Content development of the VISION-Cog: a novel tool to assess cognitive impairment in visually impaired older adults in Singapore

Abstract

Objectives

Current cognitive screening and diagnostic instruments rely on visually dependent tasks and are, therefore, not suitable to assess cognitive impairment (CI) in visually impaired older adults. We describe the content development of the VISually Independent test battery Of NeuroCOGnition (VISION-Cog)–a new diagnostic tool to evaluate CI in visually impaired older Singaporean adults.

Design

The content development phase consisted of two iterative stages: a neuropsychological consultation and literature review (stage 1) and an expert-panel discussion (stage 2). In stage 1, we investigated currently available neuropsychological test batteries for CI to inform constructions of our preliminary test battery. We then deliberated this battery during a consensus meeting using the Modified Nominal Group technique (stage 2) to decide, via agreement of five experts, the content of a pilot neuropsychological battery for the visually impaired.

Setting

Singapore Eye Research Institute.

Participants

Stakeholders included researchers, psychologists, neurologists, neuro-ophthalmologists, geriatricians and psychiatrists.

Outcome measure

pilot VISION-Cog.

Results

The two-stage process resulted in a pilot VISION-Cog consisting of nine vision-independent neuropsychological tests, including the modified spatial memory test, list learning, list recall and list recognition, adapted token test, semantic fluency, modified spatial analysis, verbal subtests of the frontal battery assessment, digit symbol, digit span forwards, and digit span backwards. These tests encompassed five cognitive domains–memory and learning, language, executive function, complex attention, and perceptual-motor abilities. The expert panel suggested improvements to the clarity of test instructions and culturally relevant test content. These suggestions were incorporated and iteratively pilot-tested by the study team until no further issues emerged.

Conclusions

We have developed a five-domain and nine-test VISION-Cog pilot instrument capable of replacing vision-dependent diagnostic batteries in aiding the clinician-based diagnosis of CI in visually impaired older adults. Subsequent phases will examine the VISION-Cog’s feasibility, comprehensibility and acceptability; and evaluate its diagnostic performance.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study follows a robust multistep methodology to ensure a valid development process.

• The pilot VISually Independent test battery Of NeuroCOGnition-Cog was developed using both inductive and deductive approaches of iterative expert consultation and literature review to ensure comprehensiveness and clinical relevance.

• The decision-making process was guided by panel consensus method to reduce the risk for bias and assure strong content validity.

• To ensure cultural appropriateness, we changed the protocols of some existing tests.

• The results would be more generalisable if Malay-speaking and Tamil-speaking participants were included.

Introduction

Cognitive impairment (CI) is the fifth-leading cause of disability for the older adult population¹ and imposes a significant physical, psychological and social burden on patients, caregivers, families and society.² In addition, the disease exerts a heavy economic toll at the individual and societal levels, estimated to be US$818 billion per year globally.² With 2 billion people estimated to be aged ≥60 years worldwide by 2050,³ the number of individuals with CI is expected to triple by 2050.⁴ Given this high and increasing prevalence, valid and reliable assessment tools that can accurately diagnose CI are we essential so that appropriate healthcare resources may be designated for the clinical management of older individuals with CI.

Currently, brief cognitive screening tools, such as the Mini-Mental Status Examination (MMSE) and Montreal Cognitive Assessment (MoCA), are widely used to screen patients for CI.⁵ Those with suspected CI may be further evaluated with neuropsychological test batteries, such as the Vascular Dementia Battery (VDB) or Repeated Battery for the Assessment of Neuropsychological Status (RBANS), in clinical settings.^{6 7} However, these tests have vision-dependent items that rely heavily on visual inputs,^8–10 which can overestimate CI in visually impaired individuals.^11–17

Moreover, the comorbid visual impairment (VI) and CI is common, with one in every three older adults with CI also having VI^{18 19}; importantly, patients with VI have been shown to perform better on visually independent cognitive tests.¹¹ As such, many CI screening tools have been adapted for visually impaired individuals by excluding vision-related questions.^{8 20} For example, the MoCA-blind was developed by omitting four vision-dependent items in the visuospatial, executive and naming domains.²⁰ However, as a result of item deletion, these amended tests may have different psychometric and diagnostic properties compared with their original versions, raising concerns over their validity.⁸ Indeed, as compared with the full MoCA version, the MoCA-blind was reported to have only 44% sensitivity in detecting mild CI, notwithstanding its excellent specificity of 98%.²⁰ Despite some previous attempts,^{8 21} there is no validated neuropsychological test battery currently available that is designed specifically for establishing CI among older adults with VI.

