Furop. J . clin. Invest. 5, 289-295 ( 1 9 7 5 )
The Growth Hormone Response to Insulin Induced Hypoglycaemia in Anorexia Nervosa and Control Underweight or Normal ,Subjects H. Brauman and F. Gregoire Service de Biologie Clinique and Institut de Psychiatrie, Hzpital Universitaire Brugmann, Brussels, Be 1gium Received: June 10, 1974, and in revised form: December 9, 1974 Swrimary. Hypoglycaemic stimulation of growth hormone (GH) secretion has been measured in 12 anorexia nervosa patients, 1 1 adolescents approximately matched for weight and age, 14 underweight adults with no history of anorexia nervosa and 27 normal patients. The results showed a high proportion of blunted growth hormone responses among the anorexia nervosa patients as compared to normal subjects, underweight adolescents or adults. In the group of underweight adults, diminished growth hormone responses were most frequently seen in patients with depressive illness. The complex pathogenesis of the diminished growth hormone response in anorexia nervosa involves many factors; undernutrition itself, possible hypothalamic insufficiency related to the psychopathological background and other endocrine abnormalities like hypogonadism, relative hypercorticism and hypothyroidism. The interference of the nutritional, psychological and endocrine factors could not be dissociated in our investigation or from the review of the literature on this subject. It is suggested that hypothalamic insufficiency plays a prominent role in the growth hormone hyporesponsiveness in anorexia nervosa. Although diminished, the stimulated growth hormone response remains higher in anorexia nervosa than in true hypopituitarism. This phenomenon coupled to a preserved or even increased suprarenal function enables one to differentiate cachexia resulting from anorexia nervosa and hypopituitarism leading eventually to cachexia.
Key oords: Hypoglycaemia, growth hormone, anorexia nervosa, undernutrition.
The introduction of radioimmunoassays of hypophyseal hormones has opened a wide field of study of the dynamics of hypophyseal function in health and disease. One of the most powerful and reliable ways to assess the functional capacity of the hypothalamo - hypophyseal axis is the growth hormone (GH) response to insulin induced hypoglycaemia. Since the first demonstration by Roth, Glick, Yalow and Berson (1963) of the stimulatory effect of rapidly decreasing levels of plasma glucose on growth hormone secretion, an extensive amount of data has been published concerning this test in many different clinical situations. Soon after the introduction of this valuable diagnostic test, attention was directed to the possible relationship between growth hormone secretion and the actual nutrttional state: adaptation of growth hormone secretion to longstanding starvation and the relation of the adaptative process to the initial cause and qualitative aspects of malnutrition. Moreover, the causal link which many clinicians established between hypopituitarism and cachexia sometimes led to questioning of the functional integrity of the pituitary in cases of cachexia. Patients with anorexia nervosa constituted an interesting group for investigation from the point of view of the hypophyseal responsiveness in res-
pect to the questions raised above. Few data have appeared until now concerning this particular aspect of this disease. The purpose of our paper is to report the growth hormone response to hypoglycaemia in anorexia nervosa, and to compare it to the response observed in subjects of the same degree of leanness but without the clinical features of anorexia nervosa and in normal subjects. This comparison was undertaken with the hope that it would permit definition of the exact role of leanness p e r se in an eventual growth hormone hyporesponsiveness.
Material and Methods The growth hormone response to insulin induced hypoglycaemia has been maasured on blood samples drawn before and every 30 minutes after a rapid I.V. injection of 0.1 unit of insulinlkg of body weight. The total duration of the test was 120 minutes in most of the cases and has been performed on hospitalized patients, early in the morning after an overnight resting and fasting period. The test has been performed in four groups of patients. Group I consisted of 12 females considered as having -anorexia nervosa on the basis of the follow-
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa
290
Table 1 . Anorexia nervosa patients. G1: glucose in mg/100 ml; GH: growth hormone in ng/ml before and at different times in minutes after insulin injection
GH response to insulin induced hypoglycaemia
Clinical data Patient
age in
weight in kg
height in cm
Duration of the disease in years
years
other 0' diagnosis G1 GB
20' 30' G1 G1 GH
45' G1
G1
60' GH
G1
90' GH
120'
G1
GH
7.6
50
5.5
0.8
60
0.0
delirious episode s
75
4.8
65
50
4.2
45
45
8.3
50
2
none
ao
7.6
30
31
9.8
35
35
7.6
50
163
I
none
76
5.0
35
25
7.0
3a
50
4.5
51
39
161
6
laxative I 0 0 abuse
2.0
60
35
5.0
60
65
16.0
75
19
38
151
I
none
70
2.8
4a
25
9.6
4a
50
3.2
45
Fon.
16
39
I65
2
none
94
4.0
25
32
32.0
59
67
14.6
7a
10.0 a4
15.2
7
Thir.
19
38
168
2
none
65
8.0
40
20
10.0
34
36
23.0
42
16.0
50
13.0
a
Cre.
37
33
157
I2
laxative abuse
90
1.0
42
25
8.0
40
50
5.0
65
7.0
75
-
9
Geu .
21
31
I63
2
de 1 ir iou s a7 episode
8.0
65
52
13.0
37
sa
8.0
60
5.0
70
4.0
I
Lis.
15
33
I60
1
2
Dec
.
15
33
163
3
Wat.
15
38
4
Bas.
29
5
Thie.
6
10
Lem
.
16
36
I55
I
suicidal attempts
92
3.8
60
46
16,8
30
35
24.0
50
11
Par,
17
41
156
1
none
85
6.6
-
69
10.6
60
65
16.4
80
6.4
a7
8.4
12
Coe.
