The Munich vulnerability study on affective disorders: microstructure of sleep in high-risk subjects

Vulnerability markers for affective disorders have focused on stress hormone regulation and sleep. Among rapid eye movement (REM) sleep, increased REM pressure and elevated REM density are promising candidates for vulnerability markers. Regarding nonREM sleep, a deficit in amount of and latency until slow wave sleep during the first half of the night is a characteristic for depression. To further elucidate whether changes in the microstructure of sleep may serve as vulnerability markers we investigated the premorbid sleep composition in 21 healthy high-risk proband (HRPs) with a positive family history for affective disorders and compared HRPs with a control group of healthy subjects (HCs) without personal and family history for psychiatric disorders. The sleep electroencephalogram (EEG) was conventionally scored and submitted to a quantitative EEG analysis. The main difference in sleep characteristics between HRPs and HCs was an abnormally increased REM density. Differences in the spectral composition of sleep EEG were restricted to an increased power in the sigma frequency range. Since the HRP group comprised six unrelated and 15 related subjects we controlled for sibling effects. We could replicate the increased REM density in the group of HRPs whereas elevated power in the low sigma frequencies persisted only with approaching significance. The present study further supports elevated REM density as putative vulnerability marker for affective disorders. However, sleep EEG in our group of HRPs did not show slow wave sleep abnormalities. Ongoing follow up investigations of HRPs will clarify whether the observed increase in sigma EEG activity during nonREM sleep is of clinical relevance with respect to the likelihood to develop an affective disorder.     

The Munich vulnerability study on affective disorders: premorbid polysomnographic profile of affected high-risk probands.

BACKGROUND: The most characteristic alterations in the sleep electroencephalogram (EEG) during major depression are a shortened latency to rapid eye movement (REM) sleep and an elevated REM density. Because these changes persist in remission, they might represent vulnerability markers. To identify vulnerability markers, we investigated premorbid sleep EEG parameters in healthy high-risk probands (HRPs) with a positive family history of affective disorders. METHODS: We identified 136 depressed inpatients from referrals to our hospital who had first-degree relatives with an affective disorder as well as first-degree relatives with no current or lifetime history of psychiatric disorders. The latter (the HRP group) were investigated by polysomnography. During the follow-up period, 20 HRPs developed an affective disorder. Their premorbid sleep data were analyzed. RESULTS: Premorbid sleep EEG of affected HRPs showed an increased REM density (total night and first REM period) compared with the control group without personal or family history of a psychiatric disorder. CONCLUSIONS: Increased REM density can be observed not only in patients with depression, but also in their healthy relatives. Moreover, it is predictive for the onset of a psychiatric disorder. Therefore, it can be recommended as a possible endophenotype of these diseases.     

Temporal coherence in ultradian sleep EEG rhythms in a never-depressed, high-risk cohort of female adolescents.

BACKGROUND: Previous work has indicated that low temporal coherence of ultradian sleep electroencephalographic rhythms is characteristic of depressed patients and of depressed women, in particular. It may also be evident in one quarter of those at high risk, based on a family history of depression. METHODS: The present study evaluated temporal coherence of sleep electroencephalographic rhythms in 41 adolescent girls with a maternal history of depression (high risk) and 40 healthy controls (low risk). The entire sample was followed clinically every 6 months for 2 years. RESULTS: Temporal coherence was significantly lower among the high-risk girls than in controls. Regression analyses predicted group from coherence values and correctly classified 70% of the high-risk group with a false-positive rate of 5% among controls. Moreover, 54% of the high-risk girls were identified with extreme low coherence. On clinical follow up, 14 girls showed depressive symptoms, 9 in the high-risk group (22.5%) and 5 controls (12.2%). Six met DSM-IV criteria for first-episode major depressive disorder, five high-risk and one control. Most importantly, 41% of those identified as having the most abnormal coherence values either showed symptoms of depression or met diagnostic criteria upon follow up. CONCLUSIONS: Low temporal coherence is evident in adolescent girls at high risk for depression. The more abnormal the coherence, the greater the risk of a first episode of major depressive disorder within 2 years of sleep study, approximately 10 times greater than in controls.     

Electroencephalographic sleep in spousal bereavement and bereavement-related depression of late life.

