Skip to main content
Download PDF

Factors associated with fall risk increasing drug use in older black and white men and women: the Health ABC Study

BMC Geriatrics volume 24, Article number: 773 (2024) Cite this article

Abstract

Background

Most older adults use medications that may increase falls, often defined as fall risk increasing drugs or “FRIDs”. Two definitions for FRIDs, the Centers for Disease Control and Prevention’s (CDC) Stopping Elderly Accidents, Deaths & Injuries (STEADI-Rx) and Swedish National Board of Health and Welfare (SNBHW) definitions, are widely accepted, though include different FRIDs in their definitions. Whether factors associated with FRID use in older adults differ by definition is unknown.

Methods

We hypothesized that factors for FRID use will vary by FRID definition in 1,352 community-dwelling older Black and White adults with medication information in the Health, Aging and Body Composition Study (Health ABC; 2007–08 clinic visit; 83.4 ± 2.8 years; 54.1% women; 65.1% White). Multivariable logistic regression and multivariable negative binomial regression, progressively entering groups of covariates (demographics, lifestyle/behavior factors, and multimorbidity), modeled FRID use (yes/no) and count.

Results

Of 87.0% participants using SNBHW FRIDs, 82.9% used cardiac medications, with lower use of all other FRIDs (range:1.1-12.4%). Of 86.6% participants using STEADI-Rx FRIDs, 80.5% used cardiac medications, with lower use of all other FRIDs (range:1.1-16.1%). Participants with FRID use by either definition were more likely to have chronic health conditions, a hospitalization in the prior year, higher non-FRIDs medication counts, higher Center for Epidemiologic Studies Depression Scale (CES-D) scores, and less physical activity (all p < 0.05). Participants with STEADI-Rx FRID use had poorer vision and higher Modified Mini-Mental State (3MS) scores. In multivariable logistic regression for SNBHW use, hypertension, body mass index (BMI), 3MS scores, and non-FRID count were positively associated with FRID use and poorer vision and Digit Symbol Substitution Test (DSST) scores were negatively associated. In addition to SNBHW factors, higher CES-D scores were associated with STEADI-Rx FRID use. In multivariable negative binomial regression, hypertension, higher BMI, CES-D scores, and non-FRID count were associated with higher FRID count and sleep problems with lower FRID count for both definitions. Higher 3MS and lower DSST scores were associated with higher STEADI-Rx FRID count. Women had lower SNBHW FRID count after adjustments.

Conclusions

Risk factors for FRID use in older adults differ slightly by STEADI-Rx and SNBHW FRIDs definition, but are largely similar.

Peer Review reports

Introduction

Medications with potential to increase a person’s risk of falling are referred to as “fall-risk increasing drugs” (FRIDs). Only two definitions for FRIDs have been used widely. In 2010, the Swedish National Board of Health and Welfare (SNBHW) created a list of 13 clusters of drugs classified as FRIDs [1]. In 2020, the Centers for Disease Control and Prevention (CDC) in partnership with the University of North Carolina Eshelman School of Pharmacy and School of Medicine created a list of 9 categories of drugs that are FRIDs under the program STEADI-Rx (Stopping Elderly Accidents, Deaths & Injuries) [2].

Regardless of definition, the use of FRIDs in older adults (≥65 years) have become increasingly more common. For example, according to data from the Medical Expenditure Panel Survey (MEPS), FRID use showed an increase from 57% in 1999 to 94% in 2017 under the STEADI-Rx definition, possibly driven by the rise in antidepressant, anticonvulsant, and antihypertensive medication use [3]. The increase in FRIDs has also coincided with an approximate linear trend in fall-related mortality from 29.4 per 100,000 in 1999 to 63.3 per 100,000 in 2017, a 115% increase [3, 4]. Current research has not described the reason for these concurrent increases, though FRID use may be a contributing factor.

Additionally, the demographics, lifestyle factors [5], and multimorbidity [5] of older adults who use FRIDs remain largely unexamined, with limited information under the STEADI-Rx definition and no known study using the SNBHW. Based on population data from 1999 to 2017, FRID use was high in those ≥75 years old (79.2%) and those 65–74 years old (72.9%) [3]. Women (77.0%) and men (74.0%) also had a high percentage of receiving ≥1 FRID [3]. Moreover, under-represented ethnic/race groups of older adults that were Native American (80.0%) and Black (78.4%) reported higher FRID use vs. older White (75.8%) and Asian/Pacific Islander (66.1%) adults [3]. However, major gaps remain in understanding sociodemographic risk factors, including education and income, of older adults who use FRIDs, and many demographics associated with FRID use are only available in a non-generalizable subset of the population or in non-U.S. populations, including demographics of individual FRID use [5,6,7,8,9,10,11,12,13]. These factors are important to describe since the population of older adults who use FRIDs could be the target population for deprescribing efforts to reduce falls and fall injury risk in older adults. Thereby, our results will support future recruitment efforts in deprescribing research.

Therefore, the objective of the current study is to, for the first time, assess the association of a wide range of risk factors, including demographics of age, race, sex, and education, lifestyle characteristics, and disease-related factors, with the use of FRIDs, including both FRID classes and counts of FRID classes, in older Black and White men and women from the Health, Aging and Body Composition Study (Health ABC). We hypothesized that demographics, lifestyle/behavior factors, and types of multimorbidity will be associated with FRID use according to both the CDC’s STEADI-Rx and the SBNHW definitions, and that these factors will vary by FRID definition.

