Content Information
National and statewide foodborne illness trends
Foodborne diseases have been constantly changing over the years. Diseases such as typhoid fever, tuberculosis and cholera that were common over a century ago have been replaced with diseases such as E. coli, Salmonella, and Noroviruses. This change can be attributed to evolving new microbes, the changing environment, changing food-production practices and consumption habits, as well as advances in laboratory technology, to name a few.
More than 250 diseases have been transmitted through food, causing more than 76 million cases of foodborne disease each year in the United States, including approximately 325, 000 hospitalizations and 5,000 deaths.
Many diseases that can be transmitted through food are reportable diseases in Iowa, required by Iowa Administrative Code [641] Chapter 1 (refer to chapter 3 Understanding the Law for this code). The trends of five of the most common reportable diseases both nationally and in Iowa – Shigella, Salmonella, Campylobacter, E. coli 0157:H7, and Hepatitis A are discussed in this chapter. The majority of cases in Iowa are individual sporadic cases and have not been linked to other cases or found to be part of a foodborne outbreak.
In this chapter, a graph for each of these five diseases displays the number of cases reported in Iowa by month versus the threshold level for that month. Reported cases include all sporadic and outbreak-related cases. The threshold level is the highest number of sporadic cases that would be expected to be reported each month to the Iowa Department of Public Health (IDPH). The threshold level was calculated using two standard deviations above the five year moving average of cases for each month. To detect outbreaks in these trends, the five year moving average was calculated using only the sporadic cases and excluding all cases that were outbreak-related. If outbreak cases were included, the standard deviations would become quite wide and anything out of the ordinary would not be easily detected.
This type of graph shows what IDPH would normally see versus what they actually did see. In most cases, if the outbreaks are large enough, the disease can be easily detected on the graph as the point at which the number of cases reported (red line) exceeds the threshold (yellow area). Not all outbreaks are big enough to exceed the threshold, and are detected via other methods.
Trends in reported foodborne outbreaks in Iowa are also discussed in this chapter. Some involve an enteric disease that as an isolated case is not normally reportable in Iowa. However, an outbreak of any disease is reportable (e.g. norovirus, Staphylococcus aureus).
Campylobacter Trends
Campylobacter is recognized as the most common cause of bacterial gastrointestinal illness in humans. Incidence from 2005 Food Net data is about 12.7 Campylobacter cases per 100,000 population diagnosed in the United States. However, experts believe the actual number of cases is 38 times what is reported. Nationally, 80 percent of reported cases are foodborne. The incidence of Campylobacter reported in Iowans in 2005 was 18.1 cases per 100,000 persons, which was similar to the incidence in 2004 of 18.8 cases per 100,000 persons. These two rates are the highest incidence in Iowa in the past 15 years, refer to Figure 1. Reasons for the increase are unknown. Campylobacter outbreaks were reported in 2004 and 2005 which may contribute to the higher incidence rates.
Campylobacter infection rates in Iowa in 2005 were higher in males (21.5 cases/100,000 persons) than in females (14.7 cases/100,000 persons). Overall, males under the age of ten had the highest incidence rates in Iowa at 37.7 cases per 100,000 persons. Females over the age of 80 had the lowest at 5.3 cases per 100,000 persons. Figure 2 shows age and gender specific incidence rates.
Over the past several years, Iowa has seen a consistent seasonal pattern in Campylobacter cases. The incidence begins to increase in the spring, peaks in summer and decreases in the fall, with the lowest number reported in December and January.
Figure 3 below demonstrates this pattern and shows the reported cases by month versus the threshold level. Both in 2004 and 2005, a Campylobacter outbreak was reported in Iowa due to the consumption of raw milk contaminated with Campylobacter jejuni. The only other Campylobacter foodborne outbreak reported to IDPH was an outbreak associated with a picnic in June of 2001. The outbreak in 2005 caused the number of cases reported in Iowa to exceed the threshold level but the two outbreaks reported in years prior to 2005 did not exceed the threshold level. The threshold level was exceeded in several instances in 1996-1998 but no outbreaks were reported or identified.
Escherichia coli O157:H7 Trends
Preliminary FoodNet data from 2005 shows the national incidence of E. coli O157:H7 infections around 1.06 per 100,000 persons. Eighty-five percent of E. coli infections are foodborne-related. In 2005, the incidence of E. coli O157:H7 in Iowa was above the national incidence at 3.4 cases per 100,000 persons. (Figure 4)
As with Campylobacter, Iowa sees a consistent seasonal pattern of E. coli O157:H7 cases (Figure 5). The majority of cases occur May through September with the peak number of cases around July of each year. Figure 5 also shows the reported E. coli O157:H7 cases in Iowa by month versus the threshold level.
