Associations between the Falls Efficacy Scale International (FES-I) and poor strength and balance in community-dwelling older people

ABSTRACT

Concern about falling (CaF) is associated with higher falls risk in community-dwelling older adults. Our aim was to determine whether CaF (as measured by the Falls Efficacy Scale International version (FES-I)) is associated with poor functional mobility and balance in older adults. Sensitivity and specificity analysis was carried out to test for associations. FES-I score had good to excellent sensitivity when predicting timed up and go (TUG) (87.1%), five times sit to stand time (FTSS) (82.9%) and gait speed (GS) (81.4%) indicative of poor mobility and balance. Moderate specificity was also observed when predicting GS (55.2%) and FTSS (62.3%); a low-to-moderate specificity was observed when predicting TUG (50.0%). A FES-I score indicating CaF showed high specificity and moderate specificity in identifying those with poor scoring. Further work is needed to prospectively assess the relationship between the FES-I and poor mobility and balance.

KEYWORDS:

• Falls
• concerns about falling
• older adults
• gait
• balance

Introduction

Approximately 30% of community-dwelling adults over the age of 65 falls each year (Ganz et al., 2020; Rubenstein, 2006) with around half experiencing more than one fall per year (Hawk et al., 2006). Falls are commonly associated with gait and balance abnormalities (GABAb) (Pirker & Katzenschlager, 2017), alongside deficits in strength. Falls not only have the potential to cause serious immediate health issues, such as head injuries and/or fractures, but are also significant source of morbidity and mortality (Burns et al., 2016; James et al., 2020) and may lead to long-term issues with mobility and a concern about falling (CaF).

CaF is a psychosocial construct encompassing multiple fall-related difficulties, including anxiety, loss of confidence and impaired self-efficacy (Parry et al., 2013). CaF has been reported to have a higher prevalence in women, and increases in incidence with age (Scheffer et al., 2008). Previous iterations of the Falls Efficacy Scale have also been found to be a predictor of future falls (Hadjistavropoulos et al., 2007; Yardley et al., 2005), which may be mediated by impairments in physical performance (Kumar et al., 2016). The amelioration of CaF through exercise (Kruisbrink et al., 2021; Travers et al., 2019) further suggests a link between CaF and GABAb, themselves a predictor of future falls, mediating the relationship between CaF and falls in community-dwelling older adults (Hadjistavropoulos et al., 2010).

The use of a pre-existing questionnaire that assesses CaF, such as the Falls Efficacy Scale International version (FES-I), as a predictive tool has the potential to streamline the identification of those with strength and balance (therefore at risk of falling) and potentially providing them with an intervention to improve any deficits. The FES-I is well established and commonly used in both clinical and research arenas (Whipple et al., 2018).

Previous questionnaires have been developed and implemented targeting the identification of adults at risk of age-related conditions such as frailty (Morley et al., 2012) or sarcopenia (Malmstrom et al., 2016).

Our aim was to determine whether CaF, as measured by the FES-I, is associated with GABAb in community-dwelling older adults attending a falls prevention service as determined by commonly used gait and balance tests.

Materials and methods

Participants

Routine clinical data (including gait and balance measures and FES-I scores) were collected from consecutive attendees at North Tyneside Community Falls Prevention Service, full details of which are published elsewhere (Parry et al., 2016). The service aims to provide community-dwelling older adults (aged 60+) a multifactorial, multi-intervention service for falls and syncope prevention and management. Whilst attending the falls clinic, appropriate participants may be offered a place on the free Age UK North Tyneside Strength and Balance 12-week course. Caldicott approvals were locally granted, while ethical approval was confirmed as not required by North of England Commissioning Service given the routine anonymised clinical data collection.

Measures

The FES-I was used as a measure of CaF (Yardley et al., 2005), given its reliability, validity, and robustness (Hill et al., 2014). The FES-I is a 16-item questionnaire in which participants score their concerns around falling in specific scenarios from 1 (not at all concerned) to 4 (Very concerned). A score of greater than 23 has previously been used as an indicator of significant concerns around falling (Delbaere et al., 2010).

The falls clinics physiotherapist was responsible for completing a comprehensive physical assessment, including the functional outcome measures below:

1. Gait Speed (GS) is measured by walking speed over a distance of 3 m, recorded in seconds. Previous reviews have found a gait speed of 0.8 m/s to be a cut-off in predicting poor clinical outcomes (Beck Jepsen et al., 2022).

