- Email: [email protected]
Thyroid function in bulimia nervosa
Psychoneuroendocrinology, Vol. 21, No. 3, pp. 249-261, 1996 Published by Elsevier Science Ltd Printed in Great Britain 0306-4530/96 $15.00 + .00
Pergamon
PII: S0306-4530(96)00002-9 THYROID FUNCTION
IN BULIMIA NERVOSA
Margaret Altemus 1, Marion Hetherington 2, Bingham Kennedy 3, Julio Licinio 3 and Philip W. Gold 3 1Laboratory of Clinical Science, DIRP, NIMH, Bldg 10, Rm 3D41 MSC 1264, 10 Center Drive, Bethesda, MD 20892-1264, USA; 2Department of Psychology, University of Dundee, Dundee DD1 4HN, UK; and 3Clinical Neuroendocrinology Branch, DIRP, NIMH, Bldg 10, Rm 3S231, 10 Center Drive Bethesda, MD 20892, USA
(Received 4 October 1995; in final form 8 December 1995)
SUMMARY Basal thyroid-stimulating hormone (TSH) and thyroid hormone levels were evaluated in 18 women with bulimia nervosa during a period of active bingeing and vomiting and again after 7 weeks of abstinence from these behaviors and compared to measures in 27 control women. In 10 of the patients and 11 of the controls, the TSH nocturnal surge was calculated from hourly TSH measurements obtained in the afternoon from 1500 to 1900h and in the night from 2300 to 0400h. During the bingeing phase of the illness patients had lower total triiodothyronine (T3) values than controls (p < .001). After 7 weeks without binge eating or purging, patients had lower T3, total thyroxine (T4), free triiodothyronine, free thyroxine (FT4), reverse triiodothyronine and thyroid-binding globulin (TBG) values compared to controls (p < .01) and significant reductions in T3, T4, FT4 and TBG compared to themselves in the active phase of the illness (p < .02). The reduction in thyroid hormone levels was not due to a reduction in the nocturnal thyrotropin surge, since surge values did not differ between normals and patients at either phase of the illness. Bulimics in the bingeing phase of the illness showed a positive correlation between caloric intake and TSH values (p < .01), suggesting that food bingeing may stimulate thyroid activity. In sum, these results show a substantial reduction in thyroid hormone levels after 7 weeks of abstinence from bingeing and vomiting behaviors. Published by Elsevier Science Ltd. Keywords--Bulimia; Binge eating; Thyroid hormone; TSH; TSH surge.
INTRODUCTION Bulimia nervosa is a psychiatric syndrome characterized by recurrent episodes of binge eating followed by purging of the food, usually by self-induced vomiting, to prevent weight gain. Patients with this disorder are distressed by their eating pattern, but have great difficulty resisting urges to binge. Onset of the syndrome is usually preceded by a period of dieting (Mitchell et al., 1986) and accordingly often occurs in women with anorexia nervosa (Herzog et al., 1988). Prevalence of the disorder in young women is estimated at 1-2% (Drewnowski et al., 1988; Schotte & Stunkard, 1987) and the relapse rate after treatment is Address correspondence and reprint requests to: Building 10, R o o m 3D41, 9000 Rockville Pike, Bethesda, MD 20892, U S A (Tel: 301 396 3421; Fax: 301 402 0188; e-mail: [email protected]
gov). 249
250
M. Altemus et al.
high (Herzog et al., 1991). A recent report suggests that the incidence of bulimia nervosa may be increasing (Hoek et al., 1995). A number of studies have reported a reduction in serum total triiodothyronine (T3) in normal-weight patients with bulimia, while reductions in serum free thyroxine (FT4) and thyroid-stimulating hormone (TSH) are seen less consistently (Kiyohara et al., 1988; Obarzanek et al., 1991; Pirke et al., 1985). Reductions in T3 may be due to caloric restriction, since many normal-weight bulimics have lost weight from a premorbidly heavier state (Fairburn & Cooper, 1984; Garfinkel et al., 1980), and even mild underfeeding is known to lower T3 (Garrel et al., 1984; Visser et al., 1987). Women with bulimia typically diet between binges and compared to non-dieting women matched for body size maintain their weight on fewer calories per day (Gwirtsman et al., 1989; Obarzanek et al., 1991) and have a lower resting metabolic rate (Devlin et al., 1990; Obarzanek et al., 1991). In addition, normal-weight bulimics have other metabolic signs of semi-starvation, including low fasting glucose levels, increased beta-hydroxybutyric acid and free fatty acid concentrations and reduced norepinephrine responses to orthostatic challenge (Devlin et al., 1990; Obarzanek et al., 1991; Pirke et al., 1985). Accordingly, changes in thyroid activity in bulimia nervosa seem to resemble changes seen in starvation and fasting, which usually fit a euthyroid sick pattern (Chopra et al., 1975; Chopra & Smith, 1975; Fichter et al., 1986; Portnay et al., 1974; Suda et al., 1978; Wartofsky & Burman, 1982). Prior studies of bulimic women are conflicting, but most have noted reductions in T3 while TSH levels are usually normal (Devlin et al., 1990; Kiyohara et al., 1988; Obarzanek et al., 1991; Pirke et al., 1985). Similarly, underweight subjects with anorexia nervosa have reductions in T3 and elevations in reverse triiodothyronine (rT3) which correct with refeeding, while TSH and FT4 typically do not differ from normals and do not change throughout the illness and recovery (Croxson & Ibbertson, 1975; DeRosa et al., 1983; Leslie et al., 1978; Miyai et al., 1975; Moore & Mills, 1979; Moshang et al., 1975). Alternatively, in the light of a number of studies suggesting alterations in hypothalamic stimulation of luteinizing hormone (Devlin et al., 1989; Pirke et al., 1987; Weltzin et al., 1994) and adrenocorticotropic hormone (Mortola et al., 1989) secretion in bulimia nervosa, decreased thyroid activity may be due to hypothalamic or central hypothyroidism, which is characterized by loss of the nocturnal TSH surge, low levels of FT4 and reduced or normal basal concentrations of TSH (Bartalena et al., 1987; Caron et al., 1986; Rose et al., 1990). The nocturnal TSH surge is thought to provide the major stimulatory signal for thyroid hormone production by the thyroid gland. Most often, central hypothyroidism results from hypothalamic and pituitary lesions (Caron et al., 1986; Fukuda & Greer, 1975; Rose et al., 1990), but can also occur in very severe non-thyroidal illness (Bartalena et al., 1990a,b; Romijn & Wiersinga, 1990), fasting (Romijn et al., 1990), depression (Bartalena et al., 1990c) and Cushing's disease (Bartalena et al., 1991). In order to further characterize the thyroid abnormalities in bulimia nervosa and the hormonal response to recovery from the illness, we measured basal thyroid hormones and the nocturnal thyrotropin surge in patients during a phase of bingeing and vomiting, again after 7 weeks of abstinence from the behaviors, and in controls. Because patients with bulimia nervosa often have derangements of hydration and fluid regulation (Demitrack et al., 1992), and volume shifts may influence levels of proteins and peptides in plasma (DeCostre et al., 1971), we also measured plasma renin activity and aldosterone in bulimic patients in both phases of the illness and in control subjects.
Thyroid Function in Bulimia Nervosa
251
MATERIALS AND M E T H O D S
Subjects Eighteen normal-weight bulimic women and 27 control women were included in this study. Bulimic women met DSM-IIIR (American Psychiatric Association, 1987) criteria for bulimia nervosa and had a mean percentage average body weight (ABW) of 95% (range 80117%). ABW is the percentage of the mean weight of individuals of the same sex, age and height (Society of Actuaries and Association of Life Insurance Medical Directors of America, 1979). Mean age was 25 years (range 20-34). Mean duration of illness was 7.4 years (range 2-13) and reported mean frequency of binge eating prior to admission was 27 episodes per week (range 10-75). Bulimic subjects abstained from taking any medications for at least 1 month before admission and had not had a weight change of more than 10% in the preceding 6 months. The highest lifetime ABW for this group of subjects was 104% (range 89-126%). One patient had a history of anorexia nervosa. No patient had received prior hormonal therapy. Patients continued their binge eating and vomiting during the first week of hospitalization. All subjects binged at least once per day during the first week of hospitalization and vomited after each binge. Patients were then closely supervised to enforce abstinence from bingeing and vomiting for the next 7 weeks of hospitalization. Patients selected their own food throughout the hospitalization, although they were required to keep their weights within 1 kg of their admission weights. They were treated with supportive psychotherapy, behavioral therapy and nutritional counseling. Controls were admitted to the hospital the day before the study in order to allow accommodation to the setting. Controls also selected their own food. Controls were in good health and taking no medications and had no history of eating disorders, dieting behavior or excessive exercise as assessed by clinical interview and responses to the Eating Inventory questionnaire (Stunkard & Messick, 1985). Controls were similar to patients in age (mean 23 years; range 18-35) and ABW (mean 98%; range 84-118%). All bulimic subjects and controls were caucasian. All subjects had normal screening blood tests, including electrolytes, renal and hepatic functions, and complete blood counts. The study was approved by the Institutional Review Board of the National Institute of Mental Health and all subjects gave informed consent before participation in the study.
Clinical data Caloric and macronutrient intake of the bulimic patients was calculated during the first and seventh weeks of hospitalization by subtraction of uncomsumed calories from calories presented, as described previously (Peterson et al., 1986). Twelve of the controls were also hospitalized for an additional 4 days to obtain caloric intake data. The Eating Inventory (Stunkard & Messick, 1985) was used to assess restrictive and uncontrolled eating behaviors in both bulimic women and control women. The Hamilton Depression Rating Scale was used to quantitate depressive symptoms (Hamilton, 1967).
