In a study of 59 elderly medically ill in-patients, 35% were found to have significant depressive symptomatology, as detected by the Geriatric Mental Status Schedule (GMSS). Of two screening methods, the Geriatric Depression Scale (GDS) performed satisfactorily but detection by nurses was poor. Acknowledgement of depression in medical casenotes was low. Training of nurses might improve detection; otherwise a mood-rating scale such as the GDS should be incorporated into routine practice.

SPECT is effective in the diagnosis of neuropsychiatric disorders such as OCD, and the pathological changes in brain metabolism detected by SPECT may be reversed by both drug therapy and psychotherapy.

Consultant in Old AgeThree areas relevant to prognosis are briefly reviewed. They concern recent outcome studies; some prognostic factors associated with outcome; and finally some suggestions, based on recent findings, for improving the prognosis.First some preliminary remarks. The topic here is depressive illness in the elderly, the operational definition of which is summarized under DSM-III-R (APA, 1987) criteria for major depressive disorder (MDE). Next, patient selection will inevitably influence outcome and most studies concern inpatients. What constitutes an adequate follow-up period? The pre-ECT literature-the nearest there is to a true natural history of depression-tells us that roughly a half of severely depressed patients recovered spontaneously, although often only after long periods of morbidity. Summarizing three studies, I have found (Baldwin, 1988a) that the lowest mean time to recovery was 18 months and the longest about four years. OUTCOME STUDIES Different outcome criteria are used in different studies, which makes comparisons difficult. The number of episodes over a given period, the interval between episodes, from which the 'morbid time period' can be derived, course of symptoms, pattern of symptoms, mortality rates and social or occupational functioning are all measures of outcome. Social recovery is an important but neglected area.Using standard treatments around 75-80% of patients with depressive illness will recover from an individual episode (Baldwin and Jolley, 1986). The longer term outlook is less rosy anci, for all ages, a note of pessimism has recently crept in (Lee and Murray, 1988;Murphy, 1983).The classic studies of Post (1962Post ( , 1972 that around 60% of elderly depressed patients requiring hospitalization had, after up to six years' follow-up, either remained well or developed relapses followed by full recovery. Interestingly, the advent of tricyclic antidepressants between these two studies appeared to make little difference, although selective entry of more 'difficult' cases into the later study may have been responsible. Recently two further studies, one of inpatients over an average of five years (Baldwin and Jolley, 1986) and the other of ECT-treated patients over three years (Godber et al., 1987), have been carried out. I have attempted to compare the results of these four studies (Table 1). The assessment of outcome is broadly similar, being a measure of the course of symptoms. Depressive invalidism refers to partial recovery, with or without further depressive episodes, but with residual neurotic symptomatology. The results are surprisingly similar and allow reasonably robust deductions about the clinical course of old age depressive illness. One of the few studies which combined a prospective approach with modern research instruments is that of Murphy (1983). Her study included community patients as well as inpatients. Only 44 patients out of a cohort of 124 were well at one year. When the outcome groups were dichotomized into 'good' and 'poor' only 35% were assigne...

This paper presents evidence documenting risks for affective problems and disorders among minor children of alcoholics. Subjects studied were 98 youths from alcoholic-headed and control families participating in a risk for alcoholism research project (Johnson & Rolf, see next paper in this issue). Depressive affect questionnaires were completed by the mothers of the youths as well as the youths themselves. ANO VA's and ANCO VA's (covarying for age) indicated significant group differences on both maternal and youth ratings with more affective problems shown for the offspring of alcoholics. Outliers with extreme depressive affect scores occurred most frequently for the youths in the high risk for alcoholism group. Findings are discussed with regard to risk group membership, gender, source of information, and the data's implications for further analyses as well as/or the design of future studies.