Risk Factors for Shiga Toxin-producing Escherichia Coli Infections in Preadolescent Schoolchildren in Buenos Aires, Argentina

Introduction: Shiga toxin-producing Escherichia coli (STEC) infections are the leading cause of hemolytic uremic syndrome (HUS). STEC is the most common cause of acute kidney disease, responsible for 20% of renal transplants in Argentina. Methodology: In 2007, an epidemiological survey was conducted among 883 students from the fifth and sixth years of elementary education in the public schools of San Martin City, Buenos Aires, Argentina. Degree of exposure to the known risk factors previously detected in the region as primary causes of STEC infections was evaluated. Risk factors assessed included consumption of hamburgers, poor personal hygiene, and exposure to various types of drinking and recreational water. The study was designed to evaluate exposure to risk factors for STEC infections among different socioeconomic groups. Results: Ninety-five percent of children surveyed reported consumption of hamburgers. Most of these hamburgers were precooked. Children of high and medium strata attended private swimming-pools, while children from the low stratum attended public pools. Only 30.2% of students washed their hands after going to the toilet and only 43.5% reported hand-washing before eating. Conclusions: Students demonstrated high levels of exposure to identified risk factors for STEC infections. Reduction of these risks will require cultural changes aimed at decreasing morbidity caused by food-borne infections. Institutional framework must provide the necessary resources to implement these changes and emphasize the importance of good personal hygiene. Health education must be implemented to increase food safety awareness of the consumers.


Introduction
Shiga toxing-producing Escherichia coli (STEC) is one of the most important groups of emerging foodborne pathogens of worldwide distribution and the principal cause of hemolytic uremic syndrome (HUS).This syndrome, characterized by acute renal failure, hemolytic anemia and thrombocytopenia, is usually secondary to acute infectious diarrhoea [1].Most cases of HUS are identified in children younger than 16 years old with a high incidence in children up to five years of age [2][3][4][5][6].In Argentina, HUS is endemic with over 450 sporadic cases recorded per year.HUS is directly responsible for 20% of pediatric renal transplants in Argentina, and represents a critical area for further research [7][8].Almost all diarrhoea-associated HUS cases in children are due to STEC infections belonging to O157:H7 or non O157:H7 serotypes [9].The normal habitat of STEC is the intestinal tract of ruminants and other animals; particularly sheep and cattle serve as reservoirs [10].The prevalence of STEC in ruminants is reported to be 66.6% in sheep, 56.1% in goats, and 21.1% in cattle [11].The main route of human exposure is fecal-oral, usually due to fecal contamination of foods.In humans, less than 100 bacteria per gram of food are necessary to establish STEC infection [12].The World Health Organization (WHO) has classified illness from STEC as a food-borne disease (FD) [13].
Transmission of disease can also occur through direct contact of humans with carrier animals, or through person-to-person transmission.Food is identified as the source of infection in approximately half of all reported cases [14].Previous studies have evaluated the importance of various risk factors for infection in humans.Contamination of ground meat can occur during processing or preparation and represents a potentially significant source of STEC.Bacteria are destroyed at 72°C, but it is necessary to ensure that this temperature penetrates the whole of the meat prior to consumption [15].Epidemiological studies have identified a number of useful measures to prevent possible human exposure to food-borne STEC infections.These include thorough cooking of meat (particularly ground meat), avoidance of crosscontamination during food preparation, access to safe drinking water, consumption of pasteurized milk and juices, as well as public education on the importance of good personal hygiene and cleanliness of eating utensils [16,17].These measures complement actions aimed at lowering the prevalence of STEC in cattle [18], observing good hygiene practices during slaughter in meat processing plants, and careful preparation of meat for distribution to consumers [19].
The relationship between exposure to STEC and associated morbidity from diarrhoea among different socioeconomic groups has been extensively demonstrated [20].Unsafe water, poor sanitation, and poor personal hygiene are three of the 10 identified risk factors for exposure to STEC.Worldwide, approximately 1.7 million deaths per year from infectious diarrhoea are attributable to these three factors [21].Recent epidemiological results highlight the importance of socioeconomic risk factors in rates of disease associated with E. coli O157 [22].Nevertheless, how socioeconomic level relates to these individual risk factors is unknown and warrants further evaluation [23].
The objective of the current study was to evaluate degree of exposure to STEC in the most vulnerable group: young schoolchildren.This goal was accomplished by assessing exposure to known risk factors for the development of HUS and their relation to the socioeconomic level of schoolchildren from urban areas in Buenos Aires.