Blood sampling Fasting blood samples drawn at 0900h after 1 h in the supine position were used to determine basal levels of hormones. Bulimic patients had these blood samples drawn after 1 week of bingeing and purging daily on the inpatient unit and again after 7 weeks of normal eating on the inpatient unit. Serum for T3 and total thyroxine (T4) was assayed within 24 h,
252
M. Altemuset al.
and blood for other thyroid hormone assays was collected in tubes with EDTA as an anticoagulant. Blood was centrifuged within 6 h and plasma stored at --70°C until assayed. For the subjects who had nocturnal TSH sampling, blood was drawn on the same day as the basal samples through an indwelling IV catheter with a heparin lock at hourly intervals between 1500 and 1900h and between 2300 and 0400h. This protocol of blood sampling provides an accurate assessment of the nocturnal serum TSH surge (Caron et al., 1986). Subjects were able to ambulate during the day and care was taken not to disturb sleep during the night-time blood draws. Calculation of the nocturnal TSH surge The nadir of TSH was designated as the lowest average of three consecutive values in the blood samples taken between 1500 and 1900h. The peak TSH was designated to be the highest average of three consecutive values in blood samples taken between 2300 and 0400h. The nocturnal TSH surge was calculated by subtracting the nadir TSH value from the peak value, dividing the difference by the nadir value, then multiplying the result by 100 in order to express the surge as a percentage rise over the nadir. Hormone assays T4, T3 and thyroid-binding globulin (TBG) were measured in fresh serum using an immunofluorescence assay for T4 (Abbott Laboratories, Irving, TX, USA), and radioimmunoassays for T3 (Kallestad Laboratories, Austin, TX, USA) and TBG (Ciba-Corning, Medfield, MA, USA). Intra-assay and inter-assay coefficents of variation for these three assays were less than 5%. Two bulimics and eight controls did not have TBG measurements. All other thyroid hormone assays were performed on frozen plasma. TSH was measured in duplicate using a sensitive immunoradiometric assay (CIBA-Serono, Walpole, MA, USA) having a limit of detection of 0.2 mU/L. Baseline TSH samples for half of the bulimics and half of the controls were assayed in each of two runs. To measure the TSH surge, active and abstinent samples from each patient were run together with samples from one or two control patients in each assay run. The intra-assay and inter-assay coefficients of variation were less than 5% and 6.2%, respectively. FT4 and free triiodothyronine (FT3) were measured by radioimmunoassay in single runs using commercial kits (Diagnostic Products Corporation, Los Angeles, CA, USA). Intra-assay coefficents of variation for both assays were less than 5%. rT3 was also measured by radioimmunoassay using a commercial kit (Serono Diagnostics, Norwell, MA, USA). Half of the bulimic samples and half of the control samples for rT3 were measured in each of two runs. Intra-assay and inter-assay coefficients of variation were less than 5% and 3%, respectively. Plasma renin and aldosterone were also measured in 15 patients and 10 controls by radioimmunoassay (Hazelton Laboratories, Fairfax, VA, USA). All samples were run in a single assay. Intra-assay coefficients of variation were less than 10% for both assays. Statistics Results were analyzed by two-tailed t-tests for paired and unpaired data. Pearson product moment correlations were calculated to determine relationships between variables. The slope of the regression line of two variables which were significantly related was determined by the least squares method and significant differences between slopes were determined using 95% confidence intervals. All values are given as mean _+SD. The level of significance was taken as p < .05.
Thyroid Function in Bulimia Nervosa
253
RESULTS
Clinical Variables Bulimic patients scored significantly higher than controls of all three subscales of the eating inventory: cognitive restraint: 12.7 _ 3.7 vs. 6.5 _+4.1 (t = 3.8, p < .001); disinhibition: 12.9 _+2.9 vs. 3.7 _ 2.9 (t = 9.7, p < .0001); and hunger: 11.1 _ 2.9 vs. 3.9 _+2.6 (t = 9.7, p < .0001). HamiltOn Depression Scale ratings of bulimic patients revealed a moderate degree of depression on admission ( 1 6 . 2 _ 7.5), which did not change after 6 weeks of hospitalization (15.0 _ 10.0, t = .05, p = .96). There was no significant difference between bulimics and controls in age (25 _+4 years vs. 23 _+4 years, t = - 1.1, p = NS) or %ABW (95.0 _+ 11% vs. 98 _ 10%, t = .95, p = NS). The weight of the bulimics did not change significantly during the 7 weeks of hospitalization. Mean daily caloric intake during the week of bingeing and vomiting was 12,216 _+6021 kcal/day. Frequency of purged meals (binges and smaller meals) during the first week of hospitalization did not differ from the reported frequency of binges prior to admission, when subjects were not asked to distinguish binges from smaller size purged meals (27 _ 20 vs. 21 _+8 binges/week, t = - 1.0, p = .34). As described in a separate report, the proportion of calories consumed as carbohydrate was similar in actively bulimic women and controls, but actively bulimic women consumed a significantly smaller proportion of calories as protein and a significantly greater proportion as fat (Hetherington et al., 1994). Caloric intake of the bulimics during the seventh week of hospitalization tended to be lower than mean caloric intake of the weight-matched controls, who were hospitalized for 4 days to obtain caloric intake data (1778 +_243 kcal/day vs. 1957 _ 296 kcal/day, t = 1.7, p = .09). The proportion of macronutrients consumed did not differ between bulimic women during the seventh week of hospitalization and controls.
Thyroid Hormone Values Thyroid hormone results are displayed in Table I. Compared to controls, patients with bulimia nervosa studied during the active phase of bingeing and vomiting showed evidence of the low T3 syndrome. Specifically, these patients had a significantly lower serum T3 (t = 3.8,p < .001) and no significant difference in FT3 (t = 1.5,p = .14), T4 (t = 1.5,p = .12) and FT4 (t = 1.4, p = .16). Their rT3 (t = .13, p = .90) and T B G (t = .7, p = .52) values also did not differ from controls. The basal 0800h TSH (t = .35, p = .72), the circadian TSH nadir (t = .48, p = .63) and peak (t = - . 2 8 , p = .78), and the calculated nocturnal TSH surge (t = - 1.3, p = .20) all did not differ significantly from those of controls. In contrast to the low T3 syndrome seen during active bingeing and vomiting, when the patients were re-studied after 6 weeks of supervised restraint from all pathological eating behaviors, they showed evidence of decreased thyroid function, including significant reductions compared to controls in T3 (t = 7.5, p < .001), FT3 (t = 3.0, p < .01), T4 (t = 5.4, p < .001), FT4 (t = 3.1, p < .01) and rT3 (t = 3.1, p < .01). In addition, serum TBG was significantly decreased (t = 2.2, p = .03). Despite these indices, suggestive of hypothyroidism during abstinence from bingeing, there was no change in the basal 0800h TSH (t = --.3, p = .76), the circadian TSH nadir (t = - . 6 , p = .52) or peak (t = - 1 . 0 , p = .26), or the nocturnal TSH surge (t = --1.1, p = .30). Direct longitudinal comparison in bulimic patients of thyroid function tests performed during bingeing and vomiting and again after supervised restraint from these behaviors showed significant reductions in T3 (t = 6.0, p < .001), FT4 (t = 2.6, p --- .02), T4 (t = 6.2, p < .001) and rT3 (t = 3.2, p < .01) and a t, end toward reduction in FT3 (t = 2.0, p = .06)
~N .~-
~~
cm ,,~
+I
+I
+I
+I
+I
+I
N
tN
¢II ,--~
[.,, I_. °~
oO +I
+I
-I-I
~0
~
oo
+I
+I
-I-I
+I
+I
÷I
+I
+I
+I
+I
+I
+I
+I
+I
"¢:I
++
. ,,,,,i ,-,,i
•
+I
+I
~
+I
~. +I
v
+
0
v
+I
+I
+I
+I
+I
+!
+I
+I
+I
÷I
4-I
+I
~.~.
.~"
¢,,)
2
..~:~
+,
+I
+I
+I
+I
++
::j
+I
÷I
+I
-I-I
++
-I-I
~
+I
+I
v
o
÷I
.~
v
Thyroid Function in Bulimia Nervosa
255
am 3 ¸
.J ::)
[] []
E
El []
-1¢/) I-
El []
[]
[]
[]
o
I
!
10000
20000
30000
CALORIES PER DAY Fig. 1. Relationship between 0800h plasma TSH and average daily caloric consumption in women in the active phase of bulimia nervosa.
and T B G (t = 1.88, p = .08) during the period of normal eating. H o w e v e r , there w a s no significant change in basal 0800h TSH, the circadian T S H nadir (t = -- 1.5, p = .16) or peak (t = - 1.3, p = .89), or the calculated nocturnal T S H surge (t = --.13, p = .89) values.
Renin and Aldosterone Values In the bingeing phase of the illness, c o m p a r e d to controls, w o m e n with bulimia nervosa had elevated levels of both plasma renin activity (0.75 _+ 0.08 ng/1 s vs. 0.23 _+ 0.06 ng/1 s, t = 2.1, p = .05) and aldosterone (510 _+ 410 pmol/l vs. 150 _+ 60 pmol/1, t = 2.6, p = .02). These elevated levels fell significantly after cessation of binge eating and purging (renin: 0.75 _+ 0.08 n g / l s vs. 0.24 _+ 0.19 ng/l s, t = 3.0, p < .01; aldosterone: 510 _+ 410 pmol/1 vs. 170 _+ 88 pmol/1, t = 3.5, p < .01), resulting in levels not significantly different from normals (renin: t = .2, p = .8; aldosterone: t = .6, p = .53).
Relationship Among Variables During the period of active bingeing and v o m i t i n g there w a s a positive correlation between the daily caloric intake and basal 0800h T S H (n = 18, r = .68, p < .01) (Fig. 1), T S H nadir (n = 10, r = .72, p < .02) and T S H peak (n = 10, r = .83, p < .01), but not the nocturnal T S H surge (n = 10, r = .19). There were no significant correlations between caloric intake and any measures of T S H secretion in non-bingeing bulimics or controls. There were no significant correlations between other thyroid indices and caloric intake in patients in
256
M. Altemus et al.
either phase of the illness or in controls. Binge frequency in the active phase of the illness was not significantly correlated with any thyroid indices. There was a positive correlation between FF4 and the TSH surge in controls (n = 11, r = .58, p < .05) and in abstinent bulimics (n = 10, r = .73, p < .02), but not in active bulimics. The slope of the regression line did not differ significantly between controls and abstinent bulimics. There were no significant correlations between depression ratings and any of the thyroid hormone measures or the nocturnal TSH surge.