Methods

Participants

The Health, Aging and Body Composition Study (Health ABC) was a longitudinal cohort study of Black and White, initially well-functioning, community-dwelling older men and women from Pittsburgh, PA and Memphis, TN to investigate body composition and factors related to functional limitations and disability (n = 3,075; aged 70–79 years; 52% women; 42% Black at baseline in 1997-98). Participants were recruited from a random sample of White Medicare Fee-For-Service beneficiaries and all age-eligible Black Medicare beneficiaries and community residents. To be eligible, participants had to report no difficulty walking ¼ mile, climbing 10 steps, or performing activities of daily living; had no cancer or active treatment in <3 years; and planned to remain in the study area for at least 3 years. This study was approved by institutional review boards at the University of Pittsburgh and University of Tennessee Health Science Center, and written informed consent was obtained from each participant prior to participating [14, 15]. Participants were followed for 17 years (until 2012-13) and contacted yearly to obtain information of interest. By the 2007-08 visit, 1,035 participants were deceased, 44 withdrew, and 113 were missing for other reasons. Of the 1,883 participants remaining in the Health ABC Study in 2007-08 (year 11), 527 had either phone or proxy visits with no medication data. Of 1,356 with clinic visits, home visits, or mixed visits, 1,352 participants were included in the analysis with no missing medication data (Fig. 1).

Fig. 1
figure 1

Participant flow chart for medication data at the 2007-08 Health ABC visit

Full size image

Covariates - factors of interest

Several key factors, including lifestyle and behavioral factors, as well as multimorbidity, were adjusted for in our analyses due to past literature associations with the factors or outcomes of interest. All covariates used in our analysis were measured at the 2007-08 clinic visit unless noted otherwise.

Age, race (White or Black), sex, education (less than high school, high school graduate, and postsecondary education), and site (Pittsburgh/Memphis) were self-reported at baseline [16]. Having non-Medicare health insurance (at the 2006-07 clinic visit), family income (<$10k, $10k-$25k, $25k-$50k, >$50k; at the 2001-02 clinic visit), and marital status (married, widowed, divorced, separated, never married; at the 2000-01 clinic visit) were self-reported [16]. Lifestyle and behavioral factors included smoking status (current, past, never), alcohol consumption (drinks/week), physical activity (kilocalories/week walking and stair climbing), and body mass index (BMI; kg/m2) [16]. Weight (kg) was measured on a standard balance beam scale, and height was measured by a stadiometer to the nearest 0.1 cm [17].

Multimorbidity covariates for this analysis included self-reported fall history reported from the question “In the past 12 months, have you fallen and landed on the floor or ground?”, knee or leg pain over the past year (y/n), any difficulty with activities of daily living (y/n), vision (excellent/good, fair, poor/blind), and sleeping problems (y/n) defined by restless sleep at least some of the time (1–2 days) [18,19,20]. Hospitalization in the previous year was self-reported through a questionnaire [16]. Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression Scale (CES-D) score [21]. Cognitive function was measured using the Digit Symbol Substitution Test (DSST) and Modified Mini-Mental State (3MS) Examination [22]. Diabetes was defined using fasting glucose ≥126 mg/dL and participants taking hypoglycemic medications [19]. Physiological hypertension was defined as measured systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg [15, 20]. Insufficient renal function was defined by cystatin C ≥1 mg/dL [23]. Ankle brachial index <0.9 indicated peripheral arterial disease (PAD) and ≥1.3 indicated arterial stiffening [24].

Medication counts of the number of non-FRIDs were defined by subtracting the count of FRIDs used from the total medications a participant used. Medication inventory occurred during the clinic visits using “brown bag” review method. Participants were asked to bring to clinic all prescription medications that had been taken during the month before the clinic visit. Non-prescription medications were not collected at the 2007-08 clinic visit.

Outcomes: FRID class use & count of FRIDs

Outcomes included either FRID class use (yes/no) or a continuous count of FRIDs under both (1) the SNBHW definition and (2) the STEADI-Rx definition. FRID classes include 13 classes under the SNBHW definition and 9 classes under the STEADI-Rx definition [2, 5]. These medications were coded using the Iowa Drug Information System (IDIS) Drug Vocabulary and Thesaurus codes [25] (Table 1).

Table 1 FRIDs lists and their medication classes for the SNBHW and the CDC STEADI-Rx definitions
Full size table

The medication classes in the STEADI-Rx definition included anticonvulsants, antidepressants, antihistamines, antihypertensives (hypotensive agents, alpha adrenoreceptor antagonists (alpha-blockers), vasodilating agents, calcium channel blockers, diuretics, beta-blockers, angiotensin-converting-enzyme (ACE) inhibitors, and angiotensin II receptor blockers (ARBs), antipsychotics, benzodiazepines, muscle relaxants, opioids, and hypnotics and sedatives. Though the STEADI-Rx definition does not list specific drug names that fall under each medication class, the SNBHW definition uses the Anatomical Therapeutic Chemical (ATC) Classification System, which defines drug names under each code. Specifically, the following ATC drug codes are assessed for the SNBHW definition: antidepressants (N06A), antipsychotics (N05A; N05AN excluded), opioids (N02A), anxiolytics (N05B), hypnotics and sedatives (N05C), vasodilators used in cardiac diseases (C01D), antihypertensives (C02), diuretics (C03), beta-blockers (C07), calcium channel blockers (C08), dopaminergic agents (anti-Parkinson drugs; N04B), agents acting on the renin-angiotensin system (RAS acting agents; C09), and alpha-blockers (G04CA) [5]. Antihypertensives classified by code C02 included in our analysis were centrally acting, ganglion-blocker, and peripherally acting antiadrenergic agents, agents acting on arteriolar smooth muscles, and antihypertensives for pulmonary arterial hypertension (Additional file 2 – Supplementary Table 3). The classifications of STEADI-Rx drugs were determined using pre-set drug classes developed in the Health ABC Study along with consultation of the ATC codes and a pharmacist investigator for the study (Additional file 2 – Supplementary Table 2) [15]. Each FRID class was encoded for our study using ingredient codes. All participants with diabetes taking hypoglycemic medications were also using another FRID and therefore were classified as FRID users though hypoglycemic medications were not included in the count of FRIDs.