2. Timed Up and Go (TUG), in which the time taken for a participant to stand from a seated position, walk 3 m and return to sit in the chair, is measured in seconds, with a time slower than 14 s being considered a predictor of falls (Williams & Nyman, 2018).

3. The Five Times Sit to Stand (FTSS) test, in which the time taken for a participant to stand from a chair five times without the aid of arms, is measured in seconds, with a time slower than 15 s being associated with adverse falls and gait and balance issues (Buatois et al., 2008).

All three physical tests are discussed within the World Falls Guidelines Montero-Odasso et al., (2022) as a strong recommendation as part of a gait and balance assessment (with FTSS included in short physical performance battery test), and gait speed and TUG being recommended to be used to predict falls risk.

Analyses

Analysis was carried out using SPSS for Windows (Version 25, SPSS, Inc., Chicago, IL, USA). Receiver operating characteristics (ROC) curves were used to estimate the sensitivity and specificity of FES-I scores predicting the presence of a poor outcome for a condition. The area of the graph under the ROC (AUROC) curve was also calculated to provide an estimation of how well the FES-I score discriminated between condition outcomes, with positive and negative likelihood ratios also being generated. Due to the output of the ROC curves within SPSS an FES-I score of 23.5 was used as a cut-off point.

Results

Demographics

Nine hundred and ninety-one patients’ data were reviewed (353 males, 638 females). The mean age was 74.5 years (SD = ±8.3 years), mean FES-I score was 28.7 (SD = ±10.2), while 57% reported at least one fall in the year prior.

AUROCs for GS (<0.8 m/s), TUG (>14s) and FTSS (>15s) are shown in Figure 1, with AUROCs for each measure and their 95% confidence intervals, sensitivity, specificity and positive likelihood ratios provided in Table 1.

Figure 1. Receiver operating characteristic curves of physical function tests: gait speed (A), five time sit to stand (B); timed up and go (C).

Table 1. Sensitivity and specificity analyses generated from receiver operating characteristic curves.

Physical performance measure	Area under receiver operating curves (95% CI)	Sensitivity (%) with FES-I 23.5	Specificity (%) with FES-I 23.5	Positive likelihood ratio	Negative likelihood ratio
Gait speed (<0.8 m/s)	0.75 (0.72, 0.78)	81.4	55.2	1.92	0.43
FTSS (>15s)	0.80 (0.78, 0.83)	82.9	62.3	2.20	0.27
TUG (>14s)	0.78 (0.74, 0.81)	87.1	50.0	1.74	0.26

Gait speed

A FES-I score of 23.5 or more had a sensitivity of 81.4% and a specificity of 55.2% in those with a GS of <0.8 m/s, with the likelihood ratio of 1.92 for a positive test, and 0.43 for a negative test.

Ftss

A FES-I score of 23.5 or more had a sensitivity of 82.9%, with a specificity of 62.3% at the >15s cut-off, with a likelihood ratio of 2.20 for a positive test, and 0.27 for a negative test.

Tug

A FES-I score of 23.5 or more had a sensitivity of 87.1% with a specificity of 50% at the >14s cut-off, with a likelihood ratio of 1.74 for a positive test, and 0.26 for a negative test.

Discussion

The relationship between falls and CaF is complex, with existing reviews showing that exercise (particularly strength and balance training) ameliorates both (Kruisbrink et al., 2021; Kumar et al., 2016). Identification and treatment of those with GABAb that predispose to falls is critical in falls prevention. A fall or near fall event often increases CaF in older adults (Ellmers et al., 2022), which may in turn lead to the adoption of behaviours that increase the risk of falls. Older adults that are fearful of falling are more likely to have a lower gait speed, shorter stride length and to spend more with double limb support (i.e. time when both feet on the floor supporting body weight) (Delbaere et al., 2009; Ellmers et al., 2020; Tersteeg et al., 2012). This change in gait is often accompanied by older adults focusing more on the environment around them whilst walking (Ellmers & Young, 2019), requiring what may previously have been an automatic process to become an action that requires deliberate control. The combination of these factors may increase the risk of a person falling over and increase CaF. Long-term concerns about falling may also lead to activity avoidance, causing deconditioning and deficits in strength, which in turn leads to increased CaF.

