The effect of neuroendocrine secretion on brain morphology and EEG sleep in patients with eating disorders

Harmful effects of functional hypercortisolism: a working hypothesis

Functional hypercortisolism (FH) is caused by conditions able to chronically activate hypothalamic-pituitary-adrenal axis and usually occurs in cases of major depression, anorexia nervosa, bulimia nervosa, alcoholism, diabetes mellitus, simple obesity, polycystic ovary syndrome, obstructive sleep apnea syndrome, panic disorder, generalized anxiety disorder, shift work, and end-stage renal disease. Most of these states belong to pseudo-Cushing disease, a condition which is difficult to distinguish from Cushing's syndrome and characterized not only by biochemical findings but also by objective ones that can be attributed to hypercortisolism (e.g., striae rubrae, central obesity, skin atrophy, easy bruising, etc.). This hormonal imbalance, although reversible and generally mild, could mediate some systemic complications, mainly but not only of a metabolic/cardiovascular nature, which are present in these states and are largely the same as those present in Cushing's syndrome. In this review we aim to discuss the evidence suggesting the emerging negative role for FH.

Sleep in eating disorders

Sleep research on eating disorders has addressed two major questions: (1) the effects of chronic starvation in anorexia nervosa and of rapidly fluctuating eating patterns in bulimia nervosa on the sleep regulating processes and (2) the search for a significant neurobiological relationship between eating disorders and major depression. At present, the latter question appears to be resolved, since most of the available evidences clearly underline the notion that eating disorders (such as anorexia and bulimia nervosa) and affective disorders are two distinct entities. Regarding the effects of starvation on sleep regulation, recent research in healthy humans and in animals demonstrates that such a condition results in a fragmentation of sleep and a reduction of slow wave sleep. Although several peptides are supposed to be involved in these regulatory processes (i.e. CCK, orexin, leptin), their mode of action is still poorly understood. In opposite to these experimentally induced sleep disturbances are the findings that the sleep patterns in eating disorder patients per se do not markedly differ from those in healthy subjects. However, when focusing on the so-called restricting anorexics, who maintain their chronic underweight by strictly dieting, the expected effects of malnutrition on sleep can be ascertained. Furthermore, at least partial weight restoration results in a 'deepening' of nocturnal sleep in the anorexic patients. However, our knowledge about the neurobiological systems (as well as their circadian pattern of activity) that transmit the effects of starvation and of weight restoration on sleep is still limited and should be extended to metabolic signals mediating sleep.

Slow-wave sleep and ventricular size: a comparative study in schizophrenia and major depression

BACKGROUND

A slow-wave sleep (SWS) deficit and a shortened rapid eye movement (REM) sleep latency are commonly described in schizophrenia and depression. In addition, a close association between ventricular system measurements and SWS are reported in both disorders; however, a comparative, combined polysomnographic and computed tomographic investigation is lacking.

METHODS

In the present post hoc study we analyzed the electroencephalographic sleep pattern and the ventricular brain ratio (VBR) of 14 drug-naive schizophrenic patients and of 14 depressed patients who were drug-free for a sufficient time period.

RESULTS

Whereas the depressives showed the often described SWS and REM sleep changes, these measurements were quite normal in the schizophrenics. The VBR values were similar in both patient groups and exceeded in 71% of the schizophrenics and in 50% of the depressives the cutoff point of a "normal" VBR value. A close association between SWS and VBR was found in the depressives but not in the schizophrenics.

CONCLUSIONS

The results of the present study of drug-naive schizophrenic patients and of depressed patients indicate only a minor overlap between the pathophysiological changes observed in both disorders. Therefore, the present investigation adds evidence for the assumption that schizophrenia and depression are etiopathogenetically distinct entities, rather than representing points on a continuum of liability.

Is anorexia nervosa a neuropsychological disease?

