This article goes deeper into a spore-forming pathogen that, like Bacillus cereus, depends on time and temperature control, and complements microbiology in the meat industry.
Table of contents
A pathogen that is not eliminated by cooking
Clostridium perfringens is a Gram-positive, anaerobic, spore-forming bacterium that lives in soil and in the intestines of people and animals. Raw meat and poultry usually carry it, and added ingredients contribute too: the USDA-FSIS stabilization guideline cites a survey in which C. perfringens spores were isolated from 80% of 54 spices and herbs analyzed. Its presence in a raw product is not a problem in itself. The problem starts after cooking.
Cooking destroys the vegetative cells of C. perfringens and of almost all the competing flora, including Salmonella and E. coli, but the spores survive. If the dish is then cooled slowly, the spores germinate in a food with almost no competing flora and the cells multiply very rapidly. FSIS indicates that its optimum growth temperature is between 43 °C and 47 °C, with an upper limit close to 54 °C, and that in ground meat at that temperature the generation time can be about 7 minutes. That is why this pathogen is associated less with a contaminated ingredient than with a process failure: a large pot that takes hours to come down in temperature. The same mechanism of spores that survive cooking appears in Bacillus cereus, although with a different toxin and different typical foods.
The CPE enterotoxin: the illness occurs in the intestine
Unlike Staphylococcus aureus, which forms its toxin in the food, C. perfringens causes illness because a large number of live cells is consumed. Some of them survive passage through the stomach and, when they sporulate in the small intestine, release C. perfringens enterotoxin (CPE). Only strains carrying the cpe gene cause the typical food poisoning; that gene can be on the chromosome or on plasmids, as defined in ISO 15213-2. That is why a high C. perfringens count in a food indicates a risk, but does not by itself confirm that the strain is enterotoxigenic.
The clinical picture is characteristic. According to CDC, symptoms appear 6 to 24 hours after eating, are mainly diarrhea and abdominal cramps, and usually last about a day. It is usually mild, but CDC warns that in young children dehydration can progress quickly, and the only death from this agent recorded in the EU in 2024 occurred in a residential home for older adults.
What the data say: few outbreaks, many cases per outbreak
United States
CDC estimates that C. perfringens causes nearly 900,000 foodborne illnesses each year in the United States. Outbreaks cluster in November and December, linked to holiday meals, and in hospitals, school cafeterias, prisons, nursing homes and catered events, where keeping large volumes at the right temperature is difficult. The foods most often implicated are poultry, beef and pork and meat gravies.
A CDC analysis published by Grass, Gould and Mahon (2013) reviewed confirmed outbreaks between 1998 and 2010: 289 outbreaks, 15,208 illnesses, 83 hospitalizations and 8 deaths, with a median of 24 illnesses per outbreak. Among outbreaks attributed to a single food type, beef was the most common (46%), followed by poultry (30%) and pork (16%). Restaurants were the most frequent place of preparation (43%), followed by caterers (19%), private homes (16%) and prisons (11%). The most commonly reported contributing factors were leaving food at room temperature for several hours (39% of outbreaks), slow cooling (38%) and inadequate hot holding (32%).
European Union
The EFSA and ECDC One Health report recorded in 2024 a total of 65 outbreaks due to C. perfringens toxins in 10 Member States, with 2,104 cases, 9 hospitalizations and 1 death. The mean size was 32.4 cases per outbreak, compared with 8.2 in Salmonella outbreaks. Four very large outbreaks accounted for 538 cases: two in residential institutions (prison or boarding school), one in a school canteen and one at a mass event with temporary catering. In the 24 strong-evidence outbreaks, the foods were mainly meat and meat products and mixed foods, and the report identifies time and temperature abuse during storage as the main contributing factor for this agent. The 2024 figure is 53.6% lower than in 2023, although EFSA asks for caution in interpreting the data on bacterial toxins from France, which vary greatly from year to year.
