Identifying latent classes of adults at risk for skin cancer based on constitutional risk and sun protection behavior.

OBJECTIVE This study used latent class analysis (LCA) to identify homogeneous subgroups of individuals at risk for skin cancer on the basis of their risk factors and sun protection habits to understand patterns of sun exposure and sun protection behaviors. METHOD Adults (n = 725), screened to be at risk for skin cancer, were recruited through primary care physicians and completed the mailed baseline survey as part of a skin cancer prevention trial. LCA was used to identify homogeneous subgroups, on the basis of skin cancer risk and usual sun habits, by sex. LCA solutions were then validated by assessing if class predicted differences in sun exposure, recent sunburn, outdoor activities, attitudes toward tanning, and demographics. RESULTS Females and males were divided into four and three classes, respectively, that reflected the level of sun protection habits and showed that these habits were greater for those with higher constitutional risk for skin cancer. The classes seemed to represent a continuum of sun protection efforts rather than distinct patterns of protection behaviors. Females were distinguished on their use of all habits assessed whereas males, who reported less use of sun protection overall, only differed in their use of sunscreen. Females using more protection reported less sunburn whereas males using less protection reported less sunburn. However, all subgroups reported significant annual prevalence of sunburn (including mild) of 48% to 83%. CONCLUSION LCA can distinguish subgroups of at-risk adults that are relevant and valid. This technique is recommended for targeting intervention efforts when individual tailoring is not feasible.