Therefore, our aim is to develop and validate the VISually Independent test battery Of NeuroCOGnition (VISION-Cog), a new diagnostic tool to evaluate CI in visually impaired older adults using a multiphase, mixed-methods design. Unlike short CI screening tools such as the MoCA, the VISION-Cog will be a carefully constructed comprehensive battery of neuropsychological tests capable of evaluating both the presence and the severity of CI. In this paper, we report on the content development, via literature review and expert discussion, of the VISION-Cog (phase 1). We hypothesised that at the end of the content development stage, a pilot version of the VISION-Cog would cover relevant domains of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), using a series of culturally relevant and vision-independent tests.

Methods

Overall development of the VISION-Cog

The VISION-Cog is the result of several qualitative and quantitative developmental phases: (1) Content development, in which the content of a pilot test battery was generated via a literature review and consultation with field experts; (2) Pilot study to fine-tune the battery and finalise the VISION-Cog; (3) Determination of the VISION-Cog’s diagnostic performance in a cohort of visually impaired older adults in Singapore and (4) External validation of the VISION-Cog’s diagnostic properties in a different Singaporean population of visually impaired older adults. In this paper, we describe the content development phase (phase 1), which consisted of two iterative stages, including a neuropsychological consultation and literature review (stage 1) and an expert panel discussion (stage 2).²²

Patient and public involvement

Patients or the public were not involved in the design, conduct, reporting or dissemination plans of our research.

Stage 1: neuropsychological consultation and literature review

As cognition is a complex concept that encompasses multiple neurocognitive domains,²³ including memory, language, executive function, complex attention, perceptual-motor and social cognition, the construction of the VISION-Cog was guided by both inductive and deductive approaches. Our deductive approach consisted of a comprehensive literature review and subsequent consideration of cognitive domains and tests proposed in existing neuropsychological batteries, which provided us with a conceptual framework with which to develop the VISION-Cog. Concurrently, our inductive approach, where we were guided by the clinical experience of our expert clinical psychologist (KD) who performs neuropsychological assessments in a public health setting, gave us the flexibility to remove, add or adjust cognitive domains and neuropsychological tests to cater to the unmet clinical needs of psychologists and physicians in the Singapore context.

The comprehensive literature review of available neuropsychological batteries used to assess CI in older adults in Singapore was conducted using the PubMed database from inception to 13 October 2020. The core keywords included neuropsychological battery and cognitive impairment and Singapore, among others. Our full search strategy is reported in online supplemental table S1. The bibliographies of included articles were handsearched to obtain additional relevant papers. Papers were included if they (a) evaluated CI using neuropsychological batteries and (b) included patients from Singapore. The literature review did not exclude any article type (eg, review) or study design (eg, cross-sectional). Articles were excluded if they (a) used cognitive screening tools (eg, MMSE, MoCA) only, (b) assessed other psychiatric disorders (eg, depression, schizophrenia) or (c) reported on a paediatric population. TAV, EF and PG assessed the titles and abstracts according to the predefined inclusion and exclusion criteria. If information was inadequate in the abstract, the full-text articles were retrieved for further assessment. Data extraction was performed by TAV and checked for accuracy by EF and PG. Details of the content of neuropsychological batteries, including cognitive domains, neuropsychological tests and test duration, were obtained from the articles and bibliographies. At the end of stage 1, 16 potential neuropsychological tests and their protocols (see the Results section for a full description) were compiled, ready for stage 2.

Stage 2: expert panel discussion

The first consensus meeting with a multidisciplinary expert panel was held online via zoom on 18 May 2021 for a 3-hour period. The purpose of the meeting was to agree on a neuropsychological battery suitable for assessing cognitive status in visually impaired older adults. The expert panel consisted of one neuropsychologist, one neurologist, one neuro-ophthalmologist, one geriatrician and one psychiatrist/public health expert. A modified nominal group technique (NGT), which is a structured face-to-face group interaction that allows opinions from each participant to be considered by the rest of the panel members, was used to conduct the consensus meeting.²⁴ The experts did not participate in any planning related to the study before the consensus meeting.

