Socioeconomic Factors Associated With Increased Risk of Falls in Older Adults

Abstract

Objectives: The purpose of this study is to use the National Health Interview Service (NHIS) database to assess the impact of socioeconomic factors on fall risk in older adults in the US. Methods: The NHIS database was retrospectively analyzed to assess several variables related to falls and resulting injuries in older adults. The database covered the period from 2000 to 2018. Results: Our study identified several correlations that have not been previously reported in the literature. Welfare recipiency was associated with more injuries from falls (aOR = 18.8, 95% CI: 1.9-185.6, P = .01), while US citizenship was associated with a lower risk of falls in older adults in the US (aOR = .4, 95% CI: .2-.9, P = .02). Discussion: These findings highlight the importance of evaluating patients for social determinants of health (SDOH) in both the inpatient and outpatient setting. Based on this data, we believe SDOH should be incorporated into fall risk assessments.

“Welfare recipiency was associated with an increased risk of falls, while US citizenship was associated with a lower risk of falls in older adults in the US.”

Introduction

Falls among older adults (age >65) accounted for 3 million emergency department (ED) visits in 2019. ¹ Falls were responsible for 34 000 deaths in this patient population in 2019 and are associated with a medical cost of US$50 billion annually. ¹ Additionally, social factors including income, education, and employment status, have been shown to have an impact on the health behaviors of individuals.^2-4

Several studies have previously examined the risk of falls in older adults, including the influence of socioeconomic factors.^5-8 Kim et al. ⁵ found that older adults in Korea who were less educated and unemployed had a higher risk of falls. Similarly, Paiva et al. ⁶ found that less schooling and lower income were associated with a decline in physical functioning among older adults in Brazil. A study in China found that worse self-rated economic status was associated with a higher risk of falls among older adults. ⁷ A cross-sectional study of older adults in the United States (US) found that low levels of education and income were associated with an increased risk of frailty status, which was associated with an increased risk of falls. ⁸

These studies have largely focused on patient populations outside of the United States. In addition, although prior studies examine certain socioeconomic factors including level of education and income, there are other variables that should be considered. Therefore, there remains a need to examine the relationship between socioeconomic factors and falls in older adults in the United States. The purpose of this study is to use the National Health Interview Service (NHIS) database to assess the impact of socioeconomic factors on fall risk in older adults in the US.

Methods

This study is a cross-sectional analysis of the National Health Interview Survey (NHIS) database, which is a survey collecting data on the health, health care access, and health behaviors of the civilian, non-institutionalized U.S. population. ⁹ Data are collected by census interviewers and interviews are conducted via computer-assisted personal interviewing using a mix of face-to-face interviews for initial questions and phone calls for follow ups. Harmonized data for several available variables of interest were extracted through the Integrated Public Use Microdata Series (IPUMS) Health Survey database for the period spanning the years 2000-2018. ⁹ Given that the data were publicly available, this study was determined to be exempt from institutional review board approval.

In accordance with the definition set by the American Geriatrics Society, we defined our population as individuals > 65 years of age. ¹⁰ Three variables were chosen from the database to determine rates of falls in the target population. Although these variables included adult individuals 18 and older, we chose to only include individuals aged 65 and older in our analysis. The variable “FELYR,” which reports if a person fell in the past 12 months, was obtained from a sample of adults who did not have a balance or dizziness problem in the past 12 months, as defined in the database. This variable was included only in the 2016 questionnaire. Additional fall variables included “FELINJURYR” (defined in 2008 and 2016) and “INJCAUSFALL” (defined in 1997-2014). “FELINJURYYR” was defined as an individual who had an injury resulting from a fall in the past 12 months while “INJCAUSFALL” was defined as the number of times an individual was injured by a fall in the past 3 months. Both of these variables were included only for specific subsets of individuals surveyed. Specifically, the “FELINJURYYR” variable was defined in 2008, for sample adults who have fallen in the past 12 months, and in 2016, for sample adults who have fallen at least once per month in the past 12 months, this variable indicates whether the person had an injury as a result of a fall. ⁹ For persons who had at least one injury/poisoning episode during the past 3 months, the INJCAUSFALL variable reports the number of times a person's injury or poisoning episode was caused by a fall in the past 3 months. ⁹

