- Email: [email protected]
Growth hormone and somatomedin-C in bulimia
Psychoneuraendocrinology,Vol. 13, No. 4, pp. 359-362, 1988
0306-4530/88 $3.00 + 0.00 © 1988 Pergamon Press plc
Printed in Great Britain
SHORT COMMUNICATION GROWTH HORMONE AND SOMATOMEDIN-C IN BULIMIA ALAN B. LEVY 1.and WILLIAM B. MALARKEy2 1Department of Psychiatry and 2Department of Medicine, The Ohio State University, Columbus, Ohio, U.S.A. (Received 15 June 1987; in final form 4 November 1987)
SUMMARY Somatomedin-C (SOM-C) concentrations are regulated by circulating growth hormone (GH) concentrations; however, other factors, such as nutrition, also influence SOM-C concentrations. We evaluated the GH-SOM-C axis in seven normal-weight female bulimics one day after hospital admission, and in seven age-, sex-, and weight-matched normal controls. Subjects were medicationfree for at least one month. Fasting morning serum GH concentrations were higher in all bulimics (range 2.5-13.3 ng/ml) than in all controls (range <1.0-1.8 ng/ml). The mean (+SD) maximum GH response to TRH (500 Ixg) was greater in the bulimics (12.9+4.9 ng/ml) than in the controls (3.7+2.7 ng/ml) (p<.001). Despite this GH elevation, the mean (+SD) SOM-C concentration was comparable in the bulimics (2.0+0.6 U/ml) and the controls (1.6+0.8 U/ml). This suggests that SOM-C generation is resistant to the elevated circulating GH in bulimia and that SOM-C is not inhibiting GH secretion in the pituitary-hypothalamic axis.
INTRODUCTION SOMATOMEDIN-C (SOM-C) is one of a family o f insulin-like.peptides which mediate the peripheral g r o w t h effects o f h u m a n . g r o w t h h o r m o n e (GH). While circulating S O M - C concentrations are largely dependent on GH secretion (Phillips & Vassilopoulou-Sellin, 1980), other factors, such as nutrition, may be influential. For instance, acute starvation (Clemmons e t al., 1981) and protein-calorie malnutrition (Grant et al., 1973) are associated with low somatomedin activity. Anorexia nervosa (AN), characterized by both starvation and caloric deprivation, also appears to be associated with diminished SOM-C activity, despite elevated circulating GH concentrations (Rappaport et aL, 1980). Normal-weight patients with bulimia have also been found to show high basal circulating GH (Kiriike et al., 1987), though they are neither starved nor calorie-deprived. Whether patients with bulimia have a SOM-C profile similar to A N has yet to be investigated. METHODS We studied seven normal weight females with bulimia, diagnosed by DSM-ITI criteria (American Psychiatric Association, 1980), on the day after hospital admission. All also met DSM-m-R criteria for bulimia (American Correspondence to be addressed to: Dr. Alan B. Levy, Department of Psychiatry, Ohio State University, 473 West 12th Avenue, Columbus OH 43210, USA. 359
360
A. B. LEVYand W. B. MALARKEY
Psychiatric Association, 1987). They were compared to seven age, sex, and weight-matched normal controls without a history of psychiatric illness as determined by psychiatric interview. Neither subjects nor controls had a history of endocrinopathy, and all had been medication-free, including oral contraceptives, for at least one month. Subjects and controls had fasting blood samples taken at 1000 h for basal GH, estradiol, and SOM-C determinations, followed by the infusion of Protirelin (rRH) (500 lig). Blood samples then were taken every 15 rain for 90 min, and were centrifuged within 60 min of clotting. Estradiol and GH were assayed immediately, while the samples for SOM-C assay were immediately frozen at -20* C. The GH and estradiol radioimmunoassays (RIA) had intra- and interassay coefficients of variation (CV) which were less than 10%, determined with control plasma