Table 4 Subgroup analyses with respect to duration of bisphosphonates.

From: Bisphosphonates and risk of cancers: a systematic review and meta-analysis

Subgroupsa

<1 year

≥1 year

Number of studies

RR (95% CI)

I2

P for heterogeneity

Number of studies

RR (95% CI)

I2

P for heterogeneity

Oesophageal cancer

5

1.22 (0.83–1.78)

66%

0.02

5

0.99 (0.79–1.24)

26%

0.25

Gastric cancer

3

1.11 (0.91–1.34)

0%

0.56

3

1.15 (0.75–1.75)

76%

0.02

Colorectal cancer

5

1.04 (0.88–1.24)

65%

0.02

7

0.84 (0.71–1.04)

89%

<0.01

Bladder cancer

2

1.59 (0.53–4.75)

74%

0.05

2

0.98 (0.64–1.48)

23%

0.25

Pancreas cancer

2

1.25 (0.44–3.52)

89%

<0.01

2

0.99 (0.58–1.68)

65%

0.09

Renal cell carcinoma

1

1.06 (0.73–1.53)

NA

NA

1

1.10 (0.78–1.56)

NA

NA

Breast cancer

4

0.90 (0.840.97)

0%

0.88

5

0.78 (0.630.98)

92%

<0.01

Endometrial cancer

2

0.89 (0.61–1.30)

0%

0.34

2

0.55 (0.27–1.12)

94%

<0.01

Ovarian cancer

2

0.75 (0.21–2.73)

43%

0.18

2

0.73 (0.41–1.30)

90%

<0.01

Prostate cancer

1

0.90 (0.79–1.02)

NA

NA

1

0.85 (0.760.95)

NA

NA

Lung cancer

1

1.10 (1.00–1.21)

NA

NA

1

1.01 (0.92–1.10)

NA

NA

Melanoma

1

0.90 (0.73–1.10)

NA

NA

1

1.09 (0.92–1.29)

NA

NA

  1. NA not available.
  2. aOther cancers (small-bowel, liver, cervical cancer, lymphoma or leukaemia and oral cancer) were not listed in the table because the number of studies was too few to analyse.
  3. The bold values represent P value < 0.05.