DETERMINANTS OF CARER STRESS IN ALZHEIMERS-DISEASE

Objectives. To explore the impact of subgroups and individual symptoms of non-cognitive disturbance on the carers of Alzheimer's disease pat ients. Design. Cross-sectional study using clinically valid scales to assess patient symptomatology and self-report questionnaires to measur e carer variables. Setting. Old age psychiatry outreach services in So uth and Central Manchester.Subjects. 100 patients with Alzheimer's dis ease living at home and their carers. Main outcome measures. Subjectiv e burden and distress in carers. Results. Separate statistical analyse s were performed for subgroups and individual symptoms of non-cognitiv e disturbance. For subgroups, multivariate analyses identified depress ion and behavioural disturbances in patients as significant predictors of subjective burden in carers. Carer distress was predicted by depre ssion, psychosis and cognitive impairments in patients and carer gende r. For individual symptoms of non-cognitive disturbance, three feature s of depression in patients (mood-related signs, physical signs and be haviour changes), walking disruptions and the patient-carer relationsh ip predicted of subjective burden in carers. Variance in the level of carer distress was accounted for by sleep disruptions, hallucinations and mood-related depressive features in patients and carer gender. Con clusion. The findings confirm that the non-cognitive features of Alzhe imer's disease are stressful for carers and indicate specific relation ships between mood-related and behavioural signs of depression, walkin g and sleep disruptions and hallucinations in patients and adverse car er outcomes. Patient depression and the mood-related signs of depressi on in particular were the most consistent and powerful predictors of p sychological morbidity in carers. Intervention strategies need to iden tify and target troublesome behaviours in patients and aim to either c hange these behaviours or alter the way carers respond to them. Thus, interventions need to be symptoms-rather than services-led and are lik ely to require multidisciplinary and multi-agency approaches. (C) 1998 John Wiley & Sons, Ltd.