The elucidation of sleep regulation is not an easy task. On one side, there is a multitude of solid yet disparate data, on the other side, the topic is tempting for engaging in wild speculation, particularly with respect to the functions of sleep. Models may exert a moderating influence by mediating between the two extremes. However, also they navigate between the risk of banality in reformulating the obvious, and the peril of fancy in losing touch with empirical reality.

The analysis of motivational systems underlying temporal organisation in animal behaviour has relied primarily on two conceptual functional frameworks: Homeostasis and biological clocks. Homeostasis is one of the most general and influential concepts in physiology. Walter Cannon introduced homeostasis as a universal regulatory principle which animals employ to maintain constancy of their ‘internal milieu’ in the face of challenges and perturbations from the external environment. Cannon spoke of “The Wisdom of the Body”, the collective of responses designed to defend the ideal internal state against those perturbations.

The adult human typically exhibits a monophasic sleep-wake cycle, i.e., remains awake and alert for approximately 16 hours and then sleeps for 8 hours. Recent experiments have provided new insights in the role of the endogenous circadian pacemaker in this consolidation of sleep and wakefulness.

Sleep deprivation studies had shown previously that sleepiness and alertness are co-determined by a process which keeps track of the history of sleep and wakefulness and the circadian pacemaker, which keeps track of time. During every day life and during sleep deprivation both processes change simultaneously and their relative contribution to alertness and sleep propensity cannot be assessed under these conditions.

Our current understanding of the physiological mechanisms underlying depressive disorders is not only based on behavioral, neuroendocrine and pharmacological studies in depressed humans, but also on experimental studies in a wide variety of animal models of depression. Ideally, the two approaches should operate in close interaction, each providing additional information to the other approach. However, in practice the animal model approach seems to be rather independent from the human studies. In a critical evaluation of the available animal models of depression, Willner concluded that none of the models fulfilled the criteria of a sufficient face, construct and predictive validity. Although this evaluation was made ten years ago, we feel that the situation has improved very little since that time. Most animal models fail to sufficiently mimic both the etiology and the symptomatology of human depressive disorders. With respect to the etiology, stress and major life events are generally considered to be an important factor in the development of depression. Most of the animal models however use stressors which bear little or no relationship to the biology of the species, i.e. to the situations an animal may meet in its everyday life in a natural habitat. Moreover, these models do not pay attention to the temporal dynamics of the disease. In humans, the disease is characterized by its gradual onset, which is often preceded by symptoms of anxiety. Moreover, a relatively large number of patients suffers from recurrent episodes of depression, which tend to occur with decreasing intervals and increasing duration and severity. If we want to improve our knowledge of the causal mechanisms of depression, animal models which allow an experimental analysis of the temporal dynamics of the disease are essential.

Cognitive and behavioural processes may constitute a risk for onset and persistence of depression. People who become depressed frequently show enduring negative cognitions which predispose them to depression. In addition, interpersonal processes are supposed to contribute to the etiology and maintenance of depression. Depression-prone persons are presumed to display deficient or problematic social behaviours that elicit negative reactions in others, finally resulting in withdrawal by family and friends.

About 60% of human communication is non-verbal. An ethological approach may therefore contribute to reveal behavioural and cognitive vulnerability factors for depression onset or persistence. Various studies support this presumption: High levels of patients' observed behaviour indicating involvement in the interaction between depressives and clinicians at admission are related with persistence of depression.

In 1975 van den Burg and van den Hoofdakker hypothesized that depressed patients might be ‘overaroused.’ This suggestion is consistent not only with their seminal observations on the antidepressant effects of total sleep deprivation in depression, but with the short, fragmented, and shallow sleep of depressed patients, lowered arousal thresholds, hyperactivity of the hypothalamus-pituitary adrenal (HPA) axis, and elevated core body temperature commonly found in some patients during the sleep period.