18
36
161
2
none
7a
5.0
32
13
-
25
33
3.0
35
1.8
52
3.0
Mean
82.7
4.9
45.6 35.2
SEM
3.0
0.7
ing.criteria: amenorrhoea, weight well under the normal weight limits by any standard and the presence of a distinctive psychopathology as evaluated by a careful psychi-atric examination (Feighner e t aZ., 1 9 7 2 ) . Cold intolerance and constipation were common complaints. Laxative abuse was noticed in some patients. All had strange feeding patterns or refused food under all kinds of pretext. No clear qualitative and quantitative idea of food intake could be revealed by inquiry. None of the patients was of abnormally short stature (see Table 1 ) . Any organic cause of cachexia has been excluded by proper examination in each case. The duration of the disease at the time of investigation was at least 1 year. Some of the patients of this group have been investigated at an age where the disease is uncommon (patients 4 and 8, Table I ) . These patients had in fact a very long history of anorexia nervosa and following the records their disease could have started in late adolescence. In patients 1 and 9 Table 1 , anorexia ne'rvosa was complicated by deliridus episodes and patient 10 made suicidal attempts. Laboratory data showed: high normal cholesterol levels for this age group (mean 206 mg/100 ml, range 145-285 mg/100 ml; normal mean 197 and range 135-259 mg/100 ml) , high normal serum' iron levels
4.4
4.7
11.4 42.6 49.1 2.3
3.4
3.6
1 1 . 1 56.7
2.1
7 . 1 66.2
4.3 1 . 1
4.1
7.1
1.4
(mean 146 pg/100 ml, range 100-190 pg/lOO ml; normal mean 127 and range 59-210 ug/lOO ml), no anaemia (mean haematocrit 4 1 X , range 35-48,5 X), high normal total serum lipid levels (mean 810 mg/ 100 ml, range 650-1000 mg/100 ml; normal mean 692 and range 404-1092 mg/100 ml), low normal serum protein concentration (mean 6.8 g/lOO ml, range 5.5-8.5 g/lOO ml, normal mean 7.5 and range 6 . 7 8 . 3 g/lOO ml) and normal kaliaemia except in patient 10 Table 1 . Although no real suspicion of hypothyroidism existed in these patients the basal metabolic rate was low in most of the cases (range -34 to + 1 9 ) , the protein bound iodine (PBI) was however normal (mean 6 pg/IOO ml, range 4 . 8 - 7 . 5 pg/IOO ml; normal mean 6.5 and range 4-8 pg/IOO ml), the radioinnnunoassayable triiodothyronine was low normal (mean 108 ng/100 ml, range 79-143 ng/100 ml, normal mean 121 and range 82-165 ng/100 ml) and so was the radioimmunoassayable thyrotrophin (TSH) (mean 1 . 6 pU/ml, range 1 . 2 - 2 . 5 pU/ml, normal mean 3 . 3 and range 0 . 4 - 1 0 pU/ml). Group 11 consisted of 14 adults (Table 2 ) chosen because of they were underweight to an extent comparable to group I. The duration of their underweight state is less clearly documented than in the anorexia nervosa group. This group is heterogeneous as far as the clinical diagnosis
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa
291
Table 2 . Underweight adults. All patients are women except patient 7 . The significance of G 1 and GH see Table 1 Clinical data Patient
age in years
weight in kg
height in cm
GH response to insulin induced hypoglycaemia 0' G1 GH
20' G1
C1
GH
45' G1
G1
GH
G1
CH
G1
120' GH
neurotic 90 depression
4.8
54
36
4.8
47
41
5.2
40
4.8
50
4.8
Duration of Diagnose the disease in years I
30'
90'
60'
I
Que
40
40
I60
>
2
Cap
11
31
I55
many
Silicosis 95 Allergic rash
13.3
70
50
19.8
45
40
18.0
40
33.0
22
42.0
3
Moe
39
40
I59
many
Toxicomania
85
9.0
48
35
1.0
40
45
21.0
55
16.0
60
12.0
4
KeU
39
43
I60
=nY
Alcoholism
70
14.0
50
-
0.0
58
-
32.0
40
60.0
50
40.0
5
Cir
31
40
I60
many
Schizophrenia
75
16.0
45
32
11.0
38
43
32.0
40
62.0
66
50.0
6
Har
46
31
165
I
Neurotic 80 26.0 Depression
37
21
11.0
48
60
14.0
70
18.0
15
13.0
7
Boa
40
55
I97
many
Toxi82 comania Neurotic Depression
3.0
48
IS
3.0
10
10
4. 0
58
5.0
68
11.0
0
All
46
35
I60
many
Neurotic Depression
-
6.0
52
27
4.0
42
41
50.0
61
50.0
66
50.0
9
Cio
48
49
I62
2
85
10.2
40
63
15.2
68
14
22.0
75
20.0
63
19.0
10
Gi 1
67
31
155
many
Idiopathic 84 Sprue
6.2
46
30
64.0
35
45
24.8
60
26.4
-
-
I1
Mah
46
41
113
many
Psychastenia
95
1.0
47
28
26.0
25
38
10.0
42
5.0
52
4.0
12
Vek
52
36
I57
2
Neurotic 85 Depression
4.0
36
33
11.0
31
42
10.0
47
6.0
64
4.0
13
Fae
71
48
I55
many
Malnutrition
83
5.2
13
58
13.0
52
55
12.0
50
9.6
65
5.6
14
V.Ro
34
47
160
10
Malnutrition
80
3.0
30
20
2.0
22
45
34.0
a3
40.0
81
14.0
a5
9 . 1 48.3 34.9
40 45.5
20.6
55
25.5
60
20.7
5.5
4
Leucosis
Mean SEM
is concerned. Eight patients were hospitalized in the psychiatric ward for various reasons given in Table 2 . They had secondary depression, no organic disease and no specific nutritional deficiency according to the laboratory data. Two patients (no. 2 and 10, Table 2) had an organic disease severe enough to explain emaciation but in 2 patients (no. 13 and 1 4 , Table 2) , there was no obvious cause of the underweight state. It should be noted that triiodothyronine (mean 120 ng/100 ml, range 82-165 ng/100 ml) and TSH (mean 2 . 5 pU/ml, range 1-5 pU/ml) were in the normal limits. Group III (Table 3) consisted of 1 1 boys and girls referred to the hospital because their height and weight were well below that one expected for their chronological age. They were *
2.1
1.7
3.2
3.9
14.1 4.3
4
4
3.4 3.7
5
otherwise normal and have.been included here because they matched approximately, in age and weight, the anorexia nervosa patients. Group N consisted of 2 2 adults and 5 adolescents, with normal weight and stature and no history of mental or endocrine disease. They have been taken as a reference group as far as growth hormone response to insulin induced hypoglycaemia is concerned (Table 3 ) . Growth hormone has been measured by radioimmunoassay, using the iodination technique of Hunter and Greenwood (1962) for labelling, and the Glick e t a t . (1965) separation technique for bound and free hormone with only minor modification. Both separation techniques gave comparable results in our hands (Copinschi et aZ., 1972). Human growth
H. 'Brauman and F . Gregoire: Growth Hormone in Anorexia Nervosa
292
Table 3 . Upper p a n e l : s h o r t s t a t u r e and underweight a d o l e s c e n t s . Lower p a n e l : normal a d u l t s (22 and a d o l e s c e n t s ( 5 ) . G1 and G H see T a b l e I
age
C l i n i c a l data weight height i n kg i n cm
GH r e s p o n s e t o i n s u l i n induced hypoglycaemia
Sex
20'
0'
30'
45'
60'
90'
120'
G1
GH
G1
G1
GH
G1
G1
GH
G1
GH
G1
GH
56-110
0-10.8
25-65
20-66
1-34
25-80
37-92
8.1-37.6
48-96
5.4-40.0
55-99
0.8-31.0
I I
I I
10
II
10
9
II
II
I1
I1
9
8
mean
91.9
5.9
42.0
42.2
12.9
56.1
66.0
23.9
71.3
22.6
82.3
16.8
SEM
4.9
1.0
4.1
4.1
3.3
5.7
3.5
2.8
5.0
3.2
5.1
4.2
in
F
M
years range
n
range
n
13-16
29-40
136-151
I1
I I
I I
15-74
h6-84
148-197
27
27
27
6
5
16 I I
mean SEM
71-110 0.2-18.6
10-75 10-52
0-36
19-76
20-93
20-78
28-106 6.2-78
30-110
3-60
27
27
27
27
27
27
27
27
27
25
25
27
93.9
4.9
45.5
30.5
10.8
44.5
56.9
44.6
69.0
31.3
82.8
22.7
2.9
0.8
3.2
2.1
2.0
2.8
3.2
3.6
3.4
3.3
3.3
2.9
mar A nglml
GH
75
I I O
I
50 0'
I
I
I
I
O
0
0
0
I
0
25
0 O L
30
I
I
60
90 Time (minl
Fig. I . Mean growth hormone (GH, upper panel) and mean glucose (Gl, lower panel) response curves after insulin injection: in anorexia nemosa (A), underweight adolescents ( A ) , underweight adults (0) and normal patients ( e ) . The vertical bars represent the SEM
2
A
0
I
120
Fig. 2 . Maximal increment of growth hormone (max A GH) during insulin induced hypoglycaemia as a function of basal GH level in anorexia nervosa (A) , underweight adolescents (A) and underweight adults (0). The vertical dotted line represents the upper limit for our normal basal GH level, the horizontal dotted line represents the lower limit of our normal values of max A GH
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa hormone (hGH) Wilhelmi 705 A was used as a standard and for labelling. Plasma samples of each group have been assayed by both techniques in order to exclude any systematic variation in the results. Glucose was assayed, by the method of Hoffman (1937) adapted to autoanalyzer, on blood samples collected at the same intervals as for growth hormone and on additional samples obtained at 2 0 and 45 minutes. In most of the patients plasma cortisol was also measured by the method of Vermeulen and Vanderstraeten (1964) on blood samples collected at the same intervals as for growth hormone. No significant differences of the cortisol basal levels or the shape of the cortisol response curves have been seen among the different groups of patients and therefore the results will not be reported.
