Many foods have the potential to cause foodborne illness; however, some pose a higher risk. Data were collected through the Foodbook study, a population-based telephone survey conducted between 2014 and 2015 that assessed 10,942 Canadians' food exposures using a 7-day recall period. The 19 foods included in the survey were identified as high risk for common foodborne pathogens in Canada. Results were analyzed by age group, gender, region of residence, income, and education. Consumption proportions of high-risk foods ranged from 0.4% (raw oysters) to 49.3% (deli meats). Roughly 94% of the population reported consuming one or more high-risk food in the past week. Certain high-risk food behaviors were associated with demographic characteristics. High-risk adults such as those 65 years or older still report consuming high-risk foods of concern, including deli meats (41.8%), soft cheeses (13.7%), and smoked fish (6.3%). Consumption of certain foods differed between genders, with males consuming significantly more deli meats, hot dogs, and raw or undercooked eggs and females consuming significantly more prebagged mixed salad greens. The overall number of high-risk foods consumed was similar, with both genders most frequently consuming three to five high-risk foods. High-risk food consumption was seen to increase with increasing household income, with 14.2% of the highest income level consuming six-plus high-risk foods in the past week, compared with 7.1% of the lowest income level. If a respondent had heard of a risk of foodborne illness associated with a food, it did not affect whether it was consumed. Additional consumer food safety efforts put in place alongside current messaging may improve high-risk food consumption behaviors. Enhancing current messaging by using multifaceted communications (e.g., social media and information pamphlets) and highlighting the large incidence and severity of foodborne illnesses in Canada are important strategies to improve behavior change.
Most Canadians reported eating 1+ high-risk food in the past 7 days.
High-risk adults 65 years and older reported similar high-risk food consumption as the low-risk group.
Knowing risks associated with a food did not affect consumption of high-risk foods.
Foodborne illnesses present a substantial health burden in Canada. It has been estimated that 4.0 million cases of domestically acquired foodborne illnesses (i.e., acquired within Canada) occur each year in Canada, as well as more than 11,500 hospitalizations and 238 deaths (71, 72). Foodborne illness prevention requires vigilance and cooperation among all members of the farm-to-fork continuum; however, consumers are often considered the last line of defense. Along with potentially unsafe food handling practices in the home, high-risk food consumption behaviors are an important factor contributing to foodborne illness (37). It is important for individuals to be informed of and follow correct food safety practices in the home. It is also important that consumers are aware of the increased risks associated with specific food items so that they can make informed decisions about what products to consume.
Many foods have the potential to cause foodborne illness. However, some foods pose a higher risk. Depending on how the food item is grown, processed, or transported, there may be increased opportunities for contamination or pathogen growth. For example, foods that are consumed raw, such as unpasteurized milk, cheese, and juices, or raw vegetables, including sprouts and leafy greens, lack a kill step to remove pathogens that may be present (2, 46). Generally, many food items may not be considered risky if consumed fully cooked; however, the risk presents itself when the item is deliberately consumed raw or undercooked, for example, preparing a meal with undercooked or raw eggs or consuming raw animal products such as raw beef, fish, and oysters (35, 38, 75). In addition, some foods are higher risk because of the processing they undergo (e.g., ground beef and pâté or meat spreads). In these cases, pathogens present on the surface can be spread among the meat when it is ground or made into a paste (36, 38). Similarly, with deli meats, cross-contamination can happen if there are pathogens present on the surface of the meat or on the tools processing the item (e.g., a deli slicer), introducing pathogens to the meat when it is sliced (40). This may also be the case for fruits such as melons, where pathogens can be transferred to the inside of the fruit as it is cut, especially if the fruit is not washed or is difficult to wash sufficiently, as with cantaloupe (29). Food handling practices in the home are important when considering the consumption of store-bought breaded chicken products, which have been associated with Salmonella outbreaks in the past (49). At the time of the Foodbook study, these products were commonly sold as a raw meat product; however, because they were par fried before being frozen, they appeared cooked. As a result, it has been estimated that 40% of consumers considered these products to be precooked (23) and may not have taken the proper food safety steps when handling or cooking these raw meat products.
