Main

The rapid tryptophan depletion (RTD) paradigm (Young et al. 1985; Delgado et al. 1990) allows researchers to study the consequences of 5-HT system deactivation. In humans and animals, RTD has been shown to cause significant (but transient) reductions in the levels of tryptophan (TRP) in the blood, and to reduce the amount of 5-HT in the brain (for review, see Moore et al. 2000). RTD typically consists of a day's worth of a low-TRP diet, followed by ingestion of a TRP-free amino acid load, and RTD, but not the control TRP-containing condition, has been associated with myriad clinical and physiological effects. Relapse of depressive mood following RTD in recently remitted depressed patients has been reported (Delgado et al. 1990). In healthy subjects, mood responses as well as other behavioral measures have been somewhat inconsistent (Young et al. 1985; Barr et al. 1997).

The sleep abnormalities often associated with depression are consistent with the notion of a pathophysiologic deficiency in serotonergic function in depression (Maes and Meltzer 1995). There is strong preclinical and clinical evidence that serotonin plays a role in suppressing REM sleep (McGinty and Harper 1976; Honda and Semba 1994; Luebke et al. 1992; Portas et al. 1996).

RTD apparently disinhibits REM sleep measures in serotonin reuptake inhibitor (SRI)-treated patients in full remission (Moore et al. 1998) and in healthy male control subjects (Bhatti et al. 1998), using two different strengths of RTD (100% and 25%). In SRI-treated patients, both RTD doses also reversed the reduced total sleep time and prolonged sleep latency observed at baseline. In normal males, both RTD doses significantly decreased REM latency compared with baseline, as well as significantly reduced total and free plasma TRP concentrations.

Agonists at the 5-HT1A receptor, such as ipsapirone, have been shown to inhibit REM sleep both in healthy subjects and in depressed patients (Gillin et al. 1994; Gillin et al. 1996; Seifritz et al. 1998), and this effect presumably reflects predominantly post-synaptic receptor activity. In the present study, we examined the competing effects of 5-HT depletion and 5-HT stimulation on sleep. We predicted REM sleep measures with RTD plus placebo would be significantly enhanced compared with baseline as shown in previous studies with RTD alone, and that the combination of RTD and ipsapirone would produce REM measures with values intermediate between those obtained with either 5-HT manipulation alone.

METHODS

Subjects

Ten healthy male subjects between the ages of 20 to 39 years (28.7 ± 1.9, mean ± SD) were recruited from the general public in the San Diego region through the UCSD Mental Health Clinical Research Center (MHCRC). Eight Caucasian-Americans and two African-Americans completed the study after giving written informed consent (approved by UCSD Human Subjects Committee). Routine MHCRC screening methods, and inclusion and exclusion criteria have been described elsewhere (Bhatti et al. 1998). Patients self-reported no personal or family history of psychiatric disorders.

Procedures

Nocturnal sleep studies were conducted in the sleep laboratory in the San Diego Veteran's Affairs Medical Center (SDVAMC). Bedtime was set as close as possible to the subject's customary bedtime on all nights; care was taken to ensure night-to-night consistency in this regard. The adaptation night screened for sleep disorders. All pre-bedtime tasks and questionnaires were completed at approximately the same time relative to bedtime each night. Pre-challenge baseline recordings were run the night directly preceding the challenge night. On RTD challenge nights, a TRP-free RTD was ingested at about 5:30 P.M.; then, in double-blind fashion, 10 mg ipsapirone or placebo was administered 20 min prior to bedtime. These schedules were intended to maximize the effects of each of these treatments on sleep, based upon our experience. Ipsapirone has a plasma half-life of 1–2 h and its sleep effects appear to be greatest when taken at bedtime. The maximal tryptophan-depleting effects of the RTD (as measured by plasma TRP levels, which are roughly coincident with the sleep effects we have seen) occur within 5–10 h after ingestion of the drink. Challenge nights were separated by one week.

Polysomnographic Recordings

Sleep recordings were conducted and scored according to standard technique, and as described and defined previously (Moore et al. 1998). Sleep variables were Total Sleep Time, Sleep Efficiency, Sleep Latency, Wakefulness after sleep onset, REM minutes, REM percent, REM latency, REM latency-corrected, REM density.

Rapid Tryptophan Depletion Paradigm

The RTD protocol was conducted virtually identically to previous studies with one exception, that is, the composition of the RTD was altered slightly, lessened in strength from 100% (102.5 gm) to 85% and adjusted to the individual's body weight, in a fixed ratio (85%/70kg). Comparing pre-RTD baseline to 5-h post-RTD tryptophan levels, this composition decreased plasma TRP similarly to the 100% strength RTD. Low-TRP diet, free and total plasma TRP analysis were conducted as in previous studies. Mood scales (Visual Analog Scale, [VAS]; Profile of Mood States [POMS]; a symptom checklist, and perceived sleep quality questionnaires were filled out on multiple occasions.