samples which fell in the low, middle, and high regions of the standard curve. The SOM-C RIA was performed in a single assay by Nichols Institute (Los Angeles, CA), with serum from Ortho Diagnostics (Raritan, NJ) as standard. The intra- and interassay CV's were 6.4% and 10.3% respectively. The basal GH and SOM-C values were not normally distributed: The Wilcoxon rank sum test was used to test for differences in basal GH. A log transformation was performed followed by Student's two-tailed t test to test for differences in SOM-C. Differences in the maximum TRH-stimulated GH response and basal estradiol were tested by Student's two-tailed t test. Pearson correlations were determined among SOM-C, GH, and estradiol concentrations across subjects. RESULTS The age (mean + S D ) o f our bulimics (27.3+7.4 years) was similar to that o f the controls (22.9+2.9 years). Mean ( + S D ) ideal body weight (%IBW) was also similar between bulimics (95.6+ 11.3%) and controls (99.3+5.6%). The mean ( + S D ) duration o f illness (83.1+85.4 months) and weekly frequency of bingeing (20.0+ 24.4) indicates that a moderately severely ill group o f bulimics were studied. As expected, the mean ( + S D ) Eating Attitudes Test fEAT) (Garner & Garfinkel, 1979) score was higher for the bulimics (57.4+20.5) than for the controls (9.3+5.4) ( p <.001). The basal G H concentration of all the bulimic subjects (range 2.5-13.3 ng/ml) was greater than the highest basal G H level of the control subjects (range < 1 . 0 - 1 . 8 ng/ml) (p < .01, Wilcoxon rank sum test). Moreover, the bulimics had a greater m a x i m u m mean (+ SD) G H response to TRH (12.9+4.9 ng/ml) than that of the controls (3.7+2.7 ng/ml) (p<.001). Both estradiol and S O M - C concentrations were s i m i l a r in the b u l i m i c s and the controls (Table I). Pearson correlations failed to demonstrate any relationship between S O M - C and basal GH, maximum GH, or estradiol levels. DISCUSSION In this study, normal-weight women with bulimia were found to have higher basal serum GH and a greater G H response to TRH stimulation than normal female controls. Despite elevated GH concentrations, the serum S O M - C concentrations in the bulimics were similar to those of the controls, suggesting that other factors have a significant inhibitory influence on S O M - C levels in bulimia.
TABLE I. COMPARATIVEHORMONE LEVELS FOR BULIMICS AND CONTROLS.
*p<.01 tp<.001
Subject Group
N
Bulimics Controls
7 7
Growth Hormone (ng/ml) Basal Post-TRH 6.6+4.4* 1.3 + 0.3
12.9+4.9 t 3.7 + 2.7
SOM-C (U/ml)
Estradiol (ng/ml)
2.0+0.6 1.6 + 0.8
82.3+65.4 55.3 + 12.5
SERUM G H AND SOM-C IN BULIMIA
361
GH secretion is probably the most potent stimulator of SOM-C release (Phillips & Vassilopoulou-Sellin, 1980). Somatomedin levels are low in patients with GH deficiency (Hall, 1971) and respond in a dose-dependent fashion to exogenous GH (Phillips & VassilopoulouSellin, 1980). Corticostemids in large doses have been reported to reduce SOM-C concentrations (Elders et al., 1975). While bulimics are known to have elevated adrenal activity (Levy & Dixon, 1987), mildly supraphysiologic cortisol concentrations have been reported not to influence SOM-C (Phillips et al., 1975). Estrogen, thyroid hormones, glucose, and insulin all may affect SOM-C (Almgriest et aL, 1961; Yde, 1964; Audhya & Gibson, 1975; Phillips & VassilopoulouSellin, 1980; Dawson-Hughes et al., 1986); however, the concentrations of these substances are not generally different in bulimics and controls. Thus, these factors were pmbably not influencing SOM-C in our subjects. The diets of the bulimics and controls were not known prior to hospitalization. No subject was fasting prior to testing; the bulimics ate normal hospital