Erich Seifritz

Clinical and Physiological Consequences of Rapid Tryptophan Depletion

We review here the rapid tryptophan depletion (RTD) methodology and its controversial association with depressive relapse. RTD has been used over the past decade to deplete serotonin (5-hydroxy-tryptamine, or 5-HT) in humans and to probe the role of the central serotonin system in a variety of psychiatric conditions. Its current popularity was stimulated by reports that RTD reversed the antidepressant effects of selective serotonin reuptake inhibitors (SSRIs) and monoamine oxidase inhibitors (MAOIs) in remitted patients with a history of depression but not in patients treated with antidepressants which promote catecholaminergic rather than serotonergic neurotransmission (such as tricyclic antidepressants or buproprion). However, RTD has inconsistent effects in terms of full clinical relapse in depressed patients. Pooling the data from all published reports, patients who are either unmedicated and/or fully remitted are much less likely to experience relapse (7 of 61, or ∼9%) than patients who are recently medicated and partially remitted (63 of 133, or ∼47%; although, the numbers here may reflect patient overlap between reports). Recently remitted patients who have been treated with non-pharmacological therapies such as total sleep deprivation, electroconvulsive therapy, or bright light therapy also do not commonly show full clinical relapse with RTD. We briefly review RTD effects in other psychiatric disorders, many of which are treated with SSRIs. There is accumulating evidence to suggest that RTD affects central serotonergic neurotransmission. Nevertheless, many questions remain about the ability of RTD to reverse the beneficial effects of SSRIs or MAOIs, or to induce symptoms in unmedicated symptomatic or asymptomatic patients.

Sleep deprivation and bright light as potential augmenters of antidepressant drug treatment—Neurobiological and psychometric assessment of course

The present study was designed to investigate the clinical efficacy of trimipramine with adjunct sleep deprivation (SD) or bright light (BL) and to evaluate psychometric and neurobiological variables that might be of predictive value for treatment response. We used (1) the combined dexamethasone-corticotropin releasing hormone test (DEX-CRH test) to characterize alterations of the hypothalamic-pituitary-adrenal (HPA) system; (2) polysomnography to evaluate sleep disturbances; and (3) a standardized test battery to assess cognitive psychomotor functions after study initiation and after 5 weeks of treatment. The overall response rate (> or = 50% decrease in score on Hamilton Rating Scale for Depression [HRS]) was 55% (N = 42). The response rate in the group with trimipramine monotherapy (N = 14) was 79%, whereas in the groups with adjunct SD (N = 14) and BL (N = 14), respectively, it was only 43%. All three groups showed significant improvement at the end of the third week of treatment. Neither of the adjunct treatments hastened the onset of antidepressant action as measured by HRS. A significantly higher proportion of nonresponders than responders (p < .05) had HPA dysregulation, disturbed rapid eye movement (REM) sleep (REM latency, REM% first third of night) and decreased non-REM sleep (% stage 2). The non-responders showed significantly more corticotropin (ACTH) secretion after CRH stimulation in the DEX-CRH test than the responders and a less rapid normalization of the neuroendocrine dysregulation (cortisol secretion) (p < .01). In addition, REM latency was significantly shorter in the BL group than in the monotherapy group and estimated duration of illness significantly longer in the SD group than in the monotherapy group. REM latency, percentage of REM sleep during the first third of the total sleep period, percentage of non-REM sleep stage 2 and ACTH release after a DEX-CRH challenge predicted response across all three treatment groups. The neurobiological symptoms were unevenly distributed, among the three groups, thus creating heterogeneity in these measures. This heterogeneity may have contributed to the different treatment response rates as defined by psychopathology (HRS). In contrast, the neuropsychological tests and some of the sleep-EEG investigations revealed different response patterns for different groups: The onset of improvement in simple cognitive functions and in sleep continuity was earlier in the adjunct treatment groups. This study underlines the need for a multidimensional approach including use of neurobiological and neuropsychological measures to identify the therapeutic profiles of different treatment strategies and predictors of outcome.

