University Department of Psychiatry, Oxford
Correspondence: Professor P.J. Cowen, University Department of Psychiatry, Warneford Hospital, Oxford OX37JX
Declaration of interest Funding from the Medical Research Council.
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Aim To test the hypothesis that women with a history of major depression would show impaired regulation of brain 5-HT function during a period of dieting-induced tryptophan depletion.
Method Women with and without a history of major depression were placed on a daily 1000 kcal (approximately 4200 kj) diet for three weeks. Before the diet and in the final week we measured fasting plasma tryptophan levels and the prolactin response to an intravenous tryptophan challenge.
Results Dieting lowered plasma tryptophan levels equivalently in women with and without a history of depression. In women without a history of depression, dieting also increased the prolactin response to tryptophan. This increase did not occur in women with a history of depression.
Conclusions Women with a history of depression showed impaired regulation of brain 5-HT function in response to dieting.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
METHOD |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Study design
Subjects were placed on a three-week calorie-controlled diet of 1000 kcal
(approximately 4200 kJ) per day, as previously described in Anderson et
al (1990). During the
diet, subjects completed a daily diary recording what they ate, and rated
three 100 mm visual analogue scales (VAS), happy, sad and irritable (on
a scale of 0-100: 0, not at all; 100, extremely). Neuroendocrine testing was
carried out before dieting began and at the end of the third week of dieting.
The onset of dieting was delayed by one week after the first neuroendocrine
test to ensure that both tests took part in the early- to mid-follicular phase
of two consecutive menstrual cycles. We have previously found that subjects
who fail to lose more than 2 kg in weight during this diet do not reliably
increase their prolactin response to tryptophan (details available from author
on request). Subjects who failed to achieve this weight reduction were
therefore excluded from the present study and did not undergo the second
tryptophan test.
Neuroendocrine testing
Neuroendocrine testing was carried out after an overnight fast. Subjects
were brought to the research unit at 09.00 and an indwelling venous cannula
was inserted and heparinised. A baseline period of 30 minutes elapsed during
which three venous samples were taken for plasma prolactin and total
tryptophan measurement. Following this, 5 g of intravenous tryptophan
was infused over a 15-minute period, and further blood samples were taken for
the following 105 minutes. Every 30 minutes subjects rated themselves for
nausea (the main adverse effect of tryptophan infusion) on a 100 mm VAS.
During the second neuroendocrine test, the dose of intravenous tryptophan
given was reduced in proportion to the weight loss that had been achieved
during the diet. This was to ensure that individual subjects received the same
mg/kg dose of tryptophan in the post-dieting test as they had in the pre-diet
test. For administrative reasons it was not possible to blind the staff
carrying out the neuroendocrine testing to subject status.
Biochemical measures
Blood samples for measurement of total tryptophan and prolactin levels were
separated immediately after collection and stored at -30°C. All assays
were carried out blind to subject status. Plasma tryptophan levels were
determined by high-performance liquid chromatography with amperometric
detection. Plasma prolactin levels were measured by using a standard
immunoradiometric assay (reagents provided by Netria, London), with inter- and
intra-assay coefficients of variation of 4.8% and 2.4% respectively.
Statistical analysis
Baseline plasma tryptophan levels were measured by taking the mean of the
three baseline values before tryptophan infusion. Plasma tryptophan levels
after the infusion were measured as areas under the curve (AUCs) with
subtraction of baseline values extrapolated from Time 0. Nausea ratings were
measured as peak increase over baseline. Plasma prolactin levels after the
tryptophan infusion generally showed a brief increase followed by a decline.
For this reason peak prolactin change from baseline was taken as the key
response measure, although the AUC was also calculated. The peak values were
log transformed in order to approximate to a normal distribution. All the
above measures were analysed with a repeated-measures analysis of variance
(ANOVA) with diet (pre-diet v. post-diet) as the main
within-subjects factor and diagnosis (recovered from depression v.
