Dementia and minor cognitive impairment prevalence

Globally, 46.8 million people live with dementia, and this is expected to increase to 115.4 million by 2050.^{Reference Prince, Wimo, Guerchet, Ali, Wu and Prina1,2} Minor cognitive impairment (MCI), which may later develop into dementia, is also prevalent in up to 42% of the global population over 60 years of age.^{Reference Hu, Yu, Sun, Zhang, Wang and Qin3}

Within the UK alone, there are over 850 000 people with dementia and it has become the leading cause of death.^{Reference Prince, Knapp, Guerchet, McCrone, Prina and Comas-Herrera4,5} By 2025, this number is expected to rise to 1 million.^{Reference Prince, Wimo, Guerchet, Ali, Wu and Prina1} MCI is also estimated in between 5 and 20% of people over 65 years of age in the UK,⁶ with one in ten having a chance of developing further dementia.⁷

Dementia and MCI in minority populations

Because of the impact of globalisation, many Western countries, including the UK, have formed significant ethnic minority populations; currently 14% of the UK population identify as an ethnic minority,⁸ and this is estimated to increase to 20% by 2051.^{Reference Lievesley9} This includes non-English speakers. Currently, there are an estimated 88 languages other than English spoken in the UK, with over 8% of the population not having English as a first language, and over 864 000 people struggling or unable to speak it.⁸

In countries such as the UK, there is also a growing ageing ethnic minority population.^{Reference Richards, Brayne, Dening, Abas, Carter and Price10} Whereas previously many ethnic minorities would return to their home country as they grew older, many people from minority backgrounds who migrated in between the 1950s and 1970s as young adults are now choosing to permanently reside in the UK in their old age.^{Reference Mukadam, Cooper and Livingston11}

These older ethnic minorities are at particularly high risk for receiving a diagnosis of dementia.^{Reference Richards, Brayne, Dening, Abas, Carter and Price10,Reference Chui and Gatz12} They are highly susceptible to risk factors known to increase the likelihood of dementia, showing higher prevalence for diseases and conditions such as diabetes, heart disease, hypertension and obesity.^{Reference Adelman, Blanchard, Rait, Leavey and Livingston13–Reference Seabrooke and Milne15} Furthermore, ethnic minorities are more likely to come from low socioeconomic backgrounds, which is associated with their likelihood of developing dementia.^{Reference Yaffe, Falvey, Harris, Newman, Satterfield and Koster16,Reference Fischer, Yeung, Hansen, Gibbons, Fornazzari and Ringer17}

Cognitive testing in ethnic minorities

For an early and accurate diagnosis of dementia, the use of cognitive tests in the diagnostic process is vital.^{Reference Panegyres, Berry and Burchell18} However, a key limitation of cognitive tests available to us is that they are designed for people from European cultures, fluent in the English language.^{Reference Tuerk and Sauer19,Reference Fillenbaum, Heyman, Williams, Prosnitz and Burchett20}

Therefore, when administering cognitive tests to ethnic minorities, there is reduced sensitivity and specificity; we see higher instances of false positive scores (where dementia or MCI is incorrectly detected by the test in a cognitively healthy person) and false negative scores (where a person's dementia or MCI is not detected by the test) within ethnic minorities compared with their English-speaking European counterparts.^{Reference Regan21–Reference Khan and Tadros24}

To overcome this issues, cognitive tests are often translated into target languages. However, translation alone does not address these increased rates of false positive and false negative scores that may also arise from cultural differences. Ethnic minorities may not be familiar with Western concepts such as questions about the Western calendar, Western names or Western historical and general knowledge, which are used for assessing cognitive domains such as orientation, language and memory tasks.^{Reference Parker and Philp23} Therefore, when assessing for dementia, cultural bias can significantly affect the validity of these cognitive tests.^{Reference van de Vijver and Tanzer25,Reference Van De Vijver26}