DISCUSSION Patients with bulimia nervosa exhibited the low T3 syndrome while actively bingeing and vomiting. When re-studied after supervised restraint from bingeing and vomiting, there were more pronounced reductions in thyroid hormone levels; all peripheral indices of thyroid gland function were significantly reduced in this phase compared to controls. Moreover, patients studied during supervised restraint also showed a significant reduction in all peripheral indices when compared to themselves in the bingeing phase. Throughout, all indices of TSH secretion were normal. Moreover, even though peripheral indices of thyroid function shifted significantly with the phase of the illness, indices of TSH secretion did not change. The finding of reductions in T3, T4, F r 3 and FT4 during the non-binge eating phase agrees with previous studies of bulimics after 1-3 weeks of inpatient hospitalization (Kiyohara et al., 1988; Obarzanek et al., 1991; Pirke et al., 1985). There have been no comparable studies of thyroid function in bulimics during a documented active phase of the illness; however, Devlin et al. (1989) found normal levels of T3 and T4 in a mixed group of outpatient and newly hospitalized bulimic women. Because serum TBG levels also fell during abstinence from binge eating, a component of the reduction in T3 and T4 levels during abstinence is attributable to a reduction in the bound fraction of hormones. A reduction in TBG levels during the non-binge eating phase may also have contributed to the observed reductions in FT3 and FT4 values, since current immunoassay kits often underestimate concentrations of free thyroid hormone and results can vary with changes in bound thyroid hormone concentration (Nelson et al., 1994; Wong et al., 1992). The fall in TBG during abstinence from binge eating may be due to relative nutritional deprivation or to dilutional factors. As a group, the bulimic women appeared to have significant volume contraction during the phase of bingeing and vomiting, based on the renin and aldosterone levels. It is possible that some component of the change in thyroid hormone as well as TBG levels from the active phase to the abstinent phase could be dilutional. Although there are few data concerning the effect of dehydration on plasma hormone levels, diurnal changes in plasma proteins and peptides have been linked to volume shifts over the course of the day (DeCostre et al., 1971). On the other hand, one study has reported no change in plasma cortisol after 48 h of water deprivation in sheep (Matthews & Parrott, 1991), and in a 24 h serial blood sampling study in bulimic women there was no change in plasma cortisol between active and abstinent phases of the illness (Altemus et al., 1992). The thyroid hormone data in both phases of the illness appear to be explained at least in part by the euthyroid sick syndrome, in which reductions in T3 are not accompanied by elevations in TSH (Carter et al., 1974). On the other hand, the reduction in T3 in the euthyroid sick syndrome is thought to result from decreased metabolism of T4 to T3 (Chopra
Thyroid Function in Bulimia Nervosa
257
et al., 1975), with increases in rT3, while TSH and FT4 remain at normal levels (Bermudez et al., 1975; Chopra et al., 1974). Bulimic women in this study did not fit this pattern since they did not have increases in rT3 at any point. The bulimic women in this study also did not exhibit the cardinal feature of the syndrome of central hypothyroidism, since the TSH surge was preserved in both phases of the illness. On the other hand, bulimic women show delayed plasma TSH responses to synthetic thyrotropin-releasing hormone (TRH) (Coiro et al., 1990; Gwirtsman et al., 1983; Kiriike et al., 1987; Kiyohara et al., 1988; Levy et al., 1988), which are qualitatively similar to responses seen in patients with central hypothyroidism (Beck-Peccoz et al., 1985). The lack of change in indices of TSH secretion in the abstinent bulimic patients despite substantial changes in thyroid hormone values over the course of hospitalization is also consistent with a primary impairment of thyroid gland function or a reduction in the bioactivity of endogenous TSH. One study which reported the Tr3 response to TRH infusion in bulimics (Beck-Peccoz et al., 1985) found an impaired Tr3 response to the TSH elevation following TRH infusion, again consistent with reduced biological activity of TSH or impaired thyroid gland responses to endogenous TSH. The reduction in thyroid activity in bulimia nervosa, particularly during the non-binge eating phase, may be appropriately adaptive since this study and others (Garfinkel et al., 1980; Obarzanek et al., 1991) have noted a reduced caloric intake in bulimics during a period of restraint from binge eating when compared to normals matched for body size. In the face of this decreased energy intake, the set point for feedback regulation of TSH may be lowered to reduce thermogenesis as occurs in starvation and anorexia nervosa (Croxson & Ibbertson, 1975). In addition to presenting with weights which were on average 5.2 +_5.0 kg below their lifetime highest body weight, bulimic women may have lost a small amount of lean body mass over the course of this study, since their weight did not change while they substantially increased their hydration. The strong correlations between indices of TSH secretion and caloric intake during the bingeing phase suggests that food ingestion plays a role in creating the relative increase in thyroid hormone levels during the active phase of the illness. Although much of the binge food is vomited and never actually absorbed, we speculate that preabsorbtive effects of food ingestion such as insulin secretion and the associated activation of the sympathetic nervous system (Leblanc, 1992), both of which have been shown to be greatly exaggerated during binge meals (Kaye et al., 1989), could stimulate thyroid hormone activity during the phase of bingeing and vomiting. Data that norepinephrine is a potent stimulus to TRH release (Engler et al., 1982; Palkovits et al., 1980) suggest that preabsorptive sympathetic activation could act to mitigate a central hypothyroidism induced by food restriction. In support of this model, bulimic patients have low cerebrospinal fluid and plasma norepinephrine levels during restraint from bingeing and normal levels during periods of bingeing and vomiting (Kaye et al., 1990). Further indirect evidence that preabsorbtive insulin responses can increase thyroid hormone activity comes from data that, compared to other macronutrients, carbohydrates preferentially induce insulin release and that only reduction of the proportion of carbohydrate in the diet can lower thyroid hormone activity (Davidson & Chopra, 1979; Spaulding et al., 1976). As a corollary, we have shown previously that the phase of active bingeing and vomiting is also associated with a significantly higher resting metabolic rate, even if weight is held stable (Altemus et al., 1991). In this regard, one could construe the bingeing and vomiting of bulimia nervosa to, in part, serve the function of increasing thyroid function and metabolic
258
M. Altemus et al.
rate without weight gain, and avoiding the reduction in thyroid hormone activity known to be associated with dieting. Additional longitudinal studies are needed to determine whether this apparent increase in thyroid hormone activity during the active phase of the illness is also associated with relative enhancement of mood or cognitive function. Our demonstration of state-related changes in thyroid hormone levels may help to explain the discrepant findings in the literature where frequency of bingeing and vomiting and duration of hospitalization before testing were not well controlled. Detection of abnormalities in the present study was also probably enhanced by selecting patients with at least a moderate severity of illness and careful screening of normals to exclude any with dieting behaviors. Although 7 weeks should be adequate time to come to a steady state as far as half-life of the thyroid hormones is concerned, it is possible that the adjustment to cessation of bingeing and vomiting occurs over a longer time period and the altered state observed here is temporary and resolves at a later point. In summary, these data indicate that normal-weight bulimic women have reduced thyroid hormone levels during abstinence from bulimic behaviors and that partial correction of thyroid hormone abnormalities occurs during periods of active bingeing and vomiting. These findings are consistent with reports of decreases in resting metabolic rate and reduced caloric requirements during abstinence from bulimic behaviors. Low thyroid hormone activity and a low metabolic rate could predispose these patients to weight gain and thereby exacerbate their preoccupation with weight control. Restrained caloric intake may also stimulate hunger, which is associated with onset of binges (Hetherington et al., 1994). Moreover, partial correction of hypothyroidism by bingeing and vomiting may be physiologically reinforcing and contribute to the development of this condition during periods of dieting as well as the propensity for relapse after treatment.
Acknowledgements:We wish to acknowlege the helpful review of this manuscript by Bruce Nisula, M.D. and the assistance of Yung-Mei Leong in manuscript preparation.
REFERENCES Altemus M, Hetherington M, Flood M, Licinio J, Nelson M, Bernat A, Gold P (1991) Decrease in resting metabolic rate during abstinence from bulimic behavior. Am J Psychiat 148:1071-1072. Altemus M, Licinio J, Jolkowsky L, Gold P (1992) Stimulation of the hypothalamic-pituitary-adrenal axis by bulimic behaviors. Proceedings of the 145th Annual Meeting American Psychiatric Association, Washington, DC. American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. Revised. American Psychiatric Association, Washington, DC. Bartalena L, Martino E, Brandi L, Falcone M, Pacchiarotti A, Ricci C, Bogazzi F, Grasso L, Mammoli C, Pinchera A (1990) Lack of nocturnal serum thyroptropin surge after surgery. J Clin Endocrinol Metab 70:293-296. Bartalena L, Martino E, Falcone M, Buratti L (1987) Evaluation of the nocturnal serum thyroptropin (TSH) surge, as assessed by TSH ultrasensitive assay, in patients receiving long term L-thyroxine suppression therapy and in patients with various thyroid disorders. J Clin Endocrinol Metab 65:1265-1271. Bartalena L, Martino E, Petrini L, Velluzzi F, Loviselli A, Grasso L, Mammoli C, Pinchera A (1991) The nocturnal serum thyrotropin surge is abolished in patients with adrenocorticotropin (ACTH)dependent or ACTH-independent Cushing's syndrome. J Clin Endocrinol Metab 72:1195-1199. Bartalena L, Pacchiarotti A, Palla R, Antonagelli L, Mammoli C, Monzani F, DeNegri F, Panichi V, Martino E, Baschieri L (1990) Lack of nocturnal serum thyrotropin (TSH) surge in patients with chronic renal failure undergoing regular maintenance hemofiltration: A case of central hypothyroidism. Clin Nephrol 34:30-34.
