Fig. 1

Probability of reporting use of each CAM modality conditional on latent class membership. CAM use assessed in 2002–2003, since a first primary breast cancer diagnosis in 1996–1997, among cases in the Long Island Breast Cancer Study Project, Long Island, New York. Probabilities of class membership identified through latent class analysis; number of classes determined based on model stability, fit statistics, and interpretability of latent classes; high cumulative dose refers to using each modality in excess of the median cumulative dose; abbreviation: CAM, complementary and alternative medicine