In the summer of 1996, an unusual number of E. coli O157:H7 cases were reported in eastern Iowa but no common source was found between them. The increase may have been due to improved testing. In September of 1997, over 20 students from a school in northern Iowa became ill but no exposure was identified and it was unknown if there was a link to the food served at the school cafeteria.
The increase in cases in the summer of 2000 was due to a large outbreak associated with extreme cross-contamination at a restaurant. Though two outbreaks of E. coli O157:H7 were reported in the summer of 2002, the number of persons affected was small and the threshold level was not exceeded.
In the early summer of 2004, several Iowa E. coli cases matching a five-cluster multi-state E. coli O157:H7 outbreak were identified. The pulse-field gel electrophoresis (PFGE) patterns of the five clusters differed only by a single band. A case-control study of one of the clusters showed a strong association between consumption of ground beef in the home and the illness.
The incidence rates of E. coli O157:H7 in Iowa in 2005 are higher in males (3.7 cases/100,000 persons) than in females (2.9 cases/100,000 persons). Overall, males under the age of 10 have the highest rates at 14.2 cases per 100,000 persons. Females in this same age group have the second highest incidence rates at 8.6 cases per 100,000 persons. Figure 6 shows age and gender specific E. coli O157:H7 incidence rates.
Hepatitis A Trends
In 2005, the incidence of hepatitis A in the United States based on provisional MMWR 2005 data was 1.4 cases per 100,000 people. Five percent of hepatitis infections are known to be foodborne. However, the origin of 50% of hepatitis A infections is unknown, which does not exclude foodborne illness as a possibility. Most hepatitis A occurs in community-wide outbreaks, during which infection is transmitted from person to person in households and extended families.
In the past the incidence of hepatitis A in the United States has varied in a cyclical pattern, with large increases approximately every 10 years, followed by decreases to less than the previous baseline incidence. IDPH has been documenting reported cases of hepatitis A since 1961, when a record-high 1,986 cases were reported. Rates of hepatitis A in Iowa since that year have run much lower, indicating a low incidence of the virus. This may be due to mandatory vaccination of high risk groups. The number of cases has ranged from a low of 40 in 2003 to a recent high of 490 cases in 1997. The incidence of hepatitis A in Iowa in 2005 was lower than the national incidence at 0.7 cases per 100,000 persons. (Figure 7)
Figure 8 shows the reported hepatitis A cases versus the threshold level for that month. During the years of 1995 through 1998, Iowa had ongoing outbreaks/clusters of hepatitis A, with significantly higher numbers in Polk and Woodbury counties. During this outbreak those at highest risk were methamphetamine users, but other routes of transmission included transmission from infected food handlers. Outbreaks in the western part of the state peaked in February of 1996 but gained ground in central Iowa through 1997 and 1998. One documented outbreak of Hepatitis A occurred at an elementary school in central Iowa. By the end of this outbreak, 44 students, two school employees, and three secondary cases were identified.
At the end of the 1990s, hepatitis A vaccine was more widely used and the number of cases reached historic lows. It is estimated that one-third of Americans have evidence of past infection (immunity).
The hepatitis A incidence rates were similar for females and males in 2005, at 0.66 and 0.82 cases per 100,000 persons, respectively. Males 60-69 years of age had the highest incidence at 2.56 cases per 100,000 persons in Iowa. Females 0-29 years of age and males age 50-59 and over the age of 70 had the lowest rates of hepatitis A in Iowa in 2005. (Figure 9)
Salmonella Trends
Preliminary 2005 FoodNet data shows the national incidence of Salmonella to be 14.6 cases per 100,000 persons. Ninety-five percent of all cases are reported to be foodborne-related. In Iowa, the incidence was 13.9 cases per 100,000 persons, slightly lower than the national incidence. (Figure 10)
Females and males in Iowa in 2005 had a very similar overall Salmonella incidence with 13.9 cases per 100,000 persons versus males at 13.4 cases per 100,000 persons. Females and males under the age of ten had the highest incidence of 28.1 and 38.8 cases per 100,000 persons respectively. Males over the age of 80 also had a high incidence of 20.2 cases per 100,000 persons. (Figure 11)
Over the last nine years, Salmonella has been seasonal, though the pattern has not been as well defined as for Campylobacter and E. coli 0157:H7 trends. Figure 12 shows reported Salmonella cases versus the threshold level. Major outbreaks are easily identified on this graph as the areas where the number of cases reported exceeded the threshold level.
