Perceived neighborhood walkability is associated with recent falls in urban dwelling older adults

Key points

1. 01Land use mix was the only walkability construct significantly associated with recent falls (OR = 0.55, P = .048) after controlling for covariates.
2. 02Older adults who fell found it harder to walk to a transit stop (mean 2.73 vs 2.12, P < .05) and to stores (mean 1.70 vs 2.15, P < .05) compared with those who did not fall.
3. 03Intrinsic fall risk factors including higher BMI, polypharmacy, foot numbness, and vision problems were more prevalent in the fall group.
4. 04Physical activity levels did not differ between fallers and non-fallers, suggesting the environment itself, not activity avoidance, may drive the association.
5. 05Physical therapists prescribing walking programs should assess neighborhood walkability and direct patients to safe local walking resources.

How it was conducted

Design

Cross-sectional survey study with group comparison (recent fall vs no-fall) and logistic regression

Participants

132 urban-dwelling adults aged 65 or older without cognitive dysfunction or uncontrolled comorbidity, recruited at two medically underinsured health clinics in Flint, Michigan

Fall definition

Self-reported fall in the 4 weeks prior to the survey; defined as coming to rest inadvertently on the ground or another lower level

Walkability measure

Neighborhood Environment Walkability Scale-Abbreviated (NEWS-A), 5 subscales: land use mix, street connectivity, walking/cycling facilities, pedestrian/traffic safety, crime safety

Primary outcome

Recent fall incidence (yes/no) as the dependent variable in logistic regression, controlling for age, physical activity, vision impairment, medications, and BMI

Analysis

ANOVA for between-group comparisons; logistic regression for associations between each walkability construct and falls

What they found

• 45.4% of participants (60 of 132) reported falling at least once in the 4 weeks prior to the survey.
• Land use mix was significantly lower in the recent fall group (mean 2.54, SD 0.65) vs the no-fall group (mean 2.84, SD 0.68), Cohen's d = 0.45, P = .012, 95% CI -0.53 to -0.06.
• After controlling for covariates, land use mix was the only walkability construct significantly associated with falls: beta = -0.60, OR = 0.55, 95% CI 0.30 to 1.00, P = .048.
• Transit stop accessibility item: recent fall group mean 2.73 (SD 1.30) vs no-fall group mean 2.12 (SD 1.21), Cohen's d = 0.48, P < .05.
• Store proximity item: recent fall group mean 1.70 (SD 0.93) vs no-fall group mean 2.15 (SD 1.25), Cohen's d = 0.27, P < .05.
• Street connectivity was not significantly associated with falls: beta = -0.42, OR = 0.66, 95% CI 0.41 to 1.05, P = .081.
• Walking/cycling facilities: beta = -0.06, OR = 0.94, P = .796; pedestrian/traffic safety: beta = -0.16, OR = 0.86, P = .594; crime safety: beta = 0.16, OR = 1.17, P = .468.
• The logistic regression model including land use mix correctly classified 70.4% of participants, Nagelkerke R2 = 0.26.
• Recent fall group had significantly higher BMI (mean 33.57, SD 8.86 vs 27.94, SD 5.41, P < .001), more vision problems (61.7% vs 59.7%, P = .004), foot numbness (58.6% vs 29.2%, P = .001), polypharmacy (86.4% vs 66.7%, P = .009), and assistive device use (38.3% vs 14.9%, P = .006).

Limitations

• Cross-sectional design prevents causal inference; it is unclear whether poor walkability causes falls or whether fallers avoid walking and thus perceive their environment as less accessible.
• Fall history was based on a short 4-week recall window and was self-reported, which may underestimate true fall incidence.
• Sample recruited from two underinsured health clinics in one low-income urban area (Flint, Michigan), limiting generalizability to other urban or higher-income settings.
• The study relied on perceived rather than objectively measured walkability, and perceptions may be influenced by fall-related fear rather than actual environmental conditions.

Why it matters

For patients

Older adults living in urban areas who find it hard to walk to local stores or bus stops may be at higher risk of falling and should discuss this with their healthcare provider.

For clinicians

Physical therapists prescribing walking programs should routinely ask older adult patients about the ease of walking to stores and transit stops, and provide information on safe, accessible local walking options.

For readers

This study provides preliminary evidence that land use mix, a modifiable environmental factor, is independently associated with fall risk in urban older adults, highlighting a potential target for both clinical screening and community-level advocacy.