To obtain nationally representative estimates for analysis involving FELYR and FELINJURYR, subpopulation sample weights (SAMPWEIGHT—Sample Person Weight, is a “variable that represents the random selection of a sample person in the household to complete a supplement survey,” as described in IPUMS database) were used, because these weights define the sampling of the smallest subset (FELYR in 2016 or FELINJURYR variable in 1997-2014). For “INJCAUSFALL,” sampling weights were PERWEIGHT (Final basic annual weight), which is described in IPUMS database as representing “the inverse probability of selection into the sample, adjusted for non-response with post-stratification adjustments for age, race/ethnicity, and sex using the Census Bureau's population control totals. For each year, the sum of these weights is equal to that year's civilian, non-institutionalized U.S. population”.

Statistical Analysis

All variables were described by raw n (number of persons in the survey), raw percentage (sample percentage), and weighted estimates with 95% confidence intervals (nationally representative estimates).

Multivariable logistic regressions were used to investigate risk factors of falling or injury due to falling. Three binary outcome variables (“Fell during past 12 months”—No/Yes, “Had injury from fall, past 12 months”—No/Yes, and “Number of times injured by fall”—0-none, 1-one or more) were regressed on several predictor variables. Predictor variables included Age (in years, continuous), Sex (categorical), Main Racial Background (categorical with 4 categories), Hispanic ethnicity (categorical with 2 categories), Born in the United States (categorical with 2 categories), U.S. citizenship (categorical with 2 categories), Educational attainment (categorical with 3 categories), Ever worked (categorical with 2 categories), Above or below poverty threshold (categorical with 2 categories), Total combined family income (categorical with 3 categories), Received income from welfare (categorical with 2 categories), Family’s home owned or rented (categorical with 3 categories), Needed but couldn’t afford medical care (categorical with 2 categories). The levels of the categorical variables are listed in the tables. Odds ratios, adjusted for the predictor variables in logistic regression models (AORs), with 95% confidence intervals (CI) are reported. Statistical significance was set at P = .05. P-values were reported to three decimal places. Percentages and AOR were reported to one decimal places.

All analyses were performed with R (R Core Team (2020) Vienna, Austria). The variables of interest were read using R package ipumsr. ¹¹ The analysis was performed with R package survey.^12,13

Results

Falls within the Past 12 Months (FELYR)

The survey sample consisted of 5976 respondents and was defined only in the 2016 survey (Table 1). In this sample, 704 (11.8%) respondents reported that they fell during the past 12 months vs 5272 (88.2%) who did not (Figure 1). Based on survey sampling design (sampling weight, variable SAMPWEIGHT), the national estimate of the proportion (weighted %) of people who fell is 11.0% (95% Confidence Interval (CI): 10.0-12.1) vs 89.0% (95% CI: 87.9-90.0) who did not. The multiple logistic regression model indicated that an additional one year of age increases the odds of a fall during the past 12 months by 4.0% (Adjusted Odds Ratio (aOR) = 1.0, 95% CI: 1.0-1.1, P < .001). The odds of a fall during the past 12 months for females increased by 31.0% (aOR = 1.3, 95% CI: 1.0-1.7, P = .03) compared to males. US citizens have lower odds of a fall during the past 12 months compared to non-US citizens (61.0% lower; aOR = .4, 95% CI: .2-.9, P = .02).

Table 1.

Fall During the Past 12 Months (FELYR).

