The amount of autistic people in the United States

February 19, 2015 8:45 AM

If autism is diagnosed in 1 of 68 children, and there are 381 million people in the country, can I use these two facts to find out how many people in the US are autistic?
posted by motioncityshakespeare to Science & Nature (10 answers total)

No, since autism diagnosis rates have greatly increased over time.
posted by deadweightloss at 8:50 AM on February 19, 2015 [2 favorites]

Also, autism diagnosis is different from autism -- e.g., some non-autistic people are misdiagnosed and some autistic people are undiagnosed. You'd at least have to figure out whether the former is roughly equal to the latter to use diagnosis rates as a factor in the equation.
posted by Etrigan at 9:00 AM on February 19, 2015

Yeah, diagnosis rate and incidence rate aren't the same thing. To use that statistic, you have to assume that (a) all autistic children are diagnosed as such, (b) all children with an autism diagnosis are truly autistic, and (c) the incidence rate of autism has not changed significantly over the lifetime of an average American.

As deadweightloss notes, the diagnosis rate has increased greatly, but at one point there was some debate over whether this increase is solely due to an increase in awareness of a condition that had always been present at a constant level in the population, or whether some new environmental factor is also partly responsible for increasing the incidence rate. I don't know what the current thinking is on this front, though.
posted by Johnny Assay at 9:03 AM on February 19, 2015

Not necessarily, since it isn't really making an apples to oranges comparison. Your calculation assumes the following:

1) The definition of autism remains the same throughout time. (But it hasn't).

2) The diagnosis rate of autism is constant.

3) All people with autism are diagnosed with it. (Not likely, and using which definition?).

4) The demographics of the United States are constant throughout time. (Nope)

All in all, this method might be a "back of the envelope calculation" to an interesting intellectual exercise.
posted by oceano at 9:08 AM on February 19, 2015

Response by poster: Well that's a wrap. I assumed as much, because as much as I researched I can't find a number that shows how many people in the US have autism, just the incidence rate. Thank you all.
posted by motioncityshakespeare at 9:18 AM on February 19, 2015

According to the CDC, autism prevalence in North America is ~ 1%. That might get you an estimate that is closer to the 'true' prevalence than the 1 in 68 number. Calculating a prevalence statistic that isn't limited to just children is difficult, for many of the reasons mentioned above. The data to calculate a prevalence statistic for the entire country doesn't' exist, and would be fairly difficult to piece together. These folks recently estimated a global prevalence of 0.76%
posted by anthropophagous at 9:35 AM on February 19, 2015 [1 favorite]

To be pedantic about the other proposition in your question, you mean 318 million people, yes?
posted by psoas at 12:24 PM on February 19, 2015

The U.S. population is slightly under 319 million, also, not 381.
posted by Slinga at 1:22 PM on February 19, 2015 [1 favorite]

In addition to what the other posters have said, your two facts aren't quite enough on their own:

1. 1 in 68 children are diagnosed
2. There are 319 million people in the US

Not all of those people are children, of course, so you'd have to know how many children comprise that 319 million number.
posted by pdb at 2:11 PM on February 19, 2015

Also, it's easy to imagine different mortality rates in autistic and non-autistic children, so the adult incidence might not be the same even in the absence of the other issues already suggested.
posted by Bruce H. at 9:37 PM on February 19, 2015