Access to safe drinking water is crucial to health and development issues. It is estimated that about a quarter of people in the home lack safely managed services of drinking water, and nearly half of the world's population lacks safety management of health facilities (1, 2). More than 200 million people in rural areas in China do not have access to safe drinking water due to limited water services and lack of complete water treatment facilities, compared to almost all cities with well-equipped water treatment facilities (3, 4). In addition, the accumulation of human and animal waste in rural areas and the abuse of chemical fertilizers and pesticides have exacerbated the problem of drinking water safety in rural areas (5, 6).

Due to China's geographical and economic constraints, many rural areas receive water through small centralized water supply systems (SCWS) (4). This type of water supply is provided and managed by the villages themselves, with few regular monitoring reports and a lack of proper water purification and disinfection facilities (7, 8). Yunnan Province, where the study area is located, has one of the lowest qualified rates of water quality in China (7). In 2013, a survey of 1,067 rural drinking water safety projects in Yunnan province showed that only 14.34% had complete treatment facilities, and only 66.83% of treated water samples were qualified (7). It is hard to find reliable and affordable safe water in rural areas. As a result, the burden of providing safe water falls on households. Tengchong, is dotted with hot springs and geothermal fields whose discharge can contaminate groundwater and surface water, putting residents' health at risk (9, 10).

In developed countries, tap water is subject to strict quality control and management (11, 12). In most cases, tap water can be directly drunk (12, 13). However, in China, due to poor water quality and secondary pollution, direct drinking tap water has certain risks (14). The habit of drinking boiled water, which greatly reduces the risk of diarrhea and other gastrointestinal diseases in China (15). However, there are still many rural residents with poor health habits and low health awareness, who drink unboiled tap water or even raw water. A survey conducted in Tengchong County found that the infection rate of Entamoeba histolytica among hospitalized patients was twice the average for the Chinese population, more than half were rural residents, and was highly associated with drinking unboiled tap water (16). Point-of-use household water treatment (HWT) is often considered a stop-gap solution, yet in many low-income rural areas, where households are still remain responsible for treating drinking water (2). Based on the evaluation criteria used, ceramic, bio-sand, and domestic water filters are considered to be the most effective and have the greatest potential for widespread and sustainable use in improving household water quality to reduce water-borne diseases and deaths (17).

Previous studies in developed countries have extensively investigated the influencing factors of residents' drinking water behavior and drinking water choice, including the use of bottled water, tap water, and filters (13, 17, 18). However, there are few studies on residential water use behavior in China (15, 19). In fact, access to safe drinking water and healthy water use behavior has become the primary demand and goal of the Chinese public and government. With the deepening of the aging degree, the outflow of the young rural labor force, there are a large number of rural left-behind middle-aged and elderly residents (20, 21), and they are low income, poor living conditions, inadequate social security, education level is generally not high, is a vulnerable group that cannot be ignored. To ensure drinking water quality and guide residents to drink correctly, it is vital to explore the influencing factors of drinking water behavior and choices for drinking water providers and the government.

On this basis, this research randomly selected the elderly residents of six townships, in Tengchong, Yunnan province as the research object to carry out a cross-sectional survey. Possible influencing factors such as gender, age, ethnicity, income, education, access to drinking water, and sanitation were investigated. We innovatively included residents' perceptions of drinking water in our study to provide a more comprehensive analysis. This study investigated the current situation of drinking behavior and filter use intention of middle-aged and elderly residents under the SCWS and further explored the influencing factors.

Drinking water behavior reflects residents' awareness and philosophy of health and drinking water sanitation. However, few studies have examined the factors that influence drinking water behavior in China, especially in rural areas where drinking water safety is a major concern. We conducted a cross-sectional survey in six villages in Tengchong County, Yunnan Province, where water quality is poor and SCWS is common in China, to understand the drinking of unboiled tap water by middle-aged and elderly rural residents and the possibility of promoting filters. The results showed that a large number of middle-aged and elderly residents also drank unboiled tap water to some extent (35%). In addition, only 7.6% of the residents in our survey have any drinking water filtration facilities in their homes, and only 29.8% of them are willing to use filters. Many factors affect residents to drink unboiled tap water and the willingness to use filters, so explore the reasons and formulate measures and strategies to eliminate the behavior of middle-aged and elderly rural residents to drink unboiled tap water. At the same time, understanding the willingness of residents to treat water with filters is of great significance for the construction of rural drinking water safety.

