Introduction

It is estimated that 285 million people worldwide are visually impaired, with blindness affecting 39 million and low vision further 246 million individuals. An estimated 90% of visually impaired individuals live in developing countries, while older persons comprise 82% of individuals living with blindness [1]. The Malaysian National Eye Survey 1996 reported a prevalence of blindness and low vision in the general population as 0.2% and 2.44% respectively [2]. However, the prevalence of blindness and low vision has increased to 1.2% and 5.4% respectively based on the 2014 National Eye Survey (NES 2) with the majority of both groups aged 50 years old and above [3]. The causes of low vision include glaucoma, diabetes, age-related macular degeneration, cataracts, retinal disease and uncorrected refractive errors. The relationship between increasing age and visual impairment as well as its observed increasing prevalence are concerning for a developing country like Malaysia as it is experiencing rapid population ageing [4].

An expanding older population is also expected to lead to an increase in the number of people vulnerable to a decline in health and function or frailty [5]. Frailty has been defined as a clinically recognizable state of increased vulnerability resulting from age-associated decline in reserve and function across multiple physiological systems [6]. Fried and colleagues have proposed an operational definition, or ‘phenotype,’ of frailty based on five criteria commonly used by researchers [7]. The criteria are slow gait speed, low physical activity, unintentional weight loss, self-reported exhaustion and muscle weakness. Individuals who fulfil three or more of these five criteria are considered frail.

Visual impairment in older adults has been associated with increased risk of adverse health outcomes including functional [8] and cognitive decline [9]. However, little is known about the relationship between visual function (vision and contrast sensitivity) and frailty. A study conducted in Hong Kong reported an independent association between visual acuity and frailty detected with the FRAIL scale [10]. The Beaver Dam Eye Study in Australia also found an association between frailty and poor visual function [5]. Another cross-sectional study involving community-dwelling older adults in England reported that an increased risk of developing the state of pre-frail or frail among older adults with poor visual acuity over a four- year follow-up period [11]. The latter study utilized self-reported symptoms of visual difficulties or disease. Another study which included objective determination of visual acuity suggested that those living with visual impairment are three times more likely to be pre-frail and four times more likely to be frail [12]. All but one study used visual acuity as a measure of visual function but not contrast sensitivity. The importance of contrast sensitivity as a influencing factor for the quality of vision is gaining prominence. Previous studies have detected the effects of reduced contrast sensitivity on activities of daily living like difficulty with reading [13], face recognition [14], driving [15], and mobility [16].

Early detection of frailty is vital to prevent complications which will otherwise burden both health and social care systems. This study was conducted to determine the potential relationship between visual function and the development of frailty among community-dwelling older adults.

Methods

This was a cross-sectional sub-study of the Malaysian Elders Longitudinal Research (MELoR) conducted from June 2013 to December 2015. The study population consisted of 8769 individuals aged 55 years and older selected from the electoral rolls of the parliamentary constituencies of Petaling Jaya North, Petaling Jaya South and Lembah Pantai through simple random sampling. Of these, only 5815 participants had a contactable address, and 3334 met the inclusion criteria and had invitations sent out. Written informed consent was obtained. The study was conducted in accordance with the Declaration of Helsinki and ethical approval was granted by the Medical Research Ethics Committee of the University of Malaya Medical Centre (ethics approval number MREC 943.6).

Best-corrected visual acuity at four meters was determined for each eye with the LogMAR chart. Contrast sensitivity was measured binocularly at one meter with the Pelli Robson chart. The luminance of the measurement environment was in accordance with the manufacturers’ recommendations. All participants were literate. Impaired vision was defined as a Log MAR score of less than 0.3 [17]. For contrast sensitivity, a cut-off value of 1.65 was utilized, below which individuals were considered to have poor contrast sensitivity [5]. Participants with impaired vision or poor contrast sensitivity were referred to the ophthalmology clinic for further evaluation and management.

