Accounting for age of onset and family history improves power in genome-wide association studies

Abstract

Genome-wide association studies (GWASs) have revolutionized human genetics, allowing researchers to identify thousands of diseaserelated genes and possible drug targets. However, case-control status does not account for the fact that not all controls may have lived through their period of risk for the disorder of interest. This can be quantified by examining the age-of-onset distribution and the age of the controls or the age of onset for cases. The age-of-onset distribution may also depend on information such as sex and birth year. In addition, family history is not routinely included in the assessment of control status. Here, we present LT-FHþþ, an extension of the liability threshold model conditioned on family history (LT-FH), which jointly accounts for age of onset and sex as well as family history. Using simulations, we show that, when family history and the age-of-onset distribution are available, the proposed approach yields statistically significant power gains over LT-FH and large power gains over genome-wide association study by proxy (GWAX). We applied our method to four psychiatric disorders available in the iPSYCH data and to mortality in the UK Biobank and found 20 genome-wide significant associations with LT-FHþþ, compared to ten for LT-FH and eight for a standard case-control GWAS. As more genetic data with linked electronic health records become available to researchers, we expect methods that account for additional health information, such as LT-FHþþ, to become even more beneficial.Currently, most case-control GWASs are conducted with a regression model where the outcome is the case-control status or occasionally the age of onset of disease. 20 In this paper, we have opted for using the phrase age of onset over age at first diagnosis because they commonly refer to the same underlying thing, i.e., when a diagnosis is given. Recently, researchers have proposed several methods that leverage additional information to improve the power to detect genetic associations without having to increase the number of genotyped individuals. These include multivariate methods that leverage shared environmental or genetic correlations between traits and diseases [21][22][23][24][25] as well as methods that account for age of onset. [26][27][28][29] Perhaps the most fruitful development has come from methods that leverage family information to increase statistical power to identify associations, such as genome-wide association study by proxy (GWAX) 30,31 and liability-thresholdmodel-based approach. 32 The liability threshold model conditioned on family history (LT-FH) 32 estimates the

Phenotype integration improves power and preserves specificity in biobank-based genetic studies of MDD

“…A striking exception is MTAG.FamHist, which preserves specificity. This extends previous observations that careful methods can exploit family history to improve genetic studies 43,44,55 . However, MTAG is highly sensitive to input phenotype selection, and therefore requires extensive domain knowledge, just like previous approaches including combining multiple depression measures 13 and GWAS by proxy 43 .…”

Leveraging family history in genetic association analyses of binary traits

“…Such a scenario was demonstrated by our interaction model in the simulations. In addition, LT-FH has been extended to LT-FH ++ that adopts an age-dependent liability threshold model, which accounts for disease age of onset in addition to family history [ 17 ].…”

Dynamic association of the first identifiable symptom with rapidity of progression to first-episode psychosis

“…Multimodal approaches to risk prediction, exemplified by recent applications of machine learning (Koutsouleris et al, 2021; Rosen et al, 2021), can increase prognostic capacity by integrating clinical, environmental, and biological markers of risk. Ultimately, prognoses for the rapidity of illness progression could be further refined by integrating characteristics of the first identifiable symptom with other risk factors such as familial history and polygenic scores (Pedersen et al, 2021). Until these approaches are implemented and accessible to all, simple associations between early symptoms and the rapidity of illness progression can help tailor the intensity and timing of early interventions in at-risk individuals.…”