Plausible genetic instruments
The authors address an important research topic in their mendelian randomisation study investigating the association of genetically predicted testosterone levels with risk of thromboembolism, heart failure and myocardial infarction. I congratulate them in their comprehensive research article and significant contribution to the field.
It is clear from the article that the authors fully appreciate the importance of considering valid genetic instruments in such study. They make effort in their methods to ensure that any such genetic variants 'should be unrelated to factors potentially confounding any association with the outcomes, including baseline age, body mass index, socioeconomic status (Townsend index and educational level), and lifestyle factors (smoking and drinking) at Bonferroni corrected significance.' Furthermore, appropriate caution is offered in the discussion, with explanation that 'variants predicting testosterone from the SHBG gene region are open to the pleiotropic effects of sex hormone binding globulin.' Credit is also offered by the authors to 'the estimates from JMJD1C variants', which 'could be least biased by sex hormone binding globulin because JMJD1C is probably relevant to male fertility'.
I was fascinated to find after a quick search on Wikipedia (
JMJD1C - Wikipedia) that the JMJD1C protein is a candidate histone demethylase and possibly a coactivator for transcription factors, suggestive of widespread physiological effects
beyond testosterone. Further investigating the corresponding gene on the PhenoScanner curated database of results from genome-wide association studies (
PhenoScanner) predominantly highlights associations with platelet traits, rather than testosterone.
It may rather be that variants related to the JMJD1C gene are more likely to be exerting pleiotropic effects (unrelated to testosterone) on risk of risk of thromboembolism, heart failure and myocardial infarction. Consequent bias in the mendelian randomisation estimates might have major implications for the results and indeed the conclusions drawn in this study.