The U.S. Centers for Disease Control and Prevention provided estimates of annual domestically acquired foodborne illnesses attributed to specific food commodities based on reported outbreaks. The largest proportion of these illnesses (51.1%) was attributed to plants, with 22.3% of illnesses attributed to leafy greens alone (56). A further 41.7% were attributed to land animals (13.8% dairy, 12.2% meat, and 9.8% poultry), and 6.1% were attributed to aquatic animals (56).
In 2014, a systematic literature search was conducted by Nesbitt et al. (53) found survey results from 26 Canadian publications from 1998 to 2011 regarding Canadians' consumer food safety knowledge and behavior. Questions in nine of these surveys assessed knowledge and/or consumption of high-risk foods. These questions included asking about consumers' knowledge or beliefs about risky foods, rating foodborne illness risks related to specific foods, or identifying foods that are high risk (53). The list of foods determined to be high risk varied across surveys; however, food items identified as high risk were consistent across many studies, with frequently cited foods including meats and poultry, followed by fish and seafood, dairy products, eggs, and produce. Raising awareness of high-risk foods was identified as an area requiring targeted consumer food safety education (53).
Some populations are at a higher risk of developing foodborne illness or are more likely to develop a severe illness requiring medical treatment (65). Age is one factor that can alter the risk of developing a foodborne illness. Adults aged 65 years and older are at a higher risk because their immune systems may no longer be as effective at fighting off infections (41, 69). Children 5 years old and younger are also more susceptible because their immune systems are still developing and cannot fight off infection as effectively (32, 67).
This article provides an in-depth description of consumption patterns in Canada for foods considered higher risk because of their greater potential to be contaminated with common foodborne pathogens in Canada. These include four bacterial pathogens (Escherichia coli, Salmonella, Campylobacter jejuni, and Listeria) and one foodborne virus (norovirus). These pathogens account for roughly 80% of foodborne illnesses in Canada (72).
Data on food consumption were collected as part of the Foodbook study, a national population survey conducted in 2014 to 2015 (58). The objectives of this article are to frame high-risk food consumption data around high-risk age groups and other demographic groups and evaluate patterns in high-risk food consumption as a whole. This article also combines the consumption of high-risk foods with the knowledge of their risk, which is important for informing food safety programs and education efforts.
MATERIALS AND METHODS
Data were collected as part of the Foodbook study, a population-based telephone survey conducted over a 1-year period between April 2014 and April 2015. The Foodbook study included questions on food consumption, animal and water exposures, consumer food safety knowledge and behaviors, and recent acute gastrointestinal illness. Households were randomly selected from a sampling frame of telephone numbers that consisted of landlines (70% listed and 10% random digit dialing) and cell phones (20%). Surveys were collected over a 12-month period across all 13 provinces and territories and four age groups (0 to 9, 10 to 19, 20 to 64, and 65+ years). Individuals were excluded if they could not speak the supported languages (English, French, Inuktitut, and on-demand verbal translation for other languages), if they did not have a listed landline or cell phone number, or had traveled outside their province or territory of residence during the 7 days before the interview. The Foodbook study was reviewed and approved by Health Canada and the Public Health Agency of Canada's Research Ethics Board (REB 2013-0025), as well as by the Newfoundland and Labrador Health Research Ethics Authority to meet a unique provincial legal requirement (HREB 13.238). Additional details on sampling frame, participant selection, questionnaire administration, and data collection are outlined in the original Foodbook Report (58). A full data set and copy of the questionnaire are available through the Government of Canada's Open Data Platform (59). Demographic information collected included age, gender, province or territory of residence, education level, and household income level. A module of questions regarding consumer food safety practices and knowledge was added to the Foodbook study from November 2014 to April 2015. The questions from this module were asked only of survey participants 18 years or older. These questions were developed in consultation with food safety stakeholders (50).
