Sleep Disturbance in Children with Growth Hormone Deficiency Masaharu Hayashi, MD, Masayuki Shimohira, MD, Sumitaka Saisho, MD, Kazuhiko Shim ozawa, MD and Yoshihide Iwakawa, MD

We examined the effects of growth hormone (GH) deficiency on sleep development by performing all-night polysomnography in three female children with GH deficiency (GHD). The percentage of REM sleep seemed to be reduced before the treatment in 2 cases, and human GH (hGH) compensation slightly increased it. Submental twitch movements (mTMs), i.e., body movements during sleep localized in the submental muscle and lasting less than 0.5 seconds, were commonly disturbed in the three patients. Rapid eye movements in REM sleep (REMs) were reduced before the therapy in one case, this decrease being reversed on hGH compensation. REMs also seemed to increase after hGH treatment in the other two cases. Dopamines and cholinergic muscarinic agonists can cause GH release, while mTMs and REMs might be related to dopaminergic and cholinergic systems in the human brain. It is intriguing that GHD, and the disturbance of mTMs and REMs coexisted in children with GHD. Since a relatively poor social outcome in patients with GHD has been. reported, even after hGH compensation, it is important to monitor their neurological development by means of evaluation of their sleep disturbance. Key words: Growth hormone, growth hormone deficiency, dopamine, all-night polysomnography, sleep disturbance. Hayashi M, Shimohira M, Saisho S, Shim ozawa K, Iwakawa Y. Sleep disturbance in children with growth hormone deficiency. Brain Dev 1992; 14: 170-4

Several neurotransmitters in the central nervous system (CNS) are involved in the release of human growth hormones (hGH) through the regulation of the secretion of growth hormone releasing factors and somatostatins [1]. The secretion of hGH is regularly associated with the first episode of stage 4 sleep [2] and this sleep-related growth hormone secretion is influenced by various exogenous substances modulating neurotransmitters in CNS [1]. A relatively poor social outcome was reported in patients. with congenital growth hormone deficiency (GHD), though hGH compensation induced satisfactory physical growth in them [3,4]. It is known that most GHD is caused by perinatal accidents [5 , 6] and MRI demonstrated the transection of the pituitary stalk, which is suspected to be a sequela of the perinatal insult in some

From the Department of Pediatrics, Tokyo Medical and Dental University, Tokyo. Received for pUblication: October 2, 1991. Accepted for publication: March 11, 1992. Correspondence address: Dr. Masaharu Hayashi, Department of Pediatrics, Saitama Medical School, 38 Morohongo, Moroyama, Saitama 350-04, Japan.

cases of congenital GHD [7]. However, detailed neurological monitoring has not been performed yet. There have been few sleep studies on patients with GHD [8-11], even though sleep itself has an intimate relationship with growth hormone secretion [1]. Sleep parameters, such as the proportion of REM sleep, body movements (BMs) and rapid eye movements (REMs) during REM sleep, are closely connected with monoaminergic neuron systems in the brain [12]. In this investigation, we examined the effect of GH defiCiency on sleep development, by performing all-night polysomnography (pSG) in children with GHD. PATIENTS AND METHODS Three female children with GHD were examined (Table 1). Informed consent was obtained from all patients and their parents. A diagnosis of GH deficiency was established on the basis of the failure of plasma hGH to respond in general GH provocative tests. All patients were loaded with 0.1 U/kg rapidly active human insulin, 600 m/kg arginine and 10 mg/kg I-dopa. In cases 2 and 3, the nocturnal GH secretion profile was also examined. The peak value of nocturnal GH secretion is presented as the

Table 1 Clinical findings in the three cases Case

NA

1 2 3

(-)

(+) (+)

BH (em)

SD

CA

BA

SmC (U/ml)

133 128 116

-3.8 -2.4 -1.9

By 8m 11y 5m 9y 1m

12y 8m 8y2m 7y6m

0.62 1.37 l.29

Insulin 12 4~

It

The peak value ofpGH (ng/ml) Arginine I-dopa Sleep 10 6 6

2~

11

6

NT 12 11

NA: neonatal asphyxia, BH: body height, SD: standard deviations for Japanese children in similar age groups, CA: chronological age, BA: bone age estimated according to the atlas of Greulich and Pyle [13], SmC: somatomedin C, pGH: plasma growth hormone, NT: not tested. A diagnosis of growth hormone deficiency was established on the basis of the failure of plasma GH to reach 5 ng/ml in at least one provocative test, as described in detail under Patients and Methods. ~ indicates a low response of plasma GH in tests.