meals in the 24 hours prior to testing, the precise composition of which was not recorded. Immtmoreactive SOM-C concentrations are influenced by a variety of factors (Chatelaine t aL, 1983). These include methods of blood collection and storage, with serum samples yielding higher SOM-C concentrations than EDTA plasma samples. Because our determinations were made on serum samples, the values should not be compared with laboratory values based on EDTA plasma. Serum variability was minimized by centrifuging and freezing (-20* C) all samples within 60 min of clotting. Poor nutrition reportedly reduces SOM-C concentrations. Clemmons et al. (1981) demons.tvated a rapid reduction in SOM-C during fasting, which began to normalize in the first three days after refeeding. Grant et al. (1973) found protein-calorie malnourished children to have low somatomedin activity, despite elevated GH concentrations. Similarly, low somatomedin activity despite high basal GH concentrations has been described in low weight patients with AN (Rappaport et al., 1980). Thus, poor nutrition or starvation may inhibit SOM-C activity despite elevated GH concentrations. While this is in part due to the production of SOM-C activity inhibitors measured in the bioassay of somatomedin, actual SOM-C concentrations are known to be reduced as well (Clemmons et al., 1981). The possibility that nutritional factors may prevent elevated SOM-C concentrations despite high GH concentrations in bulimics should be considered. Similar to low-weight anorectics, bulimics tend to demonstrate a delayed TSH response to TRH, elevated basal and TRHstimulated GH, and elevated B-hydroxybutydc acid concentrations (Pirke et al., 1985; Kiriike et al., 1987; Levy et al., 1988). Because these responses generally normalize in refed and weightrecovered anorectics, they may be regarded as metabolic and endocrinologic indicators of malnutrition (Garfinkel et al., 1975; Vigersky et al., 1977). If indeed malnutrition is influencing the GH-SOM-C axis in bulimics, it appears to be acting at two distinct levels, first, it appears that SOM-C is unable to provide appropriate positive feedback on somatostatin at the hypothalamus and/or negative feedback on GH at the pituitary to prevent the GH elevations Berelowitz et al., 1981). Second, the elevated GH appears unable to appropriately stimulate SOM-C production. Further research is necessary to identify the relationship between GH, SOM-C, and nutrition in bulimia.
Acknowledgements: This study was supportedin part by grants from the BremerFoundation,Ohio Department of MentalHealth, and NIH throughGCRC RR-34. The authorsthank Julie Luznyfor manuscriptpreparation. REFERENCES Almgriest S, IRkosD, Luft R (1961) Studieson sulfationfactor (SF) activityof human serum. Acta Endocrinol 37: 138-147.
362
A.B. LEVY and W. B. M A L A R K E Y
American Psychiatric Association (1980) DSM-III : Diagnostic and Statistical Manual of Mental Disorders, 3rd edition. APA, Washington DC. American Psychiatric Association (1987) DSM II1: Diagnostic and Statistical Manual of Mental Disorders, 3rd edition (revised). APA, Washington DC. Audhya TK, Gibson KD (1975) Enhancement of somatomedin titres of normal and hypopituitary scra by addition of L-triiodothyronine in vitro and at physiological concentxations. ProcNatlAcadSci USA 72: 604-608. Berelowitz M, Szabo M, Frohman L, firestone S, Chu L, Hintz RL (1981) Somatomcdin-C mediates growth hormone negative feedback by effects on both hypothalamus and the pituitary. Science 212: 1279-1281. Chatelain PG, Van Wyk JJ, Copeland KC, Blethen SL, Underwood LE (1983) Effect of in vitro action of serum proteascs or exposure to acid on mcasurable immunoreactive somatomedin-C in serum. J Clin