Many of the dramatic seasonal changes that occur in animals are triggered by changes in the length of night. Changes in the duration of melatonin secretion, which usually occurs exclusively at night and therefore lasts longer in winter than in summer, often act as a chemical transducer of the effects of seasonal changes in night-length on animals' behavior and physiology. Cells in sites that regulate seasonal behaviors, such as those associated with breeding, appear to possess melatonin receptors and interval timing mechanisms that enable them to detect melatonin and measure the duration of the nightly interval when it is present. These sites then trigger changes that are appropriate for the season that is indexed by the measured length of the interval.

Sleep electroencephalography in depressed patients reveals many signs of disrupted sleep, like long sleep latency, frequent awakenings, reduced amounts of time spent in the sleep stages 3 and 4, and early morning wakefulness. Upon total deprivation of sleep for one night, many patients experience an unexpected alleviation of their depression, which usually lasts until the subsequent sleep period. Attempts have been made to explain these changes of mood to result from induced changes in sleep physiological mechanisms. Such attempts can roughly be categorized in two classes. One class of hypotheses concerns proposed disturbances in circadian sleep control (i.e. the timing of sleep is inappropriately controlled), the other class concerns postulated disrupted homeostatic sleep control (i.e. the intensity of sleep is inappropriately controlled). For both types of theoretical approaches data have been published which are consistent with the hypotheses as well as data which are not.

Seasonal affective disorder (SAD) is characterised by recurrent episodes in autumn and winter of depression, hypersomnia, augmented appetite with carbohydrate craving, and weight gain, and can be successfully treated with bright light. Circadian rhythm hypotheses (summarized in) have stimulated research into the pathophysiology of SAD, postulating that:

1. 1.The illness is a consequence of delayed phase position,

2. 2.It is correlated with diminished circadian amplitude, or

3. 3.It results from changes in the nocturnal duration between dusk and dawn e.g. of low core body temperature or melatonin secretion. Light is considered to act directly on the circadian pacemaker (‘Process C’) and not on sleep dependent processes (‘Process S’). Thus successful treatment of SAD must act via mechanisms within known retinohypothalamic pathways. Conversely, emergence of SAD symptoms may reflect inappropriate neurobiological response to decreasing daylength.

Light therapy (LT) has become increasingly popular in various countries around the world in the last decade. For instance, according to a recent survey carried out in Germany in 1992, 13% (n = 56) of all German psychiatric hospitals (n = 422) used LT for different treatment indications and another 8% indicated their interest to do so. Among university facilities LT is even more popular, with a percentage of 57%. Although the most frequently used treatment indication for LT is seasonal affective disorder (SAD) or its subsyndromal form (S-SAD) it is apparent that other forms of depression e.g. non-seasonal forms, either acute or chronic are also a target for this new treatment modality. There is a number of studies supporting the use of LT for SAD (for review), however there are just a few studies for non-seasonal depression or for the other treatment indications (for review).

Since the description of SAD the prevalence of this disorder has been of particular interest. The more so, because early studies, indicated the correctness of the on theoretical grounds - expected interaction between latitude and prevalence rate of SAD. More eleborate prevalence studies, mainly in the USA indeed showed a positive correlation between prevalence rate and latitude. In Europe, only few studies have appeared. Although the results of most of these studies are preliminary and research-methods and criteria vary across studies, tentative conclusions are that the prevalence rates in Europe are considerably lower than in the USA, but that there is a similar trend in the direction of an increase of prevalence at higher latitudes.8"10 In the present study, the prevalence of SAD in the Netherlands is investigated.

The common treatment modalities for depression, i.e. pharmacotherapy and psychotherapy have the significant disadvantage of at least a three- to four-week time lag between initiation of treatment and amelioration of mood. Total sleep deprivation (TSD) in contrast, leads to an immediate antidepressant effect in 60% of the patients. However, it has gained only little clinical relevance as usually the improvement is only transient and almost regularly reversed by the next nights of sleep. A procedure preserving the antidepressive effect of sleep deprivation would therefore be of high clinical relevance.

