Numerous observational studies have attempted to identify risk factors for infection with SARS-CoV-2 and COVID-19 disease outcomes. Studies have used datasets sampled from patients admitted to hospital, people tested for active infection, or people who volunteered to participate. Here, we highlight the challenge of interpreting observational evidence from such non-representative samples. Collider bias can induce associations between two or more variables which affect the likelihood of an individual being sampled, distorting associations between these variables in the sample. Analysing UK Biobank data, compared to the wider cohort the participants tested for COVID-19 were highly selected for a range of genetic, behavioural, cardiovascular, demographic, and anthropometric traits. We discuss the mechanisms inducing these problems, and approaches that could help mitigate them. While collider bias should be explored in existing studies, the optimal way to mitigate the problem is to use appropriate sampling strategies at the study design stage.

A key assumption in Mendelian randomisation is that the relationship between the genetic instruments and the outcome is fully mediated by the exposure, known as the exclusion restriction assumption. However, in epidemiological studies, the exposure is often a coarsened approximation to some latent continuous trait. For example, latent liability to schizophrenia can be thought of as underlying the binary diagnosis measure. Genetically driven variation in the outcome can exist within categories of the exposure measurement, thus violating this assumption. We propose a framework to clarify this violation, deriving a simple expression for the resulting bias and showing that it may inflate or deflate effect estimates but will not reverse their sign. We then characterise a set of assumptions and a straightforward method for estimating the effect of SD increases in the latent exposure. Our method relies on a sensitivity parameter which can be interpreted as the genetic variance of the latent exposure. We show that this method can be applied in both the onesample and two-sample settings. We conclude by demonstrating our method in an applied example and reanalysing two papers which are likely to suffer from this type of bias, allowing meaningful interpretation of their effect sizes.

Background The prevalence of atrial fibrillation (AF) is increasing with an aging worldwide population, yet a comprehensive understanding of its causes and consequences remains limited. We aim to assess the causes and consequences of AF via a bidirectional Mendelian randomization (MR) analysis. Methods We used publicly available genome-wide association study (GWAS) summary data, centralized and harmonized by an open GWAS database. We assessed the genetically predicted effects of 5048 exposures on risk of AF, and the genetically predicted effects of genetic liability to AF, on 10 308 outcomes via two-sample MR analysis. Multivariable MR analysis was further conducted to explore the comparative roles of identified risk factors. Results MR analysis suggested that 55 out of 5048 exposure traits, including four proteins, play a causal role in AF (P <1e-5 allowing for multiple comparisons). Multivariable analysis suggested that higher body mass index, height and systolic blood pressure as well as genetic liability to coronary artery diseases independently cause AF. Three out of the four proteins (DUSP13, TNFSF12 and IL6R) had a drug prioritizing score for atrial fibrillation of 0.26, 0.38 and 0.88, respectively (values closer to 1 indicating stronger evidence of the protein as a potential drug target). Genetic liability to AF was linked to a higher risk of cardio-embolic ischaemic stroke. Conclusions Our results suggest body mass index, height, systolic blood pressure and genetic liability to coronary artery disease are independent causal risk factors for AF. Several proteins, including DUSP13, IL-6R and TNFSF12, may have therapeutic potential for AF.