Dioxins and polychlorinated biphenyls (PCBs) belong to persistent organic pollutants (POPs) listed in the Stockholm Convention. Dioxins are a group of 75 polychlorinated dibenzo-p-dioxins (PCDD) and 135 polychlorinated dibenzofurans (PCDF) which are unintentionally produced during incomplete combustion or manufacturing of chlorinated substances and pesticides and can be found in hazardous waste. PCBs refer to a large family of 209 congeners. Unlike dioxins, PCBs were intentionally used in capacitors, transformers, paint or in cement additives for their non-flammability, stability, and high dielectric constants properties until 1977 in the USA and internationally banned in 2004. The half-lives of PCBs in humans ranges from a few months for the congener PCB-77 to 22–42 years for PCB-189. Dioxins exhibit the same variability in persistence with a median half-life of 7 years for 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) or more than a decade for Hexachlorodibenzo-p-dioxin (HxCDD) (Milbrath et al., 2009). This resistance to degradation led to a biomagnification in food chains and an ubiquitous contamination (Kelly et al., 2007). Consequently, the main exposure route in the general population comes from high fat food products.

TCDD and PCBs are classified as carcinogenic to humans by the International Agency for Research on Cancer (IARC). Most evidence for TCDD was established for all-cancers combined from occupational cohorts and rat studies, through a mechanism involving the Aryl hydrocarbon Receptor (AhR) (IARC, 1997). PCBs cause malignant melanoma and positive associations were found for non-Hodgkin lymphoma and breast cancer in human studies (IARC, 2013). A group of dioxin-like PCBs (DL-PCBs) also share the common toxicological mechanism as dioxins, activating AhR whereas non-dioxin-like PCBs (NDL-PCBs) can link to various receptors such as constitutive androstane receptor (CAR) and pregnane X receptor (PXR) involving metabolic dysfunctions or glucocorticoid receptor (Kafafi et al., 1993; JECFA, 2016). Thus, several PCBs also have endocrine disrupting properties which may impact hormone-dependent cancers, obesity and cardiovascular diseases (Diamanti-Kandarakis et al., 2009). Dioxins-like compounds are endocrine disrupting chemicals which may not be characterized by a monotonic dose-response, indicating that effects at high doses may not be extrapolated at low doses (Birnbaum, 2012; Vandenberg et al., 2012).

There is epidemiological evidence for increased all-cause and cancer mortality related to exposures to dioxins or PCBs in one occupational cohorts (Kogevinas et al., 1997; Manuwald et al., 2012; Wang et al., 2013; Yi et al., 2014) and accidental mass poisoning (Consonni et al., 2008; Warner et al., 2011; Li et al., 2015; Onozuka et al., 2020). A limitation of generalizing observations from occupational cohorts or populations exposed due to accidental poisoning is that the exposure levels are much higher than those to which the general population is exposed. As an example, following the Yusho incident in 1968 in Japan, which led to massive contamination of rice oil with dioxins and PCB, average total PCB/dioxins concentrations in blood sampled in 1995 were 3–14 times higher than those of controls from the general population (Masuda et al., 1998). The World Health Organization stated that occupational exposure of workers was “1-3 orders of magnitude higher than the blood levels measured in the general population” (WHO, 1998).

In the general population, studies are scarce and reported inconsistent relationships between non-occupational exposure to dioxins or PCBs and all-cause mortality (Lin et al., 2012; Fry and Power, 2017; Fiolet et al., 2021). Only one Swedish cohort assessed dietary exposure to PCBs, with a positive association for cardiovascular mortality and no association for cancer mortality in the general population (Donat-Vargas et al., 2020).

The main objective of the present study was to investigate the association between dietary exposure to dioxins, DL-PCBs and NDL-PCBs and all-cause mortality, cancer-specific and cardiovascular-specific mortality in a prospective cohort including 451,390 adults with 46,627 deaths, based in nine European countries.