Prevalence and associated factors of cognitive impairment among the elderly population: A nationwide cross-sectional study in China

Background Cognitive impairments are associated with increased risk for progression to dementia. In China, limited surveys have been conducted to estimate the national prevalence and risk factors associated with cognitive impairment in China. This study aims to assess the national prevalence and modifiable risk factors for cognitive impairments in the Chinese elderly population. Methods This cross-sectional study was based on the 2018 China Health and Retirement Longitudinal Study. The Mini Mental State Examination (MMSE) is recommended to test for cognitive impairment. Univariate and multivariate logistic regression models were used in assessing risk factors for cognitive impairments in the Chinese elderly population. Results A total of 3768 participants aged 60 years or older were enrolled in this study. The national prevalence of cognitive impairments was 22.24% in China, and the prevalence of cognitive impairment was higher in the south-west region than in the north region (29.94 vs. 16.53%, p < 0.05). The risk for cognitive impairments was higher in the following participants: not married or not living with spouse relative to married with spouse present (OR = 1.39, 95% CI, 1.15–1.70; p = 0.001), nap duration of ≥ 90 min relative to 30–60 min (OR = 1.54, 95% CI, 1.20–1.98; p = 0.001), sleep duration of ≥ 8 h relative to 6–8 h (OR = 1.73, 95% CI, 1.29–2.31; p < 0.001), and depression relative to no depression (OR = 1.67, 95% CI, 1.41–1.97; p < 0.001). The risk of cognitive impairment was lower in participants living in the urban areas relative to the rural areas (OR = 0.57, 95% CI, 0.47–0.69; p < 0.001) and consuming alcohol once a month relative to never consuming alcohol (OR = 0.69, 95% CI, 0.51–0.94; p = 0.02). Conclusion Cognitive impairment prevalence was high in the Chinese elderly population. The potentially modifiable risk factors for cognitive impairment should be further assessed in the development of interventions for the elderly Chinese population.


Background
Dementia is arguably the most feared and devastating disease affecting the elderly population; and is a leading cause of disability and dependence of aged individuals, worldwide (1). According to the World Alzheimer Report, in 2019, more than 50 million individuals suffered from dementia globally, and this number is estimated to surge to 152 million by 2050 (2). Cognitive impairment is associated with increased risk of disability, increased health expenditures, and progression to dementia (3). Problems in memory, language, thoughts, and judgment are more prominent than normal age-related changes in patients with cognitive impairments, although the basic daily activities are not disrupted (4). Cognitive impairment ranges from mild to severe, and is one of the most common and disabling non-motor symptoms in elderly individuals. Mild cognitive impairment (MCI) is also known as cognitive impairment without dementia. It is considered a preclinical transitional stage between healthy aging and dementia, and gradually progresses to dementia in nearly 10-30% of patients with MCI (5). The updated estimated prevalence of MCI is 15.5%, and the prevalence of severe cognitive impairment (dementia) is approximately 6.0% in individuals aged 60 years or older in China (6). Notably, no effective medication is currently available for the treatment of cognitive impairment. Therefore, identifying the etiologies of cognitive impairments and suppressing the incidence are more important than treating them following their onset.
In the past decades, rapid demographic and epidemiological transition, have led to an aging population in China. Thus, the country has a large number of people with cognitive impairments. Many studies have investigated the prevalence and risk factors for cognitive impairments in China's general population (6)(7)(8)(9)(10). Older age, being female, rural residence, illiteracy, living without a partner, smoking, hypertension, hyperlipidemia, diabetes, heart disease, and cerebrovascular disease are the main risk factors for dementia and MCI (6)(7)(8)(9)(10). However, the prevalence and risk factors are vary among cities or regions of China, and limited surveys have been conducted to estimate the national prevalence and risk factors associated with cognitive impairments. These inconsistencies require further study for a realistic estimate.
In the present study, by conducting a nationwide crosssectional survey, we aimed to estimate the prevalence of cognitive impairments and its modifiable risk factors in the elderly in China. The findings can help to enhance understanding of cognitive impairments and strategies for protecting the elderly population against cognitive decline.

