0021-972X/79/4906-0805$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1979 by The Endocrine Society
Vol. 49, No. 6 Printed in U.S.A.
Catecholamine Metabolism in Primary Anorexia Nervosa HOWARD A. GROSS, C. RAYMOND LAKE,* MICHAEL H. EBERT, MICHAEL G. ZIEGLER,f AND IRWIN J. KOPIN Section on Experimental Therapeutics, Laboratory of Clinical Science, National Institute of Mental Health, Bethesda, Maryland 20014
were more than 20-25% below ideal, patients' blood pressures and pulse rates, plasma levels of norepinephrine, and 24-h urinary excretion of 3-methoxy-4-hydroxyphenol glycol and homovanillic acid were lower than those of a group of normal volunteers. After weight gain, these parameters increased to nearnormal levels. At no time was plasma dopamine-/?-hydroxylase activity abnormal. The results suggest that abnormalities in catecholamine metabolism in primary anorexia nervosa are caused by starvation, and that neuronal functions dependent on aminergic neurotransmission may be altered as a result. (J Clin Endocrinol Metab 49: 805, 1979)
ABSTRACT. Abnormalities in neuroendocrine function and sympathetic nervous system activity appear to be present in primary anorexia nervosa. Hypothalamic catecholamines are involved in control of endocrine function and norepinephrine is released from sympathetic nerve endings. Because of possible abnormalities in catecholamine metabolism, plasma levels of norepinephrine and urinary excretion of homovanillic acid and 3-methoxy-4-hydroxyphenyl glycol were studied in female patients with primary anorexia nervosa before and after significant clinical improvement and compared with normal female volunteers. During the phase of the disease in which body weights
S
EVERAL observations suggest that patients with primary anorexia nervosa (PAN) have abnormalities in catecholamine metabolism in both the central and sympathetic nervous systems. The neuroendocrine {e.g. amenorrhea) and thermoregulatory abnormalities which have been documented recently in PAN (1) suggest the presence of hypothalamic dysfunction during the cachectic phase of PAN (2), and are consistent with an abnormality in sympathetic neuronal function (3, 4). A number of studies in animals suggest that noradrenergic and dopaminergic pathways in the hypothalamus play an important role in the normal regulation of appetite (5, 6) and endocrine function. Disordered central nervous system catecholaminergic function has been hypothesized in affective disorders (7), and there are several similarities between the psychiatric symptoms found in PAN and those of depressed patients (8). In the present study, patients with PAN were evaluated before and after weight gain for abnormalities of plasma tyrosine, norepinephrine (NE), and dopamine-/?-hydroxylase (DBH), as well as urinary excretion of 3-methoxy-4-hydroxyphenyl glycol (MHPG), the major metabolite of
NE in the brain, and homovanillic acid (HVA), the major metabolite of dopamine (DA) in the brain. Materials and Methods Fifteen female patients with PAN who were admitted to the Clinical Center at the NIH and 39 normal female volunteers in the same age range were the subjects for this study. After thorough medical and psychiatric evaluations, written informed consent was obtained and the treatment program was initiated. Six patients were evaluated both before and after a weight gain of 8-13 kg. Blood pressure and pulse rate were recorded in 15 ill and 13 improved patients. The patients ranged in age from 12-32 yr, with a mean age (±SEM) of 22 ± 2 yr; upon admission they met specific criteria for anorexia nervosa (9). The body weights ranged from 20.2-39.6 kg. The patients were amenorrheic, had lost at least 25% of their body weight, and were more than 15% below their ideal body weights, as determined from Metropolitan Life Tables and the University of Iowa Growth Charts for use in pediatric practice. The age of onset was less than 30 yr and occurred in the absence of medical illness or psychosis (primary affective disorder or schizophrenia). Typically, they had a distorted attitude toward food, with denial of nutritional needs, etc. Lanugo, bradycardia, periods of overactivity, episodes of bulemia, and vomiting were often present. All patients were placed on a low monoamine diet, which excludes chocolate, coffee, tea, colas, vanilla, oranges, nuts, bananas, and cheese. After a 3-day period of adjustment to the research unit, urine was collected during two sequential 24-h intervals. During the markedly anorectic phase of the illness, the patients consumed a diet containing vitamin supplements and at least 875 cal/day. The total caloric intake (±SEM) was
Received April 2, 1979. Address requests for reprints to: Dr. Howard A. Gross, National Institute of Mental Health, Building 10, Room 2S243, 9000 Rockville Pike, Bethesda, Maryland 20205. • Present address: Departments of Psychiatry and Pharmacology, Uniformed Services Medical School, Bethesda, Maryland 20014. f Present address: Departments of Medicine and Pharmacology, University of Texas Medical Branch, Galveston, Texas 77550.
