OBJECTIVETo assess variables associated with different perceived global health ranks in an older population.

MATERIALS AND METHODSThe Blue Mountains Eye Study was a populationbased survey of residents aged >or=49 years in an area west of Sydney, Australia. Questions relating to demographic and socioeconomic indicators, limitations of daily living activities, medical history, hearing problems, exercise, smoking and alcohol consumption were asked in a standardised questionnaire at interview. Comprehensive eye examinations were performed. Self-rated health was assessed from a standard question and categorised as excellent, good, fair or poor. Associations were assessed using logistic regression models.

RESULTSAfter exclusions, 714/3589 (19.9%) participants rated their health as excellent, 1969 (54.9%) as good, 766 (21.3%) as fair and 140 (3.9%) as poor. A wide array of study variables assessed showed significant associations with self-rated health. Compared with a self-ranking of good or excellent health, variables significantly associated in a multivariate model with a rank of poor health included: regular use of community support services, more than one hospital admission in the last 12 months, perceived inability to go out alone, difficulty walking, or history of angina, asthma or cancer. Variables significantly associated with a rank of fair health were broader and included socioeconomic status, sensory impairment, tinnitus, many chronic diseases and negative health behaviours, including smoking and heavy alcohol consumption.

CONCLUSIONSOur findings suggest that different sets of variables may influence people to rank their global health at different perceived levels.