Study group
In total, 883 questionnaire surveys targeted at children in the 5th and 6th years of primary school (ages 10 to 12 years) were sent by professional surveytakers.The children attended 14 of the 56 state schools in an urban area of Buenos Aires, San Martin district, during 2007.This district is representative of the urban ring of Buenos Aires.Ethical approval for the study was obtained from the Department of Promotion and Prevention of Public Health Services of San Martin district.
The study was stratified by socioeconomic level because of the known association between the prevalence of childhood diarrhoea and economic inequalities.Documentary information was obtained from the National Institute of Statistics and Census (INDEC), which defines the socioeconomic levels of the district of San Martin as high, medium-high, medium, medium-low and low [24] (Figure 1).
The 56 tuition-free public schools of the district were geo-referenced and included in the distribution maps of socioeconomic level through a system of Cartesian coordinate axes.
For this study, the five socioeconomic levels (INDEC) were partially combined to create three groups: high (high and medium-high), medium (medium) and low (medium-low and low).Students representing 25% of each socioeconomic level were selected by simple random sampling from each school.

Survey
Grade, school attended, and area were recorded for each student.Exposure to known multiple factors was assessed.The following factors were included in the assessment:

Statistical analysis
Statistical analysis data was processed by the program EpiInfo 2002 ver.3.2 (CDC-WHO).The statistical analysis of the variables under study, as classified by socioeconomic level, was by the Chi square independence test and the Z test to compare two proportions.We estimated odds ratios with 95% confidence intervals (CI) [25].

Study group characteristics
Ranking the schools according to socioeconomic distribution revealed that of the 56 public schools identified in the district, 12 (21%) belonged to the high level, 18 (32%) to the middle level, and 26 (46%) to the low level [24].Based on the previously described sampling method, information was collected for all children present on the day of the survey-takers' visit in three schools from the high socioeconomic level (110 respondents), five schools in the middle level (394 respondents), and six schools in the low level (379 respondents) (Table 1).
No significant differences were found among socioeconomic levels when comparing the distribution of schools sampled within the district (p = 0.963), the respondent groups (p = 0.071), or the two grades surveyed (p = 0.065).

Exposure to risk factors
Hamburger consumption: Of students surveyed, 94.6% (835/883) said they consumed hamburgers.Confidence intervals for types of hamburger consumed and the absolute and relative frequency of consumption are presented in Table 2.
Analysis of the data by socioeconomic status showed significant differences in the overall consumption of hamburgers, with the highest proportion in the high and middle levels (p = 0.001).There was no significant difference (p = 0.11) in the types of hamburgers consumed among the socioeconomic groups (Table 2).
Personal hygiene: Of the children surveyed, 815/883 (92.3%) washed their hands while 68/883 (7.7%) did not.Time of hand-washing is detailed in Table 6.In addition, 499/883 (56.5%) of respondents reported that they never washed their hands before eating.Children from low socioeconomic levels washed their hands more frequently than those from other socio-economic groups did (p = 0.000).Out of 379 students from the low level, 206 (54.0%) reported washing their hands regularly as compared to 31 of 110 (28.0%) and 147 of 394 (37.0%) from the high and medium-level groups, respectively.The odds ratio for washing hands before eating was 36.02 (CI 4.94-262.64).

