Though studies on HEV IgG seroprevalence among blood donors [5–7] or epidemiologically exposed subjects  are available from France, there is a lack of data with respect to pregnant women. Indeed, pregnant women are usually followed as outpatients by gynecologists and HEV is still considered as a rare and exotic disease this setting. Analysis of reported seroprevalence rates has also to take into account the use of assays with different sensitivities and the geographical origin of recruited subjects. Indeed, by using more sensitive assays, recent HEV IgG seroprevalence studies conducted in France have produced much higher results than earlier studies . Significant epidemiological differences also exist within the same country: previous studies have reported seroprevalence rates among blood donors five-fold higher in south-west (16.6%)  than in north of France (3.2%), though using the same assay . This finding was confirmed by a higher frequency of clinically apparent HEV infections in the south of France . The present study confirms this north-south gradient in a population of pregnant women, with much higher rates in south-east France. Reasons for this striking difference are still unknown, and the present study failed to identify life style factors that may explain this gradient. Seroprevalence rates reported here may appear lower than those recently reported in other French populations. However, we found HEV IgG positivity associated with age and it is noteworthy from previous studies that the probability of being exposed to HEV increases with age [6, 7]. Our finding of a 29.3% of HEV seroprevalence in south-east France with the Wantai assay is in line with the results from Mansuy et al. who reported a rate over 40% in blood donors from south-west France aged 28 to 37 years .
The high HEV seroprevalence, particularly in south of France, at the onset of the pregnancy in women who report no history of liver disease confirms that most infections are subclinical or unrecognized. Concerning the incidence of HEV infection among pregnant women in western countries, a rate of 0.67% of IgM anti-HEV was shown in a Spanish cohort during the first trimester of pregnancy  but no clinical symptoms and normal aminotransferases levels were reported, suggesting either false-positive results or silent forms of infection. In the present study, no markers of acute infection and no seroconversion were observed during pregnancy. This absence of infection even in an area with a significant circulation of the virus, like south-east of France may be due to the relatively short time span between samplings (6 months) but also to dietary measures taken by pregnant women to prevent infectious diseases such as toxoplasmosis that could have protect them also from HEV. However, the number of questionnaires returned at the end of pregnancy was too small to assess this hypothesis.
Finally, this prospective study assessed for the first time the HEV prevalence and incidence during pregnancy in France and suggests that this infection is a rare occurrence during pregnancy, even in high endemic regions of western countries.