Table 3. Multivariable Mendelian Randomization Estimates (Odds Ratio With 95% CI) for Heart Failure Without and With Adjustment for CAD Risk.

	HDL-Cholesterol	LDL-Cholesterol	Triglycerides	CAD
Heart failure
Without adjustment for CAD risk	0.97 (0.87–1.08)	1.17 (1.06–1.28)*	1.19 (1.04–1.37)†	-
With adjustment for CAD risk	1.02 (0.92–1.13)	0.95 (0.85–1.06)	1.06 (0.93–1.20)	1.73 (1.47–2.04)‡
Abdominal aortic aneurysm
Without adjustment for CAD risk	0.65 (0.51–0.85)*	1.75 (1.40–2.17)*	0.94 (0.68–1.28)	-
With adjustment for CAD risk	0.69 (0.53–0.88)*	1.46 (1.12–1.91)*	0.84 (0.61–1.16)	1.60 (1.06–2.42)†

Analyses were also performed for abdominal aortic aneurysm as a negative control. The CAD column indicates the association between genetically predicted coronary artery disease and the outcome of interest (ie, heart failure or abdominal aortic aneurysm) while accounting for the genetic associations of the lipid fractions in a multivariable analysis. If the regression coefficients for the lipid measurements remain unchanged on adjustment for CAD, then the effects of lipids on the outcome do not operate via CAD. If the regression coefficients attenuate to the null, then the effects of lipids on the outcome are entirely mediated via CAD. CAD indicates Coronary Artery Disease; HDL, high-density lipoprotein; and LDL, low-density lipoprotein.

^*P<0.003

^†P<0.05

^‡P<10⁻⁸