An evaluation of age-varying genetic effects underlying body-mass index and blood pressure in the UK Biobank.

The proportion of trait-associated SNPs with evidence for age-interaction effects varied substantially by trait.

• 10.3% of BMI discovery SNPs had evidence for an age-interaction effect at P < 0.05.
• 44.7% of pulse pressure (PP) discovery SNPs had evidence for an age-interaction effect at P < 0.05.
• Age-interaction enrichment was described as 'trait dependent', suggesting different genetic architectures across traits.
• The study examined BMI and three blood pressure traits: systolic, diastolic, and pulse pressure.

An increased rate of change in genetically predicted pulse pressure across the age period is associated with higher susceptibility to peripheral artery disease (PAD).

• Interaction odds ratio = 2.71 for the association between rate of change in genetically predicted PP and PAD.
• P-value = 1.82x10-13.
• 95% confidence interval: 2.08-3.53.
• This was identified using a modified inverse-variance weighted (IVW) MR analysis framework.

The study conducted age-stratified GWAS in 2-year age strata across individuals aged 40-69 years in the UK Biobank.

• Sample sizes were up to N = 26,330 per age stratum.
• Traits examined included BMI and three blood pressure traits: systolic blood pressure, diastolic blood pressure, and pulse pressure.
• A meta-regression approach was used to systematically identify SNPs with evidence for age interaction effects.
• Both trait-associated GWAS signals and additional loci genome-wide were examined for age-interaction effects.

Modified inverse-variance weighted MR analyses highlighted the influence of age on the genetically predicted relationship between pulse pressure and adverse cardiovascular outcomes.

• Cardiovascular and cardiometabolic outcomes examined included type-2 diabetes, stroke, peripheral artery disease, heart failure, coronary heart disease, and atrial fibrillation.
• The modified IVW analysis incorporated SNP*Age interaction terms to assess how age modifies genetically predicted trait-outcome relationships.
• Results indicated that PP showed the strongest evidence for age-varying genetic effects influencing cardiovascular disease risk among the traits studied.
• The approach extends conventional MR by describing the effect of SNP*Age interaction on exposure-outcome relationships.

Conventional GWAS approaches assume that genetic associations are constant across different ages, an assumption this study evaluated and found to be violated for multiple traits.

• GWAS are conventionally conducted in cohorts spanning a wide age-range with the assumption of constant genetic associations.
• Age-varying genetic associations have implications for the interpretation of genetic effects in downstream applications such as Mendelian randomization.
• The study found differential enrichment of age-interaction effects that was trait dependent, challenging the constant-effect assumption.
• All summary data generated in the project has been made accessible to the research community as a resource.