- Email: [email protected]
Decreased plasma tryptophan concentration in major depression: relationship to melancholia and weight loss
Journal
of Affectice
Disorders,
20 (1990)
185
185-191
Elsevier
JAD 00756
Decreased plasma tryptophan concentration in major depression: relationship to melancholia and weight loss 1-M. Anderson, MRC
M. Parry-Billings
Unit of Clinical Pharmacology and
I, E.A. Newsholme
and Uniuersity Department
t Cellular Nutrition Research
Group, Department
of Psychiatry
I, J.R. Poortmans Littkmore
of Biochemistry,
Hospital,
Littlemore,
Unrversity of Oxford,
1
and P.J. Cowen Oxford OX4 4XN,
Oxford OXI 3QlJ.
U.K.
U.K.
(Received 25 April 1990) (Revision received 12 July 1990) (Accepted 17 July 1990)
Summary Plasma total tryptophan (TRP) concentration was significantly lower in 31 patients with major depression compared to a healthy control group. The ratio of plasma TRP concentration to that of other branch chain amino acids (the TRP: BCAA ratio) was also decreased. Further analysis revealed that the decrease in plasma TRP and TRP : BCAA ratio was most apparent in patients with major depression and melancholia. Overall, women but not men had significantly decreased plasma tryptophan concentrations, perhaps because of a contributory effect of weight loss; this latter effect, however, could not be distinguished clearly from a diagnosis of melancholia. Our data suggest that in some depressed patients, reductions in plasma tryptophan availability may contribute to abnormalities in brain 5hydroxytryptamine function.
Key words: Tryptophan;
Depression;
Melancholia;
Introduction There is continued interest in the role of brain serotonin (5-hydroxytryptamine, 5-HT) pathways in the aetiology of depression (Coppen, 1967; Cowen, 1988) and the mediation of antidepressant effects of pharmacological treatments (deMon-
Address for correspondence: Dr. I.M. Anderson, MRC Unit of Clinical Pharmacology and University Department of Psychiatry, Littlemore Hospital, Littlemore, Oxford OX4 4XN, U.K. 0165-0327/90/$03.50
0 1990 Elsevier Science Publishers
Branch
chain amino
acids;
5-Hydroxytryptamine
tigny and Blier, 1984; Delgado et al., 1990). The synthesis of brain 5-HT is dependent on the availability of the 5-HT precursor, tryptophan (TRP) (Fernstrom, 1983), and several studies have measured the circulating concentration of TRP in depressed patients. The results of these studies have been conflicting with reports of both normal (Riley and Shaw, 1976; Msller et al., 1979) and decreased TRP concentrations (DeMyer et al., 1981; Joseph et al., 1984; Maes et al.,-1987). An important consideration is that the availability of TRP for brain 5-HT synthesis is determined by two factors: first, the amount of TRP in the
B.V. (Biomedical
Division)
186
plasma unbound to albumin (plasma free TRP) (Curzon, 1979,1988); and second, the total amount of plasma TRP in relation to branch chain amino acids (BCAA) that compete with TRP for transport across the blood-brain barrier (the TRP: BCAA ratio) (Fernstrom, 1973, 1979). One reason for the conflicting results of TRP studies in depressed patients may be that low plasma TRP concentrations are found only in particular clinical subgroups. In a recent study, for example, we found that both plasma total TRP and the TRP : BCAA ratio were significantly decreased in a small group of patients with DSM-III major depression and melancholia (Cowen et al., 1989). This reduction was most clearly apparent in female patients who reported weight loss during the affective episode and is consistent with our studies in normal volunteers that have shown decreased plasma TRP concentrations following weight loss through dieting (Anderson et al., 1989, 1990; Goodwin et al., 1990). Interestingly, Maes et al. (1987) found that plasma total TRP and TRP : BCAA ratio were significantly decreased in patients meeting diagnostic criteria for DSM-III major depression with melancholia or psychosis but not in non-melancholic depressives. Since many of the melancholic and psychotic patients would have been expected to have experienced significant weight loss the relative importance of depressive subgroup and weight loss as determinants of reduced plasma TRP availability is unclear. The purpose of the present study was to confirm and extend our original observations on plasma TRP and TRP : BCAA ratio in major depression and to investigate further the role of weight loss, sex and melancholic subtype as determinants of reduced plasma TRP availability in depressed patients. Methods
Subjects We studied 31 depressed patients (16 male and 15 female) who on the basis of a semi-structured interview met DSM-III criteria for major depression, Sixteen subjects met DSM-III criteria for major depression with melancholia. The mean score for the group on the 17-item Hamilton De-
pression Rating Scale (HDRS) (Hamilton, 1967) was 22.8 (range 15-34). All subjects had been free of antidepressant medication for at least 4 weeks and most for considerably longer. Five subjects were taking benzodiazepines for night sedation but no other medication was allowed. The subjects studied included 12 patients whose data have been previously reported (Cowen et al., 1989). Weight loss was assessed using the weight loss rating item on the Beck Depression Inventory (BDI) (Beck et al., 1961) which asks subjects how much weight they have lost ‘recently’ (explained to subjects as weight loss in the last month). A specially designed observer-rated questionnaire (available from the authors) was also administered, again with the aim of detecting weight loss over the previous month. Before completing these assessments subjects were weighed and informed of the result. Both assessments were used to decide likely weight loss. Five patients had weight loss of 5-10 lbs, eight patients had weight loss of 11-15 lbs and two patients had weight loss of more than 15 lbs. These groups did not differ significantly in plasma TRP measures and so were combined to make a single ‘weight loss’ group (i.e., weight loss of > 5 lbs in the last month; n = 15). The results of this group were compared with a ‘no weight loss group’ (i.e., weight loss of less than 5 lbs in the last month; n = 16). An absolute estimate of weight loss for each subject was also made. Overall male patients reported weight loss during the last month of 5.3 + 1.8 lbs and females, 5.6 + 1.4 lbs (mean + SEM). For comparison with the patients we selected 31 control subjects (16 male, 15 female) with no history of psychiatric disorder and no change in weight in the last 2 months. As far as possible they were age- and weight-matched with the patient group (mean age of patients 33.5 years, range 20-55 years; mean age of controls 33.0 years, range 21-53 years; mean weight of patients 149 lbs, range 89-230 lbs; mean weight of controls 152 lbs, range 99-226 lbs). Sampling procedure and biochemical analysis Both patients and controls were sampled after an overnight fast. Plasma was separated from venous blood by centrifugation and stored at -20°C until assay. The concentrations of BCAA
187
(valine, leucine and isoleucine) were determined by an enzymatic method (Livesey and Lund, 1980) on neutralised perchloric acid extracts of plasma. The TRP concentration was determined on trichloroacetic acid extracts using the fluorometric method of Denkla and Dewey (1967) with the modification described by Bloxam and Warren (1974). Assays were carried out blind to diagnosis.
