336
1331
MEDICAL SCIENCE
Treatment of short normal children with hormone—a cautionary tale?
41 short normal children were randomly allocated either to daily injections of growth hormone (rhGH) at 30 lU/m2 per week or to no treatment. 6 months of rhGH therapy produced up to 76% loss of fat mass and up to 25% increase in lean body mass (LBM). These changes were significantly different from those in the untreated group. LBM was the main determinant of resting energy expenditure (REE) expressed as kJ/24 h. REE expressed as kJ/kg LBM per 24 h correlated negatively with height, which was responsible for 66% of the variance in REE kJ/kg LBM per 24 h. Short children therefore expend more energy than tall childen in fulfilling basic metabolic needs. After 6 months REE kJ/24 h increased significantly in treated children. However, treated children did not differ significantly from untreated children in REE kJ/kg LBM per 24 h. rhGH does not therefore seem to have a specific effect upon REE. The possibility that rhGH produces profound metabolic effects should limit its use in otherwise healthy children until the mechanism of action is more clearly elucidated. 1990;
growth
has been the collection of growth data.2-5 However, the clinical features of GHD in both adults and children classically include relative obesity, which disappears with hormone replacement,6,7 whereas acromegalic patients have much less body fat than do normal individuals.8 In view of its known lipolytic9 and anabolic effects10 rhGH could induce substantial and possibly stressful metabolic changes in short but otherwise healthy children with normal body
composition. The Wessex Growth Study is a longitudinal, prospective, community-based study of the growth patterns of all children in two neighbouring health districts found during a 2 year period to be on or below Tanner’s 3rd height centile" at school entry .12 Most previous work has been based on 15 IU 1m2 per week,24 the standard replacement dose for growth hormone deficiency. Short children without growth hormone deficiency probably need a greater dose than this,22 and up to 40 mg IU/m2 per week has been used.5 In the Wessex study a randomised cohort of children is being treated with 30 IU/m2 per week. Anthropometric, metabolic, and psychometric variables are being monitored. Here we report the baseline data collected during the first year of the rhGH trial, with emphasis on body composition and resting energy expenditure (REE) because these may be useful markers for metabolic and nutritional effects.
Introduction The
of recombinant
growth hormone (rhGH) as a encouraging growth in short children without classical growth hormone deficiency (GHD) is being investigated. These children seem to have normal patterns of endogenous growth hormone secretion, albeit at lower levels than do tall children.1 The emphasis in these studies use
means
of
ADDRESS: Departments of Paediatrics, Human Nutrition and Medicine 2 (Endocrinology), Southampton General Hospital and Southampton University, Southampton, UK (J. M. Walker, MRCP, S A. Bond, BSc, L D Voss, MA, P. R. Betts, FRCP, S. A. Wootton, PhD, A. A. Jackson, FRCP). Correspondence to Dr J. M. Walker, University Department of Paediatrics, Level 8, Addenbrooke’s Hospital, Cambridge CB2 2QQ, UK.
1332
Patients and methods Local and Committee
approval and study.
Safety of Medicines ethical committee parental consent were obtained for this
on
informed
Study population 41 children smaller than or equal to -20 standard deviation (SDS) but without GHD or other known disorders were randomly allocated to either treatment with rhGH (’Genotropin’ Kabi-Pharmacia, Milton Keynes, UK) at 30 IU/m2 per week scores
(n = 21) or to no treatment (n = 20). The rhGH was given by daily subcutaneous injection with an ’Auto Injector’ (Owen Mumford Ltd, Woodstock, Oxon, England). The children were aged 69-83 yr at randomisation, and the sex distribution was the same in the two groups.
Growth Each of the 41 study children has a control matched for age, sex, and school class. All 82 children are being measured every 6 months by one trained anthropometrist with a portable electronic stadiometer (Holtain Ltd, Crymych, Wales).
Body composition Body composition, as assessed by anthropometry, was recorded randomisation and every 6 months. Skinfold was measured by one observer in the standard manner with Harpenden skinfold calipers at the biceps, triceps, subscapular, and suprailiac sites in each child. Body density13 and body fat14 were calculated as previously reported. Lean body mass (LBM) was calculated by subtracting the derived fat mass from body weight. To avoid bias, measurements were made without access to previous data. at
of energy requirements during childhood REE was also measured in 20 healthy local schoolchildren whose age-range was the same as that among the short children and whose heights were between the 3rd and 97th Tanner centiles.11 REE was measured by indirect calorimetry with an open-circuit ventilated hood system. All studies were done between 0730 and 1000 h after an overnight fast and the room temperature throughout was maintained at 22-24°C. To induce the resting state the child was given a story or music to listen to through personal stereo headphones whilst supine in bed. Once a steady reading was obtained, recordings were made for at least 30 min. All the children were cooperative and able to complete the study. Air was drawn through a ventilated hood (volume 50 litres) at a flow rate that was adjusted to maintain at 0,5-0,65% the concentration of CO2 in air leaving the hood; the flow rate was usually 30-40 litres/min. The flow rate was measured by use of a vortex-shedding flowmeter and the temperature of the expired air was continuously measured with a thermistor probe at the point of gas sampling. Inspired, expired, and calibration gases were continuously sampled and analysed for 02, CO2, and N2 by a respiratory quadropole mass spectrometer (V. G. SpectraLab, Middlewich, England), which was calibrated before and after each test. Gas concentration, flow, and temperature were processed by microcomputer (Apple IIe) to give O2 consumption and CO2 production corrected to standard temperature and pressure. REE was determined from O2 consumption and respiratory exchange ratio according to Jécquier15 and expressed as kj/24 h and as kj/kg LBM per 24 h.
