Abstract
The elderly in Peru face significant barriers in healthcare, notably in detecting cognitive impairment and dementia. These difficulties are exacerbated by the scarcity of validated and standardised cognitive assessment instruments for this age group. The Montreal Cognitive Assessment (MoCA) has proven to be a useful tool for the early detection of dementia, evaluating eight domains of cognitive functions, including: visuo-spatial and executive function, naming, memory, attention, language, abstraction, and orientation. This study aims to standardise the Spanish version of the Montreal Cognitive Assessment (MoCA) for the elderly in Lima, addressing the critical need for culturally and demographically adapted cognitive evaluation tools in Peru. The test was administered to 338 ambulatory and homebound elders from three institutions: San Miguel District Municipality, San José Obrero Polyclinic in Barranco, and EDMECON in Surco. The study provides normative data and cutoff scores for the Peruvian elderly population, facilitating the clinical application of the MoCA in Peru and potentially other Spanish-speaking countries. Our results indicate high orientation scores and low delayed recall performance, possibly highlighting cognitive strengths and weaknesses in our sample. Moreover, age and education significantly influenced cognitive performance, with education being the strongest predictor. We discuss our findings in relation to the use of appropriate cut-off points and considerations of cultural sensitivity relevant to the Peruvian context.
1 INTRODUCTION
Dementia represents a set of pathologies whose common characteristic is the death of neuronal cells, triggering a degenerative and irreversible process affecting the nervous system 1,2. Dementia is a clinical syndrome that involves irreversible loss of cognitive functions, especially memory, significantly compromising daily activities 3. Diagnosis requires a substantial decrease in patient autonomy and quality of life compared to a premorbid stage, not attributable to other medical or psychiatric causes 4. Neurocognitive disorders are among the most prevalent age-related disorders worldwide and lead to substantial morbidity and mortality 5–7.
The prevalence of dementia among individuals aged 60 years and older is approximately 8.5% in Latin America 8 with projections indicating an increase from 7.8 million in 2013 to 27 million by 20509,10. The Global Burden of Disease (GBD) study shows that neurodegenerative diseases in Peru, measured in Disability Adjusted Life Years (DALY), rose from the fifth to the third largest contributor to neurological diseases 11. A recent study in Peru 12,13 found a 6.85% prevalence of dementia in 1,532 individuals over 65 in Lima, with Alzheimer’s disease (AD) as the predominant diagnosis. Plausible factors contributing to these findings include the ageing of the population, poor access to health services, low diagnostic capacity of dementia in primary care services, and limited recognition of symptoms by family members and caregivers 13. Given the recent increase in life expectancy, it is plausible that AD incidence will rise, potentially tripling by 20504,14.
There is no cure for dementia, only drugs to slow its progression 15. For this reason, early diagnosis is beneficial for patients and their caregivers and can result in substantial cost savings for healthcare systems 16. In fact, early diagnosis provides numerous benefits to the patient: (a) greater and more effective adherence to pharmacological and rehabilitative treatments, (b) the possibility of participating in treatments that are still experimental, and (3) the possibility of organising one’s life in relation to the progression of the disease 17.
Diagnosing dementia in Latin America faces critical issues: tests are rarely performed in primary care. This leads to a high number of cases of dementia that go unnoticed and misdiagnosed 18. In this context healthcare workers can identify dementia with a 66% positive predictive value after a few hours of training 18.
Several cognitive screening tools for MCI and dementia 19,20 have low sensitivity for early detection 21. Most tests are complex, needing trained personnel and specialised equipment 22, often unavailable in primary care. Consequently, MCI is frequently unrecognised in primary care 22.
Promoting research on short cognitive tests validated for Peru 23 is essential, similar to neuropsychological test validation projects in other countries (for example, the NEURONORMA projects in Spain 24 and Colombia 25). To date, several short cognitive tests have been validated in the context of Lima (Peru), including the Clock Drawing Test (CDT) 26,27; the Addenbrooke’s Cognitive Examination (ACE) 28; the INECO frontal screening test (IFS) 29 and the memory impairment test (M@T) 30,31. However, the MoCA test (“https://mocacognition.com/“) has not been validated in the Peruvian context.
MoCA is presented as a brief, reliable, and valid test for neurocognitive evaluation given its high degree of sensitivity and specificity for MCI and dementia 22,32. Currently, the MoCA is available in multiple translated versions, including adaptations for individuals with visual impairments, as well as video and telephone formats. These adaptations improve the accessibility and usability of the test, particularly in longitudinal studies, by minimising sample bias and mitigating participant attrition 33. Thus, it arises as a promising alternative for the detection of early cognitive impairment by general practitioners 34. In the Latin America region, there is a tendency to use the original Spanish version of the MoCA test due to the scarcity of studies related to its validity and cross-cultural adaptation. This situation may result in practices that are not in compliance with international diagnostic standards 35. The objective of this study is to standardise the Spanish version of the Montreal Cognitive Assessment Test (MoCA) for the elderly in Lima-Peru, assessing the impact of age, sex, and educational level. In addition, the norms for the Lima elderly population will be developed taking into account age and educational level.
