Risk factors for falls among adults with intellectual disabilities: A narrative review.

BACKGROUND
The prevalence of falls involving people with intellectual disabilities (ID) is high in comparison with the general population. There has been little evidence to date on the contributing risk factors. The objective of this review was to identify risk factors for people with intellectual disabilities.


METHOD
Literature searches were conducted using electronic databases to explore evidence on the subject, and narrative synthesis was employed to analyse the results.


RESULTS
Seven risk factors were identified: decreasing physical ability, epilepsy, paretic conditions, impulsiveness, previous falls, incontinence and non-use of assistive equipment. Thematic analysis identified factors across the four concepts: the person, the situation, ongoing and protective factors.


CONCLUSION
Factors for falls involving people with intellectual disabilities are dynamic and multifactorial. Some are specific to the population; however, further research is required to develop the understanding of the possible reasons. The findings have implications across policy, education, practice and research.

POPE Et al. Sherrard, Tonge, and Ozanne-Smith (2001) proposed falls as a common cause of injury (including fractures) for people with intellectual disability and the suggested high prevalence is an area of concern that requires to be more fully understood. Pal, Hale, Mirfin-Veitch, and Claydon (2014) suggested reduced confidence and restricted mobility as a consequence of a fall can result in secondary health problems such as anxiety. This is an important issue as people with intellectual disability are already widely recognized to experience high levels of lifestyle-related health conditions (McGarry, 2014). Whilst legislation and guidance have been developed based on the needs of the general ageing population (Department of Health, 2009; National Institute of Health and Clinical Excellence (NICE) 2013), there is limited reference to the specific risks related to adults with intellectual disability. Willgoss, Yohannes, and Mitchell (2010) undertook the first systematic literature review regarding people with intellectual disability and falls risk, with four contributory factors identified: increasing age, epilepsy, mobility and behavioural. The review concluded that research was limited in both quality and quantity to enable a more robust exploration. Recommendations included using different methodologies with large samples and smaller, indepth qualitative studies. Further, it was suggested that increasing the understanding of risk factors would enable the development of interventions amendable to addressing the factors with the most potential to be modified and minimized (Lord, Ward, Menz, & Close, 2007).
The aim of the current review was to build upon the findings of Willgoss et al. (2010) in exploring the factors that contribute to falls in adults with intellectual disability. The objectives were as follows: • To identify if the known risk factors associated with falls involving adults with intellectual disabilities remain relevant; • To identify any new risk factors from 2010 onwards, relevant to adults with intellectual disabilities; • To critically appraise evidence to identify future implications for research, clinical practice, policy and education

| ME THODS
A 12-step strategy enabled a structured approach to documenting of the search strategy for publication (Kable, Pich, & Maslin-Prothero, 2012). A search was undertaken using the following electronic databases: AMED, CINAHL, MEDLINE and PsycINFO. These databases were chosen due to their focus on healthcare research. In addition, ASSIA, Cochrane Library and Google Scholar were used to supplement the database search. The first search was undertaken in November 2017 followed by a further one in July 2019 to ensure all publications have been included. First, a literature search of electronic databases using the terms detailed in Table 1 was undertaken.
Intellectual disabilities (ID) will be the preferred terminology as it is utilized by both the ICD and DSM classifications. However, the search also covered other terms including learning disabilities, mental retardation and developmental disorder.
Articles were then reviewed by the first author against agreed inclusion and exclusion criteria and agreed and confirmed by the research team to ensure rigour. Papers were first reviewed against their title and full available abstract on the database entry.