control) as a between-subjects factor. The prolactin data were also plotted
against time and analysed by repeated-measures ANOVA with time as a further
within-subject factor. The daily VAS scales were measured as change from
baseline, and the resulting AUC values of controls and of those recovered from
depression over the three-week dieting period were analysed with an unpaired
t-test (two-tailed). Body weight and weight loss of the two subject
groups were also compared using unpaired t-tests.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
The recovered patients tended to be heavier than controls, but the two groups achieved similar absolute and percentage weight losses (Table 1). There was a main effect of dieting on mean baseline plasma tryptophan levels (F=11.8, P=0.002) but no main effect of diagnosis (F=0.93, P=0.35) or interaction of diet with diagnosis (F=0.26, P=0.61). Plasma tryptophan levels fell in both groups after dieting (recovered from depression 10.5 (s.e.=0.60) to 9.30 (s.e.=0.72) µg/ml; controls 11.0 (s.e.=0.47) to 10.1 (s.e.=0.43) µg/ml). Following tryptophan infusion there was a substantial increase in plasma tryptophan values. The ANOVA of the AUC of tryptophan post-infusion levels showed no main effect of diet (F=0.21, P=0.65) or diagnosis (F=0.71, P=0.41) and no significant diet-by-diagnosis interaction (F=0.004, P=0.95). An ANOVA of the peak nausea ratings showed no main effect of diet (F=0.78, P=0.39) or diagnosis (F=0.01, P=0.92), and no interaction between diet and diagnosis (F=1.0, P=0.33). Mean peak nausea ratings for controls during the pre-diet tryptophan infusion were 13.1 mm (s.e.=5.3), compared with 13.6 mm (s.e.=4.3) during the post-diet infusion. The corresponding values for the recovered depression group were 10.0 mm (s.e.=4.2) and 18.2 mm (s.e.=9.2) respectively.
The ANOVA of the peak prolactin values showed no main effect of diet (F=0.69, P=0.42) or diagnosis (F=2.9, P=0.1), but there was a significant interaction between diet and diagnosis (F=5.7, P=0.025). The prolactin responses to tryptophan of the two groups were very similar prior to dieting. After dieting, however, peak prolactin responses were increased in the controls but not in the recovered group (Fig. 1). The repeated-measures ANOVA including time as a within-subject factor showed no main effect of diet (F=0.74, P=0.40) or diagnosis (F=2.12, P=0.16). However, there were significant interactions between diet and diagnosis (F=7.46, P=0.012) and diet and time and diagnosis (F=1.95, P=0.040) (Fig. 2). The AUC analysis of prolactin response showed no significant effect of diet (F=3.30, P=0.082) or diet-by-diagnosis interaction (F=3.24, P=0.085). However, the mean (s.e.) of the AUC prolactin (over the 120 min infusion period) response was increased by dieting in controls (10495 group (3728) to 23 320 (8836) mIU x min/l, P=0.045) but not in the recovered group (6495 (2729) to 6553 (2207) mIU x min/l, P=0.99).
|
|
VAS ratings showed a significant reduction in happy scores over the three-week period of the diet in the recovered group relative to controls (-233.6±80.7) mm x day v. -26.8 (s.e.=30.8) mm x day, P=0.03). However, no significant differences were found between the two groups on ratings of sad (recovered group 45.5 (s.e.=25.4) mm x day v. controls 36.3 (s.e.=45.5) mm x day, P=0.86) or of irritable (-3.6 (s.e.=77.8) mm x day v. 37.2 (59.0) mm x day, P=0.68).
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Baseline 5-HT function in depressed women
The women who had recovered from depression did not differ from the
controls in their pre-dieting prolactin response to tryptophan challenge.
Although the presence of major depression is associated with blunted endocrine
responses to tryptophan (Power & Cowen,
1992), our previous work has indicated that these responses
normalise with clinical recovery
(Upadhyaya et al,
1991). This finding suggests that, if there is a trait abnormality
in brain 5-HT function in women vulnerable to major depression, it
may be apparent only at times when 5-HT function is under some form
of biochemical stress, such as decreased precursor availability.
It is worth noting, however, that women with acute major depression who have particularly marked weight loss can exhibit enhanced prolactin response to tryptophan (Cowen & Charig, 1987). This suggests that the difference between recovered women and controls found in the present study is one of degree, and that with sufficient restriction of tryptophan availability women recovered from depression might also show some up-regulation of 5-HT-mediated endocrine responses.
Mechanism of dieting-induced 5-HT up-regulation
The precise adaptive mechanisms that underlie the increased prolactin
response to tryptophan following dieting in healthy women are not established.
We have obtained evidence for a functional up-regulation of post-synaptic
5-HT2 receptors following dieting
(Cowen et al, 1996),
but the prolactin response to tryptophan appears under normal circumstances to
be mediated by post-synaptic 5-HT1A rather than
5-HT2 receptors (Smith
et al, 1991). Moreover, in animal experimental studies a
low tryptophan diet causes an up-regulation in 5-HT2 but
not 5-HT1A receptor function
(Franklin et al,
1999). It is possible that, during dieting, receptors other than
the 5-HT1A receptor play a role in tryptophan-induced
prolactin release. Another possibility is that compensatory changes in
presynaptic 5-HT mechanisms may be involved. For example, rats
maintained on a low tryptophan diet exhibit changes in 5-HT synthesis
consistent with activation of tryptophan hydroxylase
(Gil-Ad et al,
1976).