In 1996, the threshold was exceeded when a large outbreak occurred after a community turkey dinner. Several hundred persons developed illness and over 20 persons required hospitalization. Clinical and food samples grew identical strains of Salmonella Thompson along with the diarrheal toxin producing Bacillus cereus. From the amount of Salmonella grown from the food samples, it is estimated that the number of bacteria, ingested per person, may have been as high as several million! The epidemiologic and laboratory data pointed to the mishandling of turkey and broth, which allowed bacterial growth and toxin production as causative factors.
The next point on the graph where reported cases exceeded the threshold was in May 1998, when a national Salmonella Agona outbreak occurred due to contaminated cereal. Two to three million pounds of plain toasted oat breakfast cereals produced under 39 brand names were voluntarily recalled. An increase in cases in the summer of 2002 was due to two large outbreaks, one foodborne and one that began as foodborne then evolved into a community-spread outbreak.
In the spring of 2002, the threshold was slightly exceeded. At this time, seven Iowans became ill with Salmonella Enteritidis after attending a software users conference at a hotel in Texas.
The obvious spike in August and September 2002 can be attributed to two known outbreaks. The first was a Salmonella Berta outbreak linked to a restaurant. Thirty-three stool specimens were collected. Of those, 32 were identical matches to each other by pulse-field gel electrophoresis testing. Evidence indicted that mashed potatoes and gravy or a combination of food items in a buffet may have been contaminated, held and reused days later in the same buffet. From August to September of 2002, a Salmonella Newport outbreak was also occurring. Early in this outbreak, 17 cases were linked to eating at a fast-food restaurant where six ill food handlers were identified. Many other cases occurred due to secondary or household spread.
More recently, in the spring of 2004, a Salmonella Typhimurium variant Copenhagen outbreak occurred in people eating at a wedding reception. Chicken breasts were the implicated food item through both the epidemiological and environmental investigations. In the summer of 2004, another Salmonella Typhimurium variant Copenhagen outbreak occurred in a child-care center but it was not foodborne-related.
Shigella Trends
Preliminary 2005 FoodNet data shows the national incidence of Shigella to be 4.7 cases per 100,000 persons. Experts believe the actual number is 20 times greater than what is reported. Since 1995, the United States has seen a decreasing incidence of Shigella. Twenty percent of national cases are foodborne. The 2005 Iowa incidence of Shigella was 3.5 cases per 100,000 persons. (Figure 13)
Characteristically, Shigella causes large, periodic outbreaks nationally and in Iowa. (Figure 14) In 2000, Iowa saw an increase in cases, mainly due to person-to-person spread. In the summer of 2001, 69 cases occurred as the result of an inadequately disinfected wading pool. In the late winter/early spring of 2004, a person-to-person outbreak occurred in a child-care center. Two more person-to-person outbreaks occurred in 2005, one occurred within a daycare facility and the other could be attributed to person-to-person spread. Only two foodborne outbreaks due to Shigella have been reported in Iowa – one in 1995 and one in 1984.
Overall, Iowa Shigella incidence is almost identical in females (2.99 cases per 100,000 persons) and in males (2.81 cases per 100,000). Children under the age of ten have the highest Shigella rates in Iowa. Females under the age of 10 have the highest incidence, at 12.05. (Figure 15)
Trends in reported foodborne outbreaks in Iowa
From 1990 to 2005, 140 foodborne outbreaks were reported to the IDPH, affecting over 6,000 persons. Figure 16 shows the number of outbreaks reported by year since 1990. In all of them, Center for Acute Disease Epidemiology staff participated to some extent in the outbreak investigation, whether to provide a simple telephone consultation or as the primary investigator. Day-to-day foodborne illness complaints received by counties and the Iowa Department of Inspection and Appeals are not included in these numbers. The increase in reported outbreaks over the last several years may be due to several factors. Staffing and education of the public health workforce has also increased in response to the events of September 11, 2001, resulting in more capacity to find and investigate food-related incidents.
Over the past 16 years, 54.3 percent of foodborne outbreaks reported to the IDPH were either suspected or confirmed to be viral (refer to Figure 17). Bacterial etiology was confirmed or suspected in 33.6 percent of the outbreaks. In almost 10 percent, a specific etiology was never identified. Less than two percent of outbreaks reported to IDPH, were from chemical and parasitic etiologies. One outbreak had two etiologies (bacterial and viral).
Norovirus has caused most of the outbreaks in Iowa over the past several years, but many of them may not have been foodborne outbreaks. The percentage truly due to person-to-person spread versus point source food contamination with this virus is unknown. In many norovirus outbreaks in Iowa, it has been impossible to implicate a specific food causing the outbreak. Due to its epidemiology – low infectious dose, the ability to be spread person-to-person via fecal-oral or aerosol formation after projectile vomiting, and its ability to survive on environmental surfaces for weeks – it is likely that the majority of outbreaks resulted from person-to-person spread and were not foodborne.
