The correlation between family food handling behaviors and foodborne acute gastroenteritis: a community-oriented, population-based survey in Anhui, China

Study populations and sampling design

A multistage stratified cluster sampling method was applied to select subjects in Anhui Province in 2015. Survey areas were selected based on environmental conditions, population size, economic level, culture, and lifestyle. In the first stage, 6 cities (Hefei, Fuyang, Wuhu, Suzhou, Xuancheng, and Chizhou) were selected randomly from a total of 17 cities (Hefei, Wuhu, Ma-anshan, An-qing, Chaohu, Bengbu, Huaibei, Liu-an, Tongling, Huainan, Xuancheng, Chuzhou, Huangshan, Fuyang, Haozhou, Suzhou, and Chizhou) in Anhui Province. One district and one county from each of the six cities was then randomly selected. In the second stage, two streets and two towns were selected randomly from each district and each county, respectively. In the third stage, one community and one administrative village was selected from each street and town, respectively, resulting in a total of 24 sites (12 communities and 12 administrative villages). In the fourth stage, 1600 households were randomly sampled from the 24 sites. Of the 1600 households selected, 84 declined participation, resulting in 1516 (94.8%) households that completed interviews.

Data collection

Approximately 100 to 150 different households were interviewed each month. Over a 12-month period, 1516 households were surveyed. In each household, the individual celebrating his/her birthday next was selected to participate in the survey. For those younger than 12 years old, the questions were answered by parents or guardians. Interviews were conducted by trained investigators using a questionnaire designed specifically for this study. Data gathered from the questionnaire included: (1)general demographic characteristics; (2) occurrence of acute gastroenteritis over the past 4 weeks; (3) behavior of food purchasing, handling, storage, and consumption. All subjects were asked whether they had experienced diarrhea or vomiting in the four weeks prior to the survey. In cases where diarrhea or vomiting were reported, details with respect to possible causes, the suspected food(s), and the origin of suspected foods were recorded. Information related to the behavior of food purchasing, handling, storage, and consumption were obtained from subjects ≥18 years old, who were the primary food handlers in the family.

Case definition

Acute gastroenteritis was defined as diarrhea and/or vomiting for a period of at least 24 h within the 4 weeks preceding the interview. Diarrhea was defined as at least three loose-stool bowel movements within a 24-h period. Vomiting was recorded only if it was coupled with at least one other symptom (abdominal pain, nausea, or fever). Cases of diarrhea or vomiting caused by irritable bowel syndrome, colitis, colon diverticulitis, pregnancy, chemotherapy/radiotherapy, medication, food allergy, Crohn’s disease, and excess alcohol were excluded from the data set. Respondents who experienced symptoms of foodborne acute gastroenteritis were referred to as the “case group”. Respondents who did not experience symptoms of foodborne acute gastroenteritis were referred to as the “control group”.

Statistical analysis

The database was constructed using EpiData 3.1 software (EpiData, Odense, Denmark), and statistical analyses were performed using SPSS Software V.18.0 (SPSS Inc., Chicago, IL, USA). All qualitative data were expressed as frequency (percentage). The incidence of foodborne acute gastroenteritis was calculated by dividing the number of subjects reporting foodborne acute gastroenteritis in the four weeks prior to the survey by the total number of subjects. The proportion of those exposed to inappropriate food handling behavior was calculated as the number of subjects who engage regularly in inappropriate food handling behavior relative to the total number of subjects. A chi-square test was used to compare the following: (1) the incidence of foodborne acute gastroenteritis associated with different socio-demographic characteristics; (2) the proportion of those exposed to inappropriate food handling separated by gender; and (3) the proportion of those exposed to inappropriate food handling among cases of foodborne acute gastroenteritis and among those who did not report illness. Unconditional logistic regression was used to explore the association between food handling and foodborne acute gastroenteritis and to compute odds ratios and 95% confidence intervals (95% CI). Linear regression was used to diagnose the collinearity of independent variables through tolerance value and the variance inflation factor (VIF). A Tolerance value < 0.1 or a VIF- value > 10 was considered indicative of collinearity. A P-value < 0.05 was considered significant. Two-tailed tests were conducted.

