PrEP Demographics

Following up on the previous post, some people on the Facebook PrEP group were asking about general demographics of PrEP users. This data is again from the six-city study of users of a MSM-targeted real time dating site, and is for the week ending 2015-03-27. The original request was for age, race, gender, orientation, and gender identity. The data comes from a website targeted at MSM and I am limited to the fields and free text the website provides; gender and gender identity are really discussed on the site and I did not capture orientation. From some quick survey work, it looks like the vast majority of users report being either gay or bisexual. I’ll give results for overall PrEP reporting, age, and ethnicity here.

Overall, there were 13905 unique users observed. Out of that, 271 (1.9%) reported being on PrEP, either through free-text comments (e.g. “On PrEP”) or using the relatively new HIV status of “Negative/PrEP.” The overall trendline of free text disclosures dropping and controlled vocabulary disclosures increasing continued.

The average age was 28.3 years, although there were 6721 missing values, either people who chose not to disclose or had obviously incorrect ages. The average age for individuals disclosing being on PrEP is slightly lower, 27.3 years, although this difference is not significant (t(145.18) = -1.58, p = .12).

For race, there are 9 explicit races; the top three are Caucasin (“White) (49.8%), Afro-american (“Black”) (21.9%), and “Mix” (7.9%). American Indian, Asian, Latino, Middle Eastern, Other and Pacific Islander make up the rest, with 378 reporting no race. Just looking at Afro-Americans and Caucasians, .7% of Afro-Americans report being on PrEP, and 2.6% of Caucasians report being on PrEP, a highly significant difference (x^2(1, n=9977) = 34.3, p < .001).

A few random caveats here. First, this is the total population seen. I really should be only looking at new users in this analysis. From prior work I know that once a profile is set up, it is rarely changed so people may have gone on PrEP and not updated their profile. Second, this is all self-reported data. There are many, many reasons to elide or embellish the truth on these sites (Ellison, et. al. “Profile as Promise.” New Media and Society, and Wohlfeiler’s work on the different stake holders), so this is probably a lower rather than upper bound on PrEP usage. Third, this is not a random sample of US cities; the cities were chosen for particular demographic features (e.g. very MSM-accepting, large Afro-American populations, etc.) and I miss some major population centers.