Cohort differences in dependency in activities of daily living, self-perceived health and GP contact in older people over a seven-year timespan

Total populations aged 75 years an over, drawn from the age-sex register of the General Practice for Melton Mowbray in 1981 and 1988, were surveyed on health and sociodemographics by trained fieldworkers. For this study, participants were subdivided into seven year birth cohorts, born 1885-1891, 1892-1898, 1899-1905 and 1906-1912. Age-cohort and age-period logistic regression models were fitted to quantify trends in dependency (defined by human or technological help in at least one of: mobility about the home getting in and out of bed getting in and out of a chair dressing and eating), in self-perceived health defined by response to the question "For your age, in general, would you say that your health was good, fair or poor", and in self-reported GP contact in the home or surgery in the past month. Five-year survival and predictors of GP contact were also modelled. With more than 90% response, numbers interviewed at home were 1124 in 1981 and 1500 in 1988. Prevalence of dependency in the newer cohorts was slightly reduced compared to earlier cohorts surveyed at comparable age, but perceptions of health were less favourable in the newer cohorts, this difference being statistically significant for women. Less than good self-perceived health was confirmed as a predictor of five-year mortality in people aged 75-81 years, and an independent predictor of GP contact. For men, each succeeding cohort had higher rates of GP contact, whilst rates were stable for women. Associations of self-perceived health with GP contact were strongest for men aged 75-81 years in 1988, these men also having higher risk of contact and worse 5-year survival compared to men aged 75-81 years in 1981.