In addition, our analysis is based on cross-sectional and case control studies, which are retrospective in nature, and therefore subject to the biases inherent in retrospective studies.