Search

Listeria in food - is zero tolerance necessary?

Listeria monocytogenes is widespread in the environment and has frequently been found on meat, fish and poultry and their products. It is also found on vegetables and it is particularly prevalent on decaying vegetation and grasses. Food processing plants often harbour the bacteria in cool damp areas. In spite of this widespread occurrence, the incidence of disease in humans from its ingestion with food is relatively rare. There are fewer than three cases per million population in the UK each year and the number of reported cases in Australia is about the same. However, when disease does occur the consequences may be very serious.

Infection of humans by Listeria monocytogenes can result in listeriosis, a disease whose symptoms include septicemia, meningitis and abortion. Cells of L. monocytogenes, ingested with food, cross the barrier of the intestinal tract and are then engulfed by white blood cells and transported to different parts of the body. L. monocytogenes is able to grow and multiply within cells and spread directly from cell to cell. Strains differ in their pathogenicity.

To bring about a systemic infection, pathogenic strains must be able to overcome the twin defences of the intestinal barrier and the protection from specific immune responses within the body. Experiments with inoculated mice have shown the importance of both defence systems; and that high doses are needed to establish disease when given orally. These results with mice are consistent with dose-response relationships derived from estimates of human exposure to L. monocytogenes ingested with food.

People with an altered or impaired immune system, such as those who are very young, very old, or who are pregnant, are most susceptible. In recent years, the proportion of cases associated with pregnancy seems to have fallen in some countries although that for other cases, mainly in the instance of immunocompromised adults, has risen. Outbreaks of the disease have been linked to the consumption of a number of contaminated foods including coleslaw, soft ripened cheese, paté, jellied pork tongue, smoked fish, and smoked mussels. L. monocytogenes is relatively easily destroyed by heat so the risk of infection is eliminated in fully cooked food.

Other sources of human infection by listeriosis exist, including infection from the mothers of newly born infants during birth.

Setting standards
Public health authorities have recognised for some time that contaminated food is a source of listeriosis but the low incidence of disease made it difficult to set appropriate standards.

US authorities applied the strictest criterion on ready-to-eat foods, requiring them to be recalled if L. monocytogenes was detected - this is in effect a "zero tolerance". Other countries, including the UK, either had no formal microbiological standards with regard to this microorganism in most ready-to-eat foods, or did not have a "zero tolerance" criterion. Instead they undertook a monitoring role to detect problem foods. They also required industry to improve handling and storage conditions and advised those consumers at greatest risk - pregnant women and the immunocompromised - not to consume foods which were considered 'hazardous'.

Which foods pose the greatest threat?
The foods most often implicated in both sporadic and outbreak cases are those that
  • are ready-to-eat and are not heated directly before consumption;
  • have been stored for some time under chilled conditions; and
  • are the kinds of food in which L. monocytogenes can grow to large numbers.

These types of food constitute the greatest hazard and should be targeted for special attention.

It is the ability of L. monocytogenes to grow at common refrigeration temperatures that has been of particular concern. However between 0°C and 5°C, the growth rate of L. monocytogenes is very sensitive to small changes in temperature, pH, aw and nitrite and considerably different amounts of growth are possible on apparently "similar" products.

How much is too much? Setting realistic levels
In spite of efforts to establish a "zero tolerance", surveys in different countries indicate ready-to-eat products are still contaminated. The actual criterion for a "zero tolerance" is set by the type of detection method used. For example, if the criterion is based on not detecting L. monocytogenes in one 25g sample, the criterion is <0.04 cells/gram. Meeting the "zero tolerance" does not mean that the food is "free of the pathogen". Furthermore, in countries without a "zero tolerance", the incidence of human listeriosis is similar to that in countries where a "zero tolerance" has been established.

