CI for 10 readings is only ~2.3 mmHg [26]. A variable number of 5-min measurement periods on each day and in each sleep state was obtained per child. To examine the significance of our observed results unbalanced analyses of variance (BMDP 3V) [2] were performed with the infant as a random variable and with day and sleep state and different interaction terms as possible fixed effect variables. The model gave a good tit to observed data and made it possible to estimate the relative influence of day and sleep state on upper arm and calf BP and HR.
205
Results The observed values (and S.D.) for HR and arm and calf BP in AS and QS on day 1 and 5 are summarised in Table 1. The differences between sleep states for each day and the increase between days for each sleep state are also given. There was a large individual variation in absolute HR values. HR increased from day 1 to 5 in both AS and QS and was higher in AS than in QS on both days. The difference between sleep states was greater on day 1 than day 5 and the increase from day 1 to 5 was greater during QS than AS. BP was much less variable between individuals than HR. Between days I and 5 observed arm SBP rose 9.5 and 8.1 mmHg while arm DBP rose only 5.5 and 5 mmHg in AS and QS, respectively. In the calf the increase in both SBP and DBP between days 1 and 5 was less than in the arm. There was little difference in arm SBP or DBP between sleep states on either day but leg SBP and DBP differed more especially on day 1. Since the number of 5-min recording periods per baby was different and there were individual differences in resting HR and BP the increases observed could be
TABLE
I
Observed heart rate and blood pressure results with SD. states and the rise from day AS. active
sleep; QS,
I
showing showing the differences between sleep
to day 5.
quiet sleep, S.D.,
standard
deviation.
Number
of observations: day I AS. II = IX:
day I Q’S, n = 23; day 5 AS. n = 40; day 5 QS n = 19.
AS
Heart
rate
tbpm)
(mmHg)
(mmHg)
(SD.)
diff
118.7
f
9.2
112.2
f
II.6
-6.5
day 5
122.5
f
8.3
118.2
f
13.9
-4.3
3.8
6.0
day I
62. I
f
2.8
62.4
f
3.3
-0.3
day 5
71.6
f
7.0
70.5
zt 6.9
-1.1
rise Calf SBP
QS
day I rise
Arm SBP
(S.D.)
9.5
8.1
day I
66. I
??
5.3
61.8
f
4.5
-4.3
day 5
68.9
f
6.3
67.8
zt 6.5
-1.1
rise
2.8
6.0
Arm DBP
day I
35.8
f
I.7
35.3
+z 2.6
-0.5
tmmHg)
day 5
41.3
f
4.8
40.3
?? 4.5
-1.0
rise
5.5
5.0
Calf DBP
day I
38.0
f
3.5
35.2
f
4.2
-2.x
0mnHg)
day 5
39.0
*
3.7
37.8
f
4.6
-1.2
rise
I .o
2.6
206
due to a skew distribution of the number of measurement periods between the individuals. To examine the significance of our observed results unbalanced analyses of variance were performed (see Subjects and Methods). In the first analysis both sleep stage, day and an interactive term between sleep state and day were included as possible fixed effects in the model. The result given in Table II, panel A showed
TABLE
II
Results of the unbalanced pressure.
analyses
of variance
for heart
rate and arm and calf systolic
SW’. systolic blood pressure; DBP, diastolic blood pressure: err, var., error variance; Variable
Parameter
Estimate
A. Heart rate
err. var. constant
47.18 118.14
7.00 2.85
day sleep day and sleep infant
-1.79 -2.32 0.34 68.03
0.77 0.76 0.77 34.26
err. var. constant day sleep
47.34 118.20 -1.82 -2.38
7.02 2.83 0.76 0.75
infant
67.20
33.88
err. var. constant
Il.65 66.59
I .73 1.51
day sleep infant
-4.27 -0.01 19.15
0.38 0.37 9.57
err. var. constant
2.70
day sleep
18.23 65.72 -I .98 -1.47
0.47 0.46
infant
16.81
8.77
err. var. constant
6.51 38.20
0.96 0.94
day sleep infant
-2.64 -0.03 7.18
0.28 0.28 3.69
err. var. constant day sleep
II.05 37.42 -0.84 -0.90
1.64 0.82 0.37 0.36
infant
4.83
2.78
B. Heart
rate
C. Arm SBP
D. Calf SBP
E. Arm DBP
F. Calf DBP
SD.