Statistical methods

Descriptive statistics were calculated and stratified by FRID use (yes/no) for both definitions to determine statistically significant differences. T-tests or Wilcoxon rank-sum tests were used for continuous variables, and Chi-Square tests or Fisher’s Exact tests for categorical variables. The prevalence of each FRID class was calculated for the sample and included overlap in participants taking ≥1 medication. We also calculated the prevalence of each FRID class for participants taking only 1 FRID class. Risk factors associated with FRID class use and count of FRID classes were modeled using multivariable logistic regression and negative binomial regression, respectively. For the primary analysis using multivariable logistic regression modeling FRID class use (yes/no), covariate factors of interest were entered in groups, where covariates were removed at p≥0.1 except for age, race, and sex which were included regardless of significance since these demographic characteristics are associated with other exposures. The secondary analysis using negative binomial regression to model the count of FRIDs entered covariates in groups, with covariates removed at p≥0.1. Age, race, and sex were also included in the count of FRIDs models regardless of significance. For both the primary and secondary analyses, models were built progressively starting with unadjusted models and progressed by adding age, race, and sex for minimally adjusted models. The models were further adjusted by groups of covariates (1) Model 1: other demographics, (2) Model 2: Model 1 + lifestyle and behavioral factors, and finally (3) Model 3: Model 2 + multimorbidity and count of non-FRIDs. Final models included the same covariates for both definitions for comparability of results, including non-significant variables if significant for the final model of either definition. Collinearity of all covariates was assessed prior to the modeling. Two separate sensitivity analyses were done with variables excluded from initial model building due to high missingness: (1) forcing marital status (26% missing) and income (8% missing) into the final models, and (2) forcing cystatin C (10% missing) and peripheral arterial disease (19% missing) into the final models. All analyses were conducted using SAS software (version 9.4; SAS Institute, Cary, NC) with LOGISTIC and GENMOD procedures to obtain the main results.

Results

Of 1,352 participants, 1,176 (87.0%) were using FRIDs under the SNBHW definition, and 1,171 (86.6%) were using FRIDs under the STEADI-Rx definition in Chi-Square analysis. Of users, 23 were classified as ‘yes’ under the SNBHW definition to ‘no’ under the STEADI-Rx definition, and 18 were classified as from ‘yes’ under STEADI-Rx definition to ‘no’ under the SNBHW definition. Univariate descriptive statistics showed participants with FRID use vs. no use under the SNBHW definition were more likely to be married, have a hospitalization in the prior year, lower physical activity from walking and stair climbing, a higher BMI, a higher prevalence of hypertension, poor renal function, diabetes, higher CES-D scores, and take more non-FRIDs. In addition to these factors, except for income, participants with FRID use under the STEADI-Rx definition were less likely to report vison impairment (poor/very poor/blind vs. excellent vision) and have a higher 3MS score (Additional file 1 – Table 1).

Medication prevalence

Of FRID users as defined by the SNBHW, overall cardiac medications were used by 82.9% of participants (Additional file 2 – Supplementary Table 1). More specifically, RAS acting agents were used by 43.5% of participants, with similar prevalence of some other antihypertensive medications: diuretics, beta-blockers, and calcium channel blockers (Table 2). Prevalence of alpha-blocker, antihypertensive, and nitrate use was much lower. Of the other FRID classes defined by the SNBHW, antidepressants were used by 12.4% of participants, followed by opioids, benzodiazepines, sedative hypnotics, anti-Parkinson drugs, and antipsychotics. Moreover, 334 participants (24.7%) were using just one of the SNBHW FRID classes, primarily consisting of antihypertensive medications. A higher number of women than men used SNBHW antihypertensives. Of FRID users as defined by the STEADI-Rx definition, overall cardiac medications were used by 80.5% of participants, which could similarly be described for comparability with the SNBHW definition as antihypertensives (6.5%), RAS acting agents (43.4%), diuretics (40.8%), beta-blockers (38.6%), calcium channel blockers (33.1%), alpha-blockers (9.5%), and nitrates (4.5%) (Additional file 2 – Supplementary Table 1, Table 2). The prevalence for all other FRIDs use like opioids, antidepressants, muscle relaxants, benzodiazepines, anticonvulsants, antihistamines, sedative hypnotics, and antipsychotics was much lower than overall cardiac medication use (Table 2). Lastly, 707 participants (52.3%) were using just one of the STEADI-Rx FRID classes, which also primarily consisted of antihypertensives. A higher number of women than men also used STEADI-Rx antihypertensives.

Table 2 SNBHW and STEADI-Rx FRID class prevalence (%): Overall and for 1 FRID class only
Full size table

Multivariable logistic regression

In the final multivariable analysis using the SNBHW definition (Table 3), higher BMI, physiological hypertension, higher 3MS score, and higher number of non-FRIDs were associated with higher odds for FRID use. Poor vs. excellent vision and higher DSST score were associated with lower odds of FRID use. Except for CES-D score, findings were consistent for the fully adjusted multivariable analysis using the STEADI-Rx definition (Table 3). CES-D score was associated with higher odds of FRID use using the STEADI-Rx definition.

