Abstract
Falling is the leading cause of unintentional injury and death among older Americans (CDC, 2022). However, current fall prevention/prediction programs and research address “fall risk factors'' without distinguishing risk from difficulty, which is not ideal because the treatments targeting risk factors may be different from those targeting difficulty factors. In other fields, risk is defined as the product of the likelihood and consequence of adverse events, while difficulty is the strain on internal sensorimotor capacity. I have employed a Visual Inverted Pendulum (VIP) balancing paradigm which permits experimental manipulation of risk and difficulty independently of each other. My paradigm requires participants to use a joystick to balance an animated VIP upright, within invisible fall boundaries. Joystick manual control of a VIP circumvents the reflex-like processes that normally contribute to human upright stance, which has been modeled as balancing a biomechanical inverted pendulum, thereby unmasking a potentially overlapping cognitive mechanism sensitive to risk. In two studies (EXPERIMENTS 1 & 3), participants performed a cumulative 900 seconds of balancing on each of the two consecutive days. On Day 1 (D1) of EXPERIMENT 1, the fall boundaries were at ±90°, and the number of falls and quality of joystick performance were recorded, and on Day 2 (D2), the fall boundaries were, without participants’ knowledge, narrowed to ±45°. The initial number of falls on D2 was significantly greater than that at the end of D1, showing that risk increases at ±45°. However, by the end of D2, fall incidence declined significantly, to a level statistically similar to that at the end of D1. Such equivalent final performance indicates that balancing within ±45° boundaries is not more difficult than within ±90°, in the sense that reserve sensorimotor capacity can overcome the higher risk of balancing at ±45° boundaries. EXPERIMENT 2 determined that fall boundaries at ±22.5° made avoiding falls more difficult, but not impossible. Thus, ±22.5° was chosen for the D2 boundaries of EXPERIMENT 3. Initially, a significant increase was found in the likelihood of falling relative to that at the end of D1 (±90° boundaries), indicating increased risk. The number of falls declined significantly during D2, but the final number of falls was still significantly higher than that at the end of D1. Thus, given the same time to adapt as in EXPERIMENT 1, participants could not fully regain the proficiency of balancing at ±22.5° which they had shown at ±45°. Participants had insufficient reserve capacity for balancing at ±22.5° boundaries, at least with the amount of practice permitted. Distinct patterns of joystick control strategy were observed on D2 between EXPERIMENTS 1 and 3, which indicated that a lower limit on the decision time for joystick commands was responsible for the increased difficulty with ±22.5° compared to ±45° fall boundaries. In summary, the results reveal that the ±45° VIP boundaries increased the level of risk but not difficulty, whereas the ±22.5° boundaries increased both. Thus, risk and difficulty are separable in VIP balancing. If such a distinction exists also in bipedal balancing, it would indicate that fall prevention/prediction strategies customized for risk and difficulty factors could have better outcomes than current programs.