Data Analysis

Within-subjects repeated measures analyses of variance (ANOVA) were performed on one or more variables (mood, plasma tryptophan, polysomnographic sleep measures) compared across four conditions (RTD plus ipsapirone, RTD plus placebo, and both pre-challenge baseline nights). All variables were initially examined for normality of distribution using Kolmogorov statistics. Data are expressed as means ± s.e.m. Level of significance was set at α < 0.05. For sleep measures and mood data, n = 10; for plasma data, n = 9.

RESULTS

Plasma Tryptophan Measures

The combination of treatments (RTD plus ipsapirone, RTD plus placebo) was well tolerated by all subjects. Total and free plasma TRP concentrations (μmol/l) were significantly reduced 5 h post-ingestion after both RTD trials (p < .001 repeated measures ANOVA with pairwise pre versus post-RTD comparison). Free plasma TRP decreased by 78% and 80%, respectively, in the two challenges. Total TRP decreased by 92% after both challenges.

Polysomnographic Sleep Measures

As shown in Table 1, ANOVAs revealed significant differences across the four nights of sleep on measures of REM-latency (F(1,9) = 22.84, p < .001) and REM-latency-corrected (F(1,9) = 11.03, p < .01). There were no significant differences seen in REM min, REM%, nor REM density, or of total sleep time and sleep latency. Post hoc pairwise comparisons showed that the significant differences were due to sleep effects only when the RTD plus ipsapirone condition was compared with its preceding baseline recording night. RTD plus placebo did not differ significantly from baseline on any sleep measure.

Table 1 Summary of Sleep Statistics for the Pre-Challenge Baseline and RTD Challenge Nights in the Placebo and Ipsapirone Conditions
Full size table

Mood Measures

There were no significant short-term effects on mood of neither RTD nor ipsapirone (data not shown). There were also no significant findings from any item on the subjective sleep questionnaire (data not shown).

DISCUSSION

Free and total plasma TRP levels were significantly lowered 5 h post-RTD compared with baseline levels, to an extent comparable to other published studies, including our own. However, the REM sleep findings of our earlier RTD study in healthy subjects were not replicated, and were partially consistent with other reports (Voderholzer et al. 1998). We had predicted that: 1) after RTD plus placebo, REM measures would be disinhibited compared with baseline conditions, and 2) after RTD plus ipsapirone, REM measures would approximate baseline values. Neither prediction was borne out. RTD plus placebo in this study did not significantly shorten REM latency, and REM min, REM%, and REM density were not significantly increased. All significant changes from baseline of sleep in this study were essentially restricted to the REM-suppressing effects of 10 mg ipsapirone (prolonged REM-latency and REM-latency-corrected on RTD plus ipsapirone night).

It is unclear why these data are inconsistent with previous reports. Here, after an 85% RTD, REM latency was 69 ± 5 (mean ± s.e.m.), whereas in the Bhatti et al. (1998) study after the 100% RTD, mean REM-latencies were 40 ± 11, and 48 ± 12 after the 25% RTD. Baseline REM latencies did not differ significantly between the studies, nor did subject age. Examining REM-latency changes from baseline across both studies, in our opinion, each of the 21 subjects can be classified as either a responder or a non-responder to RTD. In the previous RTD-only study, after the 100% RTD, five of 11 subjects demonstrated a sleep onset REM period (SOREMP, REM latency < 20 min). Two of the 11 had a SOREMP after the 25% RTD. In the RTD plus ipsapirone study reported here, there were no SOREMPs. Participants in both studies had been screened in the same fashion, including for family history of affective illness.

Pronounced individual and gender-related differences in rates of 5-HT synthesis are suggested by PET data (Nishizawa et al. 1997). We wonder if individual regulatory differences in the serotonin system itself (possibly genetic) may determine whether or not a subject “responds” to RTD. If so, such differences in vulnerability to RTD may have implications for the plethora of conflicting findings in healthy subjects using a variety of behavioral measures, for example, mood (e.g., Young et al. 1985; Barr et al. 1997), anxiety and panic (Goddard et al. 1995; Klaassen et al. 1998), aggression (Cleare and Bond 1995; Bjork et al. 1999), food choices (Oldman et al. 1994; Young et al. 1988), as well as REM sleep. The variability in the RTD-induced REM responses of healthy subjects contrasts dramatically with the consistent REM-disinhibiting effects of RTD in depressed patients treated with SRIs (Moore et al. 1998) or monoamine oxidase inhibitors (Landolt et al. 2000).

In summary, this study failed to demonstrate that the REM-suppressing effects of ipsapirone, a 5-HT1A agonist, could counteract the REM sleep-disinhibiting effects of tryptophan depletion. However, the study failed to do so because neither the REM sleep findings nor the subjective mood findings of our previous RTD report in healthy men were replicated, despite significant and comparable plasma depleting effects of RTD. We believe there are important individual differences in vulnerability to RTD. The medical or psychiatric factors that may be associated with this vulnerability are unknown.