Endocrinol Metab 56: 376-383. Clemmons DR, Klibanski R, Underwood LE, McArthur JW, Ridgeway EC, Beitins IZ, Van Wyk JJ (1981) Reduction of plasma immunoreacfivc somatomedin-C during fasting in humans. J Clin Endocrinol Metab 53: 1247-1230. Dawson-Hughes B, Stern D, Goldman J, Rcichlin S (1986) Regulation of growth hormone and somatomcdin-C secretion in post-menopausal women: effect of physiologic estrogen replacement. J Clin Endocrinol Metab 63: 424-432. Elders MI, Wingfield BS, McNatt NIL, Clarke JS, Hughes ER (1975) Glucocorticoid therapy in children: effect on somatomedin secretion. Am JDis Child 129: 1393-1396. Garfinkel PE, Brown GM, Stanccr HC, Moldofsky H (1975) Hypothalamic pituitary function in anorexia nervosa. Arch Gen Psychiatry 32: 739-744. Garner DM, Garfinkel PE (1979) The Eating Attitudes Test: an index of the symptoms of anorexia nervosa. Psychol Med 9: 273-279. Grant DB, Hambley J, Bccker D, Pimstone BL (1973) Reduced sulfation factor in undernourished children. Arch Dis Child 48: 596-600. Hall K (1971) Effect of intravenous administration of human growth hormone on sulfation factor activity in serum of hypopituitary subjects. Acta Endocrino166: 491-497. Kiriike N, Nishiwaki S, Izumiya Y, Macda Y, Kawakita Y (1987) Thyrolxopin, prolactin and growth hormone responses to thyrotropin-relcasing hormone in anorexia nervosa and bulimia. Biol Psychiatry 22: 167-176. Levy AB, Dixon KN (1987) DST in bulimia without endogenous depression. Biol Psychiatry 22: 783-786. Levy AB, Dixon KIN,Malarkey WB (1988) Pituitary response to TRI-Iin bulimia. Biol Psychiatry 23: 476-484. Phillips LS, Hcfington AC, Doughaday WH (1975) Steroid hormone effects on somatomcdin. I. Somatomedin action in vitro. Endocrinology 97: 780-786. Phillips LS, Vassilopoulou-Sellin R (1980) Somatomedins. N Engl J Med 30: 438-446. Pirke KM, Pahl J, Schweigcr V, Warnhoff M (1985) Metabolic and endocrine indices of starvation in bulimia: a comparison with anorexia nervosa. Psychiatry Res 15: 33-39. Rappaport R, Prevot C, Czernichow P (1980) Somatomcdin activity and growth hormone secretion. Acta Pediatr Scand 69: 37-41. Vigersky RA, Anderson AE, Thompson RH, Loriaux DL (1977) Hypothalamic dysfunction in secondary amcnorrhca associated with simple weight loss. N Engl J Med 297:1141-1145. Yde H (1964) The growth-hormone-dependentsulfation factor in serum of untreated diabetes. Lancet ii: 624-626.
0306-4530/88 $3.00 + 0.00 © 1988 Pergamon Press plc
Printed in Great Britain
SHORT COMMUNICATION GROWTH HORMONE AND SOMATOMEDIN-C IN BULIMIA ALAN B. LEVY 1.and WILLIAM B. MALARKEy2 1Department of Psychiatry and 2Department of Medicine, The Ohio State University, Columbus, Ohio, U.S.A. (Received 15 June 1987; in final form 4 November 1987)
SUMMARY Somatomedin-C (SOM-C) concentrations are regulated by circulating growth hormone (GH) concentrations; however, other factors, such as nutrition, also influence SOM-C concentrations. We evaluated the GH-SOM-C axis in seven normal-weight female bulimics one day after hospital admission, and in seven age-, sex-, and weight-matched normal controls. Subjects were medicationfree for at least one month. Fasting morning serum GH concentrations were higher in all bulimics (range 2.5-13.3 ng/ml) than in all controls (range <1.0-1.8 ng/ml). The mean (+SD) maximum GH response to TRH (500 Ixg) was greater in the bulimics (12.9+4.9 ng/ml) than in the controls (3.7+2.7 ng/ml) (p<.001). Despite this GH elevation, the mean (+SD) SOM-C concentration was comparable in the bulimics (2.0+0.6 U/ml) and the controls (1.6+0.8 U/ml). This suggests that SOM-C generation is resistant to the elevated circulating GH in bulimia and that SOM-C is not inhibiting GH secretion in the pituitary-hypothalamic axis.
