Cohort study of ageing from Bagé (SIGa-Bagé), Brazil: profile and methodology.

BACKGROUND

The Bagé Cohort Study of Ageing is a population-based cohort study that has recently completed the first follow-up of a representative sample of older adults from Bagé, a city with more than 100,000 inhabitants located in the state of Rio Grande do Sul, Brazil. This is one of the first longitudinal studies to assess the impact of primary health care coverage on health conditions and inequalities. Our aim is to investigate the prevalence, incidence and trends of risk factors, health behaviours, social relationships, non-communicable diseases, geriatric diseases and disorders, hospitalisation, self-perceived health, and all-cause and specific-cause mortality. In addition, we aim to evaluate socioeconomic and health inequalities and the impact of primary health care on the outcomes under study.

METHODS/DESIGN: The study covers participants aged 60 or over, selected by probabilistic (representative) sampling of the urban area of the city of Bagé, which is covered by Primary Health Care Services. The baseline examination included 1593 older adults and was conducted from July 2008 to November 2008. After eight to nine years (2016/2017), the first follow-up was conducted from September 2016 to August 2017. All participants underwent an extensive core assessment programme including structured interviews, questionnaires, cognitive testing (baseline and follow-up), physical examinations and anthropometric measurements (follow-up).

RESULTS

Of the original participants, 1395 (87.6%) were located for follow-up: 757 elderly individuals (47.5%) were re-interviewed, but losses in data transfer occurred for 22. The remaining 638 (40.1%) had died. In addition, we had 81 (5.1%) refusals and 117 (7.3%) losses. Among the 1373 older adults who were followed down, there was a higher proportion of female interviewees (p=0.042) and a higher proportion of male deaths (p=0.001) in 2016/2017. There were no differences in losses and refusals according to gender (p=0.102). There was a difference in average age between the interviewees (68.8 years; SD ±6.5) and non-interviewees (73.2 years; SD ±9.0) (p<0.001). Data are available at the Department of Social Medicine in Federal University of Pelotas, Rio Grande do Sul, Brazil, for any collaboration.