805
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:38 For personal use only. No other uses without permission. . All rights reserved.
JCE&M • 1979 Vol49 • No 6
GROSS ET AL.
806
an average of 1460 ± 224 cal/day during the 4-day interval comprising the 2 days immediately before the urine collection interval and the 2 days during which urine was collected. The diet contained 54.2 ± 8.2 g protein, 61.3 ± 12.1 g fat, and 180 ± 20.8 g carbohydrate. Tube feeding was threatened but was not found to be necessary during the initial 5 days. After 10 days of evaluation, the patients were treated over the next 4-7 weeks with behavior modification therapy, with tube feeding occasionally required to maintain food intake. During the last 2 days of treatment, two 24-h urine collections were again obtained. During the last 4 days, the average food intake included 2790 ± 161 cal, consisting of 94.3 ± 7.6 g protein, 111 ± 9.3 g fat, and 357 ± 21.5 g carbohydrate. The control subjects studied under the same conditions for the urinary metabolite studies consisted of nine normal female volunteers with a mean age of 21 ± 3 years. They also were placed on a low monoamine diet 48 h before beginning two 24-h urine collections. These females had no medical or psychiatric illnesses and were at normal weights for their ages and heights. The urine was acidified with 3% 6 N HC1 and refrigerated during the collection period, after which aliquots were frozen and kept at —20 C. MHPG and HVA were assayed by a gas chromatography-mass spectrometry assay utilizing deuterated internal standards (10). For biological evaluation of sympathetic neuronal function (11), blood samples were obtained for determination of plasma levels of NE and DBH activity from the PAN patients while severely anorectic (n = 13) and/or after significant weight gain (n = 8). For controls, blood samples were obtained from 39 normal female subjects aged 23 ± 1 years. The patients and normal subjects were asked to lie supine in a quiet room and the needle of a heparin-lock was inserted into a suitable vein in the nondominant arm. When the subjects were relaxed, as judged subjectively and by pulse rate decrement to a stable rate (usually by 20 min after venipuncture), a basal (resting supine) blood sample of 12 ml was obtained. With subjects upright, a second 12-ml blood sample was taken after 5 min. Using an isometric hand grip with a pressure gauge, the subjects were asked to maintain 30% of their maximal grip for 5 min, at which
time the third blood sample was obtained. Blood pressures and pulse rates were measured at the times blood was sampled. Blood samples were collected in chilled glass tubes containing 2 ml anticoagulant (acid-citrate-dextrose) and centrifuged, and the plasma was removed, frozen, and stored at —70 C. NE was assayed by a radioenzymatic method previously described (11). In this assay, [3H]epinephrine is formed from endogenous NE by the phenylethanolamine-JV-methyltransferase-catalyzed transfer of a [3H]methyl group from [3H]S-adenosyl methionine to the amine group of NE. Tyrosine levels in plasma were determined by the micromethod of Weng et al. (12). DBH was assayed by the technique of Weinshilboum and Axelrod (13) using phenylethylamine as substrate in the presence of 3.3 X 10~6 M CuSO4 at pH 5.5 using 4 /xl plasma in a total incubation volume of 300 jul. DBH activity is expressed in units; 1 U is the equivalent of 1 nM phenylethanolamine generated/h of incubation. The patients were weighed each morning in the same hospital gown on the same scale before breakfast and after voiding.
