Fall risk prediction in older adults: a generalized linear mixed-effect model analysis based on physical performance...

Area of Science:

• Gerontology
• Biomedical Engineering
• Public Health

Background:

• Falls are a significant cause of morbidity and mortality in older adults, especially those with osteoarthritis (OA).
• Traditional fall-risk assessments using categorical thresholds may not fully capture the complex nature of fall risk.
• Continuous physical performance measures offer a more nuanced approach to assessing fall risk.

Purpose of the Study:

• To develop a probabilistic model for predicting fall risk in older adults.
• To examine the association between continuous physical performance measures (5-Times Sit-to-Stand test, gait speed, lower-limb power) and fall risk.
• To incorporate key clinical and demographic variables (age, sex, OA, hearing impairment, pain) into the fall-risk model.

Main Methods:

• Utilized 13 years of follow-up data from the National Health and Aging Trends Study (NHATS) with 18,023 participants (65+ years).
• Employed a generalized linear mixed model (GLMM) with a binomial distribution to model fall occurrence.
• Analyzed continuous variables: 5-Times Sit-to-Stand (5-STS) duration and power, gait speed, adjusted for age, sex, OA, hearing impairment, and pain.

Main Results:

• The 5-Times Sit-to-Stand (5-STS) power model demonstrated the best fit (R² = 0.7667), indicating greater lower-limb strength is protective against falls (OR = 0.87).
• Longer 5-STS duration (OR = 1.03) and slower gait speed (OR = 1.45) were associated with increased fall risk.
• Continuous measures of physical performance provided significant predictive value for fall risk.

Conclusions:

• Continuous lower-limb performance measures, specifically 5-STS power, offer superior predictive value for fall risk compared to traditional categorical assessments.
• Findings support individualized, data-driven strategies for fall prevention in older adults.
• Further research is recommended to validate these findings in diverse cohorts and explore additional confounders for enhanced clinical applicability.