Statistical analysis Least squares regression analysis, Student’s t test, rank sum tests, and multivariate regression analysis were calculated by use of the
’Statsgraphics’ (STSC Inc, Rockville, Maryland, USA) package.
Results
Resting energy expenditure (REE) REE was determined in 10 treated and 9 untreated children before randomisation and every six months. As part of a wider study
Growth and body composition data for the first 6 months available for 18/21 treated and 19/20 untreated children. In the treated group 2 children have withdrawn from the study (1 because of dislike of injections, and 1 because of failure to give injections and attend for follow up), and 1 was known to have not complied well with the regimen. 1 untreated child also did not attend for follow-up. These 4 children were excluded from the analysis. were
Growth The untreated and control children continued to grow at a rate appropriate for their height whereas, even allowing for the error of measurement (SD for a single height measurement was 0-25 cm16), those treated with rhGH showed a sharp increase in growth velocity. The mean change in growth velocity SDS (GVSDS) over the first 6 months was from -0-60 SDS to +4-59 SDS (p< 0-001), which was equivalent to a mean increase in growth rate of 4-28 cm/year (p < 0-001), no significant changes in GVSDS occurred in the untreated children (p > 0 1 for both GVSDS and growth rate). 12-month data are available from 13 treated and 15 untreated children, and again the latter show no significant change in GV. Those on rhGH show a sustained increase in GV throughout the first year of, on average, 3 83 cm/year or from - 0 60 SDS to + 4.24 SDS.
Body fat - 10l
Individual children
Fig 1-Percentage change in body mass (lower) at 6 mo.
fat
(upper)
and lean
body
Children treated with rhGH showed a significant loss of fat mass of up to 76% (fig 1), the mean loss being equivalent to 0-68 kg of fat from an initial fat mass of 2-33 kg. In contrast the untreated children, whose mean initial fat mass was 2 41
1333
expressed as kj/24 h. However,
when REE was per unit LBM substantial variation remained.
expressed Stepwise multivariate analysis of these results revealed that height alone contributed 66% of the variance of REE expressed as kj/kg LBM per 24 h, shorter children using more energy per unit of LBM than tall children. Age and weight, widely used predictors of REE,1’ did not account significantly for a greater proportion of the variance. Height seems therefore to be the principal variable contributing to REE kj/kg LBM per 24 h. Figure 2 shows this relation and for clarity removes the non-significant influence of age by using height SDS as the index for height, an SDS 0 being the 50th centile for height." =
r.rvT
-4
-3
-2
-1
0
1
2
3
4
Height SDS
REE and treatment with rhGH
Fig 2-Relation between resting energy expenditure expressed as kg/kg LBM per 24 h and height SDS.
After 6 months REE (kj/24 h) increased significantly in the treated children by, on average, 480 kJ/24 h or 12% of the pre-randomisation value (fig 3). In the untreated group it did not alter significantly. However, when expressed per unit LBM the increase in REE with treatment disappeared, and after 6 months there was no significant difference in REE kj/kg LBM per 24 h between the two groups (p > 0’ 1).
(p < 0 001)
an average of 0-34 kg (p < 0-001 for difference between treated and untreated group). The mother of 1 girl was concerned because her daughter became visibly fatter and she has since entered puberty. 2 others had a major change in social circumstances during this time, and 1 of these also had a marked rise in his growth velocity of 3-3
kg, gained
cm/year. Lean body mass
Discussion
(LBM)
This
LBM increased in both groups (fig 1, lower), but more so (p < 0-001) in the treated group, which had an increase of up to 25%. The mean gain was 2-81 kg in the treated and 0-8 kg in the untreated group. 12-month data are available for 13 treated and 15 untreated children. In most children the changes in body fat and LBM persisted, and in some, including those who had shown only slight changes at 6 months despite good compliance, they were greater at 1 year than at 6 months.