Discussion
This study evaluated exposure of schoolchildren population to known risk factors for developing HUS [9].The study group was comprised of children up to 15 years old represent one of the populations known to be at high risk [6,15,17].
One risk factor strongly associated with HUS is meat consumption, particularly ground meat.A previous study using a case-control design showed an association between infection with STEC O157 and consumption of incompletely cooked hamburgers in the week prior to illness [26].This study showed that 94% of children surveyed consumed hamburgers from a variety of sources.Differences among socioeconomic strata were in the frequency of consumption and not in the method of preparation.
In the last two generations, the consumption of precooked food has become widespread due to convenience, affordability, palatability and strong advertisement.This has resulted in substantial changes in handling methods for ground meat that place consumers at higher risk of exposure to undercooked meat.Changes in food handling such as inadequate cooking can contribute to the survival of pathogens in the meat with associated development of food-borne diseases.In Argentina, where meat consumption is estimated to be 63 kg/person/year [27], consumption of fast food increases the risk of exposure to pathogens, including STEC.
In the present study, nearly 80% of children consumed commercially prepared precooked hamburgers that were sold at stores regulated by the agency for Hazard Analysis and Critical Control Point (HACCP) and evaluated by health authority activities.Large numbers of children (60%) from all socioeconomic groups also reported consumption of homemade hamburgers.Meat handling and preservation methods at home can also contribute to the risk of cross-contamination with an associated increase in the number of pathogenic organisms in the meat [9,28].Preparation methods of homemade hamburgers would naturally vary among different homes.Therefore, the role of the government in educating the general public in proper food handling is crucial to reducing the risk of meat contamination in the target population.The Argentine Food Code legislation bans the sale of previously ground beef in retail stores [29].However, the beneficial effects of this law depend on the efficiency of meat inspection by regulatory authorities [27].
The population studied showed minimal exposure to untreated drinking water.Only 6.3% consumed well water that could be of poor quality and represent potential risk.Consumption of other beverages consisted primarily of soft drinks that do not carry an associated risk for HUS.A smaller percentage of children reported consumption of juices.Consumption of commercially prepared non-pasteurized apple juice has been associated with cases of HUS in the United States [30].In Argentina, commercial juices are pasteurized and should not represent a potential source of contamination.
In Argentina, some cases of HUS are associated with contaminated recreational water [13,31].Children generally play for long periods of time in recreational waters and are more likely to intentionally or accidentally swallow water [32].The current study group showed high exposure to family-owned swimming pools.Although initial water quality may be safe, inadequate maintenance of pools represents a potential risk factor.Thus implementation of adequate chlorination protocols is necessary to reduce disease associated with contaminated pond water.Educating individuals regarding proper pool maintenance and its role in disease prevention is critical.In high and medium strata, this probable risk situation is reduced to the half in holyday, when children break to move to the sea.Moreover, fewer children in low strata use the      recreational open water and therefore depend on the sanitary quality of the pools.Public swimming pools depended on current required regulatory inspections by qualified authorities to ensure water quality.The positive impact of such regulatory measures on health and safety has been previously documented [33,34].The importance of hygiene measures, particularly hand-washing, in the prevention of disease, including acute infantile diarrhoea, has been previously documented [34].In Argentina, hand-washing has been identified as a protective factor in a multivariate analysis for STEC infection [9].Therefore, in the current study, personal hygiene habits were assessed based on frequency and time of hand-washing.In the present study, schoolchildren generally practiced hand-washing but many of them did not wash their hands prior to eating food.Since hand-washing before eating is critical to the prevention of food-borne diseases, those who fail to wash their hands before eating are at risk of exposure.
Hand-washing should be considered an important preventive measure and strongly encouraged.However, most of the schools visited did not have the necessary elements of hygiene, such as soap in the bathrooms.Behavioral changes in these groups are unlikely without adequate infrastructure and appropriate institutional support.
All the identified risks are preventable through appropriate health and consumer education of children and their families.Consumer demand for a safe food supply must be addressed with an institutional framework maintained under governmental supervision.The government of Argentina has sporadically undertaken such prevention campaigns, but these measures have not demonstrated long-term impact on the incidence of disease.Short-term positive results, such as the decreased incidence of infant diarrhoea during the campaign to prevent cholera disease (Morales R, unpublished results, 2006), are occasionally reported.Unfortunately, none of these campaigns have resulted in lasting behavioral changes.
Given the endemic nature of HUS in Argentina, isolated and local campaigns are unlikely to reduce incidence of the disease.Implementation and sustainment of a national program for prevention and control will provide the greatest benefit to the general population.Health education programs should be directed at increasing awareness of food safety and adequate personal hygiene.The educational concepts that reduce the incidence of diarrhoea in developing countries are outlined in strategies promoted by the Pan-American Health Organization (PAHO) and the World Health Organization (WHO) [35].
Encouraging individuals to take positive measures for improving food safety and hygiene measures is highly desirable and critical in the prevention of foodborne diseases.Reduction of risk also depends on cultural dietary changes that will help decrease the burden of infectious disease in developing countries [21].
More than 200 food-borne diseases with high morbidity and mortality rates in children are currently recognized.It is imperative to develop coordinated and integrated actions to decrease illness associated with these diseases.These illnesses also contribute to increased costs within the health-care system, have a strong social impact, and create distrust in the community regarding safety in the food chain.In particular, treatment costs associated with HUS significantly exceed the investment in prevention [8].We contend that focusing on health risks is key to prevention of disease.Correctly implemented, highquality health programs that do not discriminate based on cultural and socioeconomic conditions will improve overall child health [36].Such risk reductions and sustainable development may also help in reducing inequalities in the society.These changes are not only the responsibility of governments but the concern of all individuals within the population [21].The current study demonstrates that Argentinian children are at high risk of exposure to known risk factors for the development of HUS.These findings confirm the need for improved educational programs targeted at improving personal hygiene, establishment of correct meat handling and cooking techniques, and maintenance of safe recreational water.Further studies are needed to identify appropriate educational methods targeted at different cultural and socioeconomic groups in the prevention food-borne disease.

Figure 1 .
Figure 1.Distribution of schools surveyed in various socioeconomic levels of San Martin, Buenos Aires, Argentina

Table 1 .
Distribution of schools and students surveyed according to socioeconomic level

Table 2 .
Hamburger consumption in students aged 10-12 years from different socioeconomic levels NR: No response a Frequencies considered on multiple responses b Considered as buying raw c Bought cooked for consumption

Table 3 .
Consumption of water and source in different socioeconomic strata a No significant difference between socioeconomic groups.

Table 4 .
Combinations of drink consumption in different socioeconomic strata a : Situations not covered by the previous options with low frequency (≤ 3).NR: No response

Table 5 .
Recreational water used according to socioeconomic level

Table 6 .
Hand-washing patterns and frequency in schoolchildren related to socioeconomic level a Corresponds to multiple responses from the total survey b Situations not covered by the previous options and low frequency (≤ 3).