Exposure to L. monocytogenes is more common than is often realised. It has been estimated that, in the late 1980s in the US, about 10 colony-forming-units (cfu) were consumed in ready-to-eat food at least 10 to 100 times per capita every year. Estimations for Germany suggest that, on average, every person is exposed to more than 1000 cfu of L. monocytogenes in food about 24 times a year. Examination of foods implicated in both sporadic cases and outbreaks have typically had more than 1000 cfu of L. monocytogenes/g. Sometimes the number of L. monocytogenes exceeded 1-10 million/g. There are examples where ready-to-eat foods sampled at retail outlets carried 1000 and more cfu of L. monocytogenes/g without evidence of them causing human infection.

The relationship between infection and the number of L. monocytogenes cells ingested has been calculated from the number of listeriosis cases in Germany and from the measured incidence and numbers of L. monocytogenes in smoked fish. Conservative estimates indicated that all cases, including the third which occurred at or near birth in neonates, were due to the consumption of smoked fish alone, and that 20% of the population was at increased risk because of an impaired immune system.

It appeared from this analysis that, even when over 500,000 cfu of L. monocytogenes were ingested in a meal, only one case of listeriosis could be expected to result from the ingestion of close to 17,000 such meals. From a serving of smoked fish containing 100 cfu of L. monocytogenes/g the chance of a susceptible person acquiring listeriosis was calculated to be <1 in 1,000,000. The probability of the ingestion of a single viable L. monocytogenes cell causing infection was estimated to be about one in ten thousand million. The probability of even a high risk person suffering listeriosis appears, therefore, to be extremely low unless large numbers of L. monocytogenes are ingested.

A concentration of not more than 100 cells of L. monocytogenes/g at the time of consumption is of low risk to consumers, but general advice is that foods with more than this number (100/g) should be rejected. For foods in which growth can occur between the time of sampling and consumption, lower levels may be set. In order to set such a low cut-off level, knowledge of the growth characteristics of the micro-organism in the food under the normal storage and distribution conditions is needed. Although predictive models for the growth of L. monocytogenes can be helpful these tend to yield "worst case" scenarios where the amount of growth may be over-estimated because:
  • factors which reduce growth in some foods may not be included in the models
  • little information is available for storage at temperatures below 4°C
  • listeria growth may cease soon after the other microflora present reaches its maximum.
A criterion of not more than 100 cells of L. monocytogenes could, more appropriately, be applied only to foods that are to be eaten without a prior listericidal treatment, such as cooking, and if they are:
  • foods which have been implicated in listeriosis, or
  • foods in which L. monocytogenes can grow to large numbers, or
  • foods where considerable contamination is suspected.

Summary
There is little evidence that consumption of foods containing small numbers of L. monocytogenes (<100 cfu/g) causes listeriosis. Microbiological criteria should only be applied to a food when there is evidence of a hazard to health and where the establishment of the criterion will reduce the health risk. Foods containing <100 cfu/g of L. monocytogenes and in which the micro-organism does not grow, do not present a hazard to health. The establishment of a "zero tolerance" will lead to the rejection of food without concomitant reduction in human illness.

Good Hygienic Practices and HACCP are essential in producing a safe food and education and training can help producers and consumers apply safe handling practices. Inspection may provide evidence that the operations and practices used can consistently give a safe food while microbial testing can be used to indicate that Good Hygienic Practices and HACCP have been effectively applied. Public health strategies often include guidelines to highly susceptible groups on the selection of suitable types of food and on its safe handling.

Providing consumers at greatest risk with advice on how to avoid eating certain hazardous foods is more likely to protect consumers from infection than the application of bacterial standards for L. monocytogenes. Sampling and testing foods cannot guarantee their safety.

For further information contact:
Dr Patricia Desmarchelier
Brisbane Laboratory
Food Science Australia
Telephone 07 3214 2000

Food Safety and Hygiene
Prepared by Keith Richardson and Beverley George
Food Science Australia
PO Box 52, North Ryde 1670. Tel +61 2 9490 8397 Fax +61 2 9490 8499
Email enquiries@csiro.au