50
I
T
Statistics
Comparisons between patient and control groups were assessed by Student’s unpaired t-test (two-tailed). The weight of the different variables in predicting plasma TRP concentration was determined by stepwise multiple linear regression. Correlations were carried out using Pearson’s product moment. Results The patients had a significantly lower mean plasma total TRP and also a significantly decreased TRP : BCAA ratio compared to controls. The BCAA concentration did not differ between patients and controls, so that the reduction in TRP : BCAA ratio was attributable to the decrease in plasma TRP concentration (Table 1). Subdivision of the patients by sex showed that plasma TRP was significantly reduced in women but not men, and the same was true of the TRP : BCAA ratio (Table 1). In patients with weight loss, plasma TRP concentration was significantly reduced, an effect not observed in patients without weight loss. In both groups the TRP : BCAA ratio showed similar reductions of borderline statistical significance (P <
TABLE
CON
DEP
(31)
(31)
M-W-
M-W+
M+W-
M+W+
(9)
(6)
(7)
(9)
Fig. 1. Plasma TRP concentrations in depressed patients (DEP) and controls (CON). The patients were subdivided according to the presence ( + ) or absence (- ) of melancholia (M) and weight loss (W). The number in parentheses is the number of patients in each category. * * P c 0.005; * * * P i 0.001.
0.09) (Table 2). Further subdivision by sex revealed that in women but not men, weight loss was associated with significantly reduced plasma TRP and TRP : BCAA ratio (Tables 3 and 4). Patients with melancholia had significantly reduced plasma TRP concentration and TRP: BCAA ratio compared to controls (Table 2) and the same pattern of results was seen when male and female melancholic depressives were consid-
1
PLASMA CONCENTRATIONS OF TRYPTOPHAN (TRP) AND BRANCH LEUCINE, ISOLEUCINE) IN CONTROLS AND DEPRESSED PATIENTS n
All subjects Females Males
TRP (PM)
CHAIN
AMINO
ACIDS
Ratio (TRP
BCAA (PM)
(BCAA)
(VALINE,
: BCAA)
Controls
Patients
Controls
Patients
Controls
Patients
Controls
Patients
31 15 16
31 15 16
45.4+ 2.0 46.9 f 3.5 44.0 f 2.3
38.2* 1.6 * * 36.9* 2.1 * 39.7 f 2.4
376.1 f 16.3 344.1 k 23.8 406.1 * 20.2
363.5 f 12.9 357.6 f 16.0 369.1 k 20.4
0.128 +0.008 0.145_+0.013 0.112 * 0.007
0.109 + 0.005 * 0.105 f 0.006 * * 0.112~0.009
Values are mean + SEM. * P < 0.05, * * P < 0.01, less than controls.
188 TABLE
2
PLASMA CONCENTRATIONS OF TRYPTOPHAN (TRP) AND BRANCH CHAIN LEUCINE, ISOLEUCINE) IN CONTROLS AND DEPRESSED PATIENTS ACCORDING
AMINO ACIDS (BCAA) TO SUBGROUP
: BCAA)
Subjects
n
TRP (PM)
BCAA (PM)
Ratio (TRP
Controls
31
45.4k2.0
376.1 + 16.3
0.128+0.008
Patients No weight loss Weight loss No melancholia Melancholia
16 15 15 16
40.9 35.5 42.2 34.7
382.3 343.5 349.9 376.3
0.109+0.007 0.108 + 0.008 0.125kO.008 0.094 + 0.005 * * *
* 2.2 +2.1 * * + 2.5 * 1.5 * * *
+ k + +
14.9 20.7 22.3 13.6
(VALINE.
Values are mean + SEM. * * P i 0.002, ***P
ered separately (Tables 3 and 4). Subdivision of subjects into those with melancholic and nonmelancholic depression with and without weight loss yielded small cell sizes, but plasma TRP was significantly reduced in patients with melancholic depression without concomitant weight loss and
TABLE
further reduced in patients with both weight loss and a diagnosis of melancholia. Subjects with weight loss who did not meet criteria for melancholia did not have significantly decreased plasma TRP concentrations compared to controls (P > 0.1) (Fig. 1).
3
PLASMA CONCENTRATIONS OF TRYF’TOPHAN LEUCINE, ISOLEUCINE) IN FEMALE CONTROLS
(TRP) AND BRANCH CHAIN AMINO ACIDS (BCAA) (VALINE, AND DEPRESSED PATIENTS ACCORDING TO SUBGROUP
: BCAA)
Female subjects
n
TRP (PM)
BCAA (PM)
Ratio (TRP
Controls Patients No weight loss Weight loss No melancholia Melancholia
15
46.9+ 3.5
344.1 + 23.8
0.145+0.013
44.0 32.2 41.2 33.1
382.0,25.8 341.3 f 19.7 357.9 + 30.7 357.4+ 16.0
0.117+0.007 0.097 * 0.009 * * 0.118f0.008 0.094 + 0.008 * *
6 9 7 8
+ * + f
1.9 2.7 * * 3.0 2.3 * *
Values are mean k SEM. * * P < 0.005, less than controls.
TABLE
4
PLASMA CONCENTRATIONS OF TRYI’TOPHAN (TRP) AND BRANCH CHAIN AMINO LEUCINE, ISOLEUCINE) IN MALE CONTROLS AND DEPRESSED PATIENTS ACCORDING
ACIDS (BCAA) TO SUBGROUP
: BCAA)
Male subjects
n
TRP (PM)
BCAA (PM)
Ratio (TRP
Controls Patients No weight loss Weight loss No melancholia Melancholia
16
44.Ok2.3
406.1 k 20.2
0.118+0.007
10 6 8 8
39.1 40.6 43.1 36.3
382.5 + 346.7 + 343.0 f 395.1+
0.105+0.011 0.125 f 0.014 0.131 kO.013 0.093 + 0.006 *
Values are mean + SEM. * P i 0.05, less than controls.