Demographic characteristics of respondents

Males comprised 56.5% of the subjects. Over 60% of subjects were between 18 years and 50 years old. The proportions of subjects aged ≤17 years and ≥ 50 years were 5.5 and 10.1%, respectively. Most of the subjects (79.0%) had a secondary or tertiary education. Almost 57.7% of respondents lived in rural areas and 42.3% in urban areas. Approximately 49.5% of the respondents reported that their annual household income was within the range of 30,000–89,999 CNY (4387–13,161 USD), while 33.7% reported an annual household income greater than 90,000 CNY (13,161 USD). The remaining 17.0% of the subjects reported an annual household income of less than 30,000 CNY (4387 USD). Approximately 19% of the respondents were peasant migrant workers who commuted to the city from the countryside to work in the construction industry for six months or more, while approximately 13.7% of the respondents were peasant farm workers in rural areas. The percentage of respondents working as service staff, medical staff, and laborers (e.g. factory workers) were 10.1, 15.4, and 11.3%, respectively. Finally, teachers accounted for 8.9% of the subjects, while students and children accounted for 4.7% of the subjects.

Incidence and distribution of foodborne acute gastroenteritis

Suspected food and food origin

The status of food handling in the family environment

As can be seen in Table 3, from a total of 1440, the proportion of respondents who answered “no” to the items: “often thoroughly heat milk”, “often thoroughly heat leftovers stored in the refrigerator”, “often thoroughly heat cooked food purchased from outside”, “often thoroughly heat green beans”, and who answered “yes” to the item “often eat raw seafood” were 75.6, 41.6, 71.3, 17.3, and 15.0%, respectively. The proportion of respondents who answered “no” to “often thoroughly boil disinfection kitchen utensils” and “often store leftovers in the refrigerator” were 72.2 and 32.5%, respectively. In addition, approximately 29.0% of respondents often purchase stale raw vegetables, meat, and other ingredients. Stale, raw vegetables were defined as those that had been maintained in the market for a prolonged period of time before being purchased, and the edges of the leaves had begun to turn yellow and rot. Approximately 26.9% of respondents reported that they often buy expired packaged food. Almost 23.6% of respondents often store raw meat and cooked meat in the same container, while 19.8% of respondents answered “no” to “often thoroughly clean raw vegetables or meat purchased from the market before cooking”.

Male and female respondents differed significantly in their answers to the following food handling behaviors: “often thoroughly heat cooked food purchased from outside” (X² = 26.518, P < 0.001), “often thoroughly heat leftovers stored in the refrigerator” (X² = 173.063, P < 0.001), “often thoroughly cook green beans” (X² = 38.091, P < 0.001), “often store leftovers in the refrigerator” (X² = 18.820, P < 0.001), “often thoroughly clean raw vegetables or meat purchased from outside” (X² = 4.950, P < 0.05), “often store raw meat and cooked meat in the same container” (X² = 15.288, P < 0.001), and “often purchase stale raw vegetables, meat and other ingredients” (X² = 17.222, P < 0.001).

Comparison of food handling behaviors between the case group and the control group

Of the 1440 respondents, 153 reported experiencing foodborne acute gastroenteritis in the previous four weeks and are referred to as the “case group”. The remaining 1287 respondents are referred to as the “control group”. As seen in Table 4, the proportions of those that answered no to “often thoroughly heat milk” (X² = 14.768, P < 0.001), “often store leftovers in the refrigerator” (X² = 5.875, P < 0.05), “often thoroughly heat leftovers stored in the refrigerator” (X² = 29.597, P < 0.001), and “often thoroughly cook green beans” (X² = 21.581, P < 0.001) and yes to “often eat raw seafood” (X² = 55.298, P < 0.001) and “often store raw meat and cooked meat in the same container” (X² = 53.764, P < 0.001), were significantly higher in the case group compared to the control group.

Multivariable logistic regression analysis

A multivariable logistic regression was conducted to analyze the association between food handling behaviors and foodborne acute gastroenteritis. Gender, age, occupation, residence, education, economic status, eating out, and food handling behaviors were independent variables, while foodborne acute gastroenteritis was the dependent variable. As shown in Table 5, after eliminating the effects of demographic characteristics on foodborne acute gastroenteritis, the behaviors entered into the regression were: “often thoroughly heat milk before eating” (P < 0.001, OR = 4.665, 95%CI = 2.526–8.617), “often thoroughly heat leftovers stored in the refrigerator before eating” (P < 0.001, OR = 3.416, 95%CI = 2.139–5.454), “often store leftovers in the refrigerator” (P < 0.05, OR = 1.775, 95%CI = 1.169–2.696), “often store raw and cooked meat in the same container” (P < 0.001, OR = 4.291, 95%CI = 2.722–6.765), “often thoroughly cook green beans before eating” (P < 0.001, OR = 2.859, 95%CI = 1.798–4.545), and “often eat raw seafood” (P < 0.001, OR = 3.250, 95%CI = 2.136–4.946). There were no obvious collinear issues between independent variables.

