Table 1.

Subgroup analysis for the association of rs6232 and rs6234/rs6235 with obesity

rs6232	Random-effects OR (95% CI)	P-value		Heterogeneity		χ2-test for difference	No. of studies/cohorts	Sample size
		Classical	Global	I2 (95% CI)	P-value
Ethnicity
White Caucasian	1.14 (1.06–1.24)	0.001	1.08 × 10−5	44 (7–66)	0.015	0.485	22	125 579
East Asian							0	0
Hispanic	1.52 (0.65–3.55)	0.335	0.182	67 (0–90)	0.049		3	3679
African	0.83 (0.32–2.13)	0.694	0.810	a	0.597		2	2026
Cohort age-group
Child/adolescent	1.53 (1.22–1.93)	2.84 × 10−4	3.68 × 10−5	25 (0–68)	0.250	3.00 × 10−4	6	7077
Adult	1.10 (1.03–1.17)	0.006	0.001	22 (0–54)	0.182		21	124 207
Cohort size
≤1000	1.14 (0.92–1.40)	0.223	0.351	0 (0–71)	0.536	0.756	7	5983
>1000	1.16 (1.06–1.26)	0.003	1.17 × 10−5	53 (21–72)	0.003		20	125 301
Population-based recruitment
No	1.19 (1.02–1.38)	0.030	7.15 × 10−4	53 (13–75)	0.012	0.222	13	29 799
Yes	1.12 (1.03–1.22)	0.010	0.004	28 (0–62)	0.157		14	101 485
rs6234/rs6235
Ethnicity
White Caucasian	1.09 (1.04–1.14)	2.04 × 10−4	4.28 × 10−9	64 (44–77)	1.25 × 10−5	5.13 × 10−7	23	163 385
East Asian	1.00 (0.98–1.03)	0.961	1.00	22 (0–67)	0.274		5	66 721
Hispanic	1.13 (0.99–1.27)	0.050	0.103	0 (0–85)	0.579		4	4711
African	1.16 (0.92–1.46)	0.222	0.115	59 (0–86)	0.064		4	4764
Cohort age-group
Child/adolescent	1.13 (1.00–1.29)	0.053	5.61 × 10−4	68 (29–86)	0.004	0.005	7	10 296
Adult	1.06 (1.02–1.09)	0.001	6.35 × 10−5	66 (49–77)	4.19 × 10−7		29	229 285
Cohort size
≤1000	1.02 (0.91–1.15)	0.69	0.576	44 (0–77)	0.096	0.439	7	6363
>1000	1.07 (1.04–1.11)	2.56 × 10−5	3.48 × 10−7	71 (58–80)	1.71 × 10−9		29	23 3218
Population-based recruitment
No	1.11 (1.05–1.17)	1.60 × 10−4	3.65 × 10−6	76 (61–85)	1.35 × 10−7	0.001	16	92 067
Yes	1.05 (1.00–1.09)	0.040	0.005	46 (8–68)	0.014		20	14 7514

OR, odds ratio; CI, confidence interval.

^aAnalysis not possible if <2 degrees of freedom.