Results Detailed values of growth hormone and glucose levels observed during the provocative tests are presented in Table 1 for the anorexia nervosa patients and in Table 2 for the group of underweight adults. Table 3 gives the range and average values of growth hormone and glucose observed in the underweight adolescents and the control group. A few results stated as "0" mean in fact that in repeated assays no significant difference between these samples and the buffer control could be detected. The mean growth hormone and glucose response curves for each group of patients are represented in Fig. 1 . As it can be seen from Fig. 1, the shape of the growth hormone response curve differs from group to group despite very comparable average glucose response curves. The anorexia nervosa patients have a mean response curve distinctly lower than normal. Two patients only, of this group, had a growth hormone response curve above the lowest value of the normal range. The mean response curve of the 1 1 underweight adolescents is also flatened as compared to normal but not to such extent as seen in the previous group. Only 1 of the 1 1 patients of this group showed a frankly subnormal growth hormone response whereas the response of the ten others was in the lower part of the normal range, Contrasting with the other groups, the underweight adults show a much greater heterogeneity of growth hormone response to insulin induced hypoglycaemia (Table 2 ) . The mean growth hormone response curve for this group (Fig. 1) is also flatened as compared to normal and to the same extent as the mean response curve seen in the underweight adolescents. As it has been suggested that high basal levels of growth hormone induce an impairment of growth hormone response to a further acute stimulation, we looked for a negative correlation between the maximal increment of growth hormone (max. A GH) and the basal level. It appears from Fig. 2 that such a negative correlation does not exist for our underweight patients (group I, I1 and 111). There
293
is only 1 patient of the 37 included in this analysis whose poor growth hormone response can be related to a basal level of 2 6 ng/ml, which is abnormally high. This patient belonged to the group of underweight adults (patient 6 Table 2 ) .
Discussion It appears from our study that a high proportion of anorexia nervosa patients showed a poor growth hormone response to hypoglycaemic stimulation when compared to normal patients. This blunting of the growth hormone response does not seem to be directly related to the decrease in body weight: it was not observed in the two other groups of patients who were underweight to comparable degrees and no correlation could be found between response and absolute body weight or degree underweight. Such blunting has been reported in a few cases of anorexia nervosa: by Marks et aZ. (1965) in 2 cases and by Landon et al. (1966) in 3 out of 5 cases. These authors suggested that this phenomenon was probably related to the high growth hormone basal level, which is frequently found in fasting (Glick et al., 1965), chronic starvation (Alvarez et az., 1972) or improper food balance (Pimstone et al., 1966; Samuel and Deshpande, 1972). None of our anorexia nervosa patients had a high growth hormone basal level and this cause of unresponsiveness can be ruled out in our series. Moreover, Alvarez et al. (1972) observed, in a series of undernourished patients, supranormal basal levels and supranormal growth hormone response to insulin induced hypoglycaemia, Jacquet et al. (1970) observed in lean subjects supranormal basal levels and normal response to stimulation by arginine. Independently of any increasing effect on basal growth hormone level, improper food balance and starvation could have induced a relative unresponsiveness in the anorexia nervosa patients. Indeed, protein-poor diets have been shown by Merimee and Fineberg (1972) and Samuel et al. (1972) to diminish the stimulated growth hormone response. However, anorexia nervosa patients are known to avoid carbohydrates and fats especially but not meat (Slater and Roth, 1969) and in our patients almost normal serum protein levels hardly support any specific lack of proteins. Chronic starvation might have led to a diminished reserve of growth hormone as suggested by the observations made by Wilson (1954), Russfield and Somers (1965) in humans, and by Srebnik et al. (1959) , and Meites and Feil (1965) in animals. This last factor cannot entirely be excluded in our series. As no fully satisfactory explanation, related to chronic undernutrition, may be advanced for the blunting of the growth hormone response, one has to turn to other hypotheses. Hypothalamic nuclei are the site of production of hypophyseal hormone releasing factors and dopaminergic substances (Toivola and Gale, 1972; Himswath et a t . , 1972; Parkhie and Johnson, 1973; Harris, 1973). Experimental destruction of the ventromedial nucleus in primates results in a situation which closely resembles the one observed
294
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa
1, weight gain 8 kg) and in that case, the shape of in our anorexia nervosa patients: normal basal the growth hormone response curve did not change. levels of growth hormone and poor response to inIn three other cases (patients 1 , 9 and 10 Table I ) , sulin induced hypoglycaemia (Abrams et at., 1966). the treatment did not increase weight and growth There are other psychopathological situations than hormone response remained unchanged. As far as anorexia nervosa, not necessarily associated with other endocrine functions are considered, Russell undernourishment, which have been recently demon(1965) showed for example that amenorrhoea in strated to have impaired growth hormone secretion: anorexia nervosa was very resistant to treatment certain cases of depressive illness, as described and concluded that malnutrition accounts for only by Sachar et a t . (1971, 1972), and in maternal some abnormalities of hormone output and that the care deprived dwarfs (Powell et aZ., 1967). It likely site of the disturbance is the hypothalamus. might be pointed out that the relative lowering Force feeding does not of course correct the underof the mean growth hormone response curve seen in lying mental trouble probably responsible for the our group of underweight adults is precisely due endocrine abnormalities. to the poor growth hormone response seen in some The differential diagnosis between cachexia of the depressed patients belonging to this group leading eventually to hypopituitarism and true (see Table 2). hypopituitarism with cachexia may be difficult Although this hypothesis is attractive and in border line cases. The persistence of a normal supported by a recent study performed in 5 cases suprarenal function and the absence of real hypoof anorexia nervosa by Mecklenburg et aZ. (1974), thyroidism, as seen in anorexia nervosa, makes it should not lead one to neglect the possible the diagnosis of primary hypopituitarism most influence of other endocrinological factors which may play a role in growth hormone unresponsiveness. unlikely. Moreover, although frequently diminished, the growth hormone response in anorexia nervosa Hypercorticism inhibits the growth hormone resremains higher than in true hypopituitarism and ponse to insulin induced hypoglycaemia (Frantz cachexia occurs only rarely even in longstanding and Rabkin, 1 9 6 4 ) . Sachar et a;. (1971) found a hypopituitarism. relationship between increased cortisol production rate and growth hormone unresponsiveness in depressed patients. It is possible that an increase References, of the cortisol production rate, a parameter which has not been evaluated in our patients, was present Abrams, R.L., Parker, M.L., Blanco, S., Reichlin, in some of them. Their normal basal cortisol level S., Daughaday, W.M.: Hypothalamic regulation of and cortisol response to hypoglycaemia does not growth hormone secretion. Endocrinology 78, exclude a high cortisol production rate as sug605 (1966) gested by the report of Boyar et a2. (1974). Alvarez, L.C., Dimas, C.A., Castro, A., Rossman, A normal thyroid function is also necessary L.G. , Vanderlaan, E.F. , Vanderlaan, W.P. : Growth for a normal basal and stimulated growth hormone hormone in malnutrition. J. clin. Endocr. 