Blood pressure rhythms during the perinatal period in very immature, extremely low birthweight neonates

Early Human Development 56 (1999) 49–56 www.elsevier.com / locate / earlhumdev

Blood pressure rhythms during the perinatal period in very immature, extremely low birthweight neonates a a, a Gabriel Dimitriou , Anne Greenough *, Vasiliki Kavvadia , Stephanos Mantagos b a

Children Nationwide Regional Neonatal Intensive Care Centre, King’ s College Hospital, London SE5 9 RS, UK b Department of Paediatrics, University of Patras School of Medicine, Patras, Greece

Received 12 March 1999; received in revised form 15 June 1999; accepted 22 June 1999

Abstract The aim of this study was to investigate if blood pressure (BP) rhythms were present in the perinatal period in very immature infants. Twenty-two infants, median gestational age 24–28 weeks, who had indwelling arterial lines with undamped arterial BP waveforms, were studied. The infants were all receiving intensive care under constant conditions. The hourly mean, systolic and diastolic BPs on days 2 and 7 were examined. A cosinor analysis of the mean BP was performed examining period lengths of 4, 8, 12, 16, 20, and 24 h to determine whether ultradian and / or circadian rhythms existed. On day 2, but not day 7, the mean and systolic BPs showed significant variation and circadian and ultradian rhythms were demonstrated. We suggest that maternal influences may be responsible for the BP rhythms noted in very immature infants on day 2.  1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Immaturity; Blood pressure; Circadian rhythms

1. Introduction In human fetuses, circadian rhythms in heart rate and body movements have been *Corresponding author. Tel.: 1 44-171-346-3037; fax: 1 44-171-924-9365. 0378-3782 / 99 / $ – see front matter  1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S0378-3782( 99 )00034-1

50

G. Dimitriou et al. / Early Human Development 56 (1999) 49 – 56

reported [1–4]. It has been argued that such rhythms could be either imposed by the maternal environment or evidence of the development of the fetal nervous system, including a body clock [5]. Circadian rhythm in blood pressure has been clearly documented in adults with a prominent fall at night, such that on average the blood pressure is then 15–20 mmHg lower than in the afternoon [6]. Investigation of full term newborns in the perinatal period, however, suggests that a circadian rhythm of blood pressure is present only in a minority [7]. In addition, a serial study carried out over the first 3 months demonstrated a circadian rhythm in heart rate only appeared between 15 and 30 days of postnatal life [8]. Those data [7,8] suggest that the circadian rhythms seen in the fetus are more likely to be due to maternal influences. If that explanation is correct then one would predict circadian rhythms would be demonstrated in very immature infants immediately after birth but, with decreasing maternal influence, disappear within the next few days as the infants’ postconceptional age would still be less than term. The aim of this study was to test that hypothesis by investigating if BP rhythms were present on days 2 and 7 after birth in infants less than or equal to 28 weeks of gestational age.

2. Methods Infants born at less than or equal to 28 weeks of gestation and of birthweight less than 1000 g were considered for this study. Infants who, on days 2 or 7, received therapies which could have affected their blood pressure, that is inotropes, blood volume expanders, neuromuscular blocking agents, indomethacin or diuretics, were excluded. Only data from those infants who on day 2 had an indwelling arterial cannula with an undamped arterial BP waveform were used in the analysis. Arterial cannulation had been performed for clinical reasons, that is because of the infant’s requirement for respiratory support on admission to the neonatal intensive care unit (NICU). An umbilical artery was cannulated with a 4 or 5 french gauge catheter and patency maintained by a slow infusion (1 ml / h) of a heparinized 0.45% saline solution, which was changed every 24 h. The arterial line was connected to a transducer (Medex, Viamed, Keighley, UK). The rated sensitivity of the transducer was 5 mV/ V per mmHg61%, which meant no amplifier gain adjustment was required when used with a Horizon 2000 monitor. The Horizon 2000 automatically offsets the static pressures in the transducer dome when the operator initiates zeroing and searches for a flat signal (less than 2 mmHg variation). If the signal was found to be flat for 5 s, the monitor offsets any static transducer output to produce a zero mmHg display. The monitor allows a continuous display of the arterial waveform and minute-by-minute changes in the systolic, diastolic and mean BP. The arterial waveform was considered to be damped and hence inaccurate, if there was no dicrotic notch and less than a 10-mmHg difference between the systolic and diastolic BP [9,10]. The nurses recorded the systolic, diastolic and mean BPs on observation charts every hour.

