This question was answered by Steven Hayes, here:
Some of the highlights include:
The ACT literature is just now getting large enough to consider whether its impact is differential based on SES, ethnicity, gender, nationality, and the like. So far the news is good.
…A six hour ACT workshop with patient education works significantly better than a six hour patient education workshop alone in producing changes in diabetes self-management and blood glucose (at 3 month follow-up). The study was done at a public health clinic in a poor and largely Latino and Asian section of East Palo Alto. The percentage of minority participants was 76.5%.
…A 9 hour ACT protocol reduced seizures in epileptics 96% (90& were seizure free at a one yar follow up), while an attention placebo had no effect. The participants were all poor South African blacks living in a residential center.
…As for gender, in all of the studies so far ACT works as well for women as men, except for one analogue pain study in which it worked better for women. By the way, there is some indication that ACT helps with racial prejudice directly.
In Lillis & Hayes (2007) undergraduates enrolled in two separate classes on racial differences were exposed Acceptance and Commitment Therapy and an educational lecture drawn from a textbook on the psychology of racial differences in a counterbalanced order. Results indicate that only the ACT intervention was effective in increasing positive behavioral intentions at post and a 1-week follow-up. These changes were associated with other self-reported changes that fit with the ACT model.
Overall, so far as we can tell, ACT works well in different ethnic, cultural, national, and socio-economic groups; and works for both men and women. We shall learn more as the data comes in, but it is pretty cool that ACT researchers are already refusing to limit their work for middle class majority populations. A similar thing can be said for cognitively disabled populations, as the psychosis data are showing.