Clinical Endocrinology (1 977) 6 )347-359.

NYCTOHEMERAL SECRETION OF GROWTH HORMONE IN NORMAL CHILDREN OF SHORT STATURE A N D IN CHILDREN WITH HYPOPITUITARISM AND INTRAUTERINE GROWTH RETARDATION P. M . HOWSE, P. H. W . R A Y N E R , J . W . WILLIAMS,* B. T . RUDD,* P. V. BERTRANDE,? C . R . S . THOMPSON$ A N D L . A. J O N E S $

Institute of Child Health, University of Birmingham, *Department of Clinical Endocrinology, Central Birmingham Health District, f Department of Mathematical Statistics, University of Birmingham and $ Department of Neurophysiology, University of Aston (Received 8 December 1975; revised I 1 November 1976; accepted 1 December 1976) SUMMARY

A continuous blood sampling technique has been used to monitor human growth hormone (GH) during sleep in fourteen normal short children (age range 6.5 - 15.0 years), twelve hypopituitary children (2.8 - 17.3 years), three children with psychosocial GH deficiency (4.0- 13.0 years), and three children with intrauterine growth retardation (9.5-11.3 years). The mean GH level of a 5 h sleep period (22.3003.30 hours) was used to represent the GH response to sleep. The GH response to insulin induced hypoglycaemia (IST) was also determined. In normal short children there was a significant relationship between 5 h mean GH levels and chronological age. The curve defining this relationship was similar to the third centile linear growth velocity curve. The 5 h mean GH levels of the hypopituitary and psychosocial GH deficiency children were more than 2 SD below the age related mean established for normal short children. The children with intrauterine growth retardation demonstrated values which were more than 2 SD above the age related mean. The evaluation of serum growth hormone (GH) during sleep, has recently been used as a test of pituitary function in growth retarded children (Eastman & Lazarus, 1973; Wise et al., 1975). Interpretation of the response is difficult because normal GH values during sleep have not been established. We have used a continuous blood sampling technique to define the range of nocturnal GH levels for normal short children. The sleep GH response of children with hypopituitarism, psychosocial GH deficiency and intrauterine growth retardation (IUGR) has been similarly investigated and is reported with reference to the range established for normal short children. METHODS AND MATERIALS The following groups of children were selected for study: short normal (Group I), hypopituitary (Group II), psychosocial GH deficiency (Group HI), and IUGR children (Group Correspondence: Dr P. M . Howse, Institute of Child Health, Francis Road, Birmingham B16 8ET.

347

3 48

P.M . Howse et al.

IV). All of the children had been referred t o the Birmingham Children’s Hospital for investigation of short stature; with two exceptions (Table 1) all were more than two standard deviations (SD) below the fiftieth centile for height. The following criteria were used to assign the children to the above groups: Group I (normal short children): (a) parental short stature (father < 157 cm or mother < 152 cm); (b) a linear growth velocity less than the third centile value; (c) absence of criteria listed below for other groups. Group I1 (hypopituitarism): (a) a progressive ‘fall off in linear growth velocity over 2 years; (b) marked immaturity of body form and facies, truncal obesity and delayed dentition; (c) other trophic hormone deficiency. Group 111 (psychosocial GH deficiency): (a) a social history suggestive of emotional deprivation, in addition to (a) and (b) of Group 11. Group IV (IUGR): (a) low birth weight for gestational age; (b) linear growth velocity similar t o Group I(b); (c) ‘bird’ headed facies, clinodactyly, loss of sequence in addition to delay in bone maturation.

