
82 Epidemiology of Electromagnetic Fields
Nevertheless, if glioma were not the short-latency and fast-growing disease that is
currently believed, reverse causation might contribute to explain the risk reduction in
regular mobile phone users. Moreover, if prodromal symptoms were a better and earlier
predictor of a subsequent diagnosis of glioma than of meningioma, a temporal bias would
also accommodate the differences between the glioma and meningioma studies in the
amount of risk reduction observed among recent users. In fact, there is limited support to
these hypotheses: in a longitudinal study of the association between previous hospital