Frailty was determined by the Fried’s phenotype which included the components of weight loss, weak handgrip strength, slow walking speed, reported exhaustion and low physical activity. The presence of weight loss was defined as an unintentional weight loss of ≥ 4.5 kg for the past one year. Grip strength was assessed using the Jamar handheld dynamometer, with measurements obtained from each hand using Roberts protocol [18]. Participants were required to sit with shoulders fully adducted and elbows held at 90 degrees of flexion. The maximum handgrip strength from the six measurements, three from each hand, was used for analysis. Weak handgrip strength was determined using the 25th centile cut-off of less than 15 kg in women and less than 20 kg in men [18]. Walking speed was measured over 15 feet (4.75 m) at usual pace, while a slow walking speed was defined using the 75th centile cut-off of over seven seconds. Exhaustion was defined using a score of 3 or 4 from the Centre for Epidemiological Studies Depression (CESD) scale. Participants were asked the following question: I feel full of energy these days, with the answer categories: 1 'Often', 2 'Sometimes', 3 'Not often', 4 'Never’. A score of 3 or 4 points was defined as positive for presence of exhaustion. Physical activity was measured with the international physical activity questionnaire (IPAQ). The kilocalorie expenditure per week was estimated based on a published formula. Participants were considered to have low physical activity when they engage in less than three days of vigorous-intensity activity per week or less than five days of moderate-intensity activity per week [19]. Frailty was defined as the presence of three or more of the five components. Pre-frailty was defined as the presence of one or two components and non-frail was defined as having none of the five components. The presence of falls was identified through a self-reported history of at least one fall in the preceding 12 months.

Data were analyzed using SPSS version 25 (SPSS, Inc., Chicago, IL, USA). The quantitative variables were summarized in terms of mean and standard deviation while the qualitative variables were summarized as frequencies and percentages. The independent sample t-test and Pearson’s chi-squared test were used to determine differences between group. All significant values were set at p < 0.05. Multinomial regression was used to determine the odds ratio with 95% confidence intervals.

Results

Out of the 1565 participants who completed the interview session for the demographic data in the MELoR study, 1332 participants had complete vision data, 1278 participants had complete contrast sensitivity data and 1274 participants completed frailty assessments.

Compared to participants with normal vision, those with impaired vision were older and more likely to have no formal or primary education, diabetes mellitus (41.3%) and hypertension (38.9%). Significant ethnic differences existed, with Malay ethnicity associated with increased likelihood of impaired vision (43.3%) (Table1).

Table 1 Demographic data according to vision and contrast sensitivity categories
Full size table

In comparison with participants with normal contrast sensitivity, those with poor contrast sensitivity were older, and more likely to have no formal or only primary education and diabetes. A history of falls was significantly more reported among participants with poor contrast sensitivity (Table 1).

Seventy-three (5.73%) participants were classified as frail, 1006 (90.98%) pre-frail and 40 (3.14%) non-frail. No significant relationship existed between frailty and visual impairment (p = 0.241). However, individual component comparisons found slow gait speed (p < 0.001), weak handgrip (p < 00.001) and exhaustion (p = 0.028) to be significantly more common among those with visual impairment (Table 2). Significant associations existed between frailty and impaired contrast sensitivity (p = 0.001) (Table 3). Additionally, four out of five components: slow gait speed (p < 0.001), weak handgrip (p < 0.001), exhaustion (p = 0.032) and low physical activity (p = 0.042) were associated with poor contrast sensitivity.

Table 2 Frailty status and components according to visual performance
Full size table
Table 3 Frailty status and components according to contrast sensitivity status
Full size table

Multinomial regression analysis revealed no significant association between impaired vision with pre-frail or frail compared to non-frail individuals. Poor contrast sensitivity was significantly associated with the presence of frailty (OR 5.34, 95% CI 1.71–16.69). The association remained after further adjustment for age and diabetes (OR 5.13, 95% CI 0.92–19.14). The association was, however, no longer significant when controlled for level of education (p = 0.07) (Table 4).

Table 4 Multivariate models for frailty and contrast sensitivy
Full size table

Discussion

Contrast sensitivity is closely associated with frailty in our population of older urban-dwelling adults. Reduced visual acuity was, on the contrary, not associated with frailty. After controlling for age and diabetes, older adults with impaired contrast sensitivity remained 5.13 times more likely to be frail according to Fried’s criteria. Impaired contrast sensitivity was also associated with increased risk of self-reported falls.

