Journal of

@1994 British Association for Psychopharmacology

Psychopharmacology 8(1) (1994) 8-13

Effect of acute tryptophan depletion on mood and appetite in healthy female volunteers A. D. Oldman, A. E. S. Walsh, P. University Department of Psychiatry

Salkovskis,

and MRC Unit

of Clinical

D. A. Laver and P. J. Cowen

Pharmacology, Littlemore Hospital, Oxford OX4 4XN, UK

healthy female subjects received the following three drinks in a double-blind, semi-balanced, crossover design: (a) 50 g of amino acids without L-tryptophan (LTP); (b) 50 g of amino acid with LTP (balanced); (c) plain water. Compared to both the balanced amino acid mixture and plain water, the LTP drink significantly lowered plasma total and free tryptophan at 4.5 h. However, compared to the two control conditions, there was no effect of the LTP drink on subjective ratings of mood or hunger. Similarly, the LTP drink did not alter significantly either total calorie intake or the macronutrient content of a test meal 5 h after drink ingestion. Twelve

Key words: L-tryptophan; appetite; hunger; mood; calories; macronutrient

Introduction

Conversely, lowering the availability of LTP to the brain decreases 5-HT synthesis (Biggio et al., 1974; Moja et al., 1989). One way of achieving this effect is to administer a mixture of amino acids from which LTP has been omitted. This manoeuvre decreases the supply of LTP to the brain through two mechanisms: (a) by lowering plasma tryptophan concentration; and (b) by restricting brain entry of LTP through competition of the other amino acids for transport across the blood-brain

5-hydroxytryptamine (5-HT) pathways influence several important behaviours in animals and humans. There is much evidence, for example, that brain 5-HT neurones are involved in the regulation of mood, with numerous studies indicating that patients with depressive disorders have impaired brain 5-HT function; in addition, drugs that are helpful in treating depression often facilitate 5-HT neurotransmission (for reviews, see Cowen and Anderson, 1991; Deakin, 1991). Brain

barrier (Young et al., 1985). This method of lowering brain applied to both animals and humans.

There is also strong evidence that 5-HT neurones are involved in the control of appetite, perhaps principally by influencing processes of satiety (Blundell and Hill, 1987). Drugs that increase brain 5-HT function decrease food intake in both animals and humans (Kennet and Curzon, 1991; Godall et al., 1992) while the converse has been demonstrated for 5-HT receptor antagonists, though the latter effect depends to a large extent on the experimental conditions employed (Silverstone and Schuyler, 1975; Stallone and Nicolaidis, 1989). It has also been suggested that changes in brain 5-HT function may have consequences for macronutrition selection in that increasing brain 5-HT neurotransmission may tend selectively to reduce carbohydrate intake (Wurtman and Wurtman, 1979). This effect, however, has not been found in all studies. The synthesis of 5-HT depends on the conversion of the amino acid precursor of 5-HT, L-tryptophan (LTP), to 5-hydroxytryptophan. This process is catalysed by tryptophan hydroxylase, an enzyme confined to 5-HT neurones (Lovenberg, Jequier and Sjoerdsma, 1968). Under normal physiological conditions, tryptophan hydroxylase is unsaturated with LTP; accordingly altering the availability of LTP to the brain changes brain 5-HT synthesis (Ashcroft, Eccleston and Crawford, 1965; Fernstrom and Wurtman, 1971). Administration of LTP to rats, for example, increases brain 5-HT synthesis and release (Gessa et al., 1974; Morris et al., 1987; Curzon, 1981; Sharp, Bramwell and Grahame-Smith, 1992).

5-HT function has been In rats, Biggio et al. (1974)

demonstrated that an LTP-free mixture of essential amino acids decreased brain 5-HT turnover. In addition, we have shown recently that a similar amino acid treatment decreased the 5-HT release evoked by electrical stimulation of the dorsal raphe nucleus in vivo (Gartside, Cowen and Sharp, 1992). In humans, Young et al. (1985) reported that a 100 g LTP-free amino acid mixture markedly decreased plasma total and free LTP. This effect was associated with a small but significant lowering of mood in healthy male volunteers. In a subsequent study, Young et al. (1988) found that LTP depletion produced by the amino acid mixture resulted in a modest but significant decrease in protein intake in male subjects. The purpose of the present study was to examine the effect of acute LTP depletion on mood, appetite and food intake in healthy female volunteers. Previous studies by ourselves and others indicate that females may be more vulnerable than males to the effects of LTP depletion on brain 5-HT function (Anderson et al., 1990). We therefore predicted that LTP depletion would significantly lower mood and increase food intake in our subjects.