Study sample and data cleansing
To investigate the prevalence of cognitive impairments in the elderly, the China Health and Retirement Longitudinal Study (CHARLS) 2018 follow-up dataset was downloaded and analyzed. The CHARLS was hosted by Peking University and approved by the Peking University Ethics Review Committee (IRB00001052-13074). Initiated in 2008, the CHARLS uses probabilities proportional to the size method to sample aging populations aged 45 years and above in the whole of China, with follow-ups in 2011, 2013, 2015, and 2018. The data were obtained from 150 counties and 450 villages in 28 provinces. The dataset from CHARLS are representative and of high quality. Detailed description and specific study design of the CHARLS project can be accessed from previous publications (11, 12) or its official website (http://charls. pku.edu.cn/). In the 2018 survey, the cognitive functions of participants aged 60 years and above were evaluated with the Mini-Mental State Examination (MMSE) questionnaire (13,14). The flowchart of data cleansing is displayed in Supplementary Figure 1. In the present study, a total of 3,768 participants were analyzed.

Assessment of cognitive function
Cognitive functions were assessed using the MMSE questionnaire, which comprises 30 items with scores ranging from 0 to 30. The questionnaire is widely used in epidemiology for cognitive function assessment. The MMSE consists of seven aspects: orientation to time (five items), orientation to place Frontiers in Public Health frontiersin.org . /fpubh. .
(five items), registration (three items), attention and calculation (five items), recall (three items), and language (nine items). Assessment was performed by two well-trained staff through a face-to-face interview with the native dialect. According to previous studies (10,(13)(14)(15)(16), the cutoff of MMSE was set at 16/17 for illiterate individuals, 19/20 for individuals with 1-6 years of education, and 23/24 for individuals with at least 7 years of education. An MMSE score lower than the above-described cutoff values indicated cognitive impairment.
Depression was assessed using the Center for Epidemiological Studies Depression Scale-10 questionnaire (14,20). Medical histories were mainly based on the self-reports of the respondents. Given differences in living and cultural habits, the living localities of the participants were categorized into six regions in the same manner as previous studies did (17, 21): north (Shanxi, Hebei, Beijing, Tianjin, and Inner Mongolia), northeast (Jilin, Liaoning, and Heilongjiang), east (Jiangsu, Fujian, Shanghai, Shandong, Zhejiang, Jiangxi, and Anhui), northwest (Qinghai, Shanxi, Xinjiang, and Gansu), southwest (Sichuan, Chongqing, Yunnan, and Guizhou), and south-central (Hunan, Henan, Guangdong, Hubei, and Guangxi).

Statistical analysis
Data of clinical characteristics were summarized as proportions (%) according to data type. Descriptive statistics were used to investigate the prevalence of cognitive impairments in different groups. Moreover, univariate binary logistic regression was used in scanning risk factors for cognitive impairments. Then, correlation coefficients were calculated using Spearman tests to prevent collinearity before multivariate regressions were performed. The correlations among the covariates were low (Pearson correlation coefficients < 0.5, Supplementary Figure 2), and thus no variable was deleted (22). Variables with p of < 0.05 in univariate logistic regression were included in multivariate logistic regression analysis. Therefore, age, marital status, residence, alcohol consumption, afternoon napping, sleep duration, depression and liver disease were included in the multivariable logistic regression model. All the analyses and figures were made using R 4.0.1 (R Foundation for Statistical Computing, Vienna, Austria). A p-value of < 0.05 (two-sided) indicated statistical significance.

Results
Baseline characteristics and prevalence in the grouped population  Table 1. The overall prevalence of cognitive impairment was 22.24% (95% CI, 20.94-23.60). In total, 3,768 participants comprised 2,451 (65.0%) males and 1,317 (35.0%) females, and 81.05% of the participants were married. About 30.8% of people lived in urban areas, 60.7% in rural areas, and the remaining in the urban-rural integration zone. The participants were divided into groups based on cigarette use, alcohol consumption, nap duration, and sleep duration, as shown in Table 1.

