In this study, we conducted comprehensive MR analyses to investigate the causal effects of seven serum lipid biomarkers on the risk of early AMD. We found that higher HDL-C and ApoA levels increased the risk of early AMD, whereas LDL-C, ApoB, and TG appeared to be associated with decreased risk of early AMD. The important value of this study is to fill the gap in the causal effects of lipid biomarkers on the risk of early AMD. More importantly, these MR findings shed light on the different roles of lipid subfractions on early AMD and aid us in the understanding of lipid metabolism in the initiation and formation of drusen in the early stages, as well as the potential utility of blood lipid-modifying therapies in preventing and treating AMD.
Compared with previous MR studies, the direction of causal effect size was the same as the causal effect of lipid on intermediate and late AMD, that genetically elevated HDL-C levels increased the risk of AMD[17–19]. Han et al. found that high-density HDL-C and ApoA1 levels increased the risk of intermediate, GA, CNV, and advanced AMD subtypes[19]. In our study, the effect size of HDL-C levels on early AMD was 1.25 per 1-mmol/l increase (95%CI: 1.22-1.70), which appeared smaller than that on intermediate AMD (OR=1.34 per 1-mmol/l increase, 95% CI: 1.20-1.49) [19]. Notably, there was a robust association between HDL-C and early AMD, specifically using the inverse-variance weighted, weighted median, weighted mode, and MR Egger methods. Besides, as the major apolipoproteins in HDL-C particles, our study also supported that higher ApoA levels increase the risk of early AMD, with OR 2.04 (95%CI: 1.50-2.77, P=6.27×10−6, Table 2). With the ever-increasing knowledge of HDL-C, it shows two side to health. On the one hand, HDL-C induces reverse cholesterol transport associated with plaque regression and have anti-inflammatory[21] and antioxidant properties[22, 23], which could strengthen the endothelial function and decrease the risk of atherosclerosis[24]. On the other hand, the adverse effect of a high HDL-C level on AMD may partially reflect the dysfunction of HDL. It has become apparent that under certain conditions such as cardiovascular disease[25], aging[26], or an acute phase response, HDL-C could have pro-inflammatory and pro-oxidant roles that damage cholesterol efflux[27, 28]. Consequently, oxidation products, such as peroxidation lipids, gradually accumulate in the retina and Bruch’s membrane, which contribute to the development of AMD. Although a pilot study by Vavvas et al. over 12 months reported that high-dose atorvastatin resulted in regression of drusen deposits in 10 of 23 patients[29], the impact of lipid-modifying therapies on AMD risk is currently unknown. Our results emphasize that the potential effects of lipid-modifying drugs on AMD and related phenotypes should be further investigated.
Consistent with previous observational studies, higher TG levels reduce the risk of early AMD[10, 12]. In our study, raised TG levels were associated with decreased risk of early AMD, with OR 0.77 (95%CI: 0.71-0.84, P=5.02×10−10); the effect size was robust on MR-PRESSO outlier-corrected tests. In previous observational studies, LDL-C and CHOL levels were not associated with any early AMD characteristics[12]. In MR studies, the causal association between LDL-C and advanced AMD was insignificant in both Europeans and Asians[17, 18]. However, Han X et al. found that the association of LDL-C was primarily with GA (OR=0.70, 95%CI: 0.59-0.83, P=3.8×10−5) and intermediate AMD (OR=0.77, 95%CI: 0.67-0.87, P=6.5×10−5), but not strong enough with nAMD[19]. Similarly, ApoB (OR=0.76, 95%CI: 0.69-0.85) and CHOL (OR=0.81, 95%CI: 0.70-0.94) were also associated with decreased risk of intermediate[19]. In our study, lipids biomarkers, LDL-C (OR=0.90, 95%CI: 0.85-0.96, P=2.03×10−3) and ApoB (OR=0.52, 95%CI: 0.38-0.72, P=6.60×10−5) present a protective role in the development of early AMD. However, there was no robust association of LDL-C and ApoB with early AMD in weighted median analyses, suggesting that there is no consistent relationship between LDL-C and early AMD. Besides, we find no evidence of the association between Lp(a) and AMD risk.
Dyslipidaemia has been involved in the formation of drusen, which is characterized in the early stage of AMD. Several lines of evidence support a role for dyslipidemia in AMD pathogenesis. First, histological evidence identified that drusen contains lipid material[8, 30]. Second, animal experiments have demonstrated that impaired macrophage cholesterol efflux through HDL-mediated reverse cholesterol transport may lead to a pro-angiogenic status in AMD[31, 32]. Third, genetic associations study has identified risk variants in genes involved in lipid metabolism and the transfer of lipids among lipoproteins, such as hepatic lipase C (LIPC), lipoprotein lipase (LPL), cholesterol ester transferase (CETP), ABC binding cassettes A1 (ABCA1) and apolipoprotein E (APOE), which have all been shown to influence the course of AMD pathobiology[33–35]. Moreover, previous MR studies found the dyslipidemia is causality from intermediate to late AMD[17–19]. Additionally, this study emphasis that serum lipid levels also have a causal effect on early AMD.
This study is mainly based on early AMD and provides additional insights into the causal role of seven serum lipid biomarkers. A strength of this study is that we used large-scale datasets with standard protocols to measure lipid biomarkers; this allowed us to systematically evaluate the effects of lipids on AMD risk. Compared with traditional studies, MR findings are less likely to be affected by confounding or bias from reverse causation. At the same time, our results should be interpreted in light of the limitations of the study. First, as this study was based on European ancestry participants, the generalizability of our findings to other ethnic groups requires further investigation. Second, in the MR framework, the genetically predisposed biomarker changes were assumed to have linear and lifetime effects on AMD risk. The potentially non-linear relationships and short-term effects of these biomarkers are unclear. Third, although biochemical evidence suggests that retina-specific lipid transport is facilitated by proteins regulating systemic lipid metabolism[36], retina-specific lipid concentrations would be more relevant measurements for understanding AMD risk. Further studies are needed to investigate the effects of retina-specific lipid metabolism on AMD risk. Besides, blood lipid levels change throughout life. Therefore, long-term follow-up studies are needed to understand the impact of dyslipidemia on AMD risk and progression over time. Finally, using multiple variants is that the analysis does not point to a single mechanism being causal for AMD risk. Nevertheless, our study has important research and clinical implications.