Unipolar depression: is it divisible into autonomous subtypes?

Depression spectrum disease has been defined as an illness in which a first-degree family member has alcoholism and/or antisocial personality. Pure depressive disease may be considered as the remainder of the depressive illnesses or more rigorously as depression in a person who has a family history of depression but no alcoholism. Evidence is presented that the course of the illness is different in the two groups. Depression spectrum disease is more variable, with more personality problems and interpersonal conflict. Preliminary data indicate that depression spectrum disease may be linked to such genetic markers as C3 or alpha-haptoglobin.

The validity of neurotic-reactive depression. New data and reappraisal.

A family history of alcoholism can be used as a validating factor in the diagnosis of reactive-neurotic depression. Not only is this true but there are clear data that indicate the presence of positive symptoms that can be used to make the diagnosis. A set of criteria based on previous research is presented for the diagnosis of neurotic-reactive depression. These criteria are based on a clustering of certain symptoms, events, and traits in patients with neurotic-reactive depression. The patients showed stormy life-styles, some specific symptoms, personality abnormalities, presence of life events before the onset of depression, and a family history of alcoholism. They had relatively few hospitalizations for depression and responded poorly to specific antidepressant treatment.

Suicide, attempted suicide, and relapse rates in depression.

Suicide, attempted suicide, and relapse rates in 519 depressives were examined, comparing the effects of different treatments. After six months, suicide attempts were seen significantly less frequently in the ECT groups (0.8%) than in the antidepressant group (4.2%) or the "adequate" antidepressant subgroup (7.0%) Fewer suicide attempts occurred in the ECT group compared to the antidepressant group among both in those who had attempted suicide prior to admission (0% vs 10%) and in those who had not (1.1% vs 3.6%). A history of attempted suicide showed a greater risk of both suicide (2.9%) in the following year and subsequent suicide attempt (5.9%). A depressive diagnosis may be as good a predictor of suicidal behavior as a history of attempted suicide. Relapse rates did not differ between treatment groups.

Mortality in depressed patients treated with electroconvulsive therapy and antidepressants.

The treatments of 519 depressed patients hospitalized from 1959 to 1969 were compared in a three-year follow-up study with particular reference to mortality. The electroconvulsive therapy (ECT) group had a significantly lower mortality than the inadequate antidepressant treatment group (P less than .05) and the group that received neither ECT nor antidepressants (P less than .025). Although the adequate antidepressant treatment group had a low mortality, statistically significant differences between this and other treatment groups could not be documented. Nonsuicidal deaths (P less than .005), and particularly myocardial infarctions (P less than .01), were significantly more frequent in the inadequately treated group compared to the adequately treated group. The superiority of adequate treatment is especially striking among men and among the older age groups. The results underscore the importance of adequate treatment of depression, especially in the older man.

Bipolar illness. The sex-polarity effect in affectively ill family members.

We examined the sex ratios of unipolar and bipolar affective disorder in 14 studies of bipolar illness. The data are highly consistent in demonstrating that (1) females are more frequently affected than males; (2) among affected females, the ratio of unipolar depression to bipolar illness is about 2:1; whereas (3) among affected males, the ratio of bipolar to unipolar illness is approximately 1:1. This curious discrepancy between the sexes may hold a clue to understanding familial transmission as well as heterogenicity in bipolar illness.

Newer experimental methods for classifying depression. A report from the NIMH collaborative pilot study.

A total of 83 patients receiving diagnoses of major depressive disorder in the pilot phase of the National Institute of Mental Health Collaborative Study of the Psychobiology of Depression were used to evaluate two newer methods of classifying depressive disorders on the basis of family history or course. A third subtype based on family history, nonfamilial depression, was compared with the two subtypes originally proposed by Winokur and colleagues, pure depression and depression spectrum diseases. The system for classifying depressions on the basis of course and antecedent disorder, primary vs secondary depression, was also compared. These data from the pilot study indicate that two newer systems for classification have some predictive, construct, and content validity, but both are in need of further investigation before they become accepted methods for the classification of depressive disorders.

Circadian analysis of plasma cortisol levels before and after dexamethasone administration in depressed patients.

