|Nutrition Evidence Library|
Risky Foods: To what extent do US consumers eat raw or undercooked animal foods?
Moderate, clear and consistent evidence shows that the consumption of raw or undercooked animal-source food products is relatively common in the US, especially for eggs and egg-containing products and ground beef products.
Overall strength of the available supporting evidence: Strong; Moderate; Limited; Expert Opinion Only; Grade not assignable For additional information regarding how to interpret grades, click here.
Evidence Summary Overview
A total of eight studies were reviewed regarding the extent to which US consumers eat raw or undercooked animal foods. All of the studies (one meta-analysis, one systematic review and six cross-sectional studies) received neutral quality ratings.
In their direct observation study of US household meal preparers, Anderson et al (2004) found that 61% of those who prepared a chicken entrée undercooked the chicken. In this study, 46% of those who chose to prepare meatloaf undercooked the ground beef. In contrast, Dharod et al (2007b) documented that almost none (7%) of the Puerto Rican household meal preparers included in their study undercooked the chicken. Lopez Osorio et al (2008) found that US consumers were more likely than Argentinean and Spanish consumers to prefer beef steaks to be cooked rare. However, Trepka et al (2007) found in their study that only 3.5% of Women, Infants and Children (WIC) participants liked their meat cooked medium-rare or rare.
Studies reviewed have found that among diverse US study populations, raw or undercooked animal-derived products are widely consumed (Bryd-Bredbenner et al, 2008; Patil et al, 2005; Trepka et al, 2007). Bryd-Bredbenner et al (2008) reported that among a large sample of college students, a substantial number reported consuming a variety of risky foods, such as homemade cookie dough containing raw eggs (53%), fried eggs with runny or soft yolks (33%), sushi (29%), raw sprouts (29%), raw oysters, mussels or clams (11%) and rare hamburgers (7%). Trepka et al (2007) found that among female African-American WIC clients, 24.7% reported usually eating undercooked eggs, 51.6 percent of pregnant women reported “sometimes,” or “frequently,” eating hot dogs or deli meats since becoming pregnant without first reheating them, and 35.5% reported eating soft cheeses and blue-veined cheeses sometimes or more frequently since becoming pregnant. In addition, almost 12% reported consuming hamburgers with pink or red color inside, and only 62% reported always using boiling water before preparing infant formula.
The prevalent consumption of undercooked eggs detected in localized studies is confirmed by a systematic review (Redmond and Griffith, 2003) and the meta-analysis by Patil et al (2005). Based on US surveys conducted between 1977 and 2000, Redmond and Griffith (2003) report that the prevalence for this practice has ranged from 5% to 56%, with the most recent surveys suggesting that as many as half of the US population may consume undercooked or raw eggs.
Raw milk consumption has been associated with serious foodborne outbreaks in the US. Kaylegian et al (2008) examined raw milk consumption practices in a sample formed predominantly of dairy farmers from upstate New York. As many as 45.3% reported having consumed raw milk during the previous year. The main reasons for consuming raw milk were taste, convenience and cost. Concerns related to health hazards associated with raw milk consumption were expressed by 38.2% of the raw milk and 73.2% of the pasteurized milk consumers.
Evidence Summary Paragraphs
Anderson et al, 2004 (neutral quality), a cross-sectional study compared consumer food-handling behaviors with the Fight BAC! Consumer food-safety recommendations. Ninety-nine subjects (92 women and seven men) were randomly recruited by telephone, and videotaped in their home while preparing a meal. Videotapes were coded according to Fight BAC! recommendations, a food safety survey was administered and temperature data was collected. The authors found that many subjects undercooked the meat and poultry entrees and very few subjects used a food thermometer. More specifically, 61% of those who prepared a chicken entrée undercooked the chicken, and 46% of those who chose to prepare meatloaf undercooked the ground beef. Overall, subjects did not follow the Fight BAC! recommendations for safe food handling.
Byrd-Bredbenner et al, 2008 (neutral quality), a cross-sectional survey assessed risky eating behaviors among 4,343 (female, 65%; male, 35%) young adults enrolled in 21 colleges and universities located in 17 US states (mean age 19.92±.67 years). Students across the US, enrolled in introductory courses, were invited to complete an online food safety survey between January and October, 2005. A calculated mean risky eating score of 5.1±3.6 indicated college students consume some risky foods (53% consumed raw homemade cookie dough; 33% consumed fried eggs with runny or soft yolks; 29% consumed sushi; 29% raw sprouts; 11% raw oysters, clams, or mussels; and 7% consumed hamburgers cooked rare). Men ate significantly more risky foods than women (P<0.0001), white participants engaged in significantly more risky eating behaviors than non-white participants (P<0.001). Students had strong feelings of food safety self-efficacy (4.1±0.6), were between the contemplation and preparation stage-of-change (2.7±1.2), believed food poisoning was somewhat of a threat (3.1±0.8) and had modest food safety knowledge.
