We've actually just finished doing two major studies that looked directly at that issue. We were using the Canadian community health survey, which is the largest national health survey, and we were looking at using the raw data in a secret data centre at U of T that Michael Baker is in charge of.
We were looking at the impact on a variety of health outcomes based on income level. Some of it we were able to break down by gender—some of it we weren't—and in a wide variety of health outcomes, like depression, attempted suicides, these things are far more pronounced among lower-income groups than they are among higher-income groups, and within the lower-income groups they are more pronounced among women than among men.
I didn't bring numbers because I didn't think we were going to go that route, but your hypothesis is very clearly supported by the single largest Canadian survey in the area.