Pituitary-adrenal regulation in healthy subjects and in depressed patients is very dynamic. Interpretation of results of the 1-mg dexamethasone suppression test has usually depended on the result of a single blood cortisol level measurement obtained in the morning or afternoon. We analyzed circulating cortisol concentrations by obtaining blood samples at 20-minute intervals for 24 hours before and after dexamethasone administration in depressed patients. The results illustrate the variability in patterns of escape from the effects of dexamethasone among depressed patients; they also indicate the influence of the sampling time on the test results and thus on the relationship of the test result to various clinical classifications. Finally, these results provide the basis for understanding the consequences of alternative sampling strategies.

The induction of mania. A natural history study with controls.

Previous reports have indicated that tricyclic antidepressants (TCAs) induce mania, but the studies suffer from lack of control groups. This study included 137 unipolar and 157 bipolar patients, all having two or more admissions to the same hospital. In many cases, the same patients were treated at one time with TCAs and at another time with no somatic therapy at all. Most patients who switched (20/23) were bipolar. Twenty-seven bipolar patients receiving no treatment had a 41% switch rate (rate/person). They had 32 admissions, and the rate of switch while receiving no treatment was 34% (rate/admission). Twenty-six patients receiving tricyclics had a 28% switch rate; for these, there were 30 admissions and the switch rate was 23%. Thus, it appears that the rate of induction of mania by TCAs is not greater than what one would expect from the natural history of the illness itself. Validity of these findings is attested to by the fact that lithium carbonate and neuroleptic treatment, as expected, significantly prevented the induction of mania.

Suicide in subtypes of major affective disorder. A comparison with general population suicide mortality.

We investigated the risk of suicide among 705 patients with primary unipolar depression, 302 patients with secondary unipolar depression, and 586 patients with bipolar affective disorder (BAD). The suicide rates among the study subjects were compared with that of the general population of Iowa, the area from which the subjects were selected. An increased risk of suicide was found in all psychiatric groups, except the group of female patients with BAD, which was associated with a lower risk of suicide than unipolar disorders. Thirty suicides (73.2%) occurred during the first two years of follow-up; this trend was particularly pronounced in female patients with primary unipolar depression and male patients with BAD.

Hypothalamic-pituitary-adrenal axis activity in depressive illness. Its relationship to classification.

Serum cortisol response to the 1-mg overnight dexamethasone suppression test was studied in 221 depressed patients and 109 nondepressed psychiatric controls. Nonsuppression distinguished patients with primary unipolar depression (65/146) from patients with secondary unipolar depression (0/42) and nondepressed controls (0/109). Furthermore, nonsuppression distinguished the three familial subtypes of primary unipolar depressive illness: familial pure depressive disease (FPDD; 38/50 patients), sporadic depressive disease (SDD; 24/55 patients), and depression spectrum disease (3/41 patients). Moderate elevations in baseline serum cortisol levels were found in FPDD, SDD, and bipolar depression. Medication did not affect the results. The data suggest that the depressive syndrome is composed of separate illnesses, each of which has a distinctive pattern of hypothalamic-pituitary-adrenal axis activity during the depressed state as well as a specific clinical and familial psychiatric history.

The Iowa record-linkage study. II. Excess mortality among patients with organic mental disorders.

Of 543 patients with organic mental disorders hospitalized at the University of Iowa Psychiatric Hospital, Iowa City, during a ten-year period, 87 died. This mortality was significant based on a control population. Patients of all ages were at risk for early death, especially those younger than 40 years. Risk was greatest during the first two years of follow-up; thereafter the observed death rate approached the expected rate. Patients were at special risk for death from "natural" causes, particularly cancer and heart disease among women, and influenza or pneumonia or "other" natural causes among men. During the first two years of follow-up, men were also at risk for death from accidents or suicide. Women with alcohol- and drug-related psychoses were at risk for death early in follow-up, but the diagnosis was not associated with risk from "unnatural death" in either sex.

The Iowa record-linkage study. I. Suicides and accidental deaths among psychiatric patients.

