Multiple studies have confirmed marked associations between SNPs in the
CFH- and
ARMS2 gene, more recently also the
C3 gene, and the progression to advanced stages of AMD.
23–25,28 However, the investigation of whether the relation between genotypic variants and the progression to early disease is different from that with progression to late AMD has frequently been neglected. Our analyses seem to corroborate a hypothesis formulated in a previous analysis of prevalent AMD cases in the MARS baseline examination.
26 There, we had observed an association between
CFH-rs1061170 and the presence of early AMD, while
ARMS2-rs10490924 seemed mainly related to prevalent late AMD.
26 On stratifying between progression to early and late stages in our cohort, we now show that
CFH-rs1061170 was indeed mostly involved in early progression, that is, the onset of drusen and pigmentary abnormalities, while the influence of
ARMS2-rs10490924 was mainly detectable for the progression to neovascularization and/or atrophy. Our findings concerning the association between the
C3 variant and early or late progression were inconclusive and lacked statistical significance. Associations of
C3 with the development of early AMD have been addressed only in a report from the Rotterdam Study.
28 However, these authors had to pool prevalent and incident cases and patients from a further case-control study to produce a result with some statistical significance. Contrary to findings of others, our results do not indicate a risk-increasing effect of
C3 with regard to late-stage progression.
25,33 We note, however, that our results are based on different study design, low statistical power and a short follow-up period.