Methods Volunteers Twelve female volunteers between the ages of 19 and 27, and of normal weight for height (BMI range 17.4-29.0) were recruited 8

9

through advertisement. All volunteers were in good health, had no history of psychiatric disorder, no dietary restriction (two subjects were non-strict vegetarian) and no history of eating difficulties. The mean Beck Depression Inventory (BDI) score was 6.42 (range 0-17). All women had a regular menstrual cycle.

Experimental design Subjects were required to attend the research unit for three complete mornings, each a week apart such that no session fell within the pre-menstrual period of their cycle (i.e. each session fell within days 3-12, with these days being confirmed before the test sessions). The three conditions consisted of the amino acid drink without tryptophan (T - ), the balanced control drink (B) and a control session where subjects ingested a drink of water with no amino acids (placebo). Because of the consistency and taste of the amino acid mixture, it was necessary to use a semi-balanced design with placebo always administered first in order to ensure that the remaining two conditions remained double-blind. Order effects were therefore gauged on the basis of T - and B drinks, with the placebo condition used as an anchor for the group comparison. Measures

and

was designed to be nutritionally balanced (B). The experimental drink was of the same composition with the exclusion of the 1.15 g of L-tryptophan (T - ). Amino acid amounts were as follows: L-alanine, 2.75 g; L-arginine, 2.45 g; L-cysteine, 1.35 g; glycine, 1.6 g; L-histidine, 1.6 g; L-isoleucine, 4.0 g; L-leucine, 6.75 g; L-lysine monohydrochloride, 5.5 g; L-methionine, 1.5 g; L-phenylalanine, 2.85 g; L-proline, 6.1 g; L-serine, 3.45 g; L-threonine, 3.45 g; L-tyrosine, 3.45 g; Lvaline, 4.45 g; L-tryptophan, 1.15 g. These are as described by Young et al. (1985), based on the ratios of amino acids found

in human milk. Amino acids were mixed with 300 ml refrigerated water, and the resulting suspension was flavoured with 12.5 mg saccharin and blackcurrant. Procedure Subjects arrived at 8.30 a.m. after an overnight fast, and spent the test session in a testing room. At 8.40, baseline VASs, POMS and food preferences were taken, followed by 10 ml fasting plasma. At 9.00 the drink was administered. Between 9.30 a.m. and 1.30 p.m., VASs were administered once per hour. At 1.30 rating scales were given followed by the second 10 ml blood

sample ( + 4.5 h).

During the test sessions, mood was measured using the Profile of Mood States Questionnaire (POMS), which had been amended to examine current mood, and 100 mm visual analogue scales (VAS). The parameters were despondent/sad, anxious, irritable and tense. Appetite was measured using a free-choice, buffet-style test meal accompanied by VAS and food preference questionnaires before and after the meal. All food items were weighed to the nearest 0.1g before and after the meal. The macronutrient and energy value of each meal was then calculated from these data using specifically designed software. Food preference questionnaires were those of Blundell and Hill (1987), and consisted of one free-choice and one forced-choice checklist. The latter examined the preference for carbohydrate over

The test meal

was

presented at 1.40, and volunteers were left Rating scales were completed

to determine the end of the meal.

the meal was over, and then volunteers were sent home with a booklet of rating scales to complete before and after their evening meal and lunch next day. once

Biochemical

measures

Blood samples (10 ml) were taken into heparin tubes prior to the drink (’0’ min) and 4.5 h after the drink. Plasma samples were separated by centrifugation. Plasma total and free tryptophan were determined by fluorometric detection according to the method of Bloxam and Warren (1974) and Bloxam, Hutson and Curzon (1977).

protein food items. The way in which a typical meal was made up was in Table 1.

as

detailed

Data analysis All data were analysed by analysis of variance using the BMDP

Dietary manipulation

p2V statistical package or the ANOVA suite on Supastat (SPA). This analysis was group (order 1 versus order 2). Greenhouse

The two amino acid mixtures were made up as drinks. The control drink contained 16 amino acids (52 g of amino acids in total),

Geisser probability of the analysis.

Table 1

Composition of

test meal

’Choice determined by food preference ratmgs completed by volunteer prior

to

study.

was

used where repeat

measures were

part

10

Results Biochemical measures Plasma concentrations of total and free tryptophan (TRP) were analysed for all three conditions. For levels of total plasma TRP, the ANOVA yielded a significant main effect of treatment (F2,22=33.77, p