Joe Maranville, Qiao Fan, Tien Yin Wong, Ching–Yu Chen and Heiko Runz
Only a few years ago, doctors would advise their patients that elevated blood levels of high-density lipoprotein cholesterol (HDL-C), then termed the “good cholesterol”, were beneficial and would protect them against coronary heart disease. This belief has been called into question, however, as neither genetics nor clinical trials could demonstrate that raising HDL-C levels would protect against cardiovascular disease. Our study, published recently in the IJE, casts further doubt on this “not-so-good-anymore” cholesterol by showing that genetic variants that cause higher HDL-C levels also increase the risk for age-related macular degeneration (AMD).
AMD is the leading cause of vision loss in the elderly worldwide, and in the United States it affects more than twice as many people as Alzheimer’s disease. A long line of epidemiological studies indicated a relationship between HDL-C and AMD. However, results of these studies have been inconsistent, and researchers could concur neither on a directionality of effects nor whether the assumed relationship was direct or indirect.
More recently, research into human genetics has shed further light on this issue by discovering, through genome-wide association studies (GWAS), that genetic variants in a few genes that raise HDL-C levels also elevate AMD risk. For instance, through GWAS in Europeans, a common allele (rs3764261) near the lipid-regulating gene CETP was found to be associated not only with higher plasma HDL-C levels, but also with AMD risk (odds ratio, 1.15). This was further supported by findings from East Asians that a coding variant in CETP (D442G) that increases plasma HDL-C levels also strongly increased the risk of exudative AMD (per allele odds ratio, 1.7).
Our study further solidifies the link between HDL-C and AMD by demonstrating that, in general, common polymorphisms that modify plasma HDL-C levels also have an impact on AMD risk, and that genetically increased levels of HDL-C increase the risk of AMD.
We chose to test 185 common genetic variants associated through GWAS with plasma lipid levels that had independently also been analysed for their impact on AMD risk. We then applied Mendelian randomisation, a method that leverages the random allocation of DNA variants at birth to test for a causal relationship between a modifiable risk factor (HDL-C) and a disease outcome (AMD risk). The association between HDL-C and AMD remained robust using three alternative Mendelian randomisation models, while variants that modify other lipid species fractions in the blood, particularly “bad” low-density lipoprotein cholesterol and triglyceride levels, did not co-associate with AMD risk.
These observations are consistent with a recent similar study involving people of predominantly European descent. Importantly, our results agree with and extend on this work through direct analysis of the full European-ancestry summary statistics and replication in an independent sample of East Asian ancestry. In total, our results are based on a sample of more than 40 000 people with European and East Asian ancestry, establishing that the link between HDL-C and AMD risk is generalisable beyond a single ethnicity.
Our study thus proposes HDL-C as a potentially key risk factor during AMD pathogenesis. However, this comes with several unknowns. First, the mechanisms that link HDL-C and AMD remain entirely unclear. For instance, we do not yet know whether the culprit for the association is in the blood or whether yet unknown lipid-regulatory mechanisms specific to the eye may play a role. Second, we observed considerable differences in the extent by which HDL-C variants affect AMD risk across populations, loci and subtypes of AMD — the reasons for this remain unexplained. Finally, there are limitations in the Mendelian randomisation approach used in our study. Specifically, we do not yet know how well the insights gained from evaluating the effect of genetic (i.e. lifelong) exposure to elevated HDL-C levels on AMD risk translate into changes to HDL-C levels during one’s lifetime through environmental exposures. Furthermore, our study was designed to assess effects of HDL-C on risk of developing disease, rather than effects on disease activity in people who already have or are in the process of developing AMD.
The latter point will be particularly important for assessing whether drugs that reduce HDL-C levels could be of benefit to treat or prevent AMD. Of potentially more importance is that drugs that elevate HDL-C levels — ranging from statins that mildly increase HDL-C, to niacin that elevates HDL-C by more than 20% — are routinely being used in the treatment of cardiovascular disease. It can thus be expected that all “eyes” will be on a new generation of drugs that target CETP to elevate HDL-C levels and reduce cardiovascular events. The results of a large Phase III cardiovascular outcome trial in about 30 000 patients for one of these drugs, anacetrapib, have recently been released.
It will be interesting to see clinical data emerge that could help to further assess putative risks caused by HDL-C-raising therapies, as well as to validate the suitability of Mendelian randomisation to assist therapy development.
Fan Q, Maranville JC, Fritsche L, et al. HDL-cholesterol levels and risk of age-related macular degeneration: a multiethnic genetic study using Mendelian randomization. International Journal of Epidemiology, dyx189, https://doi.org/10.1093/ije/dyx189.
Joe Maranville is an associate principal scientist at Merck Research Labs in Boston in the USA. He has a background in human genetics (PhD) and clinical pharmacology (certified by American Board of Clinical Pharmacology). His work focuses on the application of statistical genetics to inform drug development, especially integrative analyses of intermediate traits and clinical outcomes.
Qiao Fan is a senior research fellow at the Centre for Quantitative Medicine, Duke-NUS Medical School, in Singapore. Her research focuses on gene mapping, gene–environment interactions, causal inference, risk prediction modelling using high-dimensional data, and genomics of common eye diseases.
Ching-Yu Cheng is an associate professor at Duke-NUS Medical School and principal clinician scientist at Singapore Eye Research Institute. His work involves a variety of epidemiological and clinical research into age-related ocular diseases, as well as identification of susceptibility genes for complex ocular phenotypes, using both genome-wide association approaches and next-generation sequencing.
Tien Wong is professor and medical director of Singapore National Eye Center, Duke-NUS Medical School, National University of Singapore. He is a retinal specialist and leads a broad-based research program comprising epidemiological, clinical and translational studies of retinal diseases, such as diabetic retinopathy and age-related macular degeneration.
Heiko Runz is the head of translational genetics at Merck Research Labs in Boston. He is a board-certified medical geneticist and translational scientist with a track record in clinical, Mendelian and complex genetics. His current research spans from translating genetic discoveries into function to informing clinical drug development programs on medical and business needs.