Yeah, even the ones where an effect was found have a very low GRADE score (framework for assessing whether conclusions of studies should be used as guidelines) that is affected by risk of biases, directness of evidence, consistency and publication bias.