Results Patients with marked weight loss during a period of anorexia nervosa had significantly lower blood pressures (supine and standing) and heart rates while standing or during exertion than the control subjects (Table 1). Anorectic patients who had gained weight had significantly higher heart rates and systolic pressures than the markedly underweight patients, but the increments in diastolic pressure and heart rate during exertion were not altered significantly. The increase in pulse rate elicited by the isometric hand grip was less in patients with PAN, both before and after weight gain, than in the controls. While supine or standing or during exertion, the markedly anorectic patients had plasma levels of NE that were significantly lower than those in control subjects or in the patients who had gained weight (Fig. 1). The plasma
TABLE 1. Cardiovascular studies in anorexia nervosa Pulse rate (beats/min)
Blood pressure (torr)
Age (yr)
Subjects
Controls (39) 111 (15)
23 ± 1 22 ± 2
Improved (13)
23 ± 2
Wt (kg) Supine
Standing
Exercise
72.5 ± 1.7 (39)
89.9 ± 2.3" (36)
117.0 ± 5.6° (14)
67.7 ± 2.9d
78.6 ± 2.5rf
92.3 ± 5.1
77.5 ± 2.8
89.9 ± 3.4
98.0 ± 4.1e
Supine
Standing
108.0 ± 1.6* 67.0 ± 1.6* 92.0 ± 2.0rf 58.0 ± 2.0 101.0 ± 2.0' 61.0 ± 2.0*
108.0 ± 1.8* 73.0 ± 1.3 91.0 ± 3.0rf 65.0 ± 2.0 102.0 ± 2.0" 70.0 ± 2.0
54.5C 32.5 ± 1.1 43.3 ± 1.9'
Values given are the mean ± SE; the number of subjects is in parentheses. n Different from ill patients, P < 0.005. 6 Different from ill patients, P < 0.001. c Represents ideal body weight for age, height, and sex from Metropolitan Life Insurance Tables and University of Iowa Growth Charts for use in pediatric practice. d Different from improved, P < 0.02. e Different from ill patients, P < 0.01. 'Different from controls, P < 0.01. 8 Different from controls, P < 0.025. h Different from controls, P < 0.05.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:38 For personal use only. No other uses without permission. . All rights reserved.
807
CATECHOLAMINES IN ANOREXIA NERVOSA levels of NE in the improved patients were not significantly different from those of the controls. DBH activities were similar in controls, anorectic patients, and recovered patients (629 ± 79, 703 ± 167, and 583 ± 166 U, respectively). The mean weight of anorectic patients (32.5 ± 1.1 kg) was significantly (P < 0.01) less than the patients who had gained weight (43.3 ± 1.9 kg). Six anorectic patients who gained 10.9 ± 0.8 kg during treatment were studied both before and after weight gain. The increases in plasma NE after weight gain were
|
|BASAL
E%3 STAND
B&&GRIP o 3 o
apparent in every patient at rest, when standing, and during exertion but did not seem to be related to weight gain. In six anorectic patients in whom plasma tyrosine levels were measured before and after weight gain, the mean levels [21.8 ± 3.5 and 20.1 ± 5.4 jUg/ml (±SD), respectively] were not significantly different from the levels found in normal nonfasting adults (12). Urinary excretion of both MHPG and HVA were significantly lower in ill anorectic patients than in the normal female control group (Table 2). Weight gain was attended by an increase in HVA excretion in all patients and an increase in MHPG excretion in 9 of the 10 patients studied both before and after adequate weight gain. The mean increases in excretion of MHPG (2.50 ± 0.54 jumol/ day) and HVA (6.7 ± 1.20 /imol/day) were both highly significant (P < 0.001) and resulted in attainment of mean excretion levels that were not significantly different from normal. In 3 patients, however, MHPG excretion remained lower than that found in any of the normal subjects. Discussion
tr O
z
Vol. 49, No. 6 Printed in U.S.A.