Table 2 Total sleep time and percentages of individual sleep stages Sleep stage (% of total sleep time) 3 4

Total sleep time (min)

I

2

Normal girls (1) Case 1. pre post

505.0 440.8

3.0 ± 1.3 17.0· 9.0·

48.7 ± 6.2 37.7 39.1

5.2 ± 1.4 15.4· 9.1·

16.5 ± 3.3 15.7 21.3

25.6 ± 3.8 14.3. 21.6

Normal girls (2) Case 2. pre post Case 3 pre post

2.3 606.6 662.3 502.4 516.7

47.9 ± 4.8 45.6 43.9 55.4 52.6

3.1 ± 1.2 5.3 7.5 6.0 5.2

16.7 ± 4.2 16.4 11.4 15.7 15.4

29.3 ± 4.8 29.7 30.5 17.0. 20.3

Case

± 1.4 3.0 6.7 5.9 6.4

REM

pre: PSG before the treatment, post: PSG after hGH compensation, Normal girls (1), (2): the data for normal girls aged from 13 to 15 years, and from 6 to 9 years, respectively, reported in William's normal standards [15] .• and. indicate significant increases and decreases, respectively, exceeding the mean ± 2SD for the standards.

Table 3 Frequencies of body movements during sleep and rapid eye movements in REM sleep mTM (/hour) Controls (n = 6)

47.4

±

14.8

GM (/hour) 9.0

±

1.8

REMs (/minute) 5.1

±

2.1

Case 1 pre post

62.1 33.6

8.7 8.3

l.l4.12.23

Case 2 pre post

21.n 12.7+

8.7 9.4

3.48 4.90

Case 3 pre post

15.0~

10.7 10.5

6.32 6.90

15.0~

mTM: twitch movements of the submental muscle, GM: gross movements, REMs: rapid eye movements during REM sleep, pre: PSG before the treatment, post: PSG after hGH compensation. The data for controls are expressed as means ± SD and ~ indicates a significant decrease, exceeding the mean ± 1.5SD in controls.

item, Sleep, in Table L The children were prepubertal and exhibited at least a one-year delay in skeletal age (bone age estimated according to the atlas of Greulich and Pyle [13]), as shown in Table L Their intelligence quotients, chromosomal study results, brain CT scans, and thyroid

and adrenal functions were normal_ Case 1 suffered from an additional anti-diuretic hormone deficiency, but she had received suitable replacement therapy two years before the present study _ Pituitary-derived hGR (pithGR), recombinant methionyl hGR (met-hGR) and recombinant methionine-free hGR (r-hGR) were prescribed in cases 1,2 and 3, respectively. We performed PSGs before and five days after the first administration of hGR in all cases. The total doses of hGR administered, when the second PSG was performed, were 8 IV of pit-hGR, 14 IV of met-hGR and 10 IV of r-hGR in cases 1, 2 and 3, respectively. PSGs comprised two channel electroencephalograms, electrooculograms with electrodes attached to the outer canthi, and surface electromyograms of the submental muscle, rectus abdominus muscle, and four or more other muscles in the trunk and limbs_ The sleep stage was determined every twenty seconds according to a standard system [14]. We calculated the proportions of each sleep stage, BMs during sleep and REMs in REM sleep_ BMs during sleep were classified into two types, as follows: gross movements (GMs), which involved the body trunk and lasted for more than 2 seconds, and twitch movements (TMs), which were localized in one muscle and lasted less than 0.5 seconds. The proportion of each sleep stage was

Hayashi et al: Sleep in GH deficiency

171

Control

Case 1

Case 2

Case 3

'71

..

'.1

,

'''I

.. .. ..• .,

131

::I

'21

J:!

"1

0

..

~ 'II

.a E ::I

Z

7'

.0

., .., n

21

'0 2

2

.. IIEII

Sleep stages

.. IIEII

2

2

.. IIEII

2

2

.. IIEII

2

Twitch movements (Submental muscl.)

Control

Case 1

Case 2

Fig 1 The occurence pattern of twitch movements of the submental muscle against each sleep stage. 0-0 : submental twitch movements (mTMs) on PSG before the treatment, .... : mTMs on PSG after the treatment.

Case 3

.0

.. .. •

70

::I

0

s:

..

'0

Q.

•

.a E

.0

::I

Z

"0

ao 2.

,. 2

a

.. IIEM

Sleep .tage.

2

a

.. IIEM

2

2

.. 11111

2

3

.. IIEII

Gross movement.

compared with Williams' normal standards [15] _ Regarding BMs and REMs, we also performed PSGs in 6 healthy age-matched children.

172 Brain & Development, Vol 14, No 3, 1992

Fig 2 The occurence pattern of gross movements against each sleep stage. "'-I> : gross movements (GMs) on PSG before the treatment, I>.-li>. : GMs on PSG after the treatment.