53.2 + 3.6 i4.1 + 1.8 *
19.2 45.4 33.9 20.9
(VALINE,
189
The above findings were supported by the results of a stepwise multiple linear regression analysis with TRP as the dependent variable and melancholia, weight loss and sex as the independent variables. The only significant prediction of plasma TRP concentration was a diagnosis of melancholia (R2 = 0.19, P = 0.014). A similar finding was obtained where the TRP : BCAA ratio was used as the dependent variable ( R2 = 0.29, P = 0.002). Correlation analyses revealed that plasma TRP did not correlate significantly with absolute weight loss in the patients as a whole (r = -0.15, P = 0.41) or in male patients (1. = 0.14, P = 0.62) but there was a significant negative correlation with weight loss in female patients (I = -0.64, P = 0.01); however, there was no relationship between plasma TRP and severity of depression as measured by HDRS score (r = 0.11, P = 0.55) or between HDRS score and TRP: BCAA ratio (r = -0.11, P = 0.55). Discussion
The present data confirm our previous findings that plasma total TRP concentration is decreased in major depression and that, as a consequence of this, the TRP: BCAA ratio is also lowered. This suggests that during some depressive disorders, the availability of plasma TRP for brain 5-HT synthesis is reduced. While depressed women who reported weight loss had the lowest plasma tryptophan concentrations, our data suggest that a diagnosis of melancholic depression, rather than sex or presence of weight loss, is the major determinant of low plasma TRP in depressed patients. The findings of a reduced plasma TRP concentration and TRP : BCAA ratio in melancholic patients are in agreement with the data of Maes et al. (1987). However, weight loss is one of the diagnostic criteria for melancholia and it is important, if possible, to distinguish between the effects of weight loss and melancholia on plasma TRP. This is especially pertinent given our previous findings that in both normal men and women who lose weight by dieting, plasma TRP concentration and TRP : BCAA ratio are significantly reduced (Anderson et al., 1989, 1990; Goodwin et al., 1990). In the present study weight loss, as
judged by our ratings, was also associated with decreased plasma TRP in depressed patients. However, in the group of patients with weight loss but no melancholia, plasma TRP was not significantly reduced, though a trend in this direction was apparent. The small numbers in this comparison may be a limiting factor. In addition overall plasma TRP was significantly reduced in depressed women but not depressed men. However, some caution is needed in interpretation of this gender difference as, in contrast to our data, a number of studies have reported that plasma TRP is lower in women than men (Armstrong and Stave, 1973; Hagenfeldt et al., 1984). Taken together our findings suggest that weight loss is not by itself sufficient to explain reduced plasma TRP concentrations in depression, though it may be a contributory factor, especially in women, in whom weight loss correlated inversely with plasma TRP concentration. However, distinguishing the role of weight loss from melancholia in females in our study is made difficult because of the high concordance of reported weight loss with melancholia in women (seven out of nine subjects) compared to men (two out of eight subjects). The results of the multiple regression analysis also support the importance of a diagnosis of melancholia as a correlate of low plasma TRP in depressed patients. Further, the presence of melancholic depression even in the absence of reported weight loss was still associated with significantly lowered plasma TRP concentrations. Previous studies of plasma total TRP and TRP: BCAA ratio in depressed patients have yielded contradictory findings (see Introduction), some of which may be explained if reductions in plasma TRP concentration are particularly confined to patients with melancholic depression. Interestingly Curzon et al. (1979) reported lower total and free plasma TRP concentrations in depressed patients with retardation. It is important, however, to acknowledge the limitations of our study design, particularly with regard to our attempt to distinguish weight loss and melancholia. Our weight loss assessment in depressed patients depended on retrospective reports and accordingly our clinical measures may be too imprecise to disentangle weight loss reliably from a diagnosis of melancholic depression. A
190
further difficulty is that we were not able to measure plasma free TRP concentrations which also influences the availability of plasma TRP for brain 5-HT synthesis (Curzon, 1979, 1988). Some studies in depressed patients, for example, have found decreased plasma free TRP concentrations but normal total TRP levels (Coppen et al., 1973). We found a significant decrease in plasma TRP in depressed women, but not men. This sex difference, which was apparent even though the proportion of patients with melancholia and weight loss was similar in both sexes, might be a further source of contradictory results between studies. Interestingly, Meltzer (1989) also reported that decreased plasma TRP concentrations were confined to female depressives. The lower plasma TRP concentrations in depressed women in our study might be explained if the effects of melancholia and weight loss on plasma total TRP concentration were additive because, as mentioned above, more depressed women than men had both melancholia and weight loss. Another possible explanation is that women are more vulnerable than men to factors that tend to lower plasma TRP. In our dieting study in healthy subjects, for example, we found that women showed a greater decrease in plasma TRP than men, despite achieving a similar degree of weight loss (Anderson et al., 1990). Furthermore, Delgado et al. (1989) reported that a chronic low TRP diet in normal volunteers reduced plasma TRP more in women than in men. What might be the relationship of reduced TRP availability to the syndrome of depressive illness? It is possible that in some patients low plasma TRP levels and TRP: BCAA ratio may lead to reduced brain 5-HT function and subsequently the development of depressive illness. However, another possibility, which we favour, is that low plasma TRP availability may be an epiphenomenon of depression resulting from factors associated with melancholic illness, for example weight loss and cortisol hypersecretion (Badawy, 1977; Maes et al., 1990). While, in this hypothesis, reduced plasma TRP availability is a secondary consequence of certain features of depression, it could nevertheless be an important additional complication. Both we and others have obtained evidence from neuroendocrine challenge tests that
brain 5-HT function is abnormal in depression independent of the presence of low plasma TRP concentrations (Heninger et al., 1984; Cowen and Charig, 1987; Meltzer, 1989; Deakin et al., 1990). Accordingly an additional reduction in plasma TRP concentration and TRP : BCAA ratio could further compromise brain 5-HT function and lead to a worsening of the depressive state. In this respect it is of interest that a low TRP : BCAA ratio may predict response to drugs that enhance brain 5-HT function (Mnrller et al., 1980, 1990) and that TRP supplements have been shown to potentiate the effects of certain antidepressants, notably monoamine oxidase inhibitors (Coppen et al., 1963; Paykel and Hale, 1986). Acknowledgements I.M.A. was a Medical Research Council Training Fellow. M.P.-B. was the recipient of a Science and Engineering Research Council Studentship. The study was supported by grants from the Oxford Region and District Health Authorities. References Anderson, I.M., Crook, W.S., Gartsidc, S.E., Parry-Billings, M., Newsholme, E.A. and Cowen, P.J. (1989) Effect of moderate weight loss on prolactin secretion in normal female volunteers. Psychiatr. Res. 29, 161-167. Anderson, I.M., Williams, J., Parry-Billings, M., Newsholme, E.A., Fairbum, C.G. and Cowen, P.J. (1990) Dieting lowers plasma tryptophan in normal volunteers. Br. J. Clin. Pharmac. 29, 591P-592P. Armstrong, M.D. and Stave, U. (1973) A study of plasma free amino acid levels. II. Normal values for children and adults. Metabolism 22, 561-569. Badawy, A.A.-B. (1977) Minireview: the functions and regulations of tryptophan pyrrolase. Life Sci. 21, 755-768. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J.E. and Erbaugh, J.E. (1961) An inventory for measuring depression. Arch. Gen. Psychiatry 4, 561-567. Bloxam, D.L. and Warren, W.H. (1974) Error in the determination of tryptophan by the method of Denkla and Dewey. A revised procedure. Anal. Biochem. 60, 621-624. Coppen. A.J. (1967) The biochemistry of affective disorder. Br. J. Psychiatry 113, 1237-1264. Coppen, A., Shaw, D.M. and Farrell, J.P. (1963) Potentiation of the antidepressive effect of a monoamine oxidase inhibitor by tryptophan. Lancet i. 79-81. Coppen. A., Eccleston, E.G. and Peet, M. (1973) Total and free tryptophan concentration in the plasma of depressive patients. Lancet ii, 60-63.
