Associations between diabetes, body mass index and frailty: The Western Australian Health In Men Study

Objectives: To investigate whether diabetes and obesity are associated with frailty independently, and to determine the proportion of frailty cases attributable to each factor. Study design: Prospective cohort study of 4219 older men assessed in 2001–04 (time-point 1, T1), of whom 1939 were reassessed in 2008–09 (time-point 2, T2). Frailty was defined as positive responses on three or more of the five domains on the FRAIL scale: fatigue, difficulty climbing a flight of stairs (resistance), difficulty walking 100 m (ambulation), >5 illnesses, or >5% weight loss. We explored associations of diabetes and obesity with frailty using binary logistic regression, and estimated population attributable fractions for diabetes and obesity as risk factors for frailty. Main outcome measures: Associations of obesity and diabetes with frailty. Results: At T1, 15.5% of participants (n = 652) were frail, 15.4% (n = 651) had diabetes, and 15.1% (n = 636) were obese (BMI ≥ 30 kg/m²). At T2, 22.9% (n = 444) were frail, 19.8% (n = 383) had diabetes and 9.1% (n = 176) were obese. In multivariable models, diabetes was associated with frailty at T1 (odds ratio [OR] = 2.1, 95% confidence interval [CI] = 1.7–2.6) and T2 (OR = 2.1, CI = 1.6–2.8). Obesity was associated with frailty at T1 (OR = 1.7, CI = 1.3–2.2) only. Both diabetes (OR = 1.5, CI = 1.1–2.2) and obesity (OR = 1.9, CI = 1.3–2.9) at T1 were independently associated with the development of frailty between T1 and T2. Approximately the same proportion (5%) of new frailty cases were attributable to diabetes or to obesity. Conclusions: Diabetes and obesity are modifiable risk factors which independently carry equal risk for the development of frailty in older men. Interventions targeting these risk factors may have the potential to reduce frailty risk.