Cohort demographics
	Raw n	Raw %	Weighted %/mean (95% CI)	aOR (95% CI)	P-value
Output variable
Fell during past 12 months
No	5272	88.2	89(87.9-90.0)
Yes	704	11.8	11(10-12.1)
Predictor variables
Age
Age	5976	—	73.2(73-73.4)	1.0 (1.0-1.1)	.000
Sex
Male	2647	44.3	47.0(45.3-48.7)	Ref
Female	3329	55.7	53.0(51.3-54.7)	1.3 (1.0-1.7)	.030
Main racial background
White	5047	84.5	83.8(82-85.5)	Ref
Black/African-American	582	9.7	10.3(9-11.7)	1.2 (.8-1.7)	.433
Asian or Pacific Islander	280	4.7	5.0(4.2-5.8)	1.2 (.5-2.5)	.694
Other race	67	1.1	1.0(.6-1.4)	1.8 (.6-4.8)	.276
Hispanic ethnicity
Not Hispanic	5568	93.2	91.3(89.8-92.7)	Ref
Hispanic	408	6.8	8.7(7.3-10.2)	1.4 (.7-2.5)	.344
Born in the United States
No	668	11.2	14.5(12.9-16.3)	Ref
Yes	5304	88.8	85.5(83.7-87.1)	1.2 (.7-2.1)	.485
U.S. citizenship
No, not U.S. citizen	117	2.0	2.7(2.1-3.3)	Ref
Yes, U.S. citizen	5857	98.0	97.3(96.7-97.9)	.4 (.2-.9)	.022
Educational attainment
At most some college	4212	70.7	69.6(67.9-71.3)	Ref
Bachelor’s degree	968	16.3	17.1(15.9-18.5)	1.4 (1.0-2.0)	.071
Master’s, Professional, or Doctoral Degree	774	13.0	13.3(12.1-14.5)	1.4 (1.0-2.1)	.087
Ever worked
No	282	5.8	6.0(5.1-7.0)	Ref
Yes	4540	94.2	94.0(93.0-94.9)	1.4 (.8-2.5)	.201
Above or below poverty threshold
At or above poverty threshold	5049	92.4	93.4(92.4-94.3)	Ref
Below poverty threshold	414	7.6	6.6(5.7-7.6)	1.0 (.6-1.7)	.912
Total combined family income
US$0-US$49,999	3014	57.9	49.7(47.6-51.7)	Ref
US$50,000-US$99,999	1417	27.2	29.9(28.2-31.6)	.9 (.6-1.2)	.321
US$100,000 and over	778	14.9	20.5(18.7-22.3)	.8 (.5-1.3)	.465
Received income from welfare/public assistance, previous calendar year
No	5934	99.7	99.7(99.5-99.8)	Ref
Yes	20	0.3	.3(.2-.5)	.3 (.0-2.3)	.225
Family’s home owned or rented
Owned or being bought	4767	80.0	83.4(82.0-84.8)	Ref
Rented	1049	17.6	14.7(13.4-16.0)	1.3 (.9-1.9)	.122
Other arrangement	142	2.4	1.9(1.5-2.4)	1.1 (.6-2.0)	.852
Needed but couldn’t afford medical care, past 12 months
No	5857	98.1	98.4(98-98.8)	Ref
Yes	116	1.9	1.6(1.2-2.0)	1.9 (.9-4.0)	.119

Figure 1.

Percent of Individuals Experiencing Fall-Related Outcomes.

Injury Resulting From a Fall in the Past 12 Months (FELINJURYYR)

Overall, 817 of 1697 respondents (48.1%) reported having an injury resulting from a fall within the past 12 months (Table 2). This variable was defined in the 2008 and 2016 surveys. Females had an increased odds of injury resulting from a fall in the past 12 months compared to males (aOR = 1.8, 95% CI: 1.3-2.4, P < .001). In addition, individuals who self-identified as “other race” for racial background had a higher odds of injury resulting from a fall in the past 12 months (aOR = 3.8, 95% CI: 1.2-12.0, P = .03).

Table 2.

Had an Injury from a Fall in the Past 12 Months (FELINJURYR).