Dharod et al, 2007b (neutral quality), a cross-sectional study, applied the Hazard Analysis Critical Control Points (HACCP) model at the household level to identify sanitation and food handling "Critical Control Points" for home prepared "Chicken and Salad" using direct observations and microbiological indicators. A sample of 60 Puerto Rican women recruited in inner city Hartford, Connecticut, were provided chicken breasts (CB), lettuce and tomatoes (LT) and spices to prepare a meal in their home kitchens; food and kitchen surface samples were collected during stages of food preparation and tested for total and coliform counts, and presence of pathogenic microorganisms; observed food handling behaviors were compared with microbial testing results. The authors observed that no participants used a thermometer to check whether the CB was adequately cooked [most determined doneness using cooking time and visual change in texture and color of meat and some (20%) tasted meat to determine doneness]. However, temperature measurements by research staff on meat showed that 93% of participants cooked the CB to an adequate temperature.
Kaylegian et al, 2008 (neutral quality), a cross-sectional survey determined raw milk consumption beliefs and practices among New York State dairy producers and farm workers. An eight-question survey was developed to collect information on demographics, previous household milk consumption practices, reasons for consuming or not consuming raw milk, whether raw milk was supplied to others in the community, demographics of community raw milk consumers and concerns about raw milk consumption practices. Data set was adjusted to only include dairy producers and farm workers so that 150 responses were analyzed from 336 mailed surveys. Regarding demographics of raw milk consumers, dairy producers represented the majority (89.7%) of raw milk drinkers while 10.3% were farm workers; 72% of raw milk consumers reported living on the farm; raw milk consumers were more likely (P<0.05) than pasteurized milk consumers to be associated with smaller farms; about 64% of the raw milk consumers were between 21 and 65 years of age and about 16% were less than 10 years old. In terms of their milk consumption habits, most (76.5%) raw milk drinkers indicated that they had been drinking unpasteurized milk for more than 21 years, 2.9% for six to 10 years and 5.9% for less than five years; the 68 raw milk consumers represented 45.3% of survey respondents and they obtained raw milk from the producers’ bulk tank; 68 (45.3%) respondents reported consuming fresh raw milk from the farm; of 68 raw milk drinkers, 33 (50%) obtained milk solely from the farm and 33 (50%) also purchased some commercially processed (e.g., pasteurized) milk from a store. The average quantity of milk consumed per week did not differ much between raw and pasteurized milk households; consumption was 4.1 gallons per week and 3.5 gallons per week, respectively. The primary reasons that 66 raw milk drinkers gave for consuming raw milk included taste (56, or 84.8%), convenience (53, or 80.3%) and cost (38, or 57.6%). About 11% noted other reasons, such as “the family likes it better,” “freshness,” “they ran out of store milk,” “they want the higher fat for butter making,” or that it “was from grass-fed cows.” 39 (29.8%) farms provided raw milk to the community. Concerns related to health hazards associated with raw milk consumption were expressed by 38.2% of the raw milk and 73.2% of the pasteurized milk consumers.
López Osorio et al, 2008 (neutral quality), a cross-sectional study designed to predict the optimum cooking temperatures of beef based on acceptance or rejection using survival analysis statistics. Data from 306 subjects from Argentina, Spain and the US were segmented by age groups (young and middle-aged adults) and stated preference for degree of doneness (rare, medium and well-done). Subjects were asked to look at pictures from the American Meat Science Association (AMSA) Color Guide and decide if these were undercooked, okay or overcooked. Survival analysis statistics were applied to the data to predict optimum internal cooking temperatures. The 95% CI were: 75±6.2°C, 78±4.3°C and 82±2.6°C, for consumers stating a preference for rare, medium and well-done beef, respectively. The 55°C picture of the AMSA Color Guide was rejected as meat undercooked by almost all consumers, including those who stated they preferred ‘‘rare” beef. At the other extreme, the 82°C picture was rejected as meat undercooked by 29% of those consumers who stated they preferred their beef ‘‘well-done,” but not all consumers found the 82°C picture to be overcooked; 65% of those who stated they preferred ‘‘rare” beef found this picture to be overcooked. The middle-aged consumers tended to have lower rejection probability (16%) than the younger consumers (23%) due to the beef being overcooked. US consumers were more likely than Argentinean and Spanish consumers to prefer beef steaks to be cooked rare. Country of residence and age group had little influence on optimum temperatures.
Patil et al, 2005 (neutral quality), a meta-analysis of 20 studies evaluated United States consumers' consumption of raw or undercooked foods, knowledge of proper food safety practices and reported behaviors, based on demographic differences (gender, ethnicity, age, education, geographic region and metropolitan vs. non-metropolitan area). Findings from the studies were combined using meta-analysis methods to estimate percentages of consumers engaging in risky behaviors, such as consumption of raw food, poor hygiene and cross-contamination, separated by various demographic categories. Consumer knowledge of safe handling practices did not correspond with reported use of the practices, suggesting that knowledge is a poor indicator of behavior. Compared with women, men reported greater consumption of raw or undercooked foods (26.7%); mid-age adults consumed more raw food (except milk, 24.7%) than did young adults and seniors; high-income individuals reported greater consumption of raw foods (29%); the highest raw ground beef and egg consumption (29%) were found in the US Mountain region; more people consumed raw or undercooked eggs (47%) than consumed raw or undercooked ground beef (21%), shellfish (12%) and raw milk (2.1%); consumption of raw or undercooked food varied by gender, ethnicity, age, income, education level and region.
