| Summary: | The aim of this thesis was to study of the validity of validity of different frailty phenotypes among community-dwelling older people aged 65+ in northwest Russia, and to develop a risk-prediction model for adverse outcomes within this target population. The data for my PhD project were obtained from the prospective cohort Crystal study of community-dwelling individuals aged 65 and older living in the Kolpino district of St. Petersburg. Initially, a validation study was conducted to evaluate and compare three frailty screening instruments when applied to this population: the frailty phenotype model (the Fried model); the cumulative deficit approach (the Puts model); and the 15-item Groningen Frailty Indicator, a self-assessment questionnaire (the Steverink-Slaets model). As results, none of these models was valid for predicting 5-year mortality and disability rates, or 2.5-year mental and physical decline rates. Further research was performed to identify valid markers to predict adverse outcomes in this population. We found that the classical cardiovascular risk factors lose their predictive power beyond the age of 65 in Russian population. We discovered that poor physical performance is the only factor that was associated with a loss of autonomy and mental decline in our population aged 65 years and older. A new Crystal risk model was developed. Older adults were identified at higher risk when two or more of the following components are present: poor physical performance, low muscle mass, poor lung function, anemia (model 1) or poor physical performance, low muscles mass, poor lung function, and a combination of anemia with high C-reactive protein levels and high B-type natriuretic peptide levels (model 2). Both risk scores were externally validated in a second cohort of adults 80 years and older. Model 2 was associated with a higher risk of cardiovascular mortality and had a better predictive capacity compared with model 1. These risk prediction models might help patients and physician to make decisions about ...
|
|---|