191 Cowen, P.J. (1988) Recent views on the role of 5-hydroxytryptamine in depression. Curr. Opin. Psychiatry 1, 56-59. Cowen, P.J. and Charig, E.M. (1987) Neuroendocrine responses to intravenous tryptophan in major depression. Arch. Gen. Psychiatry 44, 958-966. Cowen, P.J., Parry-Billings, M. and Newsholme, E.A. (1989) Decreased plasma tryptophan in major depression. J. Affect. Disord. 16, 27-31. Curzon, G. (1979) Relationships between plasma, CSF and brain tryptophan. J. Neural Transm. Suppl. 15, 81-92. Curzon, G. (1988) Feeding, stress, exercise and the supply of amino acids to the brain. In: G. Huether (Ed.), Amino Acid Availability and Brain Function in Health and Disease. Springer-Verlag, Berlin, pp. 39-49. Curzon, G., Kantamaneni, B.D., Lader, M.H. and Greenwood, M.H. (1979) Tryptophan disposition in psychiatric patients before and after stress. Psychol. Med. 9, 457-463. Deakin, J.F.W., Pennell, I., Upadhyaya, A.J. and Lofthouse, R. (1990) A neuroendocrine study of 5-HT function in depression: evidence for biological mechanisms of endogenous and psychosocial causation. Psychopharmacology 101, 8592. Delgado, P.L., Chamey, D.S., Price, L.H., Landis, H. and Heninger, G.R (1989) Neuroendocrine and behavioural effects of dietary tryptophan restriction in healthy subjects. Life Sci. 45, 2323-2332. Delgado, P.L., Chamey, D.S., Price., L.H., Aghajanian, G.K., Landis, H. and Heninger, G.R. (1990) Serotonin function and the mechanism of antidepressant action: reversal of antidepressant-induced remission by rapid depletion of plasma tryptophan. Arch. Gen. Psychiatry 47, 411-418. deMontigny, C. and Blier, P. (1984) Effects of antidepressant treatment on 5-HT neurotransmission: electrophysiological and clinical studies. Adv. Biochem. Psychopharmacol. 39, 223-240. DeMyer, M.K., Shea, P., Hendrie, H.C. and Yoshimura, N.Y. (1981) Plasma tryptophan and five other amino acids in depressed and normal subjects. Arch. Gen. Psychiatry 38, 642-646. Denkla, W.D. and Dewey, H.K. (1967) The determination of tryptophan in plasma, liver and urine. J. Lab. Clin. Med. 69, 160-169. Fernstrom, J.D. (1979) Diet induced changes in plasma ammo acid pattern: effects on the brain uptake of large neutral ammo acids and on brain serotonin synthesis. J. Neural. Transm. Suppl. 15, 55-67. Femstrom, J.D. (1983) Role of precursor availability in control of monoamine synthesis in brain. Physiol. Rev. 63,484-546. Femstrom, J.D., Larin, F. and Wurtman, R.J. (1973) Correlations between brain tryptophan and plasma neutral ammo acids following food consumption in rats. Life Sci. 13, 517-524.
Goodwin, G.M., Cowen, P.J., Fairbum, C.G., Parry-Billings, M., Calder, P.C. and Newsholme, E.A. (1990) Decreased plasma tryptophan levels in dieting. Br. Med. J. 300, 14991500. Hagenfeldt, L., Bjerkenstedt, L., Edman, G., Sedvall, C. and Wiesel, F.A. (1984) Amino acids in plasma and CSF and monoamine metabolites in CSF: interrelationship in healthy subjects. J. Neurochem. 42, 833-837. Hamilton, M. (1967) Development of a rating scale for primary depressive illness. Br. J. Sot. Clin. Psychol. 6, 278-296. Heninger, G.R., Charney, D.S. and Stemberg, D.E. (1984) Serotonergic function in depression: prolactin response to intravenous tryptophan in depressed patients and healthy subjects. Arch. Gen. Psychiatry 41, 398-402. Joseph, M.S., Brewerton, T.D., Reus, V.I. and Stebbins, G.T. (1984) Plasma t_-tryptophan/neutral ammo acid ratio and dexamethasone suppression in depression. Psychiatr. Res. 11, 185-192. Livesey, G. and Lund, P. (1980) Enzymic determination of branched-chain amino acids and 2-oxoacids in rat tissues-transfer of 2-oxoacids from skeletal muscle to liver in viva. B&hem. J. 188, 7055713. Maes, M., De Ruyter, M., Hobin, P. and Suy, E. (1987) Relationship between the dexamethasone suppression test and the L-tryptophan/competing amino acid ratio in depression. Psychiatr. Res. 21, 323-335. Maes, M., Schotte, C., Scharpe, S., Martin, M. and Blockx, P. (1990) The effects of glucocorticoids on the availability of L-tryptophan and tyrosine in the plasma of depressed patients. J. Affect. Disord. 18, 121-127. Meltzer, H.Y. (1989) Serotonergic function in depression. Br. J. Psychiatry 155. Suppl. 8, 25-31. Moller, S.E., Kirk, L. and Honor& P. (1979) Free and total plasma tryptophan in endogenous depression. J. Affect. Disord. 1, 69-76. Moller, SE., Kirk, L. and Honor&, P. (1980) Relationship between plasma ratio of tryptophan to competing ammo acids and the response to L-tryptophan treatment in endogenously depressed patients. J. Affect. Disord. 2, 47-59. Meller, SE., Beth, P., Bjerrum, H., Bojholm, S., Butler, B., Folker, H., Gram, L.F., Larsen, J.K., Lauritzen, L., Loldrup, D., Munk-Andersen, E., Odum, K. and Rafaelsen, O.J. (1990) Plasma ratio of tryptophan/neutral amino acids in relation to clinical response to paroxetine and clomipramine in patients with major depression. J. Affect. Disord. 18, 59-66. Paykel, E.S. and Hale, A.S. (1986) Recent advances in the treatment of depression. In: J.F.W. Deakin (Ed.), The Biology of Depression. Gaskell Press, London, pp. 153-173. Riley, C.J. and Shaw, D.M. (1976) Total and nonbound tryptophan in unipolar illness. Lancet ii, 1249.
of Affectice
Disorders,
20 (1990)
185
185-191
Elsevier
JAD 00756
Decreased plasma tryptophan concentration in major depression: relationship to melancholia and weight loss 1-M. Anderson, MRC
M. Parry-Billings
Unit of Clinical Pharmacology and
I, E.A. Newsholme
and Uniuersity Department
t Cellular Nutrition Research
Group, Department
of Psychiatry
I, J.R. Poortmans Littkmore
of Biochemistry,
Hospital,
Littlemore,
Unrversity of Oxford,
1
and P.J. Cowen Oxford OX4 4XN,
Oxford OXI 3QlJ.
