The risk for nursing home admission (NHA) did not change in ten years--a prospective cohort study with five-year follow-up.

In an aging population the burden on health care systems depends on the proportion of lifetime spent in good or poor health. The objective of this study was to examine the effect of a ten-year cohort difference on NHA, indicating changes in lifetime spent without severe disabilities. Additionally, important risk factors for NHA were identified. The data were obtained from two cohort studies of elderly people. Cohort A (1991-1993) comprised 74+ patients from 20 German general practices and cohort B (2002-2003) 70+ patients from 14 general practices. The merged sample consisted of 2301 community dwelling patients that contacted their general practitioner within a 12-month period during the respective enquiry period. After an initial assessment at study entry, participants were monitored over a five-year period respective NHA and death. The Cox proportional hazards model was used including socio-epidemic data, state of health, chronic diseases, dementia, health system usage, and social support. The ten-year cohort-difference was no predictor of NHA within a 5-year period. Significant influencing variables were: age (OR 1.10), living with others (OR 0.59), no auxiliary person (OR 1.69), mild forgetfulness (OR 2.12), clear cognitive impairment (OR 3.74), severe cognitive disturbance (3.61), loss of memory (11.83), walking difficulties (OR 1.53), impaired vision (OR 1.90), and cancer (OR 0.22). This study could not find a cohort effect on NHA. With regard to increased life expectancy the findings do not support the compression of morbidity hypothesis. The identified influencing variables contribute to the understanding of NHA risk factors.