Factors associated with falls in older adults with cataracts

The aim of the present study was to evaluate factors associated with falls in community-dwelling older adults diagnosed with cataracts. An analytical, cross-sectional study was conducted with a sample of community-dwelling older adults residing in the Federal District of Brazil. Interviews and assessment tools were administered, such as the Timed Up and Go test, Short Physical Performance Battery (SPPB), Biodex Balance System, Katz Index, Lawton Scale, Minnesota Leisure Time Physical Activity Questionnaire and Mini Mental State Examination. Statistical analysis involved binary logistic regression. One hundred forty-two older adults (85 with cataracts) participated in the study (mean age: 69.39 ± 5.67 years). Falls were associated with the female sex (OR: 4.45) and sub-maximum score on the SPPB (OR: 3.53) among patients with cataracts, whereas multimorbidity (OR: 5.10) was the risk factor risk for older adults without cataracts. The data suggest different risk factors for falls among older adults diagnosed with cataracts.


Introduction
Cataract is the most common cause of blindness and one of the most prevalent degenerative diseases among older adults 1 .It is estimated that among the 20 million people affected by severe visual impairment throughout the world, one third of the cases is the result of cataract 2 .According to data from the 2013 Brazilian National Health Survey, cataract affects 28.7% of the older population in the country 3 .
Cataract can exert a negative impact on postural balance, mobility and functional capacity and is related to the fear of falls, a greater number of morbidities, depressed mood and cognitive decline 4,5 .As postural balance depends on the proper integration of vestibular, proprioceptive and visual information, any impairment in one of these sensory systems can lead to deficiencies in the postural adjustments needed to maintain one's balance 6,7 .
Older adults with cataracts have blurred vision, which results in a reduction in visual acuity as well as reduced sensitivity to contrast and the perception of colors 1 .When combined with poor lighting, cataracts favor the occurrence of tripping and slips on irregular surfaces on the ground of the environment (e.g., puddles, carpet edges, doorjambs, etc.) 8 .Indeed, studies report an association between visual impairment and falls in older adults 8,9 .
There is a consensus in the literature regarding the significant improvement in visual acuity following cataract surgery 10,11 .However, the effect of surgery on the risk of falls is inconclusive 10 ; there are reports of both a reduction 12 and an increase in such risk 11 .Other factors, such as lifestyle, the use of medication, multimorbidity and level of physical activity, can affect the risk of falls among older adults with cataracts, even after recovering their vision 11 .
Besides visual impairment, other intrinsic and extrinsic factors may be related to the risk of falls among community-dwelling older adults, such as urinary incontinence, depression, fear of falls, balance abnormalities, cognitive deficiency, polypharmacy and functional limitations 13 .The literature offers numerous studies addressing the multidimensional aspects of the risk of falls in older adults living in the community 14,15 and those in nursing homes 16,17 .However, few researchers have addressed groups with specific problems, such as older adults with cataracts.Therefore, investigations are needed to examine the relationship between visual impairment and falls, as many studies do not take into consideration the influence of multiple confounding factors that may be associated with this outcome 18 .
This investigation was based on the hypothesis that older adults with cataracts have different risk factors for falls than those without cataracts.Thus, the aim of the present study was to evaluate factors associated with falls in community-dwelling older adults with cataracts.