Before the meeting, a demonstration of the 16 potential neuropsychological tests was prerecorded and sent to the panel for review, along with the preliminary list of appropriateness criteria outlined in the Results section of stage 1. The experts were required to privately note down their comments on (1) the suitability of the appropriateness criteria, (2) the feasibility and acceptability of the 16 tests and (3) potential test modifications needed.

During the meeting, the experts were invited to propose additional appropriateness criteria and tests for consideration. The panel subsequently discussed, voted and agreed on a final list of appropriateness criteria of the VISION-Cog. Moving on, each neuropsychological test was discussed in turn and additional tests were proposed. Based on the discussion, the experts were asked to priority rank each of the tests separately for each domain of cognition and each method of administration (tactile or auditory) via an anonymous survey. Only domains (memory, executive function and attention) with more than one tactile-dependent or one auditory-dependent test were included in this ranking exercise. The priorities were converted into ranking scores, with the highest priority receiving a score of 1, the second-highest priority a score of 2, etc. Subsequently, all ranking scores for each test were tabulated to achieve a final total score. The tests were then compiled into a final list with the lowest total scores equivalent to having the highest priorities.

After the tabulated ranking results were shown to the panel during the second round of discussion, each panel member was required to vote on (1) the inclusion of the test and (2) any modification needed. The three voting options for test inclusion were ‘Yes’, ‘Yes but optional’ and ‘No’. If there was disagreement, a further discussion took place to resolve the discrepancies and find a compromise, followed by another vote. In this second voting round, the content validity ratio (CVR) was used to quantify consensus. The formula of CVR is (N_e–N/2)/(N/2), in which N_e is the number of experts answering ‘Yes’ and N is the total number of experts.²⁵ CVR varies between 1 and −1 with a higher score indicating better agreement between panel members. As our expert panel consisted of five members, a minimum CVR value of 0.99, or unanimous consensus, was required.²⁵ This process was repeated for every test until the expert panel reached an agreement on the optimal test battery. Finally, bilingual experts (n=4) fluent in both English and Mandarin were asked to consider if any of the tests needed specific cultural adaptation to ensure they retained content validity.

Based on the experts’ suggestions, we formulated the modified spatial memory test (MSMT) to include three subtests, immediate recall, delayed recall and delayed recognition. In the immediate recall and delayed recall subtests, the participants were asked to recall the locations of nine target items (cork, lighter, key, coin, bottle opener, clothes peg, safety pin, screw and padlock) placed in a wooden case immediately and after approximately 15 min, respectively. To ensure that the test remained meaningful and relevant in Singapore, we sought to identify culturally appropriate replacement target items (online supplemental appendix 1).²⁶ The items were subsequently purchased from commonly available shops in Singapore and presented to a group of 45 cognitively normal older adults whose primary language was either English (n=23) or Mandarin (n=22). All participants were blindfolded to remove visual input. For each item, participants were then instructed to use both hands to feel the item. An accurate response was defined as either (1) correct naming of the item or (2) correct description plus agreement with the name given. A predefined threshold for a culturally appropriate item was deemed as ≥80% accurate responses. For the MSMT-delayed recognition, participants were required to correctly recognise the 9 target items from a list of 18 items, including 9 targets and 9 distractors, given in random order. Selection of appropriate distractor items was performed in biweekly consensus meetings attended by psychologists (KD and SYQ) and clinical researchers (TAV, EF, PG and ELL). The above processes were performed iteratively until we formed nine suitable target items for the Immediate Recall and Delayed Recall subtests and nine suitable distractors for the Delayed Recognition subtest. Details of the study protocol are outlined in online supplemental appendix 1.

A postmeeting discussion was conducted via email to deliberate inquiries raised by the experts during the first consensus meeting. As a result of this feedback, the study team identified an additional test, spatial analysis and provided online supplemental data (test demonstration, pilot study results) to the expert members to aid their final decision. To come to the final decision, each panel member was required to discuss the suitability of proposed tests and subsequently vote on (1) inclusion of the test and (2) any modification needed. This process was completed iteratively until a consensus agreement was reached and no further issues were identified. At this stage, the pilot VISION-Cog was deemed to have optimal content validity.