| Quality appraisal
Quality appraisal of the papers that met the criteria was undertaken by the research team using the Critical Appraisal Skills Programme TA B L E 1 Search terminology S1 AND S2 AND S3 Add search limits February 2009-October 2017 Full text English language "injur*" OR "fractur*" OR "fear of falling" OR "accidental fall*" OR "slip*" OR "trip*" OR "fall*" OR "stumble" OR "loss of balance" OR "unintentional injury" OR "trauma" risk factor*" OR "rate of" OR "risk*" OR "epidemiology" intellectual disabili*" OR "learning disabilit*" OR "developmental delay" OR "cognitive impairment*" OR "mental handicap" OR "mental retardation" OR "developmental disabilit*" POPE Et al.
(CASP) (Critical Appraisal Skills Programme, 2013). Each paper was scored as follows zero-not met as no details were provided, 1-partially met if some details were provided and 2-fully met if details were provided Voss and Rehfuess (2013). A score of 17 or above demonstrates strong qualities with scores under 10 being considered weak. Papers scoring between this will indicate good qualities with some limitations.
Following initial review, the remaining papers were reviewed thematically and commonly recurring rubrics were identified. This was due to the papers included having both qualitative and quantitative data, a narrative synthesis approach enabled themes to be drawn out.
In developing a common rubric, if the p value > .05, results data were extracted. A preliminary examination identified few factors demonstrating p > .01. Odd ratios of > 1 indicating risk factors and < 1 suggesting a protective factor. Risk factors which appeared in more than one paper and meeting the p value are set out in Table 2.
Thematic analysis process started with each paper being reread, with extracts identified based on the relevancy to the review questions. An inductive method was used to develop coding of key themes. Themes were re-visited throughout a two-week period, re-reading the extracts to provide opportunity to ensure themes were consistent. All themes were confirmed by the research team (Appendix 1). Seventeen full-text records remained following screening by title and abstracts using the inclusion criteria. Following a review of full texts, ten papers were excluded as they did not meet the study criteria, leaving seven papers for full quality appraisal. Reasons for papers being excluded were as follows: duplication of data (n = 1), not meeting the review questions (n = 5) and not meeting the inclusion criteria (n = 4). One further paper was identified and included during July 2019 search.

| Study characteristics
Eight papers met the inclusion criteria. Paper publication dates ranged from 2009 to 2019. One paper (Chiba et al., 2009)  Accommodation varied across the studies from supported living to residential facilities; no study was conducted in a hospital setting.
Three papers included qualitative methodologies using focus groups

| Quality appraisal
CASP scores for papers ranged from 13 to 18. Cox et al., (2010) and Four themes were identified: the person, the situation, ongoing factors and risk protectors. (2018) identified risk of falls increased from 3.5% to 6.6% for people with intellectual disability over 55 in a ten-year period in comparison with 1.7% to 3.2% in the general population. Chiba et al., (2009) and Cox et al., (2010) noted that intellectual disability alone was not an identifiable risk factor. Cahill et al., (2014) proposed the limitations of cognitive insight as the risk factor rather than intellectual disability per se. Enkelaar et al., (2013) identified that people with mild intellectual disability were more exposed to situations that they may not have insight to predict and resolve, thereby placing themselves at risk of falls.

| Risk factors for the individual
Three studies identified that females with intellectual disability fell more than males Hsieh et al., 2012;Pal et al., 2014). One study identified gender as not being statistically significant (Cox et al., 2010), with the possible reasons for gender differences not identified in any of the papers. falls history (Cahill et al., 2014;Cox et al., 2010 andPal et al., 2014).
Epilepsy was identified as a significant risk factor in three studies (Chiba et al., 2009;Cox et al., 2010;Finlayson et al., 2010). Cahill et al. (2014) found that people who fall during seizures are incapable of initiating protective mechanisms.

| The situation
Falls can occur spontaneously and immediately. Two studies identified a number of risk factors at the moment of a fall; increased mobility, impulsivity and location. Cox et al., (2010) andEnkelaar et al., (2013) found no significant difference in the balance and gait between people with intellectual disability who fell and those who did not. Enkelaar et al., (2013) found impulsivity and distractibility to be important as a contributory factor related to falls. An ability to initiate protective mechanisms, such as "cushioning" the fall, was identified by Cahill et al., (2014).
Falls were more likely to occur indoors and during the daytime (Pal et al., 2014). Enkelaar et al., (2013) found that people with intellectual disability who fell indoors performed significantly worse in balance and gait assessments. People with intellectual disability were more likely to take risks in their home environment as it was perceived to be "less risky" than outdoors when encountering  Chiba et al., (2009) Finlayson et al., (2010) found that carers felt many falls experienced by some people with intellectual disability were not preventable and were often described as "clumsy". Carers stated falls occurred mainly due to slips, trips and loss of balance (Cox et al., 2010). Carers were more likely to record "unspecified activity"

| Factors for ongoing risks
(i.e. the person was doing an activity which could not be categorized) in their documentation than the general population (Axmon et al., 2018). This means the assessment of falls was made difficult as the information recorded was not specific.
The inappropriate or non-use of assistive equipment was iden-  where more likely to be protected from falls by carers.