Effects of a Tryptophan-Free Amino Acid Drink Challenge on Normal Human Sleep Electroencephalogram and Mood

BACKGROUND Serotonin has been implicated in the regulation of sleep and mood. In animals a tryptophan-free amino acid drink (TFD) challenge has been found to reduce brain serotonin. We hypothesized this TFD would produce alterations in electroencephalographic (EEG) sleep commonly associated with depression, i.e. an enhancement of rapid eye movement (REM) sleep, and adversely affect mood ratings in humans. METHODS We investigated the effects of a TFD challenge in 11 healthy male subjects on EEG sleep and mood (assessed by Profile of Mood States). All subjects received on separate occasions an experimental drink containing approximately 100 g of an amino acid mixture (100% TFD) and a control drink containing one fourth strength (25% TFD) of the experimental drink 5 hours prior to sleep (6:00 PM). RESULTS Both drinks significantly decreased plasma tryptophan levels 5 hours postchallenge (11:00 PM). Both drinks significantly decreased REM latency, and the 25% TFD also increased REM time and REM% compared to baseline. No significant changes were found in subjective ratings of depression; however, subjects reported confusion and tension and a decrease in elation, vigor, and friendliness compared with baseline. CONCLUSIONS These TFD findings further support the involvement of serotonin deficiency in EEG sleep findings commonly seen in depression.

Microsleep during partial sleep deprivation in depression

BACKGROUND Sleep deprivation (SD) exerts a beneficial effect on mood and sleep in about 60% of depressed patients usually followed by a relapse into depression after the recovery night. Short phases of sleepiness, especially naps in the early morning, may be responsible for this phenomenon. METHODS To evaluate the effect of short, even ultrashort phases of sleep-microsleep (MS) during partial sleep deprivation (PSD) on mood, cognitive psychomotor performance (CPP), and sleep, an electroencephalograph (EEG) was continuously recorded over 60 hours in 12 patients with major depression. Subjective mood was assessed by a visual analogue scale and CPP by a letter cancellation test. RESULTS The results illustrate that in depressed patients during PSD the amount of MS is increased, predominantly in the early morning, which was subjectively unrecognized and not observed by nursing staff. Patients with a low cumulative amount of MS during PSD improved significantly in mood, CPP, and sleep pattern compared to the patients with a high amount of MS who showed only slight changes. CONCLUSION Therefore, accumulated MS may influence the SD-induced positive effects in depressed patients.

The 5-HT1A agonist ipsapirone enhances EEG slow wave activity in human sleep and produces a power spectrum similar to 5-HT2 blockade ☆

The REM sleep-suppressing effect of postsynaptic 5-HT1A stimulation has been well established. Here we investigate the effects of the 5-HT1A agonist ipsapirone (10 and 20 mg) on sleep EEG power spectra during non-REM sleep in nine healthy humans. At the lower dose, slow wave activity (SWA; EEG power in the delta (1-4.5 Hz) range) was significantly enhanced. At the higher dose, where side-effects occurred, the enhancement in SWA was not significant. The spectral profile was characterized by a bimodal increase of power in the lower delta and in the theta (5-8 Hz) frequencies, and by troughs at 4 Hz and at 11 Hz, a pattern compellingly similar to that reported for a 5-HT2 antagonist (seganserin). We propose that the spectral data following the lower ipsapirone dose reflect a net decrease of neuronal activity at 5-HT2 receptors, mediated through stimulation of somatodendritic autoreceptors in the raphe nuclei (presynaptic) and/or through stimulation of postsynaptic 5-HT1A receptors colocalized with 5-HT2 receptors. The spectral non-REM sleep EEG profile might be used to investigate central 5-HT function in humans.

Human sleep EEG following the 5-HT1A antagonist pindolol: possible disinhibition of raphe neuron activity

The sleep electroencephalogram (EEG) was used to assay central effects of pindolol (10 and 30 mg p.o.), a mixed beta(1/2)-adrenoceptor/5-hydroxytryptamine (5-HT)(1A/1B) receptor blocker, in humans. Compared to placebo, pindolol produced a dose-related suppression of rapid-eye-movement (REM) sleep, including a prolongation of REM latency, and a decrease of REM time and REM density. At the higher dose, it also reduced EEG spectral power during non-REM sleep in portions of the delta, theta, and alpha frequencies (1.125-5.125 Hz, 7.125-9.625 Hz). By contrast, betaxolol (20 mg p.o.), a selective beta1-antagonist devoid of serotonergic affinity, affected neither REM sleep nor EEG power. REM sleep is, in part, under the inhibitory control of serotonergic neurons projecting from the dorsal raphe nucleus to pontine cholinergic/cholinoceptive cells. The EEG power spectrum induced by pindolol tended to be opposite to what has previously been reported for ipsapirone, a 5-HT1A agonist. Therefore, the present data, tentatively, are consistent with the contention that pindolol inhibits, possibly selectively, somatodendritic 5-HT1A autoreceptors in humans and may antagonize self-inhibition of midbrain raphe nuclei 5-HT neurons.