Flaherty et al^{Reference Flaherty, Moises Gaviria, Pathak, Mitchell, Wintrob and Richman27} explored this in their research on the five dimensions of cross-cultural equivalence that can be used to minimise cultural bias and measurement errors in cross-cultural testing.^{Reference Flaherty, Moises Gaviria, Pathak, Mitchell, Wintrob and Richman27} These dimensions are content, conceptual, criterion, semantic and technical equivalence, and should be maintained by tests when they are adapted for different language and culture backgrounds.^{Reference Flaherty, Moises Gaviria, Pathak, Mitchell, Wintrob and Richman27}

In the context of cognitive tests for dementia and MCI technical equivalence (how the test is administered is culturally appropriate for the person being assessed, e.g. language the test is administered in) and semantic equivalence (once translated, the meanings of the test questions are the same as the original test questions, e.g. accurate translation of questions) are retained through simply translating the cognitive test into a target language.^{Reference Flaherty, Moises Gaviria, Pathak, Mitchell, Wintrob and Richman27}

However, without accounting for cultural differences as well, there is still a loss of three of the five dimensions. These are content equivalence (the questions of the test are culturally relevant to the cultural context of the person being assessed), criterion equivalence (after accounting for culture, the test as a whole still assesses for dementia and MCI) and conceptual equivalence (after accounting for culture, the individual test questions still accurately assess the cognitive domains they were meant to at the original level of difficulty).^{Reference Flaherty, Moises Gaviria, Pathak, Mitchell, Wintrob and Richman27}

To ensure that all five dimensions of cross-cultural equivalence are maintained, solutions such as altering cut-off scores of cognitive tests for different ethnic minorities and non-English-speaking groups has been suggested.^{Reference Bohnstedt, Fox and Kohatsu28,Reference Milani, Marsiske, Cottler, Chen and Striley29} However, this has still been shown to reduce the sensitivity and specificity.^{Reference Bohnstedt, Fox and Kohatsu28,Reference Milani, Marsiske, Cottler, Chen and Striley29} With developing and piloting new cognitive tests not being the most feasible solution, adapting existing cognitive tests for different languages and cultures has been proposed.^{Reference van de Vijver and Tanzer25,Reference Van De Vijver26,Reference Guillemin, Bombardier and Beaton30}

The Montreal Cognitive Assessment

The Montreal Cognitive Assessment (MoCA) is a cognitive test, developed to screen and assess for dementia and MCI.^{Reference Nasreddine, Phillips, Bédirian, Charbonneau, Whitehead and Collin31} This tool was developed by Nasreddine et al,^{Reference Nasreddine, Phillips, Bédirian, Charbonneau, Whitehead and Collin31} and is currently used to aid physicians primarily in detecting MCI, a state that can progress to dementia. The test lasts 10 minutes and is given a score out of 30. There are 13 questions that assess the cognitive domains attention, fluency, memory, language, orientation, visuospatial abilities and executive functioning (see Table 1).

Table 1 Questions of the Montreal Cognitive Assessment

A validation study conducted in 2005 found that the MoCA is a better substitute for the assessment of MCI and dementia than the Mini-Mental State Examination (MMSE).^{Reference Nasreddine, Phillips, Bédirian, Charbonneau, Whitehead and Collin31} Since its inception, the MoCA has been used to assess for several variations of dementia and associated diseases such as Parkinson's disease, vascular dementia and Huntington's disease.^{Reference Dalrymple-Alford, MacAskill, Nakas, Livingston, Graham and Crucian32–Reference Videnovic, Bernard, Jaglin, Shannon, Fan and Jaglin34}

A recent systematic review of cross-cultural applicability of the MoCA investigated the effect language and cultural diversity have on assessing for MCI, using this cognitive screening tool.^{Reference O'Driscoll and Shaikh35} The review reported a wide range of cut-off scores cross-culturally (both across countries and within regions), making it difficult to identify cut-off scores to differentiate between cognitively healthy individuals and those with MCI, as scores could be considered either cognitively healthy or indicative of MCI depending on which cut-offs are used. Furthermore, the review indicates very few studies having conducted validation studies on their adapted measures, meaning sensitivity and high false positive rates. Thus, the review highlights how it can be difficult to identify recommended adjusted cut-offs according to culture, age and education level.^{Reference O'Driscoll and Shaikh35}