3 4 , secretion (Brauman and Corvilain, 1968; Brauman 400 (1972) ot a t . , 19n). There is some evidence, as reviewed by Crisp (1970), that in anorexia nervosa an adapt- Boyar, R.M., Kapen, S., Finkelstein, J.W., Fukushima, D.K. , Weitzman, E.D. , Hellman, L. : Deive process to the hypocaloric diet leads to a lineation of the luteinizing hormone and corsecondary type of thyroid hypofunction. This would tisol abnormalities in anorexia nervosa. Abbe an additional contributory factor for growth hormone unresponsiveness and in fact the TSH and stract 33, SOC. clin. Invest. (1974) Brauman, H., Corvilain, J.: Growth hormone response the triiodothyronine levels were to some extent lower than normal in our anorexia nemosa patients. to hypoglycemia in myxedema. J. clin. Endocr. 28, 301 (1968) Oestrogens and gonadotrophins seem to play an enhancing role on basal and stimulated secretion Brauman, H., Smets, P., Corvilain., J.: Comparative of growth hormone (Frantz and Rabkin, 1965; Hillstudy of growth hormone response in normal subjects man et at., 1968; Merimee et aZ., 1967). Unresand in patients with primary myxedema or hyperponsiveness has been observed in cases of inthyroidism before and after treatment. J. clin. duced hypogonadism (Simon et at., 1967); gonadal Endocr. 3 6 , 1162 (1973) insufficiency has been quantitatively documented Copinschi, G., Wolter, R., Brauman, H.: Reproin anorexia nervosa (Russell, 1965) and therefore ductibility of plasma growth hormone deterthe possible interference of this factor can be minations between different laboratories. Clin. taken in account. chim. Atta 38, 207 (1972) Crisp, A.H. : Anorexia nervosa. "Feeding disorder!', Many factors seem thus to play a role in the relative unresponsiveness of the anorexia nervosa llnervous malnutrition" or "weight phobia"? Wld. patients : undernutrition, possible hypercorticism, Rev. Nutr. Diet 12, 452 (1970) relative hypothyroidism, depression of the gonadal Feighner, J.P., Robins, E., Guze, S.B., Woodruff, function and finally an ill defined hypothalamic R.A. Jr., Winokur, G., Munoz, R.: Diagnostic criteria for use in psychiatric research. Arch. insufficiency related to the mental disease. If undernutrition should play the prominent role, then gen. Psychiat. 26, 57 (1972) one could expect a better responsiveness after reFrantz, A.G., Rabkin, M.T. : Human growth hormone. feeding. We had no opportunity to retest the Clinical measurement, response to hypoglycemia patients after a substantial gain of weight during and suppression by corticosteroids. New Engl. force feeding except in one case (patient 2 Table J. Med. 271, 1375 (1964)
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa Frantz, A.G., Rabkin, M.T.: Effect of estrogen and sex difference on secretion of human growth hormone. J. clin. Endocr. 25, 1470 (1965) Glick, S.M., Roth, J . , Yalow, R.S., Berson, S.A.: Immuno-assay of human growth hormone in plasma. Nature (London) 199, 784 (1963) Glick, S.M., Roth, J . , Yalow, R.S., Berson, S.A.: Regulation of growth hormone secretion. Recent Prog'r. Hormone Res. 21, 241 (1965) Harris, F.: Congenital hypothalamic lesion leading to growth hormone deficiency and destruction of the appetite satiety centre. Proc. roy. SOC. Med. 66, 217 (1973) Hayek, A . , Crawfard, J.D.: L-Dopa and pituitary secretion. J. clin. Endocr. 34, 764 (1972) Herbert, V., Lau, K.S., Gottlieb, C.W., Bleicher, S.J.: Coated charcoal immunoassay of insulin. J. clin. Endocr. 25, 1375 (1965) Hillman, J . , Hanrmond, J . , Sokola, J., Reiss, M.: Changes in plasma growth hormone levels in retarded children. J. ment. Defic. Res. 12, 294 (1 968) Himsworth, R.L., Carmel, P.W., Frantz, A.G.: The location of the chemoreceptor controlling growth hormone secretion during hypoglycemia in primates. Endocrinology 91, 217 (1972) Hoffman, W.S.: Rapid photoelectric method for determination of glucose in blood and urine. J. biol. Chem. 120, 51 (1937) Hunter, W.M., Greenwood, F.C.: Preparation of iodine-131 labelled growth hormone of high specific activity. Nature (London) 194, 495 ( 1962) Jacquet, Ph., Scornavachi, J.C., Vague, Ph., Vague, J.: Variations du taux plasmatique de l'hormone somatotrope humaine, au cours du test 5 l'arginine, dans les obisiths et les maigreurs. Ann. Endocr. (Paris) 31, 80 (1970) Parkhie, M.F.. , Johnson, H.D. : Hypothalamic growth hormone releasing factor: release and synthesis after exposure to a high ambient temperature. Proc. SOC. exp. Biol. (N.Y.) 142, 311 (1973)
295
Russell, G.F.M.: Metabolic aspects of anorexia nervosa. Proc. roy. SOC. Med. 58, 811 (1965) Russfield, A.B., Somers, S.C.: Malnutrition and tropic hormone storage in hypophysis. Arch. Path. 75, 564 (1963) Sachar, E.J., Finkelstein, J., Hellman, L.: Growth hormone response in depressive illness. Arch. gen. Psychiat. 25, 263 (1971) Sachar, E.J., Mushrush, G., Perlow; M., Weitzman, E.D., Sassin, J.: Growth hormone response to L-Dopa in depressed patients. Science 178, 1304 (1 972) Samuel, A.M., Deshpande, U.R.: Growth hormone levels in protein caloric malnutrition. J. clin. Endocr. 35, 863 (1972) Sherman, L., Kim, S., Benjamin, F., Kolodny, H.D.: Effect of chlorpromazine on serum growth-hormone concentration in man. New Engl. J. Med. 284, 72 (1971) Simon, S., Schiffer, M., Glick, S.M., Schwartz, E.: Effect of medroxyprogesterone acetate upon stimulated release of growth hormone in men. J. clin. Endocr. 27, 1633 (1967) Slater, R., Roth, M.: Clinical Psychiatry. Third edition, p. 124. London: Baillisre, Tindall and Cassell 1969 Srebnik, H.H., Nelson, M.M., Simpson, M.E.: Reduced growth hormone content in anterior pituitary of rats on protein free diets. Proc. SOC. exp. Biol. (N.Y.) 101, 97 (1959) Toivola, P.T.K., Gale, C.C.: Stimulation of growth hormone release by microinjection of norepinephrin into hypothalamus of baboons. Endocrinology 90, 895 (1972)
H. Brauman Service de Biologie Clinique HEpital Universitaire Brugman I, av. J.J. Crocq B-1020 Brussels Belgium
The Growth Hormone Response to Insulin Induced Hypoglycaemia in Anorexia Nervosa and Control Underweight or Normal ,Subjects H. Brauman and F. Gregoire Service de Biologie Clinique and Institut de Psychiatrie, Hzpital Universitaire Brugmann, Brussels, Be 1gium Received: June 10, 1974, and in revised form: December 9, 1974 Swrimary. Hypoglycaemic stimulation of growth hormone (GH) secretion has been measured in 12 anorexia nervosa patients, 1 1 adolescents approximately matched for weight and age, 14 underweight adults with no history of anorexia nervosa and 27 normal patients. The results showed a high proportion of blunted growth hormone responses among the anorexia nervosa patients as compared to normal subjects, underweight adolescents or adults. In the group of underweight adults, diminished growth hormone responses were most frequently seen in patients with depressive illness. The complex pathogenesis of the diminished growth hormone response in anorexia nervosa involves many factors; undernutrition itself, possible hypothalamic insufficiency related to the psychopathological background and other endocrine abnormalities like hypogonadism, relative hypercorticism and hypothyroidism. The interference of the nutritional, psychological and endocrine factors could not be dissociated in our investigation or from the review of the literature on this subject. It is suggested that hypothalamic insufficiency plays a prominent role in the growth hormone hyporesponsiveness in anorexia nervosa. Although diminished, the stimulated growth hormone response remains higher in anorexia nervosa than in true hypopituitarism. This phenomenon coupled to a preserved or even increased suprarenal function enables one to differentiate cachexia resulting from anorexia nervosa and hypopituitarism leading eventually to cachexia.