Table 3 Risk factors for FRID use (yes/no) from the final multivariable logistic regression models
Full size table

Sensitivity analyses for multivariable logistic regression

Being never married vs. married (OR = 0.34, 95% CI: 0.14, 0.83) was associated with a lower odds of FRID use for the STEADI-Rx definition and for the SNBHW definition (OR = 0.34, 95% CI: 0.14, 0.81) in the final multivariable logistic regression models. Income, insufficient renal function, and PAD were not related to either the SNBHW definition or the STEADI-Rx definition.

Multivariable negative binomial regression

BMI, physiological hypertension, CES-D score, and number of non-FRIDs were associated with higher incidence rates of FRIDs count using both the SNBHW and STEADI-Rx definition (Table 4). Sleep problems were associated with lower incidence rates for both definitions, though women had a lower incidence rate of FRIDs count for only the SNBHW definition. Additionally, under the STEADI-Rx definition, 3MS score was associated with a higher incidence rate of FRIDs count and DSST was associated with a lower incidence rate (Table 4).

Table 4 Risk factors for Count of FRIDs (#) from the final multivariable negative binomial regression models
Full size table

Sensitivity analyses for multivariable negative binomial regression

No association was found in the final multivariable negative binomial models between marriage or income and the count of FRIDs. However, insufficient renal function was associated with a higher rate of the count of FRIDs using both the SNBHW definition (IRR = 1.21, 95% CI: 1.10, 1.34) and the STEADI-Rx definition (IRR = 1.18, 95% CI: 1.07, 1.30), and PAD was associated with FRIDs count for the SNBHW definition (IRR = 1.15, 95% CI: 1.03, 1.28) but not the STEADI-Rx definition.

Discussion

To our knowledge, our study was the first to describe the association of a wide range of risk factors for FRID use in older Black and White men and women and to compare the risk factor associations across two common FRID definitions, SNBHW and STEADI-Rx. In community-dwelling older Black and White adults, having hypertension and higher BMI, higher 3MS score, and number of non-FRIDs were associated with a higher odds of FRID use and higher rate of FRID counts, suggesting that multimorbidity in older adults is associated with FRID use. Additionally, the number of medications may be higher in older adults with higher BMI [26] due to the comorbidity associated with obesity [27]. Older adults with higher BMI also may be more likely to have hypertension [28]. One previous Health ABC Study showed that BMI, 3MS score, hypertension, and number of non-FRIDs, may be higher with antihypertensive users than non-antihypertensive users, though these comparisons were not statistically tested [15]. The SNBHW and STEADI-Rx definitions of FRIDs classified antihypertensives similarly overall as described in the methods.

We found some differences in statistical significance between factors using the two definitions, although ORs and IRRs were generally similar. Only one factor (women) had different associations for FRID use (y/n) and the count of FRIDs when using the SNBHW and the STEADI-Rx definition, which may be due to differences in the medication classes included within each definition. For example, the SNBHW definition included anti-Parkinson drugs, which were not included in the STEADI-Rx definition, though the prevalence of those using those drugs were low. Medication classes found in the STEADI-Rx but not the SNBHW definition included muscle relaxants, antihistamines, and anticonvulsants, which all had fairly low prevalence also. Higher CES-D scores were associated with STEADI-Rx FRID use, and lower DSST scores were associated with SNBHW FRID use. Women vs. men had a lower SNBHW FRIDs count, which may reflect more antihypertensive-only users for women than men in this definition for our participants (see Table 4). Higher depressive symptoms may be associated with higher FRID counts since depression in older adults is also associated with chronic diseases such as cardiovascular disease, chronic pain and fatigue [29] which have higher FRID use of antihypertensives, opioids and anticonvulsants [30,31,32]. Participants with sleep problems had a lower rate of FRIDs count, which seems unusual given that antihypertensives were the most common FRID and cardiovascular disease is often associated with sleep issues [33].

Lastly, we found some additional unexpected risk factors for FRID use. Participants with higher 3MS scores had a higher rate of STEADI-Rx FRIDs count, and higher DSST scores associated with a lower rate of FRIDs count. DSST scores were in the expected direction showing participants with higher DSST scores were associated with lower FRIDs count. 3MS scores were in the opposite direction and showed participants with lower cognition were associated with lower FRIDs count. The DSST has often been associated with psychomotor speed, while the 3MS included components of global cognition like orientation, concentration, language, praxis, and immediate and delayed memory [34]. It may have been that the medications for those with lower cognition were more closely scrutinized, resulting in lower FRID use. These results also suggest other FRIDs besides antihypertensives may be responsible for associations as prior Health ABC Study showed no difference in 3MS scores among antihypertensive users vs. non-antihypertensive users [15]. In other studies, opioid and sedative hypnotic use were associated with a higher risk of cognitive impairment [35, 36]. Poor vision was associated with a lower odds of FRID counts for both definitions. One study has suggested that both poor visual input and correcting poor visual input should be monitored in older adults, as both were associated with an increased risk of falls [37]. Therefore, older adults with poor vision are possibly more closely monitored for multiple FRID use to not exacerbate fall risk, resulting in lower use. These unexpected risk factors for FRIDs should be investigated further to confirm our findings.