For this analysis, the respondents were grouped into three age groups: 0- to 5-year-olds, 6- to 64-year-olds, and 65+-year-olds. By doing so, comparisons could be made between high-risk age groups (children younger than 5 years and older adults) and to one low-risk group (e.g., 6- to 64-year-olds). In addition, the province or territory of residence variable was combined to represent six main regions of Canada: British Columbia, the Prairie region (Alberta, Saskatchewan, and Manitoba), Ontario, Québec, the Atlantic region (New Brunswick, Nova Scotia, Prince Edward Island, and Newfoundland and Labrador), and the North region (Yukon and Northwest Territories and Nunavut). The highest level of education completed was collected from participants older than 25 years. The results were grouped into three categories: (i) high-school diploma or less, (ii) trade or college diploma, and (iii) bachelor's degree and above. Finally, participants were asked which of the following categories included their total household annual income. The four categories were (i) less than $30,000; (ii) $30,000 or more, but under $60,000; (iii) $60,000 or more, but under $80,000; and (iv) $80,000 or more.
High-risk food selection
Outbreak investigations and consumer food safety information have helped to identify foods that present a potential higher risk for foodborne illness. Foods were classified as high risk based on information provided by Health Canada, including fact sheets for foods considered high risk for common foodborne pathogens and those considered high risk for consumption by vulnerable populations (31, 33, 34). Canadian food recalls and foodborne illness outbreaks within the last 5 years were also reviewed to determine whether additional or novel food sources should be included. Tahini (captured in Foodbook as “tahini, halva or other products made from sesame seeds”) was included as a high-risk food because of frequent recalls issued since 2015 associated with Salmonella and it has been identified as an unusual food source for Salmonella contamination by the World Health Organization's International Food Safety Authorities Network (4, 5, 8, 13, 76). Prebagged mixed salad greens have also been associated with numerous recalls since 2015 because of Listeria and E. coli contamination and therefore were included as a high-risk food (6, 9, 10, 60).
In total, 19 foods were considered high risk: unpasteurized milk, soft cheese, unpasteurized milk cheese, store-bought breaded chicken, ground beef consumed raw or undercooked, raw beef, deli meats, hot dogs, pâté or meat spread, raw oysters, smoked fish, raw fish, unpasteurized juice, romaine lettuce, prebagged mixed salad greens, melons, sprouts, raw or undercooked eggs, and tahini. Foodbook study respondents were asked whether they had consumed each food item in the last 7 days.
For a subset of 10 foods, the frequency of consumption was also asked of respondents who had consumed the food item in the last 7 days. For example, if a respondent said yes to consuming sprouts in the past 7 days, the interviewer then asked how many times they had consumed sprouts in the past 7 days. Those foods were unpasteurized milk, unpasteurized milk cheese, ground beef consumed raw or undercooked, raw beef, deli meats, raw oysters, romaine lettuce, prebagged mixed salad greens, sprouts, and tahini.
High-risk food classification
The 19 high-risk foods included three dairy products, nine meat or seafood products, five produce items, and two other foods. Four foods were included because they were high risk for Salmonella contamination: raw or undercooked eggs, store-bought breaded chicken, melons, and tahini (7, 29, 49, 75, 76). Romaine lettuce was included because it is high risk for E. coli contamination (21). Six high-risk foods for Listeria contamination were soft cheeses, deli meats, hot dogs, pâté or meat spreads, smoked fish, and raw fish (45, 62, 66). Raw oysters were included because they are high risk for norovirus contamination (44). An additional seven foods were considered high risk for more than one bacterial pathogen, including ground beef consumed raw or undercooked, raw beef, and unpasteurized juices, all of which are considered high risk for E. coli, Campylobacter, and Salmonella contamination (2, 28, 70). Unpasteurized milk and unpasteurized milk cheese are high risk for Salmonella, E. coli, Listeria, and Campylobacter contamination, and prebagged mixed salad greens are high risk for both E. coli and Listeria (9, 10, 20).
Knowledge of high-risk foods
During a 6-month period of the Foodbook study (November 2014 to April 2015), respondents ages 18 years and older (those most likely to be responsible for food preparation) were also asked whether they had heard of risks of foodborne illness associated with the consumption of a set list of foods. These foods consisted of both high-risk foods (hamburger, chicken, deli meats, frozen chicken nuggets, unpasteurized milk, soft unpasteurized cheese, unpasteurized juice, alfalfa sprouts, raw oysters, and raw eggs) and lower-risk or control foods (pasteurized milk and hard cheese).