Although spousal bereavement in late life is common and frequently leads to major depression, the boundary between bereavement without a depressive syndrome and bereavement-related depression has been insufficiently studied from a physiological perspective. Because other forms of depression are associated with physiological changes, including sleep, we have attempted to clarify the relationship of bereavement and bereavement-related depression by investigating electroencephalographic (EEG) sleep in 31 elderly volunteers with recent spousal bereavement, stratified by the presence (n = 15) or the absence (n = 16) of major depression (Research Diagnostic Criteria). Entry into the study was limited to volunteers without a personal history of psychiatric disorder. As hypothesized, bereaved subjects with major depression had significantly lower sleep efficiency, more early morning awakening, shorter rapid eye movement (REM) latency, greater REM sleep percent, and lower rates of delta wave generation in the first nonREM (NREM) period, compared with bereaved subjects without depression. Furthermore, the sleep of bereaved subjects with single-episode major depression resembled that of elderly patients with recurrent unipolar major depression (n = 15) on measures noted above. Sleep in bereavement without depression was similar to that of 15 healthy control subjects (neither bereaved nor depressed). These findings suggest that the current DSM-III-R concept of uncomplicated bereavement is not confirmed, as the sleep patterns of subjects who develop a depressive syndrome in the context of bereavement, many of whom might be considered to have "uncomplicated bereavement" by DSM-III-R standards, are identical to sleep patterns found in major depressive episodes. To our knowledge, this is the first study of EEG sleep in spousal bereavement with and without major depression.     

Sleep in spousally bereaved elders with subsyndromal depressive symptoms.

Spousal bereavement in late life frequently leads to major depression. However, many people suffer from "minor" depressive symptoms that entail considerable suffering even in the absence of syndromal major depression. We describe longitudinal electroencephalographic (EEG) sleep and clinical evaluations in 14 elderly, recently spousally bereaved subjects who were experiencing subsyndromal depressive symptoms. While subjects did not meet diagnostic criteria for syndromal major depression, they did have mildly elevated scores on the Hamilton Rating Scale for Depression (mean = 10.6, range = 8-16) at the time of initial sleep studies (T1), which were carried out, on average, 5.5 months after loss of the spouse. Entry into the study was limited to volunteers who did not have a personal history of major depression or psychiatric disorder. Twelve subjects underwent followup clinical and EEG sleep evaluations (T2), 9.9 months after spousal loss. Fifty percent continued to show depressive symptoms at 6-month followup. Test-retest comparisons of sleep and clinical measures were made with a group of sex- and age-matched control subjects who were neither bereaved nor depressed. EEG sleep measures did not significantly correlate with time from loss of spouse, severity of depressive symptoms, or subjective sleep quality. Analysis of variance with repeated measures detected a significant group X time interaction effect for delta sleep ratio (decreasing in controls but increasing in the bereaved).     

Electroencephalographic sleep of adolescents with major depression and normal controls

Forty-nine, mostly outpatient (86%), nonbipolar adolescents, aged Tanner stage III to 18 years, with a current diagnosis of major depressive disorder and 40 adolescents without current presence or history of psychiatric disorder were studied polysomnographically for three consecutive nights. Sleep latency was significantly longer in the depressive groups. The nonendogenous depressive patients exhibited significantly more awake time and lower sleep efficiency during the sleep period. No significant group differences were found for first rapid eye movement (REM) period latency, REM density, or any other REM sleep measures. Age correlated significantly with REM latency and delta sleep time, especially among depressive patients. No significant correlations between sleep measures and severity of illness were found. It appears that the classic REM sleep findings associated with the adult depressive syndrome are not present among depressive adolescents, indicating a later ontogeny for these abnormalities.     

Cholinergic REM sleep induction test in subjects at high risk for psychiatric disorders.

The influence of the cholinergic agonist RS 86 on electroencephalographic (EEG) sleep was investigated in 21 healthy members of families identified as being at high risk for psychiatric disorders and in 17 healthy control subjects without any personal or family history of a psychiatric illness. In comparison to the placebo night, the administration of RS 86 led to a shortening of rapid eye movement (REM) latency in both groups. This effect, however, was much more pronounced in the high-risk group, whereas in the control subjects the arousal system and the slow-wave sleep during the first nonREM period were more affected. These observations suggest that the cholinergic action on sleep regulating mechanisms has differing preferential targets in high-risk probands and in control subjects.     

