Norwalk Virus Infection and Disease Is Associated with ABO Histo–Blood Group Type

To the Editor—Hutson et al. [1] report an association, in voluntary subjects experimentally infected under laboratory conditions, between ABO phenotype and the risk of symptomatic Norwalk virus infection. We have found a similar association in an outbreak of Norwalk-like virus (NLV) infection that affected British troops. During May 2002, a British military field hospital located in Bagram, Afghanistan, experienced an outbreak of incapacitating illness, which eventually proved to be the result of NLV infection [2]. The infection was introduced by an index-case subject and, subsequently, rapidly spread through the medical unit. No virological investigations were available in Afghanistan, but NLV was identified by electron microscopy, polymerase chain reaction, and a novel antigen-detection ELISA, in all 8 of the patients who were evacuated to Europe by aircraft equipped with medical devices. The unit represented a small, closed community, and the movement of personnel in and out of the unit was restricted until the outbreak had been controlled. Thirty-five (45.5%) of 77 hospital staff developed symptoms that fulfilled the clinical-case definition. Because all British soldiers have their blood group recorded on their medical records, we were able to calculate odds ratios for symptomatic infection, in relation to blood group. The relationship between ABO group and symptomatic disease is reported in table 1. Because of both the close living conditions within the field hospital and the high number of symptomatic individuals, which is particularly important since restriction of movement had been invoked, it is reasonable to assume that all personnel were exposed to an infecting dose of NLV during the outbreak. Odds ratios of symptomatic infection in relation to blood groups. Similar to Hutson et al. [1], we found that individuals of blood group B appeared to have reduced susceptibility to symptomatic NLV infection, although statistical significance was not reached. The severity of the disease, ascertained from hospital records and questionnaires, appeared to be heightened in individuals of blood group O. Two individuals suffered from incapacitating illness that appeared to be life threatening during the early stages and required artificial ventilation. Both individuals were of blood group O. Our findings are the results of a reactive-outbreak investigation performed in adverse field conditions. To have been able to perform virological studies on the entire cohort is desirable, but such studies were impractical, given the circumstances. Our observations are of interest, however, suggesting that individuals of blood group B may indeed be less susceptible to symptomatic NLV infection. The implications for operational military forces are unclear.