Abstract
Purpose: :
Inconsistent associations between vitamin C and NVAMD in observational studies may reflect measurement errors and uncontrolled confounding. Investigation of variants in genes influencing plasma response to dietary factors provides an unconfounded assessment ("Mendelian randomization") of associations with disease. We investigated the association of a snp in the gene encoding the sodium dependent vitamin C-transporter (SLC23A1) with NVAMD in the EUREYE study.
Methods: :
EUREYE was conducted in seven European countries with ethics approval obtained at each centre. Participants gave written informed consent, were interviewed for risk factors and underwent fundus photography (images independently graded at the Rotterdam centre). Plasma Vitamin C was measured using an enzyme-based assay stabilized with metaphosphoric acid. Other antioxidants (carotenoids and vitamin E) were measured by reverse phase HPLC. DNA was genotyped for rs33972313 using KASPar technology.
Results: :
rs33972313 and plasma vitamin C were available for 2179 people (2081 controls with no AMD and 98 with NVAMD). The genotype prevalence in controls was 95.1% (G:G), 4.9% (G:A) and 0% A:A. Vitamin C levels in controls were lower in G:A compared to G:G, difference of -7.8 mmol/L, 95% CI (-13.6, - 1.9) p=0.02. There were no genotype differences in levels of other antioxidants. The unadjusted ORs of NVAMD for G:A compared to G:G were 1.56, 95% CI (0.65-3.73) p=0.3 and 2.29, 95% CI (1.24-4.22) p=0.01 in analyses adjusting for factors associated with plasma vitamin C, (age, sex, centre, smoking, alcohol, diabetes, cardiovascular disease and systolic blood pressure). The OR was unchanged when plasma vitamin C was added to the model. There was no association between vitamin C and NVAMD in models either including or excluding rs33972313.
Conclusions: :
The increased OR of the rs33972313 A allele for NVAMD supports the role of vitamin C in the aetiology of this condition.
Keywords: age-related macular degeneration • antioxidants • genetics