Background

Dementia is a leading cause of disability in people older than 65 years worldwide, including China, which induces huge challenges for policy makers, healthcare professionals, and family members. Considering no effective treatment for dementia, as well as brain pathology which begins years before onset of objective cognitive symptoms and may be irreversible by the time of diagnosis, many investigators have shifted their focus toward delaying dementia in persons who are in preclinical phases of the disease. Mild cognitive impairment (MCI), referring to cognitive decline from a previous level of functioning both subjectively and by objective evidence, represents the preclinical, transitional stage between healthy cognitive aging and dementia, and affects 10–15% of the population over the age of 65. Although 20–30% of persons with MCI will revert to normal at subsequent follow-up, there is a 5–10% annual rate of progression to dementia in those with MCI, which is much higher than the 1–2% incidence per year among the general population. Moreover, it has been suggested that approximately 50% will progress to dementia in 5 years. MCI represents what researchers and clinicians regard as a “window” in which it may be possible to intervene and delay development to dementia. It is thus imperative to screen for MCI and clarify potential influencing factors for MCI in old population at risk in large-scale study in efforts to improve cognitive functioning and delay progression to dementia.

In addressing cognitive screening tools, the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA) are the most commonly used methods in cognitive impairment detection in both clinical and research fields. It was widely identified that MoCA was superior to MMSE in the detection of MCI as the MMSE had lower sensitivity among multiple study settings. Furthermore, the MoCA showed differences in cognitive profile even in those performing in the normal range on the MMSE and would appear to be a useful brief tool to assess cognition in those with MCI, particularly where the ceiling effect of the MMSE is problematic. Similar studies were carried out in China, however, these studies were done in single region with small sample size, thereby suffered from a lack of representation and reliablity and studies in particular to compare the MMSE and MoCA in the detection of MCI among community-based samples are rare. Therefore, studies in multiple regions are further warranted to confirm the concordance between MMSE and MoCA in the identification of MCI, which may yield different and novel findings because of large sample size.

Additionally, to understand potential and modifiable risk factors to cognitive complaint is to some extent crucial for defense, treatment and intervention in the precarious state of MCI, thereby may delay progression to dementia. Researches to date have identified several factors, such as age, gender, educational and occupational attainment, marriage, income, psychological well-being, physical exercise, social engagement, diet and history of chronic diseases, but some of these findings were controversial, which might be attributable to varied countries of study origin, and the heterogeneity in research methods, including the age range included and the use of different cognitive assessment methods and diagnostic criteria. Especially education had strong influence on MMSE and MoCA performance, and the unpredictable effects of those with more education performing poorer relative to those with less education was observed. It is also necessary to distinguish whether there is disparity in potential factors for cognition when applying different cognitive screening tools to the same population.

Taken together, present study aims to determine the correlation and agreement between MMSE and MoCA in detecting MCI, and to test their differences in influencing factors for MCI among Chinese middle-aged and older population attending baseline survey of the Community-based Cohort Study on Nervous System Diseases in urban and rural areas of four provinces. Findings in this study may yield profound implications for the selection of cognitive measures and MCI management.

Methods

Study population

Data in the present study were derived from the baseline of the Community-based Cohort Study on Nervous System Diseases, an ongoing and longitudinal study established in 2018–2019 by National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention, which focused on potential factors associated with risks of three nervous diseases, including epilepsy for subjects aged > 1 year, and Alzheimer’s disease (AD) and Parkinson’s disease in ≥55 year-old population. Participants without such diagnosed diseases at enrollment were recruited using a multistage stratified random sampling approach in Hebei, Zhejiang, Shaanxi and Hunan province, respectively. Two cities and two counties were randomly selected in each province. Urban and suburban neighborhoods within the cities, and townships and villages within the counties were selected randomly. In each community, all members meeting the inclusion and exclusion criteria of any of three nervous diseases in a randomly selected household were interviewed. Protocol of this project was reviewed and approved by Medical Ethics Committee of National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention (No. 2017020, 6 November 2017). And written informed consent was obtained for each participant.

Present study targeted at subjects recruited in the cohort of AD. The eligible samples for inclusion were (1) 55 years old and older, (2) resident population living in the sampled community, (3) absent of clinically diagnosed AD, and (4) free of comorbid conditions that could affect assessment, such as congenital or acquired mental retardation, diagnosed MCI, and visual/hearing abnormalities even after correction. Subjects with completed data of sociodemographic characteristics, disease history, cognitive examination, psychological evaluation, and survey of basic abilities of daily living were selected to participate in the present study. According to the definition of MCI, we excluded subjects because of their inability to perform basic activities of daily living involving eating, dressing, bathing, toileting, grooming, transferring bed or chair, walking across a room, and urinary or fecal continence (n = 71). For participants locating in the part of <P₁ or > P₉₉ (P: percentile) of sleep duration distribution in each age group, we used the corresponding P₁ and P₉₉ to replace those of <P₁ or > P₉₉, respectively. Finally, a total of 4923 participants were involved in the analysis.

Cognitive assessment

All participants underwent cognitive assessment using Chinese version of the MMSE and the MoCA in present study. Both instruments were valid and reliable among Chinese by taking cultural and linguistical differences into account. MMSE and MoCA were conducted strictly face to face following the guidelines and protocols by trained investigators and were completed during 5–10 min and 10–15 min, respectively.

The MMSE is a 30-point questionnaire used extensively in clinical and research settings to measure cognitive impairment, including simple tasks in a number of areas: the test of time and place, the repeating lists of words, arithmetic such as serial subtractions of seven, language use and comprehension, and basic motor skills. The MoCA is another 30-point test covering eight cognitive domains, and details on the specific MoCA items had been introduced by Nasreddine et al.. The cultural and linguistic modifications of MoCA Beijing version we used from the original English version were also concretely described.

Cognitive function of different domains were evaluated according to items of each test. Details on the components and corresponding maximum scores for each domain were shown in Table 1. The sum of included item points was the subscore of cognitive domain. Dysfunction of cognitive domain was defined as any incorrect test of included items, and cutoffs were listed in Table 1.

Table 1. Cognitive domains assessed by the MMSE and MoCA

The sum of all item points produced total scores of MMSE and MoCA, respectively, ranging from 0 to 30. A higher score indicates better cognitive function. When the education years of the participants were no more than 12 years, 1 point was added on their MoCA total score (if < 30). MCI was identified using education-specific cutoff points of total scores of MMSE and MoCA, respectively. MMSE ≤19 for illiterate individuals, ≤22 for participants with elementary school education, and ≤ 26 for those with middle school education and above. According to Chinese MoCA norms, ≤13 for illiterate individuals, ≤19 for individuals with 1–6 years of education, and ≤ 24 for those with 7 or more years of education.

Sociodemographic and health-related characteristics

Questionnaires were used to collect information on age, gender, educational level, current employment status, household income, residence area, current smoking, alcohol intake during last year, sleep duration covering daytime napping and full-night sleep, and disease histories of hypertension, diabetes, stroke and myocardial infarction by trained investigators. Additionally, a self-report assessment to identify depression in the elderly was performed using the Geriatric Depression Scale (GDS) 30-point version, and depression was defined if GDS score > 11. All these parameters were further grouped for data analysis (age: 55–64, 65–74 and ≥ 75 years; gender: male and female; educational level: below elementary school, elementary school, middle school, high school and above; monthly household income per capital: < 1000, 1000–3999, and ≥ 4000 Chinese yuan; statuses of current employment, current smoking and alcohol intake last year: yes and no; residence area: urban, suburban, county town, and village according to the administrative divisions; sleep condition: yes and no depend on if meeting age-specific sleep duration recommendations; disease history of hypertension, diabetes, stroke or myocardial infarction: yes and no; and depression: yes and no).

Statistical analysis

Continuous variables were presented as mean ± standard deviation (SD) and median, P₂₅ and P₇₅ were also calculated in order to evaluate presence of ceiling/floor effect in MMSE and MoCA tests, while categorical variables were expressed as n (%). Because of the non-normal distribution, non-parametric Wilcoxon rank-sum test or Kruskal-Wallis analysis was performed to test differences in distribution of MMSE or MoCA total score by sociodemographic and health-related factors. If the difference was significant among three subgroups and above, multiple comparison was conducted by Student-Newman-Keuls. The percentage for the relative standard deviation (RSD%) [(SD/mean) × 100] was calculated to examine inter-individual variance of the MMSE and MoCA total scores in the whole population, assuming that greater RSD% indicates better detection of cognitive heterogeneity of the sample. The MMSE and MoCA RSD% index obtained were further compared by means of Wilcoxon signed-rank test. Prevalence of MCI by various factors was compared by Chi square test and Cochran-Armitage trend test if appropriate. And trends in proportions of subjects with MMSE-identified cognitive domain dysfunction across subscore strata of corresponding MoCA cognitive domain were also analyzed by Cochran-Armitage trend test. Scatter plot and Spearman correlation coefficient were applied to explore the correlation between MMSE and MoCA total scores. The agreement between MMSE and MoCA to detect MCI was obtained by Kappa value. Multiple logistic regression was employed, with MCI (yes vs. no) as dependent, and age, gender, employment status, household income, residence area, smoking, sleep condition, hypertension history and depression as independent variables, to explore the potential association of sociodemographic and health-related factors with MCI risk assessed by MMSE and MoCA, respectively. Predictors were simultaneously included in the regression model based on the significance of differences in MCI prevalence by studied factors in present study, and the evident influence of gender on MCI in previous studies. No collinearity between predictors was detected in both final models (tolerance: 0.79–0.99 and VIF: 1.01–1.27 for MMSE; tolerance: 0.80–0.98 and VIF: 1.02–1.26 for MoCA). A value of p < 0.05 was considered significant. Statistical analysis was carried out using SAS 9.4 (SAS Inc., Cary, NC, USA).

Results

Characteristics of study population

A total of 4923 subjects aged 55 years and more were included in this study (Table 2), in which those aged 55–64, 65–74 and ≥ 75 years accounted for 41.5, 40.7 and 17.8%, respectively. More than half of participants were female (56.1%). Around 18.2% of subjects completed the education of high school and above. The majority of participants were unemployed (82.8%), which included retired subjects. And the proportions of subjects with moderate monthly household income per capital and meeting the recommended age-specific sleep duration were 61.2 and 68.0%, respectively. People who smoked currently and drank alcohol last year accounted for 15.6 and 17.1%, respectively. The rates of people with reported disease history of hypertension, diabetes, stroke and myocardial infarction were 31.8, 9.7, 2.0 and 1.9%, respectively. And 8.6% subjects had self-reported depression in this study.