Evidence for central nervous system, and more particularly cortical, etiology of anorexia nervosa is reviewed. Topics covered are neuropsychiatric comorbidity, inheritance patterns, the neurobiology of body-image disturbance and of the eating function, perinatal and alcoholic insult to the brain, neurochemical and neuroelectric disturbance, anatomic and metabolic brain imaging, and neuropsychological impairment. It is concluded that there is indeed an important neuropsychological etiological dimension to anorexia nervosa. The profile most frequently associated with anorexia nervosa is right posterior hypometabolism, followed by right anterior hypermetabolism, both associated with right-sided abnormal electroencephalogram spiking. It is also proposed that bulimia consists of a "positive" neurological subtype and that restricting anorexia represents a "negative" neurological subtype. Priorities for further research into anorexia nervosa are specified to include twin adoption studies, brain electrical topography studies, postmortem histological studies, and experimentally inspired neuropsychological studies.

Sleep electroencephalographic patterns and cranial computed tomography in anxiety disorders

Sleep electroencephalogram (EEG) and cranial computed tomography (CT) were performed in 15 drug-free inpatients with an anxiety disorder. The sleep EEG of the patients was characterized by a decreased sleep efficiency and a shortened REM latency. Under visual inspection, four (27%) patients showed ventricular enlargement, while an increased ventricular brain ratio (VBR) was measured in five (33%). No clear associations between EEG sleep measures and ventricular size became obvious, although there is good evidence in the literature that sleep continuity, slow-wave sleep, or REM sleep are closely related to ventricular and/or sulcal widening in various neuropsychiatric disorders.

All-night electroencephalographic sleep and cranial computed tomography in depression. A study of unipolar and bipolar patients

All-night electroencephalographic (EEG) sleep recording and cranial computed tomography were performed in 24 inpatients with major depression (14 unipolar, 10 bipolar). The patients showed the characteristic "depression-like" EEG sleep alterations and their ventricular brain ratio (VBR) was increased compared with the control subjects. No major differences were found between the unipolar and the bipolar groups. There was a close and positive association between the VBR values and several measures of slow wave sleep. It is hypothesized that this relationship is due to an altered function of the limbic-hypothalamic-pituitary-adrenocortical axis in depression that affects both EEG sleep and brain morphology.

Nocturnal sleep in Huntington's disease

Nocturnal sleep was studied in 16 inpatients with Huntington's disease. In comparison with healthy controls, patients exhibited a disturbed sleep pattern with increased sleep onset latency, reduced sleep efficiency, frequent nocturnal awakenings, more time spent awake and less slow wave sleep. These abnormalities correlated in part with duration of illness, severity of clinical symptoms, and degree of atrophy of the caudate nucleus. Patients showed an increased density of sleep spindles.

Nocturnal sleep EEG in patients with HIV infection

Nocturnal sleep was studied in 14 human immunodeficiency virus (HIV)-positive patients without opportunistic infections of the central nervous system. Seven patients had no bodily complaints at the time of the investigation. Patients exhibited an impaired nocturnal sleep with longer sleep onset latency, reduced total sleep time, reduced sleep efficiency, and more time spent awake and in stage 1. Stage 2 sleep was significantly decreased; in 2 cases, sleep spindle density was extremely low. REM latency was reduced and correlated negatively with depressive symptomatology, while the percentages of REM and slow wave sleep were normal. No significant differences in sleep parameters were present among patients in different stages of the illness, or between patients with and without bodily complaints. Ventricular size and sulcal width on computed tomography scans correlated with sleep variables indicating reduced sleep quality, and with REM density. Decreased tryptophan plasma levels were associated with shorter and less efficient sleep, and with reduced stage 2 sleep. The findings demonstrate that sleep EEG investigations can be valuable for detecting and monitoring central nervous system affection in HIV-positive individuals.

Brain morphology and sleep EEG in patients with Huntington's disease

In 12 patients with Huntington's disease, the relationship between brain morphology, nocturnal sleep EEG, and clinical variables was studied. Global cerebral atrophy did not significantly correlate with sleep parameters, whereas atrophy of the caudate nuclei was associated with reduced slow wave sleep and increased time spent awake. Several clinical parameters (e.g., anergia and thought disturbance scores of the Brief Psychiatric Rating Scale, illness duration and global clinical assessment) showed significant correlations with global cerebral atrophy. Similar studies in other neuropsychiatric disorders demonstrate associations between sleep alterations and brain morphological changes of different localizations, thus pointing to a complex relationship between both. It can be hypothesized that the caudate nuclei may be involved in sleep regulation; indirect evidence from studies with positron emission tomography (PET) point in the same direction.