Oince the discovery in the late 1950's of imipramine (Tofranil®), the first tricyclic antidepressant (TCA), and iproniazid, the first monoamine oxidase inhibitor (MAOI), many other TCAs and MAOIs, now being considered the classical antidepressants, have become available. Their effectiveness has been shown in numerous studies, although according to nowadays standards well designed studies (placebo-controlled, defined diagnostic groups, the use of standardized rating scales) with these compounds are relatively scarce.

From the early 1980's the so called modern antidepressants have been introduced: the selective serotonin re-uptake inhibitors (SSRIs), the reversible selective monoamine oxidase-A inhibitors (RIMAs) and a variety of other compounds. All these drugs have been registered after their effectiveness had been shown in well designed, placebo-controlled studies. In defining the efficacy of antidepressants, registration authorities consider two aspects important: statistical evidence and clinical relevance.

Although less undisputed in biological psychiatry than formerly, there are basically still two premisses which underly most hypotheses in biological depression research. These are deeply rooted in the medical way of thinking and have guided the construction of DSM-III. The first is that depressive symptoms and complaints cluster in one or more discrete syndromes, and the second is that these correspond to separate nosological entities, characterized by specific biological derangements and pathogenetic processes. Below a number of issues are raised which cast doubts on the validity of these premisses and on the progress made in pinpointing the derangements looked for.

Unequivocal results demonstrating a causal relationship between a disturbance in circadian rhythms and depression have not yet been reported (reviews). However, acute mood changes, such as the antidepressive effect of sleep deprivation, diurnal variations of mood and their interrelationship, are commonly put forward as evidence of the importance of circadian dysregulations in affective disorders. The purpose of the present study is to obtain more insight in the mechanisms underlying these mood changes. The results will be discussed in the context of a recently postulated non-chronobiological explanation.

Earlier studies have suggested that the relationship between diurnal variation of mood and the response to total sleep deprivation (TSD) is clear and unambiguous: improvement of mood during the day prior to TSD (a positive diurnal variation) is followed by a positive response (mood improvement) to TSD, while no improvement or deterioration of mood during the day prior to TSD (a negative diurnal variation) may result in no, or even a negative, TSD response (for references see Van den Hoofdakker). However, these conclusions were based on the results from cross-sectional studies, comparing single TSD effects across individuals. Comparison of sleep deprivation effects within individuals, however, revealed that the course of mood during the day prior to TSD is irrelevant for the TSD response. Accordingly, a favourable response to TSD appeared to be related to the patient's propensity to show diurnal mood variations per se, irrespective of their direction.

Since Rosenthal et al published their first study about the effects of light treatment on mood in patients suffering from winter depression, many research centres have become interested in the study and treatment of seasonal affective disorders (SAD), one of them being the department of Biological Psychiatry in Groningen. In 1987, we started the first experiments under the inspiring direction of Prof. Rudi van den Hoofdakker and hundreds of patients have participated in various investigations ever since.

According to interpersonal theories on depression, the type of interaction between depression-prone subjects and their social environment plays a causal role in the development and course of depression (e.g. Coyne et al.). So far, interpersonal theories have been tackled mostly by psychometrical methods. However, non-verbal behaviour plays an important role in human social interactions. It is assumed that 60-65% of human communication is non-verbal. Ethological observations have shown that non-verbal interpersonal behaviour of depressed subjects, as assessed prior to treatment, is related to treatment-response or subsequent course of depression. These results are in line with an interpersonal approach of depression.

In manic-depressive patients treated with lithium salts the transport of choline over the erythrocyte membrane is strongly inhibited, resulting in dramatically increased erythrocyte choline levels (for review see ref. 1). Whether or not there is a relationship between this effect and treatment response is not clear. Data on this issue are scarce possibly because the measurement of treatment response in lithium prophylaxis is very difficult and time consuming. Also the effect on erythrocyte choline is specific for lithium and not for manic-depressive illness. We will address here the question of a possible relation between erythrocyte choline and clinical effects of lithium.