Validity and Responsiveness of the FRAIL Scale in Middle-Aged Women

To assess the validity and responsiveness of the FRAIL scale in middle-aged women, who are a group at high risk of developing frailty.

Longitudinal cohort study from 1998 to 2013.

Australia.

10,412 women born in 1946-1951 from the Australian Longitudinal Study on Women's Health (ALSWH).

Frailty was measured by the 5-item FRAIL scale and assessed every 3 years from 1998 to 2013. Face validity was examined by assessing relationships with age. Spearman correlation quantified the relation of each item of the FRAIL scale with the scale's total score. Adjusted logistic regression models assessed the construct validity of frailty in 1998 predicting depression (10-item Center for Epidemiologic Studies Depression Scale Short Form) and disability (needing help with daily tasks) in 2013. Mortality was recorded from 1998 up to December 31, 2014. Survival analysis was done using Cox proportional hazards models. Finally, responsiveness was examined by measuring the relationship between changes in self-rated health and changes in FRAIL score between 2 subsequent surveys.

Frailty increased with age such that 5.8% of women were frail at age 50 and 11.3% at age 66. Each component of the FRAIL scale was correlated with the total FRAIL score (rho = 0.13-0.82, all P < .001). Compared to being healthy, women who were frail in 1998 had an increased likelihood [odds ratio (95% confidence interval [CI])] of being depressed [2.77 (2.12, 3.63)] or disabled [6.87 (4.84, 9.77)] in 2013 with a hazard ratio (95% CI) for death of 2.01 (1.40, 2.87). Having a deficit in each of the 5 items in 1998 also increased the likelihood of being depressed or disabled in 2013. Changes in self-rated health were associated with total FRAIL score changes.

The FRAIL scale is valid for use in longitudinal studies of middle-aged women. Studying trajectories of frailty from the middle-age to older-age population may yield insights into risk factors for poorer health in this population.