Thyroid Function in Bulimia Nervosa
259
Bartalena L, Placidi G, Martin E, Falcone M, Pellegrini L, Dell'Osso L, Pacchiarotti A, Pinchera A (1990) Nocturnal serum thyrotropin (TSH) surge and the TSH response to TSH-releasing hormone: dissociated behavior in untreated depressives. J Clin Endocrinol Metab 70:650-655. Beck-Peccoz P, Amir S, Menezes-Ferreira M, Faglia G, Weintraub B (1985) Decreased receptor binding of biologically inactive thyrotropin in central hypothyroidism: effect of treatment with thyrotropin releasing hormone. New Engl J Med 312:1085-1090. Bermudez F, Surks M, Oppenheimer J (1975) High incidence of decreased serum triiodothyronine concentration in patients with n-on-thyroidal disease. J Clin Endocrinol Metab 41:27-40. Caron P, Nieman L, Rose S, Nisula B (1986) Deficient nocturnal surge of thyrotropin in central hypothyroidism. J Clin Endocrinol Metab 62:960-964. Carter J, Eastman C, Corcoran J, Lasarus L (1974) Effect of severe chronic illness on thyroid function. Lancet 2:971-974. Chopra I, Chopra U, Smith S, Reza M, Solomon D (1975) Reciprocal changes in serum concentrations of 3,3',5'-triiodothyronine (reverse T3) and 3,3',5'-triiodothyronine (T3) in systemic illnesses. J Clin Endocrinol Metab 41:1043-1049. Chopra I, Smith S (1975) Circulating thyroid hormones and thyrotropin in adult patients with proteincalorie malnutrition. J Clin Endocrinol Metab 40:221-227. Chopra I, Solomon D, Chopra U~ Young R, Teco GC (1974) Alterations in circulating thyroid hormones and thyrotropin in hepatic cirrhosis: evidence for euthyroidism despite subnormal serum triidothyronine. J Clin Endocrinol Metab 39:501-511. Coiro V, Capretti L, Volpi R, d'Amato L, Marchesi C, DeFerri A, Rossi G, Bianconi L, Marcato A, Chiodera P (1990) Growth hormone responses to growth hormone-releasing hormone, conidine, and insuline-induced hypoglycemia in normal weight bulimic women. Neuropsychobiology 23:8-14. Croxson M, Ibbertson H (1975) l o w serum triiodothyronine (t3) and hypothyroidism in anorexia nervosa. J Clin Endocrinol Metab 44:167-174. Davidson M, Chopra I (1979) Effect of carbohydrate and noncarbohydrate sources of calories on plasma 3,5,5'-triiodothyronine concentrations in man. J Clin Endocrinol Metab 48:577-581. DeCostre P, Buhler U, DeGroot LJ, Refetoff S (1971) Diurnal rhythm in total serum thyroxine levels. Metabolism 20:782-791. Demitrack M, Kalogeras K, Altemus M, Pigott T, Listwak S, Gold P (1992) Plasma and cerebrospinal fluid measures of arginine vasopressin secretion in patients with bulimia nervosa and in healthy subjects. J Clin Endocrinol Metab 74:1277-1283. DeRosa G, Corsello S, DeRosa E, Casa SD, Ruffilli M, Grasso P, Pasargiklian E (1983) Endocrine study of anorexia nervosa. Exp Clin Endocrinol 82:160-172. Devlin M, Walsh B, Kral J, Heymsfield S, Pi-Sunyer F, Dantzic S (1990) Metabolic abnormalities in bulimia nervosa. Arch Gen Psychiat 47:144-148. Devlin M, Walsh B, Roose S, Katz J, Linkie D (1989) Gonadotropin patterns in eating disorders. Psychiat Res 28:11-24. Drewnowski A, Hopkins S, Kessler R (1988) The prevalence of bulimia nervosa in the US college student population. Am J Public Health 78:1322-1325. Engler D, Chad D, Jackson I (1982) Thyrotropin-releasing hormone in the pancreas and brain of the rat is regulated by central noradrenergic and dopaminergic pathways. J Clin Invest 69:1310-1320. Fairburn C, Cooper P (1984) The clinical features of bulimia nervosa. Br J Psychiat 144:238-246. Fichter M, Pirke K, Holsboer F (1986) Weight loss causes neuroendocrine disturbances: experimental study in healthy starving subjects. Psychiat Res 17:61-72. Fukuda J, Greer M (197) The effect of basal hypothalamic deafferentation on nycthemeral rhythm of plasma TSH. Endocrinology 975749-752. Garfinkel P, Moldofsky H, Garner D (1980) The heterogeneity of anorexia nervosa: bulimia as a distinct subgroup. Arch Gen Psychiat 37:1036-1040. Garrel D, Todd L, Pugeat M, Calloway D (1984) Hormonal changes in normal men under marginally negative energy balance. Am J Clin Nutr 39:930-936. Gwirtsman H, Kaye W, Obarzanek E, George D, Jimerson D, Ebert M (1989) Decreased caloric intake in normal-weight patients with bulimia: comparison with female volunteers. Am J Clin Nutr 49: 86-92. Gwirtsman H, Roy-Byrne P, Yager J, Gerner R (1983) Neuroendocrine abnormalities in bulimia. Am J Psychiat 140:559-563.
260
M. Altemus et al.
Hamilton M (1967) Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 6:278-296. Herzog D, Keller M, Lavori P (1988) Outcome in anorexia nervosa and bulimia. J Nerv Ment Dis 176:131-143. Herzog D, Keller M, Lavori P, Sacks N (1991) The course and outcome of bulimia nervosa. J Clin Psychiat 52suppl:4-8. Hetherington MM, Altemus M, Nelson ML, Bernat AS, Gold PW (19~4) Eating behavior in bulimia nervosa: multiple meal analysis. Am J Clin Nutr 60:864-873. Hoek HW, Barelds AIM, Bosveld JJF, van der Graaf Y, Limpens VEL, Maiwald M, Spaaij CJK (1995) Impact of urbanization on detection rates of eating disorders. Am J Psychiat 152:1272-1278. Kaye W, Gwirtsman H, George D (1989) The effect of bingeing and vomiting on hormonal secretion. Biol Psychiat 25:768-780. Kaye W, Gwirtsman H, Goerge D, Jimerson D (1990) Disturbances of noradrenergic systems in normal weight bulimia: relationship to diet and menses. Biol Psychiat 27:4-21. Kiriike N, Nishiwaki S, Izumiya Y, Maeda Y, Kawakita Y (1987) Thyrotropin, prolactin, and growth hormone responses to thyrotropin-releasing hormone in anorexia nervosa and bulimia. Biol Psychiat 22:167-176. Kiyohara K, Tamai J, Kobayashi N, Nakagawa T (1988) Hypothalamic-pituitary-thyroidal alterations in bulimic patients. Am J Clin Nutr 47:805-809. Leblanc J (1992) Cephalic phase response to food and its effect on feeding behavior. In: Anderson G, Kennedy S (Eds) The Biology of Feast and Famine. Academic Press, San Diego, pp 47-60. Leslie R, Isaacs A, Gomez J, Raggatt P, Bayliss R (1978) Hypothalamic-pituitary-thyroid function in anorexia nervosa: influence of weight gain. Br Med J 2:526-528. Levy A, Dixon K, Malarkey W (1988) Pituitary response to TRH in bulimia. Biol Psychiat 23:476484. Matthews S, Parrott R (1991) Dehydration, but not vasopressin infusion, enhances the adrenocortical response of sheep to corticotropin-releasing hormone or restraint. Acta Endocrinol 125:556-562. Mitchell J, Hatsukami D, Pyle R, Eckert E (1986) The bulimia syndrome: course of the illness and associated problems. Compr Psychiat 27:165-170. Miyai K, Yamamoto T, Azukizawa M, Ishibashi K, Kumahara Y (1975) Serum thyroid hormones and thyrotropin in anorexia nervosa. J Clin Endocrinol Metab 40:334-338. Moore R, Mills I (1979) Serum T3 and T4 levels in patients with anorexia nervosa showing transient hyperthyroidism during weight gain. Clin Endocrinol 10:443-449. Mortola J, Rasmussen D, Yen S (1989) Alterations of the adrenocorticotropin-cortisol axis in normal weight bulimic women: evidence for a central mechanism. J Clin Endocrinol Metab 68:517-522. Moshang T, Parks J, Baker L, Vaidya V, Utiger R, Bongiovanni A, Syder P (1975) Low serum triiodothyronine in patients with anorexia nervosa. J Clin Endocrinol Metab 40:470--473. Nelson JC, Weiss RM, Wilcox RB (1994) Underestimates of serum free thyroxine (T4) concentrations by free T4 immunoassays. J Clin Endocrinol Metab 79:76-79. Obarzanek E, Lesem M, Goldstein D, Jimerson D (1991) Reduced resting metabolic rate in patients with bulimia nervosa. Arch Gen Psychiat 48:456-462. Palkovits M, Zaborszky L, Feminger A, Mezey E, Fekete M, Herman J, Kanyicska B, Szabo D (1980) Noradrenergic innervation of the rat hypothalamus: experimental biochemical and electron microscope studies. Brain Res 191:161-171. Peterson R, Kaye W, Gwirtsman H (1986) Comparison of calculated estimates and laboratory analysis of food offered to hospitalized eating disorder patients. J Am Diet Assoc 86:490--492. Pirke K, Fichter M, Schweiger U, Fruth C, Streitmatter A, Wolfram G (1987) Gonadotropin secretion pattern in bulimia nervosa. Int J Eating Disord 6:655-661. Pirke K, Pahl J, Schweiger U, Warnhoff M (1985) Metabolic and endocrine indices of starvation in bulimia: a comparison with anorexia nervosa. Psychiat Res 15:33-39. Portnay G, O'Brian J, Bush J, Vagenakis A, Azizi F, Arky R, Ingbar S, Braverman L (1974) The effect of starvation on the concentration and binding of thyroxine and triiodothyronine in serum and on the response to TRH. J Clin Endocrinol Metab 39:191-194. Romijn A, Adrianse R, Brabant G, Prank K, Endert E, Wiersinga W (1990) Pulsatile secretion of thyrotropin during fasting: a decrease of thyrotropin pulse amplitude. J Clin Endocrinol Metab 70:1631-1636.
Thyroid Function in Bulirnia Nervosa
261
Romijn J, Wiersinga W (1990) Decreased nocturnal surge of thyrotropin in nonthyroidal illness. J Clin Endocrinol Metab 70:35--42. Rose S, Manasco P, Pearce S, Nisula B (1990) Hypothyroidism and deficiency of the nocturnal thyrotropin surge in children with hypothalamic-pituitary disorders. J Clin Endocrinol Metab 70:1750-1755. Schotte D, Stunkard A (1987) Bulima vs. bulimic behaviors on a college campus. J Am Med Assoc 258:1213-1215. Society of Actuaries and Association of Life Insurance Medical Directors of America (1979) Build Study. Society of Actuaries, Chicago, IL. Spaulding S, Chopra I, Sherwin R, Lyall S (1976) Effect of caloric restriction and dietary composition on serum T3 and reverse T3 in man. J Clin Endocrinol Metab 42:197-200. Stunkard AJ, Messick S (1985) The Three Factor Eating Questionnaire to measure dietary restraint, disinhibition and hunger. J Psychosomatic Res 29:71-83. Suda A, Pittman C, Shimizu T, Chambers J (1978) The production and metabolism of 3,5,3triiodothyronine and 3,3,5-triiodothyronine in normal and fasting subjects. J Clin Endocrinol Metab 47:1311-1319. Visser T, Lamberts S, Wilson J, Docter R, Hennemann G (1987) Serum thyroid hormone concentrations during prolonged reduction of dietary intake. Metabolism 27:405-409. Wartofsky L, Burman K (1982) Alterations in thyroid function in patients with systemic illnesses: the euthyroid sick syndrome. Endocrine Rev 3:164-217. Weltzin TE, Cameron J, Berga S, Kaye WH (1994) Prediction of reproductive status in women with bulimia nervosa by past high weight. Am J Psychiat 151:136-138. Wong TK, Pekary E, Hoo GS, Bradley ME, Hershman JM (1992) Comparison of methods for measuring free thyroxin in nonthyroidal illness. Clin Chem 38:720-724.