Resting energy expenditure (REE) before treatment A correlation of 0-8 (p < 0-001) was found between REE expressed as kj/24 h and LBM in normal children. Stepwise multivariate analysis revealed that LBM was the main determinant of REE when expressed in absolute units (kj/24 h) and accounted for 75 % of the variance. Addition of age, height, and weight into the model did not significantly improve the relation and when brought into the equation could account only for a further 1 % of the variance in REE
- 200
Fig
L
kJ/24h
3-Mean change in REE at 6 months when expressed or as kJ/kg LBM per 24 h.
kJ/24 h
as
study has shown that in short but otherwise normal children, the striking increase in growth velocity produced by high doses of rhGH is accompanied a profound alteration in body composition characterised by a loss of body fat of up to 76% and a gain in LBM of up to 25%. Despite the known lipolytic and anabolic properties of growth hormone, the effect of rhGH on body composition in short normal children has not been extensively reported. Hindmarsh and Brook, who examined only triceps and subscapular skinfolds, reported significant decreases after 6 months in children treated with doses of 12-20 IU 1m2 per week, with return to baseline levels at 12 months.i8 Even studies using higher doses of rhGH than we did here make no reference to body composition,s which seems surprising in view of the striking clinical changes. The children became skinny, with obvious loss of adipose tissue from all areas of the body but especially from the limbs and face. The girls in particular looked innappropriately muscular. The increase in growth velocity after 6 months was greater than seen in other similar studies and is presumably dose related,3,4 though it may be too soon to be certain of this difference. Taken together, these fmdings give the impression that the increase in growth velocity and the deposition of metabolically active tissue are being fuelled by body fat stores and may therefore ultimately be subject to substrate limitation. LBM is not confined to skeletal muscle and the clinically obvious alterations in its distribution may reflect similar changes in the visceral organs. Therefore asssessment of these changes should be included in longterm monitoring. The effects at cellular level, where rhGH influences cell hyperplasia in particular,19 will be considerably more difficult to monitor reliably in vivo. When measured with care REE is a well-validated, easily reproducible value, from which basic energy needs can be derived and compared for individuals and populations. There is a dearth of data on energy expenditure for children because of technical reasons and difficulties with patient cooperation. Our methods seem to have overcome these difficulties and the results are consistent with the limited data available. They confirm that in childhood, as in
1334
adults,2° REE in kj/24 h increases in proportion to LBM. However, considerable variation remained when REE was expressed per unit LBM and the data show that one of the principal determinants of REE in kj/kg LBM per 24 h during childhood is height. Short children expend more energy fulfilling their basic metabolic requirements than do tall children. The metabolic basis of this relation between REE and LBM is not clear. Because all the studies were done in a thermoneutral environment, the relation is unlikely to be accounted for by an increase in thermogenesis in the shorter children as a function of an increased surface area to volume ratio. Thermogenesis of itself is not a determinant of REE, and the range of surface area within the group is not sufficient to account for the observed two-fold drop in REE kj/kg LBM per 24 h. A more likely explanation for the difference is a change in body composition and proportions with increasing height. Linear growth is associated with a disproportionate increase in muscle relative to visceral tissue. As visceral tissue contributes about 80% to REE,21 shorter children, with their smaller muscle mass, would have a higher visceral to skeletal muscle ratio and hence a higher REE kj/kg LBM per 24 h. There is clearly a need for more detailed work relating body composition, including visceral to skeletal tissue ratios, to metabolic demands. Within the individual, treatment with rhGH increases REE (in kj/24 h), which seems to be due to the associated rise in LBM, a finding that again emphasises the close relation between these two factors. There is as yet no evidence that rhGH has any other specific effect upon REE. Our findings provide insights into the mechanisms of linear growth. Only when these are understood more fully can rational decisions be made about appropriate modes of intervention to improve growth performance. Meanwhile the profound changes observed in body composition, together with energy needs that seem to be greater than those of tall children, suggest that treatment of short but otherwise healthy children with rhGH may be a "stress". We have, as yet, no evidence that this stress is adverse or that the changes in body composition are not reversible with time. Thus the trial continues. J. M. W. and L. D. V. are supported by Kabi-Pharmacia UK Ltd. The Auto Injector devices were donated by Owen Mumford Ltd, Woodstock, UK. REFERENCES 1. Albertsson-Wikland K, Rosberg S. Analyses of 24-hour growth hormone profiles in children: relation to growth. J Clin Endocrinol Metab 1988; 67: 493-500. 2. Hindmarsh PC, Pringle PJ, Di Silvio L, Brook CGD. Effects of 3 years of growth hormone therapy in short normal children. Acta Paediatr Scand 1990; 366 (suppl): 6-12. 3. Ackland FM, Jones J, Buckler JMH, Dunger DB, Rayner PHW, Preece MA. Growth hormone treatment in non-growth hormone-deficient children: effects of stopping treatment. Acta Paediatr Scand 1990; 366 (suppl): 32-37. 4. Cowell CT. Effects of growth homrone in short, slowly growing children without growth hormone deficiency. Acta Paediatr Scand 1990; 366 (suppl): 29-30. 5. Bougnères PF. High-dose growth hormone treatment of non-growth hormone-deficient children: preliminary results after 2 years. Acta Paediatr Scand 1990; 366 (suppl): 38-40. 6. Tanner JM, Whitehouse RH. The effect of human growth hormone on subcutaneous fat thickness in hyposomatotrophic and panhypopituitary dwarfs. J Endocrinol 1967; 39: 263-75. 7. Salomon F, Cuneo RC, Hesp R, Sönksen PH. The effect of treatment with recombinant human growth hormone on body composition and metabolism in adults with growth hormone deficiency. N Engl J Med 1989; 321: 1787-90.
