Variance-components tests for genetic association with multiple interval-censored outcomes.

Area of Science:

• Genetics
• Biostatistics
• Epidemiology

Background:

• Large genetic databases like the UK Biobank are crucial for disease-associated genetic variant discovery.
• These databases often contain interval-censored data, posing analytical challenges.
• Existing methods may lose information by converting time-to-event data to binary outcomes.

Purpose of the Study:

• To develop a novel statistical methodology for genetic association studies using interval-censored data.
• To enable the analysis of genetic variant sets (genes, pathways) with multiple interval-censored outcomes.
• To improve the detection of genetic associations with complex diseases by leveraging comprehensive data.

Main Methods:

• Developed a statistical test to associate sets of genetic variants with multiple interval-censored outcomes.
• Employed methods that preserve information from time-to-event data.
• Utilized simulations to evaluate the power of the proposed approach compared to single-outcome methods.

Main Results:

• The proposed methodology demonstrates significant power gains over single-outcome analyses.
• The method effectively integrates information from multiple correlated phenotypes.
• Simulations confirm the enhanced ability to detect genetic associations.

Conclusions:

• The new method provides a powerful tool for genetic association studies with interval-censored data.
• It offers an advantage over traditional approaches by analyzing multiple outcomes.
• Applied to UK Biobank data, it aids in identifying genes associated with bone fracture and fall risks.