Prevalence and Risk Factor of Cognitive Impairment were Different between Urban and Rural Population: A Community-Based Study

INTRODUCTION

With the rapid aging of populations worldwide, cognitive impairment (CI) has become the most serious public health problem in developed and developing countries [1, 2]. Because the treatment for CI, especially Alzheimer’s disease, is limited, prevention becomes an effective way to delay its onset.

In our previous pilot study in Shanghai, we identified several risk factors for CI, such as high fasting glucose concentration, physical inactivity, and having more children [3]. However, this study included only adults living in urban community. There might be disparities in relevant risk factors of CI between urban and rural areas due to the big difference in economic growth, lifestyle, and accessibility to health service [4]. In addition, low education, low income, pesticide exposure, and use of tobacco are prevalent in Chinese rural areas [5, 6], which could lead to a higher prevalence of cognitive dysfunction in rural than urban areas [7, 8, 7, 8].

At present, studies directly comparing urban and rural populations were rare, especially in Chinese populations. Only a few studies suggested that there was a higher prevalence of dementia in rural areas than that in urban areas, and these may be due to difference in educational level and economic status [9–11]. We thus conducted a cross-sectional survey of CI in urban and rural areas in Shanghai to compare the difference of prevalence of CI and potential risk factors in these communities.

METHODS

RESULTS

There were 42.7% men in the rural population and 32% in the urban population. The average age was similar between two groups (64.9 versus 64.8 y) (Table 1). The rural population had more labor workers, smokers, and alcohol drinkers, higher prevalence of self-reported diabetes mellitus, hypercholesterolemia, coronary heart disease, and stroke, and lower prevalence of hypertension, relative to the urban population (p <  0.05 for all). Urban individuals were high level educated (completed senior school education) compared to rural ones (p <  0.001).

In the rural community, 329 participants (7.43% ) were diagnosed as CI. Women had a higher prevalence of CI, relation to men (8.95% versus 5.39% ; p <  0.001) (Table 2). The prevalence in the urban community is 6.54% . However, we did not observe a significant gender difference (6.82% in women versus 5.94% in men; p = 0.32).

In general, prevalence of CI was similar between rural and urban populations (p = 0.13) (Table 2). When we stratified by gender, we found that rural women had a higher prevalence compared to urban women (8.95% versus 6.82% ; p = 0.006). The prevalence of CI in urban men was not different from rural men. In both urban and rural population, the prevalence of CI increased dramatically with age, going over 10% among people aged 80 years and older. The prevalence rates for subgroups stratified by educational level and gender are presented in Fig. 1. The prevalence of CI for illiterate people was significantly higher than other groups (illiterates: 14.90% , 95% CI: (12.97, 16.85); primary school graduates: 5.69% , 95% CI: (4.66, 6.71); junior school graduates or above: 5.40% , 95% CI:(4.75, 6.05)).

Comparing associated factors between rural and urban communities, we found that age was associated with increased risk of having CI in both rural (OR = 1.06, 95% CI: 1.03–1.08) and urban populations (OR = 1.04, 95% CI: 1.01 –1.08) (p-interaction = 0.16). But the risk factors for rural populations also included female gender (OR = 2.03, 95% CI: 1.08–3.82, p-interaction = 0.04), history exposure to pesticides (OR = 4.68, 95% CI: 1.27–17.21, p-interaction = 0.99), history of head trauma (OR = 1.89, 95% CI: 1.10–3.24, p-interaction = 0.05), and history of encephalitis or meningitis (OR = 6.02, 95% CI: 1.92–18.85, p-interaction = 0.51).

In contrast, in urban populations, CI was associated with diabetes mellitus (OR = 1.79, 95% CI: 1.04–3.07, p-interaction = 0.01), having more than two children (OR = 2.39, 95% CI: 1.27–4.50, p-interaction <0.001), and lack of physical activities (OR = 2.25, 95% CI: 1.11–4.57, p-interaction = 0.01) (Table 3).

DISCUSSION

We found that the prevalence of CI was higher in rural population than urban population and rural women population had the highest prevalence of CI in our study. With regards to risk factors of CI, features associated with the rural population included women, exposure to pesticides, head trauma, and encephalitis or meningitis. While in the urban population, associated risk factors included diabetes mellitus, lack of physical activities, and having more than two children.