The effects of breathing-related sleep disorders on mood disturbances in the general population

BACKGROUND: Results of clinical studies suggest that there may be a relationship between breathing-related sleep disorders and depressive disorders. This study aims to assess the impact of breathing-related sleep disorder on major depressive disorder in the general population. METHOD: A cross-sectional telephone survey was carried out between 1994 and 1999 in the general population of the United Kingdom, Germany, Italy, Portugal, and Spain. A total of 18,980 randomly selected subjects aged 15 to 100 years and representative of the general population of their respective countries participated in the study. The questionnaire included a series of questions about sleep quality, breathing-related sleep disorder symptoms, mental disorders, and medical conditions. Data are presented using point prevalence. RESULTS: 2.1% of the subjects were found with obstructive sleep apnea syndrome at the time of the interview, and 2.5% had some other type of DSM-IV breathing-related sleep disorder diagnosis. The association of DSM-IV breathing-related sleep disorder diagnosis and major depressive disorder diagnosis was found in 0.8% of the sample. As many as 18% of individuals with a major depressive disorder diagnosis also have a DSM-IV breathing-related sleep disorders diagnosis, and 17.6% of subjects with a DSM-IV breathing-related sleep disorders diagnosis have a major depressive disorder diagnosis. Multivariate models showed that even after controlling for obesity and hypertension, the odds of having a DSM-IV breathing-related sleep disorders diagnosis was 5.26 for individuals with a major depressive disorder diagnosis. CONCLUSION: About 800 of 100,000 individuals have both a breathing-related sleep disorder and a major depressive disorder. The identification of 1 of these 2 disorders should prompt the investigation of the other disorder since nearly a fifth of them have the other disorder.     

The Munich vulnerability study on affective disorders: premorbid psychometric profile of affected individuals

OBJECTIVE: We conducted a longitudinal high-risk study to identify psychometric vulnerability markers for affective disorders. METHOD: We examined 82 healthy subjects [high-risk probands (HRPs)] with at least one first-degree relative suffering from an affective disorder. The premorbid psychometric profile of 20 HRPs who developed a psychiatric disorder during follow-up was compared with the profile of control subjects without personal and family history of psychiatric disorders matched for age and gender. RESULTS: Somatization, complaints (vegetative lability), and perception of strain are increased in HRPs who developed a psychiatric disorder. These alterations were not influenced by the time interval until the onset of the disorder. CONCLUSION: The premorbid psychometric profile in subjects at high risk for affective disorders is characterized by somatization, complaints, and elevated perception of strain. Together with previous findings our results suggest that these alterations can be regarded as potential vulnerability markers for affective disorders.     

The Maudsley Bipolar Disorder Project: the effect of medication,family history,and duration of illness on IQ and memory in bipolar I disorder

BACKGROUND: Despite the growing recognition of the importance of cognitive impairment in psychiatric disorders, the effect of clinical factors, such as medication use and family history of affective disorders, on cognition in bipolar I disorder patients still remains unclear. This study examines the contribution of known and potential predictors to both general intellectual function and memory in a representative population of bipolar I disorder patients. METHOD: Of the 425 patients receiving treatment within a defined catchment area, 63 were identified as having bipolar I disorder. Of these patients, 43 were enrolled in the study and participated in a personal interview by a psychiatrist. All patients were invited to participate in a personal interview by a psychiatrist, and information on family history, past psychiatric history, past and current treatments, duration of illness, and age at onset was collected, in addition to demographic data. Cognitive performance was assessed using the Wechsler Adult Intelligence Scale-Revised, the National Adult Reading Test, and the Wechsler Memory Test III. RESULTS: Forty-three patients with DSM-IV bipolar I disorder were enrolled into the Maudsley Bipolar Disorder Project. Patients on treatment with antipsychotic drugs had a lower current full scale IQ, lower general memory scores, and lower working memory scores. A family history of affective disorders was associated with a higher full scale IQ, but not with either general or working memory measures. Duration of illness was negatively associated with general memory scores, but had no effect on either IQ or working memory measures. CONCLUSION: Current antipsychotic medication, duration of illness, and family history of affective disorder were the most significant predictors of IQ and memory function in bipolar I disorder patients.     