The critical point: cooling
A study by CDC's EHS-Net network (Schaffner and colleagues, 2015) recalls that between 1998 and 2008 improper cooling contributed to 504 restaurant-associated outbreaks in the United States and that, between 1998 and 2002, C. perfringens caused 50 of the 102 outbreaks with a known agent linked to that failure. When measuring how restaurants cool, they found that only 65% of observations met the 2009 Food Code criterion. Large pieces of meat and stews were what cooled most slowly. Foods more than 7.6 cm deep were twice as likely to cool more slowly than required; those that nobody monitored, more than twice as likely, and those cooled without ventilation, almost twice as likely.
Foodservice: FDA Food Code 2022
The Food Code, the model FDA offers for adoption by U.S. jurisdictions, sets in section 3-501.14 a two-stage cooling process for cooked foods that require time and temperature control:
- First stage: from 57 °C (135 °F) to 21 °C (70 °F) within 2 hours or less.
- Total: from 57 °C to 5 °C (41 °F) or less within a maximum of 6 hours.
- Ingredients at ambient temperature (for example, reconstituted foods): to 5 °C within 4 hours.
Section 3-501.15 lists the accepted methods: shallow pans, smaller or thinner portions, rapid cooling equipment, an ice water bath with stirring, containers that facilitate heat transfer and ice as an ingredient, with containers uncovered or loosely covered. It is complemented by hot holding at 57 °C or above (3-501.16) and by reheating for hot holding to at least 74 °C (165 °F) for 15 seconds, within no more than 2 hours (3-403.11).
Meat industry: FSIS Appendix B
For cooked meat and poultry products, USDA-FSIS regulations require a performance standard: no growth of C. botulinum and no more than 1 log multiplication of C. perfringens during stabilization (9 CFR 318.17 and 381.150). The revised guideline of December 2021 (FSIS-GD-2021-13, the so-called Appendix B) proposes cooling options considered sufficient to meet it:
- Option 1.1: from 54.4 °C (130 °F) to 26.7 °C (80 °F) in 1.5 hours or less, and from 26.7 °C to 4.4 °C (40 °F) in 5 hours or less.
- Option 1.2: start cooling within 90 minutes after cooking, go from 48.9 °C (120 °F) to 26.7 °C in 1 hour or less and from 26.7 °C to 12.8 °C (55 °F) in 5 hours or less, continuing down to 4.4 °C.
- Option 1.3, for products cured with at least 100 ppm sodium nitrite and 250 ppm ascorbate or erythorbate: from 54.4 °C to 26.7 °C in 5 hours and from 26.7 °C to 7.2 °C (45 °F) in 10 hours.
Since the growth optimum (43 to 47 °C) falls within the first segment, between 54.4 °C and 26.7 °C, that is the one most worth recording in detail. The guideline also notes that salt, pH, water activity and antimicrobials change the margin: for example, the limiting water activity for C. perfringens growth is 0.93. Companies may use other parameters if they validate them; predictive microbiology is one of the tools for estimating how much the pathogen grows with a real cooling curve. Within a HACCP plan, cooling can be treated as a critical control point with time and temperature limits.
How it is detected in the laboratory
ISO 15213-2:2023, successor to ISO 7937
The horizontal reference method for enumerating C. perfringens is ISO 15213-2, published in November 2023, which cancels and replaces ISO 7937:2004. It is a plate count: the sample or its dilutions are pour-plated in tryptose sulfite cycloserine agar (TSC, formerly called SC) with an overlay of the same medium, and incubated anaerobically at 37 °C for 20 hours. Typical colonies are black or grey to yellowish-brown due to the formation of iron sulfide. If only spores are to be counted, a heat treatment of 10 minutes at 80 °C is applied before plating, which is now optional. The method was validated in ready-to-eat or ready-to-reheat meat products, composite foods, eggs and egg products, processed fruits and vegetables, infant formula and cereals, animal feed and environmental samples. Its practical counting limit is more than 100 CFU/g in solids with a single plate (see what CFU are).
The series is completed by ISO 15213-1, for sulfite-reducing clostridia in general, and ISO/TS 15213-3, for the detection (presence/absence) of C. perfringens.
Confirmation
A black colony on TSC is presumptive C. perfringens: other sulfite-reducing clostridia also blacken the medium. The standard defines as confirmed the isolate that, in addition, possesses the acid phosphatase enzyme, and also describes the test on sulfide indole motility (SIM) agar. The result is expressed as CFU of confirmed C. perfringens per gram. As an option, Annex D includes molecular methods to differentiate strains pathogenic to humans (carrying the cpe gene) from non-pathogenic ones. The general logic of this step is in presumptive vs. confirmed result.
