Some Discussions at NYS Assn Food Protection Conference
An excellent session at the NY State Association for Food Protection (NYSAFP) conference in Saratoga Springs this week featured two university professors (Don Schaffner, Rutgers, and Ben Chapman, NC State) who spoke about their podcast 'Risky or Not?'.
One issue they addressed at length was dose-response (DR) relationships between exposure to a single pathogen cell (but generally a series of higher doses) by ingestion and the likelihood of transmission of disease (illness outcome). Here are some points from the discussion:
The concept of 'minimum infective dose' is, in Don Schaffner's words, 'BS'. For risk analysis, one requirement is data on DR curves or families of curves (Coleman et al., 2018). Of course, whether those DR curves are from animals, human volunteers, or outbreaks, uncertainties must be addressed for extrapolation to new scenarios of exposure (Coleman, 2025).
Question: How likely is illness from one Listeria monocytogenes (Lm) cell {for healthy immunocompetent people}?
The single response from the audience was 1/10^14: 1 chance in 10 raised to the 14th power exposures or servings, 1 illness in 100,000,000,000,000 servings, one illness in 100 trillion servings; negligible risk by many criteria.Granted, Lm has caused serious illness and death under certain conditions.
CDC reported fatalities for deli meats, pasteurized ice cream/milkshakes, pasteurized soft cheeses (CDC, 2005-2020), and 2 fatalities associated with NY state raw cheese (CDC, 2017).
No outbreaks or illnesses were confirmed for fluid raw milk (CDC, 2005-2020; Stephenson et al., 2024)Lm levels associated with fatalities for hospitalized immunocompromised people in the ice cream outbreak were a million cells or more; no illness was observed in the general population (children, adults, elderly) exposed to approximately a billion Lm cells (Pouillot et al., 2016).
Question: How likely is illness from one Salmonella cell?
Responses were 1/100 and 1/10^6 (one in a million), depending on the dataset, serotype, and assumptions (Coleman et al., 2017; Oscar, 2021; Teunis et al., 2022). For one human clinical isolate administered to human volunteers, more than a billion Salmonella cells did not cause illness (Coleman and Marks, 1998; Coleman and Marks, 2017).Don Shaffner asked how many Lm are acceptable in foods that do not support Lm growth north of the NY state border, in Canada?
The response was 100 Lm per mL or gram food, whereas US imposes zero tolerance for any detectable Lm, despite very low infectivity.
Note that few people acknowledge that raw fluid milk is a food that suppresses Lm growth within the typical shelf life period, likely due to competition with the dense and diverse natural milk microbiota (Coleman et al., 2023). Yet the US recalls ready-to-eat foods when Lm is detected, without enumeration, even though 100 Lm/mL in foods not supporting growth is extremely unlikely to cause illness in either the general or more susceptible immunocompromised populations, as noted above.
Risky or Not? People make these decisions every day, sometimes using scientific evidence. Sometimes culture, ideology, politics, and economics influence decisions.
For NY state, only four illnesses were associated with raw milk over the past 10 years. For those who missed previous postings of this figure below, blue bars represent numbers of licenses NY State issued for raw milk dairies from 2005-2022 on the left axis. The orange symbols represent 0, 1, 2, or 3 outbreaks per year in this same time period on the right axis. Despite increasing access, rates of illnesses have not increaed. Neither are trends increasing in any other US state (Stephenson et al., 2024).