U.K.
U.K.
(Received 25 April 1990) (Revision received 12 July 1990) (Accepted 17 July 1990)
Summary Plasma total tryptophan (TRP) concentration was significantly lower in 31 patients with major depression compared to a healthy control group. The ratio of plasma TRP concentration to that of other branch chain amino acids (the TRP: BCAA ratio) was also decreased. Further analysis revealed that the decrease in plasma TRP and TRP : BCAA ratio was most apparent in patients with major depression and melancholia. Overall, women but not men had significantly decreased plasma tryptophan concentrations, perhaps because of a contributory effect of weight loss; this latter effect, however, could not be distinguished clearly from a diagnosis of melancholia. Our data suggest that in some depressed patients, reductions in plasma tryptophan availability may contribute to abnormalities in brain 5hydroxytryptamine function.
Key words: Tryptophan;
Depression;
Melancholia;
Introduction There is continued interest in the role of brain serotonin (5-hydroxytryptamine, 5-HT) pathways in the aetiology of depression (Coppen, 1967; Cowen, 1988) and the mediation of antidepressant effects of pharmacological treatments (deMon-
Address for correspondence: Dr. I.M. Anderson, MRC Unit of Clinical Pharmacology and University Department of Psychiatry, Littlemore Hospital, Littlemore, Oxford OX4 4XN, U.K. 0165-0327/90/$03.50
0 1990 Elsevier Science Publishers
Branch
chain amino
acids;
5-Hydroxytryptamine
tigny and Blier, 1984; Delgado et al., 1990). The synthesis of brain 5-HT is dependent on the availability of the 5-HT precursor, tryptophan (TRP) (Fernstrom, 1983), and several studies have measured the circulating concentration of TRP in depressed patients. The results of these studies have been conflicting with reports of both normal (Riley and Shaw, 1976; Msller et al., 1979) and decreased TRP concentrations (DeMyer et al., 1981; Joseph et al., 1984; Maes et al.,-1987). An important consideration is that the availability of TRP for brain 5-HT synthesis is determined by two factors: first, the amount of TRP in the
B.V. (Biomedical
Division)
186
plasma unbound to albumin (plasma free TRP) (Curzon, 1979,1988); and second, the total amount of plasma TRP in relation to branch chain amino acids (BCAA) that compete with TRP for transport across the blood-brain barrier (the TRP: BCAA ratio) (Fernstrom, 1973, 1979). One reason for the conflicting results of TRP studies in depressed patients may be that low plasma TRP concentrations are found only in particular clinical subgroups. In a recent study, for example, we found that both plasma total TRP and the TRP : BCAA ratio were significantly decreased in a small group of patients with DSM-III major depression and melancholia (Cowen et al., 1989). This reduction was most clearly apparent in female patients who reported weight loss during the affective episode and is consistent with our studies in normal volunteers that have shown decreased plasma TRP concentrations following weight loss through dieting (Anderson et al., 1989, 1990; Goodwin et al., 1990). Interestingly, Maes et al. (1987) found that plasma total TRP and TRP : BCAA ratio were significantly decreased in patients meeting diagnostic criteria for DSM-III major depression with melancholia or psychosis but not in non-melancholic depressives. Since many of the melancholic and psychotic patients would have been expected to have experienced significant weight loss the relative importance of depressive subgroup and weight loss as determinants of reduced plasma TRP availability is unclear. The purpose of the present study was to confirm and extend our original observations on plasma TRP and TRP : BCAA ratio in major depression and to investigate further the role of weight loss, sex and melancholic subtype as determinants of reduced plasma TRP availability in depressed patients. Methods
Subjects We studied 31 depressed patients (16 male and 15 female) who on the basis of a semi-structured interview met DSM-III criteria for major depression, Sixteen subjects met DSM-III criteria for major depression with melancholia. The mean score for the group on the 17-item Hamilton De-
pression Rating Scale (HDRS) (Hamilton, 1967) was 22.8 (range 15-34). All subjects had been free of antidepressant medication for at least 4 weeks and most for considerably longer. Five subjects were taking benzodiazepines for night sedation but no other medication was allowed. The subjects studied included 12 patients whose data have been previously reported (Cowen et al., 1989). Weight loss was assessed using the weight loss rating item on the Beck Depression Inventory (BDI) (Beck et al., 1961) which asks subjects how much weight they have lost ‘recently’ (explained to subjects as weight loss in the last month). A specially designed observer-rated questionnaire (available from the authors) was also administered, again with the aim of detecting weight loss over the previous month. Before completing these assessments subjects were weighed and informed of the result. Both assessments were used to decide likely weight loss. Five patients had weight loss of 5-10 lbs, eight patients had weight loss of 11-15 lbs and two patients had weight loss of more than 15 lbs. These groups did not differ significantly in plasma TRP measures and so were combined to make a single ‘weight loss’ group (i.e., weight loss of > 5 lbs in the last month; n = 15). The results of this group were compared with a ‘no weight loss group’ (i.e., weight loss of less than 5 lbs in the last month; n = 16). An absolute estimate of weight loss for each subject was also made. Overall male patients reported weight loss during the last month of 5.3 + 1.8 lbs and females, 5.6 + 1.4 lbs (mean + SEM). For comparison with the patients we selected 31 control subjects (16 male, 15 female) with no history of psychiatric disorder and no change in weight in the last 2 months. As far as possible they were age- and weight-matched with the patient group (mean age of patients 33.5 years, range 20-55 years; mean age of controls 33.0 years, range 21-53 years; mean weight of patients 149 lbs, range 89-230 lbs; mean weight of controls 152 lbs, range 99-226 lbs). Sampling procedure and biochemical analysis Both patients and controls were sampled after an overnight fast. Plasma was separated from venous blood by centrifugation and stored at -20°C until assay. The concentrations of BCAA
187
(valine, leucine and isoleucine) were determined by an enzymatic method (Livesey and Lund, 1980) on neutralised perchloric acid extracts of plasma. The TRP concentration was determined on trichloroacetic acid extracts using the fluorometric method of Denkla and Dewey (1967) with the modification described by Bloxam and Warren (1974). Assays were carried out blind to diagnosis.
