Predicting Value of Nine-Item Berg Balance Scale Among the Aged: A 3-Year Prospective Follow-Up Study

The aim was to assess the predicting value of the nine-item Berg Balance Scale (BBS-9) for falls among the community-dwelling aged.

The participants (N = 298) were divided according to BBS-9 (range: 0-36) into those scoring 0-32 sum points (risk group [RG]; n = 158) and those scoring 33-36 (non-risk group [non-RG]; n = 140). Falls were recorded by fall diaries, which subjects were asked to mail to the research assistants (RAs) monthly during the 12-month follow-up. The occurrences of falls requiring treatment were collected from the health center and hospital registers during the 12- and 36-month follow-ups.

During the 12-month follow-up, 271 falls (171 in RG and 100 in non-RG) and 29 falls requiring treatment (22 in RG and 7 in non-RG) occurred. During 36 months, there were 98 falls that required treatment (72 in RG and 26 in non-RG). The incidence of falls was higher in RG compared with non-RG (incidence rate ratio [IRR]: 1.57, 95% confidence interval [CI]: 1.23-2.01) during the 12-month follow-up (p < .001). Also, the incidence of falls requiring treatment was significantly higher in RG than in non-RG during 12 months (IRR: 2.82, 95% CI: 1.20-6.59; p = .017) and 36 months (IRR: 2.56, 95% CI: 1.63-4.01; p < .001).

BBS-9 with a cutoff score of 32/33 is an applicable tool for predicting risk of falls among the community-dwelling aged. Future studies are needed to assess the predicting value of BBS-9 among different age groups in the elderly population.