Statistical analysis was performed in Stata 15.1 using the weighted survey data. To weight the data, the census metropolitan area nearest to where the respondent lives was used, along with age group, household type, province or territory, the number of people in the household, the number of landlines and cell phones in the household, and gender to calculate the individual-level survey expansion weight. The final weighting variable was created using a poststratification step with iterative proportional ranking. High-risk food exposures were analyzed nationally, as well as by gender, age group, education level, income level, and region of residence. The frequency of consumption (in the past 7 days) for a subset of 10 foods was also analyzed, as previously mentioned. If respondents indicated having consumed a particular food in the past 7 days, the average number of times that particular food was consumed ranged from 1.16 times (raw oysters) to 6.68 times (unpasteurized milk), with an overall average of 2.55 times. As a result, the consumption frequency categories chosen were 0 times, 1 to 3 times (average), and 4-plus times (above average). The total number of high-risk foods consumed overall was also calculated by creating a variable to capture the sum of the number of high-risk foods individuals reported eating. This variable was then broken down into four categories: zero high-risk foods, one to two high-risk foods, three to five high-risk foods, and six or more high-risk foods. Finally, for each food item included in the knowledge of high-risk foods subsection, respondents were categorized into four groups: (i) those who had heard of a risk associated with the food and ate the food, (ii) those who had heard of a risk associated with the food and did not eat the food, (iii) those who had not heard of a risk associated with the food and ate the food, and (iv) those who had not heard of a risk associated with the food and did not eat the food.
Exposures reported by each demographic level were compared with the referent level in that demographic group. The level with the largest proportion of respondents was chosen to be the referent group. Females were therefore chosen as the referent group for gender comparisons, and the 6- to 64-year age group was chosen for all age group comparisons. The Prairie region (Alberta, Saskatchewan, and Manitoba), was selected as the referent for all regional comparisons. Those with an education of high-school diploma or less formed the referent group for each education comparison, and the highest income category (more than $80,000) was the referent for all income comparisons.
Significant differences in exposures between levels were assessed using an adjusted Wald test with a P value cutoff of 0.05. Statistically significant results were determined using a Bonferroni-corrected two-tailed P value test, meaning that an original alpha value of 0.05 was divided by the number of statistical tests to get the final P value for each independent variable. For example, when looking for significant results among the four income levels, the Bonferroni-corrected P value were set to 0.01 (0.05/4 = 0.0125). In that case, only results with a P value of less than 0.01 were labeled as significant. Then, 95% confidence intervals were generated for all proportions.
In total, 10,942 Canadians completed the Foodbook survey, and data were weighted to represent the Canadian population. Data regarding food safety knowledge and practices were collected from 2,413 of those respondents, during the second half of the 12-month study period (e.g., over a 6-month period).
High-risk food consumption by demographic groups
Nationally, consumption proportions of high-risk foods (in the previous 7 days) ranged from 0.4% (raw oysters) to 49.3% (deli meats; Table 1). Comparing exposure proportions between genders, males were found to eat significantly more deli meats, hot dogs, and raw or undercooked eggs compared with females (P < 0.03). Females consumed significantly more prebagged mixed salad greens (P < 0.03). Most (94%) of the population reported consuming one or more high-risk food in the past 7 days, and more than half reported consuming three or more high-risk foods (Table 2).
Young children (ages 0 to 5 years) reported significantly higher consumption of store-bought breaded chicken and melons compared with the referent group (those aged 6 to 64 years; P < 0.01; Table 1). Children were also found to eat significantly less soft cheese, raw oysters, raw fish, romaine lettuce, prebagged mixed salad greens, and raw or undercooked eggs compared with the referent group (P < 0.01). Compared with the referent group, older adults (ages 65 years and older) were found to consume significantly less unpasteurized milk, store-bought breaded chicken, deli meats, raw fish, unpasteurized juice, romaine lettuce, and prebagged mixed salad greens (P < 0.01). A significantly higher proportion of both the 0 to 5 and the 65+ age groups were found to have consumed no high-risk foods over the past week compared with the referent group (P < 0.01).
