Development of a Fried Frailty Phenotype Questionnaire for Use in Screening Community-Dwelling Older Adults

Abstract

Objectives: The aim of this study was to develop a Fried Frailty Phenotype Questionnaire (FFPQ) and a Japanese FRAIL (fatigue, resistance, ambulation, illnesses, and loss of weight) scale (FRAIL-J) and to evaluate the reliability and validity of both questionnaires in Japanese community-dwelling older adults.

Design: Cross-sectional analysis of Itoshima Frail Study (IFS).

Setting: The IFS is an ongoing community-based prospective study in Itoshima (Japan).

Participants: A total of 858 older adults age 65-75 years.

Methods: The FRAIL-J comprises 5 existing items comparable to those in original FRAIL scale but with broader utilization in the Japanese healthcare system. In FFPQ, resistance, ambulation, and loss of weight were the same as those in FRAIL-J. Fatigue was the same with exhaustion in FFP and inactivity was assessed using a yes or no question. Data including demographics, and physical and cognitive functions, and objective physical activity was collected and analyzed in relation to both questionnaires.

Results: The FFPQ and FRAIL-J showed low internal consistency (Kuder-Richardson formula 20 coefficients = 0.32 and 0.29) and good test-retest reliability (intraclass correlation coefficients = 0.79 and 0.72). The correlations ranged from -0.22 to 0.49 when correlating each item with cross-sectional outcomes. Using FFP as a criterion, the ares under the curve for FRAIL-J and FFPQ were 0.86 and 0.88, respectively. The optimal cut-off for FRAIL-J was 2, with a higher Youden index (66.7% vs 20.3% for 3) and a high negative predictive value (99.5%) but low positive predictive value (13.1%). As for FFPQ, either 2 or 3 was evaluated as cut-off because the Youden index (62.2% vs 58.5%) and negative predictive value (99.7% vs 99.2%) were similar although the positive predictive value was low (9.7% vs 33.3%). Using a 2-point cut-off, both questionnaires had slight agreement with FFP. The highest agreement (kappa = 0.42) was found between FFP and FFPQ using a 3-point cut-off.

Conclusions/implications: The FFPQ and FRAIL-J can be used for frailty screening in Japanese community-dwelling older adults.

Keywords: FRAIL; Fried Frailty Phenotype; Reliability; frailty; older adults; validity.