Key oords: Hypoglycaemia, growth hormone, anorexia nervosa, undernutrition.
The introduction of radioimmunoassays of hypophyseal hormones has opened a wide field of study of the dynamics of hypophyseal function in health and disease. One of the most powerful and reliable ways to assess the functional capacity of the hypothalamo - hypophyseal axis is the growth hormone (GH) response to insulin induced hypoglycaemia. Since the first demonstration by Roth, Glick, Yalow and Berson (1963) of the stimulatory effect of rapidly decreasing levels of plasma glucose on growth hormone secretion, an extensive amount of data has been published concerning this test in many different clinical situations. Soon after the introduction of this valuable diagnostic test, attention was directed to the possible relationship between growth hormone secretion and the actual nutrttional state: adaptation of growth hormone secretion to longstanding starvation and the relation of the adaptative process to the initial cause and qualitative aspects of malnutrition. Moreover, the causal link which many clinicians established between hypopituitarism and cachexia sometimes led to questioning of the functional integrity of the pituitary in cases of cachexia. Patients with anorexia nervosa constituted an interesting group for investigation from the point of view of the hypophyseal responsiveness in res-
pect to the questions raised above. Few data have appeared until now concerning this particular aspect of this disease. The purpose of our paper is to report the growth hormone response to hypoglycaemia in anorexia nervosa, and to compare it to the response observed in subjects of the same degree of leanness but without the clinical features of anorexia nervosa and in normal subjects. This comparison was undertaken with the hope that it would permit definition of the exact role of leanness p e r se in an eventual growth hormone hyporesponsiveness.
Material and Methods The growth hormone response to insulin induced hypoglycaemia has been maasured on blood samples drawn before and every 30 minutes after a rapid I.V. injection of 0.1 unit of insulinlkg of body weight. The total duration of the test was 120 minutes in most of the cases and has been performed on hospitalized patients, early in the morning after an overnight resting and fasting period. The test has been performed in four groups of patients. Group I consisted of 12 females considered as having -anorexia nervosa on the basis of the follow-
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa
290
Table 1 . Anorexia nervosa patients. G1: glucose in mg/100 ml; GH: growth hormone in ng/ml before and at different times in minutes after insulin injection
GH response to insulin induced hypoglycaemia
Clinical data Patient
age in
weight in kg
height in cm
Duration of the disease in years
years
other 0' diagnosis G1 GB
20' 30' G1 G1 GH
45' G1
G1
60' GH
G1
90' GH
120'
G1
GH
7.6
50
5.5
0.8
60
0.0
delirious episode s
75
4.8
65
50
4.2
45
45
8.3
50
2
none
ao
7.6
30
31
9.8
35
35
7.6
50
163
I
none
76
5.0
35
25
7.0
3a
50
4.5
51
39
161
6
laxative I 0 0 abuse
2.0
60
35
5.0
60
65
16.0
75
19
38
151
I
none
70
2.8
4a
25
9.6
4a
50
3.2
45
Fon.
16
39
I65
2
none
94
4.0
25
32
32.0
59
67
14.6
7a
10.0 a4
15.2
7
Thir.
19
38
168
2
none
65
8.0
40
20
10.0
34
36
23.0
42
16.0
50
13.0
a
Cre.
37
33
157
I2
laxative abuse
90
1.0
42
25
8.0
40
50
5.0
65
7.0
75
-
9
Geu .
21
31
I63
2
de 1 ir iou s a7 episode
8.0
65
52
13.0
37
sa
8.0
60
5.0
70
4.0
I
Lis.
15
33
I60
1
2
Dec
.
15
33
163
3
Wat.
15
38
4
Bas.
29
5
Thie.
6
10
Lem
.
16
36
I55
I
suicidal attempts
92
3.8
60
46
16,8
30
35
24.0
50
11
Par,
17
41
156
1
none
85
6.6
-
69
10.6
60
65
16.4
80
6.4
a7
8.4
12
Coe.
