Table 3.

Multivariate Analyses of Associations Between Drusen Area and Genetic Loci*

Gene/Variant	No Drusen Measured (N = 111)	Drusen Area < Median (N = 54)		Drusen Area ≥ Median (N = 74)
		OR (95% CI)†	P Value‡	OR (95% CI)†	P Value‡
CFH score	1.0 (ref)	0.93 (0.61–1.41)	0.73	1.58 (1.01–2.46)	0.04
ARMS2/HTRA1: rs10490924	1.0 (ref)	1.24 (0.72–2.15)	0.44	2.45 (1.35–4.45)	<0.001
C3 R102G: rs2230199	1.0 (ref)	1.00 (0.55–1.82)	0.99	1.57 (0.87–2.85)	0.14
ABCA1: rs1883025	1.0 (ref)	0.65 (0.31–1.38)	0.27	0.84 (0.41–1.73)	0.65
LIPC: rs10468017	1.0 (ref)	1.14 (0.61–2.12)	0.69	0.76 (0.43–1.36)	0.36
CETP: rs3764261	1.0 (ref)	1.17 (0.67–2.04)	0.57	1.28 (0.78–2.12)	0.33

^*Each genetic variant assessed in a model, controlling for age, sex, education, body mass index, and smoking.

^†OR and CI per risk or protective allele. CFH score categorized as 0, 1, 2, 3–4 risk score categories.

^‡Based on GEE using PROC GENMOD of SAS with the eye as the unit of the analysis using a logistic link and a binomial distribution with the working independence model to account for the intereye correlation. Separate analyses were performed (1) comparing eyes with measurable drusen less than the median versus eyes with no drusen and (2) comparing eyes with measurable drusen greater than or equal to the median versus eyes with no drusen. The reference category is no drusen measured.