Timed Up and Go Test can predict recurrent falls: a longitudinal study of the community-dwelling elderly in China

Abstract

Purpose

Falling is a major health problem in community-dwelling elderly individuals. The aim of the present study was to conduct a prospective investigation to evaluate the accuracy of the Timed Up and Go Test (TUGT), 4-meter walking test, and grip strength test to screen for the risk of falls and to determine a cutoff point to be used clinically.

Patients and methods

This was a prospective study that included 541 participants. The fall data were obtained via face-to-face interview, and the date, site, and circumstances of any falls were recorded. TUGTs were recorded as part of a comprehensive geriatric assessment. We collected the same data at baseline and after follow-up via comprehensive geriatric assessment.

Results

The incidence of falls of our study subjects was 20.8%. The recurrent-fall group had a fall rate of 6.8% during the follow-up year. The standard area under the curve (AUC) of our screening tool was >0.70, and hence our tool can be used for clinical purposes. After adjusting for age and gender, the AUC of TUGT became 0.642, so it cannot be used as a predictive tool for measuring any types of falls. However, when recurrent falls were adjusted for age and gender, the TUGT’s AUC improved to 0.733 and a score of 15.96 seconds is used as a cut-point to screen recurrent falls in community-dwelling elderly Chinese individuals.

Conclusion

Future falls were best predicted by TUGT in recurrent fallers at baseline. A score of 15.96 seconds is used as a cut-point to screen recurrent falls in community-dwelling elderly Chinese individuals.

Keywords:

• community dwelling
• elderly
• fall
• TUGT

Acknowledgments

The authors thank Peipei Han from the Department of Rehabilitation Medicine, Tianjin Medical University, for processing data and writing documents. This work was supported by the National Natural Science Foundation of China (81372118) and Tianjin Municipal Science and Technology Commission (grant number 16ZXMJSY00070).

Disclosure

The authors report no conflicts of interest in this work.