Utility of family history in disease prediction in the era of polygenic scores

Wolford, Brooke N.; Surakka, Ida; Se, Graham; Jb, Nielsen; Zhou, Wei; Me, Gabrielsen; Ah, Skogholt; Bm, Brumpton; Douville, Nicholas J.; We, Hornsby; Lg, Fritsche; Boehnke, Michael; S, Lee; Hm, Kang; Hveem, Kristian; Cj, Willer

doi:10.1101/2021.06.25.21259158

Cited by 2 publications

(2 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2 Lastly, our models were based on only a single PGS, although the performance of several different genome-wide PGSs (ie, those derived from statistical methods, such as metaGRS, LDpred, or PRS-CS) have shown to be nearly equivalent in CAD prediction. 26…”

Section: Discussionmentioning

confidence: 99%

Sex-Specific Survival Bias and Interaction Modeling in Coronary Artery Disease Risk Prediction

Surakka

Wolford

Ritchie

et al. 2023

Circ: Genomic and Precision Medicine

View full text Add to dashboard Cite

Background: The 10-year Atherosclerotic Cardiovascular Disease risk score is the standard approach to predict risk of incident cardiovascular events, and recently, addition of coronary artery disease (CAD) polygenic scores has been evaluated. Although age and sex strongly predict the risk of CAD, their interaction with genetic risk prediction has not been systematically examined. This study performed an extensive evaluation of age and sex effects in genetic CAD risk prediction. Methods: The population-based Norwegian HUNT2 (Trøndelag Health Study 2) cohort of 51 036 individuals was used as the primary dataset. Findings were replicated in the UK Biobank (372 410 individuals). Models for 10-year CAD risk were fitted using Cox proportional hazards, and Harrell concordance index, sensitivity, and specificity were compared. Results: Inclusion of age and sex interactions of CAD polygenic score to the prediction models increased the C-index and sensitivity by accounting for nonadditive effects of CAD polygenic score and likely countering the observed survival bias in the baseline. The sensitivity for females was lower than males in all models including genetic information. We identified a total of 82.6% of incident CAD cases by using a 2-step approach: (1) Atherosclerotic Cardiovascular Disease risk score (74.1%) and (2) the CAD polygenic score interaction model for those in low clinical risk (additional 8.5%). Conclusions: These findings highlight the importance and complexity of genetic risk in predicting CAD. There is a need for modeling age- and sex-interaction terms with polygenic scores to optimize detection of individuals at high risk, those who warrant preventive interventions. Sex-specific studies are needed to understand and estimate CAD risk with genetic information.

show abstract

Section: Discussionmentioning

confidence: 99%

Sex-Specific Survival Bias and Interaction Modeling in Coronary Artery Disease Risk Prediction

Surakka

Wolford

Ritchie

et al. 2023

Circ: Genomic and Precision Medicine

View full text Add to dashboard Cite

show abstract

“…One way to mitigate against including imprecise risk factors in risk prediction is to plan during the study design stage, and collect the correct covariates required for traditional statistical models, such as QRISK3, in large scale cohort studies. As suggested by Wolford et al 2021, biobanks should collect high quality family history risk factors for use in future prediction models. During our preliminary work on an independent external validation of QRISK3 using UK Biobank data, we found that UK Biobank lacks information on the age at which family members were diagnosed with coronary heart disease (a requirement for the QRISK3 model), as shown in Table 1.…”

Section: Collecting the Most Appropriate Risk Factorsmentioning

confidence: 99%

Clinical Prediction Models in Epidemiological Studies: Lessons from the Application of QRISK3 to UK Biobank Data

Parsons¹,

Colopy²,

Clifton³

et al. 2022

Journal of Data Science

View full text Add to dashboard Cite

Statistical models for clinical risk prediction are often derived using data from primary care databases; however, they are frequently used outside of clinical settings. The use of prediction models in epidemiological studies without external validation may lead to inaccurate results. We use the example of applying the QRISK3 model to data from the United Kingdom (UK) Biobank study to illustrate the challenges and provide suggestions for future authors. The QRISK3 model is recommended by the National Institute for Health and Care Excellence (NICE) as a tool to aid cardiovascular risk prediction in English and Welsh primary care patients aged between 40 and 74. QRISK3 has not been externally validated for use in studies where data is collected for more general scientific purposes, including the UK Biobank study. This lack of external validation is important as the QRISK3 scores of participants in UK Biobank have been used and reported in several publications. This paper outlines: (i) how various publications have used QRISK3 on UK Biobank data and (ii) the ways that the lack of external validation may affect the conclusions from these publications. We then propose potential solutions for addressing these challenges; for example, model recalibration and considering alternative models, for the application of traditional statistical models such as QRISK3, in cohorts without external validation.

show abstract

Utility of family history in disease prediction in the era of polygenic scores

Cited by 2 publications

References 45 publications

Sex-Specific Survival Bias and Interaction Modeling in Coronary Artery Disease Risk Prediction

Sex-Specific Survival Bias and Interaction Modeling in Coronary Artery Disease Risk Prediction

Clinical Prediction Models in Epidemiological Studies: Lessons from the Application of QRISK3 to UK Biobank Data

Contact Info

Product

Resources

About