18
36
161
2
none
7a
5.0
32
13
-
25
33
3.0
35
1.8
52
3.0
Mean
82.7
4.9
45.6 35.2
SEM
3.0
0.7
ing.criteria: amenorrhoea, weight well under the normal weight limits by any standard and the presence of a distinctive psychopathology as evaluated by a careful psychi-atric examination (Feighner e t aZ., 1 9 7 2 ) . Cold intolerance and constipation were common complaints. Laxative abuse was noticed in some patients. All had strange feeding patterns or refused food under all kinds of pretext. No clear qualitative and quantitative idea of food intake could be revealed by inquiry. None of the patients was of abnormally short stature (see Table 1 ) . Any organic cause of cachexia has been excluded by proper examination in each case. The duration of the disease at the time of investigation was at least 1 year. Some of the patients of this group have been investigated at an age where the disease is uncommon (patients 4 and 8, Table I ) . These patients had in fact a very long history of anorexia nervosa and following the records their disease could have started in late adolescence. In patients 1 and 9 Table 1 , anorexia ne'rvosa was complicated by deliridus episodes and patient 10 made suicidal attempts. Laboratory data showed: high normal cholesterol levels for this age group (mean 206 mg/100 ml, range 145-285 mg/100 ml; normal mean 197 and range 135-259 mg/100 ml) , high normal serum' iron levels
4.4
4.7
11.4 42.6 49.1 2.3
3.4
3.6
1 1 . 1 56.7
2.1
7 . 1 66.2
4.3 1 . 1
4.1
7.1
1.4
(mean 146 pg/100 ml, range 100-190 pg/lOO ml; normal mean 127 and range 59-210 ug/lOO ml), no anaemia (mean haematocrit 4 1 X , range 35-48,5 X), high normal total serum lipid levels (mean 810 mg/ 100 ml, range 650-1000 mg/100 ml; normal mean 692 and range 404-1092 mg/100 ml), low normal serum protein concentration (mean 6.8 g/lOO ml, range 5.5-8.5 g/lOO ml, normal mean 7.5 and range 6 . 7 8 . 3 g/lOO ml) and normal kaliaemia except in patient 10 Table 1 . Although no real suspicion of hypothyroidism existed in these patients the basal metabolic rate was low in most of the cases (range -34 to + 1 9 ) , the protein bound iodine (PBI) was however normal (mean 6 pg/IOO ml, range 4 . 8 - 7 . 5 pg/IOO ml; normal mean 6.5 and range 4-8 pg/IOO ml), the radioinnnunoassayable triiodothyronine was low normal (mean 108 ng/100 ml, range 79-143 ng/100 ml, normal mean 121 and range 82-165 ng/100 ml) and so was the radioimmunoassayable thyrotrophin (TSH) (mean 1 . 6 pU/ml, range 1 . 2 - 2 . 5 pU/ml, normal mean 3 . 3 and range 0 . 4 - 1 0 pU/ml). Group 11 consisted of 14 adults (Table 2 ) chosen because of they were underweight to an extent comparable to group I. The duration of their underweight state is less clearly documented than in the anorexia nervosa group. This group is heterogeneous as far as the clinical diagnosis
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa
291
Table 2 . Underweight adults. All patients are women except patient 7 . The significance of G 1 and GH see Table 1 Clinical data Patient
age in years
weight in kg
height in cm
GH response to insulin induced hypoglycaemia 0' G1 GH
20' G1
C1
GH
45' G1
G1
GH
G1
CH
G1
120' GH
neurotic 90 depression
4.8
54
36
4.8
47
41
5.2
40
4.8
50
4.8
Duration of Diagnose the disease in years I
30'
90'
60'
I
Que
40
40
I60
>
2
Cap
11
31
I55
many
Silicosis 95 Allergic rash
13.3
70
50
19.8
45
40
18.0
40
33.0
22
42.0
3
Moe
39
40
I59
many
Toxicomania
85
9.0
48
35
1.0
40
45
21.0
55
16.0
60
12.0
4
KeU
39
43
I60
=nY
Alcoholism
70
14.0
50
-
0.0
58
-
32.0
40
60.0
50
40.0
5
Cir
31
40
I60
many
Schizophrenia
75
16.0
45
32
11.0
38
43
32.0
40
62.0
66
50.0
6
Har
46
31
165
I
Neurotic 80 26.0 Depression
37
21
11.0
48
60
14.0
70
18.0
15
13.0
7
Boa
40
55
I97
many
Toxi82 comania Neurotic Depression
3.0
48
IS
3.0
10
10
4. 0
58
5.0
68
11.0
0
All
46
35
I60
many
Neurotic Depression
-
6.0
52
27
4.0
42
41
50.0
61
50.0
66
50.0
9
Cio
48
49
I62
2
85
10.2
40
63
15.2
68
14
22.0
75
20.0
63
19.0
10
Gi 1
67
31
155
many
Idiopathic 84 Sprue
6.2
46
30
64.0
35
45
24.8
60
26.4
-
-
I1
Mah
46
41
113
many
Psychastenia
95
1.0
47
28
26.0
25
38
10.0
42
5.0
52
4.0
12
Vek
52
36
I57
2
Neurotic 85 Depression
4.0
36
33
11.0
31
42
10.0
47
6.0
64
4.0
13
Fae
71
48
I55
many
Malnutrition
83
5.2
13
58
13.0
52
55
12.0
50
9.6
65
5.6
14
V.Ro
34
47
160
10
Malnutrition
80
3.0
30
20
2.0
22
45
34.0
a3
40.0
81
14.0
a5
9 . 1 48.3 34.9
40 45.5
20.6
55
25.5
60
20.7
5.5
4
Leucosis
Mean SEM
is concerned. Eight patients were hospitalized in the psychiatric ward for various reasons given in Table 2 . They had secondary depression, no organic disease and no specific nutritional deficiency according to the laboratory data. Two patients (no. 2 and 10, Table 2) had an organic disease severe enough to explain emaciation but in 2 patients (no. 13 and 1 4 , Table 2) , there was no obvious cause of the underweight state. It should be noted that triiodothyronine (mean 120 ng/100 ml, range 82-165 ng/100 ml) and TSH (mean 2 . 5 pU/ml, range 1-5 pU/ml) were in the normal limits. Group III (Table 3) consisted of 1 1 boys and girls referred to the hospital because their height and weight were well below that one expected for their chronological age. They were *
2.1
1.7
3.2
3.9
14.1 4.3
4
4
3.4 3.7
5
otherwise normal and have.been included here because they matched approximately, in age and weight, the anorexia nervosa patients. Group N consisted of 2 2 adults and 5 adolescents, with normal weight and stature and no history of mental or endocrine disease. They have been taken as a reference group as far as growth hormone response to insulin induced hypoglycaemia is concerned (Table 3 ) . Growth hormone has been measured by radioimmunoassay, using the iodination technique of Hunter and Greenwood (1962) for labelling, and the Glick e t a t . (1965) separation technique for bound and free hormone with only minor modification. Both separation techniques gave comparable results in our hands (Copinschi et aZ., 1972). Human growth
H. 'Brauman and F . Gregoire: Growth Hormone in Anorexia Nervosa
292
Table 3 . Upper p a n e l : s h o r t s t a t u r e and underweight a d o l e s c e n t s . Lower p a n e l : normal a d u l t s (22 and a d o l e s c e n t s ( 5 ) . G1 and G H see T a b l e I
age
C l i n i c a l data weight height i n kg i n cm
GH r e s p o n s e t o i n s u l i n induced hypoglycaemia
Sex
20'
0'
30'
45'
60'
90'
120'
G1
GH
G1
G1
GH
G1
G1
GH
G1
GH
G1
GH
56-110
0-10.8
25-65
20-66
1-34
25-80
37-92
8.1-37.6
48-96
5.4-40.0
55-99
0.8-31.0
I I
I I
10
II
10
9
II
II
I1
I1
9
8
mean
91.9
5.9
42.0
42.2
12.9
56.1
66.0
23.9
71.3
22.6
82.3
16.8
SEM
4.9
1.0
4.1
4.1
3.3
5.7
3.5
2.8
5.0
3.2
5.1
4.2
in
F
M
years range
n
range
n
13-16
29-40
136-151
I1
I I
I I
15-74
h6-84
148-197
27
27
27
6
5
16 I I
mean SEM
71-110 0.2-18.6
10-75 10-52
0-36
19-76
20-93
20-78
28-106 6.2-78
30-110
3-60
27
27
27
27
27
27
27
27
27
25
25
27
93.9
4.9
45.5
30.5
10.8
44.5
56.9
44.6
69.0
31.3
82.8
22.7
2.9
0.8
3.2
2.1
2.0
2.8
3.2
3.6
3.4
3.3
3.3
2.9
mar A nglml
GH
75
I I O
I
50 0'
I
I
I
I
O
0
0
0
I
0
25
0 O L
30
I
I
60
90 Time (minl
Fig. I . Mean growth hormone (GH, upper panel) and mean glucose (Gl, lower panel) response curves after insulin injection: in anorexia nemosa (A), underweight adolescents ( A ) , underweight adults (0) and normal patients ( e ) . The vertical bars represent the SEM
2
A
0
I
120
Fig. 2 . Maximal increment of growth hormone (max A GH) during insulin induced hypoglycaemia as a function of basal GH level in anorexia nervosa (A) , underweight adolescents (A) and underweight adults (0). The vertical dotted line represents the upper limit for our normal basal GH level, the horizontal dotted line represents the lower limit of our normal values of max A GH
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa hormone (hGH) Wilhelmi 705 A was used as a standard and for labelling. Plasma samples of each group have been assayed by both techniques in order to exclude any systematic variation in the results. Glucose was assayed, by the method of Hoffman (1937) adapted to autoanalyzer, on blood samples collected at the same intervals as for growth hormone and on additional samples obtained at 2 0 and 45 minutes. In most of the patients plasma cortisol was also measured by the method of Vermeulen and Vanderstraeten (1964) on blood samples collected at the same intervals as for growth hormone. No significant differences of the cortisol basal levels or the shape of the cortisol response curves have been seen among the different groups of patients and therefore the results will not be reported.
