Despite these results, these studies had severe methodological limitations, such as lack of randomization, lack of covariate-adjusted analysis, and potential selection bias.