INTRODUCTION SOMATOMEDIN-C (SOM-C) is one of a family o f insulin-like.peptides which mediate the peripheral g r o w t h effects o f h u m a n . g r o w t h h o r m o n e (GH). While circulating S O M - C concentrations are largely dependent on GH secretion (Phillips & Vassilopoulou-Sellin, 1980), other factors, such as nutrition, may be influential. For instance, acute starvation (Clemmons e t al., 1981) and protein-calorie malnutrition (Grant et al., 1973) are associated with low somatomedin activity. Anorexia nervosa (AN), characterized by both starvation and caloric deprivation, also appears to be associated with diminished SOM-C activity, despite elevated circulating GH concentrations (Rappaport et aL, 1980). Normal-weight patients with bulimia have also been found to show high basal circulating GH (Kiriike et al., 1987), though they are neither starved nor calorie-deprived. Whether patients with bulimia have a SOM-C profile similar to A N has yet to be investigated. METHODS We studied seven normal weight females with bulimia, diagnosed by DSM-ITI criteria (American Psychiatric Association, 1980), on the day after hospital admission. All also met DSM-m-R criteria for bulimia (American Correspondence to be addressed to: Dr. Alan B. Levy, Department of Psychiatry, Ohio State University, 473 West 12th Avenue, Columbus OH 43210, USA. 359
360
A. B. LEVYand W. B. MALARKEY
Psychiatric Association, 1987). They were compared to seven age, sex, and weight-matched normal controls without a history of psychiatric illness as determined by psychiatric interview. Neither subjects nor controls had a history of endocrinopathy, and all had been medication-free, including oral contraceptives, for at least one month. Subjects and controls had fasting blood samples taken at 1000 h for basal GH, estradiol, and SOM-C determinations, followed by the infusion of Protirelin (rRH) (500 lig). Blood samples then were taken every 15 rain for 90 min, and were centrifuged within 60 min of clotting. Estradiol and GH were assayed immediately, while the samples for SOM-C assay were immediately frozen at -20* C. The GH and estradiol radioimmunoassays (RIA) had intra- and interassay coefficients of variation (CV) which were less than 10%, determined with control plasma samples which fell in the low, middle, and high regions of the standard curve. The SOM-C RIA was performed in a single assay by Nichols Institute (Los Angeles, CA), with serum from Ortho Diagnostics (Raritan, NJ) as standard. The intra- and interassay CV's were 6.4% and 10.3% respectively. The basal GH and SOM-C values were not normally distributed: The Wilcoxon rank sum test was used to test for differences in basal GH. A log transformation was performed followed by Student's two-tailed t test to test for differences in SOM-C. Differences in the maximum TRH-stimulated GH response and basal estradiol were tested by Student's two-tailed t test. Pearson correlations were determined among SOM-C, GH, and estradiol concentrations across subjects. RESULTS The age (mean + S D ) o f our bulimics (27.3+7.4 years) was similar to that o f the controls (22.9+2.9 years). Mean ( + S D ) ideal body weight (%IBW) was also similar between bulimics (95.6+ 11.3%) and controls (99.3+5.6%). The mean ( + S D ) duration o f illness (83.1+85.4 months) and weekly frequency of bingeing (20.0+ 24.4) indicates that a moderately severely ill group o f bulimics were studied. As expected, the mean ( + S D ) Eating Attitudes Test fEAT) (Garner & Garfinkel, 1979) score was higher for the bulimics (57.4+20.5) than for the controls (9.3+5.4) ( p <.001). The basal G H concentration of all the bulimic subjects (range 2.5-13.3 ng/ml) was greater than the highest basal G H level of the control subjects (range < 1 . 0 - 1 . 8 ng/ml) (p < .01, Wilcoxon rank sum test). Moreover, the bulimics had a greater m a x i m u m mean (+ SD) G H response to TRH (12.9+4.9 ng/ml) than that of the controls (3.7+2.7 ng/ml) (p<.001). Both estradiol and S O M - C concentrations were s i m i l a r in the b u l i m i c s and the controls (Table I). Pearson correlations failed to demonstrate any relationship between S O M - C and basal GH, maximum GH, or estradiol levels. DISCUSSION In this study, normal-weight women with bulimia were found to have higher basal serum GH and a greater G H response to TRH stimulation than normal female controls. Despite elevated GH concentrations, the serum S O M - C concentrations in the bulimics were similar to those of the controls, suggesting that other factors have a significant inhibitory influence on S O M - C levels in bulimia.
TABLE I. COMPARATIVEHORMONE LEVELS FOR BULIMICS AND CONTROLS.