Incidence and distribution of foodborne acute gastroenteritis

To our knowledge, this is the first survey of foodborne acute gastroenteritis in a community population in Anhui Province, China. The monthly incidence of foodborne acute gastroenteritis recorded in the current study was 10.9%, which was higher than that reported in Italy (8.9%) [7], France (2.6%) [4], Japan (3.5%), Trinidad and Tobago (5.1%), and British Columbia, Canada (9.2%). In addition, the monthly incidence of foodborne acute gastroenteritis in the current study was greater than that previously reported in Jiangsu (4.4%) and Gansu (8.5%) [33–37] Provinces, China. Differences in illness prevalence may be related to exposure to risk factors, research design, and/or survey method.

Consistent with previous studies from China, the incidence of foodborne acute gastroenteritis differed in gender, age, education, economic level, and residence. In the current study, the incidence of foodborne acute gastroenteritis in women (15.3%) was double that of men (7.5%), subjects aged ≥50 years had the highest incidence of foodborne acute gastroenteritis (23.5%) and higher education (bachelor’s degree or above) was associated with the lowest incidence (6.0%). The incidence of foodborne acute gastroenteritis was lowest (7.5%) among respondents with a high income (total annual income greater than 90,000 CNY) and the incidence of foodborne acute gastroenteritis in rural residents (13.3%) was significantly higher than those living in urban areas (7.6%).

Meat and meat products, milk and dairy products, aquatic products, eggs and egg products, beans and bean products, and fruits and vegetable were the foods most commonly suspected of causing illness. Meat and meat products were the most commonly suspected foods (58.2%). The home (40.6%), street stalls (24.8%), and fast food restaurants (15.2%) were the most frequently reported locations where suspected foods were consumed. These results indicate that food hygiene inspection and food safety management of street stalls and fast-food restaurants must be increased and that information on proper food handling must be emphasized in the family environment.

Food handling in the family environment

Thorough heating of food, storing leftovers in the refrigerator, and the disinfection of kitchen utensils are critically important behaviors in the prevention of the growth of microorganisms [38, 39]. However, most people are not educated in the proper handling of food. Inappropriate food handling occurs in food production, food processing, storage, and consumption. In the current study, most respondents reported that they infrequently thoroughly heated milk (75.6%) and cooked food purchased from outside the home (71.2%). In addition, as many as 41.6% of respondents reported storing leftovers in the refrigerator infrequently.

The storage and processing of raw and cooked foods together is often the cause of cross contamination. For example, raw meat and cooked meat are often stored in the same container. In addition, foods containing toxic chemicals such as green beans and seafood are the cause of food poisoning outbreaks in China. The consumption of undercooked or raw foods that contain toxic chemicals, such as green beans and raw seafood, has led to many outbreaks of food poisoning [40, 41]. In the current study, 23.6% of respondents reported that they often store raw meat and cooked meat in the same container. Most respondents reported that they often thoroughly cook green beans (82.7%) and do not usually eat raw seafood (85.0%).

These results suggest that inappropriate food handling is prevalent among consumers in the family environment. Previous research has shown that a lack of food safety knowledge is detrimental to the behaviors of food handling [42–44]. These results indicate the importance of increasing education of the typical consumer on proper food handling procedures to produced long-term changes in healthier behaviors.

Food handling behaviors and foodborne acute gastroenteritis

The present study found differences in food handling behavior between respondents that reported foodborne acute gastroenteritis (case group) and those that did not (control group). The behaviors of infrequently thoroughly heating milk before eating, infrequently storing leftovers in the refrigerator, infrequently thoroughly heating leftovers stored in the refrigerator, often eating raw seafood, often storing raw meat and cooked meat in the same container, and infrequently thoroughly cooking green beans, were 1.2, 1.3, 1.6, 2.8, 2.5, and 2.0 times greater in the case group compared to the control group.