Reduced sleep spindle activity in schizophrenia patients

OBJECTIVE: High-density EEG during sleep represents a powerful new tool to reveal potential abnormalities in rhythm-generating mechanisms while avoiding confounding factors associated with waking activities. As a first step in this direction, the authors employed high-density EEG to explore whether sleep rhythms differ between schizophrenia subjects, healthy individuals, and a psychiatric control group with a history of depression. METHOD: Healthy comparison subjects (N=17), medicated schizophrenia patients (N=18), and subjects with a history of depression (N=15) were recruited. Subjects were recorded during the first sleep episode of the night with a 256-electrode high-density EEG. Recordings were analyzed for changes in EEG power spectra, power topography, and sleep-specific cortical oscillations. RESULTS: The authors found that the schizophrenia group had a significant reduction in centroparietal EEG power, from 13.75 to 15.00 Hz, in relation to both the comparison and depression groups. No significant difference in EEG power between the comparison and depression groups was identified. The authors also found a decrease in sleep spindle number, amplitude, duration, and integrated spindle activity in schizophrenia patients. Furthermore, integrated spindle activity had an effect size corresponding to 93.0% or 90.2% separation of the schizophrenia from the comparison or depression group. CONCLUSIONS: Sleep spindles are generated by the thalamic reticular nucleus in conjunction with specific thalamic nuclei and are modulated by corticothalamic and thalamocortical connections. The deficit in sleep spindles in schizophrenia subjects may reflect dysfunction in thalamic-reticular and thalamocortical mechanisms and could represent a biological marker of illness.     

Factors associated with chronic depressive episodes: a preliminary report from the STAR-D project

OBJECTIVE: To identify baseline sociodemographic and clinical factors associated with a current chronic major depressive episode (MDE). METHOD: Outpatients with major depressive disorder enrolled in 41 US primary or psychiatric care sites were divided into two groups based on self-report of current episode length (<24 or > or =24 months). Logistic regression models were used to identify factors associated with chronicity of current depressive episode. RESULTS: About 21.2% of 1380 subjects were in current, chronic MDEs. Older age, less education, lower income, no private insurance, unemployment, greater general medical illness burden, lower physical quality of life, concurrent generalized anxiety disorder, fewer prior episodes, and history of prior suicide attempts were all associated with chronic episodes. Blacks, Hispanics, and patients receiving care in primary as opposed to psychiatric care settings exhibited greater chronicity. CONCLUSION: Chronic depressive episodes are common and are associated with greater illness burden, comorbidity, socioeconomic disadvantage, and racial/ethnic minority status.     

Screening for and detection of depression, panic disorder, and PTSD in public-sector obstetric clinics

OBJECTIVE: This study assessed rates of detection and treatment of minor and major depressive disorder, panic disorder, and posttraumatic stress disorder among pregnant women receiving prenatal care at public-sector obstetric clinics. METHODS: Interviewers systematically screened 387 women attending prenatal visits. The screening process was initiated before each woman's examination. After the visit, patients were asked whether their clinician recognized a mood or anxiety disorder. Medical records were reviewed for documentation of psychiatric illness and treatment. RESULTS: Only 26 percent of patients who screened positive for a psychiatric illness were recognized as having a mood or anxiety disorder by their health care provider. Moreover, clinicians detected disorders among only 12 percent of patients who showed evidence of suicidal ideation. Women with panic disorder or a lifetime history of domestic violence were more likely to be identified as having a psychiatric illness by a health care provider at some point before or during pregnancy. All women who screened positive for panic disorder had received or were currently receiving mental health treatment outside the prenatal visit, whereas 26 percent of women who screened positive for major or minor depression had received or were currently receiving treatment outside the prenatal visit. CONCLUSIONS: Detection rates for depressive disorders in obstetric settings are lower than those for panic disorder and lower than those reported in other primary care settings. Consequently, a large proportion of pregnant women continue to suffer silently with depression throughout their pregnancy. Given that depressive disorders among perinatal women are highly prevalent and may have profound impact on infants and children, more work is needed to enhance detection and referral.     

Epidemiology of depression and its treatment in the general population

This study examines the correlates of a major depressive disorder and its treatment in the general population. The sample was composed of 6694 individuals aged between 18 and 96 years, representative of the general population of the states of California and New York (48 million inhabitants aged 18 years or older). They were interviewed by telephone using the Sleep-EVAL system. The interviews included various sleep and health topics and the assessment of DSM-IV sleep and psychiatric disorders. The 1-month prevalence of a major depressive disorder was 5.2% in the sample, and was higher in women, middle-aged and non-Hispanic white individuals. Obesity (BMI > or =30kg/m(2)), poor health status and smoking were also strongly correlated with a major depressive disorder. A total of 57.7% of depressed subjects were receiving some forms of treatment for depression: 28.3% were taking antidepressants (alone or in combination with psychiatric health care) and 29.4% received psychiatric health care (without antidepressant medication). Severity of depression, ethnicity and weight (overweight or obese) were strongly associated with the presence of treatment. A major depressive disorder is frequent in the general population. Although its identification and treatment have improved over the years, some segments of the population, namely elderly and non-white individuals are less likely to receive appropriate care.     