Consumption of high-risk foods varied by region. Compared with the Prairie region, a significantly higher proportion of individuals from Québec were found to consume soft cheeses, unpasteurized milk cheese, ground beef consumed raw or undercooked, pâté or meat spread, smoked fish, unpasteurized juice, and sprouts (P < 0.01). Residents of British Columbia consumed significantly more soft cheese, smoked fish, and raw fish compared with the referent group, and Ontario residents consumed significantly more soft cheese and unpasteurized juice compared with the referent group (P < 0.01).
Compared with the referent group (the highest income level, or those with a household income greater than $80,000), all other income categories consumed prebagged mixed salad greens significantly less (P < 0.01). The two lowest income categories (a household income of less than $30,000 and a household income of more than $30,000, but less than $60,000) also consumed significantly less soft cheese, deli meats, and romaine lettuce (P < 0.01). The lowest income category also consumed significantly less raw fish (P < 0.01). The total number of high-risk foods consumed was found to increase as income category increased (Table 2). Of those in the highest income category, 14% consumed six or more high-risk foods in the past week, compared with only 5.3% of those in the lowest income category (less than $30,000; P < 0.01; Table 2).
Those with a trade, college, or nonuniversity certificate or diploma consumed significantly more soft cheese, raw fish, romaine lettuce, prebagged mixed salad greens, melon, and tahini and had significantly lower hot dog consumption compared with the referent group (those with a high-school diploma or less; P < 0.02). Those with a bachelor's degree or a degree above the bachelor's level consumed significantly more soft cheese, smoked fish, raw fish, romaine lettuce, and prebagged mixed salad greens compared with the referent group (P < 0.02).
Frequency of consumption
The frequency of consumption for a subset of 10 foods was examined (Table 3). Survey respondents who indicated having consumed any of the subset of 10 foods in the past 7 days were asked how many times they had consumed that particular food. For nine of the foods, the most common consumption frequency was one to three times. For unpasteurized milk, the most common consumption frequency was four or more times.
Knowledge of high-risk foods
Comparing those who had heard of a risk associated with a food and those who did not, there were no significant differences in consumption of any of 12 foods, whether the food was a control item or a risky food item. Similarly, there were no significant differences in these comparisons by gender or for those aged 65 years and older compared with those ages 6 to 64 years.
Additional result tables are provided in Supplemental Appendix 1.
Purpose of the study
This study accomplished its intended purposes: to determine the prevalence of high-risk food consumption in the Canadian population and to elucidate patterns of high-risk food consumption as it relates to demographic factors. This article also combined the consumption of high-risk foods with knowledge of their risk.
High-risk foods included in the analysis
High-risk food consumption in other studies
Compared with the Nesbitt et al. (53) food consumption study that took place in Ontario's Waterloo region in 2005 to 2006, consumption has increased for many foods with the exception of raw or undercooked eggs, in which the proportion of the population consuming this food item has decreased. Of the eight foods that were directly comparable, consumption proportions were within 5% of each other for five foods. Because the Nesbitt et al. (53) study was not nationally representative, it is difficult to draw conclusions about trends in high-risk food consumption over time for the Canadian population. Although another food consumption survey would be required to draw stronger conclusions, this indicates that there may be an overall increasing trend in consumption of high-risk foods over time.
The U.S. Centers for Disease Control and Prevention conducted a food consumption survey in 2018 to 2019 in 10 of their FoodNet sites across the country. Nine foods were directly comparable to the list of high-risk foods provided here. Of these foods, consumption proportions were within roughly 5% of each other for all nine of these foods. Respondents in the U.S. survey were found to consume seven of these foods in higher proportions, and the largest difference was found in raw or undercooked eggs, with a difference of 5.6% (consumption proportions of 15.0% in Canada and 20.6% in the United States) (18).