I .44
prob.,
and diastolic
probability,
Two-tail
prob.
0.000 0.020 0.002 0.657
0.000 0.017 0.001
0.000 0.000 0.982
0.000 0.000 0.002
0.000 0.000 0.903
0.000 0.022 0.012
207
that the interaction between sleep and day did not have significant effect at the 5% level on HR. Similarly there was no interactive effect between sleep and day on any of the BP results (not given in the Table). When the analysis was run without this interactive term, we obtained the results shown in Table II, panels B to F and Figs. 1 and 2. Day and sleep had significant effects on HR (Table II, panel B, and Fig. 1). Arm SBP and DBP were significantly affected by day but not by sleep (Table II, panels C and E). However calf SBP and DBP were significantly affected both by day and sleep (panels D and F). The BP results are summarised in Fig. 2. Table III shows a comparison of the observed and predicted results (with S.D.) from the best fitting models.
w 2
0 A active 0
70
68
arm
sleep P,/I’
a quiet sleep
I
1
leg
0 1
122
120 i
arm
c 118e
/’
c % 116I
4’ I’ /’ __-I_--
114-
112 p
Fig.
leg __--
_A i
/
I
1
1
2
I
3
Day after I. The rise in heart 0 ) as estimated
( 0 __
/’
,0’
I
4
I
5
I
I
I
I
I
1
2
3
4
5
Day after
birth
rate from day from the model.
I to day 5 in active sleep ( 0 -
Fig. 2. The upper half shows the rise in arm and leg systolic
pressure
birth 0 ) and quiet sleep
from day I to day 5 in active and
quiet sleep. The lower half shows the values obtained for diastolic pressure. The values plotted estimated from the model. Note the breaks in the blood pressure scale on the .r axis.
are those
(0) heart rate and blood
111 pressure
Blood Arm Calf Arm Calf
pressure SBP SBP DBP DBP
Heart rate (bpm)
62.3 62.3 35.5 35.7
114.0
112.2
(mmHg) 62.4 61.8 35.3 35.2
P
62.1 66. I 35.8 38.0
118.7
?? 3.l
+ 1.6 f 1.6 f I.0 ??I.0
0
SD.
AS
QS
0
day I n = I8
pressure;
day I n = 23
blood
results compared
AS, active sleep; QS, quiet sleep; SBP, systolic
Observed
TABLE
62.3 65.2 35.6 37.5
118.8
P
DBP. diastolic
zt zt * f
I.6 I.6 1.0 I.0
70.5 67.8 40.3 37.8
118.2
3.l ??
70.8 66.2 40.8 37.4
117.6
f zt f f
1.6 I.6 I.0 I.0
*3.l
71.6 68.9 41.3 39.0
122.5
0
AS
day 5
of the predicted
QS SD.
deviation
n = 40
P
S.D.. standard
day 5 n = I9
pressure;
0
blood
(P) from the model.
S.D.
to those predicted
70.9 69.2 40.9 39.2
122.4
P
results
??I.5 f I.5 ??I.0 zto.9
jz2.9
SD.