| Implications for clinical practice
There are a limited number of studies analysing the efficacy of falls interventions specific to people with intellectual disability.
The limited evidence may be due to the knowledge of the risk factors from which interventions can be developed. Health professionals were found to rely on trial and error to guide their clinical reasoning (Pal et al., 2014). Therefore, the findings from this paper will increase the awareness of the falls risk factors and inform practitioner clinical decision making, thereby aiming to reduce the falls incidence.
There the introduction of a falls clinic specifically for people with intellectual disability had a 23% decrease in falls rate; this is an area requiring further study. It is important to note the similarities in risk factors within other demographics (Hendricks et al., 2005); however, caution is expressed as some factors identified are specific for people with intellectual disability.
Practitioners must also be aware of the importance of family and carer roles as at present, their knowledge of the reasons for falls appears limited to the moment of the actual fall (Cahill et al., 2014). Factors were referred to as "slips and trips" (Cox et al., 2010) and were not felt to not be preventable (Finlayson et al., 2010). Practitioners need to therefore identify how they can improve family and carer awareness as a means to improve falls prevention.  be true for people with intellectual disabilities. Those who are more mobile and able to live with increasing independence were more at risk (Cahill et al., 2014 andEnkelaar et al., 2013). Policymakers need to be aware of these findings to ensure the needs of people with intellectual disabilities is reflected.

| Implications for education
This research has the potential to add to the education of health professionals by demonstrating how the risk factors differ for people with

| Implications for future research
Eight papers in the current study were of good quality following critical appraisal, indicating the need to improve the quantity of the evidence. There are three Cochrane systematic reviews of falls: two for older people (Gillespie et al., 2012 andCameron et al., 2018) and one for people following stroke (Verheyden et al., 2013). Future study needs to be conducted that research the range of risk factors across by utilizing and integrating a range of research methodologies. There is also a need to ensure falls are researched in context of the environmental influences, a recommendation that was made in the review by Willgoss et al., (2010).
It is important to ensure people with intellectual disability have their experiences of falls included in future research; two qualitative studies in the current focused on their views and experiences (Cahill et al., 2014 andPal et al., 2014). A challenge appeared to be the ability to recruit and engage people with intellectual disability as participants.
No study within this review was able to independently recruit participants with intellectual disability. Whilst there could be a research focus on people with mild intellectual disability who are increasingly independent and fall more, caution is expressed in discrimination against those who may have difficulties independently consenting to participate and yet have valuable contributions and experiences of falls (Goldsmith & Skirton, 2015). Therefore, future research needs to consider how all people with intellectual disability can be engaged within falls research, by addressing sampling and recruitment methods and navigating the complexities of research ethics.
This study identified a number of areas with specific risk factors for people with intellectual disability. Age has been suggested as an important risk factor; however, only two studies have assessed the older intellectual disability population (Axmon et al., 2018 andEnkelaar et al., 2013).
Three studies identified female gender as a risk factor; however, further research is required to identify possible explanations if this is the case.
Four studies suggested paretic conditions, highlighting the influence of muscle weakness on the risk of falls. Future research needs to look at scenarios where a person may need to rush or act in impulsively as a factor that resulting in a fall. Pal et al., (2014) recommended further epilepsy research, particularly falls which do not appear to occur due to seizures.
This study has identified co-morbidities which could be protective factors against falls, including Autism and Down Syndrome. This is a new finding and an area requiring further study to investigate possible protective factors, thereby generating new strategies to reduce risk of falls.

| Review limitations
This study aimed to build on the initial systematic review by Willgoss et al., (2010). The authors acknowledge that it is possible that not all studies were identified through the database searches or grey literature. The included studies were published in English and therefore those published in other languages were not included. No studies were located from Africa or South America therefore the generalization of the studies across geographical locations is a limitation.
Due to the limited return of studies that met the inclusion criteria, there was limited scope to compare and contrast similar research approaches. This influenced the quality appraisal tool utilized and the method of synthesis chosen. The use of a narrative synthesis approach was adopted to minimize the risks of bias by presenting the analysis of the evidence using recognized strategies.

| CON CLUS ION
The number of risk factors identified has increased since Willgoss

CO N FLI C T O F I NTE R E S T
There are no known conflicts of interest from the authors in the submission of this manuscript.