Observational studies (informing associations) and Mendelian randomization studies (informing causality) provided mutually complementary insight and enabled a more reliable interpretation of perplexing epidemiological relationships. This figure was constructed based on the summary of evidence shown in Table 1.