New paper "Estimating dose—response relationships for vitamin D: observational and Mendelian randomization analyses" at @TheLancetEndo: thelancet.com/journals/landi…. We find genetic evidence for causal relationship between vit D and mortality, but only in people with low vitamin D. 🧵 @TheLancetEndo Despite claims of health risks of low vitamin D, most large randomized trials have failed to demonstrate reductions in risk of heart attack, stroke, or cancer amongst those given vitamin D supplements. Does vitamin D reduce disease risk, or is this correlation with no causation?

New preprint "Disentangling the effects of traits with shared clustered genetic predictors using multivariable Mendelian randomization" online at arxiv.org/abs/2109.12361 led by Fatima Batool. Summary follows! Comments welcome! When applying Mendelian randomization to assess the potential impact of an exposure that is a gene product, it is typical to use genetic variants in the gene region that encodes the gene product as instruments in the analysis - known as cis-Mendelian randomization.

Prof Miguel Hernan from @HarvardChanSPH giving the 17th Armitage Lecture in Cambridge on target trial emulation: "How do we learn what works? A two-step algorithm for causal inference from observational data." @HarvardChanSPH Common passion with Peter Armitage: evidence-based decision making.

Our paper on body mass index (BMI) and all-cause mortality using Mendelian randomization with non-linear methods is out in BMJ today: bmj.com/content/364/bm…. Thread follows... Overall, we saw a J-shaped relationship between BMI and mortality, with genetically increased BMI leading to average reduction in mortality for underweight individuals, and an average increase in mortality for overweight and obese individuals with nadir around BMI 22-25. But...

Announcing another robust method for Mendelian randomization: biorxiv.org/content/10.110…. I hear you cry - do we really need another new robust method for Mendelian randomization? Well this one is a bit different. Tweetorial follows! Most methods for Mendelian randomization assume that all genetic variants target the same causal parameter. However, as the number of variants associated with a given trait is expands into the 100s and 1000s, this assumption is increasingly unlikely.

"Association of LPA Variants With Risk of Coronary Disease and the Implications for Lipoprotein(a)-Lowering Therapies: A Mendelian randomization analysis" out today at JAMA Cardiology ja.ma/2K52Pf7. Tweetorial follows! Lipoprotein(a) is an important biomarker for coronary heart disease, and a potential target for drug development. New drugs in development can lower lipoprotein(a) concentrations by 80-90%. We investigated whether these drugs are likely to reduce coronary heart disease risk.