Catecholamine Metabolism in Primary Anorexia Nervosa HOWARD A. GROSS, C. RAYMOND LAKE,* MICHAEL H. EBERT, MICHAEL G. ZIEGLER,f AND IRWIN J. KOPIN Section on Experimental Therapeutics, Laboratory of Clinical Science, National Institute of Mental Health, Bethesda, Maryland 20014
were more than 20-25% below ideal, patients' blood pressures and pulse rates, plasma levels of norepinephrine, and 24-h urinary excretion of 3-methoxy-4-hydroxyphenol glycol and homovanillic acid were lower than those of a group of normal volunteers. After weight gain, these parameters increased to nearnormal levels. At no time was plasma dopamine-/?-hydroxylase activity abnormal. The results suggest that abnormalities in catecholamine metabolism in primary anorexia nervosa are caused by starvation, and that neuronal functions dependent on aminergic neurotransmission may be altered as a result. (J Clin Endocrinol Metab 49: 805, 1979)
ABSTRACT. Abnormalities in neuroendocrine function and sympathetic nervous system activity appear to be present in primary anorexia nervosa. Hypothalamic catecholamines are involved in control of endocrine function and norepinephrine is released from sympathetic nerve endings. Because of possible abnormalities in catecholamine metabolism, plasma levels of norepinephrine and urinary excretion of homovanillic acid and 3-methoxy-4-hydroxyphenyl glycol were studied in female patients with primary anorexia nervosa before and after significant clinical improvement and compared with normal female volunteers. During the phase of the disease in which body weights
S
EVERAL observations suggest that patients with primary anorexia nervosa (PAN) have abnormalities in catecholamine metabolism in both the central and sympathetic nervous systems. The neuroendocrine {e.g. amenorrhea) and thermoregulatory abnormalities which have been documented recently in PAN (1) suggest the presence of hypothalamic dysfunction during the cachectic phase of PAN (2), and are consistent with an abnormality in sympathetic neuronal function (3, 4). A number of studies in animals suggest that noradrenergic and dopaminergic pathways in the hypothalamus play an important role in the normal regulation of appetite (5, 6) and endocrine function. Disordered central nervous system catecholaminergic function has been hypothesized in affective disorders (7), and there are several similarities between the psychiatric symptoms found in PAN and those of depressed patients (8). In the present study, patients with PAN were evaluated before and after weight gain for abnormalities of plasma tyrosine, norepinephrine (NE), and dopamine-/?-hydroxylase (DBH), as well as urinary excretion of 3-methoxy-4-hydroxyphenyl glycol (MHPG), the major metabolite of
NE in the brain, and homovanillic acid (HVA), the major metabolite of dopamine (DA) in the brain. Materials and Methods Fifteen female patients with PAN who were admitted to the Clinical Center at the NIH and 39 normal female volunteers in the same age range were the subjects for this study. After thorough medical and psychiatric evaluations, written informed consent was obtained and the treatment program was initiated. Six patients were evaluated both before and after a weight gain of 8-13 kg. Blood pressure and pulse rate were recorded in 15 ill and 13 improved patients. The patients ranged in age from 12-32 yr, with a mean age (±SEM) of 22 ± 2 yr; upon admission they met specific criteria for anorexia nervosa (9). The body weights ranged from 20.2-39.6 kg. The patients were amenorrheic, had lost at least 25% of their body weight, and were more than 15% below their ideal body weights, as determined from Metropolitan Life Tables and the University of Iowa Growth Charts for use in pediatric practice. The age of onset was less than 30 yr and occurred in the absence of medical illness or psychosis (primary affective disorder or schizophrenia). Typically, they had a distorted attitude toward food, with denial of nutritional needs, etc. Lanugo, bradycardia, periods of overactivity, episodes of bulemia, and vomiting were often present. All patients were placed on a low monoamine diet, which excludes chocolate, coffee, tea, colas, vanilla, oranges, nuts, bananas, and cheese. After a 3-day period of adjustment to the research unit, urine was collected during two sequential 24-h intervals. During the markedly anorectic phase of the illness, the patients consumed a diet containing vitamin supplements and at least 875 cal/day. The total caloric intake (±SEM) was
Received April 2, 1979. Address requests for reprints to: Dr. Howard A. Gross, National Institute of Mental Health, Building 10, Room 2S243, 9000 Rockville Pike, Bethesda, Maryland 20205. • Present address: Departments of Psychiatry and Pharmacology, Uniformed Services Medical School, Bethesda, Maryland 20014. f Present address: Departments of Medicine and Pharmacology, University of Texas Medical Branch, Galveston, Texas 77550.
