As part of the Collaborative Study of the Psychobiology of Depression, we have examined the pretreatment growth hormone response (ΔGH) to insulin (0·1 U/kg) and the magnitude of the hypoglycaemic response in a large number of well-defined depressed patients (N = 132) and healthy controls (N = 80). After applying rigorous exclusion criteria, data were analysed from 93 patients and 66 controls for blood glucose response and from 56 patients and 52 controls for ΔGH. Depressed patients, either unipolar or bipolar, showed less of a fall in glucose than controls. A weak association was found between the magnitude of the fall in glucose and the severity of depression. No significant differences were found in values for ΔGH between the unipolar or bipolar depressed patients and controls either for males, pre-menopausal or postmenopausal females, or the total female group. These data do not support previous claims of a lowered ΔGH response to insulin in depressed patients. However, the resistance to hypoglycaemia seen in the depressed patients is consistent with previous reports.

Preliminary data are presented from the NIMH Collaborative Study on the psychobiology of depression, biological studies, dealing with relationships between the pre-treatment levels of the neurotransmitter metabolites 3-methoxy-4-hydrophenethyleneglycol (MHPG), 5-hydroxyindoleacetic acid (5-HIAA) and homovanillic acid (HVA) and the subsequent therapeutic response of depressed patients to imipramine or amitriptyline. Eighty-seven depressed patients were studied during pre-treatment and treatment periods. It has been found that (1) both low pre-treatment urinary MHPG and low CSF 5-HIAA values are associated with a response to imipramine; these relationships were not artefacts due to sex or age; (2) there were no significant relationships between pre-treatment urinary MHPG, CSF MHPG, 5-HIAA, or HVA values and the subsequent response, or failure of response, to amitriptyline; (3) there was not a bimodal distribution for CSF 5-HIAA. For both males and females, there were positive and statistically significant correlations between CSF MHPG and urinary MHPG; for the females, there were positive and significant correlations between both urinary and CSF MHPG and CSF 5-HIAA. The theoretical and practical implications of these findings are discussed.

This research is part of the NIMH-CRB Collaborative Study on the psychobiology of depression. The main objective of the research programme is to test hypotheses concerning the interaction of neurobiological mechanisms and behaviour in the depressive disorders. Part I of the report describes the rationale and the overall approach to measuring behavioural state and outcome in the research programme. Part II reports on the results of applying the behavioural methods to a comparison of the clinical phenomenology of unipolar and bipolar depression. The behavioural patterns expressed during the episode by the two groups are different. Further, the two types are shown to react differently to treatment with tricyclic drugs, reinforcing the thesis that they are qualitatively distinct forms of the depressive disorders.

There are many reports which suggest that patients with affective illness (mania and/or depression) have abnormalities in the functioning of one or more neurobiological systems. At a conference convened by the Clinical Research Branch, Division of Extramural Research Programs, National Institute of Mental Health, these findings were reviewed and some of the factors impeding movement towards a more complete and integrated view of the functioning of neurobiological systems in patients with mania or depression were identified. As a result, a multi-research centre, collaborative approach to the study of the psychobiology of affective disorders was developed. In this collaborative programme, which has now been underway for several years, the focus has been upon: (a) the assessment of the functioning of several different types of biological systems in the same patient, both before and during treatment; (b) obtaining a reasonably large number of patients and comparison subjects; and (c) the use within and across centres of standardized diagnostic categories and behavioural rating methodologies. In this paper the history, background, and rationale for this collaborative effort are reviewed. Those biological systems chosen for study are noted, and issues such as reliability and validity of diagnoses, measurement of state variables, assessment of change with treatment, and logistical and coordinating problems are discussed.

A survey was made of studies reporting population prevalence rates for the different types of neurotic disorders in order to determine the utility of these rates for familial and genetic research. Rates varied considerably among studies both for all neuroses and for particular types of neuroses, even among studies using similar methods examining similar populations. The method of case ascertainment (personal interview versus record searches) and the threshold for defining an individual as affected are important contributors to variation. The use of the published prevalence rates for quantifying familial and genetic effects on the neurotic disorders would be premature.

In a previous study (Liss, Welner and Robins, 1973), the hospital records of 212 patients who received a diagnosis of personality disorder other than antisocial personality were studied. The clinical information about these patients was used to arrive at an established psychiatric diagnosis based upon the rigorous diagnostic criteria designed for research (Feighner, Robins, Guze, Woodruff, Winokur, and Munoz, 1972). In 118 cases (56 per cent) the clinical symptoms and the course of the disorder met the required diagnostic criteria for one or more of the established psychiatric diagnoses. The remaining 94 patients (44 per cent) had too few symptoms to meet these criteria.