2 METHODS
2.1 Participants
Participants were recruited from three institutions within the Metropolitan Lima: the EDMECON medical centre, the San José Obrero Polyclinic, and the San Miguel Municipality. The ethical approval for the study was granted by the Ethics Committee for Research in Social Sciences, Humanities and Arts of the Pontifical Catholic University of Peru under approval number 145-2023-CEI-CCSSHHyAA. Participation in the study was voluntary and all participants gave their informed consent indicating their understanding of the objectives of the study and their willingness to participate.
The sample consisted of 338 seniors over 60 years of age (122 male, 216 female), including both native Spanish speakers and those who spoke Spanish as a second language (with proficiency exceeding 10 years). Sociodemographic data, including variables such as sex, age, date of birth, years of education, and medical diagnoses, were systematically collected (Table 1). The age was divided into six ranges: 60-64, 65-69, 70-74, 75- 79, and 80+ years after the research conducted by the National Institute of Informatics and Statistics (INEI) of Peru 36. Similarly, years of education were divided into three ranges: 0-6, 7-11 and 12 + years following the education levels reported by INEI 37: Primary, Secondary, and Higher.
Individuals with auditory or visual impairments, psychiatric diagnoses, concurrent cerebrovascular pathologies, intellectual disability, traumatic brain injuries, and substance abuse disorders were excluded from the sample.
2.2 Instrument
Montreal Cognitive Assessment (MoCA) is a validated cognitive screening tool renowned for its efficacy in the early detection of dementia 38. The administration of the test typically takes around 10 minutes. The test evaluates eight key cognitive domains: visuospatial and executive function, naming, memory, attention, language, abstraction, and orientation (see Table 2). The original MoCA scoring system incorporates a correction factor, adding one point to the total score for individuals with fewer than 12 years of education (28). Here, we aim at estimating the adjustments due to age and education for our population.
2.3 Procedure
Contact with participants was conducted within the context of a medical appointment at the three institutions. In the case of the Private Centre Medical EDMECON and the National Polyclinic of San José Obrero, the evaluation was performed as part of the patient’s medical evaluation, and authorisation was requested to obtain the data by informed consent. In the case of the Municipality, the data collection was parallel to the outpatient medical care provided by the institution, where elderly were visited at their homes.
2.4 Data Analysis
We examined the potential effects of age, years of education, and sex in the raw scores of MoCA for the total sample. We evaluated sex differences in the total score and each of the cognitive domains of the MoCA using the Mann-Whitney U tests. We used the Bonferroni 39 correction of p-values to control for the family-wise error rate of multiple comparisons.
We explored associations of age and years of education with the total score and each of the cognitive domains of the MoCA using the Kendall correlation coefficient (τb) and the determination coefficient (r2) (see Table 4). Here, we employed the Bonferroni correction to control for the family-wise error rate for each test. Potential differences in cognitive performance according to age group and educational level were analysed using the Kruskal-Wallis H test and Bonferroni corrected post hoc Dunn tests 40.
Finally, we computed regression-based normed scores of the total MoCA scores adjusted for age and years of education following the procedure described in the NEURONORMA project 24,41. First, raw scores were converted to percentile ranks (Pc). These ranks were rescaled to a truncated normal distribution (M: 10, SD: 3) bounded between 2 and 18 to generate Scaled Scores adjusted for age (SSa). This procedure was used for the entire sample and each age group.
Next, adjustments for years of education were modelled using a linear regression:
The estimated regression coefficient (β) here was employed to calculate Scaled Scores adjusted for age and years of education (SSae) 42:
We selected a cut-off point of 11 years of education because it represents the period of basic education (Primary and Secondary) in Peru. The age and education adjustments estimated from our sample are shown in Tables 5 and 6.
Data preprocessing, visualisations and statistical analyses were performed with Pandas 43, Seaborn 44 and Pingouin 45 software libraries of the Python programming language.
3 RESULTS
The sociodemographic characteristics of the 338 participants in terms of age, education level, and sex are shown in Table 2. The mean cognitive performance of the total sample was 18.2 ± 5.6 for the total MoCA score. The scaled scores for each subtest are shown in Figure 1.
Scaled scores of MoCA subtests.
3.1 Effects of sex, age and education
We did not find significant differences in the total scores of MoCA grouped by sex. However, we observed that men scored higher on identification (U = 10984.5, p = .023) and attention (U = 10540.5, p = .015) than women. No significant differences were found for other cognitive domains (see Table 3).
