Neurodegenerative disease and antioxidant biomarkers: A bidirectional Mendelian randomization study

Previous observational studies have suggested that antioxidant imbalance is correlated with neurodegenerative diseases, while its cause-effect remains unclear. Thus, the goal of the present study is to explore the causal relationship between 11 antioxidant biomarkers and 3 most common neurodegenerative diseases [Alzheimer's disease (AD), Amyotrophic Lateral Sclerosis (ALS) and Parkinson's disease (PD)].

A bidirectional Mendelian randomization (MR) study was performed to investigate the causal effects by using 3 main methods (Variance Weighted (IVW), Weighted Median (WM), and MR-Egger regression) in the European population. The data of 11 antioxidant biomarkers were obtained from the open database by the most up-to-date Genome-Wide Association Studies (GWAS), the summary statistics of PD and ALS were obtained from the International Parkinson's Disease Genomics Consortium (IPDGC) (33,674 cases, and 449,056 controls), and the International Amyotrophic Lateral Sclerosis Genomics Consortium (IALSC) (20,806 cases and 59,804 controls), respectively. For AD, we specifically used two recently published GWAS data, one from the International Genomics of Alzheimer's Project (IGAP) (21,982 cases and 41,944 controls), and the other from a large meta-analysis (71,880 cases and 383,378 controls) as validation data.

Based on the Bonferroni correction p < 0.0015, there was no significant causal evidence for the antioxidant biomarkers on neurodegenerative diseases, however, the reverse analysis found that AD was significantly related to the decrease in retinol (IVW: beta = -0.023, p = 0.0007; WM: beta = -0.025, p = 0.0121), while the same analysis was carried out between the AD validation database and retinol, the results were consistent (IVW: beta = -0.064, p = 0.025). Moreover, AD on Glutathione S-transferase (GST), PD on Glutathione Peroxidase (GPX) as well as PD on uric acid (UA) also indicated potential causal-and-effect associations (IVW: p = 0.025; p = 0.027; p = 0.021, respectively).

There was no sufficient evidence that antioxidant imbalance has a significant causal effect on neurodegenerative diseases. However, this study revealed that genetically predicted AD was significantly related to the decrease in retinol, which provides a new insight into previous research and indicates the possibility to regard retinol as potential biomarker for the diagnosis and progress of AD.