When an outbreak is considered confirmed
CDC accepts three criteria: isolating at least 10⁶ (one million) organisms per gram in the stool of two or more ill persons, demonstrating the enterotoxin in the stool of two or more ill persons, or isolating at least 10⁵ (one hundred thousand) organisms per gram in the implicated food. The count criteria are valid only if the sample was handled correctly, so transport and storage are part of the result. FSIS, for its part, asks that in cooling challenge studies the final measurement of bacterial load include both spores and vegetative cells.
What TAAG can contribute against this pathogen
TAAG's current catalog does not include a kit for Clostridium perfringens, for sulfite-reducing clostridia or for detection of the cpe gene. To enumerate this pathogen, the method to use is still ISO 15213-2 or another alternative method validated for the matrix. Specio 00.1 Bacteria (V-SF14) detects more than 80 spoilage bacteria in beverages, dairy, sauces and surfaces, but it is aimed at spoilage and does not declare C. perfringens among its targets, so it does not replace it.
Where TAAG does have scope is in other hazards that share the same operation of central kitchens and cooked ready-to-eat products. Specio 1.2 S. aureus (V-SF39, real-time PCR with melting curve) declares food, beverages and living and inert surfaces, and serves to monitor the pathogen associated with handling. Specio 1.3 E. coli (V-SF167) covers the same matrices and provides a hygiene indicator. For recontamination of cooked products, Ampliora 2.3 Listeria spp. and L. monocytogenes (V-SF46) declares food, beverages and surfaces, with a focus on meat and ready-to-eat foods (see RTE foods). None of them replaces the control of cooling curves, which is the main measure against C. perfringens.
Conclusion
Clostridium perfringens is not controlled by more intense cooking, because its spores withstand it, but by what happens afterwards: how long the food takes to pass through the range between 54 °C and room temperature. CDC and EFSA data agree on the profile of its outbreaks: meats and dishes prepared in large volumes, served in cafeterias, nursing homes, prisons and events, with many cases per outbreak. The FDA Food Code and FSIS Appendix B translate that risk into specific cooling curves that should be measured and recorded. In the laboratory, ISO 15213-2 makes it possible to quantify the pathogen and confirm the isolates, and the interpretation of a count depends on the sample having arrived in good condition.
About TAAG
Consult TAAG on how to complement cooling control with monitoring of hygiene and other pathogens, for example with Specio 1.2 S. aureus or Specio 1.3 E. coli.
Frequently asked questions
Why does Clostridium perfringens survive cooking?
Because it forms heat-resistant spores. Cooking kills the vegetative cells, but the spores survive, are activated by the heat and germinate if the food cools slowly. With no competing flora left, the cells multiply very rapidly between about 54 °C and room temperature.
How much time do I have to cool a cooked dish in a foodservice operation?
The FDA Food Code 2022 requires going from 57 °C to 21 °C in 2 hours or less and reaching 5 °C or less within a total of 6 hours. To achieve this it recommends shallow pans, small portions, ice baths or rapid cooling equipment.
What is the difference between the Food Code and FSIS Appendix B?
The Food Code applies to restaurants, cafeterias and foodservice operations. Appendix B is a USDA-FSIS guideline for plants producing cooked meat and poultry products, with options that limit C. perfringens growth to 1 log or less; option 1.1 requires going from 54.4 °C to 26.7 °C in 1.5 hours and then to 4.4 °C in 5 hours.
Which standard is used to enumerate Clostridium perfringens in food?
ISO 15213-2:2023, which replaced ISO 7937:2004. It is a count on TSC agar with anaerobic incubation at 37 °C for 20 hours and confirmation of typical colonies with the acid phosphatase test.
Does a high C. perfringens count prove that the food caused an outbreak?
Not by itself. CDC considers an outbreak confirmed if at least 10⁵ organisms per gram are isolated from the implicated food or 10⁶ per gram from the stool of two or more ill persons, or if the enterotoxin is detected in stool. In addition, only strains carrying the cpe gene cause the poisoning.