G. Dimitriou et al. / Early Human Development 56 (1999) 49 – 56

51

2.1. Analysis The BP data from days 2 and 7 were analyzed. On day 7, only data from infants whose arterial catheters remained indwelling and the signals undamped were used. An analysis of variance for repeated measures was applied to detect any statistically significant variation in the data. A cosinor analysis [11] using a software package (Chronolab 3.0, [12]) was performed on the blood pressure means of the population. This method is based on fitting the data to a cosine function of a fixed period. The following parameters of the fitted function were used for comparison: the mean percent rhythm (PR), the mesor (the rhythm-adjusted 24-h mean), the amplitude (the differences between the maximum and mesor values) and the acrophase (the time of the peak of the rhythm). The standard error, 95% confidence intervals and p value associated with the zero amplitude test of the group were also identified. The model reveals a 24-h rhythmic change by a circadian amplitude differing from zero. To determine whether statistically significant ultradian rhythms existed in the mean BP, cosinor analysis was performed examining period lengths of 4, 8, 12, 16, and 20 h.

2.2. Patients The BP data from 22 infants, median gestational age 26 weeks (range 23–28) and birthweight 845 g (range 566–986), who consecutively fulfilled the eligibility criteria during a 2-year period, were examined. Six of the 22 infants did not have indwelling arterial lines on day 7 and thus only the results of the remaining 16 infants were analyzed on day 7. Fourteen infants had RDS and received exogenous surfactant administration, four had congenital pneumonia and the other four respiratory distress due to severe prematurity [13]. All were ventilated conventionally on both study days with rates between 60 and 100 bpm, inspiratory times between 0.25 and 0.4 s and a positive end expiratory pressure of 3 cmH 2 O. Eight infants received continuous intravenous sedation. All the infants had their ventilator rate and inspiratory time manipulated so that they breathed synchronously with the ventilator. The infants were nursed within humidified double walled incubators at a neutral temperature [14] in the intensive care nursery. Overhead lighting was at a constant level and phototherapy given if the infant’s bilirubin exceeded that appropriate for their weight. None of the infants were enterally fed during the study. Arterial blood pressure monitoring is standard practice on the NICU. All the invasive techniques in this study were used for clinical purposes. This study was an audit of data being recorded for clinical purposes and no ethical problems were raised as the data were completely anonymized. The study did not need to be submitted to the Research Ethics Committee of King’s Healthcare NHS Trust for approval. 3. Results The systolic and mean BPs showed significant variation on the second (P , 0.05),

52

G. Dimitriou et al. / Early Human Development 56 (1999) 49 – 56

Fig. 1. The mean values of the systolic, diastolic and mean BPs of the study population for each hour are plotted for the second (upper graph) and seventh (lower graph) day. Moving averages (three points wide) were used to smooth the BP data: (d) systolic, (♦) mean, and (m) diastolic BP.

but not on the seventh, day (Fig. 1). A circadian rhythm of the mean BP was present on day 2 (P , 0.05), but not on day 7 (Table 1). Ultradian rhythms were present on the second, but not the seventh day (Table 1).

4. Discussion We have demonstrated that in very immature infants rhythms in BP were present on day 2, but not day 7, which is different from the findings in term babies [7,15]. There are many factors which influence the development of circadian rhythms. In

G. Dimitriou et al. / Early Human Development 56 (1999) 49 – 56

53

Table 1 Circadian and ultradian characteristics of the mean BP Period (h)

Percent rhythm

P value

Mesor (mmHg)

Amplitude (mmHg)

Mean

S.E.