Results

Stage 1

Several fundamental decisions were made in consultation with our expert psychologist which subsequently formed a preliminary list of appropriateness criteria of the VISION-Cog. First, the VISION-Cog should include the DSM-5’s domains of cognition, consisting of memory and learning, language, executive function, complex attention and perceptual-motor abilities,²³ with each domain containing at least one tactile-dependent and one auditory-dependent test. The social cognition domain in DSM-5 was not included in the VISION-Cog because valid auditory or tactile tests to evaluate this domain are not currently available. Like existing visually dependent batteries (eg, VDB and RBANS), the removal of the social cognition domain should not curtail the validity of the VISION-Cog. Second, the target participants should be community-dwelling older Singaporean adults aged ≥60 years who are visually impaired. Third, to achieve successful implementation, the test battery must be feasibly performed by older adults within 60–90 min.

The comprehensive literature review identified 101 articles, of which 88 full-text articles were retrieved (figure 1). A total of 57 articles were subsequently accepted according to our inclusion and exclusion criteria. From these, 12 neuropsychological batteries were extracted (online supplemental table S2). With the results of our literature search and discussion with our psychologist, we proceeded to develop the VISION-Cog by adapting the two most updated and commonly used batteries in clinical settings in Singapore, the VDB and RBANS.^{27 28} Because the VDB and RBANS contain tests that are vision-dependent, we then performed another literature search to locate appropriate tactile-dependent or hearing-dependent neuropsychological tests as potential replacements. We discovered 27 suitable tests, including 24 tactile and 3 auditory tests (online supplemental table S3). After further discussion (online supplemental appendix 2), we selected and purchased 16 potential tests for our preliminary list (table 1). No new psychometric assessments were performed as prior studies have shown acceptable reliability for all of the tests included in the battery.^{29 30}

Figure 1.

Data extraction flow chart.

Table 1.

Preliminary list of neuropsychological tests

Domain	Tactile-dependent tests	Auditory-dependent tests
Memory	Spatial memory test Pattern board test	List Learning, list recall and list recognition Story memory and story recall
Language	Adapted token test	Semantic fluency
Executive function	Block design (no time bonus) Pattern of search	Verbal subtests of the frontal battery assessment
Complex attention	Tactual formboard test Digit symbol Block design (with time bonus)	Digit span forwards Digit span backwards Auditory detection
Perceptual-motor	Object assembly	Left-right orientation test

Stage 2

At the end of the first consensus meeting, the expert panel unanimously agreed on a list of appropriateness criteria for the VISION-Cog, describing battery content, target population, test duration and test material (table 2). For the battery content, the panel commented that, unlike other audio-spatial animals, humans did not depend heavily on hearing capability to locate positions in space.³¹ This meant that the auditory-dependent test in the perceptual-motor domain was unnecessary, resulting in only tactile-dependent tests required for this domain. To shorten the test duration, expert members suggested simplifying test instructions and imposing discontinuation rules where appropriate (eg, if the participant cannot complete the easy part of the test, do not proceed to the more difficult parts).

Table 2.

Appropriateness criteria of the VISION-Cog

1	The test battery should include the DSM-5’s domains of cognition, consisting of memory and learning, language, executive function, complex attention and perceptual-motor abilities, with the exclusion of social cognition. Each domain should contain at least one tactile-dependent and one auditory-dependent test—except for the perceptual-motor domain, in which only tactile-dependent tests are required.
2	The target participants are community-dwelling Singaporean/Singapore permanent residents aged ≥60 years who have visual impairment which is defined as near binocular visual acuity of worse than N8 according to the US definition.
3	To achieve successful implementation, the test battery must be feasibly performed by older adults within 60–90 min.
4	Tactile materials should be easily reproducible, available and standardised across clinical settings.

DSM-5, Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; VISION-Cog, VISually Independent test battery Of NeuroCOGnition.

After that, the panel ranked, discussed and voted on the inclusion of the 16 preliminary neuropsychological tests (table 1). The ranking and voting results are shown in online supplemental table S4. Of the 16 originally proposed tests, a total of 9 neuropsychological tests were selected for the test battery. These include the spatial memory test and list learning, list recall and list recognition for the memory domain, adapted token test and semantic fluency for the language domain, block design (no time bonus) and verbal subtests of the frontal battery assessment for the executive function domain, and digit symbol, block design (with time bonus),³² digit span forwards and digit span backwards for the complex attention domain. For these nine tests, modifications were also suggested by the panel to ensure clarity of instructions and cultural relevance (online supplemental table S5, tests no. 1–9). The modification process is shown in details in online supplemental appendix 2.

