Introduction: Frailty is a state of increased vulnerability; its prevalence amongst people over 85 years may be as high as 26% but individuals classified as frail differ considerably in terms of their function and clinical outcomes. An important healthcare challenge is to better discriminate between frail people who are expected to do less well to facilitate targeted intervention and prognostication. This study aimed to investigate the utility of physical performance testing in frail older people and ascertain relationships with all-cause mortality.
Methods: 103 older women were followed up at home after original recruitment to the study during an acute hospital admission six months previously. Demographic information was collected and Fried frailty status was defined as three of: weight loss, weakness, exhaustion, slowness, low activity (2 = pre-frail). Physical performance was assessed using timed-up-and-go (TUG), gait speed, one-legged stand and chair rises. Mortality data was ascertained over the subsequent 18 months.
Results: Average age was 87 years; 70% were frail and 17% were pre-frail. 19% of participants attempted all four physical performance tests; 25% were unable to attempt any. Only TUG predicted mortality over 18 months (hazard ratio per second faster, 0.99; age-adjusted p-value, 0.06); there were no associations with other functional assessments or frailty status. A physical performance score was generated according to the number of assessments attempted (0–4); this was the biggest predictor of long-term mortality (hazard ratio per point gained, 0.25; age-adjusted p-value, 0.008).
Conclusions: Frailty was highly prevalent in community dwelling older women six months after discharge from hospital and was not a predictor of mortality in the next 18 months. Physical performance assessments were difficult for the study participants to complete. The simple physical performance score assessed ability and willingness to complete physical performance tests and was a powerful predictor of mortality.