Comparing Outcome Measures for Risk of Falls in Older People with Comorbidities.

This longitudinal clinical study aimed to investigate the predictive value of several outcome measures in the fall risk of older people with comorbidities

This study was part of the Smart Bear research project, a large-scale, multi-center clinical study funded by the EU (HORIZON 2020, No:857172). 

Inclusion criteria were: a) age: 65 - 80 years, b) medical history: at least 2 of the following conditions: cardiovascular disease (CVD), hearing loss, balance disorders, mild depression, mild cognitive impairment (MCI) or frailty, c) adequate cognitive function (Montreal Assessment Scale-MoCA>18/30), d) adequate mobility level.  Εxclusion criteria were: a) severe or life-threatening condition, b) deterioration of pre-existing medical condition, and c) unwillingness to cooperate.

 Participants were assessed at the baseline with the following outcome measures: a) the Falls Efficacy Scale - International (FES-I) full version, b) the Activity-specific Balance Confidence (ABC) Scale, c) the Functional Gait Assessment (FGA) test, d) the miniBESTest, and e) the Rapid Assessment of Physical Activity (RAPA). Their falls were reported over six months through the Smart Bear platform. During this period, none of the participants was involved in a balance rehabilitation program. 

The Receiver Operating Characteristic (ROC) analysis was performed using IBM SPSS 29.0, calculating the Area Under the Curve (AUC), the cut-off points, sensitivity, and specificity of each outcome measure. The data were obtained from the clinical site of the Municipality of Palaio Faliro, a clinical partner of the Smart Bear research project.  

Ninety (90) participants included in the study [62/90 females; 34/90 non-smokers; mean age:72.46, 95%CI:(71.52-73.35) years; BMI:30.26, 95%CI:(29.20-31.35); MoCA: 23.84, 95%CI (23,37-24,27)], all of whom reported balance disorder at baseline, and diagnosed with at least one of the other medical conditions (34/90 hearing loss, 65/90 CVD, 70/90 MCI, 22/90 mild depression, 6/90 frailty). 

Over six months, 29/90 participants reported a fall. In the ROC analysis, the FES-I questionnaire recorded the best accuracy in identifying future falls (AUC: 0.731, 95%CI:0.621-0.840, p0.001), followed by the ABC questionnaire (AUC: 0.683, 95%CI:0.565-0.801, p=0.002), the miniBESTest (AUC: 0.673,  95%CI:0.562-0.784, p=0.002) and the FGA test (AUC: 0.668, 95%CI: 0.557-0.780, p=0.003). The RAPA questionnaire failed to have better accuracy than flipping a coin (AUC: 0.517, 95%CI:0.394-0.641, p=0.784). 

At a cut-off point of 22/64, the FES-I demonstrated the highest sensitivity (83%) and specificity (57%) compared to the other outcome measures.

In older people with comorbidities, including balance disorder, assessing fear of falling using a self-reported questionnaire (FES-I) demonstrates superior predictive properties in detecting the risk of falls compared to other outcome measures.

Identifying the accuracy of outcome measures in predicting fall risk in older people with comorbidities will help identify high-risk individuals and allow personalized, early intervention.