Circannual rhythm of measles subsequent to immunisation

This is an epidemiological study of biennial and annual rhythms of measles in England and Wales between 1959-1994 highlighting changes following immunisation. The study describes annual rhythms starting in autumn (Week 41) and the statistical method tests the fit of a sine curve to annual data. Before immunisation the previously established biennial rhythm was seen: the two years differed in magnitude of seasonal variation, position of the peak, "visibility of school holidays" and peak breadth. After immunisation (1968) these biennial features tended to disappear, one annual rhythm becoming the predominant feature. By the late eighties adjacent years had very similar sine curve fits. About 1990 (89-91) the fit of the sine curve was no longer significant, was of low amplitude and numbers were at an all time low. In 1992 significant annual variation returned, numbers rose and magnitude of seasonal variation increased. At this time a new epidemic was being predicted on other evidence and this renewed sine curve fit may be an additional warning signal. A possible influence of normal birth rhythm on numbers of susceptibles is described.