Bengtsson B-Å, Brummer R-J, Edén S, Bosaeus I. Body composition in acromegaly. Clin Endocrinol 1989; 307 121-30. 9. Goodman HM, Schwartz J. Growth hormone and lipid metabolism. In: Knobil E, Sawyer WH, eds. Handbook of physiology, vol 4, part 2. Washington DC: American Physiology Society, 1974; 211-32. 10. Cheek DB, Hill DE. Effect of growth hormone on cell and somatic growth. In: Knobil E, Sawyer WH, eds. Handbook of physiology, vol 4, part 2. Washington DC: American Physiology Society, 1974; 8.
159-85. 11. Tanner JM, Whitehouse RH. Clinical longitudinal standards for height, weight, height velocity, weight velocity and stages of puberty. Arch Dis Child 1976; 51: 170-79. 12. Voss L, Walker J, Lunt H, Wilkin T, Betts P. The Wessex Growth Study: first report. Acta Paediatr Scand 1989; 349 (suppl): 65-72. 13. Brook CGD. Determination of body composition of children from skinfold measurements. Arch Dis Child 1971; 46: 182-84. 14. Siri WE. Body composition from fluid spaces and density. MS UCRL3349. Los Angeles: Dormer Laboratory, University of
California. 15.
Jécquier E, Felber J-P. Indirect calorimetry. Clin Endocrinol 1987; 1:
911-35. 16. Voss LD, Bailey BRJ, Cumming K, Wilkin T, Betts PR. The reliability of height measurement. Arch Dis Child 1990 (in press). 17. Schofield WN. Predicting basal metabolic rate, new standards and review of previous work. Human Nutr: Clin Nutr 1985; 39C (suppl 1): 5-41. 18. Hindmarsh PC, Brook CGD. Effect of growth hormone in short normal children. Br Med J 1987; 295: 573-77. 19. Isaksson OGP, Lindahl A, Nilsson A, Isgaard J. Mechanism of the stimulatory effect of growth hormone on longitudinal bone growth. Endocr Reviews 1987; 8: 426-38. 20. Cunningham JJ. A reanalysis of the factors influencing basal metabolic rate in normal adults. Am J Clin Nutr 1980; 33: 2372-74. 21. Bray GA, Atkinson RL. Factors affecting basal metabolic rate. Prog Fed Nutr Sci 1977: 2: 395-403.
From The Lancet Private
or
public water?
The report of the inquiry by a Committee of the Corporation of the City of London as to the water-supply to London will be read with interest alike by water consumers and by the several companies’ shareholders. As the result of the inquiry the committee who have conducted it have recommended the constitution of a special authority to purchase and take over the undertakings of the water companies, and the committee believe that this can be effected to the financial benefit of the consumers. The value of the stock of the eight companies supplying London is estimated by Mr Wood to be 33 444 506. The assumed net profits of the London companies available for dividend-ie, after providing for interest on the preference and loan capitals-amounted in the year 1889-90, according to Mr Stoneham, to 939 806. The committee are advised by Mr Scott, the City Chamberlain, that the capital required can be borrowed at 3 per cent, and thus the interest would amount to L831 794, leaving a balance of 108 012 available for sinking fund or other purposes. Thus there would be an immediate gain to the water consumers, and if the water supplies were above suspicion and the plant of the companies in good order, there can to be no doubt as to the desirability of the purchase of the undertakings of the companies on this basis.... To an extent the water companies are masters of the situation. Unless the question be settled during the next session, the effect on the quinquennial assessment will be to largely increase the compensation to be paid to them. The bargain which may be made to-day will be impossible in a short time, and it remains to be seen whether the present generation will have sufficient interest in posterity to take this burden upon themselves. The demand to be supplied by meter raises, however, other large questions which are not easy of settlement. Now, rich London pays for poor London, water being regarded as a sanitary necessity which, even in the interests of the rich, should be freely available for the poor. If the receipts of the proposed water authority are to be collected on another basis than that of the present, it is not at all evident that the fmancial prospect the Committee puts forward would be realised. ...