Anthropometry This study was designed t o provide an anthropometric index of distinction between the groups. Height (measured on an Harpenden stadiometer) weight and bone age (Tanner et al., 1972) were monitored,at three-monthly intervals, for at least 1 year prior to the investigation. Measurements were continued in GH treated children in Group I1 and in Group I11 children following their removal into care. Height (Ht) at the time of investigation and the preceeding yearly height velocity (Ht Vel) were expressed in terms of their standard deviation score (SDS) for chronological age (CA) and bone age (BA) (Tanner et al., 1966) according t o the equation, SD score = (X --R)/Sx where X i s the measurement, R is the mean of the measurements of the standardizing group for the age, and Sx the SD of the group. The mid parental height standard deviation score (PAR Ht SDS) was also calculated (Tanner et al., 1970). Each of the three SDS were used to discriminate between the groups; by combining the above scores in the following equation:

D = L 1 (height SD score ) + L 2 (mid parent height SD score) t L3 (height velocity SD score) (Armitage, 1971) was devised which, calculated for each patient, would indicate the group a further value (0) t o which each patient was most likely to belong. Pubertal development was rated according to Tanner (1955). Growth hormone (Raben and Wilhelmi preparations) was supplied by the Medical Research Council, and was injected intramuscularly twice weekly.

Continuous sampling technique A teflon cannula inserted into an antecubital vein was connected by polyvinyl chloride (PVC) tubing and a Watson Marlowe flow inducer to an automatic sample collector. Adjustment of pump speed determined the volume of sample. Thrombogenesis was prevented by binding heparin to the cannula and tubing using tridodecylmethyl ammonium chloride (TDMAC) (Kowarski et al., 1971; Grode et al., 1969; G. A. Grode, personal communication, 1972). Glutaraldehyde further stabilised the heparin binding (Lagergren & Ericksson, 1971). Overnight continuous blood sampling for GH was performed on the third evening of admission except in those children with suspected psychosocial GH deficiency, who were

Nyctohemeral growth hormone in normal short children

349

investigated on the first evening. Blood was collected in 15 min aliquots between 22.00 and 06.00 hours. An Insulin Stimulation Test (IST) was then performed (dosage 0.1 u/kg > 15 kg and 0.05 u/kg < 15 kg). Both continuous blood sampling and IST were carried out under conditions of bed rest and fasting from 19.00 hours on the evening prior to the test. The sleep pattern was monitored polygraphically and was classified according to the method of Rechtschaffen & Kales (1 968).

Laboratory methods Growth hormone was assayed using a modification of the double-antibody radioimmunoassay (Hartog et aZ., 1964), and the W.H.O. 1st International Reference Standard 661127 for GH; GH values are expressed in mu/l. The standard deviation from the mean of replicate analyses was used as the index of precision. The precision of the method was k 0.6 mu/l in the range 0-20 mu/l and 2 1.3 mu/l in the range 20-80 mu/l. Blood glucose was determined by an automated glucose oxidase technique (Hill, 1965). RESULTS Anthropometry Growth data and results of statistical analyses are shown in Tables 1 and 2. There were insufficient children (three) in Groups 111 and IV for meaningful group analysis. As Group 111 children differed only in the aetiology of their hyposomatrophism from Group 11, these groups have been combined for statistical purposes. Group IV children have been excluded from group analysis. Comparison of Groups I and 11-111 means for the Height Standard Deviation Score for chronological age (Ht SD) and the Height Velocity Standard Deviation Score for bone age (Vel SD) indicates significantly lower values in the hyposomatrophic Groups I1 and 111. Although the Mid Parental Height Standard Deviation Scores (PAR Ht SD) do not differ significantly between the groups, the PAR Ht SD in Group I children (short normal) is significantly below the expected mean of 0 (P 0.18 should be GH deficient, and with D < 0.14 should be a short normal child and this, in the main, is demonstrated. There is also an indeterminate area between 0.14 and 0.18 which is occupied by five short normal children, three of whom show puberty, and three children with psychosocial GH deficiency. Reference to Table 2 indicates that all but two children (A.H., J.S.) of the GH deficiency group, have shown a significant increase in growth velocity (> 2 cm over pre-treatment year) on GH therapy. All three children with psychosocial GH deficiency similarly improved their growth velocities, after they had been removed from home to residential school.