*p<.01 tp<.001
Subject Group
N
Bulimics Controls
7 7
Growth Hormone (ng/ml) Basal Post-TRH 6.6+4.4* 1.3 + 0.3
12.9+4.9 t 3.7 + 2.7
SOM-C (U/ml)
Estradiol (ng/ml)
2.0+0.6 1.6 + 0.8
82.3+65.4 55.3 + 12.5
SERUM G H AND SOM-C IN BULIMIA
361
GH secretion is probably the most potent stimulator of SOM-C release (Phillips & Vassilopoulou-Sellin, 1980). Somatomedin levels are low in patients with GH deficiency (Hall, 1971) and respond in a dose-dependent fashion to exogenous GH (Phillips & VassilopoulouSellin, 1980). Corticostemids in large doses have been reported to reduce SOM-C concentrations (Elders et al., 1975). While bulimics are known to have elevated adrenal activity (Levy & Dixon, 1987), mildly supraphysiologic cortisol concentrations have been reported not to influence SOM-C (Phillips et al., 1975). Estrogen, thyroid hormones, glucose, and insulin all may affect SOM-C (Almgriest et aL, 1961; Yde, 1964; Audhya & Gibson, 1975; Phillips & VassilopoulouSellin, 1980; Dawson-Hughes et al., 1986); however, the concentrations of these substances are not generally different in bulimics and controls. Thus, these factors were pmbably not influencing SOM-C in our subjects. The diets of the bulimics and controls were not known prior to hospitalization. No subject was fasting prior to testing; the bulimics ate normal hospital meals in the 24 hours prior to testing, the precise composition of which was not recorded. Immtmoreactive SOM-C concentrations are influenced by a variety of factors (Chatelaine t aL, 1983). These include methods of blood collection and storage, with serum samples yielding higher SOM-C concentrations than EDTA plasma samples. Because our determinations were made on serum samples, the values should not be compared with laboratory values based on EDTA plasma. Serum variability was minimized by centrifuging and freezing (-20* C) all samples within 60 min of clotting. Poor nutrition reportedly reduces SOM-C concentrations. Clemmons et al. (1981) demons.tvated a rapid reduction in SOM-C during fasting, which began to normalize in the first three days after refeeding. Grant et al. (1973) found protein-calorie malnourished children to have low somatomedin activity, despite elevated GH concentrations. Similarly, low somatomedin activity despite high basal GH concentrations has been described in low weight patients with AN (Rappaport et al., 1980). Thus, poor nutrition or starvation may inhibit SOM-C activity despite elevated GH concentrations. While this is in part due to the production of SOM-C activity inhibitors measured in the bioassay of somatomedin, actual SOM-C concentrations are known to be reduced as well (Clemmons et al., 1981). The possibility that nutritional factors may prevent elevated SOM-C concentrations despite high GH concentrations in bulimics should be considered. Similar to low-weight anorectics, bulimics tend to demonstrate a delayed TSH response to TRH, elevated basal and TRHstimulated GH, and elevated B-hydroxybutydc acid concentrations (Pirke et al., 1985; Kiriike et al., 1987; Levy et al., 1988). Because these responses generally normalize in refed and weightrecovered anorectics, they may be regarded as metabolic and endocrinologic indicators of malnutrition (Garfinkel et al., 1975; Vigersky et al., 1977). If indeed malnutrition is influencing the GH-SOM-C axis in bulimics, it appears to be acting at two distinct levels, first, it appears that SOM-C is unable to provide appropriate positive feedback on somatostatin at the hypothalamus and/or negative feedback on GH at the pituitary to prevent the GH elevations Berelowitz et al., 1981). Second, the elevated GH appears unable to appropriately stimulate SOM-C production. Further research is necessary to identify the relationship between GH, SOM-C, and nutrition in bulimia.
Acknowledgements: This study was supportedin part by grants from the BremerFoundation,Ohio Department of MentalHealth, and NIH throughGCRC RR-34. The authorsthank Julie Luznyfor manuscriptpreparation. REFERENCES Almgriest S, IRkosD, Luft R (1961) Studieson sulfationfactor (SF) activityof human serum. Acta Endocrinol 37: 138-147.
