This study controlled for confounding factors, such as ethnicity and type of insurance, and had a large sample size and wide geographic representation; these strengths should increase the generalizability of our reassuring findings to other populations.