One of the outcomes of this modification process was the MSMT; it comprises three subtests: immediate recall, delayed recall and delayed recognition. Response accuracy of piloted items for the MSMT—immediate recall and—delayed recall subtests ranged from 30% to 100% (online supplemental table S6). Elaboration on items that failed the threshold of cultural appropriateness (response accuracy <80%) and their reasons are outlined in online supplemental appendix 2. The nine culturally appropriate items (response accuracy ≥80%) identified in this pilot study consisted of cork 1, key 1, lighter 2, clothes peg 2, bottle opener 2, coin, padlock, screw and pencil. The setup of the immediate recall and delayed recall subtests is represented in figure 2. Next, the nine selected distractor items for the delayed recognition subtest are shown in online supplemental table S7. They were selected so that each distractor and the corresponding matched memory item have slight differences in either shape, texture or function only.

Figure 2.

Setup of the spatial memory—immediate recall and—delayed recall subtests.

Of the original 16 proposed tests, 7 tests were excluded due to time constraints, excessive cognitive burden and unreasonable dexterity challenges (online supplemental table S8). This resulted in a lack of tests in the perceptual-motor domain. Thus, we searched for other tactile tests for this domain and subsequently proposed the spatial analysis test designed by Dial et al.^{33 34} This tactile test consisted of four subtests: shape matching, puzzle construction, form matching and form matching/size transformation. We presented the test to the panel which in turn suggested additional modifications to improve the workability of this test for our consideration (online supplemental table S5, tests no. 10). We subsequently conducted a pilot study on two older adult participants to obtain their perceptions on the modifications of the modified spatial analysis test, as advised by the panel (online supplemental appendix 2). The panel discussed and agreed with full consensus that two subtests, shape matching and puzzleconstruction, were suitable to examine the perceptual-motor domain. Moreover, the other two subtests, form matching and form matching/size transformation, were acceptable to replace the block design test to be the test of choice in the executive function domain. Furthermore, the spatial analysis test (with time bonus) was deemed to be a suitable substitution of the block design (with time bonus) to examine the complex attention domain. At the end of stage 2, the pilot VISION-Cog comprised nine tactile or auditory tests spanning five cognitive domains of the DSM-5 (table 3). The full protocols of these tests are outlined in online supplemental table S9.

Table 3.

Included neuropsychological tests in the pilot VISION-Cog

Domain	Tactile-dependent tests	Auditory-dependent tests
Memory	Modified spatial memory test	List learning, list recall and list recognition
Language	Adapted token test	Semantic fluency
Executive	Modified spatial analysis (form matching and form matching/size transformation)	Verbal subtests of the frontal battery assessment
Complex attention	Digit symbol Modified spatial analysis (with time bonus)	Digit span forwards Digit span backwards
Perceptual-motor	Modified spatial analysis (shape matching and puzzle construction)

VISION-Cog, VISually Independent test battery Of NeuroCOGnition.

Discussion

Following an iterative two-stage process of neuropsychological consultation, literature review and expert panel discussion, we have formulated a list of appropriateness criteria describing the VISION-Cog’s content, target population, duration and material. More importantly, using inductive and deductive approaches, we have successfully selected nine relevant tests for the pilot VISION-Cog and identified suitable test modifications. Overall, phase 1 of the developmental process resulted in a pilot VISION-Cog with robust content validity. The VISION-Cog will be pilot tested, fine-tuned and validated in subsequent phases to become a diagnostic instrument capable of replacing vision-dependent neuropsychological batteries and supporting the clinician-based diagnosis of CI in visually impaired older adults.

