Abstract
Illness associated with Bacillus cereus may be underreported as very few of those affected seek medical attention owing to the mild nature and short duration of symptoms. For this reason there is little information on the prevalence and concentration of this pathogen in retail food products. A total of 1263 retail food samples were examined for B. cereus using the Australian Standard 1766.2.6 (1991): spread plate technique on polymyxin pyruvate egg yolk mannitol bromothymol blue agar, of which the limit of detection was log10 2.0 cfu/g. Bacillus cereus was not detected in samples of skim milk powder, sandwiches, sushi, fresh beef mince, tortillas, or shelf stable stir-fry sauces. Bacillus cereus was detected in the following food samples: uncooked pizza bases (1 of 63 samples, log10 count of 2.0 cfu/g), cooked pizzas (8 of 175, mean log10 3.4 cfu/g), cooked meat pies (7 of 157, mean log10 2.2 cfu/g), cooked sausage rolls (5 of 153, mean log10 2.6 cfu/g), processed meats (1 of 350, log10 3.3 cfu/g), and raw diced chicken (3 of 55, mean log10 4.3 cfu/g). It appears that composite food products have more positive detection samples because the numerous ingredients may introduce spores into the foods. This study provides valuable data on the distribution, prevalence, and concentration of B. cereus in selected retail products.
Introduction
The Australia New Zealand Food Standards Code Standard 3.2.2 (Food Standards Australia New Zealand, 2009) includes provisions for a two-step process of cooling foods: cool from 60°C to 21°C within 2 hours, following by cooling from 21°C to 5°C in a further 4 hours. There is potential for B. cereus to occur in re-heated foods if this cooling requirement is not followed, as re-heating temperatures are usually not high enough to destroy germination of spore-forming bacteria. The time taken for cooling and re-heating thus provides an opportunity for bacteria to increase in numbers. Composite food products may be at a higher risk of B. cereus contamination due to the use of numerous ingredients through which spores may be introduced. The objective of this study was to investigate the distribution, prevalence, and concentration of B. cereus in a broad range of retail ready-to-eat/ready-to-assemble, ready-to-re-heat, and raw foods. Foods for which temperature abuse may be a risk factor for foodborne illness, such as filled savory baked products, were selected to increase the relevance of the study, as were composite food products and foods for which the B. cereus contamination status is not known.
Materials and Methods
Sampling
A total of 1263 retail food samples were obtained from Brisbane (Queensland, Australia) retailers between February 2006 and February 2007. Random samples were drawn, based on availability within each store. No more than 10 samples were taken from each store. These samples were divided into three categories: ready-to-eat/ready-to-assemble foods, ready-to-re-heat foods, and raw foods. Ready-to-eat/ready-to-assemble foods comprised 175 pizzas, 350 cooked, cured, chilled, processed meats, 70 sushi samples, 20 sandwiches, 70 powdered skim milk samples, and 70 tortilla shells. Ready-to-re-heat foods comprised 157 frozen cooked meat pies, 153 frozen cooked sausage rolls, and 50 shelf stable stir-fry sauces. The raw food category comprised 55 chilled diced chicken samples, 63 chilled unbaked pizza bases, and 30 chilled beef mince samples. Surface temperature measurements were taken at point of sale for chilled products. The temperature of pizzas was not taken as these were made to order and sampled within half an hour of baking. Pizzas were transferred from the original container into a large sterile stomacher bag. All samples were placed on ice and received within 4 hours into the National Association of Testing Authorities (NATA) accredited laboratory of EML (a private testing laboratory) in Queensland. NATA is Australian Government–endorsed provider of accreditation. Frozen products were received frozen. All chilled and cooked samples that were received into the laboratory above 7°C were discarded. The time and temperature combination of 4 hours at 7°C was chosen based on the demonstrated inability of food-poisoning strains within the broader B. cereus group (including psychro-tolerant B. cereus strains) to grow at and below this temperature (Guinebretière et al., 2008).