Results Detailed values of growth hormone and glucose levels observed during the provocative tests are presented in Table 1 for the anorexia nervosa patients and in Table 2 for the group of underweight adults. Table 3 gives the range and average values of growth hormone and glucose observed in the underweight adolescents and the control group. A few results stated as "0" mean in fact that in repeated assays no significant difference between these samples and the buffer control could be detected. The mean growth hormone and glucose response curves for each group of patients are represented in Fig. 1 . As it can be seen from Fig. 1, the shape of the growth hormone response curve differs from group to group despite very comparable average glucose response curves. The anorexia nervosa patients have a mean response curve distinctly lower than normal. Two patients only, of this group, had a growth hormone response curve above the lowest value of the normal range. The mean response curve of the 1 1 underweight adolescents is also flatened as compared to normal but not to such extent as seen in the previous group. Only 1 of the 1 1 patients of this group showed a frankly subnormal growth hormone response whereas the response of the ten others was in the lower part of the normal range, Contrasting with the other groups, the underweight adults show a much greater heterogeneity of growth hormone response to insulin induced hypoglycaemia (Table 2 ) . The mean growth hormone response curve for this group (Fig. 1) is also flatened as compared to normal and to the same extent as the mean response curve seen in the underweight adolescents. As it has been suggested that high basal levels of growth hormone induce an impairment of growth hormone response to a further acute stimulation, we looked for a negative correlation between the maximal increment of growth hormone (max. A GH) and the basal level. It appears from Fig. 2 that such a negative correlation does not exist for our underweight patients (group I, I1 and 111). There
293
is only 1 patient of the 37 included in this analysis whose poor growth hormone response can be related to a basal level of 2 6 ng/ml, which is abnormally high. This patient belonged to the group of underweight adults (patient 6 Table 2 ) .
Discussion It appears from our study that a high proportion of anorexia nervosa patients showed a poor growth hormone response to hypoglycaemic stimulation when compared to normal patients. This blunting of the growth hormone response does not seem to be directly related to the decrease in body weight: it was not observed in the two other groups of patients who were underweight to comparable degrees and no correlation could be found between response and absolute body weight or degree underweight. Such blunting has been reported in a few cases of anorexia nervosa: by Marks et aZ. (1965) in 2 cases and by Landon et al. (1966) in 3 out of 5 cases. These authors suggested that this phenomenon was probably related to the high growth hormone basal level, which is frequently found in fasting (Glick et al., 1965), chronic starvation (Alvarez et az., 1972) or improper food balance (Pimstone et al., 1966; Samuel and Deshpande, 1972). None of our anorexia nervosa patients had a high growth hormone basal level and this cause of unresponsiveness can be ruled out in our series. Moreover, Alvarez et al. (1972) observed, in a series of undernourished patients, supranormal basal levels and supranormal growth hormone response to insulin induced hypoglycaemia, Jacquet et al. (1970) observed in lean subjects supranormal basal levels and normal response to stimulation by arginine. Independently of any increasing effect on basal growth hormone level, improper food balance and starvation could have induced a relative unresponsiveness in the anorexia nervosa patients. Indeed, protein-poor diets have been shown by Merimee and Fineberg (1972) and Samuel et al. (1972) to diminish the stimulated growth hormone response. However, anorexia nervosa patients are known to avoid carbohydrates and fats especially but not meat (Slater and Roth, 1969) and in our patients almost normal serum protein levels hardly support any specific lack of proteins. Chronic starvation might have led to a diminished reserve of growth hormone as suggested by the observations made by Wilson (1954), Russfield and Somers (1965) in humans, and by Srebnik et al. (1959) , and Meites and Feil (1965) in animals. This last factor cannot entirely be excluded in our series. As no fully satisfactory explanation, related to chronic undernutrition, may be advanced for the blunting of the growth hormone response, one has to turn to other hypotheses. Hypothalamic nuclei are the site of production of hypophyseal hormone releasing factors and dopaminergic substances (Toivola and Gale, 1972; Himswath et a t . , 1972; Parkhie and Johnson, 1973; Harris, 1973). Experimental destruction of the ventromedial nucleus in primates results in a situation which closely resembles the one observed
294
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa
1, weight gain 8 kg) and in that case, the shape of in our anorexia nervosa patients: normal basal the growth hormone response curve did not change. levels of growth hormone and poor response to inIn three other cases (patients 1 , 9 and 10 Table I ) , sulin induced hypoglycaemia (Abrams et at., 1966). the treatment did not increase weight and growth There are other psychopathological situations than hormone response remained unchanged. As far as anorexia nervosa, not necessarily associated with other endocrine functions are considered, Russell undernourishment, which have been recently demon(1965) showed for example that amenorrhoea in strated to have impaired growth hormone secretion: anorexia nervosa was very resistant to treatment certain cases of depressive illness, as described and concluded that malnutrition accounts for only by Sachar et a t . (1971, 1972), and in maternal some abnormalities of hormone output and that the care deprived dwarfs (Powell et aZ., 1967). It likely site of the disturbance is the hypothalamus. might be pointed out that the relative lowering Force feeding does not of course correct the underof the mean growth hormone response curve seen in lying mental trouble probably responsible for the our group of underweight adults is precisely due endocrine abnormalities. to the poor growth hormone response seen in some The differential diagnosis between cachexia of the depressed patients belonging to this group leading eventually to hypopituitarism and true (see Table 2). hypopituitarism with cachexia may be difficult Although this hypothesis is attractive and in border line cases. The persistence of a normal supported by a recent study performed in 5 cases suprarenal function and the absence of real hypoof anorexia nervosa by Mecklenburg et aZ. (1974), thyroidism, as seen in anorexia nervosa, makes it should not lead one to neglect the possible the diagnosis of primary hypopituitarism most influence of other endocrinological factors which may play a role in growth hormone unresponsiveness. unlikely. Moreover, although