Comparing Raw Milk Foodborne Illness Data with Other Foods

Raw milk is often described as uniquely dangerous. The usual evidence is a comparison with pasteurized milk, which produces large ratios like 840x. A more useful comparison is with other foods people eat every day, measured the same way: how many consumers get sick each year.

Raw milk: the outbreak-based estimate

Costard et al. 2017, published in Emerging Infectious Diseases, estimated that outbreaks linked to unpasteurized milk and cheese cause about 730 illnesses and 21 hospitalizations a year in the United States. Raw milk drinkers numbered roughly 9.6 million in the survey the study used, and about 11 million adults in FDA surveys from 2016 and 2019.

That puts the outbreak-based per-consumer illness rate at about 0.007%, or 1 in 13,000 to 1 in 15,000 per year, and hospitalization at roughly 1 in 450,000 to 1 in 520,000.

Leafy greens

A 2024 study in the Journal of Food Protection by Yang and Scharff estimated that leafy greens cause as many as 2,307,558 foodborne illnesses a year in the United States (90% confidence interval 1.08 million to 4.08 million). Nearly everyone eats leafy greens, so dividing by the full US population of about 335 million gives a per-consumer rate of about 0.69%, or roughly 1 in 145 per year. Even the bottom of the confidence interval works out to about 0.32%.

The same study found that 19.8% of E. coli O157:H7 illnesses in the United States are associated with romaine lettuce alone. O157:H7 is the strain behind hemolytic uremic syndrome (HUS), the most serious complication cited in raw milk warnings.

Chicken

Chicken is the leading food source of foodborne Salmonella in the United States. CDC estimates that Salmonella causes about 1.3 million domestically acquired foodborne illnesses a year, and the federal Interagency Food Safety Analytics Collaboration (IFSAC) attributes 19.1% of them to chicken, about 250,000 illnesses a year.

Campylobacter adds more. IFSAC no longer publishes Campylobacter attribution, noting that outbreak sources differ considerably from the sources of everyday cases. The most widely cited estimate, a ranking of pathogen-food combinations by Batz and colleagues (published in 2012 in the Journal of Food Protection) that combined outbreak data with expert judgment, put Campylobacter from poultry first among all pairs, at more than 600,000 illnesses a year.

Together that is roughly 850,000 illnesses a year. Spread across nearly the whole population, it comes to about 0.25%, or 1 in 400 per year. Like the leafy greens figure, it is a total-burden estimate that includes undiagnosed and non-outbreak illness.

Making the comparison fair

The leafy greens and chicken figures are total-burden estimates. It counts all illness, including cases never linked to an outbreak and cases never diagnosed. Costard’s raw milk figure is adjusted for underdiagnosis but counts outbreak-related illness only. Comparing 0.69% with 0.007% directly would overstate the gap.

The best available evidence on raw milk’s sporadic (non-outbreak) illness comes from Minnesota. Robinson et al. 2014, in Emerging Infectious Diseases, found that from 2001 to 2010, lab-confirmed sporadic patients who reported drinking raw milk outnumbered raw milk outbreak cases by about 25 to 1.

Applying that ratio nationally, adding 25 sporadic cases for every outbreak case, raises the raw milk estimate from about 730 to about 19,000 illnesses a year, or roughly 0.2% of consumers. That is likely an overstatement: Robinson counted every patient who reported drinking raw milk during the exposure period, and the authors state that these illnesses “could not be definitively linked” to raw milk. Raw milk drinkers also eat lettuce, chicken and everything else.

Raw milk outbreak illness rates, with every source linked, are summarized in raw milk statistics.

On the outbreak measure, raw milk’s per-consumer illness rate is about one-hundredth that of leafy greens. Adding the generous sporadic-illness estimate brings it up to about 0.2%, still below chicken and below even the low end of the leafy greens range. Robinson’s own Minnesota estimate goes further: counting weekly consumers only and applying larger underdiagnosis multipliers, it reached 17.3% of consumers over ten years, about 1.7% a year. That figure rests on the same unconfirmed exposures, and it still sits in the same range as the leafy greens estimate (up to about 1.2% at the top of its interval). No published estimate places raw milk in a risk category of its own.

Pasteurized dairy is not risk-free either

Costard’s study also recorded what pasteurized dairy caused. From 2009 to 2014, 11 outbreaks linked to pasteurized milk and cheese produced 102 illnesses and 87 hospitalizations. Of 17 dairy outbreak deaths in the dataset, 16 were from Listeria, and unpasteurized dairy accounted for only one Listeria illness in the entire period. Pasteurization kills pathogens present before heating, but not contamination introduced afterward during processing and packaging, a recognized route for Listeria in soft cheese.

What these numbers don’t show

• Individual risk. These are population averages across every kind of farm, from licensed and routinely tested dairies to unlicensed operations. They do not predict the risk from a specific source.
• Who is at risk. Children, pregnant women, older adults and people with weakened immune systems face higher risk of severe outcomes from these pathogens, whatever food carries them.
• The current year. Costard’s model uses 2009 to 2014 outbreak data, and the consumer counts are from 2006 to 2019.

None of this means raw milk carries no risk. It means the risk is of the same kind as, and on most available estimates smaller than, risks the public accepts from ordinary foods without calls for prohibition. For the per-serving figure behind the 840x claim, see What “840 Times More Likely to Get Sick from Raw Milk” Actually Measures.