Pergamon
PII: S0306-4530(96)00002-9 THYROID FUNCTION
IN BULIMIA NERVOSA
Margaret Altemus 1, Marion Hetherington 2, Bingham Kennedy 3, Julio Licinio 3 and Philip W. Gold 3 1Laboratory of Clinical Science, DIRP, NIMH, Bldg 10, Rm 3D41 MSC 1264, 10 Center Drive, Bethesda, MD 20892-1264, USA; 2Department of Psychology, University of Dundee, Dundee DD1 4HN, UK; and 3Clinical Neuroendocrinology Branch, DIRP, NIMH, Bldg 10, Rm 3S231, 10 Center Drive Bethesda, MD 20892, USA
(Received 4 October 1995; in final form 8 December 1995)
SUMMARY Basal thyroid-stimulating hormone (TSH) and thyroid hormone levels were evaluated in 18 women with bulimia nervosa during a period of active bingeing and vomiting and again after 7 weeks of abstinence from these behaviors and compared to measures in 27 control women. In 10 of the patients and 11 of the controls, the TSH nocturnal surge was calculated from hourly TSH measurements obtained in the afternoon from 1500 to 1900h and in the night from 2300 to 0400h. During the bingeing phase of the illness patients had lower total triiodothyronine (T3) values than controls (p < .001). After 7 weeks without binge eating or purging, patients had lower T3, total thyroxine (T4), free triiodothyronine, free thyroxine (FT4), reverse triiodothyronine and thyroid-binding globulin (TBG) values compared to controls (p < .01) and significant reductions in T3, T4, FT4 and TBG compared to themselves in the active phase of the illness (p < .02). The reduction in thyroid hormone levels was not due to a reduction in the nocturnal thyrotropin surge, since surge values did not differ between normals and patients at either phase of the illness. Bulimics in the bingeing phase of the illness showed a positive correlation between caloric intake and TSH values (p < .01), suggesting that food bingeing may stimulate thyroid activity. In sum, these results show a substantial reduction in thyroid hormone levels after 7 weeks of abstinence from bingeing and vomiting behaviors. Published by Elsevier Science Ltd. Keywords--Bulimia; Binge eating; Thyroid hormone; TSH; TSH surge.
INTRODUCTION Bulimia nervosa is a psychiatric syndrome characterized by recurrent episodes of binge eating followed by purging of the food, usually by self-induced vomiting, to prevent weight gain. Patients with this disorder are distressed by their eating pattern, but have great difficulty resisting urges to binge. Onset of the syndrome is usually preceded by a period of dieting (Mitchell et al., 1986) and accordingly often occurs in women with anorexia nervosa (Herzog et al., 1988). Prevalence of the disorder in young women is estimated at 1-2% (Drewnowski et al., 1988; Schotte & Stunkard, 1987) and the relapse rate after treatment is Address correspondence and reprint requests to: Building 10, R o o m 3D41, 9000 Rockville Pike, Bethesda, MD 20892, U S A (Tel: 301 396 3421; Fax: 301 402 0188; e-mail: [email protected]
gov). 249
250
M. Altemus et al.
high (Herzog et al., 1991). A recent report suggests that the incidence of bulimia nervosa may be increasing (Hoek et al., 1995). A number of studies have reported a reduction in serum total triiodothyronine (T3) in normal-weight patients with bulimia, while reductions in serum free thyroxine (FT4) and thyroid-stimulating hormone (TSH) are seen less consistently (Kiyohara et al., 1988; Obarzanek et al., 1991; Pirke et al., 1985). Reductions in T3 may be due to caloric restriction, since many normal-weight bulimics have lost weight from a premorbidly heavier state (Fairburn & Cooper, 1984; Garfinkel et al., 1980), and even mild underfeeding is known to lower T3 (Garrel et al., 1984; Visser et al., 1987). Women with bulimia typically diet between binges and compared to non-dieting women matched for body size maintain their weight on fewer calories per day (Gwirtsman et al., 1989; Obarzanek et al., 1991) and have a lower resting metabolic rate (Devlin et al., 1990; Obarzanek et al., 1991). In addition, normal-weight bulimics have other metabolic signs of semi-starvation, including low fasting glucose levels, increased beta-hydroxybutyric acid and free fatty acid concentrations and reduced norepinephrine responses to orthostatic challenge (Devlin et al., 1990; Obarzanek et al., 1991; Pirke et al., 1985). Accordingly, changes in thyroid activity in bulimia nervosa seem to resemble changes seen in starvation and fasting, which usually fit a euthyroid sick pattern (Chopra et al., 1975; Chopra & Smith, 1975; Fichter et al., 1986; Portnay et al., 1974; Suda et al., 1978; Wartofsky & Burman, 1982). Prior studies of bulimic women are conflicting, but most have noted reductions in T3 while TSH levels are usually normal (Devlin et al., 1990; Kiyohara et al., 1988; Obarzanek et al., 1991; Pirke et al., 1985). Similarly, underweight subjects with anorexia nervosa have reductions in T3 and elevations in reverse triiodothyronine (rT3) which correct with refeeding, while TSH and FT4 typically do not differ from normals and do not change throughout the illness and recovery (Croxson & Ibbertson, 1975; DeRosa et al., 1983; Leslie et al., 1978; Miyai et al., 1975; Moore & Mills, 1979; Moshang et al., 1975). Alternatively, in the light of a number of studies suggesting alterations in hypothalamic stimulation of luteinizing hormone (Devlin et al., 1989; Pirke et al., 1987; Weltzin et al., 1994) and adrenocorticotropic hormone (Mortola et al., 1989) secretion in bulimia nervosa, decreased thyroid activity may be due to hypothalamic or central hypothyroidism, which is characterized by loss of the nocturnal TSH surge, low levels of FT4 and reduced or normal basal concentrations of TSH (Bartalena et al., 1987; Caron et al., 1986; Rose et al., 1990). The nocturnal TSH surge is thought to provide the major stimulatory signal for thyroid hormone production by the thyroid gland. Most often, central hypothyroidism results from hypothalamic and pituitary lesions (Caron et al., 1986; Fukuda & Greer, 1975; Rose et al., 1990), but can also occur in very severe non-thyroidal illness (Bartalena et al., 1990a,b; Romijn & Wiersinga, 1990), fasting (Romijn et al., 1990), depression (Bartalena et al., 1990c) and Cushing's disease (Bartalena et al., 1991). In order to further characterize the thyroid abnormalities in bulimia nervosa and the hormonal response to recovery from the illness, we measured basal thyroid hormones and the nocturnal thyrotropin surge in patients during a phase of bingeing and vomiting, again after 7 weeks of abstinence from the behaviors, and in controls. Because patients with bulimia nervosa often have derangements of hydration and fluid regulation (Demitrack et al., 1992), and volume shifts may influence levels of proteins and peptides in plasma (DeCostre et al., 1971), we also measured plasma renin activity and aldosterone in bulimic patients in both phases of the illness and in control subjects.
Thyroid Function in Bulimia Nervosa
251
MATERIALS AND M E T H O D S
Subjects Eighteen normal-weight bulimic women and 27 control women were included in this study. Bulimic women met DSM-IIIR (American Psychiatric Association, 1987) criteria for bulimia nervosa and had a mean percentage average body weight (ABW) of 95% (range 80117%). ABW is the percentage of the mean weight of individuals of the same sex, age and height (Society of Actuaries and Association of Life Insurance Medical Directors of America, 1979). Mean age was 25 years (range 20-34). Mean duration of illness was 7.4 years (range 2-13) and reported mean frequency of binge eating prior to admission was 27 episodes per week (range 10-75). Bulimic subjects abstained from taking any medications for at least 1 month before admission and had not had a weight change of more than 10% in the preceding 6 months. The highest lifetime ABW for this group of subjects was 104% (range 89-126%). One patient had a history of anorexia nervosa. No patient had received prior hormonal therapy. Patients continued their binge eating and vomiting during the first week of hospitalization. All subjects binged at least once per day during the first week of hospitalization and vomited after each binge. Patients were then closely supervised to enforce abstinence from bingeing and vomiting for the next 7 weeks of hospitalization. Patients selected their own food throughout the hospitalization, although they were required to keep their weights within 1 kg of their admission weights. They were treated with supportive psychotherapy, behavioral therapy and nutritional counseling. Controls were admitted to the hospital the day before the study in order to allow accommodation to the setting. Controls also selected their own food. Controls were in good health and taking no medications and had no history of eating disorders, dieting behavior or excessive exercise as assessed by clinical interview and responses to the Eating Inventory questionnaire (Stunkard & Messick, 1985). Controls were similar to patients in age (mean 23 years; range 18-35) and ABW (mean 98%; range 84-118%). All bulimic subjects and controls were caucasian. All subjects had normal screening blood tests, including electrolytes, renal and hepatic functions, and complete blood counts. The study was approved by the Institutional Review Board of the National Institute of Mental Health and all subjects gave informed consent before participation in the study.
Clinical data Caloric and macronutrient intake of the bulimic patients was calculated during the first and seventh weeks of hospitalization by subtraction of uncomsumed calories from calories presented, as described previously (Peterson et al., 1986). Twelve of the controls were also hospitalized for an additional 4 days to obtain caloric intake data. The Eating Inventory (Stunkard & Messick, 1985) was used to assess restrictive and uncontrolled eating behaviors in both bulimic women and control women. The Hamilton Depression Rating Scale was used to quantitate depressive symptoms (Hamilton, 1967).