Table 2. Cognitive assessment by MMSE and MoCA by sociodemographic and health-related factors in study population

Comparison of cognitive assessment between MMSE and MoCA, and cognitive function by sociodemographic and health-related factors

Average score of cognitive test using MMSE and MoCA in total population was 25.5 ± 4.9 and 22.6 ± 6.1, respectively (Table 2), and the MoCA RSD% (26.9%) was significantly greater than that in MMSE (19.0%) (p < 0.0001). Scatter plot of Fig. 1 depicted the relationship between MMSE and MoCA total scores, and Spearman correlation coefficient was 0.8374 (p < 0.0001).

Fig. 1

_{The correlation between the MMSE and MoCA total scores in Chinese population aged ≥55 years}

There were significant differences in the distribution of both MMSE and MoCA total scores by age group, gender, educational level, current employment status, household income, residence area, alcohol intake, sleep duration condition, history of hypertension, and depression (all p < 0.0001, Table 2). Multiple comparisons further indicated that total scores of either MMSE or MoCA in subjects aged 55–64 years, those with education of high school and above, and high monthly household income per capital, and those living in urban area, were likely to be the largest among their corresponding subgroups (all p < 0.05). Percentile analysis showed presence of ceiling effect (maximum total score on the 75th percentile) for MMSE in several subgroups of 55–64 years, education of high school and above, high level of monthly household income, urban and suburban areas of residence, but for MoCA only in subgroups with high monthly household income level and living in urban area.

According to cutoffs of MCI screening by MMSE and MoCA tests, prevalence of MCI in total population was 28.6 and 36.2%, respectively (Table 2). A total of 1158/4923 (23.5%) subjects fell into MCI for both MMSE and MoCA whereas 623/4923 (12.7%) who tested normal in the MMSE actually tested positive for MCI using MoCA. Of the total sample studied, 2891/4923 (58.7%) had normal scores for both tests (Table 3). The Kappa value indicating agreement for diagnosis of MCI using MoCA versus MMSE was 0.5973 (95% CI: 0.5737, 0.6209) with p < 0.0001.

Table 3 Agreement of MMSE and MoCA to detect MCI

Significant increased trends in MCI prevalence were observed along with the ascending age groups in both MMSE and MoCA settings (p < 0.0001), in which a higher proportion was observed in those ≥75 years (41.4% for MMSE and 48.2% for MoCA), while opposite trends were found in case of household income level (p < 0.0001) (Table 2). Prevalence of MCI detected by either MMSE or MoCA was considerably higher in subjects who were unemployed, currently smoked cigarettes, and had inappropriate sleep duration, hypertension history, or depression, compared to their referred groups (all p < 0.05). In addition, significant differences in MCI prevalence in setting of MMSE or MoCA were observed among areas of residence. However, only significant differences in MCI prevalence by educational level were found in MoCA test (p = 0.0004).

Subscores of cognitive domains assessed by MMSE and MoCA

Based on the distribution of each cognitive domain subscore in total samples assessed by different items of MMSE and MoCA (Table 4), the performance of execution, repetition and registration among 75% subjects using MMSE met maximum scores, whereas executive and recall dysfunctions were found in about 75% participants by MoCA test. The function of naming was performed well in both scales. Present study further focused on cognitive domains tested by both MMSE and MoCA, and found significant increased trends in proportions of subjects with cognitive dysfunction in terms of orientation, execution, calculation, naming, repetition, visuoconstruction and recall by MMSE across strata of the corresponding cognitive domain score by MoCA (all p < 0.0001, Fig. 2).

Table 4 Subscores of different cognitive domains by MMSE and MoCA in total subjects

Fig. 2

_{Proportion of subjects with cognitive domain dysfunction by MMSE across strata of MoCA-assessed cognitive subscore. Cognitive domains included a executive function, b orientation, c calculation, d naming, e repetition, f visuoconstructional function, and g recall}

Potential factors associated with MCI risk detected by MMSE and MoCA

Sociodemographic and health-related factors associated with MCI risk by multiple logistic regression model were highly consistent between MMSE and MoCA scales (Table 5). Especially, subjects aged ≥75 years (OR = 2.073, 95% CI: 1.727, 2.489 for MMSE; OR = 1.869, 95% CI: 1.570, 2.227 for MoCA) significantly increased the risk of MCI compared to the reference of 55–64 years. The odds of MCI in females was 28.0% in MMSE and 16.3% in MoCA greater than that in males. Being employed currently and living in a family with moderate or high monthly household income per capital highly reduced risk of MCI in both scales relative to their respective control group (all p < 0.05). Current smoking was identified as a risk factor with a 37.3 and 28.8% higher odds of MCI by MMSE and MoCA, respectively, compared to no smoking. A higher likelihood of MCI was observed in subjects living in county town or village, and those with hypertension history or self-reported depression (all p < 0.05).

Table 5 Potential factors associated with the risk of MCI using multiple logistic regression model^a

Discussion

MCI is a common condition in the elderly, characterized by deterioration of memory, attention, and cognitive function that is beyond what is expected based on age and educational level, but without significant interference with ability of daily activity. Present study found that MCI prevalence in Chinese population aged ≥55 years from urban and rural areas of four provinces using MMSE and MoCA was 28.6 and 36.2%, respectively, and MMSE had good correlation with MoCA (Spearman correlation coefficient = 0.8374) and moderate agreement for detecting MCI with Kappa value of 0.5973. Moreover, increasing age, female, living in county town/village, smoking, hypertension and depression significantly increased the risk of MCI in both tests. All findings indicated serious condition of cognitive impairment along with progressive increase in the growth rate of aging population in China and huge challenges on the prevention and treatment of MCI to the society and government.

The MMSE is the most widely used cognitive screening test by physicians and researchers for general cognitive evaluation. One problem with the MMSE is its ceiling effect or limited dynamic performance range for normal individuals, which increases the likelihood that persons in predementia stages score within the normal range. Consistent with previous study, the ceiling effect (28–30 points) for MCI was less using MoCA (26.2%) versus MMSE (46.3%) in this study as clearly depicted in Fig. 1 as well as the distributions of both test scores in Table 2. The greater RSD% in MoCA (26.9%) relative to that in MMSE (19.0%) further suggested MoCA distributed samples across a broader score range with less ceiling effect and had better detection of cognitive heterogeneity of the sample. On the other hand, MoCA was developed by Nasreddine in 2005 as a brief tool to screen subjects who present with cognitive complaints and usually have normal MMSE scores. Here, 12.7% of total subjects with a normal MMSE score actually tested positive for MCI according to MoCA’s adjusted cutoff points, partly reflecting higher sensitivity for MCI in MoCA although no comparison with the gold standard method was performed. This study further focused on cognitive domain subtests by MMSE and MoCA. The observed significantly increased likelihood of incorrect MMSE executive, naming, repetition, visuoconstructional, and recall tasks across decreasing scores of MoCA corresponding tasks (Fig. 2), suggested the higher sensitivity of the MoCA in detecting dysfunctions of abovementioned cognitive domains, which may be related to more components of each domain in MoCA. Together, as indicated above, the MoCA is a better measure to screen for cognitive impairment in middle-aged and older Chinese living in communities relative to MMSE as it lacks ceiling effect and shows better sensitivity.

This study found a high strength correlation between MoCA and MMSE scores with a Spearman correlation coefficient of 0.8374. This positive relationship was highly close to that reported by the original MoCA norms study in older Chinese, which obtained good correlation between both tools with Spearman correlation coefficient of 0.83. Both consistently demonstrated adequate level of concurrent validity between MoCA Beijing version and the Chinese version of MMSE for community dwellers. Significant positive correlation between total scores of MoCA and MMSE was also found in the assessment of cognitive deficit associated with chronic diseases. The MoCA and MMSE had a Kappa value of 0.5973, indicating moderate agreement. And the agreement disparity could attribute to the difference in the functions of the instruments themselves, in which MoCA was developed in particular for MCI screening whereas MMSE was originally invented as a tool to detect and monitor the development of dementia.

Changes in criteria and differences in populations studied and methodology have produced a wide range of prevalence estimates for MCI. Previous study applied uniform diagnostic criteria to harmonize data from USA, Europe, Asia and Australia, in which MCI prevalence ranged from 5 to 36.7%, and more reliably estimate MCI prevalence, as a result, a reduced MCI prevalence (2.1–20.7%) was produced when using MMSE score of 24–27 to define MCI. Present study found that prevalence of MCI in Chinese aged 55 years and older was 28.6 and 36.2% overall using education-specific cutoffs of MMSE and MoCA, respectively. Studies in mainland China over the past 5 years that used different diagnostic criteria showed MCI prevalence ranging from 12.6 to 34.1% in old population, and all these studies were conducted in single region, conversely our study covered urban and rural areas in four provinces. Representatively, Jia et al. (2014) reported that prevalence of MCI was 20.8% for individuals aged 65 years and above across multiple regions in China. His group (2020) recently conducted a large national study across different socioeconomic and geographic regions in 12 provinces and municipalities in China and found that the overall MCI prevalence was estimated at 15.5% in people aged 60 years or older, representing 38.77 million people nationwide. We also paid attention to the prevalence of MCI by demographic and health-related factors. Similar to the large-scale study, the prevalence of MCI increased with older age, and the higher prevalence of MCI was correlated with rural residence, smoking and hypertension in both MMSE and MoCA instruments in present study (Table 2). Educational level was believed to be the strongest noncognitive factor affecting cognitive test score. Consistently, less education profoundly correlated with poorer performance of the MMSE and MoCA, showing a significant increased trend of MMSE/MoCA score with high education in this study. The results supported the findings of better performance on MoCA for those with 6 years and more education compared to those with less than 6 years education. Oppositely, in the study by Ng et al., education influenced MoCA’s test performance in unpredictable manner, those with more education performed poorer relative to those with less education, which was likely to attribute to tests of MoCA domains of naming, attention, language, abstraction, and orientation. Unexpectedly, people with higher educational level (middle school/high school and above) had a greater prevalence of MCI detected by MoCA. Educational level is one of indicators of cognitive reserve, which influences the manifestation of symptoms of cognitive impairment. People with low education theoretically display a steeper cognitive decline early in the process of aging compared to those with high level of education. Our conflicting findings might indicate that confounders such as the passion for cognitive activity and strong district-level social network may buffer the relationship between low education and cognitive impairment. This raises the need for further study to test MCI by education. We also found other factors associated with MCI, such as household income, employment status, sleep duration and depression. Overall, the high level of variability in reported MCI prevalence worldwide or nationwide may be associated with ethnic and/or regional differences, and the heterogeneity in research methods, including the use of different diagnostic criteria, and the focus of samples with different characteristics, such as age brackets, gender and educational attainment. Anyhow, these findings suggested that MCI is becoming increasingly prevalent all over the world along with the changes in lifestyle and lifespan of human beings, and the clarification of risk factors for MCI would inform specific control measures as many risk factors are modifiable.