Mean

Day 2 (n 5 22) 4 8 12 16 20 24

15.3 32.2 17.9 37.0 37.2 37.2

0.31 0.05 0.25 0.03 0.04 0.04

38.1 35.9 38.7 32.4 39.0 41.8

1.1 35.9 4.5 8.1 11.0 15.5

0.7 3.1 1.7 6.8 3.6 5.5

Day 7 (n 5 16) 4 8 12 16 20 24

18.6 21.6 30.1 30.4 30.5 30.5

0.84 0.50 0.46 0.55 0.57 0.58

42.5 41.2 36.1 33.1 28.5 22.8

1.9 3.1 6.0 9.8 15.0 21.4

0.4 2.1 7.5 10.4 15.0 20.7

a

Acrophase

(95% CI)a

Mean

(95% CI)

(0.0,0.0) (21.6,7.7) (0.0,0.0) (28.4,22.0) (214.1,8.6) (215.1,26.1)

2 68.2 2 185.6 2 103.3 2 103.5 2 235.9 2 301.9

(0,0) (2162, 2 314) (0,0) (283, 2 246) (2141, 2 309) (2179, 2 348)

(0,0) (0,0) (0,0) (0,0) (0,0) (0,0)

2 267.2 2 127.9 2 335.2 2 70 2 127.5 2 166

(0,0) (0,0) (0,0) (0,0) (0,0) (0,0)

95% Confidence intervals.

healthy adults, the acrophase of BP, heart rate, adrenal corticosteroid and catecholamine secretion coincide [16] and the sleep–wake routine is a dominant synchronizer of circadian rhythms [17]. In newborns, corticosteroid rhythm is absent [18] and sleep disorganized [19]. The acrophases of rhythms in neonates differ from those in adults, indicating that the endogenous oscillators are effective in the neonate born at term and that rhythms are not only imposed by the mother. Several authors [8,20] have suggested that the presence of the circadian rhythm, at least for heart rate, indicates maturity. Thus, it is perhaps not surprising [8,20] in the very immature infants we studied no circadian rhythm of BP was present on day 7. The effect of maturity at birth on the development of circadian rhythms is controversial [5,21]. Examination of the development of rhythmic changes in skin temperature and heart rate in infants born between 26 and 29 weeks of gestation suggested that circadian rhythms did not develop until at least 5 weeks after birth, that is a postconceptional age of 34 weeks [5]. Others, however, demonstrated circadian rhythms of rectal temperature at postconceptional ages between 28 and 34 weeks [21]. Those differences [5,21] may be due to variations in the severity of the initial illness of the two study populations, as well as the type of environment in which the infants were nursed. Initially, very immature infants are nursed in intensive care and, only when their medical condition has improved, are they transferred to a lower dependency nursery where exposure to light–dark cycles takes place and usually greater exposure to noise and handling during the daytime [5]. Those associated events provide environmental cues which may contribute to the development of circadian rhythms in light–dark periods. If the fetus were not born prematurely he or she would continue in a rhythmic environment. Hence it has been suggested some consideration should be given to earlier introduction of light–dark

54

G. Dimitriou et al. / Early Human Development 56 (1999) 49 – 56

periods, after the infant’s condition has been stabilized [5]. Nursing preterm infants such that they are exposed to periods of light and darkness may lead to subsequent benefits in sleeping pattern and weight gain [22], as well as facilitating the development of circadian rhythms [5,21]. Comparison of the roles of neurologic maturity and environmental time cues in the development of the entrained circadian sleep–wake rhythm in the preterm and term human infant suggest it is the length of exposure to environmental cues which determines entrainment [23]. In comparison to term infants, the circadian sleep–wake rhythm in preterm infants (27–35 weeks) entrained after a similar time of exposure to an environment with daily time cues, but this was at an earlier postconceptional age. The preterm infants, however, were studied once after a postconceptional age of 35 weeks and thus those data do not necessarily apply to younger infants. The immature infants we studied were exposed to constant light conditions which might impair the developing biological clock as demonstrated in animal studies [24], hence they lacked circadian rhythms in BP at 7 days. A further factor which appears to influence the timing of circadian rhythm development in infants receiving intensive care is the type of care they receive [25]. Infants who receive individual caretaking from birth are able to differentiate between day and night with regard to sleep and wakefulness by day 4, but this was delayed to day 11 in those nursed in a hospital nursery with multiple caregivers [25]. We did demonstrate circadian rhythm in the mean BP on day 2. The fetal biological clock of the post-mortem human brain is histologically detectable from mid gestation onwards [26]. Circadian rhythms have frequently been demonstrated in the fetus [1–4]. Under normal conditions the maternal circadian system entrains the fetal biological clock to the 24-h day–night cycle of the environment from 22 weeks of gestation onwards. Our identification of rhythms on day 2 but not day 7 in these infants of 23–28 weeks gestational age is in keeping with the hypothesis that the rhythms on day 2 were due to lingering maternal influences. When assessing BP recordings for rhythms it is essential to avoid artefact. The most reliable method of measuring BP in VLBW infants is via an intra-arterial catheter [10]. We excluded infants whose BP trace was damped and examined for rhythms the mean, rather than the systolic or diastolic, BP as measurement of the mean BP is considered to be accurate [10,27]. Cyclical variations in BP have been reported with bolus administration of vasoactive drugs [28], hence infants receiving such infusions were excluded. In addition, we initially studied infants on day 2 rather than day 1, to avoid any confusing effect of exogenous surfactant administration on the BP recordings [29]. We demonstrated ultradian rhythms of BP. In term infants an ultradian rhythm of heart rate with a period of 3 h was demonstrated, which disappeared at 1 month when a circadian rhythm appeared [8]. The rhythm on day 1 appeared to be endogenous, as factors including feeding routines had no effect on it [8]. It has been suggested that ultradian rhythms grow into circadian rhythms [20]. An alternative hypothesis is that over the first month the circadian components gradually and erratically come into phase with one another and a dominant frequency develops [30]. Thus, it would not be surprising to see, as we demonstrate for BP and has been seen for skin temperature and heart rate [30], both types of rhythms together in very immature infants. Our data

