Table 2. Stratified analysis of the association between statin use and AMD.

Subgroup	N	Pooled RR	95% CI	P
				Heterogeneity	Meta-regression
Any AMD	14	0.95	0.74, 1.15	<0.001	NA
Early AMD	7	0.83	0.66, 0.99	0.33
Cohort	5	0.77	0.56, 0.97	0.19	0.36
Cross-sectional	2	0.98	0.71, 1.26	0.74
Late AMD	8	0.92	0.77, 1.07	0.63
Cohort	5	0.95	0.68, 1.22	0.30	0.29
Case-control	1	0.90	0.80, 1.00	NA
Cross-sectional	2	0.83	0.15, 1.50	0.73
GA	4	1.16	0.77, 1.56	0.27
CNV	5	0.90	0.80, 0.99	0.86	0.47
Cohort	4	0.82	0.41, 1.23	0.77
Case-control	1	0.90	0.80, 1.00	NA
Study type
Cohort	7	0.98	0.77, 1.19	0.04	0.78
Case-control	5	0.90	0.55, 1.26	<0.001
Cross-sectional	2	0.95	0.70, 1.19	0.62
Region
America	7	1.03	0.84, 1.23	<0.001	0.67
Europe	5	0.90	0.50, 1.30	<0.001
Age
≥65 years	10	0.99	0.75, 1.23	<0.001	0.56
<65 years	4	0.75	0.53, 0.98	0.35
Classification criteria
WAMGS	11	0.88	0.74, 1.15	<0.001	0.31
ICGS	3	1.09	0.82, 1.37	<0.001

AMD, age-related macular degeneration; CNV, choroidal neovascularization; GA, geographic atrophy; ICGS, the International Classification and Grading System; N, number of studies; NA, not applicable because only one study; CI, confidence intervals; RR, pooled relative risk; WARMGS, the Wisconsin Age-related Maculopathy Grading System.