Our study found that the elderly were more likely not to drink unboiled tap water, and the older they were, the less likely they were to drink unboiled tap water. This may be because compared with the middle-aged, the elderly generally have lower body resistance, are more likely to be affected by various external risk factors, and suffer from a high proportion of chronic diseases. Years of treatment and medication experience will also make them pay more attention to health (27, 28). Generally speaking, the level of education often reflects the level of health knowledge (29), but in our regression model, the relationship between education level and drinking unboiled tap water is not significant (p = 0.082). This may be due to the fact that middle-aged and elderly rural residents generally have a low level of education. In the appropriate stage of education, they are trapped in the backward social and economic development in China and lack educational resources (30). Most of them do not receive systematic elementary education, resulting in generally low health literacy (31, 32). However, it should be noted that residents with higher education levels (middle school and above) still drank less unboiled tap water than those with lower education levels (primary school and below) (26.43% vs. 38.33%). In addition, residents who want to gain knowledge about drinking water are more likely not to drink unboiled tap water. This is because human behavior is heavily influenced by personal and situational motivational factors, which is the core idea of self-determination theory (33, 34). When residents want to acquire knowledge about drinking water, there is an intrinsic motivation to engage in healthy behaviors, which makes sense why residents who actively acquire knowledge about drinking water are more likely to avoid drinking unboiled tap water. Meanwhile, respondents who intake water at the regular interval were more likely to avoid unboiled tap water than those who drank only when they were thirsty. The scientific way of drinking water is at regular intervals intaking water, especially the elderly should take sufficient water every day, and those who drink only when they are thirsty often neglect the health of drinking water (35, 36), so they had more possibility of drinking unboiled tap water. We also found that residents with kidney stones were more likely to drink unboiled tap water. At present, the effect of drinking water quality on kidney stone disease has been a long-term debate (37). It is generally believed that kidney stones are related to the amount of water consumed, and the simplest and most important lifestyle change to prevent stone disease is to drink more water/fluid (38). While people who drink at regular intervals tend to intake enough water throughout the day, people who drink only when they are thirsty tend not to intake enough water. And studies have found that residents who drink hard groundwater water have a higher risk of kidney stones than those who drink treated water (39), and boiling can effectively reduce the hardness of drinking water (40). No wonder residents with kidney stones are more likely to drink unboiled tap water. We also found that residents with higher perceived health risk (PHR) were less likely to drink unboiled tap water. Previous studies have found that residents avoid unsafe water at home by boiling or filtering or purchasing bottled water based on their perception of water quality (41). Therefore, the health risk perception of drinking water can reflect residents' drinking water behavior to a certain extent.

Our study found that differences in age, education, and household income in influencing water filter use choices are both similarities and differences with western countries (18, 25, 42). These factors show the characteristics of Chinese tradition and social and economic background. Older people were more likely than middle-aged people to be reluctant to use filters because elderly people in rural China have low incomes and generally low levels of education (31), as well as traditional lifestyles and long-term use of boiled tap water. In previous studies, higher education levels were also more likely to choose alternative water (filters and bottled water) (19, 42). And our research also confirms this, with more educated residents more likely to be willing to use filters. Although family income was not significant in our regression, in univariate analysis, the proportion of residents with higher household income willing to use filters was still higher (P < 0.037). In addition, in self-perception theory, it is generally believed that a person's behavior can express his or her true attitude (43). Therefore, residents' behavior of treating water at home reflects their attitude of wanting to improve drinking water sanitation, which explains that in our survey, residents with household water treatment behavior are more likely to be willing to use filters. Many residents in China's rural areas have the habit of water storage. However, this behavior has certain health risks, and it is easy to cause secondary pollution (44). Among all the residents who can store water, 51.49% use plastic buckets with no covers to store water, while the rest use self-built water cellars and tanks, where the safety of water is not guaranteed. It can be seen that these residents who have the habit of water storage often lack the awareness of healthy water use, so they are more likely not willing to use filters to improve water sanitation. PHR was also an important factor in our second regression. Similar to the Kenyan study, as the perceived risk to water increases, households are more likely to dispose of unimproved drinking water (45), and thus more likely to be willing to use filters. Residents' satisfaction with water supply institutions often affects residents' views on domestic water supply, both in water quality perception and health risk perception, as well as residents' choice of water and willingness to treat water at the household level (25, 46). Our results showed that the higher the WAT, the lower the willingness of residents to use filters.

