The present study formed part of a larger one designed to compare the mental health of people living on a new housing estate in Croydon with that of people living in an older district of the town (Hare and Shaw, 1965). In the course of the study, information was obtained by interviewers and from general practitioner records on the health of nearly 500 families in which there was at least one parent living with one or more children (under 16 years). We were able to examine the health of the parents and their children in relation to various family factors, and it is the aim of the present paper to examine the association between health and family size.

This paper deals with information obtained from nearly five hundred families in which at least one parent was living with one or more children aged 15 or under. It examines the association between the health of fathers and mothers, and also between that of parents and children. Elsewhere (Hare and Shaw, 1965a) we showed that, in the sample population from which these families came, there was a strong tendency for persons with poor mental health to have poor physical health as well.∗ We took this to support the view that in any population there tends to be a group of persons, comprising some 10 to 15 per cent, of the whole, who are particularly prone to ill-health of all kinds and who make a correspondingly high call on the facilities of the health service. One of the objects of the present study was to try and determine if there were families, as well as persons, prone to ill-health—in other words, whether a high rate of ill-health in (say) fathers was associated with high rates in other members of the family.

The development of student health services in colleges and universities has led, during the past decade, to a growing awareness of the importance of mental health problems in student populations (see Davy, 1960; Malleson, 1963).

The Disabled Persons (Employment) Act of 1944 set up a national register of disabled persons and facilities for their industrial rehabilitation. The Industrial Rehabilitation Units of the Ministry of Labour which were provided for by this Act were described and their functions evaluated in the Piercy Report (1956). The broad scope of present concepts of rehabilitation is fully reviewed by Vaughan Jones (1961) and will not be discussed here.

Depression is the commonest of mental illnesses and fortunately it has a better outcome with treatment than any other psychiatric condition. It shows itself in a variety of guises, and as there is no universal formula for treating the different types, a classification is essential. During the past ten years the arrival of the tranquillizers and particularly antidepressants has revolutionized our view of depressive illness, yet we are still using terms which are purely descriptive, determined by administrative needs long forgotten, or loosely related to aetiology about which we still know little. The time has come to link the classification of depression with rational treatment, but before offering a scheme to do this, a brief consideration of the inadequacies of existing terminology may help to explain the need for a new scheme.

In the British Journal of Psychiatry for July, 1963, Costello and Smith described an investigation based on the prediction that “… reactive depressions would sleep more at the end than at the beginning of a set period of sleep and that endogenous depressions would sleep more at the beginning than at the end of a set period for sleep …”. Hourly observation by nurses of the sleep of 18 patients with endogenous and 41 with reactive depression failed to confirm the prediction.

The clinician is deeply concerned with the difficulties of selecting a treatment method that will accomplish the goals which he wishes to achieve for a particular patient with specific symptoms and characteristics in a particular treatment setting, knowing that it is unlikely that any one single treatment will be universally and equally effective in all situations. The research investigator should attempt to help him by furnishing objective evidence as to the specific efficacies and limitations of methods of treatment in common use under the conditions that prevail where large numbers of patients are treated—in public mental hospitals. Adequate comparative information obtained under controlled experimental conditions is essential not only to therapists but also to patients and their families and to those responsible for the support and administration of treatment centres.

Desipramine (Pertofran) is a metabolite of imipramine (Tofranil) originally isolated by Herrmann, Schindler and Pulver (1959) and found to be identical with the synthetic compound G.35025. Animal experiments (Gillette et al., 1961) suggest a more rapid action for desipramine, and the majority of clinical trials to date confirm this impression for humans (e.g. Kline et al., 1962; Krakowski, 1963). Some controlled trials, however, (Azima et al., 1962; Hollister et al., 1963) fail to establish that desipramine is either faster acting or more effective than imipramine.

Short acting barbiturates such as amylobarbitone have long been the standby for relief of anxiety—certainly any new drugs introduced for the relief of such symptoms must stand comparison with these. Furthermore such symptoms are well known to be modifiable by such variables as the attitude of the doctor. For this reason a trial of such a new drug must be carried out under blind conditions.

Oxazepam, Wy 3498 (John Wyeth, Brother Ltd.) is a benzodiazepine derivative-7-chloro-1, 3-dihydro-3-hydroxy-5-phenyl-2H-1, 4-ben-zodiazepin-2-one. The structural formula is as follows:Other Benzodiazepin derivatives in current use are chlordiazepoxide (Librium) and diazepam (Valium).

Oxypertine is a new piperazine derivative, believed to have tranquillizing, antiadrenergic and hypotensive properties. Unpublished reports have suggested that it might be useful in chronic schizophrenia, although the recent study by Calwell et al. (1964) shows that its usefulness in this condition is limited. In view of its putative tranquillizing effect, the present clinical trial was undertaken to determine, under double-blind conditions, its efficacy as an adjunctive treatment for chronic anxiety neurosis. Ratings of anxiety symptoms were the major criteria for assessing the patients' progress, but the IPAT Anxiety Scale (Cattell and Scheier, 1957) was also included as a potentially useful measure of anxiety. An investigation of its validity was a subsidiary aim of the study.

The discovery that N-allylnormorphine was a specific antagonist to morphine gave hope that an anti-barbiturate agent might be discovered. Shaw et al. (11) noted that in dogs, BB-methyl-ethylglutarimide (bemegride) dispelled the EEG changes induced by barbituric acid, and suggested that it may act as a barbiturate antagonist. Subsequent studies by Louw and Sonne (5) and Pederson (9) did not lend support to this theory, and Peacock (8) could find no specific pattern of EEG changes which could be attributed to the administration of bemegride. Ackner and Pampiglione (1) concluded that the demonstration of persistence of barbiturate-fast activity in the EEG after arousal did not support the hypothesis of a direct pharmacological antagonism, and suggested that bemegride was not a direct stimulant but caused an increased responsiveness in the patient.

Out-patient drug trials have relied largely on observer-rated improvement measures to evaluate the efficacy of psychotropic medication (4). Less frequent use (5, 7, 9) has been made of patient ratings or the combination of observer plus patient ratings. Thus a comparison of improvement ratings made by the patient and his treating doctor would be of considerable methodological interest.

There has recently been an emphasis on creating a therapeutic milieu in mental hospitals. Numerous publications reflect this trend (1, 2, 5, 6, 7, 8, 9, 11, 14, 15). In addition, studies on attitudes of nurses and patients and on some of the factors which seem important in creating a favourable ward atmosphere have become commonplace (3, 4, 10, 12, 13, 16, 17).

Rating scales are used a great deal, particularly in psychiatric medicine, for the assessment of severity of a patient's condition, to assist with diagnosis and classification, and for the comparison of results of treatments. Much time and effort is expended in the construction of such scales, in analysis of their meaning and in verification of their reliability. However, even the best scale gives unreliable results if it is not used correctly.