Risk factors for falls in community-dwelling older people with mild cognitive impairment: a prospective one-year study

Introduction

Falls are a major issue affecting both physical and mental health in older people (Terroso et al., 2014). The consequences of falls are serious and associated with mortality, morbidity, hospitalisation, and substantial economic costs (James et al., 2020). One third of cognitively healthy community-dwelling older people fall one or more times annually, and this rate doubles in people with dementia (Allan et al., 2009; Kalache et al., 2007; Tinetti, Speechley & Ginter, 1988; Wesson et al., 2013). The proportion of falls are also high in people in the early stage of cognitive impairment, ranging from 53% to 63% in older people with mild cognitive impairment (MCI) over 6 to 12 month follow-up periods (Ansai et al., 2019; Ansai et al., 2018; Davis et al., 2017; Goncalves et al., 2018).

There is now considerable knowledge of fall risk factors in cognitively healthy older people (Lord, Sherrington & Naganathan, 2021) and some studies have elucidated fall risk in older people with manifest cognitive impairment and dementia (Chantanachai et al., 2021; Dolatabadi et al., 2018; Fernando et al., 2017; Modarresi et al., 2019). However, few studies have examined fall risk in older people with MCI, a growing proportion of the older population with measurable memory loss (amnestic) and/or impairments in executive function, attention, language, and visuospatial skills (non-amnestic) that is in excess of normal age-related declines, but not sufficiently severe to affect the ability to carry out daily tasks (Petersen, 2004; Roberts & Knopman, 2013). While these changes do not affect activities of daily living, evidence suggests that subtle changes may occur well before diagnosis and understanding these early impairments will allow for timely intervention.

It has been reported that older people with MCI present with physical and cognitive deficits such as poor balance, gait, executive function and memory (Bahureksa et al., 2017; Johnson et al., 2012). There is some evidence from prospective studies that these factors contribute to fall risk in people with MCI including reduced balance (Makizako et al., 2013), impaired mobility (Ansai et al., 2018), slow choice stepping reaction time (Bunce et al., 2017), slow gait speed (Pieruccini‐Faria et al., 2020), and low levels of physical activity and functional status (Ansai et al., 2019). Reduced ability to dual-task (Ansai et al., 2018) has also been found to increase fall risk, yet impaired functioning in cognitive domains such as executive function, visuospatial abilities, attention, memory, fluency and language has not in people with MCI (Ansai et al., 2019). However, these studies have been limited by small sample sizes (Ansai et al., 2019; Ansai et al., 2018; Makizako et al., 2013), inadequate follow-up length and/or procedures (Makizako et al., 2013), unstandardised fall ascertainment (Makizako et al., 2013) or have had only a narrow focus in terms of risk factors (Bunce et al., 2017; Pieruccini‐Faria et al., 2020). In addition, poor visual contrast sensitivity has been reported to be associated with falls in cognitively healthy older people, but has not been examined in older people with MCI (Saftari & Kwon, 2018). Consequently, our understanding of fall risk in people with MCI is limited and insufficient to inform interventions specifically addressing the needs of this group.

To address this gap, we conducted a large prospective cohort study in community-dwelling older people with MCI to identify key explanatory and modifiable risk factors for falls from a comprehensive range of cognitive, mood, physical, and functional mobility measures. We hypothesised that fall risk factors in older people with MCI are multifaceted and include balance impairment, executive dysfunction and depressive symptoms based on previous work (Ansai et al., 2019; Delbaere et al., 2012; Taylor et al., 2014). By elucidating why older people with MCI fall, our findings may assist in the development of appropriately targeted fall prevention strategies.

Materials and Methods

Assessments

Results

Two hundred and sixty-six of 763 who completed the 12-month follow-up for falls met the criteria for MCI and were included in this study. During follow-up, 106 (40%) MCI participants reported at least one fall (≥1 falls) and 47 (18%) reported multiple falls (≥2 falls). Table 1 presents the demographic and medical characteristics of the fallers (≥1 falls) and non-fallers (no falls). A greater proportion of the prospective faller group reported a fall in the past year and taking psychotropic medications, compared to the non-faller group (Table 1).

Discussion

In our sample of older people with MCI living in the community, 40% reported at least one fall during a 12-month follow-up period. Higher GDS (depression) scores, poorer vision (lower contrast sensitivity), impaired balance (increased postural sway), lower levels of walking activity, and psychotropic medication use were significantly associated with falls, and from these variables, reduced vision, poorer balance and psychotropic medication use were found to be independent predictors of falls in a multivariable analysis.

The prevalence of falls reported here is consistent with one longitudinal study of older people with MCI that had a maximum follow-up period of 84-months (Pieruccini‐Faria et al., 2020). However, other studies have documented higher falls incidences. In a study of 361 older people with MCI, Davis et al. (2017) found 60% fell one or more times during 12-month follow-up, and 38% fell multiple times in this period. Similarly, in a sample of 38 people with MCI, 53% fell at least once during a 6-month follow-up period (Ansai et al., 2019; Ansai et al., 2018; Goncalves et al., 2018). The discrepancy in fall rates is likely due to differing health and socioeconomic profiles of the study cohorts, e.g. greater mobility impairment, higher proportions of females and participants with a history of falls, older age, lower levels of education, and higher rates of psychotropic medication use in other studies (Ansai et al., 2019; Ansai et al., 2018; Davis et al., 2017; Goncalves et al., 2018).