Blood sampling Fasting blood samples drawn at 0900h after 1 h in the supine position were used to determine basal levels of hormones. Bulimic patients had these blood samples drawn after 1 week of bingeing and purging daily on the inpatient unit and again after 7 weeks of normal eating on the inpatient unit. Serum for T3 and total thyroxine (T4) was assayed within 24 h,
252
M. Altemuset al.
and blood for other thyroid hormone assays was collected in tubes with EDTA as an anticoagulant. Blood was centrifuged within 6 h and plasma stored at --70°C until assayed. For the subjects who had nocturnal TSH sampling, blood was drawn on the same day as the basal samples through an indwelling IV catheter with a heparin lock at hourly intervals between 1500 and 1900h and between 2300 and 0400h. This protocol of blood sampling provides an accurate assessment of the nocturnal serum TSH surge (Caron et al., 1986). Subjects were able to ambulate during the day and care was taken not to disturb sleep during the night-time blood draws. Calculation of the nocturnal TSH surge The nadir of TSH was designated as the lowest average of three consecutive values in the blood samples taken between 1500 and 1900h. The peak TSH was designated to be the highest average of three consecutive values in blood samples taken between 2300 and 0400h. The nocturnal TSH surge was calculated by subtracting the nadir TSH value from the peak value, dividing the difference by the nadir value, then multiplying the result by 100 in order to express the surge as a percentage rise over the nadir. Hormone assays T4, T3 and thyroid-binding globulin (TBG) were measured in fresh serum using an immunofluorescence assay for T4 (Abbott Laboratories, Irving, TX, USA), and radioimmunoassays for T3 (Kallestad Laboratories, Austin, TX, USA) and TBG (Ciba-Corning, Medfield, MA, USA). Intra-assay and inter-assay coefficents of variation for these three assays were less than 5%. Two bulimics and eight controls did not have TBG measurements. All other thyroid hormone assays were performed on frozen plasma. TSH was measured in duplicate using a sensitive immunoradiometric assay (CIBA-Serono, Walpole, MA, USA) having a limit of detection of 0.2 mU/L. Baseline TSH samples for half of the bulimics and half of the controls were assayed in each of two runs. To measure the TSH surge, active and abstinent samples from each patient were run together with samples from one or two control patients in each assay run. The intra-assay and inter-assay coefficients of variation were less than 5% and 6.2%, respectively. FT4 and free triiodothyronine (FT3) were measured by radioimmunoassay in single runs using commercial kits (Diagnostic Products Corporation, Los Angeles, CA, USA). Intra-assay coefficents of variation for both assays were less than 5%. rT3 was also measured by radioimmunoassay using a commercial kit (Serono Diagnostics, Norwell, MA, USA). Half of the bulimic samples and half of the control samples for rT3 were measured in each of two runs. Intra-assay and inter-assay coefficients of variation were less than 5% and 3%, respectively. Plasma renin and aldosterone were also measured in 15 patients and 10 controls by radioimmunoassay (Hazelton Laboratories, Fairfax, VA, USA). All samples were run in a single assay. Intra-assay coefficients of variation were less than 10% for both assays. Statistics Results were analyzed by two-tailed t-tests for paired and unpaired data. Pearson product moment correlations were calculated to determine relationships between variables. The slope of the regression line of two variables which were significantly related was determined by the least squares method and significant differences between slopes were determined using 95% confidence intervals. All values are given as mean _+SD. The level of significance was taken as p < .05.
Thyroid Function in Bulimia Nervosa
253
RESULTS
Clinical Variables Bulimic patients scored significantly higher than controls of all three subscales of the eating inventory: cognitive restraint: 12.7 _ 3.7 vs. 6.5 _+4.1 (t = 3.8, p < .001); disinhibition: 12.9 _+2.9 vs. 3.7 _ 2.9 (t = 9.7, p < .0001); and hunger: 11.1 _ 2.9 vs. 3.9 _+2.6 (t = 9.7, p < .0001). HamiltOn Depression Scale ratings of bulimic patients revealed a moderate degree of depression on admission ( 1 6 . 2 _ 7.5), which did not change after 6 weeks of hospitalization (15.0 _ 10.0, t = .05, p = .96). There was no significant difference between bulimics and controls in age (25 _+4 years vs. 23 _+4 years, t = - 1.1, p = NS) or %ABW (95.0 _+ 11% vs. 98 _ 10%, t = .95, p = NS). The weight of the bulimics did not change significantly during the 7 weeks of hospitalization. Mean daily caloric intake during the week of bingeing and vomiting was 12,216 _+6021 kcal/day. Frequency of purged meals (binges and smaller meals) during the first week of hospitalization did not differ from the reported frequency of binges prior to admission, when subjects were not asked to distinguish binges from smaller size purged meals (27 _ 20 vs. 21 _+8 binges/week, t = - 1.0, p = .34). As described in a separate report, the proportion of calories consumed as carbohydrate was similar in actively bulimic women and controls, but actively bulimic women consumed a significantly smaller proportion of calories as protein and a significantly greater proportion as fat (Hetherington et al., 1994). Caloric intake of the bulimics during the seventh week of hospitalization tended to be lower than mean caloric intake of the weight-matched controls, who were hospitalized for 4 days to obtain caloric intake data (1778 +_243 kcal/day vs. 1957 _ 296 kcal/day, t = 1.7, p = .09). The proportion of macronutrients consumed did not differ between bulimic women during the seventh week of hospitalization and controls.
Thyroid Hormone Values Thyroid hormone results are displayed in Table I. Compared to controls, patients with bulimia nervosa studied during the active phase of bingeing and vomiting showed evidence of the low T3 syndrome. Specifically, these patients had a significantly lower serum T3 (t = 3.8,p < .001) and no significant difference in FT3 (t = 1.5,p = .14), T4 (t = 1.5,p = .12) and FT4 (t = 1.4, p = .16). Their rT3 (t = .13, p = .90) and T B G (t = .7, p = .52) values also did not differ from controls. The basal 0800h TSH (t = .35, p = .72), the circadian TSH nadir (t = .48, p = .63) and peak (t = - . 2 8 , p = .78), and the calculated nocturnal TSH surge (t = - 1.3, p = .20) all did not differ significantly from those of controls. In contrast to the low T3 syndrome seen during active bingeing and vomiting, when the patients were re-studied after 6 weeks of supervised restraint from all pathological eating behaviors, they showed evidence of decreased thyroid function, including significant reductions compared to controls in T3 (t = 7.5, p < .001), FT3 (t = 3.0, p < .01), T4 (t = 5.4, p < .001), FT4 (t = 3.1, p < .01) and rT3 (t = 3.1, p < .01). In addition, serum TBG was significantly decreased (t = 2.2, p = .03). Despite these indices, suggestive of hypothyroidism during abstinence from bingeing, there was no change in the basal 0800h TSH (t = --.3, p = .76), the circadian TSH nadir (t = - . 6 , p = .52) or peak (t = - 1 . 0 , p = .26), or the nocturnal TSH surge (t = --1.1, p = .30). Direct longitudinal comparison in bulimic patients of thyroid function tests performed during bingeing and vomiting and again after supervised restraint from these behaviors showed significant reductions in T3 (t = 6.0, p < .001), FT4 (t = 2.6, p --- .02), T4 (t = 6.2, p < .001) and rT3 (t = 3.2, p < .01) and a t, end toward reduction in FT3 (t = 2.0, p = .06)
~N .~-
~~
cm ,,~
+I
+I
+I
+I
+I
+I
N
tN
¢II ,--~
[.,, I_. °~
oO +I
+I
-I-I
~0
~
oo
+I
+I
-I-I
+I
+I
÷I
+I
+I
+I
+I
+I
+I
+I
+I
"¢:I
++
. ,,,,,i ,-,,i
•
+I
+I
~
+I
~. +I
v
+
0
v
+I
+I
+I
+I
+I
+!
+I
+I
+I
÷I
4-I
+I
~.~.
.~"
¢,,)
2
..~:~
+,
+I
+I
+I
+I
++
::j
+I
÷I
+I
-I-I
++
-I-I
~
+I
+I
v
o
÷I
.~
v
Thyroid Function in Bulimia Nervosa
255
am 3 ¸
.J ::)
[] []
E
El []
-1¢/) I-
El []
[]
[]
[]
o
I
!
10000
20000
30000
CALORIES PER DAY Fig. 1. Relationship between 0800h plasma TSH and average daily caloric consumption in women in the active phase of bulimia nervosa.
and T B G (t = 1.88, p = .08) during the period of normal eating. H o w e v e r , there w a s no significant change in basal 0800h TSH, the circadian T S H nadir (t = -- 1.5, p = .16) or peak (t = - 1.3, p = .89), or the calculated nocturnal T S H surge (t = --.13, p = .89) values.
Renin and Aldosterone Values In the bingeing phase of the illness, c o m p a r e d to controls, w o m e n with bulimia nervosa had elevated levels of both plasma renin activity (0.75 _+ 0.08 ng/1 s vs. 0.23 _+ 0.06 ng/1 s, t = 2.1, p = .05) and aldosterone (510 _+ 410 pmol/l vs. 150 _+ 60 pmol/1, t = 2.6, p = .02). These elevated levels fell significantly after cessation of binge eating and purging (renin: 0.75 _+ 0.08 n g / l s vs. 0.24 _+ 0.19 ng/l s, t = 3.0, p < .01; aldosterone: 510 _+ 410 pmol/1 vs. 170 _+ 88 pmol/1, t = 3.5, p < .01), resulting in levels not significantly different from normals (renin: t = .2, p = .8; aldosterone: t = .6, p = .53).
Relationship Among Variables During the period of active bingeing and v o m i t i n g there w a s a positive correlation between the daily caloric intake and basal 0800h T S H (n = 18, r = .68, p < .01) (Fig. 1), T S H nadir (n = 10, r = .72, p < .02) and T S H peak (n = 10, r = .83, p < .01), but not the nocturnal T S H surge (n = 10, r = .19). There were no significant correlations between caloric intake and any measures of T S H secretion in non-bingeing bulimics or controls. There were no significant correlations between other thyroid indices and caloric intake in patients in
256
M. Altemus et al.
either phase of the illness or in controls. Binge frequency in the active phase of the illness was not significantly correlated with any thyroid indices. There was a positive correlation between FF4 and the TSH surge in controls (n = 11, r = .58, p < .05) and in abstinent bulimics (n = 10, r = .73, p < .02), but not in active bulimics. The slope of the regression line did not differ significantly between controls and abstinent bulimics. There were no significant correlations between depression ratings and any of the thyroid hormone measures or the nocturnal TSH surge.