A diagnosis of personality disorder is commonly used in psychiatry. It is generally agreed to refer to a disorder manifested by limited adaptive flexibility and certain relatively fixed ineffectual modes of behaviour (Ausubel, 1961; Brody and Lindbergh, 1967; DSM II, 1968; Noyes and Kolb, 1958; Schneider, 1950; Small, Small, Alig and Moore, 1970; Walton, Foulds, Littman and Presly, 1970). However, of twenty or more different types of personality disorders only antisocial personality has been differentiated by rigorous criteria as a distinct diagnostic entity (Robins, 1967; Feighner, Robins, Guze, Woodruff, Winokur and Munoz, 1972; Robins, 1966).

It has been shown in a record study (Welner, Liss, Robins and Richardson, 1972), and confirmed in a follow-up study using a blind structured interview (liss, Welner, and Robins, 1972), that most of those in-patients in our studies who were discharged without a diagnosis met the criteria for a psychiatric diagnosis when rigorous criteria were used (Feighner, Robins, Guze, Woodruff, Winokur and Munoz, 1972). However, using methods and evidence from these studies, out of the 109 patients who were followed-up for a mean period of 39 months, 25 patients were conspicuous in that even after initial or repeated admissions to hospital, after their records were reviewed, and after a follow-up study, they still remained undiagnosable. It was the purpose of this study to describe and evaluate in detail this group of 25 patients with particular reference to the possible reasons why they remained undiagnosed and therefore were defined by us as ‘undiagnosable’.

Bereavement is a crisis from which few people are spared. In spite of its universality, there have been few carefully designed, well documented, systematic studies describing its natural history and its relationship to other affective disturbances.

Lindemann (6) first used the term ‘anticipatory grief’. It referred to the separation of two people with one anticipating the death of the other and preparing for it to such an extent that on the return of the other there is rejection rather than happy acceptance. In his example a soldier returning from combat complained that his wife no longer loved him and was seeking a divorce. After seeing the man, Lindemann attributed the problem to the wife's ‘anticipatory grief. It seems there could be other explanations to his example.

In a previous report records were studied of 256 in-patients who were discharged as undiagnosed, i.e. in-patients who at the time of discharge did not have a definable psychiatric illness (Welner, Liss, Robins and Richardson, 1972). In that study it was shown that when rigorous criteria for psychiatric research (Feighner, Robins, Guze, Woodruff, Winokur and Munoz, 1972) were used 68 per cent of these patients met the criteria for an established psychiatric disorder. It was concluded that: (1) The chart review diagnoses for a population of undiagnosed patients consisted of a variety of established psychiatric disorders and the population was not homogeneous. (The chart review diagnosis is a diagnosis obtained by review of the patients' hospital records and evaluating the information by using diagnostic criteria for psychiatric disorders.) (2) The most efficient way to arrive at a diagnosis was by structured rather than conventional narrative interview. This study is a follow-up study of these patients and attempts to evaluate the validity of the chart review diagnosis. A concordance between the chart review diagnosis and follow-up diagnosis supports the above conclusions. The follow-up study also served to establish diagnosis in patients who had too few symptoms initially to meet the criteria for a diagnosis.

Among patients admitted to psychiatric hospitals are some who are discharged undiagnosed. It can be expected that psychiatric patients, as patients in other medical specialties, who do not meet the criteria for any of the established diagnostic groups, will be considered as undiagnosed until the course of their illness and clinical picture makes the diagnosis evident. The undiagnosed patients are a diagnostic and therapeutic challenge. Because in a space of five years, 256 patients left undiagnosed, it was felt that this group merited investigation.

Physicians and patients frequently assume a causal connection between life events and subsequent episodes of psychiatric illness. It seems to ‘make sense’ that an illness which is to some extent manifested by disordered emotions could be caused in part by emotion-producing events. But plausibility alone is no proof of the truth of such an assumption. Realizing this, several investigators have conducted systematic studies of the interrelationships of life events and illnesses, both psychiatric and medical. Such work has been reported by Adamson and Schmale (1), Holmes, et al. (3, 7, 9, 10) Brown and Birley (2) Clayton, et al. (4) Morrison, et al. (11) Murphy, et al. (12, 13) and Hudgens, et al. (8). These authors differed regarding the specific question of whether illnesses may be caused by emotion-producing stress. The first six of the above papers presented positive evidence for such a cause-effect relationship. The last four papers reported that psychiatric patients had significantly more interpersonal conflicts than did well persons or medically ill persons, at least while their psychiatric illnesses were in progress; but the latter authors were unable to find evidence that any type of stress, interpersonal or otherwise, played a causative role in the illnesses. Disagreements among all these workers may be traced to differences in both theoretical approach and methodology.

Primary affective disorders, depressive or manic episodes in patients who have been psychiatrically well previously, or who have had episodes of mania or depression without other psychiatric illnesses (Robins and Guze, 1969), are associated with high suicide rates. The following data indicate that the suicide risk among these patients is over thirty times greater than that of the population without these disorders, and that the risk of suicide compared to other causes of death may be increased early in the course of the illness.