Methods
An observational, analytical, cross-sectional study was conducted in the Federal District of Brazil.Community-dwelling older adults were recruited through active searches at healthy ageing programs in the city of Ceilândia and ophthalmology services at two public hospitals in the Federal District.
Male and female older adults aged 60 years or older with a medical diagnosis of bilateral cataract confirmed using standardized methods 1 were included in the study.The exclusion criteria were a clinical diagnosis of neurological disease or uncorrected visual impairment and a score on the Mini Mental Health Examination suggestive of cognitive impairment (cutoff point adjusted for schooling) 19 .Older adults with paralysis or orthopedic abnormalities or the inability of remain in a standing position and walk without assistance were also excluded.
This study was conducted between December 2011 and December 2012 in compliance with recommendations for research involving human beings and received approval from the Ethics Committee of the School of Education and Research in Health Sciences, Brasília, DF, Brazil (certificate number: 0153/11).The present study is part of a larger investigation denominated: "Impact of cataract surgery on the occurrence of falls and multidimensional aspects of health: A longitudinal study involving older adults in the Federal District of Brazil".
The outcome variable (falls) was defined as an unexpected event in which an individual falls to the ground or a lower level 20 .This outcome was based on self-reports from the participants using the previous 12 months as the reference.
The other variables and respective assessment methods are listed below: -Socio-demographic variables: age, sex, marital status, reading skill, schooling, health conditions (diagnosis of cataract, having undergone cataract surgery, comorbidities, depression, number of medications and use of psychotropic agents) and lifestyle (smoking, physical activity and alcohol intake).Multimorbidity was defined as the occurrence of five or more chronic-degenerative diseases (diabetes, arterial hypertension, kidney disease, cardiovascular disease, depression, osteoporosis, etc.).These data were collected with the aid of a structured questionnaire; -Level of physical activity was determined using the Brazilian version of the Minnesota Leisure Time Physical Activity Questionnaire 21 .The classification as active or inactive was based on the criteria recommended by the American College of Sports Medicine: at least 150 minutes per week of moderate physical activity (3 METs)  or 75 minutes of vigorous or sports activity (6 METs) 22,23 ; -Physical-functional performance was evaluated at the Movement Analysis Lab of the university using the Brazilian version of the Short Physical Performance Battery (SPPB) 24 to evaluate lower limb strength, balance, gait speed and the risk of falls and the Timed Up and Go (TUG) test 25 for the evaluation of balance, functional gait capacity and the risk of falls.Cutoff points were determined using ROC analysis (not presented in this study) for the risk of falls based on the SPPB (ordinal) and TUG (numeric), which were considered binary variables.The risk of falls was also evaluated using a force plate: Biodex Balance System (BBS) 26 ; -Functional independence was measured using the Brazilian version of the Katz Index of Independence in Activities of Daily Living and the Lawton Instrumental Activities of Daily Living Scale 27,28 , which have been validated for the Portuguese language.
To evaluate associations between the outcome (falls) and the variables of interest in the older adults with and without cataract, binary logistic regression analysis was conducted using the SPSS program, version 20, to determine odds ratios (OR).The variables determined to be possible risk or protection factors regarding the occurrence of falls were incorporated in the model one by one (univariate regression analysis with "insert" method).Variables with a p-value < 0.20 were incorporated in blocks of variables grouped based on socio-demographic characteristics, health conditions, physical-functional performance and lifestyle to obtain a more representative model.Multivariate analysis was then performed based on the division of the sample into individuals with and without a diagnosis of cataract (models A and B, respectively).All models were adjusted for age.The "at least ten events" rule (successes or failures) was obeyed for each independent variable in the model.Differences between groups were evaluated using the Mann-Whitney test and chi-squared test (categorical variables).

Results
One hundred seventy-six older adults were invited to participate in the study, 144 of whom appeared for the evaluations.Two were excluded (one for not being able to see the target on the screen of the BBS and another was unable to report whether or not a fall had occurred in the previous year).Thus, 142 older adults participated in the study.Table 1 displays the characteristics of the sample stratified for the presence/ absence of cataract.No statistically significant differences were found between the two groups regarding the variables investigated, except visual acuity and the risk of falls determined using the TUG test with a cutoff point of 7.67 seconds.
The prevalence of falls in the previous 12 months was 40%.The prevalence of the practice of physical activity at the recommended level was high in the overall sample.Sub-maximum scores were found regarding the evaluations of physical-functional performance and independence on basic and instrumental activities of daily living.A tendency toward a greater prevalence rate of falls among the individuals without a diagnosis of cataract was found, but the difference between groups did not achieve statistical significance.
Table 2 displays the results of the univariate analysis considering the groups with and without cataract and the results of the analysis of differences between groups regarding the separate variables.Variables individually associated with falls are highlighted with one (p < 0.05) or two (p < 0.20) asterisks in each group.The cutoff points for the TUG test and SBBP were 7.67 seconds and 11 points, respectively.Multivariate regression models A and B represent the risk of falls in the groups with and without cataract, respectively, and were adjusted for both age (75 years or older) and sex.The female sex had fourfold greater odds of having suffered a fall than the male sex.Individuals with cataracts who had a score lower than 11 points on the SPPB had 3.5-fold greater odds of having suffered a fall than those with a higher score.Multimorbidity was associated with falls in the group without a diagnosis of cataract (Table 3).