The VISION-Cog comprises nine tests (table 3) spanning five domains of cognition (memory, language, executive function, complex attention and perceptual-motor) as defined by the DSM-5. In contrast, the VDB and RBANS^{7 28} have an excess focus on memory tasks and do not have formal tests for executive dysfunction. Thus, both batteries are commonly supplemented with the Stroop Test and Trail Making Test to examine executive function by clinicians. In view of these limitations of the VDB and RBANS, the VISION-Cog has been adapted from these two batteries and supplemented with more suitable tests for visually impaired participants so that the VISION-Cog can be more comprehensive and better suited for visually impaired participants than the VDB and RBANS. Therefore, in assessing DSM-5’s cognitive domains, our VISION-Cog is more comprehensive than the VDB and RBANS. Moreover, the expert panel functioned to ensure that the VISION-Cog has simple instructions, appropriate levels of difficulty, and a manageable duration of assessment so that it can be performed comfortably by older adults. This is indispensable because other commercially available visually independent batteries are cognitively demanding for older adults as they are intended for the vocational assessment of young adults.^{29 33}

We used the modified NGT in the first consensus meeting, which has been conducted previously to seek agreement on the content of other topics such as fall battery assessment and electronic health records, to empower the generation and discussion of opinions by all participants.^{24 35} As compared with the informal consensus method in which members may not comfortably share their views due to the influence of group dynamics, the modified NGT has an important advantage of enriching idea generation equally among the experts.³⁶ Moreover, as our expert panel was selective with a comprehensive mix of various expertise and specialties, the modified NGT is superior to the Delphi method in the efficiency of eliciting consensus.³⁶

Due to the social restrictions imposed by the Singapore government during the COVID-19 pandemic, instead of using the traditional face-to-face meeting approach, the modified NGT was conducted online. To ensure the online method did not threaten the validity of the modified NGT approach (eg, difficulty in recognising non-verbal communication cues, and reduced involvement of participants), we allocated ample resources to: (1) prepare the online meeting; (2) familiarise with online-meeting technology and (3) train the facilitator to manage group interactions. For example, before the meeting, we sent the test demonstration videos, recording forms and instruction guides to our expert panel via Google Drive, Google Docs and email platforms so that they could view the videos and note down their ideas efficiently. In addition, trial sessions were simulated beforehand for the facilitator to recognise and troubleshoot technical challenges, incorporate various modes of presentation (eg, Google Docs, PowerPoint and Zoom), and maintain interactions between participants. From our experience, the online approach had three key advantages: (1) facilitation of participation across locations; (2) elimination of travelling time and (3) reduced venue costs associated with traditional face-to-face meetings. However, some challenges such as intermittent Internet connection and unfamiliarity with zoom functions remained at the start of the meeting. These issues were successfully solved afterward and did not affect the results of the meeting. As the acceptance of electronic communication resources has sharply increased as a result of the COVID-19 pandemic, the online NGT may be a promising trend in future research.³⁷

The strengths of our study include a multistep methodological design to ensure a robust development process. Moreover, the VISION-Cog development was guided by both inductive and deductive approaches of iterative expert consultation and literature review to ensure comprehensiveness and clinical relevance. Furthermore, our use of panel consensus to guide the decision-making process reduced the risk for bias and assured strong content validity. Nevertheless, some limitations must be acknowledged. First, to ensure cultural appropriateness, we changed the protocols of some existing tests, which may have altered their psychometric properties. Hence, subsequent phases of the study will entail an evaluation for the psychometric properties of this battery to confirm its validity. Second, we did not include Malay-speaking and Tamil-speaking participants in our study. This may reduce the generalisability of our VISION-Cog in these populations. Future studies will examine the cultural adaptation of this battery for the Malay, Tamil and non-Asians populations.

In summary, our pilot five-domain nine-test VISION-Cog developed using a systematic two-stage process of literature review, expert opinion and consensus meeting may have the potential to replace vision-dependent neuropsychological batteries and aid the clinician-based diagnosis of CI in visually impaired older adults. The next step is to pilot-test this version in a sample of visually impaired older adults and assess its feasibility, comprehensibility and acceptability. Future work will explore its diagnostic potential.

Data availability statement

Data are available on reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

The study received ethical approval from the SingHealth Institutional Review Board (CIRB, Reference #2016/3089). All protocols were performed at the Singapore Eye Research Institute’s research clinic according to the Declaration of Helsinki principles. All participants were recruited from the PopulatION HEalth and Eye Disease PRofile of Elderly Singaporean (PIONEER) study, which is a population-based, cross-sectional study in Singapore that comprised older adults aged ≥60 years of Chinese, Malay, and Indian ethnicities living across Singapore. Written informed consent was obtained from all participants before participation in the VISION-Cog study.