Cholesterol-lowering drug targets reduce risk of vascular-related dementia: Mendelian randomization of 500,000 individuals

Background and Aims: Dementia is a devastating neurodegenerative disease currently affecting 60 million people worldwide, but little progress has been made in its treatment and prevention. Recent research is pointing towards a shared pathogenesis between dementia and atherosclerotic cardiovascular disease, and several shared modifiable risk factors have been suggested including atherogenic cholesterol and triglycerides. We tested whether atherogenic cholesterol- and triglyceride-lowering drug targets reduce risk of vascular-related dementia, Alzheimer’s disease, all-cause dementia, and ischemic heart disease.

Methods: We included 500,000 individuals from UK Biobank and the Copenhagen City Heart Study plus Copenhagen General Population Study. We selected genetic variants within HMGCR, NPC1L1, ANGPTL4, LPL, and CETP with effect on atherogenic cholesterol and triglycerides. Unweighted allele scores were generated, and Cox regression, one-sample Mendelian randomization, and meta-analyses performed.

Results: In meta-analyses of Cox regression results, hazard ratios for risk of vascular-related dementia/ischemic heart disease per 1 mmol/L (39 mg/dL) lower low-density lipoprotein cholesterol were 0.02(95%CI:0.008-0.04)/
0.19(0.13-0.28) for HMGCR and 0.0001(0.000-0.0007)/0.02(0.006-0.04) for NPC1L1. Corresponding HRs per 1 mmol/L (39 mg/dL) lower non-high-density (non-HDL) cholesterol were 0.22(0.17-0.29)/0.42(0.36-0.48) for CETP. Finally,
corresponding hazard ratios per halving in triglycerides were 0.70(0.38-1.28)/0.65(0.57-0.73) for ANGPTL4 and 0.96(0.60-1.31)/0.60(0.51-0.71) for LPL. In meta-analyses of Mendelian randomization results hazard ratios for risk of vascular-related dementia/ischemic heart disease per 1 mmol/L (39 mg/dL) lower low-density lipoprotein cholesterol were 0.24(95%CI:0.07-0.85)/ 0.28(0.14-0.56) for HMGCR and 0.02(0.002-0.30)/0.03(0.007-0.14) for NPC1L1. Corresponding HRs per 1 mmol/L (39 mg/dL) lower non-HDL cholesterol were 0.11(0.07-0.15)/0.42(0.35-0.49) for CETP. Finally, corresponding hazard ratios per halving in triglycerides were 0.65(0.32-1.30)/0.54(0.41-0.70) for ANGPTL4 and 1.01(0.72-1.41)/0.65(0.56-0.75) for LPL. Results for all-cause dementia were similar to those for vascular-related dementia and results for Alzheimer’s disease were null.

Conclusions: Genetic lowering of atherogenic cholesterol via HMGCR, NPC1L1, and CETP reduce the risk of vascular-related- and all-cause dementia, but not the risk of Alzheimer's disease. This is important because it suggests that these drug targets could be used in dementia prevention.