MCI is thought to be a transitional stage between being cognitively unimpaired and dementia, consensus has been reached to focus primary intervention on this population to halt dementia progression. With the increasing attention being paid to MCI, studies have been conducted in recent years in a variety of research settings to understand its influencing factors. We conducted systematic assessment of risk factors for MCI, to some extent, including demographic factors, lifestyle, psychological factors and cardiovascular risk factors. Compare to each reference group, increasing age (≥75 years), female gender, living in less urbanized areas (county town or village), current smoking, hypertension and depression considerably increased the odds of MCI detected by both MMSE and MoCA after adjustment for covariates, as reported in previous studies. Among these factors, there is no consensus on the question whether depression is the consequence or the cause for cognitive impairment in older people, but the association between depression and MCI may result in a faster progression of cognitive decline. In contrast, current employment and higher monthly household income per capital (1000–3999 and ≥ 4000 Chinese yuan) were significantly associated with lower risk of MCI, relative to unemployment and less than 1000 Chinese yuan of monthly income in present study, respectively, which was similar to previous findings also conducted in Chinese population. And the protective role of employment was attributed to increased reserve and the ability to tolerate higher levels of neuropathology thereby maintained their cognitive functioning, on the other hand, employment status would get access to higher social engagement, which was beneficial for MCI prevention.

There were several limitations in this study. First, the Chinese version of MMSE and MoCA scales and accordingly education-specific cutoffs of MCI were used in this study, which partly affected international comparison of prevalence rate and influencing factors of MCI. Second, due to limited data, we cannot analyze the impact of dietary intakes and genetic factors on cognitive impairment in this population. Additionally, a gold standard was not employed to detect MCI, as a result, this study failed to compare the sensitivity and specificity between MMSE and MoCA. Finally, false positive and false negative existed in MCI screening.

Conclusions

The findings of this study showed that MMSE and MoCA had good correlation and moderate agreement for detecting MCI in Chinese population aged 55 years and above. But MoCA had less ceiling effect for MCI and better detection of cognitive heterogeneity of the sample. High overall MCI prevalence was observed in both screenings, and residence of county town and village, current smoking, hypertension and depression were identified as modifiable risk factors for MCI except for increasing age, female gender. The cognitive function of the elderly will experience inevitable deterioration in China with the rapidly increasing aging population in near future, which poses a huge challenge for public health system and medical nursing system in China. Taken together, these findings indicate severe status of MCI in Chinese old population and provide important evidence for the establishment of specific intervention measures. Increasing public awareness of MCI and dementia, controlling MCI risk factors to delay dementia onset and boosting the implement of established strategies by authorities would effectively reduce the prevalence of MCI and dementia in China.

Background

Dementia is a leading cause of disability globally, including in China, posing significant challenges for policymakers, healthcare professionals, and families. Given the lack of effective treatments and the early onset of brain pathology years before symptoms appear, research has focused on delaying dementia in individuals during its preclinical stages. Mild cognitive impairment (MCI) represents a transitional phase between normal cognitive aging and dementia, affecting 10–15% of the population over 65. While some individuals with MCI may revert to normal cognition, an annual progression rate of 5–10% to dementia is observed, significantly higher than the general population. Approximately 50% of individuals with MCI are projected to progress to dementia within five years. MCI is considered a critical window for intervention to potentially delay dementia development, underscoring the importance of screening for MCI and identifying its influencing factors in at-risk older populations through large-scale studies.

The Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA) are widely used tools for detecting cognitive impairment in both clinical and research settings. The MoCA has generally demonstrated superior sensitivity to the MMSE in detecting MCI across various study environments, particularly by addressing the MMSE's "ceiling effect," where individuals with subtle impairments may still score in the normal range. While similar studies have been conducted in China, they often involved single regions and small sample sizes, leading to concerns about representativeness and reliability. Large-scale, multi-regional studies are therefore needed in China to confirm the agreement between MMSE and MoCA in identifying MCI within community-based samples.

Understanding potential and modifiable risk factors for cognitive complaints is crucial for the prevention, treatment, and intervention of MCI, which may delay progression to dementia. Previous research has identified various factors, including age, gender, education, occupation, marital status, income, psychological well-being, physical exercise, social engagement, diet, and chronic diseases. However, some findings have been inconsistent, possibly due to differences in study populations, methodologies, age ranges, and diagnostic criteria. Education, in particular, has a strong influence on MMSE and MoCA performance, with some studies reporting unexpected effects where individuals with more education performed poorer than those with less education. It is also important to determine if different cognitive screening tools reveal disparities in identifying potential factors for cognitive impairment within the same population.

This study aimed to determine the correlation and agreement between the MMSE and MoCA in detecting MCI. It also sought to compare how different factors influence MCI when assessed by each tool among middle-aged and older Chinese individuals participating in a community-based cohort study across urban and rural areas of four provinces. The findings from this study are expected to provide valuable insights for selecting appropriate cognitive assessment measures and for managing MCI.

Methods

Study population

Data for this study were obtained from the baseline phase of the Community-based Cohort Study on Nervous System Diseases, an ongoing longitudinal study initiated in 2018–2019 by the National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention. This cohort study investigates factors associated with nervous system diseases, specifically Alzheimer’s disease (AD) and Parkinson’s disease in individuals aged 55 years and older. Participants without pre-existing diagnoses of these conditions at enrollment were recruited using a multistage stratified random sampling approach across Hebei, Zhejiang, Shaanxi, and Hunan provinces. This involved randomly selecting two cities and two counties in each province, followed by random selection of urban/suburban neighborhoods and townships/villages. Within each community, all eligible household members were interviewed. The study protocol received approval from the Medical Ethics Committee of the National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention (No. 2017020), and written informed consent was obtained from each participant.

This study specifically focused on participants recruited for the AD cohort. Eligible individuals were aged 55 years or older, resided in the sampled community, had no clinical diagnosis of AD, and were free of comorbid conditions that could affect assessment, such as congenital or acquired mental retardation, previously diagnosed MCI, or significant uncorrected visual/hearing impairments. Participants with complete data on sociodemographic characteristics, disease history, cognitive examination, psychological evaluation, and basic activities of daily living were included. Individuals unable to perform basic daily activities (e.g., eating, dressing, bathing) were excluded (n = 71). For participants with extreme sleep durations (below the 1st percentile or above the 99th percentile for their age group), these values were replaced with the corresponding percentile values. A total of 4923 participants were included in the final analysis.

Cognitive assessment

All participants underwent cognitive assessment using the Chinese versions of the MMSE and MoCA. Both instruments have demonstrated validity and reliability in Chinese populations, incorporating cultural and linguistic considerations. MMSE and MoCA assessments were conducted strictly face-to-face by trained investigators, following standardized guidelines and protocols. The MMSE typically took 5–10 minutes to complete, while the MoCA required 10–15 minutes.

The MMSE is a 30-point questionnaire widely utilized in clinical and research settings to measure general cognitive impairment. It includes simple tasks assessing orientation to time and place, word repetition, arithmetic (e.g., serial sevens subtraction), language use and comprehension, and basic motor skills. The MoCA is another 30-point test covering eight cognitive domains. Specific details regarding MoCA items and its cultural and linguistic modifications for the Beijing version, derived from the original English version, have been previously described.

Cognitive function across different domains was evaluated based on the items within each test. The components and maximum scores for each domain were detailed in Table 1 (referring to the original document's table). The sum of included item points constituted the subscore for each cognitive domain. Dysfunction in a cognitive domain was defined by any incorrect response to included items, with specific cutoffs listed in Table 1 (referring to the original document's table).

The sum of all item points generated total scores for MMSE and MoCA, each ranging from 0 to 30. Higher scores indicate better cognitive function. For participants with 12 years or less of education, 1 point was added to their MoCA total score (if the score was less than 30). MCI was identified using education-specific cutoff points for the total scores of both the MMSE and MoCA. For the MMSE, cutoffs were ≤19 for illiterate individuals, ≤22 for those with elementary school education, and ≤26 for those with middle school education and above. Based on Chinese MoCA norms, cutoffs were ≤13 for illiterate individuals, ≤19 for those with 1–6 years of education, and ≤24 for those with 7 or more years of education.

Sociodemographic and health-related characteristics

Questionnaires were utilized by trained investigators to gather information on age, gender, educational level, current employment status, household income, residence area, current smoking status, alcohol intake during the last year, sleep duration (including daytime napping and full-night sleep), and histories of chronic diseases such as hypertension, diabetes, stroke, and myocardial infarction. Additionally, the 30-point Geriatric Depression Scale (GDS) was administered as a self-report assessment to identify depression, with a GDS score greater than 11 defining depression. All collected parameters were subsequently grouped for data analysis. Age was categorized into 55–64, 65–74, and ≥75 years. Gender was categorized as male or female. Educational level included: below elementary school, elementary school, middle school, high school and above. Monthly household income per capita was grouped into <1000, 1000–3999, and ≥4000 Chinese yuan. Statuses of current employment, current smoking, and alcohol intake in the last year were categorized as "yes" or "no." Residence area was classified as urban, suburban, county town, or village based on administrative divisions. Sleep condition was classified as "yes" or "no" depending on adherence to age-specific sleep duration recommendations. Disease histories of hypertension, diabetes, stroke, or myocardial infarction were categorized as "yes" or "no." Depression was also categorized as "yes" or "no."