Microbiological analysis
Once received into the laboratory a 10 g sample was removed from each product following the method outlined in Australian Standard 1766.2.6 (Anonymous, 1991), blended with 90 mL of 0.1% sterile peptone solution (Amyl Media, Melbourne, Australia), and stomached for 1 minute. Frozen samples were thawed overnight at 4°C before being sampled. Appropriate dilutions were plated onto polymyxin egg yolk mannitol bromothymol blue agar, and analysis for the presence of B. cereus was performed according to AS1766.2.6 (Anonymous, 1991). This standard required incubation of the polymyxin egg yolk mannitol bromothymol blue agar plates at 37°C for 24 hours followed by 25°C for a further 24 hours, and the presence of B. cereus was then confirmed using microscopy. This technique does not employ the use of a heat treatment and as such the reported result includes both spores and vegetative cells. The limit of detection was 2.0 log10 cfu/g.
Results and Discussion
Results of the microbiological testing of all samples are summarized in Table 1. No relationship was apparent between the 10 samples collected from each retail establishment. The category of food of most concern to human health was the ready-to-eat/ready-to-assemble foods. This category included sushi and processed meats that do not undergo heating processes designed to reduce microbial risk. Ready-to-re-heat foods, such as frozen cooked pies, may be relatively safer since consumers re-heat them. It should be noted, however, that re-heating by consumers is not a controlled process and cannot be relied upon to provide significant protection against B. cereus spores. A recent outbreak of Salmonella infection in the United States was associated with frozen chicken pot pies that appeared to be cooked but in fact contained raw ingredients and required thorough cooking before consumption (CDC, 2008). Frozen meat products in our survey were identified as cooked by packaging and re-heating instructions. The raw foods tested in our study should be subject to cooking in all cases. Bacillus cereus was absent in ready-to-eat sushi samples, which was surprising since this pathogen has been commonly found in rice (Andersson et al., 1995) and sushi (Adams et al., 1994; Chen et al., 2004) in other studies. Acetic acid (vinegar), a common component of sushi, can inhibit toxin production in this product (Jenson and Moir, 2003). A previous survey of sushi, however, deduced that if this product is of moderately low pH (but > 4.5) the presence of vinegar is not necessarily protective against B. cereus intoxication (Millard and Rockliff, 2003).
Limit of detection is 2.0 log10 cfu/g.
Means were only calculated for those samples with detectable Bacillus cereus.
Chilled raw diced chicken had the highest prevalence and mean count of B. cereus in this survey at 5.5% and 4.3 log cfu/g, respectively. Bacillus cereus has been detected in poultry feed (Arotupin et al., 2007), and in raw poultry meat (Abostate et al., 2006). If B. cereus spores are present in raw foods they will only pose a food safety risk if they survive the cooking process and the food is subsequently temperature abused for a sufficient time.
This study provides basic data on the distribution, prevalence, and levels of B. cereus in selected retail products. Overall, 98% of the retail samples tested were found to have counts of this pathogen < 2.0 log10 cfu/g, placing them in the satisfactory category for B. cereus in ready-to-eat foods (Food Standards Australia New Zealand, 2001). The mean count of B. cereus from samples containing > 2.0 log10 cfu/g was 3.1 log10 cfu/g. The minimum number of B. cereus cells required to cause illness has been estimated to be > log 5.0 log cfu/g (Hobbs and Gilbert, 1974), so it is unlikely that the products examined in our study are of public health concern. However, storage of food at inappropriate temperatures may allow the growth of B. cereus, which, depending on the strain and the type of food, may result in toxin production. Although there is no evidence suggesting that consumer abuse of foods during transport between retail and home contributes to B. cereus–associated illness, it would be beneficial to profile the time and temperature of frozen and chilled products from the point of sale through to domestic refrigeration and consumption.
Conclusions
The low prevalence and concentration of B. cereus in the foods examined in this survey suggests a low risk to public health. This information is of particular relevance to those risk managers in Australia gathering data on foods for which the B. cereus status is unknown.
Footnotes
Disclosure Statement
No competing financial interests exist.