DISCUSSION Patients with bulimia nervosa exhibited the low T3 syndrome while actively bingeing and vomiting. When re-studied after supervised restraint from bingeing and vomiting, there were more pronounced reductions in thyroid hormone levels; all peripheral indices of thyroid gland function were significantly reduced in this phase compared to controls. Moreover, patients studied during supervised restraint also showed a significant reduction in all peripheral indices when compared to themselves in the bingeing phase. Throughout, all indices of TSH secretion were normal. Moreover, even though peripheral indices of thyroid function shifted significantly with the phase of the illness, indices of TSH secretion did not change. The finding of reductions in T3, T4, F r 3 and FT4 during the non-binge eating phase agrees with previous studies of bulimics after 1-3 weeks of inpatient hospitalization (Kiyohara et al., 1988; Obarzanek et al., 1991; Pirke et al., 1985). There have been no comparable studies of thyroid function in bulimics during a documented active phase of the illness; however, Devlin et al. (1989) found normal levels of T3 and T4 in a mixed group of outpatient and newly hospitalized bulimic women. Because serum TBG levels also fell during abstinence from binge eating, a component of the reduction in T3 and T4 levels during abstinence is attributable to a reduction in the bound fraction of hormones. A reduction in TBG levels during the non-binge eating phase may also have contributed to the observed reductions in FT3 and FT4 values, since current immunoassay kits often underestimate concentrations of free thyroid hormone and results can vary with changes in bound thyroid hormone concentration (Nelson et al., 1994; Wong et al., 1992). The fall in TBG during abstinence from binge eating may be due to relative nutritional deprivation or to dilutional factors. As a group, the bulimic women appeared to have significant volume contraction during the phase of bingeing and vomiting, based on the renin and aldosterone levels. It is possible that some component of the change in thyroid hormone as well as TBG levels from the active phase to the abstinent phase could be dilutional. Although there are few data concerning the effect of dehydration on plasma hormone levels, diurnal changes in plasma proteins and peptides have been linked to volume shifts over the course of the day (DeCostre et al., 1971). On the other hand, one study has reported no change in plasma cortisol after 48 h of water deprivation in sheep (Matthews & Parrott, 1991), and in a 24 h serial blood sampling study in bulimic women there was no change in plasma cortisol between active and abstinent phases of the illness (Altemus et al., 1992). The thyroid hormone data in both phases of the illness appear to be explained at least in part by the euthyroid sick syndrome, in which reductions in T3 are not accompanied by elevations in TSH (Carter et al., 1974). On the other hand, the reduction in T3 in the euthyroid sick syndrome is thought to result from decreased metabolism of T4 to T3 (Chopra
Thyroid Function in Bulimia Nervosa
257
et al., 1975), with increases in rT3, while TSH and FT4 remain at normal levels (Bermudez et al., 1975; Chopra et al., 1974). Bulimic women in this study did not fit this pattern since they did not have increases in rT3 at any point. The bulimic women in this study also did not exhibit the cardinal feature of the syndrome of central hypothyroidism, since the TSH surge was preserved in both phases of the illness. On the other hand, bulimic women show delayed plasma TSH responses to synthetic thyrotropin-releasing hormone (TRH) (Coiro et al., 1990; Gwirtsman et al., 1983; Kiriike et al., 1987; Kiyohara et al., 1988; Levy et al., 1988), which are qualitatively similar to responses seen in patients with central hypothyroidism (Beck-Peccoz et al., 1985). The lack of change in indices of TSH secretion in the abstinent bulimic patients despite substantial changes in thyroid hormone values over the course of hospitalization is also consistent with a primary impairment of thyroid gland function or a reduction in the bioactivity of endogenous TSH. One study which reported the Tr3 response to TRH infusion in bulimics (Beck-Peccoz et al., 1985) found an impaired Tr3 response to the TSH elevation following TRH infusion, again consistent with reduced biological activity of TSH or impaired thyroid gland responses to endogenous TSH. The reduction in thyroid activity in bulimia nervosa, particularly during the non-binge eating phase, may be appropriately adaptive since this study and others (Garfinkel et al., 1980; Obarzanek et al., 1991) have noted a reduced caloric intake in bulimics during a period of restraint from binge eating when compared to normals matched for body size. In the face of this decreased energy intake, the set point for feedback regulation of TSH may be lowered to reduce thermogenesis as occurs in starvation and anorexia nervosa (Croxson & Ibbertson, 1975). In addition to presenting with weights which were on average 5.2 +_5.0 kg below their lifetime highest body weight, bulimic women may have lost a small amount of lean body mass over the course of this study, since their weight did not change while they substantially increased their hydration. The strong correlations between indices of TSH secretion and caloric intake during the bingeing phase suggests that food ingestion plays a role in creating the relative increase in thyroid hormone levels during the active phase of the illness. Although much of the binge food is vomited and never actually absorbed, we speculate that preabsorbtive effects of food ingestion such as insulin secretion and the associated activation of the sympathetic nervous system (Leblanc, 1992), both of which have been shown to be greatly exaggerated during binge meals (Kaye et al., 1989), could stimulate thyroid hormone activity during the phase of bingeing and vomiting. Data that norepinephrine is a potent stimulus to TRH release (Engler et al., 1982; Palkovits et al., 1980) suggest that preabsorptive sympathetic activation could act to mitigate a central hypothyroidism induced by food restriction. In support of this model, bulimic patients have low cerebrospinal fluid and plasma norepinephrine levels during restraint from bingeing and normal levels during periods of bingeing and vomiting (Kaye et al., 1990). Further indirect evidence that preabsorbtive insulin responses can increase thyroid hormone activity comes from data that, compared to other macronutrients, carbohydrates preferentially induce insulin release and that only reduction of the proportion of carbohydrate in the diet can lower thyroid hormone activity (Davidson & Chopra, 1979; Spaulding et al., 1976). As a corollary, we have shown previously that the phase of active bingeing and vomiting is also associated with a significantly higher resting metabolic rate, even if weight is held stable (Altemus et al., 1991). In this regard, one could construe the bingeing and vomiting of bulimia nervosa to, in part, serve the function of increasing thyroid function and metabolic
258
M. Altemus et al.
rate without weight gain, and avoiding the reduction in thyroid hormone activity known to be associated with dieting. Additional longitudinal studies are needed to determine whether this apparent increase in thyroid hormone activity during the active phase of the illness is also associated with relative enhancement of mood or cognitive function. Our demonstration of state-related changes in thyroid hormone levels may help to explain the discrepant findings in the literature where frequency of bingeing and vomiting and duration of hospitalization before testing were not well controlled. Detection of abnormalities in the present study was also probably enhanced by selecting patients with at least a moderate severity of illness and careful screening of normals to exclude any with dieting behaviors. Although 7 weeks should be adequate time to come to a steady state as far as half-life of the thyroid hormones is concerned, it is possible that the adjustment to cessation of bingeing and vomiting occurs over a longer time period and the altered state observed here is temporary and resolves at a later point. In summary, these data indicate that normal-weight bulimic women have reduced thyroid hormone levels during abstinence from bulimic behaviors and that partial correction of thyroid hormone abnormalities occurs during periods of active bingeing and vomiting. These findings are consistent with reports of decreases in resting metabolic rate and reduced caloric requirements during abstinence from bulimic behaviors. Low thyroid hormone activity and a low metabolic rate could predispose these patients to weight gain and thereby exacerbate their preoccupation with weight control. Restrained caloric intake may also stimulate hunger, which is associated with onset of binges (Hetherington et al., 1994). Moreover, partial correction of hypothyroidism by bingeing and vomiting may be physiologically reinforcing and contribute to the development of this condition during periods of dieting as well as the propensity for relapse after treatment.
Acknowledgements:We wish to acknowlege the helpful review of this manuscript by Bruce Nisula, M.D. and the assistance of Yung-Mei Leong in manuscript preparation.
REFERENCES Altemus M, Hetherington M, Flood M, Licinio J, Nelson M, Bernat A, Gold P (1991) Decrease in resting metabolic rate during abstinence from bulimic behavior. Am J Psychiat 148:1071-1072. Altemus M, Licinio J, Jolkowsky L, Gold P (1992) Stimulation of the hypothalamic-pituitary-adrenal axis by bulimic behaviors. Proceedings of the 145th Annual Meeting American Psychiatric Association, Washington, DC. American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. Revised. American Psychiatric Association, Washington, DC. Bartalena L, Martino E, Brandi L, Falcone M, Pacchiarotti A, Ricci C, Bogazzi F, Grasso L, Mammoli C, Pinchera A (1990) Lack of nocturnal serum thyroptropin surge after surgery. J Clin Endocrinol Metab 70:293-296. Bartalena L, Martino E, Falcone M, Buratti L (1987) Evaluation of the nocturnal serum thyroptropin (TSH) surge, as assessed by TSH ultrasensitive assay, in patients receiving long term L-thyroxine suppression therapy and in patients with various thyroid disorders. J Clin Endocrinol Metab 65:1265-1271. Bartalena L, Martino E, Petrini L, Velluzzi F, Loviselli A, Grasso L, Mammoli C, Pinchera A (1991) The nocturnal serum thyrotropin surge is abolished in patients with adrenocorticotropin (ACTH)dependent or ACTH-independent Cushing's syndrome. J Clin Endocrinol Metab 72:1195-1199. Bartalena L, Pacchiarotti A, Palla R, Antonagelli L, Mammoli C, Monzani F, DeNegri F, Panichi V, Martino E, Baschieri L (1990) Lack of nocturnal serum thyrotropin (TSH) surge in patients with chronic renal failure undergoing regular maintenance hemofiltration: A case of central hypothyroidism. Clin Nephrol 34:30-34.
