Table 2.

Subgroup analysis and sensitivity analysis for incidence of breast cancer.

Analysis	No. of studies included	No. of patients/participants included	RR (95% CI)	p-value
SUBGROUP ANALYSIS
Different vitamin D dosage
Higha	3	251/13,403	1.02 (0.80, 1.31)	0.87
Low	4	94/5,734	1.09 (0.73, 1.63)	0.66
Different follow-up duration
Longb	3	261/14,217	1.02 (0.80, 1.30)	0.88
Short	4	84/4,920	1.11 (0.73, 1.69)	0.63
Study location
High latitudec	4	84/4,920	1.11 (0.73, 1.69)	0.63
Low latitude	2	15/1,132	1.09 (0.39, 3.11)	0.87
Menopausal status
Post-	5	95/5,811	1.07 (0.72, 1.60)	0.73
Pre-d	-	-	-	-
Unknown	2	250/13,326	1.03 (0.80, 1.32)	0.83
Baseline vitamin D level
Highe	4	263/14,420	1.02 (0.80, 1.30)	0.88
Low	3	82/4,717	1.11 (0.72, 1.70)	0.63
Sensitivity analysis
Excluding studies with high risk of bias	5	88/5,161	1.15 (0.76, 1.74)	0.52
Excluding studies that involved non-study cointerventions	4	8/680	1.00 (0.25, 3.97)	1.00
Using data of the lowest dose for each study	7	345/19,137	1.04 (0.84, 1.28)	0.71
Fixed-effects model	7	345/19,137	1.04 (0.84, 1.28)	0.71

RR, risk ratio; CI, confidence interval.

^a>1,500 IU/day.

^b>12 months.

^c>50°.

^dNo study that included premenopausal women was found.

^eSerum 25(OH)D > 25 ng/ml.