frequently diminished, the growth hormone response in anorexia nervosa Hypercorticism inhibits the growth hormone resremains higher than in true hypopituitarism and ponse to insulin induced hypoglycaemia (Frantz cachexia occurs only rarely even in longstanding and Rabkin, 1 9 6 4 ) . Sachar et a;. (1971) found a hypopituitarism. relationship between increased cortisol production rate and growth hormone unresponsiveness in depressed patients. It is possible that an increase References, of the cortisol production rate, a parameter which has not been evaluated in our patients, was present Abrams, R.L., Parker, M.L., Blanco, S., Reichlin, in some of them. Their normal basal cortisol level S., Daughaday, W.M.: Hypothalamic regulation of and cortisol response to hypoglycaemia does not growth hormone secretion. Endocrinology 78, exclude a high cortisol production rate as sug605 (1966) gested by the report of Boyar et a2. (1974). Alvarez, L.C., Dimas, C.A., Castro, A., Rossman, A normal thyroid function is also necessary L.G. , Vanderlaan, E.F. , Vanderlaan, W.P. : Growth for a normal basal and stimulated growth hormone hormone in malnutrition. J. clin. Endocr. 3 4 , secretion (Brauman and Corvilain, 1968; Brauman 400 (1972) ot a t . , 19n). There is some evidence, as reviewed by Crisp (1970), that in anorexia nervosa an adapt- Boyar, R.M., Kapen, S., Finkelstein, J.W., Fukushima, D.K. , Weitzman, E.D. , Hellman, L. : Deive process to the hypocaloric diet leads to a lineation of the luteinizing hormone and corsecondary type of thyroid hypofunction. This would tisol abnormalities in anorexia nervosa. Abbe an additional contributory factor for growth hormone unresponsiveness and in fact the TSH and stract 33, SOC. clin. Invest. (1974) Brauman, H., Corvilain, J.: Growth hormone response the triiodothyronine levels were to some extent lower than normal in our anorexia nemosa patients. to hypoglycemia in myxedema. J. clin. Endocr. 28, 301 (1968) Oestrogens and gonadotrophins seem to play an enhancing role on basal and stimulated secretion Brauman, H., Smets, P., Corvilain., J.: Comparative of growth hormone (Frantz and Rabkin, 1965; Hillstudy of growth hormone response in normal subjects man et at., 1968; Merimee et aZ., 1967). Unresand in patients with primary myxedema or hyperponsiveness has been observed in cases of inthyroidism before and after treatment. J. clin. duced hypogonadism (Simon et at., 1967); gonadal Endocr. 3 6 , 1162 (1973) insufficiency has been quantitatively documented Copinschi, G., Wolter, R., Brauman, H.: Reproin anorexia nervosa (Russell, 1965) and therefore ductibility of plasma growth hormone deterthe possible interference of this factor can be minations between different laboratories. Clin. taken in account. chim. Atta 38, 207 (1972) Crisp, A.H. : Anorexia nervosa. "Feeding disorder!', Many factors seem thus to play a role in the relative unresponsiveness of the anorexia nervosa llnervous malnutrition" or "weight phobia"? Wld. patients : undernutrition, possible hypercorticism, Rev. Nutr. Diet 12, 452 (1970) relative hypothyroidism, depression of the gonadal Feighner, J.P., Robins, E., Guze, S.B., Woodruff, function and finally an ill defined hypothalamic R.A. Jr., Winokur, G., Munoz, R.: Diagnostic criteria for use in psychiatric research. Arch. insufficiency related to the mental disease. If undernutrition should play the prominent role, then gen. Psychiat. 26, 57 (1972) one could expect a better responsiveness after reFrantz, A.G., Rabkin, M.T. : Human growth hormone. feeding. We had no opportunity to retest the Clinical measurement, response to hypoglycemia patients after a substantial gain of weight during and suppression by corticosteroids. New Engl. force feeding except in one case (patient 2 Table J. Med. 271, 1375 (1964)
H. Brauman and F. Gregoire: Growth Hormone in Anorexia Nervosa Frantz, A.G., Rabkin, M.T.: Effect of estrogen and sex difference on secretion of human growth hormone. J. clin. Endocr. 25, 1470 (1965) Glick, S.M., Roth, J . , Yalow, R.S., Berson, S.A.: Immuno-assay of human growth hormone in plasma. Nature (London) 199, 784 (1963) Glick, S.M., Roth, J . , Yalow, R.S., Berson, S.A.: Regulation of growth hormone secretion. Recent Prog'r. Hormone Res. 21, 241 (1965) Harris, F.: Congenital hypothalamic lesion leading to growth hormone deficiency and destruction of the appetite satiety centre. Proc. roy. SOC. Med. 66, 217 (1973) Hayek, A . , Crawfard, J.D.: L-Dopa and pituitary secretion. J. clin. Endocr. 34, 764 (1972) Herbert, V., Lau, K.S., Gottlieb, C.W., Bleicher, S.J.: Coated charcoal immunoassay of insulin. J. clin. Endocr. 25, 1375 (1965) Hillman, J . , Hanrmond, J . , Sokola, J., Reiss, M.: Changes in plasma growth hormone levels in retarded children. J. ment. Defic. Res. 12, 294 (1 968) Himsworth, R.L., Carmel, P.W., Frantz, A.G.: The location of the chemoreceptor controlling growth hormone secretion during hypoglycemia in primates. Endocrinology 91, 217 (1972) Hoffman, W.S.: Rapid photoelectric method for determination of glucose in blood and urine. J. biol. Chem. 120, 51 (1937) Hunter, W.M., Greenwood, F.C.: Preparation of iodine-131 labelled growth hormone of high specific activity. Nature (London) 194, 495 ( 1962) Jacquet, Ph., Scornavachi, J.C., Vague, Ph., Vague, J.: Variations du taux plasmatique de l'hormone somatotrope humaine, au cours du test 5 l'arginine, dans les obisiths et les maigreurs. Ann. Endocr. (Paris) 31, 80 (1970) Parkhie, M.F.. , Johnson, H.D. : Hypothalamic growth hormone releasing factor: release and synthesis after exposure to a high ambient temperature. Proc. SOC. exp. Biol. (N.Y.) 142, 311 (1973)
295
Russell, G.F.M.: Metabolic aspects of anorexia nervosa. Proc. roy. SOC. Med. 58, 811 (1965) Russfield, A.B., Somers, S.C.: Malnutrition and tropic hormone storage in hypophysis. Arch. Path. 75, 564 (1963) Sachar, E.J., Finkelstein, J., Hellman, L.: Growth hormone response in depressive illness. Arch. gen. Psychiat. 25, 263 (1971) Sachar, E.J., Mushrush, G., Perlow; M., Weitzman, E.D., Sassin, J.: Growth hormone response to L-Dopa in depressed patients. Science 178, 1304 (1 972) Samuel, A.M., Deshpande, U.R.: Growth hormone levels in protein caloric malnutrition. J. clin. Endocr. 35, 863 (1972) Sherman, L., Kim, S., Benjamin, F., Kolodny, H.D.: Effect of chlorpromazine on serum growth-hormone concentration in man. New Engl. J. Med. 284, 72 (1971) Simon, S., Schiffer, M., Glick, S.M., Schwartz, E.: Effect of medroxyprogesterone acetate upon stimulated release of growth hormone in men. J. clin. Endocr. 27, 1633 (1967) Slater, R., Roth, M.: Clinical Psychiatry. Third edition, p. 124. London: Baillisre, Tindall and Cassell 1969 Srebnik, H.H., Nelson, M.M., Simpson, M.E.: Reduced growth hormone content in anterior pituitary of rats on protein free diets. Proc. SOC. exp. Biol. (N.Y.) 101, 97 (1959) Toivola, P.T.K., Gale, C.C.: Stimulation of growth hormone release by microinjection of norepinephrin into hypothalamus of baboons. Endocrinology 90, 895 (1972)
H. Brauman Service de Biologie Clinique HEpital Universitaire Brugman I, av. J.J. Crocq B-1020 Brussels Belgium