This is the first survey of water use behavior and willingness of middle-aged and elderly residents in rural China. In recent years, China has made great achievements in strengthening infrastructure construction and promoting projects to upgrade drinking water and toilets in rural areas. However, there are still considerable problems with drinking water for rural residents, especially the middle-aged and elderly with low education, low income, and ill-informed. A large proportion of them drink unboiled tap water. At the same time, residents showed a lower willingness to use filters. This is mainly related to the drinking water status, the health literacy of residents and the stage of social and economic development in rural China. In addition, understanding residents' perception of drinking water can provide a more comprehensive understanding of residents' drinking water behavior. Therefore, policymakers should pay more attention to the improvement of health knowledge and cultural literacy of middle-aged and elderly rural residents (47), especially the middle-aged group should not be neglected. Health education can be expanded to indirectly affect the drinking water behavior of middle-aged and elderly people by guiding families and children (48). At the same time, regular health lectures were carried out to guide and standardize the drinking water behavior of middle-aged and elderly residents, and to eliminate the behavior of drinking unboiled tap water. In addition, promote safe, effective and economic filtration and water purification facilities in rural SCWS areas to improve access to safe drinking water. Most importantly, water authorities and the CDC should be urged to strengthen the management and supervision of drinking water quality to protect public health.

It is important to mention some limitations of our study. Firstly, our analysis used a cross-sectional study design that only indicates the current domestic water behavior, and the results cannot infer causality. Secondly, limitation of sample selection, research results only represent the behavior and will of residents in this survey area, not other regions. Additionally, we only focus on middle-aged and elderly living in rural areas, but there are still large left-behind children who also face the same problem. Moreover, participants self-reported that the relevant information may be biased, and we did not explore the influence of chemical indicators of water quality on the results. These limitations also point out the direction for our further research.

We found that a certain proportion of permanent middle-aged and elderly residents in rural areas still drink unboiled tap water, especially those in middle age, and there is a low willingness to use filters. Economic and cultural level, health literacy, drinking water behavior, and drinking water perceived are the important factors influencing the willingness of middle-aged and elderly residents to drink unboiled tap water and use filters. Residents over 60 years of age who have regular drinking behavior, want knowledge about drinking water, and have a high level of PHR are more likely not to drink unboiled tap water. Residents with a middle school education or above, household water treatment behavior, and a high level of PHR are more likely to be willing to use filters. Therefore, the government and CDC should strengthen the popularization of drinking water knowledge for middle-aged and elderly rural residents, regularly carry out health education and other related measures, strengthen the regulation of middle-aged and elderly residents' drinking water behavior, eliminate the behavior of drinking unboiled tap water. Promote safe, economical and effective domestic water filtration facilities in rural SCWS areas to ensure public health safety.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Ethical Committee of Dalian Medical University. The patients/participants provided their written informed consent to participate in this study.

YD and LZ: conceptualization. YD and YP: data curation. YD: formal analysis and writing—original draft. RW, XC, JX, and MS: investigation. YD, HJ, and LZ: methodology. LZ: project administration, resources, and writing—review and editing. RW and HJ: software. XC, JX, and YC: supervision. YC and YP: validation. MS: visualization. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.