Dementia as a Risk Factor for Falls and Fall Injuries Among Nursing Home Residents

Carol van Doorn, PhD, Ann L. Gruber-Baldini, PhD, Sheryl Zimmerman, PhD, J. Richard Hebel, PhD, Cynthia L. Port, PhD, Mona Baumgarten, PhD, Charlene C. Quinn, PhD, George Taler, MD, Conrad May, MD, Jay Magaziner, PhD, MSHyg


J Am Geriatr Soc. 2003;51(9) 

In This Article


Information on fall risk factors is presented in Table 1 (prevalence rates for those observed as patient attributes, means and standard deviations for continuous variables). Of the 2,015 cohort members, 48.2% were classified as having dementia. The mean age±standard deviation of residents was 81.4±7.6, 70.4% were female, and 81.3% were white. The mean years of completed education was 10.7±3.9, and 24.1% of residents were married. Over the 2 years (730 days), residents with dementia stayed in the nursing home an average of 364.0±287.0 days, whereas residents without dementia stayed 200.5±259.0 days (P<.0001). Of the 59 facilities included in the study, 70.7% were for-profit facilities, 49.7% were chain-affiliated, 83.4% were in a metro area, and 34.4% had an Alzheimer care unit in the facilities. The mean number of beds in these facilities was 178.0±88.1.

One thousand seventeen people fell 5,438 times during the 2-year follow-up period. The rate of falls was 4.05 per person-year in the nursing home for residents with dementia, compared with 2.33 per person-year for those without dementia (P<.0001), yielding a relative risk of 1.74 (95% confidence interval (CI)=1.34-2.25) ( Table 1 ). Further analysis (not shown) demonstrated that residents with dementia were significantly less likely to self-report their falls, indicating that the calculation of fall risk for these residents is likely an underestimate. The relationship between severity of dementia (as measured using the MDS cognition scale) and falling was examined. It was found that residents who were unimpaired (scoring 0-1) were less likely to fall than those with mild or moderate cognitive impairment (scoring 2-4) (relative risk (RR)=0.67, 95% CI=0.49-0.92), but residents with severe cognitive impairment (scoring 5-10) were no more likely to fall than residents with mild or moderate cognitive impairment (scoring 2-4) (RR=0.99, 95% CI=0.80-1.21).

Table 1 presents the RR of falls for each factor. Female sex, white race, and being married were significant risk factors, conferring a 31% to 54% increase in fall risk. Having been admitted from the hospital was protective against falling, as were being bedfast, chairfast, or using an assistive device. Having an unsteady gait increased the risk of falling, as did wandering. Residents with a history of two or more falls in the past 6 months were 84% more likely to fall than those with no recent fall history. Parkinson's disease; depression; a higher score on the PGDRS behavior scale (indicating worse behavior); and taking antipsychotic, antianxiety, or antidepressant medications increased the risk of falling. The use of bed rails was protective, but the use of a trunk restraint increased fall risk. Residents in a facility with an Alzheimer care unit or a better environmental score were more likely to fall.

Confounding variables were identified by individually adding each factor described above to the bivariate model relating dementia to falls. Four variables altered the RR of falling associated with dementia (RR=1.74) by 10% or more: resident belief that he or she could improve ADLs, stroke history, presence of facility Alzheimer care unit, and number of full-time nursing aide positions per 100 beds. In multivariate analysis ( Table 2 ), controlling for these four covariates, the RR for dementia was 1.93 (95% CI=1.54-2.42).

Of 5,438 falls occurring during the 2-year follow-up period (among 1,017 people who had one or more falls), 59.5% resulted in no injury as recorded in the chart, 2.8% resulted in fracture (1.1% hip fracture only, 0.1% hip and other fracture, 1.6% other fracture only), 31% resulted in soft tissue injury, and 6.7% resulted in other injury.

Dementia was not a significant risk factor for injury among fall events (odds ratio=0.87, 95% CI=0.72-1.05), but the injurious fall rate was higher in residents with dementia because they fell more often than residents without dementia and were therefore more often at risk of injury. Residents with dementia had an injurious fall rate of 1.61 injurious falls per person-year in the nursing home, compared with 0.99 injurious falls per person-year in residents without dementia (P<.002).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.