Age-and gender-specific prevalence
As shown in Figure 1, the prevalence of cognitive impairment increased with age. The prevalence rates were 21

Gender-specific prevalence across regions
Gender-specific prevalence in different regions is displayed in Figure 2. The highest prevalence of cognitive impairments was found in the southwest region (29.94%); and the lowest, in the north region (16.53%). Difference in prevalence was found between males and females across regions. The prevalence of cognitive impairments in females was 1.68 and 1.38 times that in
Multivariate logistic regression models were used in evaluating risk factors for cognitive impairment. The final multiple logistic regression model included age, marital status, residence, alcohol consumption, nap duration, sleep duration, depression and liver disease. The prevalence of cognitive impairment was higher in participants who were not married or were not living with spouse compared with those married living with their spouses (OR = 1.39, 95% CI, 1.15-1.70; p = 0.001) and participants with nap duration of ≥90 min compared with those of nap duration of 30-60 min (OR = 1.54, 95% CI, 1.20-1.98; p = 0.001), participants with sleep duration of ≥8 h compared with those with sleep duration of 6-8 h (OR = 1.73, 95% CI, 1.29-2.31; p < 0.001), and participants with depression compared with those without depression (OR = 1.67, 95% CI, 1.41-1.97; p < 0.001). Notably, a decreased risk of cognitive impairment was observed in participants living in urban areas compared with those living in rural areas (OR = 0.57, 95% CI, 0.47-0.69; p < 0.001) and in participants with a low-alcohol consumption (less than once a month) compared with those who never consumed alcohol (OR = 0.69, 95% CI, 0.51-0.94; p = 0.02), indicating low risk of cognitive decline.