Redmond and Griffith, 2003 (neutral quality), a systematic review reviewed 88 food safety studies regarding consumer food handling in the home, published over a 26-year period. The majority of all the studies conducted (55 studies) were between 1995 and 1999. After 1999, in only two years, an additional 26 studies were completed, reflecting an increasing trend in foodborne illness incidence. Seven of 15 observational studies involved direct observations, out of which three (43%) were carried out in the US. Based on US consumer food safety surveys undertaken from 1977 to 2000, large proportions of consumers reported eating raw foods of animal origin. Since 1977, the prevalence of the consumption of undercooked hamburgers has ranged from 4% to 30% of sampled population; since 1997, some surveys have indicated that less than 5% of consumers report preference for and the consumption of medium rare and rare hamburgers. Since 1994, the prevalence of consumption of undercooked or raw eggs has ranged from 5% to 56%; the levels of consumption of such eggs appear to have been consistent from the mid-1990s to present such that up to 50% of consumers may still consume raw and undercooked eggs. One US study indicated that susceptible populations with high risk for foodborne illness continue to consume inadequately cooked runny eggs and pink beef burgers. Authors note that social desirability bias may have had the effect of reducing the prevalence of the consumption of unsafe foods, so that the actual prevalence of these practices may be higher than reported.
Trepka et al, 2007 (neutral quality), a cross-sectional study assessed baseline food safety practices among 299 adult female clients served by an inner city Miami WIC program. A 23-item self-administered questionnaire addressed food safety practices related to cleanliness, separation or avoidance of cross-contamination, proper cooking and chilling methods and avoidance of unsafe foods during pregnancy. The proportion of respondents reporting usually eating undercooked eggs was 24.7%, while 28.4% reported eating undercooked eggs at least some of the time, which was lower than reported in the Centers for Disease Control and Prevention’s (CDC) 1996 Behavioral Risk Factor Surveillance Survey (50%). Over one-half (51.6%) of the 62 pregnant women participants reported eating hot dogs or deli meats without first reheating "sometimes" or more frequently since becoming pregnant, and 35.5% reported eating soft cheeses and blue-veined cheeses "sometimes" or more frequently since becoming pregnant; both practices increasing risk of acquiring listeriosis. A high prevalence of pregnant participants ate foods that put them at risk of listeriosis at least some of the time (over one-half for hot dogs, luncheon meats or deli meats that were not reheated to steaming hot and one-third for soft cheeses, although it was unclear which food item the participants were referring to when they reported eating hot dogs, luncheon meat or deli meats). Only 3.5% of participants reported usually eating pink or under-cooked meat.
View table in new window
Research Design and Implementation Rating Summary
For a summary of the Research Design and Implementation Rating results, click here.
Anderson JB, Shuster TA, Hansen KE, Levy AS, Volk A. A camera's view of consumer food-handling behaviors. J Am Diet Assoc. 2004; 104: 186-191.
Byrd-Bredbenner C, Abbot JM, Wheatley V, Schaffner D, Bruhn C, Blalock L. Risky eating behaviors of young adults—implications for food safety education. J Am Diet Assoc. Mar 2008; 108(3): 549-552.
Dharod JM, Pérez-Escamilla R, Paciello S, Venkitanarayanan K, Bermúdez-Millán A, Damio G. Critical Control Points for Home Prepared 'Chicken and Salad' in Puerto Rican Households. Food Protection Trends 2007; 27: 544-552.
Kaylegian KE, Moag R, Galton DM, Boor KJ. Raw milk consumption beliefs and practices among New York state dairy producers. Food Protection Trends. 2008; 28: 184-191.
López Osornio MM, Hough G, Salvador A, Chambers IV E, McGraw S, Fiszman S. Beef’s optimum internal cooking temperature as seen by consumers from different countries using survival analysis statistics. Food Quality and Preference. 2008 Jan, 19 (1): 12-20.
Patil SR, Cates S, Morales R. Consumer food safety knowledge, practices and demographic differences: Findings from a meta-analysis. J Food Prot. 2005 Sep; 68 (9): 1,884-1,894.
Redmond EC, Griffith CJ. Consumer food handling in the home: A review of food safety studies. J Food Prot. 2003 Jan; 66 (1): 130-161.
Trepka MJ, Newman FL, Dixon Z, Huffman FG. Food safety practices among pregnant women and mothers in the women, infants and children program, Miami, Florida. J Food Prot. 2007; 70: 1,230-1,237.