209
Discussion Many studies on HR and BP in newborn have been performed and the values given differ considerably. We feel that this is largely due to varying awake/sleep states and the use of methods that disturb the infant. We have measured BP in relation to post-natal age and sleep state in a situation where disturbance to the infant was minima1 and have discarded measurement periods associated with any movement or change of state. The effect of ductal closure on BP has not been well documented but most of our day 1 measurements were made after 12 h of age when for many full term infants the ductus arteriosus is functionally closed. We used the median of 10 consecutive measurements made at half-minute intervals as an estimate of each measurement period. We have previously found [26] the S.D. around 1, 3 and 10 measurements to be *7.2, +4.1 and h2.3 mmHg respectively. We took great care over the choice of cuff size and position as the use of inappropriately small cuffs overestimates the true BP [ 12,151. In addition, because it was not possible to obtain exactly the same number of measurement periods in each state for each baby on both post-natal days the significance of our observed results was examined using unbalanced analyses of variance. Our values for arm SBP and DPB on day 1 are almost identical to those of Versmold [27] obtained using invasive methods and slightly lower than those of Kitterman [l4] and Moss [16]. There is no invasive data for day 5. Values reported for newborn term babies using the oscillometric or Doppler methods are often higher condithan ours [ 17,24,25]. We feel this is most likely due to less stable measurement tions than ours in awake rather than asleep infants. Calf BP is not often measured invasively in newborn infants. However, Butt and Whyte [l] found the same good correlation between posterior tibia1 and dorsalis pedis and umbilical artery measurements as between the radial and the umbilical artery. In older children Park and Guntheroth [18] found simultaneous invasive measurements of BP in the brachial and femoral artery to be the same. Using noninvasive methods the calf BP is reported to vary from 20 mmHg higher to 17 mmHg lower than the arm BP [ 10,13,19,24]. We could not confirm these reported differences. We found calf and arm BP to be the same on day 1 as did Park in a recent study [19] but calf BP was slightly lower on day 5. This difference was statistically significant in QS (see Table III, predicted results). The rise in BP from day 1 to day 5 is well documented and is due to increased cardiac output in response to the large rise in metabolic demands in the perinatal period [23] and rise in peripheral resistance and possibly to ductal closure. Calf vascular resistance increases by 20-30% in the first week of life [ 11,281. We have not been able to find equivalent data for arm vascular resistance. We found that calf BP rose between days 1 and 5 by about half the rise in the arm. This suggests a difference in the vasoreactivity of the upper and lower limb major vessels. In adults the resistance vessels of the calf and forearm have been shown to behave differently in many physiological situations, calf vasculature being less reactive than that of the forearm [4,5,22]. Similar work has not been published in infants. The lower limbs of newborns are underdeveloped and have had a poorer, lower oxygen supply than the upper limbs during fetal life. In the light of adult studies it could well be that the
210
forearm responses are more marked accounting for the greater increase in arm BP compared to the calf. Arm BP was not dependent on sleep state but the calf BP was. Similar adult values for comparison are not known. However, this finding again suggests a difference in nervous regulating mechanism between the upper and lower limb of a more central origin [22]. We did not find any interactive effect of day and sleep state. HR is known to increase during the first week of life and our findings confirm this. Our values are similar to those of Finlay [6] and Picton-Warlow [20] though lower than many reported [8,9] reflecting the sleeping state and stable conditions of the infants. HR was dependent on sleep state being IO”% lower during QS than AS. In summary, we found little difference between our Dinamap BP measurements and available invasive data. Arm and calf measurements were very similar in this group of healthy term babies. Arm and calf BP rose differentially between days 1 and 5 and were not equally dependent on sleep state. The most likely explanation for this is that the newborn infant can preferentially increase peripheral resistance in the upper extremities as does the mature adult. Acknowledgements F.C. and M.T. were supported by the Norwegian Research Council for Science and the Humanities. We thank the staff of the post-natal ward (Barsel 2, Ullevaal Hospital) and the mothers for their cooperation with the study. References I
Butt, W.W. and Whyte. H. (1984): Blood pressure monitoring in neonates: and peripheral artery catheter measurements. J. Pediatr., 105, 630-632.
2 3
Dixon, W.J. (Ed.) (1981): BMDP Statistical Software. University of California Press, Berkeley. Earley, A., Fayers, P., Ng, S., Shinebourne. E.A. and de Sweit. M. (1980): Blood pressure in the first 6 weeks of life. Arch. Dis. Child.. 55, 755-757. Essandoh. L.K., Houston. D.S.. Vanhoutte, P.M. and Shepherd. J.T. (1986): Differential effects of lower body negative pressure on forearm and calf blood flow. J. Appl. Physiol.. 61, 994-998. Essandoh, L.K., Duprez, D.A. and Shepherd, J.T. (1987): Postural cardiovascular reflexes: comparison of responses of forearm and calf resistance vessels. J. Appl. Physiol., 63. 1801-1805. Finlay. J.P.. Hamilton, R. and MacKenzie. M.G. (1984): Heart rate response to tilting in newborns
4 5 6
Comparison
of umbilical
in quiet and active sleep. Biol. Neonate, 45. l-10. Friesen, R.H. and Lichtor. J.L. (1981): Indirect measurement of blood pressure in neonates and infants utilizing an automatic noninvasive oscillometric monitor. Anesth. Analg.. 60, 742-745. Haddad. G.G.. Epstein. R.A.. Epstein, M.A.F.. Leistner. H.L. and Mellins. R.B (1980): The R-R interval and R-R variability in normal infants during sleep. Pediatr. Res., 14, 809-81 I. Harper, R.M.. Hoppenbrouwers. T.. Sterman, M.B., McGinty. D.J. and Hodgman. J. (1976): Polygraphic studies of normal infants during the first six months of life. 1. Heart rate and variability as a function of state. Pediatr. Res.. 10. 945-951. Ho, T.F.. Yip. W.C.L., Tay. J.S.H. and Wong. H.B. (1985): Ultrasonic systolic blood pressure gradient between upper and lower limbs in normal term Chinese neonates. Clin. Pediatr., 24. 146-148. Kidd. L.. Levison. H., Gemmel. P., Aharon. A. and Swyer. P.R. (1966): Limb blood flow in normal and sick newborn. Am. J. Dis. Child., 112. 402-407. Kimble, K.J.. Darnall, R.A., Yelderman, M.. Ariagno. R.. and Ream. A.K. (1981): An automated oscillometric technique for estimating mean arterial pressure in critically ill newborns. Anesthesiology. 54, 423-425.
