The present study found evidence of the presence of seasonal and pandemic influenza viral RNA in 7.2% of adult patients (≥18 years old) consulting their GP for the typical and uncomplicated symptoms of AD during the ILI and AD outbreaks in France (http://www.sentiweb.fr). We have also reported the detection of enteric viruses in half of the patients who tested positive for influenza viruses. The most frequent combination was a co-detection with two agents, primarily influenza virus B plus NoVGI.
It is to be noted that in our study the most prevalent Influenza virus was influenza virus B, detected in 8 of 10 stool specimens positive for influenza viruses. These results seem to be in agreement with previous studies about the detection of influenza virus B in patients complaining of GI symptoms. The presence of influenza virus B in gastric mucosa has been previously reported among patients with GI symptoms without concurrent respiratory symptoms . Similar results have been reported among hospitalised children infected with the influenza B virus for which abdominal pain was a dominant symptom, especially in older children . As highlighted by Kaji et al. , GI symptoms were significantly more common in adult patients with a positive throat swab for the influenza B virus (GI = 23%), and with respect to the influenza A virus (GI = 6% for A/H3N2 and 4% for A/H1N1). Previously, the influenza B virus has been reported  in 81% (17/21) of influenza positive stools of pediatric patients (<6 years of age) with concurrent respiratory and GI symptoms. Interestingly, one of the influenza virus B strains detected among these pediatric patients was viable .
In this study we have also reported the detection of A/H3N2 and A/H1N1 2009 viral RNA in the stools of two patients with AD. The detection of the A/H3N2 virus in stool samples has been previously reported in six high-risk influenza adult patients  and in three young children  reporting ILI and diarrhea. Seasonal influenza viruses detection by RT-PCR in stools has also been reported in very young children presenting with ILI and AD between the ages of 5 weeks and 9 months . Influenza virus A/H1N1 2009 was recovered from 16 (24.6%) stools of A/H1N1 2009 positive patients who were hospitalised due to the progression of acute gastroenteritis . In another study, the authors showed a positive viral culture for A/H1N1 2009 in the stool of four patients presenting the highest viral load , suggesting the fecal shedding of viable pandemic viruses.
In this study, the overall proportion of co-detection of influenza and enteric viruses was 3.6%. We detected one enteric virus in 5/10 stool specimens of influenza-positive patients. Among them, four tested positive for the influenza B virus and one enteric virus (2 NoVGI, 1 NoVGII, and 1 astrovirus), and one for influenza A/H1N1 2009 (concomitantly with NoVGI). It is to be noted that although our sample was not large enough to make conclusions that are statistically approved, we can observe that patients who tested positive for both influenza and enteric virus were older (64 years [38-80]) than patients showing a single detection of influenza viruses (30 years [24-56]) and those ones positive for enteric viruses only (35 years [28-50]). To our knowledge, until now a co-detection of influenza viruses and enteric pathogens has rarely been reported. Co-infections between rotavirus and influenza viruses (6 influenza B and 1 influenza A) have been previously reported among 2.2% of hospitalised young children with gastroenteritis . One case of co-infection with influenza A/H3N2 virus and norovirus has been reported in an elderly patient who developed diarrhea since day 3 and passed 3–4 episodes of watery/loose stool per day up to day 13 . In the present study, the duration of fever seems to be shorter among patients who tested positive for at least one enteric virus with respect to patients positive for both enteric and influenza viruses. It is difficult to interpret this result given the low number of influenza-enteric co-detections and the low number of the ‘pure’ influenza-positive cases.
Finally, the explanation of the presence of seasonal (A and B) and pandemic A/H1N1 2009 influenza viruses RNA in the stools is not clear. As previously known, the avian influenza virus prefers to bind the α-2, 3-sialic acid receptor, while human Influenza viruses frequently bind the α-2, 6-sialic acid receptor. Recent evidence indicates that both types of receptors are expressed on the surfaces of in vitro differentiated intestinal epithelial cells [25–27], suggesting that both avian and human influenza viruses have the potential to infect and replicate in human intestinal epithelial cells. Recent data confirmed that human intestinal epithelial cells can be infected by the pandemic (H1N1) viruses and H9N2 viruses isolated from both humans and birds . On the other hand, a recent study on adult hospitalised patients showed that a direct intestinal infection by seasonal influenza A viruses seems an unlikely explanation for the fecal detection of viral RNA in the patients reported . Alternative explanations of influenza virus detection in stools could be the swallowing of virus-containing nasopharyngeal secretion or extrapulmonary virus dissemination via hematogeneous circulation.
This study has several limitations. First, the total proportion of viral co-detection was likely underestimated because we did not test other diarrheal pathogens. Thus, some cases of single infection in our study could be classified as multiple infections in studies which would include these other pathogens. Second, influenza virus cultures were not performed. However, to help us evaluate whether PCR signals were false positives, positive and negative controls were included in each PCR performed. The detection of influenza B has been performed by using two different primer pairs for the NS gene, and we detected influenza A by using two independent PCR assays for the detection of M gene and H gene. Third, respiratory samples were not collected. It is to be noted that the enrolment of patients was blind to any type of information related to respiratory tract infection, thus preventing potential bias.