Statistical analysis

Continuous variables were presented as mean ± standard deviation (SD), with median, 25th percentile (P25), and 75th percentile (P75) also calculated to assess the presence of ceiling/floor effects in the MMSE and MoCA tests. Categorical variables were expressed as counts (n) and percentages (%). Due to non-normal data distribution, non-parametric tests, specifically the Wilcoxon rank-sum test or Kruskal-Wallis analysis, were employed to evaluate differences in the distribution of MMSE or MoCA total scores across sociodemographic and health-related factors. When significant differences were found among three or more subgroups, multiple comparisons were performed using the Student-Newman-Keuls test. The relative standard deviation (RSD%), calculated as (SD/mean) × 100, was used to examine inter-individual variance in MMSE and MoCA total scores across the entire population, with a higher RSD% indicating better detection of cognitive heterogeneity. The RSD% indices for MMSE and MoCA were compared using the Wilcoxon signed-rank test. The prevalence of MCI across various factors was compared using the Chi-square test, and the Cochran-Armitage trend test was applied when appropriate. Trends in the proportions of subjects with MMSE-identified cognitive domain dysfunction across subscore strata of corresponding MoCA cognitive domains were also analyzed using the Cochran-Armitage trend test. Scatter plots and Spearman correlation coefficients were used to explore the relationship between MMSE and MoCA total scores. The Kappa value was used to determine the agreement between MMSE and MoCA in detecting MCI. Multiple logistic regression models were utilized to investigate the potential association of sociodemographic and health-related factors with MCI risk, with MCI (yes vs. no) as the dependent variable. Independent variables included age, gender, employment status, household income, residence area, smoking, sleep condition, hypertension history, and depression. Predictors were simultaneously included in the regression models based on the significance of differences in MCI prevalence identified in this study and the known influence of gender on MCI from previous research. No multicollinearity was detected between predictors in either final model (tolerance: 0.79–0.99 and VIF: 1.01–1.27 for MMSE; tolerance: 0.80–0.98 and VIF: 1.02–1.26 for MoCA). A p-value less than 0.05 was considered statistically significant. Statistical analyses were performed using SAS 9.4 (SAS Inc., Cary, NC, USA).

Results

Characteristics of study population

The study included a total of 4923 subjects aged 55 years and older. Participants aged 55–64, 65–74, and ≥75 years constituted 41.5%, 40.7%, and 17.8% of the sample, respectively. More than half of the participants were female (56.1%). Approximately 18.2% of subjects had completed high school education or higher. The majority of participants were unemployed (82.8%), a group that included retired individuals. The proportions of subjects with moderate monthly household income per capita and those meeting recommended age-specific sleep durations were 61.2% and 68.0%, respectively. Individuals who currently smoked accounted for 15.6%, and those who consumed alcohol in the last year accounted for 17.1%. The reported rates of individuals with a history of hypertension, diabetes, stroke, and myocardial infarction were 31.8%, 9.7%, 2.0%, and 1.9%, respectively. Additionally, 8.6% of subjects reported having depression.

Comparison of cognitive assessment between MMSE and MoCA, and cognitive function by sociodemographic and health-related factors

The average scores for cognitive assessment in the total population were 25.5 ± 4.9 for MMSE and 22.6 ± 6.1 for MoCA. The MoCA's Relative Standard Deviation (RSD%) of 26.9% was significantly greater than that of the MMSE (19.0%) (p < 0.0001), indicating MoCA's better ability to detect cognitive heterogeneity within the sample. A scatter plot depicted a strong positive relationship between MMSE and MoCA total scores, with a Spearman correlation coefficient of 0.8374 (p < 0.0001).

Significant differences were observed in the distribution of both MMSE and MoCA total scores across various factors, including age group, gender, educational level, current employment status, household income, residence area, alcohol intake, sleep duration, history of hypertension, and depression (all p < 0.0001). Multiple comparisons further revealed that total scores for both MMSE and MoCA were generally highest in subjects aged 55–64 years, those with high school education and above, individuals with high monthly household income per capita, and those residing in urban areas (all p < 0.05). Percentile analysis indicated a ceiling effect for MMSE (maximum total score at the 75th percentile) in several subgroups, including those aged 55–64 years, individuals with high school education and above, high monthly household income, and urban/suburban residence. In contrast, a ceiling effect for MoCA was observed only in subgroups with high monthly household income and urban residence, suggesting its broader range of performance.

Based on education-specific cutoffs for MCI screening by MMSE and MoCA, the prevalence of MCI in the total population was 28.6% and 36.2%, respectively. A total of 1158 out of 4923 subjects (23.5%) were identified with MCI by both MMSE and MoCA. Notably, 623 out of 4923 subjects (12.7%) who scored as normal on the MMSE were identified as having MCI by the MoCA. Overall, 2891 out of 4923 subjects (58.7%) had normal scores on both tests. The Kappa value, which indicates the agreement for MCI diagnosis between MoCA and MMSE, was 0.5973 (95% CI: 0.5737, 0.6209) with a p < 0.0001, suggesting moderate agreement.

Significant increasing trends in MCI prevalence were observed with ascending age groups for both MMSE and MoCA settings (p < 0.0001), with the highest proportion found in those aged ≥75 years (41.4% for MMSE and 48.2% for MoCA). Conversely, opposite trends were found regarding household income level, with lower MCI prevalence in higher income groups (p < 0.0001). The prevalence of MCI detected by either MMSE or MoCA was considerably higher in subjects who were unemployed, currently smoked, had inappropriate sleep duration, a history of hypertension, or depression, compared to their respective reference groups (all p < 0.05). Additionally, significant differences in MCI prevalence based on residence area were observed for both MMSE and MoCA. However, significant differences in MCI prevalence by educational level were found only in the MoCA test (p = 0.0004).

Subscores of cognitive domains assessed by MMSE and MoCA

An analysis of cognitive domain subscores revealed that approximately 75% of subjects achieved maximum scores in execution, repetition, and registration when assessed by MMSE. In contrast, approximately 75% of participants demonstrated executive and recall dysfunctions when assessed by the MoCA. The function of naming was generally well-performed across both scales. The study further focused on cognitive domains assessed by both MMSE and MoCA, observing significantly increased trends in the proportions of subjects with cognitive dysfunction in orientation, execution, calculation, naming, repetition, visuoconstruction, and recall by MMSE across decreasing strata of the corresponding cognitive domain scores by MoCA (all p < 0.0001).

Potential factors associated with MCI risk detected by MMSE and MoCA

Sociodemographic and health-related factors associated with MCI risk, as determined by multiple logistic regression models, showed high consistency between MMSE and MoCA scales. Specifically, subjects aged ≥75 years showed a significantly increased risk of MCI compared to the 55–64 years reference group (Odds Ratio [OR] = 2.073, 95% CI: 1.727, 2.489 for MMSE; OR = 1.869, 95% CI: 1.570, 2.227 for MoCA). The odds of MCI in females were 28.0% higher when assessed by MMSE and 16.3% higher when assessed by MoCA, compared to males. Being currently employed and belonging to a family with moderate or high monthly household income per capita significantly reduced the risk of MCI in both scales relative to their respective control groups (all p < 0.05). Current smoking was identified as a risk factor, associated with a 37.3% and 28.8% higher odds of MCI by MMSE and MoCA, respectively, compared to non-smoking. A higher likelihood of MCI was also observed in subjects residing in county towns or villages, and those with a history of hypertension or self-reported depression (all p < 0.05).

Discussion

MCI is a prevalent condition among older adults, characterized by cognitive decline beyond what is expected for age and education, without significant interference with daily activities. This study found an MCI prevalence of 28.6% using MMSE and 36.2% using MoCA among Chinese individuals aged 55 years and older across urban and rural areas of four provinces. The MMSE demonstrated good correlation with the MoCA (Spearman correlation coefficient = 0.8374) and moderate agreement for detecting MCI (Kappa value of 0.5973). Factors such as increasing age, female gender, residence in county towns or villages, smoking, hypertension, and depression were significantly associated with an increased risk of MCI in both tests. These findings highlight a serious condition of cognitive impairment, which is particularly relevant given China's progressively aging population and the associated challenges for public health and healthcare systems regarding MCI prevention and treatment.

One known limitation of the MMSE is its ceiling effect, meaning individuals in pre-dementia stages may score within the normal range. Consistent with prior research, this study found a lower ceiling effect for MCI using MoCA (26.2%) compared to MMSE (46.3%), as illustrated by score distributions. The MoCA's greater RSD% (26.9%) compared to MMSE (19.0%) further indicates its superior ability to distribute scores across a broader range, thereby better detecting cognitive heterogeneity within the sample. The MoCA was specifically developed to screen individuals with cognitive complaints who might still achieve normal MMSE scores. In this study, 12.7% of subjects with a normal MMSE score were identified as having MCI by the MoCA's adjusted cutoffs, partly suggesting higher sensitivity for MCI in the MoCA, although no comparison to a gold standard method was performed. Further analysis of cognitive domain subtests revealed that the MoCA showed higher sensitivity in detecting dysfunctions in executive function, naming, repetition, visuoconstruction, and recall compared to the MMSE, likely due to a greater number of items assessing these domains in the MoCA. Collectively, these findings suggest that the MoCA is a more suitable tool for screening cognitive impairment in middle-aged and older Chinese community residents due to its reduced ceiling effect and improved sensitivity.

This study observed a strong positive correlation between MoCA and MMSE scores (Spearman correlation coefficient = 0.8374), which aligns closely with findings from the original MoCA norms study in older Chinese populations (Spearman correlation coefficient = 0.83). This consistency indicates adequate concurrent validity between the MoCA Beijing version and the Chinese version of MMSE for community-dwelling individuals. While a significant positive correlation was found, the moderate agreement (Kappa value of 0.5973) between the MoCA and MMSE can be attributed to inherent differences in their design and primary functions: the MoCA was specifically developed for MCI screening, whereas the MMSE was originally designed to detect and monitor dementia progression.

MCI prevalence estimates vary widely globally due to differences in diagnostic criteria, study populations, and methodologies. Previous international studies reported MCI prevalence ranging from 5% to 36.7%. When harmonized criteria were applied, prevalence ranged from 2.1% to 20.7%. This study found overall MCI prevalence rates of 28.6% (MMSE) and 36.2% (MoCA) in Chinese individuals aged 55 years and older using education-specific cutoffs. Studies in mainland China over the past five years reported MCI prevalence ranging from 12.6% to 34.1% in older populations, often conducted in single regions. In contrast, this study covered urban and rural areas across four provinces, enhancing representativeness. Consistent with other large-scale Chinese studies, MCI prevalence increased with age and was higher in rural residents, smokers, and individuals with hypertension in both MMSE and MoCA settings. Educational level significantly influenced cognitive test scores, with lower education profoundly correlating with poorer performance on both tests. However, an unexpected finding was a greater prevalence of MCI detected by MoCA in individuals with higher educational levels (middle school/high school and above), a finding that may indicate the influence of cognitive reserve or other confounders and warrants further investigation. Other factors associated with MCI in this study included household income, employment status, sleep duration, and depression. The high variability in reported MCI prevalence worldwide and nationwide is likely associated with ethnic/regional differences and heterogeneity in research methods, including diverse diagnostic criteria and characteristics of study samples. Nonetheless, these findings suggest that MCI is becoming increasingly prevalent globally, coinciding with changes in lifestyle and lifespan, and identifying its risk factors is crucial as many are modifiable.