In most instances, the spread was initiated by those being served, not the food preparers! In two norovirus outbreaks, for example – one at a wedding and another at a restaurant – one of the guests actually vomited at the site of the outbreak. It is suspected that some of the vomitus aerosolized and infected nearby guests, causing some of the resulting illness, while others may have gotten ill by touching items touched by the ill person with inadequately washed hands.
Of 140 outbreaks reported from 1990-2005, a specific etiologic agent was confirmed in 77 outbreaks. Figure 18 identifies the specific agent. Almost 45 percent of the outbreaks were confirmed to be norovirus or viral in nature. A quarter of them were Salmonella. Other etiologic agents were Staphylococcus food intoxication (7.7%), Bacillus spp. (7.7%), Clostridium perfringens (5.1%), E. coli 0157:H7 (3.8%), Campylobacter jejuni (3.8%), and ‘other’ (Sodium nitrite contamination 1.3%; Scrombroid/Histamine poisoning 1.3%; Trichnella spiralis 1.3%).
The locations of food preparation and food consumption were evaluated for all outbreaks reported to the IDPH from 1990-2005. In these two analyses, the location definitions are as follows.
Food establishments
- Home: kitchens in a private home.
- Institutions: hospitals, long-term care facilities, or nursing homes.
- Banquet Halls: community buildings, church halls, or hotel banquet rooms.
- Other: summer camps, food processing plants, buses, or airplanes.
- School: school or college cafeterias or food services.
- Workplace: workplace cafeterias or kitchens.
- Outdoors: parks (for food consumption only).
- Unknown: locations were not recorded.
Where the food is prepared often differs from where it is consumed. Figure 19 shows the location of food preparation in outbreaks from 1990-2005. In 55.7 percent of outbreaks reported during these years, the food was prepared in a food establishment. Seventeen percent of outbreaks were attributed to food prepared in a private home. Fewer than ten percent of outbreaks were attributed to food prepared at schools, institutions, banquet halls, or workplaces.
For the 140 foodborne outbreaks reported from 1990-2005, almost 56% of the food was prepared in food establishments, but only 34% of the food was actually consumed there. In several outbreaks, food contamination or errors with the food handling process often occurred once the food left the facilities after preparation. Figure 20 shows the locations where the food was actually consumed. In 18.9 percent of outbreaks, food was consumed at a banquet hall, 13.6 percent at a home, 10 percent at a school, and 7.9 percent at a workplace.
While most outbreaks are associated with food establishments, most cases of enteric diseases are sporadic (not part of an outbreak) and are usually caused by mishandling of food in the home.
Finding the specific culprit of an outbreak is often challenging. Often by the time the outbreak is reported any food or beverage that was left over from the event has already been discarded. When food or beverage is left over, laboratory testing may be constrained by its quality, days after the event. A specific organism may be recovered, but it could be due to the length of time the item was stored or the condition in which it was stored before testing, rather than its existence when the food was consumed and as cause of the outbreak. Figure 21 below, shows the foods implicated in foodborne outbreaks reported to the IDPH from 1990-2005. This graph includes both suspect and confirmed foods indicated by the laboratory, environmental, and/or epidemiological investigations. In 55 of the 140 outbreaks reported during this time period, no food vehicle could be determined.
This result could have been due to several factors, including multiple foods being contaminated, recall bias of the individuals involved or the small size of the sample. Beef and pork were the top food items implicated in foodborne disease outbreaks in Iowa during 1990-2005, implicated in 14.3 percent (n=20) of outbreaks. Poultry was a close second, implicated in 12.1 percent of outbreaks. Fruits and vegetables (9.3%, n=13) and salads (7.9%, n=11) were also important vehicles. Sauces and gravies and the ‘other’ category, each compromised 6.4 percent (n=9) of all foodborne outbreaks reported. The ‘other’ category included cake, powdered punch mix, chips with chili cheese dip, cookies, chips and salsa, coconut cream pie, mints, kombucha tea, and ice/water. Dishes with multiple ingredients, grains and breads, eggs and dairy products, and seafood categories each contributed to less than 6 percent.
The map below shows the counties where foodborne outbreaks reported to the IDPH occurred from 1990-2005.
Resources
Hepatitis A Transmitted by Food, Anthony E. Fiore, Food Safety Clinical Infectious Diseases 2004;38:705-15.
http://www.cfsan.fda.gov/~ebam/bam-25.html
http://www.cdc.gov/ncidod/EID/vol11no07/04-0634.htm
Preliminary FoodNet Data on the Incidence of Infection with Pathogens Transmitted Commonly Through Food – 10 States, United States, 2005. MMWR, April 14, 2006. 55(14):392-395.