Only one known study has been published assessing risk factors related to FRID use under the SNBHW definition, and no known studies have assessed risk factors related to use by the STEADI-Rx definition [5]. In the previous study using the SNBHW definition, higher total number of medications, severe falls in the previous year (defined as falls leading to emergency visits at hospitals or hospital admission as a consequence of syncope, contusion, or clinical bone fracture), and women had higher count of FRIDs used, though no other risk factors were assessed [5]. Apart from women having a higher count of FRIDs, these factors from the previous study were similar to ours, but our study additionally reported the total number of non-FRIDs and fall history over the past year instead of severe falls. Importantly, self-reported falls over the past year and severe falls in the past year likely have varying validity. These findings likely show that both less serious and more serious falls lead to higher count of FRIDs. Additionally, the most prevalent FRIDs in our study and the previous study were classes of antihypertensives and antidepressants, thereby providing some comparability between the studies [5]. One major difference between our study and this previous study is our study included only community-dwelling older adults, which may explain why some results did not align. The previous study was comprised primarily of participants living in nursing homes (76%), with community-dwelling older adults (24%) making up a smaller percentage of the participants [5]. Their findings suggest that women in nursing homes may have had higher FRID use than community-dwelling women in our study. Those in nursing homes may have unique risk factors leading to FRID use such as higher multimorbidity vs. community-dwelling older adults. Moreover, higher multimorbidity in the nursing home participants could indicate a higher rate of major depression, which is an indication for antidepressant use [38, 39], or differences in their level of pain severity, which would require different dosages and frequencies of opioid use which may also impact fall risk, with higher dosages and frequencies being associated with higher fall risk [40]. Residency in nursing homes is also associated with severe falls, which may indicate higher multimorbidity or more frailty among residents vs. community-dwelling older adults. Future studies can further explore additional measures of interest related to fall or fall injury risk to assess whether these factors can help further define the populations of older adults who use FRIDs [41]. High-risk populations, like older adults with hypertension, diabetes, chronic kidney disease, and insomnia, also need investigation for FRID use, as these are common indications of FRIDs use by prevalence of disease in older adults.

Neither age, site, nor education were significant risk factors for either definition for either FRID use (yes/no) or the FRIDs count. All of the participants in our sample were aged 79–89 years, therefore a fairly narrowly aged cohort. Future studies using a wider age range of participants are needed to further explore this finding. Additionally, our finding for education was similar to an existing study assessing anticholinergic users vs. non-users [18] but dissimilar to a study assessing antihypertensives users vs. non-users [15]. Therefore, at some of the oldest ages, education may only predict use of certain FRIDs [42].

Though studies have demonstrated that women and White older adults were more likely to fall than men and Black older adults, our study did not find that these demographics were associated with FRID use (yes/no) [43, 44]. The absence of a finding may be due to homogeneity in our populations since all participants in the Health ABC Study were Medicare eligible and able to attend a clinic visit. However, even after the implementation of Medicare Part D, which was introduced to reduce out-of-pocket medication costs and lower the cost of prescription drugs, differences still existed in cost-related nonadherence and benefits from Medicare Part D between non-Hispanic Black and White older adults [45, 46]. This lack of a differential effect for FRID use by race could be explained by differences in Medicare Part D enrollment among the Health ABC Study participants in the 2007-08 visit, leading to a more balanced sample in terms of medication coverage, and consequently medication use, than the general public. It may also be that older White and Black participants in the Health ABC Study were more balanced in indications vs. the general population. Interestingly, women had a lower rate of FRIDs count according to the SNBHW definition, though men (85.1%) and women (85.0%) ≥60 years in the US in 2015–2016 had similar use of all prescription medications (one or more) according to the National Health and Nutrition Examination Survey using the “brown bag” review method [47]. It may be that FRIDs in the SNBHW definitions or the indications of the FRIDs included are more associated with men than women. Since prevalence data for individual FRID use is not published by sex for many of the older adult populations, comparing our results to past studies is not currently possible. Further studies are needed to understand this lower use of FRIDs among women and to evaluate other racial group associations with FRID use, such as Hispanic, Native American, Asian, and Pacific Islander older adults.

Current recommendations on fall prevention provided by the United States Preventive Services Task Force (USPSTF) do not include FRID definitions. USPSTF guidelines focus on exercise interventions while regulating medication management within a level of a multifactorial intervention – multistage interventions in which modifiable risk factors are assessed first and the then subsequent interventions are tailored to each patient or participant based on the first assessment [48]. However, the guidelines do not focus specifically on FRID medication management. Our results inform future interventions on medication management of factors related to FRID use, thus providing foundational knowledge for research on FRIDs and fall/fall injury risk and later allowing for a more targeted approach in recruiting participants.

Our study had some limitations that should be noted. First, participants brought only prescription medications to the clinic visit that were used in the past 30 days. Thus, older adults who use medications as needed, like for opioids, or use over-the-counter seasonal medications, like for antihistamines, may be misclassified as FRID non-users if they had not used the medication in the previous 30 days. The brown bag review itself also may have inadvertently had the effect to reduce inappropriate drug use in participants. Second, cardiac medications constituted the majority of FRID users by either definition, thus our analyses for the counts of FRIDs used is more heavily influenced by the cardiac medications than the other FRID classes of medications. Next, our population is community-dwelling older adults who met the inclusion criteria of the Health ABC Study and were able to attend an in-person visit more than a decade after baseline, which likely resulted in a healthier cohort compared to the general population. Our results may not be generalizable to non-community-based settings, like hospitals or long-term care settings. Finally, STOPPFall (Screening Tool of Older Persons Prescriptions in older adults with high fall risk), a list of 14 medication classes defined as FRIDs that was published in December 2020, fell outside the scope of this analysis [49]. Future analyses could investigate whether the same risk factors from the SNBHW and STEADI-Rx definitions are associated with the STOPPFall definition of FRIDs.

Our study has several strengths. We evaluated many risk factors with two common FRIDs definition at the same time, which has not previously been published. Our medication inventory from the “brown bag” review method was based on participants’ use rather than prescription fill or pharmacy databases, which may be less accurate [50]. Additionally, we assessed numerous potential risk factors, including some with indications for individual FRID use (e.g., physiologic hypertension, depressive symptoms, etc.) due to the wealth of cohort data. Furthermore, the mean age in our study population was ≥80 years old, the population with the highest prevalence of FRID use [3]. Our study population also included a large percentage of Black older adults, a group typically understudied for FRIDs and fall risk factors.

