Zone 2
Call me naive but does the mitochondrial ability to improve fat burning in zone 2 exercise happen in all muscle cells when you train in zone 2 or is it specific to the cells in the muscles used? That is, is it local or global in your body? For longevity can you do just cycling or should you make sure you are doing zone 2 in your upper body muscle groups as well?
Dostarlimab, Uses/Limitations
I would appreciate a deeper dive on dostarlimab, its uses/limitations, current trials, etc. Have a relative with stage 4 colorectal cancer and am interested in its applications to later stage metastasized colorectal cancers that have the same mutations. I read the memo in which you addresses the medias sensationalization of the limited scope trial for dostarlimab and colorectal patients, and I think because it has gained so much attention it would benefit a lot of your listeners to address it more completely. Cheers
Population stratification in Mendelian randomization
Hi Peter & co., You recently said this about Mendelian randomization: "Since genetic variants are randomly allocated at birth, they – unlike LDL-C itself – are largely unaffected by behavioral, socioeconomic, physiological, or other confounding variables, which would otherwise prohibit causal inference in observational studies." The implication here seems to be that the allocation of each gene is independent from both 1) the listed confounding variables, and 2) other genes. But this doesn’t seem true in general. Membership in a subpopulation like a racial or ethnic group could be a confound, or there could be unobserved subpopulations corresponding to other groups we don't have names for. For example, admixture with Europeans accounts for both HLA haplotype variants and diabetes prevalence in a Pima Indian reservation (Knowler et al., 1988; https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1715499/), and ethnic group membership accounts for a lot of the variance in both alcoholism and dopamine D2 receptor variant (https://pubmed.ncbi.nlm.nih.gov/8095994/, https://onlinelibrary.wiley.com/doi/abs/10.1002/ajmg.1320480204). Cardon and Palmer (2003, https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(03)12520-2/fulltext) discuss these issues in detail and it seems like there was a shift towards viewing population stratification as not being a major issue, as Smith and Ibrahim (2003, in an article which I think you recommended in the past, https://academic.oup.com/ije/article/32/1/1/642797) say: "Population stratification—i.e. confounding of genotype–disease associations by factors related to subpopulation group membership within the overall population in a study—is unlikely to be a major problem in most situations," citing a couple of articles about cancer risk and the Cardon article above. But while it may not be an issue for cancer risk studies specifically, the Cardon article's more general recommendations do not seem to suggest that it's not a problem in most cases – rather, they seem to me to suggest that you need to either make sure you sample cases and controls from the same population, limit the genetic diversity of your sample, or take measures to explicitly detect and account for subpopulation structure in the genetic data. It seems to me that GWAS with huge populations should inevitably run into complex subpopulation structure, so they really should be assessing subpopulation structure in their analyses. But discussions of MR with GWAS seem to almost always frame the issue like you did — saying that the assortment of genes is random and resistant to confounds, or saying population stratification is "unlikely to be an issue" like Smith and Ibrahim did. This leads me to worry that population stratification is not mitigated or taken seriously in MR GWAS studies. But I don’t know the details of the analysis techniques people use in MR. Do they try to account for this? Either way, do you think it's something we should be thinking about when interpreting the results of MR studies, and if not, why not?