Thyroid Function in Bulimia Nervosa
259
Bartalena L, Placidi G, Martin E, Falcone M, Pellegrini L, Dell'Osso L, Pacchiarotti A, Pinchera A (1990) Nocturnal serum thyrotropin (TSH) surge and the TSH response to TSH-releasing hormone: dissociated behavior in untreated depressives. J Clin Endocrinol Metab 70:650-655. Beck-Peccoz P, Amir S, Menezes-Ferreira M, Faglia G, Weintraub B (1985) Decreased receptor binding of biologically inactive thyrotropin in central hypothyroidism: effect of treatment with thyrotropin releasing hormone. New Engl J Med 312:1085-1090. Bermudez F, Surks M, Oppenheimer J (1975) High incidence of decreased serum triiodothyronine concentration in patients with n-on-thyroidal disease. J Clin Endocrinol Metab 41:27-40. Caron P, Nieman L, Rose S, Nisula B (1986) Deficient nocturnal surge of thyrotropin in central hypothyroidism. J Clin Endocrinol Metab 62:960-964. Carter J, Eastman C, Corcoran J, Lasarus L (1974) Effect of severe chronic illness on thyroid function. Lancet 2:971-974. Chopra I, Chopra U, Smith S, Reza M, Solomon D (1975) Reciprocal changes in serum concentrations of 3,3',5'-triiodothyronine (reverse T3) and 3,3',5'-triiodothyronine (T3) in systemic illnesses. J Clin Endocrinol Metab 41:1043-1049. Chopra I, Smith S (1975) Circulating thyroid hormones and thyrotropin in adult patients with proteincalorie malnutrition. J Clin Endocrinol Metab 40:221-227. Chopra I, Solomon D, Chopra U~ Young R, Teco GC (1974) Alterations in circulating thyroid hormones and thyrotropin in hepatic cirrhosis: evidence for euthyroidism despite subnormal serum triidothyronine. J Clin Endocrinol Metab 39:501-511. Coiro V, Capretti L, Volpi R, d'Amato L, Marchesi C, DeFerri A, Rossi G, Bianconi L, Marcato A, Chiodera P (1990) Growth hormone responses to growth hormone-releasing hormone, conidine, and insuline-induced hypoglycemia in normal weight bulimic women. Neuropsychobiology 23:8-14. Croxson M, Ibbertson H (1975) l o w serum triiodothyronine (t3) and hypothyroidism in anorexia nervosa. J Clin Endocrinol Metab 44:167-174. Davidson M, Chopra I (1979) Effect of carbohydrate and noncarbohydrate sources of calories on plasma 3,5,5'-triiodothyronine concentrations in man. J Clin Endocrinol Metab 48:577-581. DeCostre P, Buhler U, DeGroot LJ, Refetoff S (1971) Diurnal rhythm in total serum thyroxine levels. Metabolism 20:782-791. Demitrack M, Kalogeras K, Altemus M, Pigott T, Listwak S, Gold P (1992) Plasma and cerebrospinal fluid measures of arginine vasopressin secretion in patients with bulimia nervosa and in healthy subjects. J Clin Endocrinol Metab 74:1277-1283. DeRosa G, Corsello S, DeRosa E, Casa SD, Ruffilli M, Grasso P, Pasargiklian E (1983) Endocrine study of anorexia nervosa. Exp Clin Endocrinol 82:160-172. Devlin M, Walsh B, Kral J, Heymsfield S, Pi-Sunyer F, Dantzic S (1990) Metabolic abnormalities in bulimia nervosa. Arch Gen Psychiat 47:144-148. Devlin M, Walsh B, Roose S, Katz J, Linkie D (1989) Gonadotropin patterns in eating disorders. Psychiat Res 28:11-24. Drewnowski A, Hopkins S, Kessler R (1988) The prevalence of bulimia nervosa in the US college student population. Am J Public Health 78:1322-1325. Engler D, Chad D, Jackson I (1982) Thyrotropin-releasing hormone in the pancreas and brain of the rat is regulated by central noradrenergic and dopaminergic pathways. J Clin Invest 69:1310-1320. Fairburn C, Cooper P (1984) The clinical features of bulimia nervosa. Br J Psychiat 144:238-246. Fichter M, Pirke K, Holsboer F (1986) Weight loss causes neuroendocrine disturbances: experimental study in healthy starving subjects. Psychiat Res 17:61-72. Fukuda J, Greer M (197) The effect of basal hypothalamic deafferentation on nycthemeral rhythm of plasma TSH. Endocrinology 975749-752. Garfinkel P, Moldofsky H, Garner D (1980) The heterogeneity of anorexia nervosa: bulimia as a distinct subgroup. Arch Gen Psychiat 37:1036-1040. Garrel D, Todd L, Pugeat M, Calloway D (1984) Hormonal changes in normal men under marginally negative energy balance. Am J Clin Nutr 39:930-936. Gwirtsman H, Kaye W, Obarzanek E, George D, Jimerson D, Ebert M (1989) Decreased caloric intake in normal-weight patients with bulimia: comparison with female volunteers. Am J Clin Nutr 49: 86-92. Gwirtsman H, Roy-Byrne P, Yager J, Gerner R (1983) Neuroendocrine abnormalities in bulimia. Am J Psychiat 140:559-563.
260
M. Altemus et al.
Hamilton M (1967) Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 6:278-296. Herzog D, Keller M, Lavori P (1988) Outcome in anorexia nervosa and bulimia. J Nerv Ment Dis 176:131-143. Herzog D, Keller M, Lavori P, Sacks N (1991) The course and outcome of bulimia nervosa. J Clin Psychiat 52suppl:4-8. Hetherington MM, Altemus M, Nelson ML, Bernat AS, Gold PW (19~4) Eating behavior in bulimia nervosa: multiple meal analysis. Am J Clin Nutr 60:864-873. Hoek HW, Barelds AIM, Bosveld JJF, van der Graaf Y, Limpens VEL, Maiwald M, Spaaij CJK (1995) Impact of urbanization on detection rates of eating disorders. Am J Psychiat 152:1272-1278. Kaye W, Gwirtsman H, George D (1989) The effect of bingeing and vomiting on hormonal secretion. Biol Psychiat 25:768-780. Kaye W, Gwirtsman H, Goerge D, Jimerson D (1990) Disturbances of noradrenergic systems in normal weight bulimia: relationship to diet and menses. Biol Psychiat 27:4-21. Kiriike N, Nishiwaki S, Izumiya Y, Maeda Y, Kawakita Y (1987) Thyrotropin, prolactin, and growth hormone responses to thyrotropin-releasing hormone in anorexia nervosa and bulimia. Biol Psychiat 22:167-176. Kiyohara K, Tamai J, Kobayashi N, Nakagawa T (1988) Hypothalamic-pituitary-thyroidal alterations in bulimic patients. Am J Clin Nutr 47:805-809. Leblanc J (1992) Cephalic phase response to food and its effect on feeding behavior. In: Anderson G, Kennedy S (Eds) The Biology of Feast and Famine. Academic Press, San Diego, pp 47-60. Leslie R, Isaacs A, Gomez J, Raggatt P, Bayliss R (1978) Hypothalamic-pituitary-thyroid function in anorexia nervosa: influence of weight gain. Br Med J 2:526-528. Levy A, Dixon K, Malarkey W (1988) Pituitary response to TRH in bulimia. Biol Psychiat 23:476484. Matthews S, Parrott R (1991) Dehydration, but not vasopressin infusion, enhances the adrenocortical response of sheep to corticotropin-releasing hormone or restraint. Acta Endocrinol 125:556-562. Mitchell J, Hatsukami D, Pyle R, Eckert E (1986) The bulimia syndrome: course of the illness and associated problems. Compr Psychiat 27:165-170. Miyai K, Yamamoto T, Azukizawa M, Ishibashi K, Kumahara Y (1975) Serum thyroid hormones and thyrotropin in anorexia nervosa. J Clin Endocrinol Metab 40:334-338. Moore R, Mills I (1979) Serum T3 and T4 levels in patients with anorexia nervosa showing transient hyperthyroidism during weight gain. Clin Endocrinol 10:443-449. Mortola J, Rasmussen D, Yen S (1989) Alterations of the adrenocorticotropin-cortisol axis in normal weight bulimic women: evidence for a central mechanism. J Clin Endocrinol Metab 68:517-522. Moshang T, Parks J, Baker L, Vaidya V, Utiger R, Bongiovanni A, Syder P (1975) Low serum triiodothyronine in patients with anorexia nervosa. J Clin Endocrinol Metab 40:470--473. Nelson JC, Weiss RM, Wilcox RB (1994) Underestimates of serum free thyroxine (T4) concentrations by free T4 immunoassays. J Clin Endocrinol Metab 79:76-79. Obarzanek E, Lesem M, Goldstein D, Jimerson D (1991) Reduced resting metabolic rate in patients with bulimia nervosa. Arch Gen Psychiat 48:456-462. Palkovits M, Zaborszky L, Feminger A, Mezey E, Fekete M, Herman J, Kanyicska B, Szabo D (1980) Noradrenergic innervation of the rat hypothalamus: experimental biochemical and electron microscope studies. Brain Res 191:161-171. Peterson R, Kaye W, Gwirtsman H (1986) Comparison of calculated estimates and laboratory analysis of food offered to hospitalized eating disorder patients. J Am Diet Assoc 86:490--492. Pirke K, Fichter M, Schweiger U, Fruth C, Streitmatter A, Wolfram G (1987) Gonadotropin secretion pattern in bulimia nervosa. Int J Eating Disord 6:655-661. Pirke K, Pahl J, Schweiger U, Warnhoff M (1985) Metabolic and endocrine indices of starvation in bulimia: a comparison with anorexia nervosa. Psychiat Res 15:33-39. Portnay G, O'Brian J, Bush J, Vagenakis A, Azizi F, Arky R, Ingbar S, Braverman L (1974) The effect of starvation on the concentration and binding of thyroxine and triiodothyronine in serum and on the response to TRH. J Clin Endocrinol Metab 39:191-194. Romijn A, Adrianse R, Brabant G, Prank K, Endert E, Wiersinga W (1990) Pulsatile secretion of thyrotropin during fasting: a decrease of thyrotropin pulse amplitude. J Clin Endocrinol Metab 70:1631-1636.
Thyroid Function in Bulirnia Nervosa
261
Romijn J, Wiersinga W (1990) Decreased nocturnal surge of thyrotropin in nonthyroidal illness. J Clin Endocrinol Metab 70:35--42. Rose S, Manasco P, Pearce S, Nisula B (1990) Hypothyroidism and deficiency of the nocturnal thyrotropin surge in children with hypothalamic-pituitary disorders. J Clin Endocrinol Metab 70:1750-1755. Schotte D, Stunkard A (1987) Bulima vs. bulimic behaviors on a college campus. J Am Med Assoc 258:1213-1215. Society of Actuaries and Association of Life Insurance Medical Directors of America (1979) Build Study. Society of Actuaries, Chicago, IL. Spaulding S, Chopra I, Sherwin R, Lyall S (1976) Effect of caloric restriction and dietary composition on serum T3 and reverse T3 in man. J Clin Endocrinol Metab 42:197-200. Stunkard AJ, Messick S (1985) The Three Factor Eating Questionnaire to measure dietary restraint, disinhibition and hunger. J Psychosomatic Res 29:71-83. Suda A, Pittman C, Shimizu T, Chambers J (1978) The production and metabolism of 3,5,3triiodothyronine and 3,3,5-triiodothyronine in normal and fasting subjects. J Clin Endocrinol Metab 47:1311-1319. Visser T, Lamberts S, Wilson J, Docter R, Hennemann G (1987) Serum thyroid hormone concentrations during prolonged reduction of dietary intake. Metabolism 27:405-409. Wartofsky L, Burman K (1982) Alterations in thyroid function in patients with systemic illnesses: the euthyroid sick syndrome. Endocrine Rev 3:164-217. Weltzin TE, Cameron J, Berga S, Kaye WH (1994) Prediction of reproductive status in women with bulimia nervosa by past high weight. Am J Psychiat 151:136-138. Wong TK, Pekary E, Hoo GS, Bradley ME, Hershman JM (1992) Comparison of methods for measuring free thyroxin in nonthyroidal illness. Clin Chem 38:720-724.