362
A.B. LEVY and W. B. M A L A R K E Y
American Psychiatric Association (1980) DSM-III : Diagnostic and Statistical Manual of Mental Disorders, 3rd edition. APA, Washington DC. American Psychiatric Association (1987) DSM II1: Diagnostic and Statistical Manual of Mental Disorders, 3rd edition (revised). APA, Washington DC. Audhya TK, Gibson KD (1975) Enhancement of somatomedin titres of normal and hypopituitary scra by addition of L-triiodothyronine in vitro and at physiological concentxations. ProcNatlAcadSci USA 72: 604-608. Berelowitz M, Szabo M, Frohman L, firestone S, Chu L, Hintz RL (1981) Somatomcdin-C mediates growth hormone negative feedback by effects on both hypothalamus and the pituitary. Science 212: 1279-1281. Chatelain PG, Van Wyk JJ, Copeland KC, Blethen SL, Underwood LE (1983) Effect of in vitro action of serum proteascs or exposure to acid on mcasurable immunoreactive somatomedin-C in serum. J Clin Endocrinol Metab 56: 376-383. Clemmons DR, Klibanski R, Underwood LE, McArthur JW, Ridgeway EC, Beitins IZ, Van Wyk JJ (1981) Reduction of plasma immunoreacfivc somatomedin-C during fasting in humans. J Clin Endocrinol Metab 53: 1247-1230. Dawson-Hughes B, Stern D, Goldman J, Rcichlin S (1986) Regulation of growth hormone and somatomcdin-C secretion in post-menopausal women: effect of physiologic estrogen replacement. J Clin Endocrinol Metab 63: 424-432. Elders MI, Wingfield BS, McNatt NIL, Clarke JS, Hughes ER (1975) Glucocorticoid therapy in children: effect on somatomedin secretion. Am JDis Child 129: 1393-1396. Garfinkel PE, Brown GM, Stanccr HC, Moldofsky H (1975) Hypothalamic pituitary function in anorexia nervosa. Arch Gen Psychiatry 32: 739-744. Garner DM, Garfinkel PE (1979) The Eating Attitudes Test: an index of the symptoms of anorexia nervosa. Psychol Med 9: 273-279. Grant DB, Hambley J, Bccker D, Pimstone BL (1973) Reduced sulfation factor in undernourished children. Arch Dis Child 48: 596-600. Hall K (1971) Effect of intravenous administration of human growth hormone on sulfation factor activity in serum of hypopituitary subjects. Acta Endocrino166: 491-497. Kiriike N, Nishiwaki S, Izumiya Y, Macda Y, Kawakita Y (1987) Thyrolxopin, prolactin and growth hormone responses to thyrotropin-relcasing hormone in anorexia nervosa and bulimia. Biol Psychiatry 22: 167-176. Levy AB, Dixon KN (1987) DST in bulimia without endogenous depression. Biol Psychiatry 22: 783-786. Levy AB, Dixon KIN,Malarkey WB (1988) Pituitary response to TRI-Iin bulimia. Biol Psychiatry 23: 476-484. Phillips LS, Hcfington AC, Doughaday WH (1975) Steroid hormone effects on somatomcdin. I. Somatomedin action in vitro. Endocrinology 97: 780-786. Phillips LS, Vassilopoulou-Sellin R (1980) Somatomedins. N Engl J Med 30: 438-446. Pirke KM, Pahl J, Schweigcr V, Warnhoff M (1985) Metabolic and endocrine indices of starvation in bulimia: a comparison with anorexia nervosa. Psychiatry Res 15: 33-39. Rappaport R, Prevot C, Czernichow P (1980) Somatomcdin activity and growth hormone secretion. Acta Pediatr Scand 69: 37-41. Vigersky RA, Anderson AE, Thompson RH, Loriaux DL (1977) Hypothalamic dysfunction in secondary amcnorrhca associated with simple weight loss. N Engl J Med 297:1141-1145. Yde H (1964) The growth-hormone-dependentsulfation factor in serum of untreated diabetes. Lancet ii: 624-626.