High-risk food consumption by demographic groups
Although there were significant differences between genders when looking at each high-risk food individually, there were no significant differences observed when comparing the overall number of high-risk foods consumed. This indicates that the difference between genders may be related more to food preferences rather than avoiding high-risk foods. In general, in this study, higher consumption proportions were seen of meats and animal products for males and greens and vegetables for females, which is consistent with other studies (52, 68). Past studies have found women to be more nutrition conscious overall than men, avoiding high-fat foods and consuming more fruits and vegetables (55, 74). Furthermore, most of those following a vegetarian diet are female, and women who do eat meat have been found to consume less of it than men (63, 73).
Exposure to store-bought breaded chicken and melon was significantly higher among children aged 0 to 5 years compared with 6- to 64-year-olds, reflecting patterns in food preferences, as well as foods targeted for consumption among children. Given that young children, specifically those 5 years and younger, are at a higher risk for foodborne illness (32, 67), this finding reinforces the importance of ensuring continued targeted public health messaging to parents or guardians of young children to maintain and increase awareness of the risks associated with those foods and how to safely prepare and consume them. Although policy change has reduced the risk associated with store-bought breaded chicken (49), melons are considered high risk if not handled and washed properly (29).
There were no high-risk food items that older adults ate significantly more of compared with the referent group. Comparing the 65+ and 6 to 64 age groups, there were similar consumption proportions of certain foods, such as soft cheeses, hot dogs, pâté or meat spread, and smoked fish. These are of particular concern because these foods present a high risk for Listeria contamination, which can have serious impacts on the health of older adults. These findings are supported by previous studies that have shown that despite being at a higher risk for infection and health complications, many older adults are still consuming high-risk foods (3, 15, 39). Continued targeted public health messaging to individuals older than 65 years is critical.
Consistent with other studies, the consumption of high-risk foods was found to increase with increasing income (1, 52, 57, 64). Many high-risk foods, including fresh produce, raw oysters, and soft cheeses, are more expensive than processed foods, and individuals with a lower income may not be able to purchase them regularly (24).
Only a small proportion (9.6%) of individuals consumed six or more high-risk foods in the study period. However, the largest proportion of respondents (49.3%) reported consuming three to five high-risk foods in the past 7 days. This indicates that many Canadians are exposed to high-risk foods multiple times each week, putting themselves at an increased risk of exposure to pathogens.
Nationally, deli meats are the most frequently reported high-risk food consumed among Foodbook respondents, and although there is a significant decrease in consumption in the 65+ age group compared with the referent group, it is still the second-highest consumed high-risk food in that age group. Deli meats have been implicated in multiple Listeria monocytogenes outbreaks and are considered a high-risk food for older adults (22). Listeria infection in these populations could lead to serious health complications. Therefore, information about the risk associated with deli meat consumption should continue to be made available and targeted to higher-risk populations.
Individuals in the 65+ age group who are living in long-term care or retirement homes, or potentially requiring more hospital stays, may have a lack of or limited choice when it comes to the food that they are served, and deli meats and other ready-to-eat foods that are risky foods for this population may be served frequently (19, 51). Therefore, the food consumption patterns in a portion of this population may be due more to what is available, and targeting messaging to long-term care providers would be more appropriate for those individuals. However, it is possible that because Foodbook only included residential and cell phone numbers, those living in an institution may have been excluded, making these findings generalizable more so to older adults not hospitalized or living in long-term care facilities.
Romaine lettuce and prebagged mixed salad greens are commonly consumed by Canadians, with almost half of the Foodbook respondents (48.8 and 46%, respectively) reporting having consumed each within the last 7 days. Recently, there have been multiple large outbreaks of E. coli associated with romaine lettuce (11, 14), and in several instances, prebagged mixed salad greens have been associated with Listeria and E. coli outbreaks and recalls (6, 9, 10, 60). These foods are part of a healthy and balanced diet, but because they are fresh produce items that have the potential to be and have been associated with bacterial contamination, and have been implicated in numerous recalls, they pose a foodborne disease risk. Therefore, communication to the public should continue with the messaging currently available, focusing on how to prevent foodborne illness, for example, by properly preparing and storing produce before consumption and avoiding the potential for cross-contamination.