Group I (short normal children) Sleep. There were multiple episodes of Stage 111-IV sleep, mean total per patient was 87 k 37.8 min. 95% of the Stage 111-IV sleep was contained withm the period 22.30-03.30 hours. GH levels during this 5 h period have been used t o define the GH response t o sleep.

Short normals

Group

A.Ca.' V.T. F.C.' H.BP M.G.'

R.K.

C.S. S.P. S.W. B.B.* M.H.

A.C.

G.B. N.M.

Patient

5.4 5.9 7.2 6.4 4 .O 6.4 2.7 5.9 6 .O 4.2 3.8 4 .O 3.3 4.3

3 .O 4 .O 5.5 5.8 5 .o 5 .O 5.6 5 .o 8.5 11.0 9 .O 11.5 11.0 11.0

6.5 7.0 7.7 7.8 8 .5 9 .O 9.5 10.2 12.2 13.2 13.6 14.6 14.7 15.0

Height SD for bone age Parental ht SD

Stage I1 puberty (Tanner, 1955).

-3.60 -1.12 - 1.46 -1.55 -0.98 1.85 -3.12 0.83 -1.71 - 1.95 0.25 1.03 -3.91 -2.45 - 1.06 -4.28 -0.13 -2.23 -6.41 - 3.60 -2.85 -0.57 0.11 -2.53 + 0.63 -0.41 - 1.64 -2.95 -3.43 - 1.97 -1.15 -2.81 -4.33 -1.42 - 1.39 -2.77 -2.60 -0.96 -1.12 0.46 - 3.54 -3.13k0.23 -0.99k0.37 -0.86f0.36

Height SD for chron. age

* Female, 'Tanner

Growth velocity (cm/yr)

Bone age (years)

Chron. age (years)

0.12 0.04 0.07 0.12 0.14 0.15 0.17 0.11 k0.02

0.13 0.15 0.01 0.09 0.18 0.07

D (see text)

Table 1. Anthropometric and hormonal data of normal short children

1.5 1.8 1.05 1.6 0.94 2.2 1.7 1.4 0.94 1.6 0.66 1.94 1.94 1.66

IST minimum

Blood glucose (rnmol/l)

Sleep peak

23 36 48 27 29 29 19 25 43 29 53 86 99 116

peak

21 18 10 17 27 10 33 40 31 31 42 31 58 23

IST

14.3 12.8 12.7 12.7 8.9 7.6 8.5 9.2 8.3 10.8 15.6 26.0 30.0 43.0

Sleep 5h mean

P1. GH (mu/l)

w

2 &

(h

5

0

Q

."a

0

v,

Intrauterine growth retardation

Psychosocial GH deficiency

R.W.A*

Hypopituitarisrn

M.H. K.N.

R.W.

M.R.*

S.H.*

R.M.

A.H. P.S. J.S.* R.B. P.T. P.Ta. A.L. D.1.r

N.R.

M.H.A S.M.n*

Patient

Group

6.0 5.1 4.5

5 .o 5.5 4.5

2 .o 2 .o 9.5

8.0 8 .O 7.O

6.5 4.8 3.5 3.2 5.1 2.5 4.6 3.6 4.5 2.3 2.4 0.6

1.o 2.8 2.0 5 .o 5 .o 6.3 7 .O 6 .O 8.5 11.7 8.O 8 .O

Bone age Growth (years) (cmlyr)

--2.9t0.37

-4.71t0.46 2 cm over the pretreatment year. The apparently inadequate growth response made by A.H. on GH therapy cannot be fully explained and is unlike reported cases (Wise et al., 1975) and A.L. in this study. However, the argument for GH deficiency in A.H. is supported by a reduction in growth velocity from the GH treatment year value of 6.4 cm t o 2.9 cm during the following year off GH therapy.

3 58

P.M . Howse et al.

