I wonder about prevalence too. About 300K in the U.S. is often quoted.
However, with the advent of using the MRI and the push to treat with CRABs, I wouldn't be surprised if the number has gone way up, compared to the 1970s, given that more people would be diagnosed. Seek and ye shall find....
I was diagnosed solely on the basis of MRI in 2004 - I did not have any other clinical test. Twenty years ago, even 10 years ago, I doubt I would have even been diagnosed with MS, even though I suspect I've had it (or whatever "it" is) for at least that long. I suspect formerly, doctors would wait and see on more borderline symptoms, reserve the diagnosis for more clear-cut cases. Some people might have a clinically isolated event and never have another neuro symptom, others might draggle along with low-grade nonspecific symptoms, like fatigue, which I had on occasion. An MS diagnosis in the US is like the kiss of death, it forever blocks you from access to health insurance coverage for ANY ailment or injury, let alone your MS, should you lose your employer insurance and don't meet eligibility for government assistance. You become a pariah as far as U.S. medical insurers are concerned. Back when there were no treatments, no good diagnostic imaging, and partly because of medical insurance here (there was also the psychological impact, tragic for a wrong diagnosis), there was an actual disincentive for a doctor to hand out the MS diagnosis in the case of mild or isolated symptoms.
So before CRABs (which was also before MRIs), doctors were probably less likely to make an MS diagnosis, particularly in the case of milder disease. Now, there is more of a "push" for diagnosis, since the general wisdom is to do an MRI and if indicated, to start this now available drug treatment earlier rather than later. I suspect, if there were a cheap, screening level blood test for MS, (like blood and urine sugar in diabetes) there would be even more MS cases found, as MRIs and spinal taps are pricey and the latter somewhat risky and unpleasant to be a screening test. Just think, if everyone complaining of something like fatigue or tingling fingers got a quick "MS blood test", how many more cases would be found?
I understand there are case reports of "subclinical" MS, that is, people only diagnosed on autopsy or diagnosed incidentally while alive, like when getting an MRI following a head injury after getting bucked off a horse or something (see Siva 2006 review article for examples). Whether those people are truly asymptomatic or their symptoms are so low-grade is unknown. Before MRIs, live subclinical cases were undiscovered. Now, they too may get added to the MS pool if the doctor decides early treatment is a good idea. That means - more cases.
This makes we wonder too about articles tracking the "natural history" of MS. Older studies tracking natural history of MS tended to indicate a worse prognosis than more recent ones. Many of the people in the early 20th century did not live more than a few years post-diagnosis. I just wonder if that is because back then, only the most severe cases were diagnosed, or were only diagnosed when they reached a certain stage. Since many people now are on CRABs, some authors report the drugs are improving overall long-term prognosis. I am not so sure, I have to wonder if in recent years, the pool of MS diagnosed persons has been "diluted" by those with milder disease that would not been diagnosed in earlier years and that has impacted observed disease course in the MS population. I doubt that confounder is controlled for in the studies.
Just some rambling thoughts on MS epidemiology. There is so much not known about this disease.