During the Foodbook study period, store-bought breaded chicken products were considered a high-risk food for Salmonella contamination because many of these products were raw chicken. Chicken, including store-bought breaded chicken, was identified as the source of 19 national outbreaks of Salmonella between January 2015 and April 2019 and was implicated as a common source of Salmonella infection before 2015 (23, 42, 45, 49). However, following changes in regulation in 2019 for frozen raw breaded chicken products, the industry is required to implement measures at the manufacturing or processing level to reduce Salmonella to below detectable amounts, meaning these products will no longer be considered high risk for Salmonella (12). This demonstrates that processing and/or regulatory changes can help reduce the risk associated with specific food products.
Knowledge of high-risk foods
Using data also captured as part of the Foodbook survey, Murray et al. (50) examined consumers' food safety knowledge and practices, including those related to high-risk foods. The current study provides an additional level of analysis to the Murray et al. study in that it combines knowledge with reported consumption practices. Murray et al. found that older respondents (aged 60 years and older) were less likely to be aware of risks associated with raw eggs, alfalfa sprouts, and unpasteurized juice. Furthermore, almost half of the older adults surveyed had not heard of the risks associated with deli meat (44), which, combined with the fact that they were consumed by roughly 42% of Foodbook respondents in the older age group, may suggest an area for increased targeted communication.
In general, Murray et al. (50) found mixed levels of awareness: most respondents had heard about risks associated with chicken and hamburger, raw eggs, unpasteurized milk, deli meats, and oysters, but fewer respondents had heard about risks associated with soft unpasteurized cheese, sprouts, unpasteurized juice, and frozen chicken nuggets. In addition, some respondents reported that they had heard of risks associated with the control food items, for example, pasteurized milk (50). These findings suggest that for a proportion of the population, there is a lack of awareness about high-risk foods. Combining this knowledge with known consumption practices allows improved targeted consumer messaging. However, our findings showed that there were no significant differences in consumption based on whether the respondent had heard of a risk associated with the food. This can be interpreted to say that for the population as a whole, consumption does not change based on having heard of risk.
There are many reasons individuals who are aware of or knowledgeable about the risk associated with consuming a particular food might continue to consume it. Generally, consumers may not believe that they are at risk of contracting a foodborne illness from food they prepared in their own home, especially those who do not perceive themselves to be at higher risk for infections or complications (77). Consumers may also be hesitant to change their behavior if they do not believe that they have previously had a foodborne illness and therefore believe they do not need to change their food consumption patterns or food preparation behaviors (77). For some foods, consumers may simply prefer the taste of food prepared in a certain way (e.g., undercooked) compared with one that follows recommended preparation practices (e.g., preferring eggs done over easy) and will make their food choices based on preference rather than food safety (77). Preparing the food in a way that decreases the risk of foodborne illness may also conflict with preestablished food preparation habits, and with older adults it may be even more difficult to change their habits or views of food safety if they have been using the same methods for years (77). Finally, food consumption practices vary between cultural and ethnic groups, and individuals may not want to change a practice to something outside of their cultural norm and may not even be aware of the risk, given that certain foods may not be available or commonly consumed within their culture (76).
Fein et al. (25) observed that mistakes in food preparation seemed to be a problem of attention and vigilance more so than knowledge or beliefs and that when reminded about potential hazards, people acted more safely. To motivate consumers to change their behavior, increasing awareness of the negative and possibly severe consequences of foodborne illnesses may help motivate consumers to practice safer food consumption and handling (77). Furthermore, messaging may be more effective coming from a source that consumers trust, such as an influential figure or, especially for vulnerable populations, their family doctor. To increase effectiveness of food safety education campaigns, it has been found that a multifaceted approach has been effective, for example, using social media to reach young people, in addition to targeted fact sheets for older adults. Specifically for older adults, it is recommended to use printed materials combined with personal contact, such as having the recommendations communicated to them by a doctor, and to focus on a limited number of practices (39). These methods may be used to better reach target or high-risk populations (15, 43). However, more research is needed to determine what may motivate consumers to follow the recommended practices.
