Are there good example publications on gwas conducted separately (controlling) for low vs high socioeconomic status families

ISG Workshop 2026 Day 2
1

Relatedly, are there good example publications on gwas conducted separately (controlling) for low vs high socioeconomic status families, like parental education (i dont mean within family gwas). I was always wondering why these analyses are that rare. The lottery of birth is defined both by which alleles are inherited and the life circumstances one is born into (both causally before conception). It can be expected that in low SES families the explained variance of most phenotypes is smaller (partially because of the selective nature of our samples).

2

SES is actually heritable - so it complicates the analysis and needs careful modelling

3

you get 2 types of estimates in samples of related individuals: within and between. The between- estimates in family-based samples are about as powered PER SAMPLE as the between-estimates in the population-based samples. But you get within-estimates “for free” in family based samples. They have lower power, as noted above

4

this is especially the case if you are analysing a sample with relatives for a meta-analysisi

5

During the lecture I received the response that also SES is actually heritable, controlling for family ses would actually bias the results. I meant, however, rather to split the GWAS. The thing is that if the GxE model we apply does not suggest differential susceptibility / diathesis stress but stronger effects in one or the other SES group, we often don’t find GxE. Usually I would expect the genetic effect to be stronger in the group with fewer resources to compensate for genetic liabilities. Now all comes together. Our samples are higher in ses than full population, and GWAS explained variance is smaller in low ses group, making it near impossible to find expected GxE pattern.

6

I responded to this: Relatedly, are there good example publications on gwas conducted separately (controlling) for low vs high socioeconomic status families, like parental education (i dont mean within family gwas). I was always wondering why these analyses are that rare. The lottery of birth is defined both by which alleles are inherited and the life circumstances one is born into (both causally before conception). It can be expected that in low SES families the explained variance of most phenotypes is smaller (partially because of the selective nature of our samples).