G. Dimitriou et al. / Early Human Development 56 (1999) 49 – 56

55

and that of Tenreiro et al. [30] argue against the hypothesis that circadian rhythms mature from ultradian rhythms [20]. These data have implications for clinical practice. Infants who are at increased risk of developing periventricular haemorrhage (PVH) have wide BP swings for a greater proportion of the time than infants who do not develop PVH [27,31]. Accurate identification of abnormality is only possible if the physiological variability of BP readings throughout the 24-h period are known. We demonstrate that even amongst very immature infants, circadian and ultradian rhythms in BP were present on day 2, which should be taken into account when assessing if an individual’s BP level is abnormal. In conclusion, very immature infants have BP rhythms on the second day after birth. We suggest that this is due to lingering maternal influences rather than evidence of an endogenous ‘clock’, as the rhythms were absent on day 7.

Acknowledgements Dr Dimitriou is supported by the Children Nationwide / Nestle’ Research Fellowship and Dr Kavvadia by the Research and Development Directorate of the South Thames (East) Regional Health Authority. We thank Sue Williams for secretarial assistance.

References [1] Patrick J, Campbell K, Carmichael L, Probert C. Influence of maternal heart rate and gross fetal body movements on the daily pattern of fetal heart rate near term. Am J Obstet Gynecol 1982;144:533–8. [2] Patrick J, Campbell K, Carmichael L, Natale R, Richardson BSO. Patterns of gross fetal body movements over 24-hour observation intervals during the last 10 weeks of pregnancy. Am J Obstet Gynecol 1982;142:363–71. [3] Visser GHA, Goodman JDS, Levine DH, Dawes GS. Diurnal and other cyclic variations in human fetal heart rate near term. Am J Obstet Gynecol 1982;142:535–44. [4] de Vries JI, Visser GH, Mulder EJ, Prechtl HFSO. Diurnal and other variations in fetal movement and heart rate patterns at 20–22 weeks. Early Hum Dev 1987;15:333–48. [5] D’Souza SW, Tenreiro S, Minors D, Chiswick ML, Sims DG, Waterhouse J. Skin temperature and heart rate rhythms in infants of extreme prematurity. Arch Dis Child 1992;67:784–8. [6] Millar-Craig MW, Bishop CN, Raftery EB. Circadian variation of blood pressure. Lancet 1978;i:795– 7. [7] Gemelli M, Manganaro R, Mami C, Rando F, De Luca F. Circadian blood pressure pattern in full term newborn infants. Biol Neonate 1989;56:315–23. [8] Ardura J, Andres J, Aldana J, Revilla MA, Aragon MP. Heart rate biorhythm changes during the first three months of life. Biol Neonate 1997;72:94–101. [9] Versmold HT, Kitterman JA, Phibbs RH, Gregory GA, Tooley WH. Aortic blood pressure during the first 12 hours of life in infants with birthweight 610 to 4222 g. Pediatrics 1981;67:607–13. [10] Weindling AM. Blood pressure monitoring in the newborn. Arch Dis Child 1989;64:444–7. [11] Nelson W, Tong YL, Lee JK, Halberg F. Methods for cosinor rhythmometry. Chronobiologia 1979;6:305–23. [12] Mojon A, Fernandez JR, Hermida RCSO. Chronolab: an interactive software package for chronobiologic time series analysis written for the Macintosh computer. Chronobiol Int 1992;9:403– 12.