The risk of foodborne acute gastroenteritis was 4.7 times greater in those that responded “no” to “often thoroughly heat milk before eating”. Milk that is not thoroughly heated prior to drinking, can increase the risk of foodborne acute gastroenteritis. Drinking cold milk can stimulate the gastrointestinal tract and cause a negative gastrointestinal reaction [45]. In addition, reports of food quality monitoring from several cities indicate that microorganisms such as Staphylococcus aurous have been detected in milk and dairy products sold in stores [46–48]. As such, milk that is not heated prior to drinking may still contain high levels of microorganisms, leading to illness.

Respondents reporting that they do not often store leftovers in the refrigerator or infrequently thoroughly heat leftovers before eating were at a 1.8 and 3.3 times greater risk for foodborne gastroenteritis. Containers used for storage, storage time, and storage temperature are associated with microbial contamination of leftovers. Reusing containers that were used for cooking to store leftovers can cause cross contamination. Rapid bacterial growth occurs in leftovers that are maintained between 4 °C (40 °F) and 60 °C (140 °F) [49], and therefore leftovers must be stored in the refrigerator within 2 h. In addition, leftovers should be heated to at least 74 °C (165 °F) prior to eating in order to reduce the risk of any prior cross contamination and/or inappropriate storage temperature [49].

Infrequently thoroughly cooking green beans before eating was associated with a 2.9 times higher risk of experiencing foodborne acute gastroenteritis in the current study. Green beans can be classified into runner beans (Phaseolus coccineus), yardlong beans (Vigna unguiculata), hyacinth beans (Purple Haricot) and snap beans (P. var. chinensis Hort.). As an important crop worldwide, green beans are low in fat, high in protein, and rich in minerals and multiplex vitamins. However, the consumption of green beans is often the cause of foodborne disease outbreaks in China. Green beans contain two main substances that cause food poisoning: saponin and phytohemagglutinin. Saponin and phytohemagglutinin are relatively resistant to heat and can are only destroyed at temperatures above 100 °C (212 °F) for ten minutes [50]. Short cooking time and low cooking temperatures increase the risk of saponin and phytohemagglutinin poisoning.

Respondents reporting that they often eat raw seafood were 3.3 times more likely to experience foodborne acute gastroenteritis. One possible reason why eating raw seafood increases the risk of foodborne diseases is that seafood has been contaminated by Vibrio parahaemolyticus before consumption. In China, the detection rate of Vibrio parahaemolyticus in seafood is 25–40% [51]. The bacteria Vibrio parahaemolyticus is a major pathogen in seafood that can cause foodborne diseases outbreaks. For example, of the 2795 foodborne disease outbreaks between 2003 and 2008, 14% were caused by Vibrio parahaemolyticus. Moreover, approximately 28% of reported outbreaks caused by Vibrio parahaemolyticus were due by aquatic products [52]. Another possible cause that eating raw seafood increases the risk of foodborne diseases is that seafood eaten contains toxins, such as ciguatoxin, histamine, and saxitoxins which can also lead to foodborne diseases [53].

Food handling in the family environment is a complex behavior. Mishandling of food can occur at each step of the food chain process, from purchasing to cleaning, cooking, consuming, and storage. Consumers often do not follow proper food handling practices. Based on our study results, the following food handling behaviors are suggested. First, foods that are not immediately eaten (such as raw seafood, green beans, milk, and leftovers), must be thoroughly heated prior to eating. Second, to reduce cross contamination, containers for cooking food and storing surplus food must be kept separately. Third, leftover food must be stored in the refrigerator in a timely manner.

Limitations

While the present study provides valuable data on the risk factors associated with foodborne acute gastroenteritis, there are limitation that must be considered. Recall bias may have affected the study results. The data collected relied on respondents accurately relaying past experiences, prior to the start of the study which may result in recall errors. In particular, data on suspected foods and origin of food may be affected. In addition, respiratory diseases can be associated with symptoms similar to those in acute gastroenteritis and therefore, some cases of acute gastroenteritis may have been incorrectly diagnosed. Some respondents that were included in the study therefore, may have had respiratory disease. There was no exclusion criteria for the “control group”. And some respondents could not be excluded from the “control group”, which may have affected the study results. The “control group” was not paired with the “case group” following some criteria (e.g. gender). The distribution of some variables (e.g. gender) may be disequilibrium between the “case group” and the “control group”.