Many findings presented here are consistent with food consumption patterns shown elsewhere, providing support for how food safety messaging has been targeted to consumers in the past (18, 52). Thus, the overarching recommendation is to continue with the current messaging. However, because similar risk behaviors are continually identified over time regardless of current messaging in place, other efforts could be put in place in conjunction with the current messaging to have an effect in changing behavior. Furthermore, this messaging could highlight the importance of safe food handling and consumption within the home, citing the high proportion of foodborne illnesses contracted at home (53, 61). In addition, more innovative and engaging communication may improve uptake of the message by consumers and empower them to change their behaviors. Building on our recommendations for continued messaging, we suggest enhancing the current messaging by employing multifaceted approaches, as mentioned in the paragraph above. Traditional strategies can be supplemented with approaches such as integrating reminders into recipes, utilizing social media on multiple platforms, or partnering with grocery stores with loyalty card programs so that those purchasing high-risk food items have access to information regarding their risk and safe handling (47). Messaging could also highlight the large incidence of foodborne illnesses in the Canadian population and the severe consequences to which they can lead. This messaging can also be used to increase awareness about novel and/or recently identified high-risk foods (e.g., flour) that were not previously thought of as being a risky product.
There were several limitations to this study. First, results are based on self-reported food consumption, which can be biased toward socially desirable responses (e.g., not reporting consumption of unpasteurized milk). Therefore, the proportion of individuals consuming some foods may be higher than what is reported. Recall bias may have also played a role; asking about food consumption over the past 7 days may be difficult for some individuals. In addition, some individuals may not be able to accurately report whether they had consumed the food for other reasons, for example, not being aware of whether a cheese is pasteurized or confusing romaine and iceberg lettuce. This may also include the respondent not being aware that a composite food may contain one of the high-risk ingredients, for example, tahini in hummus or sprouts in a salad.
The high-risk foods selected for analysis were limited to those included in the Foodbook study. Some foods, for example, raw flour, which emerged as a high-risk food in recent years (26, 48), could not be included on this list because there were no data collected regarding its consumption in the Foodbook study.
Pregnant and immunocompromised individuals are also considered a high-risk population for foodborne illness (16, 30, 41). However, pregnancy and health status were not consistently captured in the Foodbook study. Therefore, the low-risk 6- to 64-year age group likely includes a small number of immunocompromised individuals, pregnant women, or women who may become pregnant who may be altering their food consumption habits because of their health status. Future food consumption surveys should capture pregnancy and health status to identify food consumption behaviors in these groups and to detect areas of concern as they pertain to high-risk food consumption.
The results could also be biased because of the exclusion of individuals who do not have a landline or cell phone and could not participate in the study, as well as those who speak a language that did not allow them to be included in the study. There may also be an underrepresentation of older adults living in retirement or other full-time care homes that may not have had access to their own telephone, which may have influenced data regarding food consumption patterns in this population, as mentioned in “High-risk food consumption by demographic group.”
Safe food handling practices were also not considered, which, for some foods, can significantly decrease risk. For example, the risk associated with consuming hot dogs changes greatly depending on how well the item is cooked and handled. Further research into food safety practices as protective factors should be considered. For example, if an individual reports consuming a high-risk food, follow-up questions can ask about the steps they take to decrease the risk associated with the food to determine how large a risk to which they are truly exposing themselves.
The findings presented in this study show that most Canadians expose themselves to at least one high-risk food every week. These findings offer an improved understanding of which subgroups of the Canadian population are more likely to consume certain high-risk foods or consume more high-risk foods overall. In the future, these findings can be used to reinforce and guide public health messaging and develop more effective and targeted campaigns to reduce the burden of foodborne illness associated with high-risk food consumption.
The authors acknowledge the Centre for Foodborne, Environmental and Zoonotic Infectious Diseases (Public Health Agency of Canada), Outbreak Management Division and Foodborne Disease and Antimicrobial Resistance Surveillance Division, and provincial and territorial epidemiologists for their assistance with the Foodbook project; the staff at R. A. Malatest & Associates Ltd. for their expert interviewing; and the survey respondents for their participation.
Supplemental material associated with this article can be found online at: https://doi.org/10.4315/JFP-21-101.s1