50
I
T
Statistics
Comparisons between patient and control groups were assessed by Student’s unpaired t-test (two-tailed). The weight of the different variables in predicting plasma TRP concentration was determined by stepwise multiple linear regression. Correlations were carried out using Pearson’s product moment. Results The patients had a significantly lower mean plasma total TRP and also a significantly decreased TRP : BCAA ratio compared to controls. The BCAA concentration did not differ between patients and controls, so that the reduction in TRP : BCAA ratio was attributable to the decrease in plasma TRP concentration (Table 1). Subdivision of the patients by sex showed that plasma TRP was significantly reduced in women but not men, and the same was true of the TRP : BCAA ratio (Table 1). In patients with weight loss, plasma TRP concentration was significantly reduced, an effect not observed in patients without weight loss. In both groups the TRP : BCAA ratio showed similar reductions of borderline statistical significance (P <
TABLE
CON
DEP
(31)
(31)
M-W-
M-W+
M+W-
M+W+
(9)
(6)
(7)
(9)
Fig. 1. Plasma TRP concentrations in depressed patients (DEP) and controls (CON). The patients were subdivided according to the presence ( + ) or absence (- ) of melancholia (M) and weight loss (W). The number in parentheses is the number of patients in each category. * * P c 0.005; * * * P i 0.001.
0.09) (Table 2). Further subdivision by sex revealed that in women but not men, weight loss was associated with significantly reduced plasma TRP and TRP : BCAA ratio (Tables 3 and 4). Patients with melancholia had significantly reduced plasma TRP concentration and TRP: BCAA ratio compared to controls (Table 2) and the same pattern of results was seen when male and female melancholic depressives were consid-
1
PLASMA CONCENTRATIONS OF TRYPTOPHAN (TRP) AND BRANCH LEUCINE, ISOLEUCINE) IN CONTROLS AND DEPRESSED PATIENTS n
All subjects Females Males
TRP (PM)
CHAIN
AMINO
ACIDS
Ratio (TRP
BCAA (PM)
(BCAA)
(VALINE,
: BCAA)
Controls
Patients
Controls
Patients
Controls
Patients
Controls
Patients
31 15 16
31 15 16
45.4+ 2.0 46.9 f 3.5 44.0 f 2.3
38.2* 1.6 * * 36.9* 2.1 * 39.7 f 2.4
376.1 f 16.3 344.1 k 23.8 406.1 * 20.2
363.5 f 12.9 357.6 f 16.0 369.1 k 20.4
0.128 +0.008 0.145_+0.013 0.112 * 0.007
0.109 + 0.005 * 0.105 f 0.006 * * 0.112~0.009
Values are mean + SEM. * P < 0.05, * * P < 0.01, less than controls.
188 TABLE
2
PLASMA CONCENTRATIONS OF TRYPTOPHAN (TRP) AND BRANCH CHAIN LEUCINE, ISOLEUCINE) IN CONTROLS AND DEPRESSED PATIENTS ACCORDING
AMINO ACIDS (BCAA) TO SUBGROUP
: BCAA)
Subjects
n
TRP (PM)
BCAA (PM)
Ratio (TRP
Controls
31
45.4k2.0
376.1 + 16.3
0.128+0.008
Patients No weight loss Weight loss No melancholia Melancholia
16 15 15 16
40.9 35.5 42.2 34.7
382.3 343.5 349.9 376.3
0.109+0.007 0.108 + 0.008 0.125kO.008 0.094 + 0.005 * * *
* 2.2 +2.1 * * + 2.5 * 1.5 * * *
+ k + +
14.9 20.7 22.3 13.6
(VALINE.
Values are mean + SEM. * * P i 0.002, ***P
ered separately (Tables 3 and 4). Subdivision of subjects into those with melancholic and nonmelancholic depression with and without weight loss yielded small cell sizes, but plasma TRP was significantly reduced in patients with melancholic depression without concomitant weight loss and
TABLE
further reduced in patients with both weight loss and a diagnosis of melancholia. Subjects with weight loss who did not meet criteria for melancholia did not have significantly decreased plasma TRP concentrations compared to controls (P > 0.1) (Fig. 1).
3
PLASMA CONCENTRATIONS OF TRYF’TOPHAN LEUCINE, ISOLEUCINE) IN FEMALE CONTROLS
(TRP) AND BRANCH CHAIN AMINO ACIDS (BCAA) (VALINE, AND DEPRESSED PATIENTS ACCORDING TO SUBGROUP
: BCAA)
Female subjects
n
TRP (PM)
BCAA (PM)
Ratio (TRP
Controls Patients No weight loss Weight loss No melancholia Melancholia
15
46.9+ 3.5
344.1 + 23.8
0.145+0.013
44.0 32.2 41.2 33.1
382.0,25.8 341.3 f 19.7 357.9 + 30.7 357.4+ 16.0
0.117+0.007 0.097 * 0.009 * * 0.118f0.008 0.094 + 0.008 * *
6 9 7 8
+ * + f
1.9 2.7 * * 3.0 2.3 * *
Values are mean k SEM. * * P < 0.005, less than controls.
TABLE
4
PLASMA CONCENTRATIONS OF TRYI’TOPHAN (TRP) AND BRANCH CHAIN AMINO LEUCINE, ISOLEUCINE) IN MALE CONTROLS AND DEPRESSED PATIENTS ACCORDING
ACIDS (BCAA) TO SUBGROUP
: BCAA)
Male subjects
n
TRP (PM)
BCAA (PM)
Ratio (TRP
Controls Patients No weight loss Weight loss No melancholia Melancholia
16
44.Ok2.3
406.1 k 20.2
0.118+0.007
10 6 8 8
39.1 40.6 43.1 36.3
382.5 + 346.7 + 343.0 f 395.1+
0.105+0.011 0.125 f 0.014 0.131 kO.013 0.093 + 0.006 *
Values are mean + SEM. * P i 0.05, less than controls.