(Oct 25, 1890)
1331
MEDICAL SCIENCE
Treatment of short normal children with hormone—a cautionary tale?
41 short normal children were randomly allocated either to daily injections of growth hormone (rhGH) at 30 lU/m2 per week or to no treatment. 6 months of rhGH therapy produced up to 76% loss of fat mass and up to 25% increase in lean body mass (LBM). These changes were significantly different from those in the untreated group. LBM was the main determinant of resting energy expenditure (REE) expressed as kJ/24 h. REE expressed as kJ/kg LBM per 24 h correlated negatively with height, which was responsible for 66% of the variance in REE kJ/kg LBM per 24 h. Short children therefore expend more energy than tall childen in fulfilling basic metabolic needs. After 6 months REE kJ/24 h increased significantly in treated children. However, treated children did not differ significantly from untreated children in REE kJ/kg LBM per 24 h. rhGH does not therefore seem to have a specific effect upon REE. The possibility that rhGH produces profound metabolic effects should limit its use in otherwise healthy children until the mechanism of action is more clearly elucidated. 1990;
growth
has been the collection of growth data.2-5 However, the clinical features of GHD in both adults and children classically include relative obesity, which disappears with hormone replacement,6,7 whereas acromegalic patients have much less body fat than do normal individuals.8 In view of its known lipolytic9 and anabolic effects10 rhGH could induce substantial and possibly stressful metabolic changes in short but otherwise healthy children with normal body
composition. The Wessex Growth Study is a longitudinal, prospective, community-based study of the growth patterns of all children in two neighbouring health districts found during a 2 year period to be on or below Tanner’s 3rd height centile" at school entry .12 Most previous work has been based on 15 IU 1m2 per week,24 the standard replacement dose for growth hormone deficiency. Short children without growth hormone deficiency probably need a greater dose than this,22 and up to 40 mg IU/m2 per week has been used.5 In the Wessex study a randomised cohort of children is being treated with 30 IU/m2 per week. Anthropometric, metabolic, and psychometric variables are being monitored. Here we report the baseline data collected during the first year of the rhGH trial, with emphasis on body composition and resting energy expenditure (REE) because these may be useful markers for metabolic and nutritional effects.
Introduction The
of recombinant
growth hormone (rhGH) as a encouraging growth in short children without classical growth hormone deficiency (GHD) is being investigated. These children seem to have normal patterns of endogenous growth hormone secretion, albeit at lower levels than do tall children.1 The emphasis in these studies use
means
of
ADDRESS: Departments of Paediatrics, Human Nutrition and Medicine 2 (Endocrinology), Southampton General Hospital and Southampton University, Southampton, UK (J. M. Walker, MRCP, S A. Bond, BSc, L D Voss, MA, P. R. Betts, FRCP, S. A. Wootton, PhD, A. A. Jackson, FRCP). Correspondence to Dr J. M. Walker, University Department of Paediatrics, Level 8, Addenbrooke’s Hospital, Cambridge CB2 2QQ, UK.
1332
Patients and methods Local and Committee
approval and study.
Safety of Medicines ethical committee parental consent were obtained for this
on
informed
Study population 41 children smaller than or equal to -20 standard deviation (SDS) but without GHD or other known disorders were randomly allocated to either treatment with rhGH (’Genotropin’ Kabi-Pharmacia, Milton Keynes, UK) at 30 IU/m2 per week scores
(n = 21) or to no treatment (n = 20). The rhGH was given by daily subcutaneous injection with an ’Auto Injector’ (Owen Mumford Ltd, Woodstock, Oxon, England). The children were aged 69-83 yr at randomisation, and the sex distribution was the same in the two groups.
Growth Each of the 41 study children has a control matched for age, sex, and school class. All 82 children are being measured every 6 months by one trained anthropometrist with a portable electronic stadiometer (Holtain Ltd, Crymych, Wales).
Body composition Body composition, as assessed by anthropometry, was recorded randomisation and every 6 months. Skinfold was measured by one observer in the standard manner with Harpenden skinfold calipers at the biceps, triceps, subscapular, and suprailiac sites in each child. Body density13 and body fat14 were calculated as previously reported. Lean body mass (LBM) was calculated by subtracting the derived fat mass from body weight. To avoid bias, measurements were made without access to previous data. at
of energy requirements during childhood REE was also measured in 20 healthy local schoolchildren whose age-range was the same as that among the short children and whose heights were between the 3rd and 97th Tanner centiles.11 REE was measured by indirect calorimetry with an open-circuit ventilated hood system. All studies were done between 0730 and 1000 h after an overnight fast and the room temperature throughout was maintained at 22-24°C. To induce the resting state the child was given a story or music to listen to through personal stereo headphones whilst supine in bed. Once a steady reading was obtained, recordings were made for at least 30 min. All the children were cooperative and able to complete the study. Air was drawn through a ventilated hood (volume 50 litres) at a flow rate that was adjusted to maintain at 0,5-0,65% the concentration of CO2 in air leaving the hood; the flow rate was usually 30-40 litres/min. The flow rate was measured by use of a vortex-shedding flowmeter and the temperature of the expired air was continuously measured with a thermistor probe at the point of gas sampling. Inspired, expired, and calibration gases were continuously sampled and analysed for 02, CO2, and N2 by a respiratory quadropole mass spectrometer (V. G. SpectraLab, Middlewich, England), which was calibrated before and after each test. Gas concentration, flow, and temperature were processed by microcomputer (Apple IIe) to give O2 consumption and CO2 production corrected to standard temperature and pressure. REE was determined from O2 consumption and respiratory exchange ratio according to Jécquier15 and expressed as kj/24 h and as kj/kg LBM per 24 h.