Increased postural sway has been identified as a risk factor for falls in cognitively healthy older people (Chantanachai et al., 2021; Johansson et al., 2017; Kwok, Clark & Pua, 2015; Muir et al., 2010), and in a recent systematic review in older people with cognitive impairment (Chantanachai et al., 2021). The current study complements these findings and builds on previous retrospective studies conducted in older people with MCI (Liu-Ambrose et al., 2008) by confirming an association between impaired balance and prospective falls using gold standard follow-up procedures. Future research could examine the contribution of directional sway which may elucidate contrasting balance strategies between older people with and without MCI.

Our finding that reduced vision was associated with falls in older people with MCI is also in accordance with previous research conducted in cognitively healthy older people (Hong et al., 2014; Lord & Dayhew, 2001; Lord et al., 1994). Poor visual contrast sensitivity may contribute to falls via not only reducing the ability to detect environmental hazards but also by impairing balance, as seen in older people with poor visual contrast sensitivity and stereopsis (Lord, Clark & Webster, 1991), and by restricting activity due to a fear of falling (Wang et al., 2012). Interestingly, this finding contrasts with studies conducted in older people with moderate cognitive impairment, where reduced visual contrast sensitivity was not associated with falls (Taylor et al., 2014; Taylor et al., 2017; Taylor et al., 2012). It is possible people with MCI have higher levels of physical activity and greater exposure to outdoor hazards that could contribute to trips and slips. Future research could explore the relationship between vision and falls in early cognitive decline by examining visual function more thoroughly (including visual acuity, visual contrast sensitivity, depth perception, visual field loss and motion perception) as well as measuring the type of spectacles worn at the time of falls and the locations where falls occur.

Our finding of a significant association between psychotropic medication use and falls builds on the few previous studies that have reported associations between the use of psychotropic mediations and falls or fall-related injury in older people with mild to moderate cognitive impairment and Alzheimer’s Disease (Hart et al., 2019; Horikawa et al., 2005; Taylor et al., 2014), and the studies conducted in cognitively healthy older people (Seppala et al., 2018). In this study, fallers were twice as likely as non-fallers to take psychotropic medications, which are known to lead to central nervous system suppression and interfere with sensorimotor functions (e.g. balance and reaction time) (Seppala et al., 2018). Reducing psychotropic medication use and/or using non-pharmacological treatments to manage anxiety, sleep problems and depression might help to reduce falls in this population. However, future research is required to identify the best strategies for sustaining reduced psychotropic medication use (Campbell et al., 1999).

A few factors that were not included in the final multivariable model require discussion. First, the finding that depressive symptoms was a significant univariate predictor of falls is consistent with the findings of a previous study conducted in older people with MCI (Ansai et al., 2019), and studies conducted in older people with mild to moderate cognitive impairment and dementia (Allan et al., 2009; Taylor et al., 2014). Psychotropic medications are often prescribed to treat depressive symptoms and it is likely that these variables did not provide sufficiently discrete explanatory information regarding fall risk (i.e. had some shared variance). This is supported by the inclusion of depressive symptoms in a sensitivity analysis that excluded psychotropic medications as a possible predictor and suggests depressive symptoms, as well as psychotropic medication use warrants attention when considering strategies for fall prevention in older people with MCI. Second, in line with previous studies reporting low activity levels to be a fall risk factor, reduced walking activity was associated with falls in univariable analysis. The maintenance of physical activity, including walking, may benefit older people with MCI, as multiple studies have documented the beneficial effects of physical activity for cardiovascular, neurological and psychological health (Pinckard, Baskin & Stanford, 2019; Taylor, 2014). However, it should be noted that walking alone has not been shown to prevent falls (Sherrington et al., 2020).

Finally, we did not find significant associations between the measured cognitive functions and falls, including global cognitive function, executive function, memory, processing speed, and verbal fluency. Future research could further examine the relationship between specific features of MCI sub-types (e.g. amnestic MCI and non-amnestic MCI) and falls, as one previous study suggested that only non-amnestic MCI is a risk factor for falls (Delbaere et al., 2012).

The strengths of this study include the relatively large sample of people with MCI and range of quantitative assessments across cognitive, psychological and physical domains as well as health and disability, combined with multivariable modelling that helps identify key factors for falls in this population. However, the results should be considered in view of certain limitations. First, participants were relatively healthy, well-educated and lived in one urban area from one city, which may reduce generalisability of the findings. Second, there is a possibility that not all falls were recorded during follow up. The participants, due to their MCI, may have been less accurate in reporting falls, especially when a fall did not involve injuries, despite the use of gold standard fall ascertainment methods (e.g. monthly calendars and telephone calls) (Lamb et al., 2005). Third, this study did not examine risk factors for falls within subgroups of people with MCI (e.g. amnestic MCI and non-amnestic MCI), due to insufficient subgroup sample sizes.

Conclusions

This study has shown that more depressive symptoms (higher GDS scores), poorer vision (reduced contrast sensitivity) and balance (increased postural sway), lower levels of walking activity, and psychotropic medication use are associated with falls in older people with MCI. Reduced vision, poorer balance, and psychotropic medication use were identified as independent predictors of falls among older people with MCI. As these factors are potentially modifiable, the efficacy of balance training exercises, strategies to avoid psychotropic medication use and correction of vision should be considered in interventions for preventing falls in people with MCI.