The results of the correlation analysis for age and years of education are shown in Table 4. Both age and years of education were significantly correlated with the MoCA total score and its subtests. We found that age was a significant negative predictor of the total MoCA score and its subtests. In contrast, we found that years of education were a significant positive predictor of the total score of MoCA and its subtests. Age and years of education explain 12.17% and 24.67% of the total scores of MoCA, respectively.
We found a significant main effect of the age group on the total MoCA scores (H = 40, p < .001). Post hoc pairwise comparisons revealed significant differences between the 80 + (Me = 15, IQR = 7) group and the 60-64 (Me = 22, IQR = 6.5) (p < .001), 65-69 (Me = 21, IQR = 8.75) (p < .001) and 70-74 (Me = 19, IQR = 6) (p = .004) groups. Similarly, significant differences were found between the 60-64 group and the 75-79 (Me = 19, IQR = 10) group (p = .031).
Similarly, we found a significant main effect of the educational level on the total scores of MoCA (H = 79.80, p < .001). Post hoc pairwise comparisons revealed significant differences between the Primary education group (Me = 13, IQR = 6) and the Secondary (Me = 19, IQR = 5) (p < .001) and the Higher Education groups (Me = 22, IQR = 7) (p < .001). Furthermore, significant differences were found between the Secondary and Higher Education groups (p < .001).
3.2 Age and education adjusted norms
Age-adjusted scale scores obtained from our sample are presented in Table 5. Total scores, percentile values (Pc), age-adjusted scaled scores (SSa), and scores for the total sample and the five age groups are presented. Following 42, we suggest cutoff scores to classify cognitive performance in the test based on the following criteria:
Obtaining a scaled score of 10 or more and a Pc of 44-56 (or more) would not be considered impairment.
Obtaining a scaled score of 7 (one standard deviation below the mean) and a Pc of 13-43 would correspond to mild cognitive impairment.
Obtaining a scaled score of 4 (two standard deviations below the mean) and a Pc of 2-12 would correspond to moderate cognitive impairment.
Obtaining a scaled score of 2 (three standard deviations below the mean) and a Pc < 1 would correspond to severe cognitive impairment.
For the total sample, cognitive performance is considered normal or non-impaired when a score of ≥ 17 is obtained.
Years of education significantly predicted SSa (r2 = .239, p < .001) with a coefficient of 0.347. Following 24, we used this coefficient to calculate age and education adjusted scaled scores (SSae) for each combination of SSa and years of education observed in the sample. We present SSae in Table 6.
4 DISCUSSION
The purpose of this study was to standardise the Spanish version of the Montreal Cognitive Assessment (MoCA) for the elderly population in Lima. Standardising this test in Peru is crucial for an accurate cognitive evaluation, considering the high prevalence of neurocognitive disorders 8,13 and the lack of cognitive tools adapted to the sociodemographic and cultural context of the country. Furthermore, early detection of cognitive disorders such as dementia improves patient prognosis, quality of life, and reduces economic and social burdens 3, which is critical for developing countries. We examined the effect of sex, age, and education on MoCA scores. In line with previous research using MoCA 24,42, we found significant effects on overall cognitive performance of age and education, but not for sex. In addition, we found that education predicts cognitive performance better than age.
Ageing leads to functional and structural brain deterioration, associated with cognitive decline and vulnerability to diseases such as dementia 46. Research in biological ageing has identified markers that recognise differences between individuals in their ageing rate and risk of cognitive decline 47,48. A meta-analysis revealed consistent evidence of declines in cognitive abilities such as memory, processing speed, and executive function with age 49. Furthermore, there is evidence of an indirect association of accuracy in working memory tasks and ageing 50. Our results align with the literature: MoCA performance decreases as age increases. However, many measures of biological age rely on associations with chronological age, which may not always be accurate 51.
Our observed effect of education on cognitive performance is consistent with the cognitive reserve theory 52,53. This view suggests that the educational level and other forms of cognitive stimulation can help offset the effects of ageing and brain pathology. A recent analysis of global longitudinal studies of cognitive ageing supports this theory by showing that education reduces the risk of the characteristic cognitive decline observed in dementia 54. Possibly, the development of cognitive skills due to formal education such as abstract reasoning, problem solving, language, and numerical literacy affects the observed performance on the MoCA. These results raise concerns related to the multiple challenges that living below the poverty line can create, such as malnutrition, that significantly affect educational achievement in Peru 55.