Statistical analysis Least squares regression analysis, Student’s t test, rank sum tests, and multivariate regression analysis were calculated by use of the
’Statsgraphics’ (STSC Inc, Rockville, Maryland, USA) package.
Results
Resting energy expenditure (REE) REE was determined in 10 treated and 9 untreated children before randomisation and every six months. As part of a wider study
Growth and body composition data for the first 6 months available for 18/21 treated and 19/20 untreated children. In the treated group 2 children have withdrawn from the study (1 because of dislike of injections, and 1 because of failure to give injections and attend for follow up), and 1 was known to have not complied well with the regimen. 1 untreated child also did not attend for follow-up. These 4 children were excluded from the analysis. were
Growth The untreated and control children continued to grow at a rate appropriate for their height whereas, even allowing for the error of measurement (SD for a single height measurement was 0-25 cm16), those treated with rhGH showed a sharp increase in growth velocity. The mean change in growth velocity SDS (GVSDS) over the first 6 months was from -0-60 SDS to +4-59 SDS (p< 0-001), which was equivalent to a mean increase in growth rate of 4-28 cm/year (p < 0-001), no significant changes in GVSDS occurred in the untreated children (p > 0 1 for both GVSDS and growth rate). 12-month data are available from 13 treated and 15 untreated children, and again the latter show no significant change in GV. Those on rhGH show a sustained increase in GV throughout the first year of, on average, 3 83 cm/year or from - 0 60 SDS to + 4.24 SDS.
Body fat - 10l
Individual children
Fig 1-Percentage change in body mass (lower) at 6 mo.
fat
(upper)
and lean
body
Children treated with rhGH showed a significant loss of fat mass of up to 76% (fig 1), the mean loss being equivalent to 0-68 kg of fat from an initial fat mass of 2-33 kg. In contrast the untreated children, whose mean initial fat mass was 2 41
1333
expressed as kj/24 h. However,
when REE was per unit LBM substantial variation remained.
expressed Stepwise multivariate analysis of these results revealed that height alone contributed 66% of the variance of REE expressed as kj/kg LBM per 24 h, shorter children using more energy per unit of LBM than tall children. Age and weight, widely used predictors of REE,1’ did not account significantly for a greater proportion of the variance. Height seems therefore to be the principal variable contributing to REE kj/kg LBM per 24 h. Figure 2 shows this relation and for clarity removes the non-significant influence of age by using height SDS as the index for height, an SDS 0 being the 50th centile for height." =
r.rvT
-4
-3
-2
-1
0
1
2
3
4
Height SDS
REE and treatment with rhGH
Fig 2-Relation between resting energy expenditure expressed as kg/kg LBM per 24 h and height SDS.
After 6 months REE (kj/24 h) increased significantly in the treated children by, on average, 480 kJ/24 h or 12% of the pre-randomisation value (fig 3). In the untreated group it did not alter significantly. However, when expressed per unit LBM the increase in REE with treatment disappeared, and after 6 months there was no significant difference in REE kj/kg LBM per 24 h between the two groups (p > 0’ 1).
(p < 0 001)
an average of 0-34 kg (p < 0-001 for difference between treated and untreated group). The mother of 1 girl was concerned because her daughter became visibly fatter and she has since entered puberty. 2 others had a major change in social circumstances during this time, and 1 of these also had a marked rise in his growth velocity of 3-3
kg, gained
cm/year. Lean body mass
Discussion
(LBM)
This
LBM increased in both groups (fig 1, lower), but more so (p < 0-001) in the treated group, which had an increase of up to 25%. The mean gain was 2-81 kg in the treated and 0-8 kg in the untreated group. 12-month data are available for 13 treated and 15 untreated children. In most children the changes in body fat and LBM persisted, and in some, including those who had shown only slight changes at 6 months despite good compliance, they were greater at 1 year than at 6 months.