Therefore, as suggested, culturally adapting the existing MoCA would be the best approach to make it suitable for different languages and cultural contexts. In fact, the MoCA is currently widespread in use across over 50 different language and cultural versions.^{Reference O'Driscoll and Shaikh35} Yet, despite there being literature present that highlights the various adaptations of the MoCA and their development, using methods such as the use of global guidelines or the involvement of experts, there is no consensus on which questions must be culturally adapted and why.^{Reference Mirza, Panagioti, Waheed and Waheed36,Reference Maneesriwongul and Dixon37} There is also a lack of standard steps and procedures that can be undertaken to translate and culturally adapt the MoCA.

Culturally adapting cognitive tests

Waheed et al^{Reference Waheed, Mirza, Waheed, Malik and Panagioti38} proposed that every cognitive test should have its own set of specific guidelines on culturally adapting each of its questions. These would be developed through an incorporation of previous literature and feedback from those who have previously adapted it.^{Reference Waheed, Mirza, Waheed, Malik and Panagioti38}

At present, there are only one set of guidelines that have been developed for the cultural adaptation of a specific cognitive test.^{Reference Waheed, Mirza, Waheed, Malik and Panagioti38} These are for the Addenbrooke's Cognitive Examination Version III (ACE-III), a gold standard for the diagnostic accuracy of cognitive impairment and dementia.^{Reference Hsieh, Schubert, Hoon, Mioshi and Hodges39,Reference Cheung, Clugston, Croucher, Malone, Mau and Sims40} These guidelines were developed through a multi-stage qualitative study combining findings from a systematic review on the translation and cultural adaptations of the ACE-III with feedback received via questionnaires from previous adaptors of the ACE-III.^{Reference Mirza, Panagioti, Waheed and Waheed36,Reference Waheed, Mirza, Waheed, Malik and Panagioti38,Reference Mirza, Panagioti and Waheed41}

The findings identified how adaptors of the ACE-III had adapted each test question for their culture, and their rationale for how they applied changes to test questions based on cultural differences. Through these findings, steps were identified for the cultural adaptation of each ACE-III test question, and these were formatted into a set of guidelines that stated which questions of the ACE-III required cultural adaptation and how they had been adapted, with justification for the adaptation process.^{Reference Waheed, Mirza, Waheed, Malik and Panagioti38} As a demonstration of their applicability, these guidelines were then implemented, with input from focus groups, to develop an Urdu version of the ACE-III.^{Reference Waheed, Mirza, Waheed, Malik and Panagioti38}

This Urdu version of the ACE-III was then culturally validated by administering it to 25 cognitively health Urdu speakers, who were then interviewed to determine if the test questions matched their language and cultural context.^{Reference Mirza, Panagioti and Waheed41} Waheed et al^{Reference Waheed, Mirza, Waheed, Malik and Panagioti38} determined that when translating and culturally adapting a test, particularly cognitive tests that are heavily reliant on language and culture, developing cultural adaptation guidelines for the test, implementing them to create a culturally appropriate version of the test and culturally validating the test are all steps that should precede a psychometric validation of the new version of the test.^{Reference Waheed, Mirza, Waheed, Malik and Panagioti38}

Aims

In this paper, we will describe undertaking the initial portion of this methodology to develop guidelines, this time specifically for translating and culturally adapting the MoCA. Because of the widespread availability of its existing adapted versions, it was a suitable cognitive test to develop guidelines for. The research team also consisted of two of the original developers of the ACE-III guidelines (N.M. and W.W.).

We developed the guidelines for culturally adapting the MoCA through a scoping review of the literature, using a systematic search strategy and utilising feedback from those who had adapted the MoCA for their respective cultures and languages. By developing these guidelines, we endeavoured to provide a series of evidence-based instructions that future adaptors of the MoCA can incorporate into their culturally adapted protocols, with evidence to allow for the maintaining of content, criterion and conceptual equivalence.^{Reference Flaherty, Moises Gaviria, Pathak, Mitchell, Wintrob and Richman27}