Growth retarded children who demonstrate a normal GH response to IST but fail to secrete GH in response t o sleep, similar t o A.H. and A.L., have been described by Eastman & Lazarus (1973) and Wise et al. (1975). The disparity of their GH response t o the various stimuli may be interpreted on the hypothesis of multifocal, stimuli specific, receptor site GHRF release, discussed by Martin (1973). The correlation demonstrated between subnormal GH release during sleep and clinical features of hypopituitarism in the presence of an apparently adequate pituitary GH reserve, illustrates the importance of sleep GH secretion for the maintenance of normal growth. The GH response t o standard provocation tests in hypothyroidism is frequently low (Brauman & Corvilain, 1968; MacGillivray e l al., 1968). In patients with hypothyroidism and suspected hypopituitarism, the aetiology of the low GH response may be difficult to define. In this study there was a correlation between the subnormal GH response to sleep and the standard provocation tests in untreated hypothyroid (TSH deficient) suspected hypopituitarism (M.H. and S.M.) and in thyroxine corrected (TSH deficient) hypothyroid hyposomatotrophism. It may be difficult to distinguish clinically between those children with psychosocial GH deficiency and true GH deficiency in the absence of a relevant social history. Their initial IST GH response and their patterns of sleep and sleep GH secretion (delayed subnormal) are similar. However, the anthropometric analysis used in this study (D value) defines the position of the psychosocial GH deficient children at the ‘normal’ end of the hypopituitary range, i.e. overlap point with short normal children. This would be the expected position for temporary, endogenously reversible GH deficiency. Evaluation of the anthropometric analysis (D value) which was used in this study to provide an additional index of distinction between short normal and GH deficient children, has shown it t o be both accurate and sensitive. The grouping on this index correlated with that made on clinical impression and GH status at investigation. There were relatively few children in the overlap area. Analysis of these reveals that the normal short children present were those for whom anthropometric data is the least reliable, i.e. those in early puberty. The remainder included all the psychosocial GH deficient children and the previously discussed A.H. who might be expected t o occupy an intermediary position. Although the sleep test is probably the most physiological of all the tests of GH provocation, it is also the most difficult test t o perform. However, when the GH potential of growth retarded children is inappropriately or imprecisely defined by the standard provocation tests, the sleep GH response is a most valuable, diagnostic aid. The age related variation in nocturnal GH secretion which has been shown to parallel growth velocity in normal short children, in this study, is an interesting relationship, the underlying principles of which may be relevant t o the maintenance of optimum growth in GH treated hyposomatrophism.

ACKNOWLEDGMENTS

This investigation was supported by a project grant from the Medical Research Council. We wish to thank Dr W. R. Butt, Director, Clinical Endocrinology Unit, Central Birmingham Health District, for helpful advice and facilities for endocrine assays, and Dr M. A. Preece, Department of Growth and Development, Institute of Child Health, London, for his guidance and assistance in the anthropometric analysis.