Conclusion

Our study showed that of the risk factors for FRID use in the population, many were same for each definition of FRIDs. Future studies targeting FRID use and fall prevention efforts may benefit from an understanding the population of older adults that uses FRIDs.

Data availability

This study was based on the dataset from the Health ABC Study online through the National Institute on Aging’s website at https://healthabc.nia.nih.gov. The data is available to approved researchers as noted in the Health ABC Study Publication Guidelines.

Abbreviations

FRIDs::

Fall-risk increasing drugs

STEADI-Rx :

Stopping Elderly Accidents, Deaths & Injuries

SNBHW::

Swedish National Board of Health and Welfare

RAS::

Renin-angiotensin-system

BMI::

Body mass index

ADL::

Activities of Daily Living

CES-D::

Center for Epidemiologic Studies Depression Scale

3MS::

Modified Mini-Mental State Examination

DSST::

Digit Symbol Substitution Test

PAD::

Peripheral arterial disease

USPSTF::

United States Preventive Services Task Force

References

  1. Correa-Pérez A, Delgado-Silveira E, Martín-Aragón S, Cruz-Jentoft AJ. Fall-risk increasing drugs and recurrent injurious falls association in older patients after hip fracture: a cohort study protocol. Ther Adv Drug Saf. 2019;10:2042098619868640.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Ferreri SP, Blalock SJ, Robinson JM, Renfro CP, Busby-Whitehead J, Burns ER, et al. STEADI-Rx: guide for community pharmacists. National Center for Injury Prevention and Control of the Centers for Disease Control and Prevention; 2020.

  3. Shaver AL, Clark CM, Hejna M, Feuerstein S, Wahler RG, Jacobs DM. Trends in fall-related mortality and fall risk increasing drugs among older individuals in the United States,1999–2017. Pharmacoepidemiol Drug Saf. 2021;30(8):1049–56.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Drew JAR, Xu D. Trends in Fatal and Nonfatal injuries among older americans, 2004–2017. Am J Prev Med. 2020;59(1):3–11.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Milos V, Bondesson Å, Magnusson M, Jakobsson U, Westerlund T, Midlöv P. Fall risk-increasing drugs and falls: a cross-sectional study among elderly patients in primary care. BMC Geriatr. 2014;14(1):40.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Cox N, Ilyas I, Roberts HC, Ibrahim K. Exploring the prevalence and types of fall-risk-increasing drugs among older people with upper limb fractures. Int J Pharm Pract. 2023;31(1):106–12.

    Article  PubMed  Google Scholar 

  7. Doshi A, Miller GE. Trends in the Use of Diuretics to Treat Hypertension, 1997 and 2003. Agency for Healthcare Research and Quality; 2006 Mar [cited 2022 Sep 19]. Report No.: 120. https://meps.ahrq.gov/data_files/publications/st120/stat120.shtml

  8. Anderson TS, Ayanian JZ, Zaslavsky AM, Souza J, Landon BE. National trends in antihypertensive treatment among older adults by race and presence of comorbidity, 2008 to 2017. J Gen Intern Med. 2022.

  9. Lin Y-H, Huang W-Y, Chang C-C, Chen Y-F, Wu L-Y, Chang H-C, et al. Trends in the use of antimuscarinics and alpha-adrenergic blockers in women with lower urinary tract symptoms in Taiwan: a nationwide, population-based study, 2007–2012. PLoS ONE. 2019;14(10):e0220615.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Ham AC, Swart KMA, Enneman AW, van Dijk SC, Oliai Araghi S, van Wijngaarden JP, et al. Medication-related fall incidents in an older, ambulant population: the B-PROOF study. Drugs Aging. 2014;31(12):917–27.

    Article  CAS  PubMed  Google Scholar 

  11. Haasum Y, Fastbom J, Johnell K. Use of fall-risk inducing drugs in patients using Anti-parkinson drugs (APD): a Swedish Register-based study. PLoS ONE. 2016;11(8):e0161246.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Hálfdánarson Ó, Zoëga H, Aagaard L, Bernardo M, Brandt L, Fusté AC, et al. International trends in antipsychotic use: a study in 16 countries, 2005–2014. Eur Neuropsychopharmacol. 2017;27(10):1064–76.

    Article  PubMed  Google Scholar 

  13. Montastruc F, Bénard-Laribière A, Noize P, Pambrun E, Diaz-Bazin F, Tournier M, et al. Antipsychotics use: 2006–2013 trends in prevalence and incidence and characterization of users. Eur J Clin Pharmacol. 2018;74(5):619–26.

    Article  PubMed  Google Scholar 

  14. Newman AB, Haggerty CL, Kritchevsky SB, Nevitt MC, Simonsick EM, Health ABC Collaborative Research Group. Walking performance and cardiovascular response: associations with age and morbidity–the Health, Aging and Body Composition Study. J Gerontol Biol Sci Med Sci. 2003;58(8):715–20.

    Article  Google Scholar 

  15. Marcum ZA, Perera S, Newman AB, Thorpe JM, Switzer GE, Gray SL, et al. Antihypertensive Use and Recurrent Falls in Community-Dwelling older adults: findings from the Health ABC Study. J Gerontol Biol Sci Med Sci. 2015;70(12):1562–8.

    Article  CAS  Google Scholar 

  16. Sagawa N, Marcum ZA, Boudreau RM, Hanlon JT, Albert SM, O’Hare C, et al. Low blood pressure levels for fall injuries in older adults: the Health, Aging and Body Composition Study. Eur J Ageing. 2018;15(3):321–30.