Discussion
In this study, factors associated with falls in community-dwelling older adults with a diagnosis of cataract were investigated and compared to factors found for older adults without cataracts.The female sex and a score of less than 11 points on the SPPB were associated with falls in the group with cataract, whereas having five or more chronic diseases was associated with a greater chance of falls in the group without cataract.
Few analytical studies involving older adults with cataract have been conducted in Brazil.Previous investigations report associations between cataract and postural balance, fear of falls and quality of life 29,30 .However, the studies cited did not address other dimensions of the phenomenon (falls) in the population and were limited to the description of socio-demographic characteristics 31,32 .Likewise, few multidimensional studies are found in the international literature addressing the relationship between visual impairment and falls 18 .
Visual impairment can cause difficulties in detecting obstacles with little contrast, judging distances, perceiving spatial relationships and processing all visual information necessary to perform adjustments in order to maintain postural control and adequate mobility, thereby hindering the adaptation to the environmental demands required for the execution of activities of daily living 9,7 .The impact of visual impairment on falls may be greater when accompanied by other sensory and balance deficiencies 33 .
An impaired performance on visual tests is related to a worse performance on physical-functional tests. 34Older adults with cataracts and a self-reported fear of falls are less confident regarding their ability to maintain their balance and avoid falls and may therefore limit their functional activities due to the fear of falling 30 .There is a gap in knowledge regarding the effects of visual impairment on the level of physical activity among older adults.Cataract surgery is known to improve visual acuity 10 , but there is no certainty regarding a subsequent reduction in the risk of falls [10][11][12] .
Cataract is considered an independent risk factor for falls 9 and can determine more accentuated impairment with regard to balance and mobility in comparison to older adults without cataract 5 .Thus, one would expect a poorer performance on balance tests and, consequently, a greater prevalence rate of falls in individuals with cataract.However, this did not occur in the present study, as no statistically significant difference was found between groups regarding the prevalence of falls.In fact, there was a tendency toward a greater prevalence rate of falls in the group without cataract, whereas those with cataract had a worse performance on the physical-functional tests.The prevalence of falls among older adults without cataract may be associated with the occurrence of multimorbidity, as having five or more comorbidities increased the odds of falling fivefold in this group.Multimorbidity may be associated with falls 35 and frailty, which can increase the risk of unfavorable outcomes over time, such as falls, hospitalizations and fractures 36 .
Despite this apparent difference in the prevalence of falls between the groups studied, no significant differences were found between the individuals with and without cataract for any of the variables investigated, except visual acuity and performance on the TUG test.A previous case-control study conducted in Malaysia involving older adults who had suffered hip fracture found no differences between the cases with cataract and those without cataract (controls) with regard to the pre-morbid state, degree of inde- pendence for mobility, the use of gait-assistance devices or living alone 37 .On the other hand, older adults in Scotland with visual impairment at primary care services exhibited more comorbidities in comparison to those without this type of impairment 38 .The differences between studies may be related to the recruitment setting, as the sample was recruited from a hospital setting in the first study and primary care in the second.
An interesting finding of the present investigation was the high prevalence of falls in community-dwelling older adults with high levels of functioning, as demonstrated by the high levels of physical activity and sub-maximum scores for physical-functional performance and independence in terms of basic and instrumental activities of daily living.
A poor performance on a balance test, such as the TUG test, is associated with a history of falls 39 .Therefore, this test has been used to screen for older adults at risk of suffering a fall, although there is no consensus on the ideal cutoff point for the identification of this population 40 .Indeed, the use of the TUG test has been questioned with regard to its predictive capacity for falls among older adults 41 .Since older adults with higher levels of physical activity obtain better scores on postural balance tests 42 , there may be a ceiling effect for scores on balance tests 40,43 .
The participants in the present investigation had high scores on the physical-functional performance tests (TUG, SPPB and BBS).It should be stressed that the adequacy of cutoff points on balance tests for the profile of the population studied is fundamental.However, even with the establishment of adequate cutoff points on the physical-functional tests to improve sensitivity and specificity 42 , the measures for the evaluation of postural balance (TUG and BBS) were not associated with falls in the groups investigated in the present study.Moreover, the physical-functional tests employed do not include an evaluation of the sensory components of postural balance 43 .
Taken together, the present findings and those described in the literature suggest that a higher score on physical-functional tests does not necessarily imply an absence of a risk of falls.Individuals with high levels of functioning and no postural balance deficiencies continue to perform their habitual activities and could expose themselves to situations of risk 44 .Therefore, the high prevalence of physical activity among the individuals may explain the high prevalence of falls in the present sample, as 41% reported practicing physical activity at the recommended levels 19 and this percentage was even higher (52%) among the individuals with cataract.These figures are higher than the rate described in the 2013 National Health Survey, in which the proportion of older adults who practiced physical activity at the recommended level in Brazil was 13.6% 45 .