Nyctohemeral growth hormone in normal short children

3 59

REFERENCES ARMITAGE, P. (1971) Statistical Methods in Medical Research, pp. 332-340. Blackwell Scientific Publications, Oxford. BRAUMAN, H. & CORVILAIN, J. (1968) Growth hormone response to hypoglycaemia in myxoedema. Journal of Clinical Endocrinology, 28,301 -304. EASTMAN, C.J. & LAZARUS, L. (1973) Growth hormone release during sleep in growth retarded children. Archives of Disease in Childhood, 48,502-507. FINKELSTEIN, J.W., ROFFWARG, H., BOYAR, R.M., KREAM, J. & HELLMAN, L. (1972) Age related change in the twenty-four hour spontaneous secretion of growth hormone. Journal of Clinical Endocrinology and Metabolism, 35,665-670. GOTLIN, W.G., MACE, J.W. & SILVER, H.K. (1971) Raised nyctohemeral (night and day) growth hormone levels in conditions with primordial short stature. Lancet, i, 626-627. CRODE, G.A., ANDERSON, S.J., GROTTA, H.M. & FALB, R.D. (1969) Non-thrombogenic materials via a simple coating process. Transactions of the American Society for Artificial Internal Organs, 15, 1-6. HALL, K. & FILIPSSON, R. (1975) Correlation between somatomedin A in serum and body weight development in healthy children and children with certain growth disturbances. Acta Endocrinologica, 78,239-250. HARTOG, M., GAAFFER, M A , MEISSER, B. & FRASER, B. (1964) Immunoassay of serum growth hormone in acromegalic patients. British Medical Journal, ii, 1229-1232. HILL, J.B. (1965) A method for measuring deviations from equilibrium of the glucose anomers in blood. Journal of Applied Physiology, 20,749-754. KENDAL, M.G. & STUART, A. (1961) The Advanced Theory of Statistics, Vol. 11, pp. 363-365. Charles Griffin & Co., London. KOWARSKI, A., THOMPSON, R.G., MIGEON, C.J. & BLIZZARD, R.M. (1971) Determination of integrated plasma concentrations and true secretion rates of human growth hormone. Journal of Clinical Endocrinology and Metabolism, 32,356-360. LAGERGREN, H.R. & ERIKSSON, J.C. (1971) Plastics with a stable surface monolayer of cross linked heparin; preparation and evaluation. Transactions of the American Society for Artificial Internal Organs, 17, 10-12. MacGILLIVRAY, M.H., ACETO, T. Jr & FROHMAN, L.A. (1968) Plasma growth hormone responses and growth retardation of hypothyroidism. American Journal of Diseases in Childhood, 115, 173-176. MacGILLIVRAY, M.H., FROHMAN, L.A. & DOE, J. (1970) Metabolic clearance and production rates of human growth hormone in subjects with normal and abnormal growth. Journal of Clinical Endocrinology, 30, 632-638. MARTIN, J.B. (1973) Neural regulation of growth hormonc secretion. New &gland Journal o f Medicine, 288,1384-1393. RECHTSCAFFEN, A. & KALES, A. (eds.) (1968) A Manual of Standardized Termino1og.v. Techniques and Scoring System for Sleep Stages o f Human Subjects. Public Health Service, U.S. Government Printing Office, Washington D.C. TANNER, J.M. (1955) Growth in Adolescence. Blackwell Scientific Publications, Oxford. TANNER, J.M., WHITEHOUSE, R.H. & TAKAISHI, M. (1966) Standards from birth to maturity for height, weight, height velocity and weight velocity in British children. Archives of Disease in Childhood, 41,454-471,613-635. TANNER, J.M., GOLDSTEIN, H. & WHITEHOUSE, R.H. (1970) Standards for childrens heights at age 2-9 years allowing for heights of parents. Archives ofDisease in Childhood, 45, 755-762. TANNER, J.M., WHITEHOUSE, R.H., HEALY, M.J.R. & GOLDSTEIN, If. (1972) Standards for Skeletal Age. International Children’s Centre, Paris. THOMPSON, R.G., RODRIGUEZ, A., KOWARSKI, A., MIGEON, C.J. & BLIZZARD, R.M. (1972) Integrated concentrations of growth hormone correlated with plasma testosterone and bone age in prcadolescent and adolescent males. Journal of Endocrinology and Metabolism, 35, 334-337. WISE, P.H., BURNETT, R.B., GEARY, T.D. & BERRIMAN, H. (1975) Selective impairment ofgrowth hormone response to physiological stimuli. Archives of’Disease in Childhood, 50,210-2 14.

Nyctohemeral secretion of growth hormone in normal children of short stature and in children with hypopituitarism and intrauterine growth retardation.

Clinical Endocrinology (1 977) 6 )347-359. NYCTOHEMERAL SECRETION OF GROWTH HORMONE IN NORMAL CHILDREN OF SHORT STATURE A N D IN CHILDREN WITH HYPOPI...
756KB Sizes 0 Downloads 0 Views