    Article  PubMed  PubMed Central  Google Scholar 

  17. Barbour KE, Sagawa N, Boudreau RM, Winger ME, Cauley JA, Nevitt MC, et al. Knee osteoarthritis and the risk of medically treated Injurious Falls among older adults: A Community-based US Cohort Study. Arthritis Care Res (Hoboken). 2019;71(7):865–74.

    Article  PubMed  Google Scholar 

  18. Marcum ZA, Perera S, Thorpe JM, Switzer GE, Gray SL, Castle NG, et al. Anticholinergic Use and Recurrent Falls in Community-Dwelling older adults: findings from the Health ABC Study. Ann Pharmacother. 2015;49(11):1214–21.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Introducing the Health ABC Study. The Dynamics of Health, Aging, and Body Composition. [cited 2022 Apr 19]. https://healthabc.nia.nih.gov/

  20. Xue L, Boudreau RM, Donohue JM, Zgibor JC, Marcum ZA, Costacou T, et al. Persistent polypharmacy and fall injury risk: the Health, Aging and Body Composition Study. BMC Geriatr. 2021;21(1):710.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Roberts RE, Vernon SW. The Center for epidemiologic studies Depression Scale: its use in a community sample. Am J Psychiatry. 1983;140(1):41–6.

    Article  CAS  PubMed  Google Scholar 

  22. Mehta KM, Simonsick EM, Rooks R, Newman AB, Pope SK, Rubin SM, et al. Black and white differences in cognitive function test scores: what explains the difference? J Am Geriatr Soc. 2004;52(12):2120–7.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Shlipak MG, Katz R, Sarnak MJ, Fried LF, Newman AB, Stehman-Breen C, et al. Cystatin C and prognosis for cardiovascular and kidney outcomes in elderly persons without chronic kidney disease. Ann Intern Med. 2006;145(4):237–46.

    Article  CAS  PubMed  Google Scholar 

  24. Aboyans V, Criqui MH, Abraham P, Allison MA, Creager MA, Diehm C, et al. Measurement and interpretation of the ankle-brachial index: a scientific statement from the American Heart Association. Circulation. 2012;126(24):2890–909.

    Article  PubMed  Google Scholar 

  25. Marcum ZA, Perera S, Donohue JM, Boudreau RM, Newman AB, Ruby CM, et al. Analgesic use for knee and hip osteoarthritis in community-dwelling elders. Pain Med. 2011;12(11):1628–36.

    Article  PubMed  Google Scholar 

  26. Randhawa AK, Parikh JS, Kuk JL. Trends in medication use by body mass index and age between 1988 and 2012 in the United States. PLoS ONE. 2017;12(9):e0184089.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Lim Y, Boster J. Obesity and comorbid conditions. StatPearls. Treasure Island (FL): StatPearls Publishing; 2024.

    Google Scholar 

  28. Landi F, Calvani R, Picca A, Tosato M, Martone AM, Ortolani E et al. Body Mass Index is strongly Associated with Hypertension: results from the longevity check-up 7 + study. Nutrients. 2018;10(12).

  29. Goodwin GM. Depression and associated physical diseases and symptoms. Dialogues Clin Neurosci. 2006;8(2):259–65.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Rosenblum A, Marsch LA, Joseph H, Portenoy RK. Opioids and the treatment of chronic pain: controversies, current status, and future directions. Exp Clin Psychopharmacol. 2008;16(5):405–16.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Castro-Marrero J, Sáez-Francàs N, Santillo D, Alegre J. Treatment and management of chronic fatigue syndrome/myalgic encephalomyelitis: all roads lead to Rome. Br J Pharmacol. 2017;174(5):345–69.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Stewart J, Manmathan G, Wilkinson P. Primary prevention of cardiovascular disease: a review of contemporary guidance and literature. JRSM Cardiovasc Dis. 2017;6:2048004016687211.

    PubMed  PubMed Central  Google Scholar 

  33. Nagai M, Hoshide S, Kario K. Sleep duration as a risk factor for cardiovascular disease- a review of the recent literature. Curr Cardiol Rev. 2010;6(1):54–61.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Rosano C, Perera S, Inzitari M, Newman AB, Longstreth WT, Studenski S. Digit symbol substitution test and future clinical and subclinical disorders of cognition, mobility and mood in older adults. Age Ageing. 2016;45(5):688–95.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Neelamegam M, Zgibor J, Chen H, O’rourke K, Bakour C, Rajaram L, et al. The effect of opioids on the cognitive function of older adults: results from the personality and total health through life study. Age Ageing. 2021;50(5):1699–708.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Martin P, Tannenbaum C. Use of the EMPOWER brochure to deprescribe sedative-hypnotic drugs in older adults with mild cognitive impairment. BMC Geriatr. 2017;17(1):37.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Ambrose AF, Paul G, Hausdorff JM. Risk factors for falls among older adults: a review of the literature. Maturitas. 2013;75(1):51–61.

    Article  PubMed  Google Scholar 

  38. Dow B, Lin X, Tinney J, Haralambous B, Ames D. Depression in older people living in residential homes. Int Psychogeriatr. 2011;23(5):681–99.

    Article  PubMed  Google Scholar 

  39. Sheffler ZM, Patel P, Abdijadid S, Antidepressants. StatPearls. Treasure Island (FL): StatPearls Publishing; 2023.

  40. Virnes R-E, Tiihonen M, Karttunen N, van Poelgeest EP, van der Velde N, Hartikainen S. Opioids and falls risk in older adults: a narrative review. Drugs Aging. 2022;39(3):199–207.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Xu Q, Ou X, Li J. The risk of falls among the aging population: a systematic review and meta-analysis. Front Public Health. 2022;10:902599.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Naji A, Gatling JW. Muscarinic antagonists. StatPearls. Treasure Island (FL): StatPearls Publishing; 2023.

