These three terms can be confusing, so let me disuss them. Incidence involves the rate of occurrence of new cases of a condition. It should not be confused with prevalence, which is a measure of the total number of cases of disease in a population that have accumulated over time. Diagnosis involves the number of cases of a disease or disorder that are actually diagnosed; this allows for many cases to exist, to be part of the incidence and of the prevalence, but not to be diagnosed.

The latest study from the US Centers for Disease controls indicates that the US rate of incidence for autism spectrum disorders is 1 in 160 new births. This is up significantly from prior estimates done only a few years before. This large increase in incidence has generately pushed up the prevalence of autism in US society.

One of the most important questions in autism today is whether this really is an increase in incidence, or just an increase in diagnosis. Is the diagnosis just a fad that is given more freely today? Is it reflecting cases that were incident in the past and not diagnosed? Or, is it a real increase in incidence that also has increased diagnosis?

My strong belief, and that of most autism researchers I have encountered in my research, is that the dramatic rise, with some studies showing a 350 percentage point increase in autism diagnosis in 10 to 15 years, cannot be fully accounted for by other factors outside of an increasing incidence. Moreover, my beliefs have not been formed purely by statistics. I spend a lot of time talking to people about autism, which certainly may distort the accuracy of my beliefs, and the anecdotal evidence of a dramatic rise in autism is strong. More and more families around me are being impacted by autism in its various forms, particularly on the high functioning side of the spectrum.