805
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:38 For personal use only. No other uses without permission. . All rights reserved.
JCE&M • 1979 Vol49 • No 6
GROSS ET AL.
806
an average of 1460 ± 224 cal/day during the 4-day interval comprising the 2 days immediately before the urine collection interval and the 2 days during which urine was collected. The diet contained 54.2 ± 8.2 g protein, 61.3 ± 12.1 g fat, and 180 ± 20.8 g carbohydrate. Tube feeding was threatened but was not found to be necessary during the initial 5 days. After 10 days of evaluation, the patients were treated over the next 4-7 weeks with behavior modification therapy, with tube feeding occasionally required to maintain food intake. During the last 2 days of treatment, two 24-h urine collections were again obtained. During the last 4 days, the average food intake included 2790 ± 161 cal, consisting of 94.3 ± 7.6 g protein, 111 ± 9.3 g fat, and 357 ± 21.5 g carbohydrate. The control subjects studied under the same conditions for the urinary metabolite studies consisted of nine normal female volunteers with a mean age of 21 ± 3 years. They also were placed on a low monoamine diet 48 h before beginning two 24-h urine collections. These females had no medical or psychiatric illnesses and were at normal weights for their ages and heights. The urine was acidified with 3% 6 N HC1 and refrigerated during the collection period, after which aliquots were frozen and kept at —20 C. MHPG and HVA were assayed by a gas chromatography-mass spectrometry assay utilizing deuterated internal standards (10). For biological evaluation of sympathetic neuronal function (11), blood samples were obtained for determination of plasma levels of NE and DBH activity from the PAN patients while severely anorectic (n = 13) and/or after significant weight gain (n = 8). For controls, blood samples were obtained from 39 normal female subjects aged 23 ± 1 years. The patients and normal subjects were asked to lie supine in a quiet room and the needle of a heparin-lock was inserted into a suitable vein in the nondominant arm. When the subjects were relaxed, as judged subjectively and by pulse rate decrement to a stable rate (usually by 20 min after venipuncture), a basal (resting supine) blood sample of 12 ml was obtained. With subjects upright, a second 12-ml blood sample was taken after 5 min. Using an isometric hand grip with a pressure gauge, the subjects were asked to maintain 30% of their maximal grip for 5 min, at which
time the third blood sample was obtained. Blood pressures and pulse rates were measured at the times blood was sampled. Blood samples were collected in chilled glass tubes containing 2 ml anticoagulant (acid-citrate-dextrose) and centrifuged, and the plasma was removed, frozen, and stored at —70 C. NE was assayed by a radioenzymatic method previously described (11). In this assay, [3H]epinephrine is formed from endogenous NE by the phenylethanolamine-JV-methyltransferase-catalyzed transfer of a [3H]methyl group from [3H]S-adenosyl methionine to the amine group of NE. Tyrosine levels in plasma were determined by the micromethod of Weng et al. (12). DBH was assayed by the technique of Weinshilboum and Axelrod (13) using phenylethylamine as substrate in the presence of 3.3 X 10~6 M CuSO4 at pH 5.5 using 4 /xl plasma in a total incubation volume of 300 jul. DBH activity is expressed in units; 1 U is the equivalent of 1 nM phenylethanolamine generated/h of incubation. The patients were weighed each morning in the same hospital gown on the same scale before breakfast and after voiding.