Few studies to date have investigated the association between contrast sensitivity and frailty measures. The Beaver Dam Eye Study found those with greater frailty scores were likely to have impaired vision and reduced contrast sensitivity. The visual function variables utilized within the study are significantly correlated slow gait speed, low expiratory flow rate, poor handgrip strength and inability to perform chair stand. Each component of the frailty phenotype was also significantly associated with both impaired vision and reduced contrast sensitivity [15]. A 10-year follow-up study found an association between frailty and age-related cataract, which is known to affect contrast sensitivity more than visual acuity [20]. In our study, the association between reduced contrast sensitivity and the presence of frailty remained after adjustments for age and the presence of diabetes mellitus. However, the association was attenuated by adjusting for education level, implying a lower education level may account for the increased likelihood of frailty in individuals with poorer contrast sensitivity. The significant association between lower education level and both reduced visual acuity and poorer contrast sensitivity was in agreement with another similar study which established that impaired near visual acuity is associated with frailty [21]. This was an expected finding since lower education level negatively influences healthcare access and awareness [22].

Based on our study, individuals with reduced contrast sensitivity were at greater odds of being frail. This may be explained at a molecular level by the age-related reduction in the number of rods at the parafovea [23]. Contrast sensitivity measures the ability to distinguish between finer and finer increments of light versus dark, as opposed to visual acuity which measures the ability to read increasingly smaller letters on the visual acuity chart. Reduced contrast sensitivity is also associated with diabetes, which has been attributed to structural changes within the inner-retina which also affects the magnocellular and parvocellular pathways [24]. Worsening contrast sensitivity with age may also be due to neural factors like the reduction of retinal illuminance by senile miosis or optical factors such as increased higher-order aberrations [25].

In our study, the association between contrast sensitivity and frailty was independent of age and diabetes status. Contrast sensitivity appears to more accurately reflect real-life conditions from a visual perspective. It not only determines how patients see under high contrast conditions but also under real-world low contrast conditions [26]. Contrast sensitivity has a stronger association with visual tasks entailing distance judgments and mobility than conventional visual acuity testing [27, 28]. As a result, this causes worsening engagement in physical activities, thus contributing to the development of frailty.

Our cross-sectional analysis found no association between reduced visual acuity and frailty. Reduced visual acuity was only associated with slowness, handgrip and exhaustion but not physical activity and weight loss. This was similar to the findings of a large population study conducted in France by Zoler et al. which did not find a significant relationship between frailty and vision impairment [29]. This is in contrast with Swenor et al. who reported a 3.2 times increased likelihood of being pre-frail and frail with visual impairment compared to those without visual impairment [12]. Liljas et al. found participants also found a doubling in risk of becoming pre-frail and frail with visual impairment [11]. In these two studies, poor vision was self-reported while frailty was defined by the combination of pre-frail and frail. A study conducted in community-dwelling older persons in Hong Kong also found an independent association between visual impairment and frailty [10]. The FRAIL criteria rather than the Fried criteria were employed here. The differences in measurements, therefore, limits comparisons between these studies and ours.

The relationship between falls and frailty have been previously established [30,31,32]. Reduced contrast sensitivity, and not poor visual acuity, has also been found to be significantly related to the risk of recurrent falls [33]. The findings in our study are consistent with this previous study which also found that the presence of a history of falls in the past 12 months was significantly more reported in participants with poor contrast sensitivity. Falls are significantly associated with increased healthcare costs, as they may lead to long-term disabilities in the older persons which in turn are associated with increased frailty.

The low response rate within the sample population may limit the representativeness of the study with those who are frail more likely to be excluded, though it is not possible to confirm this suspicion. The Fried phenotype only considers the physical dimension and does not take into account non-physical aspects of frailty such as psychosocial aspects and cognitive impairment. The inclusion of other frailty indices should, therefore, be considered in future such studies. However, the Fried criteria remain the most widely used, hence allowing for direct comparisons with other studies. Potentially confounding variables such as health behaviour were also not controlled within this study. Causal relationship cannot be determined with a cross-sectional study design. The relationship between contrast sensitivity and frailty, therefore, needs to be confirmed in future prospective studies.

In conclusion, reduced contrast sensitivity is associated with physical frailty measured with the Fried criteria. Future studies should seek to determine whether early recognition and treatment of those with poor contrast sensitivity could potentially reduce the burden of frailty in our society.