Our finding that prevalence of CI in the rural population was higher than the urban population was consistent with the study by Jia et al. [11]. As seen in Table 1, people in rural area had a much lower educational level than people in urban area. Several studies suggested that cognitive reserve was associated with high educational level and greater lifetime exposure [13]. It was possible the rural people were more vulnerable to cognitive dysfunction because of poor cognitive reserve. Other reasons, such as low socioeconomic status and health service access, might also contribute to the high prevalence of CI in rural area [4]. In addition, prevalence of CI in women was much higher than men especially in the rural community, which was in consistent with the finding of Irvine et al., and the possible underlying mechanism might be related to less cognitive reserve and estrogen reduction in women [14].

Risk factors for CI were discovered to be different between urban and rural populations in our study. Firstly, women in rural areas, especially above the age of 80, had a much higher prevalence of CI and this was proved by other studies [11, 15]. In addition, a history of exposure to pesticides, head trauma, and encephalitis or meningitis contributed to CI in the rural population versus the urban setting. The plausible explanation was that people living in the rural community were mostly farmers (54.4% ), thus were more likely to be exposed to pesticides and liable to infectious disease or head trauma. Head injury has received a great deal of attention in CI due to the chronic traumatic encephalopathy (CTE) in athletes [16]. According to some epidemiological studies, head trauma is one of the most important environmental risk factors for cognitive decline. Neuropathological findings pointed out that brain atrophy combined with other gross features such as hippocampal sclerosis and ventricular enlargement were seen in autopsy outcomes in some athletes with CTE. CTE also shares many microscopic similarities with Alzheimer’s disease, including neurofibrillary tangles, neuropil neurites, and amyloid-β deposits [17]. In addition to repetitive mild head trauma, amyloid peptide deposition and tau pathology may also be seen after traumatic brain injury and were proved by animal models and postmortem examination [18, 19].

Compared to the rural population, diabetes mellitus and lack of physical activity were the risk factors for CI in the urban population. The proportion of people with diabetes mellitus was higher in the urban area than that in rural area in our study (15.0% versus 9.4% ), which was consistent with the study by Liu et al. [20]. The strong link between cognitive function and diabetes mellitus, combined with other findings on dietary features with dementia, calls to attention the need to foster a healthy lifestyle [21–23]. Also, our results pointed out that physical activity had direct influence on cognitive function in the urban population. Therefore, suitable activities might be an indispensable part, especially for urban residents.

Interestingly, we found that having more children (≥3) increased the risk of CI in the urban population, which was in contrast to the findings of Sundstrom and colleagues that having no children increased the risk of dementia [24]. It was possible that having more children usually means a big socioeconomic burden for parents and limited food for the whole family, which contribute to CI indirectly [25, 26]. Certainly, this finding needs to be confirmed in other studies, and further well-designed studies are needed to reveal the possible link between the number of children and CI in Chinese populations.

Our study has some limitations. First, our sample only included about 18–25% people from targeted population in two communities of Shanghai, because elderly people are usually reluctant to participate in epidemiological studies in China. Also, our sample was not randomized and only included two communities in Shanghai. It was thus not a perfect representative sample for the whole country of China. Secondly, we used MMSE score to define CI, which is less sensitive for certain cognitive domains and also has a bias with respect to education [27, 28]. The MMSE is still a globally-used instrument for cognitive screening due to its convenience and relatively good discriminant validity [29, 30]. However, further large randomized, multi-cities based studies with comprehensive cognitive assessment will be needed in the future to overcome these problems.

In conclusion, our study showed a growing prevalence of CI compared to the previous epidemiological surveys in China. There were common risk factors for CI both in rural and urban areas, such as age. However, disparity was also seen between rural and urban communities. Female gender, exposure to pesticides, history of head trauma, and history of encephalitis and meningitis contributed to CI in rural communities. While in urban areas, having more than three children, diabetes mellitus, and lack of physical activities were more important for the risk of CI. Different living environments and socioeconomic status might explain these differences to some degree. Therefore, the increasing CI prevalence may be controlled by the improvement of hygiene conditions and self-protection awareness. These findings proposed different strategies for prevention of CI in rural and urban elderly.