    Google Scholar 

  43. Geng Y, Lo JC, Brickner L, Gordon NP. Racial-ethnic differences in fall prevalence among older women: a cross-sectional survey study. BMC Geriatr. 2017;17(1):65.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Chang VC, Do MT. Risk factors for falls among seniors: implications of gender. Am J Epidemiol. 2015;181(7):521–31.

    Article  PubMed  Google Scholar 

  45. Bakk L. Racial/Ethnic differences in cost-related nonadherence and Medicare Part D: a longitudinal comparison. J Health Care Poor Underserved. 2015;26(4):1132–48.

    Article  PubMed  Google Scholar 

  46. Jamison JJ, Wang J, Surbhi S, Adams S, Solomon D, Hohmeier KC, et al. Impact of medicare part D on racial and ethnic minorities. Divers Equal Health Care. 2016;13(5):326–33.

    PubMed  PubMed Central  Google Scholar 

  47. Martin CB, Hales CM, Gu Q, Cynthia OL. Products - Data Briefs - Number 332 - February 2019. 2019 [cited 2023 Mar 6]. https://www.cdc.gov/nchs/products/databriefs/db334.htm

  48. Recommendation. Falls Prevention in Community-Dwelling Older Adults: Interventions | United States Preventive Services Taskforce. [cited 2021 Nov 2]. https://uspreventiveservicestaskforce.org/uspstf/recommendation/falls-prevention-in-older-adults-interventions

  49. Seppala LJ, Petrovic M, Ryg J, Bahat G, Topinkova E, Szczerbińska K, et al. STOPPFall (Screening Tool of older persons prescriptions in older adults with high fall risk): a Delphi study by the EuGMS Task and Finish Group on fall-risk-increasing drugs. Age Ageing. 2021;50(4):1189–99.

    Article  PubMed  Google Scholar 

  50. Caskie GIL, Willis SL, Warner Schaie K, Zanjani FAK. Congruence of medication information from a brown bag data collection and pharmacy records: findings from the Seattle longitudinal study. Exp Aging Res. 2006;32(1):79–103.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

This research was supported by the National Institutes of Health/National Institute of Aging T32AG000181, Training in Epidemiology of Aging and National Institute on Aging (NIA) Contracts N01-AG-6-2101; N01-AG-6-2103; N01-AG-6-2106; NIA grant R01-AG028050, and National Institute of Nursing Research grant R01-NR012459; This research was funded in part by the Intramural Research Program of the NIH, National Institute on Aging. This work was also supported by NIH/NIA grant R01-AG061136.

Author information

Authors and Affiliations

  1. Department of Epidemiology, School of Public Health, University of Pittsburgh, 130 N Bellefield Avenue, Suite 300, Pittsburgh, PA, 15213, USA

    Jimmie E. Roberts, Robert M. Boudreau, Kerri S. Freeland, Kristine M. Ruppert, Jane A. Cauley & Elsa S. Strotmeyer

  2. Department of Health Policy and Management, School of Public Health, University of Pittsburgh, Pittsburgh, PA, USA

    Lingshu Xue

  3. Department of Biostatistics, School of Public Health, University of Pittsburgh, Pittsburgh, PA, USA

    Jeanine M. Buchanich

  4. Division of Geriatric Medicine, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

    Jennifer A. Pruskowski

Authors
  1. Jimmie E. Roberts
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Robert M. Boudreau
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Kerri S. Freeland
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Lingshu Xue
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Kristine M. Ruppert
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Jeanine M. Buchanich
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Jennifer A. Pruskowski
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Jane A. Cauley
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Elsa S. Strotmeyer
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

JER: Participated in study concept and design, data analysis and interpretation, and preparation of the manuscript. RMB: Participated in study concept and design, data analysis and interpretation, and preparation of the manuscript. KSF: Participated in data analysis and interpretation and preparation of the manuscript. LX: Participated in study concept and design, data analysis and interpretation, and preparation of the manuscript. KMR: Participated in data interpretation and preparation of the manuscript. JMB: Participated in data interpretation and preparation of the manuscript. JAP: Participated in interpretation of data and preparation of the manuscript. JAC: Participated in study concept and design, data interpretation, and preparation of the manuscript. ESS: Participated in study concept and design, data analysis and interpretation, and preparation of the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Elsa S. Strotmeyer.

Ethics declarations

Ethics approval and consent to participate

This study used de-identified data from the Health, Aging, and Body Composition (Health ABC) Study. The NIA institutional review board approved the study protocol. All participants signed an informed consent form that was approved by the institutional review boards of the University of Pittsburgh and the University of Tennessee.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Sponsor’s role

The funding institutes had no role in the design, methods, subject recruitment, data collection, analysis, and preparation of manuscript or in the decision to submit the manuscript for publication.

Dual publication

The manuscript has not been published elsewhere and reviewed by any other journal. This abstract of this paper was presented in part at the Gerontological Society of America Annual Scientific Meeting in November 2022.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Additional file 1

Additional file 2

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Roberts, J.E., Boudreau, R.M., Freeland, K.S. et al. Factors associated with fall risk increasing drug use in older black and white men and women: the Health ABC Study. BMC Geriatr 24, 773 (2024). https://doi.org/10.1186/s12877-024-05301-w

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12877-024-05301-w

Keywords

BMC Geriatrics

ISSN: 1471-2318

Contact us