Resting energy expenditure (REE) before treatment A correlation of 0-8 (p < 0-001) was found between REE expressed as kj/24 h and LBM in normal children. Stepwise multivariate analysis revealed that LBM was the main determinant of REE when expressed in absolute units (kj/24 h) and accounted for 75 % of the variance. Addition of age, height, and weight into the model did not significantly improve the relation and when brought into the equation could account only for a further 1 % of the variance in REE
- 200
Fig
L
kJ/24h
3-Mean change in REE at 6 months when expressed or as kJ/kg LBM per 24 h.
kJ/24 h
as
study has shown that in short but otherwise normal children, the striking increase in growth velocity produced by high doses of rhGH is accompanied a profound alteration in body composition characterised by a loss of body fat of up to 76% and a gain in LBM of up to 25%. Despite the known lipolytic and anabolic properties of growth hormone, the effect of rhGH on body composition in short normal children has not been extensively reported. Hindmarsh and Brook, who examined only triceps and subscapular skinfolds, reported significant decreases after 6 months in children treated with doses of 12-20 IU 1m2 per week, with return to baseline levels at 12 months.i8 Even studies using higher doses of rhGH than we did here make no reference to body composition,s which seems surprising in view of the striking clinical changes. The children became skinny, with obvious loss of adipose tissue from all areas of the body but especially from the limbs and face. The girls in particular looked innappropriately muscular. The increase in growth velocity after 6 months was greater than seen in other similar studies and is presumably dose related,3,4 though it may be too soon to be certain of this difference. Taken together, these fmdings give the impression that the increase in growth velocity and the deposition of metabolically active tissue are being fuelled by body fat stores and may therefore ultimately be subject to substrate limitation. LBM is not confined to skeletal muscle and the clinically obvious alterations in its distribution may reflect similar changes in the visceral organs. Therefore asssessment of these changes should be included in longterm monitoring. The effects at cellular level, where rhGH influences cell hyperplasia in particular,19 will be considerably more difficult to monitor reliably in vivo. When measured with care REE is a well-validated, easily reproducible value, from which basic energy needs can be derived and compared for individuals and populations. There is a dearth of data on energy expenditure for children because of technical reasons and difficulties with patient cooperation. Our methods seem to have overcome these difficulties and the results are consistent with the limited data available. They confirm that in childhood, as in
1334
adults,2° REE in kj/24 h increases in proportion to LBM. However, considerable variation remained when REE was expressed per unit LBM and the data show that one of the principal determinants of REE in kj/kg LBM per 24 h during childhood is height. Short children expend more energy fulfilling their basic metabolic requirements than do tall children. The metabolic basis of this relation between REE and LBM is not clear. Because all the studies were done in a thermoneutral environment, the relation is unlikely to be accounted for by an increase in thermogenesis in the shorter children as a function of an increased surface area to volume ratio. Thermogenesis of itself is not a determinant of REE, and the range of surface area within the group is not sufficient to account for the observed two-fold drop in REE kj/kg LBM per 24 h. A more likely explanation for the difference is a change in body composition and proportions with increasing height. Linear growth is associated with a disproportionate increase in muscle relative to visceral tissue. As visceral tissue contributes about 80% to REE,21 shorter children, with their smaller muscle mass, would have a higher visceral to skeletal muscle ratio and hence a higher REE kj/kg LBM per 24 h. There is clearly a need for more detailed work relating body composition, including visceral to skeletal tissue ratios, to metabolic demands. Within the individual, treatment with rhGH increases REE (in kj/24 h), which seems to be due to the associated rise in LBM, a finding that again emphasises the close relation between these two factors. There is as yet no evidence that rhGH has any other specific effect upon REE. Our findings provide insights into the mechanisms of linear growth. Only when these are understood more fully can rational decisions be made about appropriate modes of intervention to improve growth performance. Meanwhile the profound changes observed in body composition, together with energy needs that seem to be greater than those of tall children, suggest that treatment of short but otherwise healthy children with rhGH may be a "stress". We have, as yet, no evidence that this stress is adverse or that the changes in body composition are not reversible with time. Thus the trial continues. J. M. W. and L. D. V. are supported by Kabi-Pharmacia UK Ltd. The Auto Injector devices were donated by Owen Mumford Ltd, Woodstock, UK. REFERENCES 1. Albertsson-Wikland K, Rosberg S. Analyses of 24-hour growth hormone profiles in children: relation to growth. J Clin Endocrinol Metab 1988; 67: 493-500. 2. Hindmarsh PC, Pringle PJ, Di Silvio L, Brook CGD. Effects of 3 years of growth hormone therapy in short normal children. Acta Paediatr Scand 1990; 366 (suppl): 6-12. 3. Ackland FM, Jones J, Buckler JMH, Dunger DB, Rayner PHW, Preece MA. Growth hormone treatment in non-growth hormone-deficient children: effects of stopping treatment. Acta Paediatr Scand 1990; 366 (suppl): 32-37. 4. Cowell CT. Effects of growth homrone in short, slowly growing children without growth hormone deficiency. Acta Paediatr Scand 1990; 366 (suppl): 29-30. 5. Bougnères PF. High-dose growth hormone treatment of non-growth hormone-deficient children: preliminary results after 2 years. Acta Paediatr Scand 1990; 366 (suppl): 38-40. 6. Tanner JM, Whitehouse RH. The effect of human growth hormone on subcutaneous fat thickness in hyposomatotrophic and panhypopituitary dwarfs. J Endocrinol 1967; 39: 263-75. 7. Salomon F, Cuneo RC, Hesp R, Sönksen PH. The effect of treatment with recombinant human growth hormone on body composition and metabolism in adults with growth hormone deficiency. N Engl J Med 1989; 321: 1787-90.