Results Patients with marked weight loss during a period of anorexia nervosa had significantly lower blood pressures (supine and standing) and heart rates while standing or during exertion than the control subjects (Table 1). Anorectic patients who had gained weight had significantly higher heart rates and systolic pressures than the markedly underweight patients, but the increments in diastolic pressure and heart rate during exertion were not altered significantly. The increase in pulse rate elicited by the isometric hand grip was less in patients with PAN, both before and after weight gain, than in the controls. While supine or standing or during exertion, the markedly anorectic patients had plasma levels of NE that were significantly lower than those in control subjects or in the patients who had gained weight (Fig. 1). The plasma
TABLE 1. Cardiovascular studies in anorexia nervosa Pulse rate (beats/min)
Blood pressure (torr)
Age (yr)
Subjects
Controls (39) 111 (15)
23 ± 1 22 ± 2
Improved (13)
23 ± 2
Wt (kg) Supine
Standing
Exercise
72.5 ± 1.7 (39)
89.9 ± 2.3" (36)
117.0 ± 5.6° (14)
67.7 ± 2.9d
78.6 ± 2.5rf
92.3 ± 5.1
77.5 ± 2.8
89.9 ± 3.4
98.0 ± 4.1e
Supine
Standing
108.0 ± 1.6* 67.0 ± 1.6* 92.0 ± 2.0rf 58.0 ± 2.0 101.0 ± 2.0' 61.0 ± 2.0*
108.0 ± 1.8* 73.0 ± 1.3 91.0 ± 3.0rf 65.0 ± 2.0 102.0 ± 2.0" 70.0 ± 2.0
54.5C 32.5 ± 1.1 43.3 ± 1.9'
Values given are the mean ± SE; the number of subjects is in parentheses. n Different from ill patients, P < 0.005. 6 Different from ill patients, P < 0.001. c Represents ideal body weight for age, height, and sex from Metropolitan Life Insurance Tables and University of Iowa Growth Charts for use in pediatric practice. d Different from improved, P < 0.02. e Different from ill patients, P < 0.01. 'Different from controls, P < 0.01. 8 Different from controls, P < 0.025. h Different from controls, P < 0.05.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:38 For personal use only. No other uses without permission. . All rights reserved.
807
CATECHOLAMINES IN ANOREXIA NERVOSA levels of NE in the improved patients were not significantly different from those of the controls. DBH activities were similar in controls, anorectic patients, and recovered patients (629 ± 79, 703 ± 167, and 583 ± 166 U, respectively). The mean weight of anorectic patients (32.5 ± 1.1 kg) was significantly (P < 0.01) less than the patients who had gained weight (43.3 ± 1.9 kg). Six anorectic patients who gained 10.9 ± 0.8 kg during treatment were studied both before and after weight gain. The increases in plasma NE after weight gain were
|
|BASAL
E%3 STAND
B&&GRIP o 3 o
apparent in every patient at rest, when standing, and during exertion but did not seem to be related to weight gain. In six anorectic patients in whom plasma tyrosine levels were measured before and after weight gain, the mean levels [21.8 ± 3.5 and 20.1 ± 5.4 jUg/ml (±SD), respectively] were not significantly different from the levels found in normal nonfasting adults (12). Urinary excretion of both MHPG and HVA were significantly lower in ill anorectic patients than in the normal female control group (Table 2). Weight gain was attended by an increase in HVA excretion in all patients and an increase in MHPG excretion in 9 of the 10 patients studied both before and after adequate weight gain. The mean increases in excretion of MHPG (2.50 ± 0.54 jumol/ day) and HVA (6.7 ± 1.20 /imol/day) were both highly significant (P < 0.001) and resulted in attainment of mean excretion levels that were not significantly different from normal. In 3 patients, however, MHPG excretion remained lower than that found in any of the normal subjects. Discussion
tr O
z
