The Importance of Source Attribution

WHO explained that estimating foodborne disease burden is complex because many hazards are not exclusively transmitted through food. Transmission depends on disease epidemiology, the type of hazard, and contextual factors (e.g., geography, seasonality, and local food consumption patterns and preparation practices). Therefore, source attribution (i.e., the proportion of disease attributable to contaminated food and to specific food groups) is critical for estimating the burden of foodborne disease.

Source attribution informs targeted disease prevention and food safety interventions, including resource allocation, to maximize public health benefits. It also enables monitoring of intervention impacts.

Study Details

The estimates of the proportions of disease burden attributable to different sources were based on a structured expert judgment (SEJ) study commissioned by WHO and supervised by the Foodborne Disease Burden Epidemiology Reference Group (FERG) for 2021–2025. According to WHO, the use of SEJ enabled the production of estimates despite gaps in public health surveillance, food consumption and monitoring, and epidemiological data that exist for many countries and hazards.

The study estimated attribution proportions for six major transmission pathways and 14 food categories across all 194 WHO Member States. A total of 29 hazards were included in the source attribution estimates (14 diarrheal, seven invasive enteric, seven parasitic, and one chemical). These hazards were selected from the 42 hazards included in the broader global burden of foodborne disease estimates, excluding those that were already considered 100 percent foodborne and attributable to a single food category.

A total of 146 experts participated in the SEJ, providing 1,436 assessments, with each contributing to one or more hazard–subregion panels.

The study included all hazards in the last attribution study conducted in 2015 to support the previous global foodborne disease burden estimates, plus ten additional hazards, and elicited estimates from a larger, more geographically diverse expert panel.

Results from the study are generally consistent with previously published estimates.

Proportion of Hazards that are Foodborne

Overall, 13 of the 29 hazards were determined to be more than 50 percent foodborne. Among enteric pathogens, Campylobacter (45–71 percent) and non-typhoidal Salmonella (59–74 percent) were predominantly foodborne in nearly all subregions, whereas the proportions of diarrheagenic E. coli (29–80 percent) and Shiga toxin-producing E. coli (STEC) (32–73 percent) that were foodborne varied by region.

Cyclospora was the only enteric protozoan primarily transmitted through food (74–95 percent).

Foodborne transmission accounted for a smaller proportion of rotavirus (1–17 percent), norovirus (19–45 percent), and hepatitis A virus (28–53 percent) compared with human-to human contact.

The parasite Toxoplasma gondii was largely attributed to foodborne transmission in most subregions, but in the African and Eastern Mediterranean regions, soilborne transmission also played an important role (34–38 percent).

Lead exposure was mainly foodborne in the Southeast Asia Region (46–59 percent), whereas attribution to food within the remaining regions was lower (7–30 percent).

In general, across all hazards, subregional differences in the relative contribution of pathways were most pronounced within the European Region and Region of the Americas; whereas in the African, Eastern Mediterranean, Southeast Asia, and Western Pacific regions, attributions were more homogeneous.

Food-Level Attributions

Food-level attribution estimates also varied across hazards and regions but showed consistent rankings for several pathogens. Poultry meat was the primary source of Campylobacter spp. in all subregions (36–74 percent), followed by beef (9–12 percent) or dairy (5–28 percent).

The top sources of Salmonella included poultry meat (22–43 percent), pig meat (8–37 percent), and eggs (15–29 percent).

Beef was the main source of STEC (15–55 percent).

Dairy contributed substantially to Listeria monocytogenes (22–57 percent) and Brucella (35–94 percent, excluding the Western Pacific region).

Vegetables, fruits and nuts, and fresh produce were deemed to be the only food sources of Cyclospora, Giardia, Cryptosporidium, Ascaris, Echinococcus granulosus, and Fasciola and Fasciolopsis. They were important sources of norovirus (approximately 30–40 percent), rotavirus (25-85 percent), hepatitis A virus (approximately 70 percent), Shigella (20–85 percent), diarrheagenic E. coli (8–60 percent), Entamoeba (90–95 percent), and lead (7–50 percent).

Shellfish was the leading food source for norovirus (22–60 percent), and finfish and shellfish for Vibrio cholerae (66–86 percent).

Grains and beans (1–33 percent), and to a lesser extent shellfish (3–20 percent), were relevant sources of lead exposure.

Interestingly, the study presented the first source attribution estimates for Trypanosoma cruzi, relevant to the Americas region. Foodborne transmission through consumption of vegetables and fruits and nuts was determined to be an important contributor to disease burden, suggesting that food safety interventions are important to prevent Chagas disease in high-risk populations.