Sleep electroencephalographic response to muscarinic and serotonin1A receptor probes in patients with major depression and in normal controls

BACKGROUND To test the hypothesis that depression is associated with an increased ratio of cholinergic to serotonergic neurotransmission, we compared the effects of pilocarpine, a muscarinic agonist, and ipsapirone, a serotonin (5-HT)1A agonist, on electroencephalographic (EEG) sleep in depressed and healthy subjects. We hypothesized, adopting the reciprocal interaction model, that the effects on REM sleep of these probes within the same individuals are negatively correlated and unmask neurobiological changes in depression. METHODS Polysomnographic recordings were obtained in 12 unmedicated patients with a current major depression and 12 normal controls. They received placebo, pilocarpine 25 mg, or ipsapirone 10 mg (orally, 15 min before bedtime, after premedication with the peripheral anticholinergic probanthine 30 mg, double blind, counterbalanced) on three occasions. RESULTS Pilocarpine shortened and ipsapirone prolonged REM latency equally in both groups. These effects were not correlated. Pilocarpine decreased slow-wave sleep and EEG delta power during the first nonREM episode more in controls than in patients, and enhanced EEG sigma power equally in both groups. Ipsapirone had no significant effects on EEG power. CONCLUSION These data do not support the postulate of muscarinic receptor up-regulation and 5-HT1A receptor down-regulation in depression. The significance of blunted delta power suppression in patients following pilocarpine warrants further investigations.

Effect of sleep deprivation on neuroendocrine response to a serotonergic probe in healthy male subjects

Neuroendocrine responses to stimulation with a selective serotonin reuptake inhibitor (citalopram) were measured to investigate the effects of all-night sleep deprivation on serotonergic function in healthy male subjects (n = 7). We studied citalopram-stimulated prolactin and cortisol plasma concentrations in a placebo-controlled cross-over protocol following sleep and sleep deprivation. Citalopram infusion (20 mg i.v. at 14:20-14:50 h) after a night of undisturbed sleep prompted robust increases in both plasma prolactin and cortisol concentrations. Following a night of sleep deprivation, by contrast, the citalopram-induced prolactin response was blunted, but the cortisol response was not significantly altered. This differential response pattern relates to the distinct pathways through which serotonin may activate the corticotrophic and the lactotrophic systems. While an unchanged cortisol response does not indicate (but also does not refute the possibility of) an altered serotonergic responsivity following sleep deprivation, the suppressed prolactin response could reflect a downregulation of 5-HT1A or 2 receptors. An alternative, not mutually exclusive, explanation points to the possibility that sleep deprivation activates the tubuloinfundibular dopaminergic system, the final inhibitory pathway of prolactin regulation.

Rapid Tryptophan Depletion Plus a Serotonin 1A Agonist: Competing Effects on Sleep in Healthy Men

In the present study, we hypothesized that the REM-suppressing effects of 5-HT1A receptor stimulation would counteract the REM-disinhibiting effects of rapid tryptophan depletion (RTD), and vice versa. We administered RTD plus ipsapirone (10 mg, p.o.) or RTD plus placebo to 10 healthy men. In contrast to our previous findings but partially consistent with other studies, RTD in combination with placebo did not produce a significant enhancement of any REM sleep measure. The combination of RTD and ipsapirone produced a significant suppression of REM sleep that was remarkably similar to the effect of ipsapirone alone. These data appear to deepen the mystery of variable and inconsistent RTD-induced responses in healthy subjects. In the case of REM sleep measures, this differs markedly from the consistent RTD-induced REM-disinhibiting effect seen in medicated depressed patients.

Effects of corticotropin-releasing factor on circadian locomotor rhythm in the golden hamster

Stress produces a reduction in the amplitude of some circadian rhythms. The neurochemical mechanisms underlying stress-induced changes in circadian rhythms are not known. To investigate a possible role of corticotropin-releasing factor (CRF) in this phenomenon, three related experiments were carried out: activity rhythms of male golden hamsters (10/14 hours light/dark entrained, lights on at 0800 h) were measured 1) following the intracerebroventricular administration of CRF (0.5, 1.0, 2.0, or 4.0 microg) at two different times of day, 2) following social stress (30-min resident-intruder confrontation), 3) and following the administration of the CRF-antagonist alpha-helical CRF9-41 (2.0 microg) prior to a 15-min resident-intruder confrontation. CRF produced a significant, dose-related decrease in circadian rhythm amplitude following administration in the morning hours, but not in the afternoon. CRF also induced transient increases in activity post injection concomitant with an activation of the hypothalamic-pituitary-adrenocortical (HPA) system. Stress similarly reduced the amplitude of activity patterns and stimulated the HPA system. The stress-induced depression of circadian rhythm amplitude was significantly attenuated following alpha-helical CRF9-41. These data suggest a role for CRF in the stress-related modulation of circadian locomotor rhythm amplitude.