56

G. Dimitriou et al. / Early Human Development 56 (1999) 49 – 56

[13] Chan V, Greenough A, Gamsu HR. Neonatal complications of extreme prematurity in mechanically ventilated preterm infants. Eur J Pediatr 1992;151:693–6. [14] Sauer PJJ, Dane JJ, Visser HKA. New standards of neutral thermal environment of healthy very low weight infants during the first week of life. Arch Dis Child 1984;59:18–22. [15] Sitka U, Weinert D, Berle K, Rumler W, Schuh J. Investigations of the rhythmic function of heart rate, blood pressure and temperature in neonates. Eur J Pediatr 1994;153:117–22. [16] Faucheux B, Kuchel O, Cuche JL, Messerli FH, Buu NT, Barbeau A, Genest JSO. Circadian variations of the urinary excretion of catecholamines and electrolytes. Endocr Res Commun 1976;3:257–72. [17] Halberg F, Scheving LE, Lucas E, Cornelissen G, Sothern RB, Halberg E, Halberg J, Halberg F, Carter J, Straub KD. Chronobiology of human blood pressure in the light of static (room-restricted) automatic monitoring. Chronobiologia 1984;11:217–47. [18] Mantagos S, Moustogiannis A, Vagenakis AG. Diurnal variation of plasma cortisol levels in infancy. J Pediatr Endocrinol Metab 1998;11:549–53. [19] Mills JN. Development of circadian rhythms in infancy. Chronobiologia 1975;2:363–71. [20] Hellbruegge TSO. The development of circadian and ultradian rhythms of premature and full term infants. In: Scheving LE et al., editors, Chronobiology, Igaku Shoin, 1974, pp. 339–41. [21] Mirmiran M, Kok JH. Circadian rhythms in early human development. Early Hum Dev 1991;26:121– 8. [22] Mann NP, Haddow R, Stokes L, Goodley S, Rutter NSO. Effect of night and day on preterm infants in a newborn nursery: randomised trial. Br Med J 1986;293:1265–7. [23] McMillen IC, Kok JS, Adamson TM, Deayton JM, Nowak RSO. Development of circadian sleep–wake rhythms in preterm and full-term infants. Pediatr Res 1991;29:381–4. [24] Cambras T, Diez-Noguera ASO. Evolution of rat motor activity circadian rhythm under three different light patterns. Physiol Behav 1991;49:63–8. [25] Sander LW, Julia HL, Stechler G, Burns PSO. Continuous 24-hour interactional monitoring in infants reared in two caretaking environments. Psychosom Med 1972;34:270–82. [26] Lydic R, Schoene WC, Czeisler CA, Moore-Ede MCSO. Suprachiasmatic region of the human hypothalamus: homolog to the primate circadian pacemaker? Sleep 1980;2:355–61. [27] Miall-Allen VM, de Vries LS, Whitelaw AGL. Mean arterial blood pressure and neonatal cerebral lesions. Arch Dis Child 1987;62:1068–9. [28] Dunster KR, Colditz PB, Joy GJ. Cyclical variation of blood pressure and heart rate in neonates [letter]. Arch Dis Child 1994;70:F77. [29] Cowan F, Whitelaw A, Wertheim D, Silverman M. Cerebral blood flow velocity changes after rapid administration of surfactant. Arch Dis Child 1991;66:1105–9. [30] Tenreiro S, Dowse HB, D’Souza S, Minors D, Chiswick M, Simms D, Waterhouse JSO. The development of ultradian and circadian rhythms in premature babies maintained in constant conditions. Early Hum Dev 1991;27:33–52. [31] Gronlund U, Korvenranta H, Kero P, Jalonen J, Valimaki IASO. Elevated arterial blood pressure is associated with peri-intraventricular haemorrhage. Eur J Pediatr 1994;153:836–41.