Exactly why patients with MS become allergic to their own brains and spinal-cord tissues is still unclear, but some clues point to a process that begins with a viral infection, perhaps even a common one. People who live in colder climates—far from the equator in either direction—are at much greater risk for developing MS than those in tropical countries. The differences in risk are enormous: People who live in one location might have a 75-times greater risk of developing multiple sclerosis. If they move when young, the risk of MS comes to resemble that of the people in the new location. However, if people migrate after age 15, they carry with them an MS risk—high or low—that corresponds to their place of origin. I have always interpreted this pattern to suggest that early exposure to a virus common in the tropics causes an infection in children that is harmless but produces lifelong immunity. On the other hand, people missing this early pattern of infection and immunity would be susceptible to an infection later in life—and these later-occurring infections might start the process that leads to MS.

This is not as far-fetched as it might seem. Poliovirus (which has no relationship to MS) has a similar epidemiological pattern. Very young children in tropical countries, perhaps with inferior sanitation, almost never suffer neurological damage when infected with poliovirus—and in that setting, almost all children do become infected early. However, at least before the advent of the vaccine, children in northern cities with good sanitation, who probably missed the early, harmless, polio infection that would have given them immunity, were instead at risk for the paralytic form of the disease if they acquired an infection with this virus later in life.