MCI is widely considered a transitional stage between normal cognition and dementia, leading to a consensus on the importance of primary intervention in this population to halt dementia progression. Recent studies have focused on understanding MCI's influencing factors across various research settings. This study systematically assessed a range of risk factors, including demographic characteristics, lifestyle, psychological factors, and cardiovascular risk factors. Compared to their respective reference groups, increasing age (≥75 years), female gender, residence in less urbanized areas (county town or village), current smoking, hypertension, and depression significantly increased the odds of MCI detected by both MMSE and MoCA after adjusting for covariates, consistent with previous research. The precise relationship between depression and cognitive impairment (cause or consequence) in older adults remains a subject of ongoing debate, but an association between depression and MCI may lead to faster cognitive decline. Conversely, current employment and higher monthly household income per capita (1000–3999 and ≥4000 Chinese yuan) were significantly associated with a lower risk of MCI, similar to findings from other Chinese populations. The protective effect of employment is thought to stem from increased cognitive reserve and the ability to tolerate higher levels of neuropathology while maintaining cognitive function, as well as providing access to greater social engagement, which benefits MCI prevention.

Several limitations were present in this study. Firstly, the use of Chinese versions of MMSE and MoCA scales, along with education-specific cutoffs for MCI, partly limits direct international comparisons of prevalence rates and influencing factors. Secondly, due to limited data, the impact of dietary intakes and genetic factors on cognitive impairment could not be analyzed in this population. Additionally, a gold standard method was not employed to detect MCI, preventing a direct comparison of sensitivity and specificity between MMSE and MoCA. Finally, the possibility of false positive and false negative results in MCI screening exists.

Conclusions

This study’s findings indicate that the MMSE and MoCA demonstrate good correlation and moderate agreement in detecting MCI among the Chinese population aged 55 years and older. However, the MoCA exhibited a lesser ceiling effect for MCI and better detection of cognitive heterogeneity within the sample. A high overall MCI prevalence was observed using both screening tools. Residence in county towns and villages, current smoking, hypertension, and depression were identified as modifiable risk factors for MCI, in addition to non-modifiable factors like increasing age and female gender. With China's rapidly increasing aging population, the inevitable deterioration of cognitive function among the elderly poses a substantial challenge to the public health and medical nursing systems. These findings highlight the severe status of MCI in the older Chinese population and provide important evidence for developing specific intervention measures. Increasing public awareness of MCI and dementia, controlling identified MCI risk factors to delay dementia onset, and boosting the implementation of established strategies by authorities are crucial steps to effectively reduce the prevalence of MCI and dementia in China.

Background

Dementia is a leading cause of disability worldwide for individuals over 65 years, including in China. This condition presents significant challenges for policymakers, healthcare professionals, and families. Given the lack of effective treatments for dementia and the fact that brain changes begin years before noticeable cognitive symptoms appear, much research now focuses on delaying dementia in individuals in the preclinical stages. Mild Cognitive Impairment (MCI) is considered a transitional phase between normal cognitive aging and dementia. It involves a decline in cognitive function, observable both subjectively and through objective tests, and affects 10–15% of the population over 65. While some individuals with MCI may improve over time, 5–10% progress to dementia annually, a rate significantly higher than in the general population. Approximately 50% of individuals with MCI are estimated to progress to dementia within five years. MCI is seen as a crucial "window" for intervention to potentially delay the onset of dementia. Therefore, screening for MCI and identifying its influencing factors in older populations are important steps toward improving cognitive function and delaying progression to dementia.

To assess cognitive function, the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA) are widely used in both clinical and research settings. The MoCA has generally been found to be more effective than the MMSE for detecting MCI, as the MMSE often has lower sensitivity. The MoCA can also identify differences in cognitive profiles even in individuals who score normally on the MMSE, making it a valuable tool for assessing cognition in MCI, especially where the MMSE's "ceiling effect" (where many high-functioning individuals score perfectly, limiting its ability to detect subtle declines) is an issue. Similar studies have been conducted in China, but many were limited by small sample sizes and single regions, which affected their representativeness and reliability. Studies specifically comparing the MMSE and MoCA for MCI detection in community-based samples remain uncommon. Further multi-regional studies are needed to confirm the agreement between MMSE and MoCA in identifying MCI, which could provide new insights due to larger sample sizes.

Understanding potential and modifiable risk factors for cognitive complaints is crucial for prevention, treatment, and intervention in the vulnerable state of MCI, potentially delaying dementia progression. Research has identified several factors, including age, gender, education, occupation, marital status, income, psychological well-being, physical exercise, social engagement, diet, and chronic diseases. However, some findings have been inconsistent, possibly due to differences in study countries, research methods, age ranges included, and the specific cognitive assessment tools or diagnostic criteria used. Education, in particular, has a strong influence on MMSE and MoCA performance, with some studies even observing unexpected effects where individuals with more education performed worse than those with less. It is also important to determine if different cognitive screening tools reveal disparities in potential factors affecting cognition when applied to the same population.

This study aims to determine the correlation and agreement between the MMSE and MoCA in detecting MCI. It also seeks to evaluate differences in factors influencing MCI when assessed by each tool among middle-aged and older Chinese individuals participating in a baseline survey of a community-based cohort study across urban and rural areas of four provinces. The findings are expected to have important implications for selecting appropriate cognitive measures and managing MCI.

Methods

The data for this study came from the baseline survey of the Community-based Cohort Study on Nervous System Diseases, an ongoing study initiated in 2018–2019 by the National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention. This cohort study investigates factors related to the risk of neurological diseases, including Alzheimer’s disease (AD) and Parkinson’s disease in individuals aged 55 years and older. Participants without pre-existing diagnoses of these diseases were recruited using a multi-stage stratified random sampling method across four Chinese provinces: Hebei, Zhejiang, Shaanxi, and Hunan. The study protocol was approved by the Medical Ethics Committee of the National Institute for Nutrition and Health, and written informed consent was obtained from each participant.

The present study specifically focused on participants recruited for the AD cohort. Eligible individuals were 55 years or older, resided in the sampled community, had no clinical diagnosis of AD or MCI, and were free of conditions that could interfere with cognitive assessment, such as intellectual disability or severe visual/hearing impairments. From the full cohort, 4923 participants with complete data on sociodemographic characteristics, medical history, cognitive assessments, psychological evaluations, and basic activities of daily living were included in the analysis. Individuals unable to perform basic daily activities were excluded.

All participants underwent cognitive assessment using Chinese versions of the MMSE and MoCA, both of which have been validated for cultural and linguistic differences in China. Trained investigators administered the tests strictly face-to-face following established guidelines, with the MMSE typically taking 5–10 minutes and the MoCA 10–15 minutes. The MMSE is a 30-point questionnaire that assesses various cognitive areas, including orientation, memory, attention, language, and basic motor skills. The MoCA is also a 30-point test covering eight cognitive domains, including executive function, visuospatial skills, naming, memory, attention, language, abstraction, and orientation. Specific items and scoring for these tests have been detailed in previous publications. Cognitive function in different domains was evaluated based on the items of each test, with a higher total score indicating better cognitive function. For the MoCA, one point was added to the total score for participants with 12 years or less of education, provided their score was less than 30. Mild Cognitive Impairment was identified using education-specific cutoff points for the total scores of both MMSE and MoCA.

Information on sociodemographic and health-related characteristics was collected through questionnaires administered by trained investigators. These included age, gender, education level, employment status, household income, residence area, current smoking status, alcohol intake, sleep duration, and histories of chronic diseases such as hypertension, diabetes, stroke, and myocardial infarction. Depression was also assessed using the Geriatric Depression Scale (GDS) 30-point version, with a score above 11 indicating depression. All collected parameters were grouped into categories for data analysis. Statistical analysis involved non-parametric tests to examine differences in MMSE or MoCA scores across various sociodemographic and health factors. The relative standard deviation (RSD%) was calculated to compare the ability of each test to detect cognitive heterogeneity, with higher RSD% indicating better detection. Chi-square tests and Cochran-Armitage trend tests were used to compare MCI prevalence and trends across factors. Spearman correlation coefficient was applied to assess the relationship between MMSE and MoCA total scores, and Kappa value determined the agreement in MCI detection. Multiple logistic regression models were used to explore associations between sociodemographic and health-related factors and the risk of MCI as assessed by each tool, adjusting for relevant variables. A p-value less than 0.05 was considered statistically significant.

Results

The study included 4923 subjects aged 55 years and older, with 41.5% aged 55–64, 40.7% aged 65–74, and 17.8% aged 75 or older. Over half of the participants (56.1%) were female. Approximately 18.2% had completed high school education or higher, and most (82.8%) were unemployed, including retired individuals. A significant portion (61.2%) had moderate monthly household income, and 68.0% met recommended age-specific sleep durations. Current smokers constituted 15.6% of the population, and 17.1% reported alcohol intake in the past year. Prevalences of reported hypertension, diabetes, stroke, and myocardial infarction were 31.8%, 9.7%, 2.0%, and 1.9%, respectively. Self-reported depression was present in 8.6% of the subjects.

The average MMSE score for the total population was 25.5 ± 4.9, while the average MoCA score was 22.6 ± 6.1. The MoCA demonstrated a significantly higher relative standard deviation (RSD%) of 26.9% compared to the MMSE's 19.0%, indicating that MoCA distributed scores across a broader range and was better at detecting varied cognitive abilities within the sample. A strong positive correlation was observed between MMSE and MoCA total scores, with a Spearman correlation coefficient of 0.8374 (p < 0.0001).

Both MMSE and MoCA total scores showed significant differences across various sociodemographic and health factors. For instance, scores were generally higher in individuals aged 55–64 years, those with high school education or above, individuals with higher monthly household income, and those living in urban areas. Analysis revealed a ceiling effect for MMSE in several subgroups (e.g., younger age, higher education, urban residence), meaning many individuals achieved the maximum score. The MoCA, however, showed a ceiling effect in fewer subgroups, primarily those with high monthly household income and urban residence, suggesting its better ability to differentiate cognitive function among higher-performing individuals.

