Studies of the natural disease course of MS are decisively dependent on the reliability of the diagnosis. Until 1983 (Poser et al. 1983; 1984), diagnosis was almost exclusively based on clinical criteria (Schumacher et al. 1965; McDonald and Halliday 1977; see also Chapter 8). In most earlier investigations into the course of the disease, diagnostic criteria were used which probably contained several sources of error when compared to pathological data. The quality of the conclusions drawn from earlier studies is limited by the inability to detect the early stages of the disease and to plan a true prospective investigation. Furthermore, it is unavoidable that individual cases cannot be followed and documented over a sufficiently long period of time by an individual observer. Even in particularly careful studies, errors of documentation may occur in 10% to 30%. Single studies describe heterogeneous groups of patients, with selection bias being introduced deliberately or due to external factors. Therefore conclusions drawn from studies of patients with relapsing remitting disease may not be compared directly with those from cases verified at post mortem or with information gained exclusively from military personnel or sampled in only one center. Despite these reservations, some facts concerning MS may be described, at least on a statistical basis.

It has been known since the first descriptions of the disease that MS may occur primarily with relapses and remissions, and may later run a (secondary) chronic progressive course, or, particularly when disease onset is in the higher age group, it may run a primarily progressive course without single separate exacerbations.