Our study found that a FES-I score indicative of CaF was associated with scores on commonly used gait and balance measures that suggest both poor physical function and increased falls risks, with high sensitivity (81.4–87.1%), and moderate specificity (50–62.3%). The observed higher sensitivity is more useful regarding the measures used within clinical practice, ensuring that fewer people are incorrectly identified as not having poor strength and balance. Although not ideal, the moderate specificity of the questionnaire may indicate that some older adults require strength and balance training that do not, potentially leading to wasted resources. An FES-I indicating CaF was approximately 15% more likely to occur in those with poor scores on all three tests, indicating a small shift in post-test probability. Of the three tests, the FES-I was not shown to be strongly predictive of poor scoring on any physical function test. This may be due to the users of the falls clinic not just being those with poor strength but also those with dizziness and syncope. Although the FES-I is useful in predicting CaF, it is not able to distinguish between those with issues with dizziness or syncope, and those without. Physical tests like the FTSS and TUG involve participants standing from a seated position, which may cause imbalance in those with above average strength, leading to the incorrect identification of poor functional mobility.

A screening method requiring older adults to self-report poor strength without completing any physical tests would require a distinction must be made between those experiencing deficits in strength, issues with dizziness, imbalance or syncope, or a combination of both. Although the specificity observed of the FES-I to predict all three tests was moderate, combining the FES-I with a simple set of questions around any issues may improve the ability of the test to distinguish between those requiring exclusively strength and balance training, or those requiring a regular assessment.

Our findings are consistent with previous studies that assessed links between CaF and poor physical function, though our comprehensive range of gait and balance measures has never been reported in this context. An association between a slower FTSS and a higher FES-I score has previously been reported (Figueiredo & Santos, 2017). The observed association between both a lower GS and TUG, and a CaF was also expected, with slower GS and TUG scores having previously been suggested to either be a deliberate effort to minimise balance difficulties (van Schooten et al., 2019) or due to a competition for cognitive processes (Herman et al., 2011; Young et al., 2016). Although users of the falls clinic are selected either via GP or through a mailed-out questionnaire, the study population is representative of community-dwelling older adults across the UK, with a similar change in the proportion of older adults being observed in North Tyneside as across the UK (ONS, 2023).

Current screening methods for exercise-based interventions utilise a multifaceted approach (Hopewell et al., 2018), requiring a large number of functional, medical and psychological tests (DiBrezzo et al., 2005), often performed in a secondary care setting. Implementation of a quick, cost-effective and easily administered questionnaire, such as the FES-I, has the potential to allow timely identification of those with GABAb and CaF that may benefit from strength and balance training.

Our findings may provide the initial step in identifying older adults that would benefit from an exercise intervention in a time- and resource-effective manner, with a high chance of finding those with GABAb, albeit with a lower ability to distinguish those without. However, the intervention, strength and balance training, is low risk and cost-effective, with benefits to all regardless of GABAb, so the risks of harm from FES-I—measured CaF in this context are negligible.

Strengths and limitations

The FES-I is an easily implemented and non-labour-intensive method that does not require specialist training to administer or self-complete, making it an ideal tool within clinical practice. Sensitivity was generally high, though specificity was moderate. The study cohort was larger than in similar previous studies (Delbaere et al., 2010); however, the data were retrospective rather than prospective.

Conclusion

CaF assessed by the FES-I is associated with poor performance on commonly used gait and balance measures including the TUG, FTSS and GS. Further work is needed to explore these relationships further, and to assess their potential use in screening for GABAb in the pursuit of improvements in falls prevention.

Availability of data and materials

The datasets used and analysed during the current study are available from the corresponding author on reasonable request

Authors’ contributions

All authors critically reviewed the final draft of the manuscript. LM drafted the final manuscript and assisted on data analysis. VS, DG, CD contributed to the study conception, data collection and analysis. ML planned and carried out statistical analysis. HT contributed to the study conception and data collection. JG aided in data collection. SP contributed with study conception and design, data collection and analysis, and co-drafting of the final manuscript.

Ethics approval and consent to participate

All methods were performed in accordance with relevant guidelines and regulations; Caldicott approvals were locally granted, while ethical approval was confirmed as not required by North of England Commissioning Service given the routine anonymised clinical data collection.

Acknowledgements

The investigators would like to thank the North Tyneside Community Falls Prevention Service team for their help with data collection and anonymisation.

Data availability statement

The data that support the findings of this study are available from the corresponding author, SP upon reasonable request.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the N/A.