..at least not in the way the term is generally used
Campylobacter
Campylobacteriosis is not much fun. I’ve had it and quite a lot of people will have a dose every year. Usually, as long as people keep taking fluids, it will go away eventually with no long-term ill effects. For a few people it can produce a nasty long-term disease known as Guillain Barré syndrome, and for fewer people still it can be fatal. Whenever you read descriptions of the disease there will inevitably be reference to something like “as few as 500 cells” is the dose which must be eaten to cause disease. Well, this is sort of true in that it is not wrong, but it doesn’t really tell anything like the whole story, and it implies that you need 500 or more cells as a dose to make you sick. This, then, goes back to the notion of the “minimum infectious dose”, some notional number that will make you sick, but if you eat one fewer then it won’t.
But, if you go back to the original human volunteer experiment, Black et al. used 800 cells as the lowest challenge dose and so there is nothing from that set of data showing that doses fewer than 800 cells would not make people sick. In fact, there is quite a lot of information to suggest that people have become sick from consuming far fewer cells than this. The problem with Campylobacter is that it is not really a hardy beast and so data for the concentration of the organism measured in foods tested some time after the event may be misleading and under-estimate the number of cells which was actually consumed.
Dose response relationship
A more logical view is to consider a dose response model where any number of cells consumed has an associated probability of causing disease and, you’d expect, the more cells consumed the greater the probability of disease. Of course, this doesn’t seem to apply to Campylobacter, but then that organism marches to the beat of its own drum. While the mathematics differ, and are possibly not yet agreed, the relationships are generally S- shaped curves with tails at high and low cell numbers. It is probably fair to say that the curves at low cell numbers have more associated uncertainty as there tends to be a dearth of data for these situations.
A study of data for Salmonella found ID50 (the dose where 50% of consumers become sick) values of the order of 40-50 cells, while the ID1 (the dose where 1% of consumers become sick) was approximately 1-2 cells (Teunis et al. (2010) Dose–response modeling of Salmonella using outbreak data, International Journal of Food Microbiology, 144 (2), 243-249).
There are many other factors which influence dose response such as immunity, biological variability, differences in the health of volunteers, age etc. which will all add to the complexity of defining such a relationship.
Low concentrations in foods
If there is a dose response relationship then there should be some examples of disease outbreaks where very low numbers of bacteria were present in the food. In fact, there are such outbreaks and I produce some data in the table below.
These numbers do have to be considered with the caveat that it is not impossible that the actual portions of foods eaten which caused disease contained a higher concentration of cells than that shown here. This might occur if the organism was not present uniformly throughout the food, and this may well be the case in some of these examples. However, with foods like ice cream there is a lot of mixing involved in its production, and it seems likely that the paprika flavouring in the crisp outbreak was also well mixed.
| Pathogen | Food | Concentration |
| Campylobacter | Raw beef liver | 3.6 MPN/g |
| Escherichia coli O157:H7 | Raw beef liver
Frozen burger patties |
0.04–0.18 CFU/g
1.45 MPN/g |
| Listeria monocytogenes | Hard cheese
Frankfurters Milk shakes |
Approximately 20 CFU/g
0.03 MPN/g 0.12-8.64 CFU/g (mean) |
| Salmonella | Paprika-flavoured crisps
Flour Tahini products Ice cream Herbal tea |
0.04–0.45 CFU/g
0.003–0.02 CFU/g 0.03–0.46 MPN/g <0.003-0.093 MPN/g 0.036 MPN/g |
Of course, people tend to eat more than a gram of food at a time and so it’s worth looking at the numbers associated with one of these outbreaks. In the case of the paprika crisps it was assumed that 100g of the crisps were consumed (which is quite a lot) and so the ingested dose between 4 and 45 cells. It was estimated that 10 million packs were eaten with a resulting 1000 cases of salmonellosis, an attack rate of 1/10,000. The probability of an individual getting sick was therefore quite low, but when lots of packets are eaten then….
There is, of course a minimum infectious dose, and that dose is 1 cell. There is no prospect of suffering from salmonellosis if you don’t consume any Salmonella cells.