5-HT adaptation and psychological response to tryptophan deficit
The lack of neuroadaptive changes in 5-HT pathways during dieting
in women recovered from depression might lead them to experience a greater
decline in brain 5-HT function than controls. In this respect it is
interesting that dieting produced a greater lowering of mood, judged by
decreases in VAS ratings of happy, in the recovered group. However, the
changes seen were not of clinical significance.
While dieting produces a modest (about 10%) but presumably sustained decrease in plasma tryptophan levels, the procedure of tryptophan depletion with a tryptophan free amino acid mixture causes an immediate but much larger decrease (at least 80%) in plasma tryptophan levels (Smith et al, 1997a,b). In women with a history of major depression who are recovered and have ceased drug treatment, acute tryptophan depletion causes a striking but temporary relapse in clinical symptoms. By contrast, acute tryptophan depletion fails to cause clinical depressive symptoms in women without a history of depression (Smith et al, 1997a,b). It is conceivable that the vulnerability of women recovered from depression to experience acute affective symptoms following the administration of tryptophan-free amino acid mixtures may be a consequence of a failure in 5-HT adaptation in response to precursor deficit. Studies which are able to measure changes in brain 5-HT neurotransmission more directly will be needed to test this hypothesis adequately.
![]() |
Clinical Implications and Limitations |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
LIMITATIONS
![]() |
ACKNOWLEDGMENTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
American Psychiatric Association (1994) Diagnostic and Statistical Manual of Mental Disorders (4th edn) (DSMIV). Washington, DC: APA.
Cowen, P.J. & Charig, E. M. (1987) Neuroendocrine responses to intravenous tryptophan in major depression. Archives of General Psychiatry, 44, 958 -966.[Abstract]
Cowen, P.J., Clifford, E. M., Walsh, A. E. S., et al (1996) Moderate dieting causes 5-HT2C receptor supersensitivity. Psychological Medicine, 26, 1155 -1159.[Medline]
First, M. B., Spitzer, R. L., Gibbon, M., et al (1997) Structured Clinical Interview for DSMIV, Axis I Disorder Clinical Version. Washington, DC: American Psychiatric Association.
Franklin, M., Craven, R. D., Dowling, B., et al (1999) Effect of a long-term tryptophan diet on the prolactin responses to the 5-HT1A and 5-HT2C agonists, 8-OH-DPAT and mCPP, in the male rat. Journal of Psychopharmacology, 13, 58 -63.[Medline]
Gil-Ad, I. F., Zambotti, F., Carruba, M.O., et al (1976) Stimulatory role for serotonergic system on prolactin secretion. Proceedings of the Society for Experimental Biology and Medicine, 151, 512 -518.[Abstract]
Power, A. C. & Cowen, P. J. (1992) Neuroendocrine challenge tests: assessment of 5-HT function in anxiety and depression. Molecular Aspects of Medicine, 13, 205 -213.[Medline]
Smith, C. E., Ware, C. J. & Cowen, P. J. (1991) Pindolol decreases prolactin and growth hormone responses to intravenous L-tryptophan. Psychopharmacology, 103, 140 -142.[Medline]
Smith, K. A., Fairburn, C. G. & Cowen, P. J. (1997a) Relapse of depression after rapid depletion of tryptophan. Lancet, 349, 915 -919.[CrossRef][Medline]
Smith, K. A., Clifford, E. M., Hockney, R. A., et al (1997b) Effect of tryptophan depletion on mood in male and female volunteers: a pilot study. Human Psychopharmacology, 12, 111 -117.
Upadhyaya, A. J., Pennell, I., Cowen, P.J., et al (1991) Blunted growth hormone and prolactin responses to L-tryptophan in depression: a state-dependent abnormality. Journal of Affective Disorders, 21, 213 -218.[CrossRef][Medline]
Walsh, A. E. S., Oldman, A. D., Franklin, M., et al (1995) Dieting decreases plasma tryptophan and increases the prolactin response to d-fenfluramine in women but not men. Journal of Affective Disorders, 33, 89 -97.[CrossRef][Medline]
Received for publication April 19, 1999. Revision received September 16, 1999. Accepted for publication September 21, 1999.