211
13 14 I5 16 17 I8 19 20 21 22
23 24 25 26 27
28
Kirkland,
R.T. and Kirkland,
J.L. (1972): Systolic blood pressure
measurement
in the newborn
in-
fant with the transcutaneous Doppler technique. J. Pediatr., 80. 52-56. Kitterman, J.A., Phibbs. R.H. and Tooley, W.H. (1969): Aortic blood pressure in normal newborn infants during the first I2 hours of life. Pediatrics, 44. 959-968. Lum, L.C. and Douglas Jones, M., Jr. (1977): The effect of cuff width on systolic blood pressure measurements in neonates. J. Pediatr.. 91. 963-966. Moss. A.J., Duffe, E.R. and Emmanouilides, G. (1963): Blood pressure and vasomotor reflexes in the newborn infant. Pediatrics, 32. 175-179. Oh, W.. Lind, J. and Gessner, I.H. (1966): The circulatory and respiratory adaptation to early and late cord clamping in newborn infants. Acta Paediatr. &and., 55, 17-25. Park, M.K. and Gunteroth, W.G. (1970): Direct blood pressure measurements in brachial and femoral arteries in children. Circulation, XLI, 231-237. Park, M.K. and Lee, D.-H. (1989): Normative arm and calf blood pressure values in the newborn. Pediatrics, 83, 240-243. Picton-Warlow, C.G. and Mayer, F.E. (1970): Cardiovascular responses to postural changes in the neonate. Arch. Dis. Child., 45. 354-59. Prechtl, H.F.R. (1974): The behavioural states in the newborn infant. Brain Res., 76. 185-212. Rusch. N.J., Shepherd, J.T., Clinton Webb, R. and Vanhoutte, P.M. (1981): Different behavior of the resistance vessels of the human calf and forearm during contralateral isometric exercise. mental stress, and abnormal respiratory movements. Circ. Res., Suppl. I. 48. 118-130. Stopfkuchen, H. (1987): Changes of the cardiovascular system during the perinatal period. Eur. J. Pediatr., 146, 545-549. Swiet. M. de, Peto, J. and Shinebourne, E.A. (1974): Differences between upper and lower limb blood pressure in normal neonates using Doppler technique. Arch. Dis. Child.. 49. 734-735. Tan, K.L. (1987): Blood pressure in full-term healthy neonates. Clin. Pediatr., 26, 21-24. Thoresen. M. and Cowan F. (1991): Dinamap blood pressure measurements in the newborn: how many - what effects? Acta Paediatr. Stand., in press. Versmold. H.T.. Kitterman. J.A.. Phibbs. R.H.. Gregory. G.A.. and Tooley. W.H. (1981): Aortic blood pressure during the first I2 hours of life in infants with birth weight 610 to 4,220 grams. Pediatrics, 67. 607-613. Wu, P.Y.K., Wong. W.H., Guerra. G.. Miranda. R.. Godoy, R.R.. Preston, B., Schoentgen, S. and Levan, N.E. (1980): Peripheral blood flow in the neonate. I. Changes in total: skin, and muscle blood flow with gestational and postnatal age. Pediatr. Res., 14. 1374-78.