In a prospective investigation of 5,412 formerly hospitalized psychiatric patients, 68 committed suicide and 38 died from accidental causes. Both figures are significantly in excess of expectations based on a relevant control population. Those at significant risk included men and women of all ages, except those older than 69 years. Comparison of standardized mortality ratios suggests relatively greater risk for women and young persons. Risk for suicide was significant for patients of both sexes with acute schizophrenia, schizophrenia, affective disorders, and alcohol and other drug abuse, for men with neuroses, and for women with depressive neuroses. Risk for accidental death was greatest for those aged from 30 to 49 years and for those with personality disorders.

The Iowa record-linkage study. III. Excess mortality among patients with 'functional' disorders.

Our investigation of the pattern of mortality among former inpatients in nine diagnostic groups was based on deaths found among 4,869 former inpatients of the University of Iowa Psychiatric Hospital, Iowa City, during a ten-year period. Comparisons were made with expected values based on a relevant Iowa control population. The first two years of follow-up was a period of great risk but not after. Excessive mortality from "unnatural" causes was found among patients of either sex with an affective disorder, schizophrenia, alcohol or other drug abuse, and personality disorders, among men with acute schizophrenia or neuroses, and among women with depressive neuroses. Women with acute schizophrenia or a psychophysiologic disorder or special symptom were at risk for a "natural" death. These findings confirm the risk of reduced life span that patients in all nine categories share.

A prospective follow-up of patients with bipolar and primary unipolar affective disorder.

OBJECTIVE: As part of the National Institute of Mental Health Collaborative Study of the Psychobiology of Depression, the comparative course of manic depressive (bipolar) and primary unipolar patients was assessed. DESIGN: Systematic evaluation using structured instruments every 6 months for a period of 5 years with the recording of remissions, new episodes, and subsequent hospitalizations. PATIENTS: The number of subjects varied somewhat depending on the analyses conducted. For a comparison of course in bipolar patients and unipolar patients, 148 bipolars were compared with 172 unipolar patients. RESULTS: Both unipolar and bipolar patients were more likely to have episodes if they had episodes prior to index admission. Likewise, prior hospitalizations predicted multiple hospitalizations in follow-up. Chronicity was significantly more prevalent among unipolar depressives but in both unipolar and bipolar patients, chronicity diminished over time. Bipolar patients were more likely than unipolar patients to have multiple episodes at the 2-year and 5-year follow-ups. In bipolar patients, there was no difference in the number of episodes in follow-up between males and females but in unipolar patients, females were significantly more likely to have subsequent hospitalizations and episodes than males. Treatment variables did not relate to these differences. A family history of mania or schizoaffective mania predicted multiple episodes in bipolar patients but not in primary unipolar depressives. A family history of all affective illness (mania, schizoaffective mania, bipolar II illness, and depression) did not predict a multiple-episode course in either bipolar or unipolar illness. In unipolar patients, the independent variables leading to multiple-episode course in follow-up are being female, an early age of onset, and prior episodes. CONCLUSIONS: As a result of this systematic follow-up study, new data add to the distinction between bipolar and primary unipolar patients both as regards number of episodes in follow-up and also as regards risk factors that are associated with the multiple-episode course.

A family study of manic-depressive (bipolar I) disease. Is it a distinct illness separable from primary unipolar depression?

OBJECTIVE: To determine whether bipolar I illness is autonomous or part of a multifactorial continuum with unipolar depression. In this study, we compare familial bipolar I illness and depression among three groups: probands with bipolar I disorder; probands with primary unipolar disorder; and controls. We also examine a continuum of severity between psychotic and nonpsychotic patients with bipolar I disorder. Considerable data suggest that bipolar I illness is distinct from unipolar illness as regards epidemiology, familial psychiatric illness, course, response to treatment, and biologic findings. METHOD: Probands were separated into bipolar I and primary unipolar depressive groups. Personally interviewed family members of these patients were compared on variables of bipolar illness or schizoaffective mania and unipolar or schizoaffective depression. A personally examined control group was compared with the relatives of the two proband groups. Similar analyses were performed using data obtained by a systematic family history method. For the same familial variables, psychotic and nonpsychotic manic probands were compared. RESULTS: Familial mania is more frequent in families of patients with bipolar disease than in controls or in families of patients with primary unipolar disorder. The latter two groups did not differ in amount of mania. Unipolar depressive illness or schizoaffective depression was higher in families of probands with bipolar and unipolar disorder than in controls. Probands with bipolar disease separated into those who had psychotic symptoms (including schizoaffective mania) and no psychotic symptoms did not differ from each other in risk for familial mania or depression. CONCLUSIONS: Bipolar I illness is a separate illness from primary unipolar illness because of an increase in familial mania. Patients with primary unipolar disease and controls show the same amount of familial mania. Lack of an increase in familial illness according to the severity of bipolar disease is against an affective continuum.