Although practicing physical activity can reduce the risk of falls 46 , practicing such activity in an outdoor environment may expose older adults to environmental barriers 47 .Moreover, Okubo et  al. (2015) found that the habit of walking did not increase the risk of falls among older adults with a low risk of falls, but was strongly associated with this outcome among those with two or more risk factors for falls 48 .
In the present study, the regression models revealed different risk factors for falls between older adults with and without cataract.In the group with cataract, the female sex and a poorer physical performance remained in the final model as factors associated with falls.
Women with cataracts were fivefold more likely to fall than men with cataracts.This may have occurred because women undergo sensory, motor and cognitive decline earlier than men 49 and are closer to the limit at which reduced muscle power could compromise their functional status 50 .Another factor to consider is the greater use of healthcare services by older women, including medical appointments, which increases the chances of a greater number of diagnoses of cataracts 51 .Women also have greater longevity and a consequent accumulation of age-related disease and also seek cataract surgery more 52 .Similar differences between women and men regarding the functional impact of cataract have been reported 53 .In general, women have more fear of falling than men and this fear is believed to play a protective role against falls 54 .The divergence between men and women may be linked to cultural factors, as women (even those who have suffered falls) continue playing their social roles, exercising activities in the home environment 52 .
In the group with cataract, a score of less than 11 points on the SPPB was considered a risk factor for falls.Balance is affected more when visual impairment is greater.Therefore, these factors tend to result in a greater incidence of falls, as reported in two previous studies involving the use of the Berg balance scale 4,55 .In the present investigation, the SPPB was the only measure of physical-functional performance associated with falls in individuals with cataract, perhaps because of its broader scope, as this battery involves muscle strength, balance and gait speed tests.Therefore, among the measures employed, the SPPB is suggested to be the most adequate for the evaluation of older adults with different levels of functioning, as it can differentiate active individuals from sedentary individuals 56 and detect differences in performance between frail and non-frail individuals 57 .In a population-based study, a statistically significant difference on the SPPB was found between recurring fallers and non-recurring fallers 58 .
The present findings and the lack of evidence favoring cataract correction surgery as the only strategy for reducing the risk of falls 59,60 underscore the importance of multidisciplinary follow up of older adults with a diagnosis of cataract to orientate these individuals and implement interventions directed at modifiable risk factors, as multidimensional strategies are effective at reducing the risk and occurrence of falls 59 .Indeed, a multidisciplinary evaluation of falls is fundamental.Certain adverse health conditions can affect one's postural control, such as those that impair the vision (e.g., cataract) or cause sensory deficiencies, as occurs in diabetes mellitus 61 , thereby increasing the risk of falls.In the context of a multidimensional evaluation, it is also important to note that variables related to the use of medications were factors associated with falls in both groups in the present sample, although not achieving statistical significance.
The limitations of the present study regard the cross-sectional design and the impossibility of performing probabilistic sampling.The strengths of this study reside in the stratified analyses to determine the existence of confounding factors and the choice of regression models based on adherence statistics of the models to the data.
The present results suggest that older adults with high levels of functioning may achieve sub-maximum scores on balance tests and still have a high prevalence rate of falls.These findings can contribute to the develop of new ideas regarding the evaluation of active community-dwelling older adults, including the use of multidimensional measures for screening in primary care services 62 .Simple measures in primary care, such as the distribution of glasses and screening with visual acuity tests in the community, could assist in the management of individuals with cataracts 63 .It is also necessary to provide education and training for health professionals that include the promotion of functional health for older adults in their integrality, breaking away from the fragmentation of work by professionals in the field 64 .

Conclusion
The present data suggest that sex and physical performance exert an influence on the risk of falls among community-dwelling older adults with cataract, whereas multimorbidity was associated with falls among those without cataract.
Physical performance tests, including tests addressing postural balance, should not be used as the only means for the identification of older adults at risk for falls, even when the cutoff points are adjusted.
Longitudinal studies are needed to confirm these findings and contribute to the development of multidisciplinary strategies for the evaluation of the risk of falls among community-dwelling older adults with a diagnosis of cataract.

Collaborations
LL Borges, LPS Paz and RL Menezes performed the field work and the latter author served as the adviser of the study.LPS Paz, LL Borges, VRFS Marães, MMF Gomes, MM Bachion and RL Menezes participated in the final draft of the manuscript.

Table 1 .
Characteristics of sample composed of community-dwelling older adults, Federal District, Brazil, 2015.

Table 1 .
Characteristics of sample composed of community-dwelling older adults, Federal District, Brazil, 2015.

Table 2 .
Univariate logistic regression of variables as potential risk or protection factors regarding falls, Federal District, Brazil, 2015.

Table 3 .
Regression models representing falls and associated factors among older adults with and without cataract, Federal District, Brazil, 2015.