Are dietary reports in a case-control study on thyroid cancer biased by risk perception of Chernobyl fallout?

BACKGROUND

In retrospective case-control studies performed following nuclear tests or nuclear accidents, individual thyroid radiation dose reconstructions are based on fallout and meteorological data from the residential area, demographic characteristics, and lifestyle as well as dietary information. Collecting the latter is a controversial step, as dietary declarations may be affected by the subjects' beliefs about their risk behavior. This report analyses the potential for such bias in a case-control study performed in eastern France.

METHODS

The study included 765 cases of differentiated thyroid carcinoma matched with 831 controls. Risk perceptions and beliefs of cases and controls were compared using Chi² tests and differences in dietary reports were analyzed using a two-way ANOVA.

RESULTS

In general, atmospheric pollution and living near a nuclear power plant were the two major risks that may influence thyroid cancer occurrence cited by cases and controls. When focusing in particular on the consequences of the Chernobyl accident, cases were more likely to think that the consequences were responsible for thyroid cancer occurrence than controls. Vegetable consumption during the two months after the Chernobyl accident was correlated with the status of subjects, but not to their beliefs. Conversely, consumption of fresh dairy products was not correlated with the status or beliefs of subjects.

CONCLUSION

We found no evidence of systematic bias in dietary reports according to the status or beliefs held by subjects about the link between thyroid cancer occurrence and Chernobyl fallout. As such, these dietary reports may be used in further studies involving individual dosimetric reconstructions.

The genetic structure of the plant pathogenic fungus Melampsora larici-populina on its wild host is extensively impacted by host domestication

Wild and cultivated plants represent very different habitats for pathogens, especially when cultivated plants bear qualitative resistance genes. Here, we investigated to what extent the population genetic structure of a plant pathogenic fungus collected on its wild host can be impacted by the deployment of resistant cultivars. We studied one of the main poplar diseases, poplar rust, caused by the fungus Melampsora larici-populina. A thousand and fifty individuals sampled from several locations in France were phenotyped for their virulence profile (ability to infect or not the most deployed resistant cultivar 'Beaupré'), and a subset of these was genotyped using 25 microsatellite markers. Bayesian assignment tests on genetic data clustered the 476 genotyped individuals into three genetic groups. Group 1 gathered most virulent individuals and displayed evidence for selection and drastic demographic changes resulting from breakdown of the poplar cultivar 'Beaupré'. Group 2 comprised individuals corresponding to ancestral populations of M. larici-populina naturally occurring in the native range. Group 3 displayed the hallmarks of strict asexual reproduction, which has never previously been demonstrated in this species. We discuss how poplar cultivation has influenced the spatial and genetic structure of this plant pathogenic fungus, and has led to the spread of virulence alleles (gene swamping) in M. larici-populina populations evolving on the wild host.