Correlates of Self-Reported Falls among Middle Age and Older Homeless Adults
Background: The United States is home to a substantial aging and homeless population. In Los Angeles, close to 40% are over 51 years of age, while 11% are 62 years and older. Falls are a significant and preventable public health issue and are experienced by up to one-third of community-dwelling older adults over 65 years of age.
Methods: Guided by the Frailty Framework among Vulnerable Populations (FFVP), this study assessed 178 homeless men and women in Los Angeles, California, aged 40-73. These participants were recruited from four agencies which served homeless populations. A structured instrument measured sociodemographic data, comorbid conditions (i.e., hypertension, diabetes, arthritis, etc.), use of assistive devices, depressive symptomology, pain, frailty, drug and alcohol use , emergency department use, social support and falls in the past year. Unadjusted relationships between these variables and falls were examined using Pearson Chi-square tests. Characteristics significant in bivariate analysis were included as potential predictors in adjusted analyses. Adjusted relationships between potential predictors and falling were assessed using logistic regression.
Results: The mean age of participants was 52.8 (ages 40-73; SD 6.87) and the majority were female (57.9%) and primarily African American/Black (66.9%), White (11.8%), and Latino/Hispanic (9.0%). In unadjusted analyses, women were more likely to fall in the past year (57.3%, p=.023), along with those who were older (p=.035); moreover, those who had a comorbid condition (82.0%) were more likely to fall (p=.002). Further, those who used an assistive device (79.6%) were more likely to fall (p<.001). Those who met the criteria for depressive symptomology were more likely to fall (p<.001) as were those who self-reported back pain (p<.001). Homeless adults who were frail were also more likely to fall (p<.001). In terms of substances, those who had drug dependency (p=.024) were less likely to fall. Last, those who had low social support were more likely to fall (p<.001) and those who had used the emergency department were more likely to fall (p=.004). In the final logistic regression, significant predictors of falling included self-reported back pain (p=.005), using an assistive device (p<.001), and those with higher levels of social support had lower odds of falling (p=.003). Non-significant predictors included gender, age, depressive symptomology, emergency department use and drug dependency.
Implications: Given that half of the sample self-reported falling within the last year, multidisciplinary health promotion interventions with community partners should be targeted for middle age and older homeless adults to decrease fall risk which could encompass building social support networks among those who utilize assistive devices; client education for fall prevention and management, along with adapting evidence-based fall prevention interventions among this population.