Balance Assessment in Community-Dwelling Elderly:
Fallers vs. Non-fallers

Nine million seniors chronically fall with 300,000 hip fractures annually while 25% of these are followed by death within a year. Falls are thought to be linked to an individual’s ability to maintain postural stability which deteriorates with increasing age [2]. Predicting falls is a very challenging task and most studies to date can only distinguish between present fallers and non-fallers. Current literature suggests “new screening tests are needed for community-dwelling older adults who are active” making fall risk prediction in active older adults possible. By means of traditional variables like peak to peak sway or RMS of the COP path susceptibility for falls cannot be predicted. More recent analysis (e.g. Hidden Markov Models) puts more emphasis on reemerging patterns of the distribution path indicating that aging and/or pathology affects variability and therefore increases risk of falling.

research

The purpose of this study was to evaluate postural balance in community-dwelling elderly. Fallers and non-fallers were compared with respect to traditional variables like peak to peak sway in AP and ML direction and number of hidden equilibria in the distribution path obtained via use of Hidden Markov Models. This study was conducted with 22 subjects (18 females, 4 males) recruited from an independent senior living community in Houston. An AMTI force plate was used to collect the data during 4 different conditions (bipedal eyes open, bipedal eyes closed, unipedal eyes open, unipedal eyes closed).

research

During all trials participants were instructed to stand as still as possible with hands on their hips for 60sec and 20sec, respectively. A painting on a wall approx. 2m in front of the participants was used as visual cue. Finally a spotter standing right behind the participants secured them in case of a potential fall.

Participants performed three trials for each condition and were able to rest in between trials at liberty. The average age was 85.7 ± 4.6 years, mean height was 163.77 ± 8.79cm with an average mass of 66.69 ± 13.21kg. No differences with respect to peak to peak sway and RMS between fallers and non-fallers were detected.

Although older participants were less stable than their younger peers this trend was not significant. However, analysis of the distribution path via Hidden Markov Models might reveal a higher number of hidden equilibria in the eyes closed vs. eyes open conditions as well as in the fallers vs. non-fallers. Additional analysis of the center of pressure distribution path via Hidden Markov Models might serve as a useful tool to detect disruptions in postural control at an early stage hence making prediction of susceptibility for falls more likely.