We only found sex differences in identification and attention, with men scoring higher than women. Current evidence supports sex differences in specific cognitive tasks, but not in overall cognitive performance: men generally perform better in mental rotation and visual–spatial working memory, while women show advantages in verbal and location memory 56. Similarly, there is evidence suggesting a faster decline in global cognition and executive function in women due to ageing 57. Our results may be compatible with this empirical evidence, showing that women score lower than men in identification and attention, possibly due to faster cognitive decline. However, our sample had an unequal sex distribution and the observed sex differences may be influenced by the cultural context 58.
We provide normative data and suggested cut-off scores for the MoCA adapted to the Peruvian population, fostering its clinical use and interpretation in the country. This pioneering study is the first to standardise the MoCA test in Peru, enriching its scientific landscape, and expanding its impact to other Spanish-speaking countries by setting a new benchmark for future studies focused on the cognitive decline of seniors 24,42,59.
One of the most notable findings of our study was the notably low score on the delayed recall test (semantic memory) and the high score on the (spatiotemporal) orientation test. The observed memory scores could be attributed to sample-specific impairments in the encoding, storage, and retrieval processes, or changes in the interaction between encoding and retrieval 60,61. In general, low delayed recall scores are associated with amnestic MCI 62 and may have negative implications on the daily activities of the evaluated individuals.
The exceptional performance in the orientation assessment indicates increased temporal and spatial acuity in our sample. This suggests preserved cognitive faculties and autonomy, relevant for excluding dementia diagnosis 63, as its loss indicates severe cognitive impairment in late-stage disease 64. The observed high orientation scores may be the result of cognitive reserve accumulated by work activity 65: Peruvians work more years on average than people in other countries and mainly under informal conditions 66. Thus, good orientation skills help maintain workplace competitiveness and secure basic economic subsistence. Future studies could explore the significance and social impacts of these findings.
To our knowledge, this study is the first post-pandemic standardisation of the test. This may explain the remarkably low scores observed in our sample compared to previous MoCA standardizations in Spanish-speaking countries 42,59. COVID-19 is thought to have direct impacts on cognitive functioning 67,68. Empirical evidence suggests that this viral infection produces changes in the orbitofrontal cortex and parahippocampal gyrus (decrease in gray matter), in the primary olfactory cortex (changes in markers of tissue damage) and a reduction in the overall size of the brain 67. Moreover, the implementation of social distancing protocols has had a profound impact on cognitive and physical activities, as well as social interactions. This disruption may have exacerbated levels of anxiety, depression, and stress responses 69,70, and has been associated with impairments in information processing speed and executive functions 71.
Hence, the results of our study are not directly comparable to any study carried out before the pandemic. We stress the need to update the original test standardisation to fit post-pandemic population characteristics. Adaptation to a post-pandemic environment may require adjustments to cognitive assessment expectations and methodologies. Such comprehensive revisions are critical to maintaining the relevance and accuracy of cognitive evaluations in this new era.
The study has limitations that must be taken into account when interpreting its results. Efforts to recruit a representative sample may still have under-represented certain demographics, affecting the external validity. Lack of long-term follow-up also limits the evaluation of the stability of the cognitive score and the prediction of future neurocognitive disorders. Another limitation is the lack of comparison with other cognitive evaluation tools and the absence of participants with a proper neuropsychological evaluation that confirms cognitive health. This affects the concurrent validity and sensitivity of the adapted MoCA compared to other instruments already available, limiting its use in the Peruvian context. Despite these limitations, the study represents an important step toward improving cognitive assessment in the elderly Peruvian population and highlights the need for more research in this field to address the cognitive health needs of the population in Latin America.
Data Availability
All data produced in the present study are available upon reasonable request to the authors.
AUTHOR CONTRIBUTIONS
LB: Conceptualisation, Methodology, Investigation, Writing - Original Draft. EG: Conceptualisation, Methodology, Investigation, Writing - Original Draft. JV: Investigation, Writing - Original Draft. KM: Investigation, Writing - Reviewing and Editing. CM: Investigation, Writing - Reviewing and Editing. RP: Conceptualisation, Methodology, Software, Formal Analysis, Writing - Original Draft. SB: Supervision, Writing - Reviewing and Editing.
FINANCIAL DISCLOSURE
None reported.
CONFLICT OF INTEREST
The authors declare no potential conflict of interests.
SUPPORTING INFORMATION
Additional supporting information may be found in the online version of the article at the publisher’s website.
ACKNOWLEDGMENTS
We acknowledge the research group Creativity, Technology and Talent (CREA TALENTUM), the Computational Cognitive Neuroscience Lab, and the Academic Direction of Social Responsibility (DARS) of the Pontifical Catholic University of Peru, for providing the resources required for this project. We also thank San José Obrero Polyclinic, San Miguel District Municipality and EDMECON, for facilitating access to participants. We thank Tiffany Sandoval for her support in the data collection process.
Abbreviations
- MoCA
- Montreal Cognitive Assessment
REFERENCES