Altered interaction between cardiac vagal influence and delta sleep EEG suggests an altered neuroplasticity in patients suffering from major depressive disorder

Jurysta F, Kempenaers C, Lancini J, Lanquart J‐P, van de Borne P, Linkowski P. Altered interaction between cardiac vagal influence and delta sleep EEG suggests an altered neuroplasticity in patients suffering from major depressive disorder. Objective: Major depressive disorder (MDD), which is associated with altered neuroplasticity and increased relative cardiac sympathic activity, enhances the risk of cardiovascular pathologies. Interaction between cardiac sympatho‐vagal indexes and delta sleep power is probably altered in MDD. Method: Sleep characteristics and cardiac sympatho‐vagal indexes of 10 depressive patients were compared to 10 control men across the first three non‐rapid eye movement (NREM)–REM cycles. Interaction between normalized high frequency (HF) and delta power bands was studied using coherence analysis. Results: Patients showed increased sleep latency, stage 1 and wake durations. No differences in heart rate variabilities were observed: Total power, HF and RR‐interval decreased from NREM to REM sleep and wakefulness in both groups. Gain value was lower in patients while coherence and phase shift were similar between groups. Modifications in HF appear 8 min before modifications in delta. Conclusion: Major depressive disorder is related to an altered link between cardiac vagal influence and delta sleep, suggesting disorders in cardiovascular controls and an altered neuroplasticity.     

Electroencephalographic sleep in psychotic depression. A valid subtype?

Electroencephalographic (EEG) sleep patterns were examined in 27 psychotic and 79 nonpsychotic subjects with major depression to evaluate the validity of the psychotic-nonpsychotic subtype dichotomy. Sleep in psychotic depression was characterized by increased wakefulness, decreased rapid eye movement (REM) sleep percentage, and decreased REM activity even after controlling for clinical differences in age, severity, and agitation. Psychotic depressive subjects also were more likely to have extremely short sleep-onset REM latencies. In psychotic depression EEG sleep varied as a function of total illness duration. Patients with recent-onset syndromes had profiles characterized by marked initial insomnia, increased stage 1 sleep percentage, and long REM latency; patients with illnesses of longer duration had extremely short REM latencies. Demonstration of selected EEG sleep variables discriminating between psychotic and nonpsychotic depression further supports psychotic depression as a distinct subtype of major affective disorder.     

Phenomenology of children and adolescents with bipolar spectrum disorders

CONTEXT: Children and adolescents who present with manic symptoms frequently do not meet the full DSM-IV criteria for bipolar I disorder (BP-I). OBJECTIVE: To assess the clinical presentation and family history of children and adolescents with BP-I, bipolar II disorder (BP-II), and bipolar disorder not otherwise specified (BP-NOS). DESIGN: Subjects and their primary caretaker were assessed by semistructured interview, and family psychiatric history was obtained from interview of the primary caretaker. SETTING: Outpatient and inpatient units at 3 university centers. PARTICIPANTS: A total of 438 children and adolescents (mean +/- SD age, 12.7 +/- 3.2 years) with BP-I (n = 255), BP-II (n = 30), or BP-NOS (n = 153). MAIN OUTCOME MEASURES: Lifetime psychiatric history and family history of psychiatric disorders. RESULTS: Youth with BP-NOS were not diagnosed as having BP-I primarily because they did not meet the DSM-IV duration criteria for a manic or mixed episode. There were no significant differences among the BP-I and BP-NOS groups in age of onset, duration of illness, lifetime rates of comorbid diagnoses, suicidal ideation and major depression, family history, and the types of manic symptoms that were present during the most serious lifetime episode. Compared with youth with BP-NOS, subjects with BP-I had more severe manic symptoms, greater overall functional impairment, and higher rates of hospitalization, psychosis, and suicide attempts. Elevated mood was present in 81.9% of subjects with BP-NOS and 91.8% of subjects with BP-I. Subjects with BP-II had higher rates of comorbid anxiety disorders compared with the other 2 groups and had less functional impairment and lower rates of psychiatric hospitalization than the subjects with BP-I. CONCLUSIONS: Children and adolescents with BP-II and BP-NOS have a phenotype that is on a continuum with that of youth with BP-I. Elevated mood is a common feature of youth with BP-spectrum illness.     

