Epidemiologic investigation of the association between herpes zoster and multiple sclerosis.

The apparent association between MS and herpes zoster reported in the literature was investigated in a community-based epidemiologic study. Five hundred ninety residents of Rochester, Minnesota, diagnosed with herpes zoster from 1945 through 1959 were followed 9,389 person-years for onset of MS. No cases were observed, whereas 0.2 case was expected, using Rochester incidence rates for MS. Calculation of the power of this study indicated that approximately 200,000 person-years of follow-up of herpes zoster patients is necessary in order to detect a threefold increase in incidence of MS with 85% probability. Conversely, 48 Rochester residents diagnosed with MS were followed for 860 person-years for subsequent onset of herpes zoster. Three cases were observed, whereas 1.8 cases were expected for a relative risk of 1.7, which is not significantly different from unity.