Cohort demographics
	Raw n	Raw %	Weighted %/mean (95% CI)	aOR (95% CI)	P-value
Output variable
Had injury from fall, past 12 months
No	880	51.9	52.1(49.0-55.3)
Yes	817	48.1	47.9(44.7-51.0)
Predictor variables
Age
Age	1697	—	76.0(75.2-76.0)	1.0 (1.0-1.0)	.658
Sex
Male	597	35.2	37.9(35.0-40.9)	Ref
Female	1100	64.8	62.1(59.1-65.0)	1.8 (1.3-2.4)	.000
Main racial background
White	1478	87.1	89.8(88.0-91.4)	Ref
Black/African-American	150	8.8	6.7(5.4-8.2)	.7 (.4-1.2)	.209
Asian or Pacific Islander	47	2.8	2.3(1.6-3.2)	1.5 (.6-3.8)	.400
Other race	22	1.3	1.2(.7-1.9)	3.7 (1.2-12.0)	.027
Hispanic ethnicity
Not Hispanic	1601	94.3	94.6(93.2-95.8)	Ref
Hispanic	96	5.7	5.4(4.2-6.8)	1.1 (.5-2.2)	.877
Born in the United States
No	147	8.7	8.6(7.1-10.4)	Ref
Yes	1550	91.3	91.4(89.6-92.9)	1.0 (.5-2.1)	.908
U.S. citizenship
No, not U.S. citizen	38	2.2	2.5(1.6-3.6)	Ref
Yes, U.S. citizen	1659	97.8	97.5(96.4-98.4)	2.2 (.8-6.0)	.143
Educational attainment
At most some college	1355	80.3	80.7(78.4-83.0)	Ref
Bachelor’s degree	184	10.9	10.0(8.5-11.7)	.9 (.6-1.5)	.731
Master’s, Professional, or Doctoral Degree	148	8.8	9.3(7.6-11.1)	1.2 (.7-2.1)	.435
Ever worked
No	115	7.4	7.5(6.0-9.1)	Ref
Yes	1433	92.6	92.5(90.9-94.0)	.8 (.5-1.4)	.382
Above or below poverty threshold
At or above poverty threshold	1282	86.6	89.2(87.4-90.9)	Ref
Below poverty threshold	199	13.4	10.8(9.1-12.6)	.8 (.5-1.3)	.367
Total combined family income
US$0-US$49,999	1157	74.5	70.6(67.5-73.6)	Ref
US$50,000-US$99,999	278	17.9	20.6(18.0-23.4)	.8 (.6-1.2)	.386
US$100,000 and over	119	7.7	8.8(7.1-10.8)	1.1 (.6-1.8)	.792
Received income from welfare/public assistance, previous calendar year
No	1681	99.5	99.8(99.5-99.9)	Ref
Yes	9	0.5	.2(.1-.5)	4.2 (.5-33.0)	.176
Family’s home owned or rented
Owned or being bought	1252	74.0	79.9(77.6-82.2)	Ref
Rented	389	23.0	17.5(15.4-19.7)	1.0 (.7-1.4)	.934
Other arrangement	52	3.1	2.6(1.8-3.5)	.9 (.4-2.1)	.745
Needed but couldn’t afford medical care, past 12 months
No	1612	95.0	95.4(94.1-96.6)	Ref
Yes	84	5.0	4.6(3.4-5.9)	.7 (.4-1.3)	.250

Number of Times Injured by a Fall in the Past Three Months (INJCAUSFALL)

Of individuals who reported having an injury in the last three months, 2714 of 4421 respondents (62.0%) reported their injury resulted from a fall one or more times within this timeframe (Table 3). This variable was defined in surveys from 1997-2014. Multiple logistic regression revealed an increased odds of one or more injury resulting from a fall within the past three months with increased age (aOR = 1.1, 95% CI: 1.0-1.1, P < .001), female sex (aOR = 1.4, 95% CI: 1.1-2.0, P = .02), family home being rented (aOR = 1.6, 95% CI: 1.1-2.3, P = .02), receiving income from welfare/public assistance within the previous calendar year (aOR = 18.8, 95% CI: 1.9-185.6, P = .01) (Figure 2). Individuals who were born in the United States had a lower odds of having an injury resulting from a fall one or more times within the past three months (aOR = .5, 95% CI: .3-.9, P = .03).

Table 3.

Number of Times Injured by a Fall in the Past 3 Months (INJCAUSFALL).

