Table 2. Subgroup analyses.

	BC recurrence					BC mortality					All-cause mortality
Study characteristics	Data number	RR (95% CI)	I2	P1	P2	Data number	RR (95% CI)	I2	P1	P2	Data number	RR (95% CI)	I2	P1	P2
Study design
Prospective	3	0.74 [0.36, 1.52]	88%	0.41		3	0.56 [0.31, 0.99]	65%	0.05		2	0.81 [0.66, 1.00]	5%	0.05
Retrospective	6	0.88 [0.68, 1.14]	69%	0.34	0.65	7	0.97 [0.76, 1.24]	69%	0.81	0.08	8	1.06 [0.98, 1.14]	10%	0.16	0.02
Sample size
<1000	4	0.77 [0.53, 1.13]	69%	0.11		3	0.55 [0.22, 1.35]	82%	0.19		1	1.09 [0.80, 1.49]	—	0.59
≥1000	5	0.89 [0.63, 1.25]	82%	0.50	0.58	7	0.90 [0.69, 1.18]	73%	0.46	0.30	9	1.00 [0.90, 1.11]	47%	0.95	0.91
Definition of BB use
Pre-diagnosis	3	0.67 [0.47, 0.94]	20%	0.02		5	0.72 [0.48, 1.08]	76%	0.11		6	1.06 [0.95, 1.18]	32%	0.28
Post-diagnosis	6	0.96 [0.75, 1.22]	78%	0.73	0.09	5	0.89 [0.62, 1.28]	76%	0.11	0.54	4	0.92 [0.78, 1.08]	28%	0.29	0.14
Follow-up duration
<5 years	3	0.83 [0.55, 1.25]	72%	0.30		6	0.92 [0.68, 1.25]	76%	0.60		6	1.03 [0.93, 1.15]	42%	0.52
≥5 years	6	0.84 [0.60, 1.17]	81%	0.37	0.96	4	0.63 [0.36, 1.09]	73%	0.10	0.23	4	0.88 [0.68, 1.15]	47%	0.35	0.27
Adjustment of menopausal status
Yes	2	0.87 [0.35, 2.11]	87%	0.75		2	0.66 [0.35, 1.24]	54%	0.20		1	0.85 [0.70, 1.04]	—	0.11
No	7	0.82 [0.63, 1.07]	73%	0.14	0.91	8	0.87 [0.66, 1.14]	76%	0.31	0.44	9	1.04 [0.94, 1.14]	29%	0.44	0.17
NOS
6	1	0.85 [0.54, 1.34]	—	0.48		1	1.03 [0.83, 1.28]	—	0.79		3	1.00 [0.90, 1.11]	0%	0.88
7	4	0.87 [0.61, 1.24]	79%	0.45		6	0.79 [0.53, 1.17]	81%	0.24		4	1.09 [0.93, 1.27]	42%	0.16
8	4	0.78 [0.47, 1.29]	84%	0.33	0.94	3	0.75 [0.56, 1.00]	8%	0.05	0.19	3	0.86 [0.71, 1.04]	12%	0.32	0.17

1. P values for subgroup significance;

2. P values for subgroup difference;

Abbreviations: BB, β-blocker; BC, breast cancer; CI, confidence interval; HR, hormonal receptors; NOS, the Newcastle–Ottawa Scale; RR, risk ratio.