Interface of panic and depression: clinical and sleep EEG correlates

Four groups of subjects were compared with respect to their clinical and demographic status and electroencephalographic (EEG) characteristics, namely: primary major depressive disorder (PRI MDD); panic disorder (Panic); "Mixed" group comprising patients meeting full syndromal criteria for MDD and panic occurring concomitantly; and normal controls. The "Mixed" (MDD + Panic) patients were characterized by earlier age of onset of psychiatric illness, longer duration of current episode, greater intensity of symptoms, higher impairment of functioning, increased miscellaneous psychopathology, and greater objective stress and anger. With respect to sleep EEG variables, PRI MDD patients were clearly different from the other three groups. The sleep profile of the "Mixed" group occupies an intermediate position between the "pure" Panic and MDD groups. Classification of the "Mixed" patients based on the discriminant function coefficients of the Schedule for Affective Disorders and Schizophrenia and sleep analysis of the "pure" groups (PRI MDD and Panic) reveals that some patients are classified as true PRI MDD while others are classified as falling somewhere along the PRI-MDD/Panic spectrum. The separation of the PRI MDD from Panic and Normals, however, is clear, suggesting that sleep can be successfully used as a physiological marker in the separation of these conditions.     

EEG sleep in Cushing's disease and Cushing's syndrome: comparison with patients with major depressive disorder.

Because patients with Cushing' syndrome (CS) and Major depressive disorder (MDD) share features of hypercortisolism and the depressive syndrome, we compared electro-encephalographic (EEG) sleep in patients with pituitary-ACTH-dependent Cushing's syndrome (Cushing's disease, CD), patients with ACTH-independent Cushing's syndrome (AICS), patients with major depressive disorder (MDD), and normal subjects. There were substantial similarities in the abnormal polysomnography profiles of patients with CD, AICS, and MDD. All three patient groups demonstrated poorer sleep continuity, shortened rapid eye movement (REM) latency, and increased first REM period density compared with normal subjects. In addition, AICS patients and MDD patients had elevated REM activity and density. These findings are discussed in terms of models of pathophysiology that relate abnormalities in sleep, mood, and hypothalamic-pituitary-adrenal function.     

Dex/CRH-test response and sleep in depressed patients and healthy controls with and without vulnerability for affective disorders

Sleep electroencephalographic (EEG) abnormalities and increased hypothalamo-pituitary-adrenal (HPA) axis activity are the most prominent neurobiological findings in depression and were suggested as potential biomarker for depression. In particular, increased rapid eye movement sleep (REM) density, deficit in slow wave sleep and excessive stress hormone response are associated with an unfavorable long-term outcome of depression. Recent studies indicate that the sleep and endocrine parameters are related to each other. This study investigated the association of sleep structure including a quantitative EEG analysis with the results of the combined dexamethasone (Dex)/corticotropin-releasing hormone (CRH)-test in 14 patients with a severe major depression, 21 healthy probands with a positive family history of depression (HRPs) and 12 healthy control subjects without personal and family history for psychiatric disorders. As expected patients with depression showed an overactivity of the HPA axis, disturbed sleep continuity and prolonged latency until slow wave sleep in the first sleep cycle. Differences in microarchitecture of sleep were less prominent and restricted to a higher NonREM sigma power in the HRP group. Dexamethasone suppressed cortisol levels were positively associated with higher NonREM sigma power after merging the three groups. We also observed an inverse association between the ACTH response to the Dex/CRH-test and rapid eye movement sleep (REM) density in HRPs, with suggestive evidence also in patients, but not in controls. This contra-intuitive finding might be a result of the subject selection (unaffected HRPs, severely depressed patients) and the complementarity of the two markers. HRPs and patients with high disease vulnerability, indicated by an elevated REM density, seem to have a lower threshold until an actual disease process affecting the HPA axis translates into depression, and vice versa. To summarize, our findings provide further evidence that the HPA axis is involved in the sleep regulation in depression. These associations, however, are not unidimensional, but dependent on the kind of sleep parameters as well as on the selection of the subjects.