Based on education-specific cutoff points, the overall prevalence of MCI in the study population was 28.6% when detected by MMSE and 36.2% when detected by MoCA. A total of 23.5% of subjects were identified with MCI by both tests, while 12.7% who scored normally on the MMSE were identified as having MCI by the MoCA, highlighting MoCA’s higher sensitivity. Conversely, 58.7% of the total sample had normal scores on both tests. The agreement between MMSE and MoCA for MCI diagnosis was moderate, indicated by a Kappa value of 0.5973. Significant trends in MCI prevalence were observed with increasing age, with a higher proportion of MCI in individuals aged 75 years and older. Conversely, MCI prevalence decreased with higher household income. Individuals who were unemployed, currently smoked, had inappropriate sleep duration, a history of hypertension, or depression showed a considerably higher prevalence of MCI compared to their respective control groups. Differences in MCI prevalence were also observed across residence areas. Notably, only the MoCA test showed significant differences in MCI prevalence based on educational level.

Analysis of cognitive domain subscores revealed that with MMSE, 75% of subjects achieved maximum scores in execution, repetition, and registration. In contrast, approximately 75% of participants showed executive and recall dysfunctions when assessed by MoCA. Naming function was generally well-performed across both scales. The study further demonstrated significant trends where increasing cognitive dysfunction, as measured by MMSE in domains like orientation, execution, calculation, naming, repetition, visuoconstruction, and recall, correlated with decreasing scores in corresponding MoCA cognitive domains.

Multiple logistic regression models identified several sociodemographic and health-related factors consistently associated with an increased risk of MCI for both MMSE and MoCA. Individuals aged 75 years and older had significantly higher odds of MCI compared to those aged 55–64. Female gender was also associated with a higher likelihood of MCI compared to males. Living in a county town or village, current smoking, having a history of hypertension, and self-reported depression were all associated with increased MCI risk. Conversely, being currently employed and having moderate or high monthly household income were associated with a reduced risk of MCI.

Discussion

Mild Cognitive Impairment (MCI) is a common condition in older adults, characterized by a cognitive decline beyond what is typical for aging, but without significantly impacting daily activities. This study found that the prevalence of MCI in Chinese individuals aged 55 years and older, living in urban and rural areas across four provinces, was 28.6% using the MMSE and 36.2% using the MoCA. The MMSE showed a good correlation with the MoCA (Spearman correlation coefficient = 0.8374), and there was moderate agreement between the two tests for detecting MCI (Kappa value of 0.5973). Factors such as increasing age, female gender, living in less urbanized areas, smoking, hypertension, and depression were consistently associated with a significantly increased risk of MCI across both assessments. These findings highlight the serious prevalence of cognitive impairment in China, especially with its rapidly aging population, and underscore the significant challenges for public health systems in preventing and treating MCI.

One known limitation of the MMSE is its "ceiling effect," meaning it has a limited dynamic range for cognitively healthy individuals, making it less effective at detecting subtle cognitive declines in preclinical stages of dementia. Consistent with prior research, this study observed a lower ceiling effect for MCI with the MoCA (26.2%) compared to the MMSE (46.3%). The higher relative standard deviation (RSD%) for MoCA (26.9%) versus MMSE (19.0%) further indicates MoCA's superior ability to differentiate cognitive function across a broader range of scores and detect cognitive heterogeneity within the sample. The MoCA was specifically developed to screen individuals with cognitive complaints who might still score normally on the MMSE. Indeed, this study found that 12.7% of participants who scored normally on the MMSE were actually identified with MCI by the MoCA, suggesting MoCA's higher sensitivity for MCI detection. Furthermore, the MoCA's higher sensitivity in detecting dysfunctions in specific cognitive domains, such as executive function, naming, repetition, visuoconstructional skills, and recall, is likely due to its more comprehensive assessment within these areas. Overall, the MoCA appears to be a more suitable screening tool for cognitive impairment in middle-aged and older Chinese community dwellers due to its reduced ceiling effect and greater sensitivity compared to the MMSE.

This study observed a strong positive correlation between MoCA and MMSE scores (Spearman correlation coefficient of 0.8374), consistent with previous findings in older Chinese populations, indicating adequate concurrent validity between the two instruments. Despite this strong correlation, the moderate agreement (Kappa value of 0.5973) for MCI diagnosis between the two tests suggests differences in their specific functions. The MoCA was designed specifically for MCI screening, while the MMSE was originally developed for detecting and monitoring dementia progression. This difference in design likely contributes to the observed disparity in agreement.

MCI prevalence estimates vary widely across studies due to differences in diagnostic criteria, populations studied, and methodologies. This study found an overall MCI prevalence of 28.6% (MMSE) and 36.2% (MoCA) in Chinese individuals aged 55 years and older, using education-specific cutoffs. These figures fall within or slightly above the range reported by other studies in mainland China. Similar to large-scale studies, MCI prevalence increased with older age, and was higher in rural residents, smokers, and individuals with hypertension. Educational level strongly influenced cognitive test scores, with higher education generally correlating with better performance on both MMSE and MoCA in this study. However, unexpectedly, individuals with higher education levels showed a greater MCI prevalence detected by MoCA, which might suggest the influence of cognitive reserve or other unmeasured factors like cognitive activity or social networks. This highlights a need for further research on the relationship between education and MCI. Other factors such as household income, employment status, sleep duration, and depression were also associated with MCI. The variability in reported MCI prevalence globally and nationally can be attributed to ethnic/regional differences and methodological heterogeneity. Nevertheless, these findings emphasize the increasing prevalence of MCI worldwide and the importance of identifying modifiable risk factors.

With growing attention on MCI as a transitional stage to dementia, focusing interventions on this population to halt progression is critical. This study systematically assessed various risk factors for MCI, including demographic, lifestyle, psychological, and cardiovascular factors. After adjusting for covariates, increasing age (≥75 years), female gender, living in less urbanized areas (county town or village), current smoking, hypertension, and depression significantly increased the odds of MCI, consistent with previous research. The relationship between depression and cognitive impairment in older adults remains debated (cause or consequence), but their association suggests a faster progression of cognitive decline. Conversely, current employment and higher monthly household income were significantly associated with a reduced risk of MCI, aligning with prior findings in Chinese populations. The protective effect of employment is thought to stem from increased cognitive reserve and higher social engagement.

Conclusions

This study found that the MMSE and MoCA demonstrated a good correlation and moderate agreement in detecting Mild Cognitive Impairment among Chinese individuals aged 55 years and older. However, the MoCA proved to be superior due to its lesser ceiling effect for MCI and better ability to detect subtle cognitive differences within the sample. A high overall prevalence of MCI was observed through both screening methods. Key modifiable risk factors identified for MCI included residence in county towns and villages, current smoking, hypertension, and depression, in addition to non-modifiable factors like increasing age and female gender. Given the rapidly growing aging population in China, the inevitable deterioration of cognitive function presents a substantial challenge for the public health and medical care systems. These findings underscore the severe status of MCI in the older Chinese population and provide crucial evidence for developing targeted intervention strategies. Increasing public awareness about MCI and dementia, effectively controlling modifiable risk factors to delay dementia onset, and robustly implementing established public health strategies by authorities could significantly reduce the prevalence of MCI and dementia in China.

Background

Dementia is a leading cause of disability for people over 65 globally, including in China. This condition creates major challenges for policymakers, healthcare providers, and families. Because there is no effective treatment for dementia, and brain changes often begin years before symptoms appear, researchers are focusing on delaying the disease in its early stages. Mild Cognitive Impairment (MCI) is considered this early, transitional stage between healthy aging and dementia. It affects 10–15% of people over 65. While some individuals with MCI improve, about 5–10% progress to dementia each year, which is much higher than the general population. Over five years, approximately 50% of people with MCI may develop dementia. This period of MCI offers a chance for intervention to potentially delay the onset of dementia. Therefore, it is important to screen for MCI and identify factors that might contribute to it in older adults.

When it comes to screening for cognitive problems, the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA) are widely used in both clinical and research settings. The MoCA has generally been found to be better than the MMSE at detecting MCI, as the MMSE often lacks sensitivity. The MoCA can also identify subtle cognitive differences even in individuals who score normally on the MMSE, making it a useful tool for assessing cognition in those with MCI, especially when the MMSE's limitations are an issue. While similar studies have been done in China, they often involved small groups from single regions, which made their findings less representative and reliable. More studies involving multiple regions are needed to confirm how well MMSE and MoCA agree in identifying MCI, as larger studies might reveal new findings.

Additionally, understanding potential risk factors for cognitive problems is crucial for preventing, treating, and intervening during the vulnerable MCI stage, which could delay progression to dementia. Research has identified several factors, such as age, gender, education, job type, marriage, income, mental well-being, physical activity, social involvement, diet, and history of chronic diseases. However, some findings have been inconsistent, possibly due to differences in study locations, research methods, age ranges included, and the cognitive assessment tools or diagnostic criteria used. Education, in particular, has a strong impact on MMSE and MoCA performance, with sometimes unexpected results where highly educated individuals might perform worse than those with less education. It is also important to determine if different risk factors for cognitive decline are found when using different screening tools on the same group of people.

This study aimed to determine how well the MMSE and MoCA correlate and agree in detecting MCI. It also explored differences in factors influencing MCI among middle-aged and older Chinese adults participating in a large community-based study across four provinces. The findings from this research could significantly impact the choice of cognitive assessment tools and how MCI is managed.

Methods

Study population

The data for this study came from the initial phase of the Community-based Cohort Study on Nervous System Diseases. This ongoing study, started in 2018–2019, investigates factors related to the risk of three neurological diseases: epilepsy (for those over 1 year old), and Alzheimer’s disease (AD) and Parkinson’s disease (for those 55 years and older). Participants without a prior diagnosis of these diseases were recruited using a specific sampling method in Hebei, Zhejiang, Shaanxi, and Hunan provinces. In each province, two cities and two counties were randomly chosen. Within these areas, neighborhoods, townships, and villages were also selected randomly. In each chosen community, all eligible household members were interviewed. The study protocol was approved by the Medical Ethics Committee of the National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention, and all participants provided written consent.

This study focused specifically on individuals recruited for the AD cohort. Eligible participants were 55 years or older, lived in the sampled community, had no clinical diagnosis of AD, and had no other conditions that could affect cognitive assessment, such as mental retardation or severe vision/hearing problems. Participants with complete data on demographics, health history, cognitive exams, psychological evaluations, and daily living abilities were included. Individuals unable to perform basic daily activities like eating, dressing, or bathing were excluded (71 people). For those with unusually short or long sleep durations, their data was adjusted. In total, 4,923 participants were included in the final analysis.

Cognitive assessment

All participants in this study underwent cognitive assessment using the Chinese versions of the MMSE and MoCA. Both tools have been proven valid and reliable for Chinese populations, taking into account cultural and linguistic differences. The MMSE and MoCA tests were conducted face-to-face by trained investigators following strict guidelines and protocols, taking approximately 5–10 minutes for the MMSE and 10–15 minutes for the MoCA.

