We estimated the magnitude and distribution of AGI in the community in China using a large-scale population-based study. Our results showed that the prevalence of AGI in the population was 2.3%, corresponding to an incidence of 0.3 episodes per person per year. In other words, our study suggests that one in every three Chinese residents suffers from AGI every year. This was the first survey provide nationwide representative estimates of the prevalence and incidence of AGI in China.
Applying the synchronized interviewing cycle and the same questionnaire, this survey covered not only north and south representative provinces in China, but also provinces representing the northwest (Gansu province) and southwest (Sichuan province), thus providing national representative data on the incidence and distribution of AGI in the population. We applied a face-to-face methodology similar to what has been used in other stucies [12, 13]. Similarly to those studies, we achieved a high average response rate of 97.2% when compared to telephone-administered surveys conducted in most developed countries. The high response rate was not only determined by the face-to-face interviewing, but also greatly benefited from cooperative street residents' committee from selected residential district or villages.
Seasonality and geography
The results showed a seasonal distribution of AGI during the study period that is in line with the normally expectation- of high incidence of AGI in summer and low incidence in winter. The data showed it’s peak in sping and summer, and lower incidence in autumn and late winter, but relative high in early winter. The variation is likely to reflectthe seasonal variation in infections with viral enteric pathogens in colder months，as seen in other studies [14–17]. China is so large that temperatures and other meteorological factors typical of different seasons vary by province during the same period. For example, when the southern region is in summer, the northen region is in early sping, and in topography descends from the west to the east. As a result, the seasonality of AGI of bacterial origin is not synchronized. An analysis stratified by province and season would be needed to clarify such variations.
We did not observe significant differences in the likelihood of having AGI in individuals with different leves of education, with different levels of income, or living in urbanized or rural areas. Travelling within two weeks was a significant risk factor, which is in accordance to what seen elsewhere .
Care seeking behavior and utilization of healthcare services was high in all age groups. Half of survyed individuals that reported AGI went to primary and lower health care institutions for medical care, which represents a higher rate of seeking medical care than estimated by other countries such as the United States (19%) , Ireland (19.5%) , Canada (20.4%) , Australia (19.5%)  and Denmark (12%) , Cuba (17.1%-38.1%)  and Argentina (26%) . It remains to be investigated if this disproportion is caused by the attention to disease, the attitude towards medicine., different medical insurance systems or differences in healthcare systems. The collection rate of biological samples in this survey was 14.3%, which is close to Canada (14.4%)  and Ireland (14.9%) , but lower than Australia (18.4 %)  and the United States (21.1%) . These results showed that the attention of hospitals in China to investigating the causes of AGI still needs to be improved, and suggest that the central and local governments should increase awareness and investment to improve the sampling rate of patients' and the rate of laboratory diagnistics, so as to better grasp the etiology of acute gastroenteritis.
The hospitalization rate of AGI in our survey was 7.9%, suggesting a high proportion of severe cases of AGI in China. This index was not included in the previous domestic survey .
Nearly 80% of cases were treated with drugs, among which 62% use antibiotics, a much higher proportion estimated in other countries (US: 8.3%, Ireland: 5.6%, Canada: 3.8%, Australia: 3.6%, Italy: 6.5% Argentina: 7%, Cuba: 6.5-18.9%) [12, 14, 18–20]. Oversuse or misuse of antimicrobials may lead to antimicrobial resistance, an important public health concern globally. These estimates suggest that there is still a long way to go to popularize the legal knowledge of the scientific use of antibiotics in China.
