This latest paper goes a long way towards nailing that down. The authors are looking at a huge and unique data set: blood draws from recruits to the US military. There are about 62 million serum samples from periodic HIV testing over the years, and in looking at the 1995-2013 data set the authors identified 955 cases of MS that developed in 10 million service members after they joined. 801 of them had enough serum samples for the study (at least three – the first one on joining, one taken as close to possible to the time of diagnosis, and at least one in between), and the team identified 1566 matched controls for them. 35 of the 801 were EBV-negative at the first sample, and 107 of the 1566 controls. All but one of those 35 became EBV-positive over time, and in fact only one person of the whole 801 was EBV-negative in the last sample before diagnosis (about a year before). Median time for development of MS after going EBV positive was about five years, and the median for MS development after the first sample was about ten years.
The study used cytomegalovirus as a control, since it’s spread in much the same way as Epstein-Barr and thus can account for environmental and behavioral effects. But there was no real correlation at all with MS and CMV, except a slightly lower rate of MS in those who were both CMV and EBV positive, which fits with previous work suggesting that the immune response to CMV attenuates the effects of Epstein-Barr. The paper also took a random 30 MS patients and 30 controls and did a comprehensive search (via antibody responses) for infection in samples both before and after MS diagnosis by basically all known human pathogenic viruses (about 200 species). The only thing that came up out of the baseline was Epstein-Barr.