The MMSE is a 30-point questionnaire widely used to measure cognitive impairment. It includes simple tasks in areas such as time and place orientation, remembering word lists, performing arithmetic like serial subtractions, using and understanding language, and basic motor skills. The MoCA is another 30-point test that covers eight different areas of cognitive function. Specific details about the MoCA items and their cultural adjustments for the Beijing version have been previously described.

Cognitive function in different areas was evaluated based on items from each test. A total score for each cognitive domain was calculated from the sum of included item points. Dysfunction in a cognitive domain was defined as any incorrect response to included items, with specific cutoffs used for each domain. The total scores for both MMSE and MoCA ranged from 0 to 30, with higher scores indicating better cognitive function. For participants with 12 years of education or less, one point was added to their MoCA total score (if it was less than 30). MCI was identified using specific cutoff points for the total scores of MMSE and MoCA, adjusted for education level. For the MMSE, scores of 19 or less for illiterate individuals, 22 or less for those with elementary school education, and 26 or less for those with middle school education and above indicated MCI. According to Chinese MoCA standards, scores of 13 or less for illiterate individuals, 19 or less for those with 1–6 years of education, and 24 or less for those with 7 or more years of education indicated MCI.

Sociodemographic and health-related characteristics

Information on participants' age, gender, education, employment, household income, living area, smoking habits, alcohol intake, sleep duration, and medical history (hypertension, diabetes, stroke, myocardial infarction) was collected using questionnaires by trained investigators. Additionally, the 30-point Geriatric Depression Scale (GDS) was used to assess depression, with a score above 11 indicating depression. All this information was grouped for data analysis. For example, age was grouped into 55–64, 65–74, and 75+ years; education into categories like "below elementary school" and "high school and above."

Statistical analysis

Data analysis involved comparing continuous variables (like scores) and categorical variables (like gender). Non-parametric tests were used to examine differences in MMSE or MoCA scores across various demographic and health-related factors. The percentage for the relative standard deviation (RSD%) was calculated to see how much scores varied, with a higher RSD% suggesting better detection of cognitive differences. MMSE and MoCA RSD% values were compared. The number of people with MCI across different factors was compared using statistical tests. Trends in cognitive domain dysfunction were also analyzed. Scatter plots and correlation coefficients were used to explore the relationship between MMSE and MoCA total scores. The agreement between MMSE and MoCA in detecting MCI was measured using a Kappa value. Finally, multiple logistic regression was used to identify factors associated with MCI risk, considering age, gender, employment, income, living area, smoking, sleep, hypertension, and depression. A p-value less than 0.05 was considered statistically significant.

Results

Characteristics of study population

This study included 4,923 individuals aged 55 years and older. The age groups were distributed as follows: 41.5% were 55–64 years old, 40.7% were 65–74 years old, and 17.8% were 75 years or older. Over half of the participants (56.1%) were female. Approximately 18.2% had a high school education or higher. The majority of participants (82.8%) were not employed, including retired individuals. About 61.2% had a moderate monthly household income, and 68.0% met recommended age-specific sleep durations. Current smokers made up 15.6% of the group, and 17.1% had consumed alcohol in the past year. The rates of reported medical histories included hypertension (31.8%), diabetes (9.7%), stroke (2.0%), and myocardial infarction (1.9%). Additionally, 8.6% of participants reported having depression.

Comparison of cognitive assessment between MMSE and MoCA, and cognitive function by sociodemographic and health-related factors

The average MMSE score for the entire group was 25.5 (±4.9), and the average MoCA score was 22.6 (±6.1). The MoCA showed significantly greater score variation (RSD% of 26.9%) compared to the MMSE (RSD% of 19.0%), indicating it was better at showing a range of cognitive abilities. A strong positive relationship was found between MMSE and MoCA total scores, with a Spearman correlation coefficient of 0.8374.

Both MMSE and MoCA scores differed significantly based on age, gender, education level, employment status, household income, living area, alcohol intake, sleep duration, history of hypertension, and depression. Specifically, individuals aged 55–64, those with high school education or above, and those with higher household incomes or living in urban areas generally had the highest scores. The MMSE showed a "ceiling effect" (many people scoring at the maximum) in several subgroups, such as younger ages, higher education, and urban residents, meaning it might not fully capture higher cognitive abilities. The MoCA, however, showed less of this ceiling effect, mainly in subgroups with high income and urban residents.

Based on the MCI cutoffs, the prevalence of MCI in the total population was 28.6% according to MMSE and 36.2% according to MoCA. Of the total participants, 23.5% were identified with MCI by both tests. Interestingly, 12.7% of individuals who scored normally on the MMSE were actually identified as having MCI by the MoCA. Overall, 58.7% of participants had normal scores on both tests. The agreement between MMSE and MoCA in diagnosing MCI was moderate, with a Kappa value of 0.5973.

MCI prevalence increased significantly with age for both tests, with the highest rates found in those aged 75 and older (41.4% for MMSE and 48.2% for MoCA). Conversely, MCI prevalence decreased as household income increased. MCI rates detected by either test were also higher in unemployed individuals, current smokers, those with inadequate sleep, hypertension, or depression. Significant differences in MCI prevalence were also observed based on the area of residence. However, differences in MCI prevalence by education level were only significant with the MoCA test.

Subscores of cognitive domains assessed by MMSE and MoCA

When looking at specific cognitive domains, the MMSE results showed that 75% of participants performed at maximum scores for execution, repetition, and registration tasks. In contrast, the MoCA revealed executive and recall difficulties in about 75% of participants. Naming ability was generally well-preserved in both assessments. The study also focused on cognitive areas assessed by both tests and found that a lower score on a MoCA cognitive domain was significantly linked to a higher chance of having a dysfunction in the corresponding MMSE domain. This trend was observed for orientation, execution, calculation, naming, repetition, visuoconstruction, and recall.

Potential factors associated with MCI risk detected by MMSE and MoCA

The demographic and health-related factors linked to MCI risk were very similar for both MMSE and MoCA. Specifically, being 75 years or older significantly increased the risk of MCI compared to those aged 55–64. Females had a higher likelihood of MCI than males (28.0% higher odds for MMSE, 16.3% for MoCA). Being currently employed and having a moderate to high monthly household income significantly reduced the risk of MCI compared to unemployment and lower income. Current smoking was identified as a risk factor, increasing the odds of MCI by 37.3% (MMSE) and 28.8% (MoCA) compared to non-smokers. A higher chance of MCI was also observed in individuals living in county towns or villages, and those with a history of hypertension or self-reported depression.

Discussion

MCI is a common condition in older adults, characterized by a decline in memory, attention, and other cognitive functions beyond what is expected for age and education, but without severe impact on daily activities. This study found that MCI prevalence in Chinese individuals aged 55 and above in urban and rural areas of four provinces was 28.6% using MMSE and 36.2% using MoCA. The MMSE and MoCA showed a good correlation and moderate agreement in detecting MCI. Factors like increasing age, being female, living in less urban areas, smoking, hypertension, and depression were all significantly linked to an increased risk of MCI in both tests. These findings highlight the serious state of cognitive impairment in China as its population ages, posing significant challenges for preventing and treating MCI.

One known limitation of the MMSE is its "ceiling effect," meaning many individuals in early stages of cognitive decline might still score within the normal range. Consistent with previous research, this study showed that the MoCA had a much lower ceiling effect for MCI (26.2%) compared to the MMSE (46.3%). The greater score variation observed with the MoCA further supports its ability to better detect a range of cognitive differences. Furthermore, the MoCA was specifically developed to screen individuals with cognitive complaints who might still score normally on the MMSE. This study found that 12.7% of participants who had normal MMSE scores were actually identified with MCI by the MoCA, suggesting MoCA's higher sensitivity for MCI. The study also noted that MoCA was more effective in detecting problems in specific cognitive areas like executive function, naming, repetition, visuoconstructional skills, and recall. Overall, these findings suggest that the MoCA is a better tool for screening cognitive impairment in middle-aged and older Chinese communities because it avoids the ceiling effect and offers better sensitivity.

This study found a strong correlation between MoCA and MMSE scores, similar to findings from other studies in older Chinese populations, indicating that both tools are generally consistent. While both tests measure cognitive function, their agreement in diagnosing MCI was moderate. This difference in agreement could be due to their differing purposes: MoCA was specifically designed for MCI screening, while MMSE was created to detect and monitor dementia.

Estimates of MCI prevalence vary widely across studies due to differences in diagnostic criteria and study populations. This study found an overall MCI prevalence of 28.6% (MMSE) and 36.2% (MoCA) in Chinese adults aged 55 and older, using education-specific cutoffs. Compared to previous studies in mainland China, which typically focused on single regions, this study covered urban and rural areas in four provinces, offering a more representative view. Similar to other large-scale studies, MCI prevalence in this research increased with age, and higher rates were associated with living in rural areas, smoking, and hypertension for both MMSE and MoCA. Education level also strongly affected cognitive test scores, with higher education generally linked to better performance on both tests. While education is considered a protective factor, this study unexpectedly found that highly educated individuals might have a higher prevalence of MCI detected by MoCA, which could warrant further research. Other factors associated with MCI included household income, employment status, sleep duration, and depression. These variations in MCI prevalence across studies highlight the influence of ethnic/regional differences and diverse research methods.

MCI is seen as a critical transitional stage towards dementia, making early intervention a key focus. This study systematically assessed risk factors for MCI, including demographic, lifestyle, psychological, and cardiovascular factors. Consistent with previous research, an increased risk of MCI was found in older adults (75+), females, those living in less urbanized areas, current smokers, and individuals with hypertension or depression. While the exact relationship between depression and cognitive impairment in older adults is still debated, their association may lead to faster cognitive decline. Conversely, current employment and higher monthly household income were significantly linked to a lower risk of MCI, similar to previous findings in Chinese populations. The protective effect of employment may be due to increased cognitive reserve and higher social engagement, both beneficial for MCI prevention.

This study had some limitations. First, using Chinese versions of MMSE and MoCA with specific education-adjusted cutoffs may limit direct international comparisons of MCI prevalence and related factors. Second, due to data limitations, the impact of dietary habits and genetic factors on cognitive impairment could not be analyzed. Additionally, the study did not use a "gold standard" to diagnose MCI, preventing a direct comparison of the sensitivity and specificity of MMSE and MoCA. Finally, any MCI screening may involve false positive and false negative results.