The estimated incidence of AGI in China of 0.28 (95%CI: 0.23-0.34) episodes per person was lower than the rate of 0.56 (95% CI: 0.56-0.57) episodes per person per year estimated by the previous survey conducted between 2010 to 2011 . However, they were similar to the estimation of 0.31 episodes per person per year calculated by reviewing scientific literature . The differences in the estimates of the different surveys may be explained by several factors. First, differences in the selection of provinces in the survey design, since China has great population diversity, which may also be reflected in diarrhea incidence and risk factors. All the provinces selected by the previous survey were southern provinces. On the contrary, our study took place in eight provinces that represent all the traditional seven regions of China . Second, in the past several years, the Chinese government has continuously improved the food safety regulation systems, developed innovative regulation mechanisms, established a structure for developing food safety standards and successfully dealt with intentional food safety issues . Risk communication was also enhanced. As examples, many ministries and departments jointly organized the Food Safety Awareness Week and Open Day. Several activities based on data generated from greatly improved foodborne disease surveillance since 2011 were also implemented. In 2013 alone, 120,000 supervision staff, more than 4,000 experts and scholars, and 35 million employees participated in activities. Hundreds of media issues nearly 20,000 news reports and over 300, 000 micro-blogging topics . These activities have also had an impact on public awareness and engagement in food safety practices. The overall food safety status has improved steadily, and may be reflected in the burden of diarrheal diseases.
Several cross-sectional surveys have been conducted in other countries to estimate the prevalence and distribution of AGI. In America [26–38], Europe [2, 39–47], Oceania[48, 49], Asia  and Africa .
The estimated incidence of AGI in China is comparable to similar retrospective studies conducted in Sweden and France, which reported 0.31 and 0.33 episodes of AGI each year, respectively, in spite of slightly different case definitions and recall periods. It is also similar to estimates of prospective follow-up studies in The Netherlands (0.28) (ranging from 0.42 to 1.66 episodes/person-year) . However, it is lower than observations in almost all the similar cross-sectional studies conducted in other countries or regions, no matter what interviewing methods they used, and higher than rate from England and Wales (0.19) . Nevertheless, comparisons between countries need to consider the varying case definitions, interviewing and sampling methods, and general differences in populations.
Our study used criterion for the identification of AGI recommended by an international collaboration focusing on burden of foodborne illnesses . However, most studies, unlike ours, used a telephone survey; some of them used random digit dialing techniques or modified version to selectparticipants [14, 15, 17, 55]. Other possible explanations for the differences in AGI incidence include cultural aspects and likelihood to answer questions openly. Chinese people are not used to talk about their health status in front of strangers, no matter if in public or in private, especially in big cities like Beijing Shanghai. Furthermore, our survey had a very high response rate. With the help of social workers in resident's committees or villager's committees, almost all the selected household participated, and we had a response rate of 97%. In contrast, the response rate in other countries was below 70%, but there is a possibility that people who have AGI recently are more inclined to respond. Other explanations for differences in reported incidence of diarrhea may include differences in risk factors such as food consumption and preparation habits, food contamination, or environmental factors.
Limitations and strengths
In this survey, the samples were weighted and standardized to make the survey sample representative. However, there are still some biases in the survey: the young and middle aged labor force in rural areas worked in the cities, while the young and middle aged labor force in cities work in the daytime; this may have resulted in more elderly people in the sample population, In addition, the selection of survey sites failed to give full consideration to China's western minority areas, such as Tibet and Xinjiang. At the same time, some groups of group-life accommodation (such as inpatients in hospitals, elderly people in nursing homes, students in long-term accommodation, prisoners in prisons and officers and soldiers in barracks) were not covered. This might have introduced selection bias. Furthermore, our study suffered from the limitations of all cross-sectional surveys. Its results reflect only real time data from the survey sample, and is difficult to make causal inferences.
Our estimates provide evidence on the incidence of AGI in the population. Due to underdiagnosis and underreporting, such data are not available from public health surveillance. However, it is crucial to demonstrate the true burden of diarrheal diseases in the population. This burden is not only of a health nature, but also social and economic. Fo example, our estimates demonstrated that at least 17% of respondants or caretakers with AGI loss a mean 3 days of work due to the illness. If extrapolated to the population, this equates to approximately 1.5 million working days loss due to AGI annually. Together with healthcare costs, this represents a substantial burden nationally. When combined with evidence on the contribution of different causative agents for the overall incidence of AGI, these estimates will form the basis to estimate the burden of foodborne diseases in China.