Discussion
In this nationwide cross-sectional study, we estimated the prevalence and associated risk factors for cognitive impairments in the Chinese elderly population. The highest prevalence of cognitive impairments was found in the southwest region of China, and the lowest was found in the north region, indicating regional differences. Several associated risk factors, including marital status, urban or rural residence, sleep and nap durations, depression, and alcohol consumption, were identified.
With regard to aging, the global elderly population is growing rapidly, especially in mainland China. According to China's Seventh National Population Census in 2020, the . /fpubh. .  Logistic regression was adopted to identify the associated independent factors of cognitive impairment. All plausible variables with P < 0.05 in univariate testing were subjected to further multivariate testing. The crude ORs were calculated in univariate regression, and the adjusted ORs were recorded using multivariate regression. OR, Odds Ratio; CI, Confidence Interval. number of elderly people over 60 years has reached 264.02 million, accounting for 18.7% of the total population (23). Rapid growth in the elderly population has stimulated interest in elucidating the causes of cognitive impairments and strategies for preventing them. Lu (16) conducted a study in Ji County of Tianjing (a rural area of northern China) and suggested that the prevalence of cognitive impairment is 38.3% (27.8% MCI and 10.5% dementia) in the overall population aged 60 years or older. After studying 96 sites from 12 provinces, Jia (6) suggested that the prevalence of cognitive impairment was 21.5% (15.5% MCI and 6.0% dementia) in the overall population aged 60 years  (27). This study found an apparent geographical variation in the prevalence of cognitive impairments in China. As shown in Figure 2, the results of prevalence distribution suggests that the incidence in western China (southwest and northwest regions) was the highest, and the prevalence in the southwest region was 1.81 times that in the north region. A meta-analysis reported that the pooled prevalence of dementia was the highest in western China (9.6%), intermediate in northern China (5.4%), and lowest in central China (3.8%) and south China (3.7%) (25). Another meta-analysis reported that the pooled prevalence of MCI was higher in western China (14.33%) than in eastern China (13.41%) (24). Dietary differences may contribute to this discrepancy. The daily diet of participants living in the north regions contains considerable amounts of milk, dairy products, and flour-based food, whereas participants living in the southwest regions consume more fruits and rice-based diet; dairy products have been confirmed to have a protective effect against cognitive impairment (28). Other differences may be attributed to the uneven economic, educational development, and different living habits across regions in China.
The identification of specific risk factors is crucial for the prevention of cognitive impairments. The prevalence of cognitive impairment was higher among elderly people, females, people who are not married or cohabitating, and people living in rural areas or western China, consistent with the findings of some previous studies (6-8, 10, 24). The present study provides further evidence in support of these risk factors. The results showed that depression is associated with the high prevalence of MCI in the Chinese elderly population, consistent with previous findings (14,23). Jia et al. (6) showed high prevalence of cognitive impairment in the Chinese elderly population with hypertension (odds ratio: dementia = 1.86; MCI = 1.62), hyperlipidemia (dementia = 1.87; MCI = 1.29), diabetes (dementia = 2.14; MCI = 1.44), heart disease (dementia = 1.98; MCI = 1.17), and cerebrovascular disease (dementia = 5.44; MCI = 1.49), but the above risk factors were not found in the present study. Medical histories were mainly based on the self-reports of the respondents, which may lead to deviation from the present results.
The present study suggested that some other risk factors are modifiable, including nap duration, sleep duration, and alcohol consumption, which are rarely explored. Sleep disturbances are common in the elderly, and approximately 50% of people aged over 65 years reported a chronic sleep complaint (29). Many studies have examined the associations between sleep duration and cognitive impairment. Moreover, The results also suggest that long durations of napping (≥90 min) and long sleeping (≥8 h) are associated with high prevalence of cognitive impairment and an afternoon nap duration of 30-60 min and night sleep of 6-8 h are associated with enhanced cognitive function. The exact biological mechanisms linking excessive sleep nap duration to cognitive impairment remain unclear (30). Further studies are needed to examine whether excessive sleeping or napping is a subtle marker of cognitive impairment in otherwise healthy elderly individuals.
Epidemiological studies have indicated that excessive alcohol consumption can induce cognitive impairments, whereas moderate consumption may reduce the risks for cognitive impairment in the elderly (31, 32). The present study suggested a similar result, that is, consuming alcohol less than once a month is associated with a lower prevalence of cognitive impairment (OR = 0.65, p = 0.005) than that in participant who never consumed alcohol. The potential factors linking low alcohol consumption to cognitive function have been attributed to flavonoids or other antioxidants, which may reduce the risk of cognitive impairment (33).
The present study has some limitations. First, we did not classify MCI and dementia. Clinical diagnosis based on the MMSE, the Montreal Cognitive Assessment, Clinical Dementia Rating score, magnetic resonance imaging, or computed tomography is necessary for MCI and dementia (6). MMSE as a single diagnostic tool is insufficient to diagnose MCI and dementia, and different diagnostic confirmation tools might result in different prevalence rates. Second, gender selection bias was found; more male participants were selected. Third, relying on the participants' self-reporting, the medical histories, nap durations, and sleep durations might not be accurate measures and might generate bias. In the evaluation of covariates, we did not employ objective monitoring devices because collecting data from a large cohort is difficult. Additionally, this cross-sectional study cannot establish causal relationships between the identified associated factors and cognitive impairment, and future longitudinal studies are needed to clarify these associations.

Conclusion
In this nationwide cross-sectional study, the overall prevalence of cognitive impairments was 22.24% in the Frontiers in Public Health frontiersin.org . /fpubh. . participants aged 60 years or over. The prevalence varied among age groups, living areas, regions and between genders. The results revealed that the modifiable risk factors for cognitive impairment were as follows: not married or cohabitating, rural residence, long duration of napping (≥90 min), long duration of sleep (≥8 h), and depression. Thus, preventive strategies for cognitive impairment are needed to improve cognitive function.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

Ethics statement
The CHARLS study was approved by Research Ethics Committees of Peking University (IRB00001052-13074). The patients/participants provided their written informed consent to participate in this study.  . /fpubh. .