Cohort demographics
	Raw n	Raw %	Weighted %/mean (95% CI)	aOR (95% CI)	P-value
Output variable
Number times injured by fall, past 3 months (0-none; 1-one or more)
0	1707	38.6	38.0(36.4-39.7)
1	2714	61.4	62.0(60.3-63.6)
Predictor variables
Age
Age	4421	—	75.6(75.4-76.0)	1.1 (1.0-1.1)	.000
Sex
Male	1587	35.9	35.6(34.0-37.1)	Ref
Female	2834	64.1	64.4(62.9-66.0)	1.4 (1.1-2.0)	.020
Main racial background
White	3896	88.1	91.3(90.5-92.1)	Ref
Black/African-American	343	7.8	5.7(5.0-6.4)	1.0 (.7-1.4)	.869
Asian or Pacific Islander	129	2.9	2.0(1.7-2.5)	.5 (.2-1.3)	.177
Other race	53	1.2	.9(.7-1.3)	1.4 (.4-4.8)	.602
Hispanic ethnicity
Not Hispanic	4091	92.5	95.2(94.5-95.7)	Ref
Hispanic	330	7.5	4.8(4.3-5.5)	.7 (.4-1.2)	.164
Born in the United States
No	486	11.0	9.2(8.3-10.1)	Ref
Yes	3925	89.0	90.8(89.9-91.7)	.5 (.3-.9)	.027
U.S. citizenship
No, not U.S. citizen	100	2.3	1.6(1.3-2.0)	Ref
Yes, U.S. citizen	4310	97.7	98.4(98-98.7)	1.3 (.4-4.1)	.664
Educational attainment
At most some college	3416	78.8	77.8(76.4-79.2)	Ref
Bachelor’s degree	541	12.5	12.9(11.8-14.1)	.8 (.5-1.2)	.278
Master’s, Professional, or Doctoral Degree	380	8.8	9.2(8.3-10.2)	.8 (.5-1.3)	.395
Ever worked
No	216	9.5	8.8(7.6-10.2)	Ref
Yes	2064	90.5	91.2(89.8-92.4)	.5 (.3-1.0)	.062
Above or below poverty threshold
At or above poverty threshold	2974	87.4	89.4(88.2-90.5)	Ref
Below poverty threshold	428	12.6	10.6(9.5-11.8)	.9 (.6-1.3)	.516
Total combined family income
US$0-US$49,999	1694	66.4	65.3(63.1-67.5)	Ref
US$50,000-US$99,999	586	23.0	23.5(21.6-25.6)	1.0 (.7-1.5)	.977
US$100,000 and over	271	10.6	11.2(9.9-12.6)	1.3 (.7-2.4)	.379
Received income from welfare/public assistance, previous calendar year
No	4309	99.4	99.5(99.2-99.7)	Ref
Yes	24	0.6	.5(.3-.8)	18.8 (1.9-185.6)	.012
Family’s home owned or rented
Owned or being bought	3403	78.1	79.0(77.3-80.6)	Ref
Rented	825	18.9	18.1(16.6-19.6)	1.6 (1.1-2.3)	.018
Other arrangement	130	3.0	3.0(2.4-3.6)	1.0 (.5-2.3)	.975
Needed but couldn’t afford medical care, past 12 months
No	4225	95.7	96.1(95.5-96.7)	Ref
Yes	191	4.3	3.9(3.3-4.5)	1.2 (.7-2.2)	.440

Figure 2.

Adjusted Odds Ratios for Select Fall Risk Factors.

Discussion

Falls continue to be a concern for older adults in the US and are associated with a wide array of risk factors. Few studies have examined socioeconomic factors associated with falls in older adults in the US. Our study found several correlations that have not been previously reported in the literature. Welfare recipiency was associated with an increased risk of falls, while US citizenship was associated with a lower risk of falls in older adults in the US. In addition, increased age and female sex were associated with an increased risk of falls in older adults, while no difference was found based on educational status. Validated models such as the Hendrich fall risk model have been developed to assess the risk of falls in older adults in the acute care setting. ¹⁴ These models take into account many physiologic and health-state associated risk factors, but do not account for the multifactorial nature of falls including social determinants of health (SDOH) and socioeconomic status (SES).

The current study found no difference in fall risk for older adults based on educational status. This is in contrast to several prior studies which have shown that lower educational status is associated with an increased risk of falls.^5,6,8,15 It is possible that having a higher level of education may result in increased awareness of the risk of falling and allow patients to be better equipped to access educational resources regarding mitigating fall risk. ¹⁵ However, in older adults with mild cognitive impairment, years of education and measures of cognitive function were not found to be associated with falls. ¹⁶ The impact of educational status on fall risk in older adults warrants further study.

Similar to prior studies that have demonstrated gender as a risk factor for falls in older adults, we found that females had a higher risk of falling and injury secondary to a fall.^17-19 Several potential explanations have been offered including that males may be less likely to report falls ¹⁹ and that females may be more likely to have altered gait mechanics ²⁰ and osteoporosis ²¹ contributing to fall risk. Similarly, increased age has been associated with an increased risk of falls.^15,22-25 We found that each additional year of age increased the odds of a fall by 4.0%. Increased age is associated with several potentially contributing factors including cognitive decline, diminished mobility, and increased frailty.