The prediction of suicide. Sensitivity, specificity, and predictive value of a multivariate model applied to suicide among 1906 patients with affective disorders.

Stepwise multiple logistic regression was utilized in an attempt to develop a statistical model that would predict suicide in a group of 1906 Iowans with affective disorders admitted to a tertiary care hospital. The risk factors identified by this approach included the number of prior suicide attempts, suicidal ideation on admission, bipolar affective disorder (manic or mixed type), gender, outcome at discharge, and unipolar depressive disorder in individuals with a family history of mania. However, the model failed to identify any of the patients who committed suicide. The results appear to support the contention that, based on present knowledge, it is not possible to predict suicide, even among a high-risk group of inpatients.

Complicated mania. Comorbidity and immediate outcome in the treatment of mania.

In a case-control study, 57 manics with antecedent or coexisting nonaffective psychiatric disorders (n = 38) or serious medical illnesses (n = 19) ("complicated mania") were compared with 114 age-, sex-, and year-of-admission-matched controls with no other disorder ("uncomplicated mania"). Significant differences emerged between the three groups in age, marital status, age at onset, number of prior hospitalizations and prior suicide attempts, organic features, and outcome measures (recovery and death rates). Patients were divided into four treatment groups based on primary mode of therapy during index admission; the groups included electroconvulsive therapy, adequate lithium carbonate, inadequate lithium carbonate, and neither treatment. Uncomplicated manics were significantly more likely to receive adequate lithium carbonate and less likely to receive inadequate lithium carbonate than were complicated manics. The latter patients had a significantly poorer immediate response to treatment overall, and to adequate lithium carbonate specifically. Seventy-eight (68.4%) uncomplicated manics had recovered ad discharge, compared with 26 (45.6%) complicated manics. Logistic regression suggested that the influence of comorbidity on outcome was more important for women than men. We conclude that complicated mania is a useful clinical construct.

Stability of psychiatric diagnosis. Schizophrenia and affective disorders followed up over a 30- to 40-year period.

Stability of diagnosis in schizophrenia and affective disorders is high. Patients selected according to research criteria for schizophrenia and affective disorders were followed up in a historical prospective study to evaluate their diagnostic stability over a 30- to 40-year period. Our follow-up, lifetime diagnosis, based on blind diagnostic assessment, showed that 92.5% of the personally interviewed schizophrenics were given the follow-up diagnosis of schizophrenia, which was significantly higher than the 78.3% found in affective disorders. However, no significant difference exists when assessment was made on available records of those who died or refused to be interviewed. These stability coefficients are discussed in light of the methodological assumptions involved in a long-term follow-up study. We concluded that the diagnostic stability in schizophrenia and affective disorders were very high and that rigorous diagnostic criteria should be maintained.

The family history method using diagnostic criteria. Reliability and validity.

Data concerning familial history of psychiatric disorders are often used to assist in diagnosis, to examine the role of genetic or nongenetic familial factors in etiology, or to develop new methods of classification. Information concerning familial prevalence may be collected by two different methods: the family history method (obtaining information from the patinet or a relative concerning all family members), and the family study method (interviewing directly as many relatives as possible concerning their own present or past symptomatology). This study compares these two methods. In general, the family study method is preferred since information is likely to be more accurate. The family history method leads to significant underreporting, but this can be minimized through the use of diagnostic criteria. This study reports on an instrument that has been developed for collecting information concerning family history and that provides criteria for 12 diagnoses--the Family History-Research Diagnostic Criteria. Using diagnostic criteria leads to greater sensitivity, but underreporting remains a major problem of the family history method.