53.2 + 3.6 i4.1 + 1.8 *
19.2 45.4 33.9 20.9
(VALINE,
189
The above findings were supported by the results of a stepwise multiple linear regression analysis with TRP as the dependent variable and melancholia, weight loss and sex as the independent variables. The only significant prediction of plasma TRP concentration was a diagnosis of melancholia (R2 = 0.19, P = 0.014). A similar finding was obtained where the TRP : BCAA ratio was used as the dependent variable ( R2 = 0.29, P = 0.002). Correlation analyses revealed that plasma TRP did not correlate significantly with absolute weight loss in the patients as a whole (r = -0.15, P = 0.41) or in male patients (1. = 0.14, P = 0.62) but there was a significant negative correlation with weight loss in female patients (I = -0.64, P = 0.01); however, there was no relationship between plasma TRP and severity of depression as measured by HDRS score (r = 0.11, P = 0.55) or between HDRS score and TRP: BCAA ratio (r = -0.11, P = 0.55). Discussion
The present data confirm our previous findings that plasma total TRP concentration is decreased in major depression and that, as a consequence of this, the TRP: BCAA ratio is also lowered. This suggests that during some depressive disorders, the availability of plasma TRP for brain 5-HT synthesis is reduced. While depressed women who reported weight loss had the lowest plasma tryptophan concentrations, our data suggest that a diagnosis of melancholic depression, rather than sex or presence of weight loss, is the major determinant of low plasma TRP in depressed patients. The findings of a reduced plasma TRP concentration and TRP : BCAA ratio in melancholic patients are in agreement with the data of Maes et al. (1987). However, weight loss is one of the diagnostic criteria for melancholia and it is important, if possible, to distinguish between the effects of weight loss and melancholia on plasma TRP. This is especially pertinent given our previous findings that in both normal men and women who lose weight by dieting, plasma TRP concentration and TRP : BCAA ratio are significantly reduced (Anderson et al., 1989, 1990; Goodwin et al., 1990). In the present study weight loss, as
judged by our ratings, was also associated with decreased plasma TRP in depressed patients. However, in the group of patients with weight loss but no melancholia, plasma TRP was not significantly reduced, though a trend in this direction was apparent. The small numbers in this comparison may be a limiting factor. In addition overall plasma TRP was significantly reduced in depressed women but not depressed men. However, some caution is needed in interpretation of this gender difference as, in contrast to our data, a number of studies have reported that plasma TRP is lower in women than men (Armstrong and Stave, 1973; Hagenfeldt et al., 1984). Taken together our findings suggest that weight loss is not by itself sufficient to explain reduced plasma TRP concentrations in depression, though it may be a contributory factor, especially in women, in whom weight loss correlated inversely with plasma TRP concentration. However, distinguishing the role of weight loss from melancholia in females in our study is made difficult because of the high concordance of reported weight loss with melancholia in women (seven out of nine subjects) compared to men (two out of eight subjects). The results of the multiple regression analysis also support the importance of a diagnosis of melancholia as a correlate of low plasma TRP in depressed patients. Further, the presence of melancholic depression even in the absence of reported weight loss was still associated with significantly lowered plasma TRP concentrations. Previous studies of plasma total TRP and TRP: BCAA ratio in depressed patients have yielded contradictory findings (see Introduction), some of which may be explained if reductions in plasma TRP concentration are particularly confined to patients with melancholic depression. Interestingly Curzon et al. (1979) reported lower total and free plasma TRP concentrations in depressed patients with retardation. It is important, however, to acknowledge the limitations of our study design, particularly with regard to our attempt to distinguish weight loss and melancholia. Our weight loss assessment in depressed patients depended on retrospective reports and accordingly our clinical measures may be too imprecise to disentangle weight loss reliably from a diagnosis of melancholic depression. A
190
further difficulty is that we were not able to measure plasma free TRP concentrations which also influences the availability of plasma TRP for brain 5-HT synthesis (Curzon, 1979, 1988). Some studies in depressed patients, for example, have found decreased plasma free TRP concentrations but normal total TRP levels (Coppen et al., 1973). We found a significant decrease in plasma TRP in depressed women, but not men. This sex difference, which was apparent even though the proportion of patients with melancholia and weight loss was similar in both sexes, might be a further source of contradictory results between studies. Interestingly, Meltzer (1989) also reported that decreased plasma TRP concentrations were confined to female depressives. The lower plasma TRP concentrations in depressed women in our study might be explained if the effects of melancholia and weight loss on plasma total TRP concentration were additive because, as mentioned above, more depressed women than men had both melancholia and weight loss. Another possible explanation is that women are more vulnerable than men to factors that tend to lower plasma TRP. In our dieting study in healthy subjects, for example, we found that women showed a greater decrease in plasma TRP than men, despite achieving a similar degree of weight loss (Anderson et al., 1990). Furthermore, Delgado et al. (1989) reported that a chronic low TRP diet in normal volunteers reduced plasma TRP more in women than in men. What might be the relationship of reduced TRP availability to the syndrome of depressive illness? It is possible that in some patients low plasma TRP levels and TRP: BCAA ratio may lead to reduced brain 5-HT function and subsequently the development of depressive illness. However, another possibility, which we favour, is that low plasma TRP availability may be an epiphenomenon of depression resulting from factors associated with melancholic illness, for example weight loss and cortisol hypersecretion (Badawy, 1977; Maes et al., 1990). While, in this hypothesis, reduced plasma TRP availability is a secondary consequence of certain features of depression, it could nevertheless be an important additional complication. Both we and others have obtained evidence from neuroendocrine challenge tests that
brain 5-HT function is abnormal in depression independent of the presence of low plasma TRP concentrations (Heninger et al., 1984; Cowen and Charig, 1987; Meltzer, 1989; Deakin et al., 1990). Accordingly an additional reduction in plasma TRP concentration and TRP : BCAA ratio could further compromise brain 5-HT function and lead to a worsening of the depressive state. In this respect it is of interest that a low TRP : BCAA ratio may predict response to drugs that enhance brain 5-HT function (Mnrller et al., 1980, 1990) and that TRP supplements have been shown to potentiate the effects of certain antidepressants, notably monoamine oxidase inhibitors (Coppen et al., 1963; Paykel and Hale, 1986). Acknowledgements I.M.A. was a Medical Research Council Training Fellow. M.P.-B. was the recipient of a Science and Engineering Research Council Studentship. The study was supported by grants from the Oxford Region and District Health Authorities. References Anderson, I.M., Crook, W.S., Gartsidc, S.E., Parry-Billings, M., Newsholme, E.A. and Cowen, P.J. (1989) Effect of moderate weight loss on prolactin secretion in normal female volunteers. Psychiatr. Res. 29, 161-167. Anderson, I.M., Williams, J., Parry-Billings, M., Newsholme, E.A., Fairbum, C.G. and Cowen, P.J. (1990) Dieting lowers plasma tryptophan in normal volunteers. Br. J. Clin. Pharmac. 29, 591P-592P. Armstrong, M.D. and Stave, U. (1973) A study of plasma free amino acid levels. II. Normal values for children and adults. Metabolism 22, 561-569. Badawy, A.A.-B. (1977) Minireview: the functions and regulations of tryptophan pyrrolase. Life Sci. 21, 755-768. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J.E. and Erbaugh, J.E. (1961) An inventory for measuring depression. Arch. Gen. Psychiatry 4, 561-567. Bloxam, D.L. and Warren, W.H. (1974) Error in the determination of tryptophan by the method of Denkla and Dewey. A revised procedure. Anal. Biochem. 60, 621-624. Coppen. A.J. (1967) The biochemistry of affective disorder. Br. J. Psychiatry 113, 1237-1264. Coppen, A., Shaw, D.M. and Farrell, J.P. (1963) Potentiation of the antidepressive effect of a monoamine oxidase inhibitor by tryptophan. Lancet i. 79-81. Coppen. A., Eccleston, E.G. and Peet, M. (1973) Total and free tryptophan concentration in the plasma of depressive patients. Lancet ii, 60-63.
