A data-driven, multivariate exploration of subjective cognitive concerns and cultural factors in mild cognitive impairment

Kristoffer Romero, University of Windsor, Windsor, Canada
Rebecca Nurgitz, University of Windsor, Windsor, Canada
Prathiba Shammi, Sunnybrook Health Sciences Centre, Toronto, Canada


Given the role of subjective cognitive complaints in diagnosing mild cognitive impairment (MCI), there is a growing body of literature delineating the specific associations between subjective cognitive impairment and performance on objective neuropsychological tests. However, few studies have explored how cultural factors relevant to people living in a diverse community (i.e., English as a second language, country of birth) impact cognitive performance and subjective estimates of cognition in older adults: If cross-cultural differences exist, there may be differential patterns or “profiles” of subjective cognitive concerns in older adults as a function of culture. In this study, rather than examine racial/ethnic group differences on standard questionnaires of subjective cognitive impairment, we conducted a retrospective chart review, examining the cognitive complaints that patients themselves reported during clinical interview using multiple correspondence analysis. By using the data provided by patients in their own words, we sought to reveal new insights into how cultural factors may interact with cognitive performance in those with MCI.

Participants and Methods:

Neuropsychological reports of 151 patients with a diagnosis of mild cognitive impairment or vascular cognitive impairment - no dementia from a tertiary memory clinic in Toronto, Canada were included in the study. For each report, we coded the presence or absence of subjective cognitive complaints grouped by cognitive domain (i.e., memory complaints, attention complaints, orientation, etc.). Culturally relevant variables such as country of birth, English as a second language, and years spent in Canada, as well as demographic variables, neuropsychological test scores, and scores on the Beck Depression Inventory were also extracted. We analyzed the subjective cognitive concerns using multiple correspondence analysis, a data reduction technique that derives latent factors and displays the data in terms of these dimensions in a common space. The result is a graphical representation of which cognitive complaints tend to be reported together, resulting in certain “profiles”. We also projected the culturally relevant, demographic, and neuropsychological test variables into the space as supplementary variables, which allowed us to see whether these variables tend to be differentially associated with certain profiles.


The multiple correspondence analysis yielded 4 dimensions that explained 61% of the variance. Of note, the first profile of cognitive complaints consisted of memory, attention, and executive functioning complaints, whereas another profile was characterized by language and visuospatial complaints. Interestingly, having English as a second language and country of birth showed modest associations with different profiles. Not surprisingly, age, estimated premorbid IQ, and levels of depressive symptoms were significantly correlated with the different profiles.


These results suggest that subtle cross-cultural differences may exist in terms of the pattern of cognitive complaints that older adults report, depending on their level of English proficiency and whether they were born in North America. Such findings may eventually help develop better tools to characterize subjective cognitive concerns more accurately for patients living in diverse, multi-cultural environments.

Category: MCI (Mild Cognitive Impairment)

Keyword 1: memory complaints
Keyword 2: mild cognitive impairment
Keyword 3: diversity