RESULTS Sleep structure The total sleep time and proportions of sleep stages are summarized in Table 2. The proportions of stages 1 and 3

in case 1 before the treatment exceeded the normal standards [IS], while the percentage of REM sleep was reduced in cases 1 and 3. These changes seemed to be improved on hGH administration (Table 2). Twitch movements ofthe submental muscle (mTMs) The frequency of mTMs both before and after the hGH therapy was reduced in cases 2 and 3, when compared with that in controls (Table 3). The distribution of mTMs as to the sleep stages was disturbed in case 1. Their occurrence was highest in REM sleep in normal children, while mTMs in case 1 mostly appeared in stage 1 sleep (Fig 1). These changes did not improve on hGH compensation (Table 1 and Fig 1). Gross movements (GMs) The frequency of GMs was within the normal range in all patients (Table 3). The occurrence of GMs in REM sleep decreased in case 1 on hGH compensation, when compared with that in controls (Fig 2). Rapid eye movements (REMs) during REM sleep REMs in case 1 were reduced, when compared with those in controls (Table 3). This decrease was ameliorated on hGH compensation. There was a tendency that REMs increased with hGH administration in the other two cases (Table 3).

DISCUSSION An episode of hGH secretion occurs within 90-120 minutes of sleep onset, which may account for 70-90% of the 24-hour secretion [2]. In rats, exogenously administrated GH increased REM sleep [16]. However, there have been few polygraphic studies regarding the sleep in patients with GHD. Guilhaum et al found a reversible deficit in stage 4 sleep recordings of 4 children with psychosocial dwarfism [8] . Taylor and Brook [9] reported a significant increase in the percentage of REM sleep in children with a genetic short stature and psychosocial dwarfism, but exhibiting normal GH secretion. According to the study by Wu and Thorpy [10], GHD children had more sleep stages 1 and 3, and less REM, as compared with age-matched normal children, and 6 of the 7 patients exhibited a decrease in stage 3 sleep after GH therapy. In adults, an increase in REM sleep time on 6 months GH treatment was noticed by Astrom et al [11]. As for the sleep structure before the treatment, case 1 exhibited increases in sleep stages 1 and 3, and a decrease in REM sleep. These are the same findings as reported by Wu and Thorpy [10]. Case 3 also showed a reduction of REM sleep. The changes in the sleep structure in cases 1 and 3 seemed to be ameliorated, though these observations might have been biased due to the first night effect [17]. The disturbance of REM sleep coincides with the

findings of others described above [10, 11]. Only in case 1 did GMs during REM sleep decrease with the treatment. In all three cases, mTMs were commonly disturbed. The distribution of mTMs as to the sleep stage was affected in case 1, while the average frequency of mTMs in whole sleep was reduced in cases 2 and 3. REMs in case 1 were decreased and this decrease was improved by the hGH therapy. REMs increased with the treatment in the other cases. Since patient 1 had the complication of anti-diuretic hormone deficiency, her hypothalamus-pituitary system might have been insulted more pronouncedly than those of the other two cases. This assumption is compatible with the finding in this study that all sleep parameters, i.e., the sleep structure, GMs, mTMs and REMs, were only impaired in case 1. There is clinical evidence of a close relationship between TMs and the dopaminergic neurons in CNS. The frequency of TMs is correlated with the dopamine metabolite, homovanillic acid, in the cerebrospinal fluid of patients with age-dependent epileptic encephalopathy [18]. The frequency and distribution pattern of TMs in sleep are affected in the dystonia syndrome, and these changes are almost completely reversed on treatment with L-dopa [19]. The occurrence of REM sleep might be mediated via the cholinergic neurons in CNS [20]. Inasmuch as it is well known that dopamines and cholinergic muscarinic agonists can cause GH release in human [1], it is intriguing that the disturbance of mTMs and REMs was observed in patients with GHD. Since perinatal accidents are thought to be involved in the pathogenesis of congenital GHD [S, 6], patients with congenital GHD might exhibit neurological abnormalities induced not only by GHD but also perinatal accidents. Some children with GHD exhibited electroencephalographic abnormalities, which could be ameliorated by treatment with hGH [S]. However, the disturbance of mTMs observed in this study was not reversed on S days hGH compensation, though the reduction of REMs in case 1 was improved by the hGH therapy. The changes in mTMs might have a closer relationship with the poor social outcome in treated patients with congenital GHD. The reliability and reproducibility of stimulated and spontaneous GH levels for identifying a child with low GH secretion have been the subject of debate [21]. Although there is a possibility that the pathogenesis in the present cases might not represent that in the typical patients with congenital GHD, our preliminary fmdings demonstrate that GH replacement may affect a specific sleep parameter in GHD children and that sleep recordings could be one way of directly monitoring the effect of GH on their neurological development.

REFERENCES 1. Dieguez C, Page MD, Sealon MF. Growth hormone neurore-

Hayashi et al: Sleep in GH deficiency 173

2. 3.

4.

