Table 3. Subgroup meta-analyses of hypothyroidism and POAG.

		Random Effects Model		Overall Effect		Test of Homogeneity
Subgroup	No. Studies	OR	95%CI	Z	P	Q	I2 (%)	P
Study design
Case–control	5	1.54	1.08, 2.21	2.36	0.018	54.37	92.6	<0.001
Cross-sectional	4	1.91	0.96, 3.79	1.84	0.066	4.47	32.8	0.215
Nested case–control/ Cohort	2	1.82	1.25, 2.64	3.11	0.002	0.01	0.0	0.920
Geographical area
North America	4	1.87	1.51, 2.32	5.74	<0.001	2.89	0.0	0.409
Europe	2	1.00	0.71, 1.41	0.01	0.991	1.04	4.2	0.307
Asia	4	1.81	1.35, 2.43	3.99	<0.001	3.85	22.0	0.279
Hypothyroidism ascertainment
Medical records	5	1.63	1.14, 2.32	2.68	0.007	51.61	92.3	<0.001
Self-report	4	1.91	0.96, 3.79	1.84	0.066	4.47	32.8	0.215
Adjustment for covariates
>3 factors	7	1.56	1.14, 2.12	2.80	0.005	52.12	88.5	<0.001
≤3 factors	4	2.09	0.89, 4.87	1.70	0.090	5.41	44.5	0.144
Publication year
1993–2010	5	1.44	0.85, 2.43	1.35	0.177	12.11	67.0	0.017
2010–2015	6	1.71	1.62, 1.80	19.45	<0.001	4.21	0.0	0.519

POAG: Primary open angle glaucoma; OR: odds ratio; CI: confidence interval