191 Cowen, P.J. (1988) Recent views on the role of 5-hydroxytryptamine in depression. Curr. Opin. Psychiatry 1, 56-59. Cowen, P.J. and Charig, E.M. (1987) Neuroendocrine responses to intravenous tryptophan in major depression. Arch. Gen. Psychiatry 44, 958-966. Cowen, P.J., Parry-Billings, M. and Newsholme, E.A. (1989) Decreased plasma tryptophan in major depression. J. Affect. Disord. 16, 27-31. Curzon, G. (1979) Relationships between plasma, CSF and brain tryptophan. J. Neural Transm. Suppl. 15, 81-92. Curzon, G. (1988) Feeding, stress, exercise and the supply of amino acids to the brain. In: G. Huether (Ed.), Amino Acid Availability and Brain Function in Health and Disease. Springer-Verlag, Berlin, pp. 39-49. Curzon, G., Kantamaneni, B.D., Lader, M.H. and Greenwood, M.H. (1979) Tryptophan disposition in psychiatric patients before and after stress. Psychol. Med. 9, 457-463. Deakin, J.F.W., Pennell, I., Upadhyaya, A.J. and Lofthouse, R. (1990) A neuroendocrine study of 5-HT function in depression: evidence for biological mechanisms of endogenous and psychosocial causation. Psychopharmacology 101, 8592. Delgado, P.L., Chamey, D.S., Price, L.H., Landis, H. and Heninger, G.R (1989) Neuroendocrine and behavioural effects of dietary tryptophan restriction in healthy subjects. Life Sci. 45, 2323-2332. Delgado, P.L., Chamey, D.S., Price., L.H., Aghajanian, G.K., Landis, H. and Heninger, G.R. (1990) Serotonin function and the mechanism of antidepressant action: reversal of antidepressant-induced remission by rapid depletion of plasma tryptophan. Arch. Gen. Psychiatry 47, 411-418. deMontigny, C. and Blier, P. (1984) Effects of antidepressant treatment on 5-HT neurotransmission: electrophysiological and clinical studies. Adv. Biochem. Psychopharmacol. 39, 223-240. DeMyer, M.K., Shea, P., Hendrie, H.C. and Yoshimura, N.Y. (1981) Plasma tryptophan and five other amino acids in depressed and normal subjects. Arch. Gen. Psychiatry 38, 642-646. Denkla, W.D. and Dewey, H.K. (1967) The determination of tryptophan in plasma, liver and urine. J. Lab. Clin. Med. 69, 160-169. Fernstrom, J.D. (1979) Diet induced changes in plasma ammo acid pattern: effects on the brain uptake of large neutral ammo acids and on brain serotonin synthesis. J. Neural. Transm. Suppl. 15, 55-67. Femstrom, J.D. (1983) Role of precursor availability in control of monoamine synthesis in brain. Physiol. Rev. 63,484-546. Femstrom, J.D., Larin, F. and Wurtman, R.J. (1973) Correlations between brain tryptophan and plasma neutral ammo acids following food consumption in rats. Life Sci. 13, 517-524.
Goodwin, G.M., Cowen, P.J., Fairbum, C.G., Parry-Billings, M., Calder, P.C. and Newsholme, E.A. (1990) Decreased plasma tryptophan levels in dieting. Br. Med. J. 300, 14991500. Hagenfeldt, L., Bjerkenstedt, L., Edman, G., Sedvall, C. and Wiesel, F.A. (1984) Amino acids in plasma and CSF and monoamine metabolites in CSF: interrelationship in healthy subjects. J. Neurochem. 42, 833-837. Hamilton, M. (1967) Development of a rating scale for primary depressive illness. Br. J. Sot. Clin. Psychol. 6, 278-296. Heninger, G.R., Charney, D.S. and Stemberg, D.E. (1984) Serotonergic function in depression: prolactin response to intravenous tryptophan in depressed patients and healthy subjects. Arch. Gen. Psychiatry 41, 398-402. Joseph, M.S., Brewerton, T.D., Reus, V.I. and Stebbins, G.T. (1984) Plasma t_-tryptophan/neutral ammo acid ratio and dexamethasone suppression in depression. Psychiatr. Res. 11, 185-192. Livesey, G. and Lund, P. (1980) Enzymic determination of branched-chain amino acids and 2-oxoacids in rat tissues-transfer of 2-oxoacids from skeletal muscle to liver in viva. B&hem. J. 188, 7055713. Maes, M., De Ruyter, M., Hobin, P. and Suy, E. (1987) Relationship between the dexamethasone suppression test and the L-tryptophan/competing amino acid ratio in depression. Psychiatr. Res. 21, 323-335. Maes, M., Schotte, C., Scharpe, S., Martin, M. and Blockx, P. (1990) The effects of glucocorticoids on the availability of L-tryptophan and tyrosine in the plasma of depressed patients. J. Affect. Disord. 18, 121-127. Meltzer, H.Y. (1989) Serotonergic function in depression. Br. J. Psychiatry 155. Suppl. 8, 25-31. Moller, S.E., Kirk, L. and Honor& P. (1979) Free and total plasma tryptophan in endogenous depression. J. Affect. Disord. 1, 69-76. Moller, SE., Kirk, L. and Honor&, P. (1980) Relationship between plasma ratio of tryptophan to competing ammo acids and the response to L-tryptophan treatment in endogenously depressed patients. J. Affect. Disord. 2, 47-59. Meller, SE., Beth, P., Bjerrum, H., Bojholm, S., Butler, B., Folker, H., Gram, L.F., Larsen, J.K., Lauritzen, L., Loldrup, D., Munk-Andersen, E., Odum, K. and Rafaelsen, O.J. (1990) Plasma ratio of tryptophan/neutral amino acids in relation to clinical response to paroxetine and clomipramine in patients with major depression. J. Affect. Disord. 18, 59-66. Paykel, E.S. and Hale, A.S. (1986) Recent advances in the treatment of depression. In: J.F.W. Deakin (Ed.), The Biology of Depression. Gaskell Press, London, pp. 153-173. Riley, C.J. and Shaw, D.M. (1976) Total and nonbound tryptophan in unipolar illness. Lancet ii, 1249.