5. 6. 7. 8.

9. 10. 11 .

12.

gulation and its alterations in disease states. Qin Endocrinol 1988;28: 109-43. Takahashi Y, Kiphis OM, Daughaday WHo Growth hormone secretion during sleep. J Qin Invest 1968;47:2079-90. Dean HJ, McTaggart TL, Fish DG, Friesen HG. The educational, vocational and marital status of growth hormonedeficient adults treated with growth hormone during childhood. Am J Dis Child 1985; 139: 1105-10. Pierson M, Leheup B, Scheitzer CL, Mangin E. The physical, psychological, educational and professional conditions of young adults given growth hormone for childhood growth hormone deficit. Acta Endocrinol 1986;279 (suppl): 130-4. Hibi I, Tanae A, Okada R. Etiological study in idiopathic pituitary dwarfism (in Japanese). Shoniko Rinsho (Tokyo) 1976;29 :849-62. Craft WH, Underwood LE, Van Wyk JJ. High incidence of perinatal insults in children with idiopathic hypopituitarism. J Pediatr 1980;96:397-402. Fujisawa I, Kikuchi K, Nishimura K, et aI. Transection of the pituitary stalk: development of an ectopic posterior lobe assessed with MR imaging. Radiology 1987; 165 :487-9. Guilhaum A, Benoit 0, Gourmelen M, Richardet JM. Relationship between sleep stage 4 deficit and reversible HGH deficiency in psychosocial dwarfism. Pediatr Res 1982; 16: 299-303. Taylor BJ, Brook CGD. Sleep EEG in growth disorders. Arch Dis Child 1986;61: 754-60. Wu RH, Thorpy MJ. Effect of growth hormone treatment on sleep EEGs in growth hormone-deficient children. Sleep 1988;11 :425-9. Astrom C, Pedersen AP, Lindholm 1. The influence of growth hormone on sleep in adults with growth hormone deficiency. Clin Endocrinol 1990;33:495-500. Hayashi M, Saisho S, Suzuki H, Shimozawa K, Iwakawa Y.

13. 14.

15. 16. 17. 18.

19.

20. 21.

Sleep disturbances in children with congenital and acquired hypothyroidism (in Japanese). No to Hattatsu (Tokyo) 1988;20:294-300. Greulich WW, Pyle SJ. Radiographic atlas of skeletal development of the hand wrist. Palo Alto, California: Stanford University Press, 1973. Rechtschaffen A, Kales A, eds. A manual of standarized terminology. techniques and scoring system for sleep stages of human subjects. Washington DC: Public Health Service, US Government Printing Office, 1968. Williams RL, Karacan I, Hurth CJ, eds. EEG of human sleep. New York· London· Sydney· Toronto: Wiley BiomedicalHealth Publications, 1974:26-68. Drucker-Colin RR, Spanis CW, Hunyogi J, Sassin JF, Mcgaugh JL. Growth hormone effects on sleep wakefulness in the rat. Neuroendocrinology 1975;18:1-8. Agnew HWJ, Webb WB, William RL. The first night effect: an EEG study of sleep. Psychophysiology 1966;2: 263-6. Suzuki H, Shimohira M, Kohyama J, Ogiso M, Iwakawa Y. A study on the correlation between cerebrospinal fluid levels of monoamine metabolites and some parameters during sleep in children with age-dependent epileptic encephalopathy. Brain Dev (Tokyo) 1987;9: 172. Segawa M, Hosaka A, Miyagawa M, Nomura Y, Imai H. Hereditary progressive dystonia with marked diurnal fluctuation. In: Eldridge R, Fahn S, eds. Advances in neurology. , Vol. 14. New York: Raven Press, 1976:215-33. Sakai K. Central mechanism of paradoxical sleep. Exp Brain Res 1984;8(suppl):3-18. Tassoni P, Cacciari E, Cau M, et aI. Variability of growth hormone response to pharmacological and sleep tests performed twice in short children. J Clin Endocrinol Metab 1990;71:230-4.

Microgyric and Necrotic Cortical Lesions in Twin Fetuses: Original Cerebral Damage Consecutive to Twinning? Cecile Bordarier, MD and Olivier Robain, MD

Extensive cortical necrosis associated with malformative microgyric-like lesions and with necrotic lesions of the white matter was observed in two male 25 week fetuses. These cases differed from previously reported cases of brain damage in monozygotic twins: both fetuses were affected and the lesions occu"ed early in pregnancy. before the end of neuronal migration, thus reSUlting in a cortical malformation associated with destructive lesions. Key words: Multicystic encephalomalacia, microgyria, brain damage, monozygous twins. Bordarier C, Robain O. Microgyric and necrotic cortical lesions in twin fetuses: original cerebral damage consecutive to twinning? Brain Dev 1992;14:174-8