Conclusions

The findings of this study indicate that the MMSE and MoCA are well-correlated and show moderate agreement in detecting MCI among Chinese individuals aged 55 and older. However, the MoCA demonstrated less of a "ceiling effect" for MCI and was better at identifying the full range of cognitive abilities in the sample. A high overall prevalence of MCI was observed in both screenings. Modifiable risk factors for MCI included living in a county town or village, current smoking, hypertension, and depression, in addition to non-modifiable factors like increasing age and female gender. As China faces a rapidly aging population, the cognitive health of its elderly will inevitably decline, posing a significant challenge for the public health and medical care systems. Overall, these findings highlight the serious prevalence of MCI in the older Chinese population and provide important evidence for developing specific intervention strategies. Increasing public awareness about MCI and dementia, controlling modifiable risk factors to delay dementia onset, and improving the implementation of existing strategies by authorities could effectively reduce the prevalence of MCI and dementia in China.

Background

Thinking problems, like dementia, are a main reason many older people around the world cannot do things easily. This is true in China, where it creates big problems for leaders, doctors, and families. There is no cure for dementia. Also, brain changes often start years before clear thinking problems show up, and these changes may not be undone once problems are seen. Because of this, many researchers are now trying to slow down thinking problems in people who are in the very early stages.

Mild cognitive impairment, or MCI, is when a person's thinking skills decline more than expected for their age. People with MCI have problems with memory or other thinking skills that can be seen by tests, but these problems do not stop them from doing their daily activities. MCI is seen as a middle stage between normal aging and dementia. It affects 10 to 15 out of every 100 people over age 65. While some people with MCI get better, 5 to 10 out of 100 people with MCI will get dementia each year. This rate is much higher than for the general public. Many experts believe that MCI is a chance to step in and slow down or stop dementia. So, it is very important to check for MCI and find out what things might cause it in older people. This can help improve thinking and slow down the move to dementia.

To check for thinking problems, two common tests are used: the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA). Many studies show that the MoCA is better than the MMSE at finding MCI. This is because the MMSE sometimes does not catch small thinking problems in people who seem normal. Studies like this were done in China, but they were often in only one area and with few people. This means the results were not always strong or widely true. There was a need for bigger studies in many areas to compare these two tests for finding MCI in common people.

Also, it is important to understand what might cause thinking problems in MCI. This can help prevent or treat them, and slow down the path to dementia. Studies have found many things linked to MCI, such as age, gender, education, job, money, well-being, exercise, social life, food, and past sicknesses. But some of these findings are not agreed upon because studies were done in different countries or used different ways of testing. Education can especially affect scores on the MMSE and MoCA. Sometimes, people with more education even scored worse than expected compared to those with less education. It is also key to know if different things affect thinking problems when using different tests on the same group of people.

This study aimed to see how well the MMSE and MoCA matched up in finding MCI. It also looked at how different things affected MCI when found by the MMSE or MoCA. This study used information from a large group of middle-aged and older people in Chinese cities and country areas. The results may help choose the best tests for thinking problems and help manage MCI.

Methods

This study used information from a larger, ongoing health study that started in China in 2018–2019. This main study looks at things that might lead to certain brain diseases. For this study, the focus was on people aged 55 and older. These were people living in selected communities who did not have Alzheimer’s disease or other conditions that might affect their test scores. People were chosen from different areas in four Chinese provinces. They were interviewed in their homes after they agreed to be part of the study.

A total of 4923 people were part of this study. They were asked about their thinking skills and other health information. People who could not do basic daily tasks like eating or dressing were not included.

All people in the study took the MMSE and MoCA tests. Both tests are known to be good and reliable in China, after being adjusted for local language and culture. Trained people gave the tests face-to-face. The MMSE took about 5–10 minutes, and the MoCA took about 10–15 minutes.

Both the MMSE and MoCA tests have 30 points. A higher score means better thinking. The MMSE checks simple tasks like telling the time and date, remembering words, doing math, and using language. The MoCA checks eight areas of thinking, including memory and planning. Both tests have parts that check how well people do in different thinking skills. The study defined thinking problems in these areas if a person got any of the test questions wrong. For the MoCA test, if a person had 12 years of schooling or less, 1 point was added to their total score if it was less than 30. MCI was found using different cutoff scores for the MMSE and MoCA based on how many years of schooling a person had. For example, lower scores meant MCI.

The study also collected information about people’s age, gender, schooling, job, household income, where they lived, if they smoked or drank, how much they slept, and if they had high blood pressure, diabetes, stroke, or heart problems. People were also asked about their feelings, and if they scored high on a depression test, they were considered to have depression. This information was then grouped to be looked at more easily.

Next, the researchers used math methods to compare the test scores and information. They looked at how average scores were different across groups. They also checked if one test was better at finding small differences in thinking. They looked at how well the MMSE and MoCA scores matched up. They also checked how often both tests found MCI in the same people. Finally, they used special math tools to see if certain things like age, gender, or health problems were linked to a higher chance of having MCI.

Results

The study included 4923 people aged 55 and older. About 4 out of 10 people were 55–64 years old, about 4 out of 10 were 65–74, and nearly 2 out of 10 were 75 or older. More than half of the people were female (56.1%). About 18 out of 100 people had finished high school or more. Most people (82.8%) did not have a job, which included those who were retired. About 6 out of 10 people had a medium household income, and about 7 out of 10 slept the recommended amount. About 1 or 2 out of 10 people smoked or drank alcohol. Many had high blood pressure (31.8%), and fewer had diabetes (9.7%), stroke (2.0%), or heart problems (1.9%). About 9 out of 100 people reported feeling depressed.

The average score for the MMSE was 25.5, and for the MoCA, it was 22.6. The MoCA showed more differences in scores among people, meaning it was better at catching varied levels of thinking. The scores from the MMSE and MoCA tests were strongly linked, showing that if someone did well on one, they likely did well on the other.

There were clear differences in scores for both MMSE and MoCA based on a person’s age, gender, schooling, job, income, where they lived, if they drank alcohol, how much they slept, if they had high blood pressure, or depression. For example, people aged 55–64, those with more schooling, higher income, or living in cities generally had higher scores. The MMSE often had many people getting perfect or near-perfect scores in some groups (a "ceiling effect"), which made it harder to see small differences. The MoCA had less of this problem.

Using the cutoff scores for MCI, the MMSE found MCI in 28.6 out of 100 people, while the MoCA found it in 36.2 out of 100 people. Out of everyone in the study, 23.5% had MCI according to both tests. However, 12.7% of people who seemed normal on the MMSE actually had MCI when tested with the MoCA. This shows MoCA found more cases of MCI. The MMSE and MoCA agreed on finding MCI about 60% of the time, which is considered a fair match.

The study found that the number of people with MCI went up with age for both tests, with older people (75 and above) having the highest rates. The rate of MCI went down as household income went up. More people had MCI if they were not working, smoked, did not sleep enough, had high blood pressure, or depression. Also, where people lived (city vs. country) affected MCI rates. However, only the MoCA test showed clear differences in MCI rates based on a person's level of schooling.

When looking at specific thinking skills, like planning or memory, the MMSE and MoCA showed that many people had problems. For example, about 75% of people had problems with planning and remembering on the MoCA test. The MMSE showed similar problems, especially when MoCA scores were lower in those areas.

Things linked to a higher chance of having MCI were mostly the same for both MMSE and MoCA tests. People aged 75 and older had a much higher risk of MCI than those aged 55–64. Females had a higher risk than males. Living in a small town or village, smoking, having high blood pressure, and having depression all raised the risk of MCI. On the other hand, having a job and a higher household income lowered the risk of MCI.

Discussion

MCI is a common problem in older people. It means their memory, attention, and other thinking skills are not as good as they used to be, but they can still do their daily tasks. This study found that MCI was present in 28.6% of older Chinese people using the MMSE and 36.2% using the MoCA. The MMSE and MoCA scores were closely linked and showed a fair agreement in finding MCI. Also, things like older age, being female, living in a small town or village, smoking, having high blood pressure, and depression all raised the risk of MCI. These findings show that thinking problems are serious as the number of older people grows in China, creating big challenges for health leaders and the government.

The MMSE is widely used, but one problem is its "ceiling effect." This means that many healthy people get a perfect score, so it's hard to see small drops in thinking skills that might point to MCI. This study found that the MoCA had less of this "ceiling effect" than the MMSE. This means the MoCA was better at finding the full range of thinking abilities and was more likely to find MCI in people the MMSE missed. For example, 12.7% of people who scored as normal on the MMSE actually had MCI according to the MoCA. The MoCA also seemed better at finding problems in specific thinking skills, like planning and memory. Overall, these findings suggest the MoCA is a better test for finding mild thinking problems in middle-aged and older Chinese adults.

This study found a strong link between MMSE and MoCA scores. This means the tests generally measure similar things. The tests also showed a fair agreement in finding MCI. The reason they don't agree perfectly is likely because the MoCA was made specifically to find MCI, while the MMSE was first designed to find and track dementia.

The number of people with MCI can change a lot in different studies, depending on how MCI is defined and who is studied. This study found MCI rates of 28.6% and 36.2%, which fall within the range of other studies. The rates in this study were similar to or higher than some large studies in China. This might be because of different study methods or the specific groups of people included. Like other studies, this one found that MCI became more common with older age. Also, living in country areas, smoking, and having high blood pressure were linked to a higher chance of MCI. While less schooling usually means poorer test scores, this study found something unexpected: people with more education sometimes had a higher MCI rate when tested with MoCA. This might be because other things, like being active in the community, help protect thinking skills. The study also found that household income, job status, sleep, and depression were linked to MCI. For example, having a job and a higher income seemed to protect against MCI.

This study had some limits. First, it used Chinese versions of the tests, so comparing MCI rates with other countries might be tricky. Second, there was not enough information to look at how diet or genetics affected thinking. Also, the study did not use a "gold standard" test for MCI, so it could not fully compare how accurate the MMSE and MoCA were. Lastly, any MCI screening can miss some cases or wrongly identify others.

Conclusions

This study found that the MMSE and MoCA tests generally agree in finding MCI in older Chinese people. However, the MoCA was better at finding a wider range of thinking abilities and was more likely to find mild thinking problems. Many older Chinese people have MCI. This study showed that living in small towns or villages, smoking, high blood pressure, and depression were things that could be changed to lower the risk of MCI. Older age and being female were also linked to higher risk.

As the number of older people in China quickly grows, their thinking health will become a bigger challenge for the public health and medical systems. These findings show that MCI is a serious issue in China's older population. They also provide important reasons to create specific ways to help. Making people more aware of MCI and dementia, controlling things that raise MCI risk, and putting good plans into action can help lower the number of people with MCI and dementia in China.