Bengtsson B-Å, Brummer R-J, Edén S, Bosaeus I. Body composition in acromegaly. Clin Endocrinol 1989; 307 121-30. 9. Goodman HM, Schwartz J. Growth hormone and lipid metabolism. In: Knobil E, Sawyer WH, eds. Handbook of physiology, vol 4, part 2. Washington DC: American Physiology Society, 1974; 211-32. 10. Cheek DB, Hill DE. Effect of growth hormone on cell and somatic growth. In: Knobil E, Sawyer WH, eds. Handbook of physiology, vol 4, part 2. Washington DC: American Physiology Society, 1974; 8.
159-85. 11. Tanner JM, Whitehouse RH. Clinical longitudinal standards for height, weight, height velocity, weight velocity and stages of puberty. Arch Dis Child 1976; 51: 170-79. 12. Voss L, Walker J, Lunt H, Wilkin T, Betts P. The Wessex Growth Study: first report. Acta Paediatr Scand 1989; 349 (suppl): 65-72. 13. Brook CGD. Determination of body composition of children from skinfold measurements. Arch Dis Child 1971; 46: 182-84. 14. Siri WE. Body composition from fluid spaces and density. MS UCRL3349. Los Angeles: Dormer Laboratory, University of
California. 15.
Jécquier E, Felber J-P. Indirect calorimetry. Clin Endocrinol 1987; 1:
911-35. 16. Voss LD, Bailey BRJ, Cumming K, Wilkin T, Betts PR. The reliability of height measurement. Arch Dis Child 1990 (in press). 17. Schofield WN. Predicting basal metabolic rate, new standards and review of previous work. Human Nutr: Clin Nutr 1985; 39C (suppl 1): 5-41. 18. Hindmarsh PC, Brook CGD. Effect of growth hormone in short normal children. Br Med J 1987; 295: 573-77. 19. Isaksson OGP, Lindahl A, Nilsson A, Isgaard J. Mechanism of the stimulatory effect of growth hormone on longitudinal bone growth. Endocr Reviews 1987; 8: 426-38. 20. Cunningham JJ. A reanalysis of the factors influencing basal metabolic rate in normal adults. Am J Clin Nutr 1980; 33: 2372-74. 21. Bray GA, Atkinson RL. Factors affecting basal metabolic rate. Prog Fed Nutr Sci 1977: 2: 395-403.
From The Lancet Private
or
public water?
The report of the inquiry by a Committee of the Corporation of the City of London as to the water-supply to London will be read with interest alike by water consumers and by the several companies’ shareholders. As the result of the inquiry the committee who have conducted it have recommended the constitution of a special authority to purchase and take over the undertakings of the water companies, and the committee believe that this can be effected to the financial benefit of the consumers. The value of the stock of the eight companies supplying London is estimated by Mr Wood to be 33 444 506. The assumed net profits of the London companies available for dividend-ie, after providing for interest on the preference and loan capitals-amounted in the year 1889-90, according to Mr Stoneham, to 939 806. The committee are advised by Mr Scott, the City Chamberlain, that the capital required can be borrowed at 3 per cent, and thus the interest would amount to L831 794, leaving a balance of 108 012 available for sinking fund or other purposes. Thus there would be an immediate gain to the water consumers, and if the water supplies were above suspicion and the plant of the companies in good order, there can to be no doubt as to the desirability of the purchase of the undertakings of the companies on this basis.... To an extent the water companies are masters of the situation. Unless the question be settled during the next session, the effect on the quinquennial assessment will be to largely increase the compensation to be paid to them. The bargain which may be made to-day will be impossible in a short time, and it remains to be seen whether the present generation will have sufficient interest in posterity to take this burden upon themselves. The demand to be supplied by meter raises, however, other large questions which are not easy of settlement. Now, rich London pays for poor London, water being regarded as a sanitary necessity which, even in the interests of the rich, should be freely available for the poor. If the receipts of the proposed water authority are to be collected on another basis than that of the present, it is not at all evident that the fmancial prospect the Committee puts forward would be realised. ...
(Oct 25, 1890)