We found that SES and several SDOH were associated with falls or fall-related injuries in older adults. People who were born in the US and those who had US citizenship had lower odds of being injured by a fall. We postulate that individuals with citizenship status may have better access to resources and more willingness to utilize the health care system. On the other hand, those without citizenship status, particularly those of lower SES, may be reluctant to seek out health care and may not have access to various governmental health care services or benefits. Interestingly, Gill et al. ¹⁵ found that individuals in South Australia who were born in non-English speaking countries had a lower risk of falls, possibly due to greater family support.

Individuals receiving income from welfare had higher odds of being injured by a fall. These individuals are typically of lower SES which may contribute to fall risk in multiple ways. Low SES is correlated with cognitive impairment which is a known risk factor for falls. ²⁶ Additionally, individuals of lower SES may have varying perceptions of and more difficulty accessing health care resources including assistive devices like canes or walkers. ²⁷ Renting a home was also associated with a higher odds of being injured by a fall. Gill et al. previously reported that individuals who reported that their home was not in need of maintenance were significantly less likely to fall, which may indicate a safer home environment.^15,28 It is possible that individuals who rent their home, particularly those of low SES, may have less than ideal living situations that increase risk of falls including outstanding repairs or maintenance and difficulty with proper accommodations for wheelchairs or other assistive devices. In addition, those who rent a home may not be allowed to make modifications to their home that have been shown to reduce fall risk such as installing guard rails on stairs or grab bars in showers. ²⁹

Fall risk assessment is complex and multifactorial. Our paper supports the notion that socio-economic factors play an important role in determining which patients are at risk for falls. This paper identifies novel demographic considerations in this realm including welfare recipiency and US citizenship among other factors such as age and biologic sex. Individual clinicians and public health programs can make use of these demographic risk factors when identifying potential beneficiaries and for creating community programs and public health policies for fall reduction interventions. This is of particular importance for those practicing Lifestyle Medicine, aspsychosocial interventions targeting nutrition,^30,31 inactivity, ³² and other modifiable risk factors, ³³ may be paramount to preventing falls and ensuring an improved quality of life for older adults. ³⁴

The widespread use of electronic health records (EHR) has allowed clinicians and public health officials to capture a broad spectrum of demographic data, which through various means (i.e., risk stratification scores, predictive algorithms and clinical determination) helps to identify individuals at risk for falling. ³⁵ Based on this data, we believe SDOH should be incorporated into fall risk assessments. Inclusion of socio-economic risk factors into EHRs, which is already available in some of the EHR programs, could be used in clinicians’ standard practice to aid in the accuracy of falls risk assessments, and aforementioned interventions. This may help prevent quality of life altering and financially costly injuries to older adults. In addition, with the advent of machine learning algorithms in assessing fall risk, incorporating these variables may help increase the accuracy and quality of these models.^36-38 However, unless identified, accurately determining patients at risk for falls will remain a difficult endeavor.^39-41 Future studies should focus on assessing the benefit of incorporating these risk factors into clinical risk assessments.

Limitations

Several limitations to this study should be noted. Given that the study was a cross-sectional analysis, we were unable to determine any causal relationships between the variables that were analyzed. Although the NHIS database is comprehensive, it is based on surveys of the population and therefore individuals must self-report answers to various questions. As a result, data may be subject to sampling bias and recall bias. In addition, certain individuals may be more reluctant to be forthcoming regarding certain information related to their health or SES. Although the database contains information associated with SES, further clarifying data such as geographic data and zip code were not elucidated. Furthermore, one of the fall variables analyzed included only individuals who reported no problems with dizziness or balance in the past year, which limits generalizability and results may differ from the general population given that both of these issues are significant risk factors for falls in older adults. Finally, the NHIS is available in only English and Spanish and consists only of the US population, which may limit generalizability to other populations that do not speak these languages within the US.

Conclusion

This study provides an analysis of falls in older adults in the US with a focus on social determinants of health and their impact on fall risk and injuries. Similar to prior studies, female sex and increased age were associated with an increased risk of falls in older adults. Contrary to prior studies, we found no statistically significant association between educational status and fall risk in this population. Several unique findings in this study related to SES include lack of US citizenship, receiving income from welfare, and renting a home. These findings highlight the importance of evaluating older adults for SDOH and the potential need to include these variables in standardized fall risk assessments.