Table 4 Subgroup analyses for MHT and CVD outcomes in observational studies.

From: A systematic review and meta-analysis of effects of menopausal hormone therapy on cardiovascular diseases

Variable

Subgroup

No. of studies

Summary estimates (95% CI)*

I2 (%)

P for heterogeneity

All-cause death

Regimen

E only

7

0.85 (0.77–0.95)

59.5

0.02

Combined EP

7

0.61 (0.34–1.09)

99.3

< 0.01

Duration

< 5 years

2

0.65 (0.25–1.64)

98.0

< 0.01

≥ 5 years

2

0.81 (0.50–1.30)

88.5

< 0.01

Timing of initiation

Early

8

0.68 (0.51–0.92)

94.6

< 0.01

Late

6

0.94 (0.73–1.21)

83.3

< 0.01

Routes of administration

Oral

2

1.01 (0.94–1.08)

0.0

0.75

Non-oral

3

0.83 (0.65–1.07)

49.1

0.14

Underlying diseases

With

3

1.26 (0.34–4.64)

91.6

< 0.01

Without

12

0.89 (0.78–1.01)

89.0

< 0.01

Recency of MHT

Past

4

0.95 (0.86–1.04)

57.8

0.07

Current

4

0.90 (0.78–1.03)

85.8

< 0.01

Study design

Cohort

15

0.90 (0.79–1.02)

88.7

< 0.01

Case–control

0

Nested case–control

0

Study quality

Good and fair

12

0.89 (0.78–1.01)

89.0

< 0.01

Poor

3

1.26 (0.34–4.64)

91.6

< 0.01

Stroke

Regimen

E only

9

1.02 (0.90–1.16)

65.9

< 0.01

Combined EP

6

1.05 (0.81–1.35)

92.0

< 0.01

Duration

< 5 years

3

1.11 (1.04–1.18)

0.0

0.43

≥ 5 years

2

1.22 (1.16–1.29)

5.0

0.30

Timing of initiation

Early

4

0.81 (0.62–1.06)

22.0

0.28

Late

6

0.91 (0.69–1.19)

59.2

0.03

Routes of administration

Oral

5

1.24 (1.11–1.39)

50.7

0.09

Non-oral

5

0.86 (0.77–0.96)

0.0

0.91

Underlying diseases

With

3

1.19 (0.27–5.26)

81.8

< 0.01

Without

10

1.00 (0.88–1.14)

69.6

< 0.01

Recency of MHT

Past

3

1.03 (0.99–1.07)

0.0

0.99

Current

3

1.17 (1.12–1.22)

0.0

0.88

Study design

Cohort

9

0.97 (0.82–1.15)

55.1

0.02

Case–control

2

0.84 (0.75–0.94)

0.0

0.44

Nested case–control

2

1.22 (1.11–1.34)

0.0

0.82

Study quality

Good and fair

10

0.99 (0.87–1.14)

72.1

< 0.01

Poor

3

1.27 (0.40–4.02)

77.3

0.01

VTE

Regimen

E only

8

0.93 (0.79–1.08)

0.0

0.51

Combined EP

6

2.21 (1.51–3.22)

90.1

< 0.01

Duration

< 5 years

4

1.23 (1.02–1.47)

0.0

0.88

≥ 5 years

2

1.19 (0.95–1.51)

0.0

0.39

Timing of initiation

Early

6

1.55 (1.26–1.92)

31.4

0.20

Late

5

1.27 (0.87–1.86)

2.9

< 0.01

Routes of administration

Oral

9

1.41 (1.19–1.67)

72.5

< 0.01

Non-oral

7

0.81 (0.60–1.09)

70.8

< 0.01

Underlying diseases

With

0

Without

12

1.32 (1.13–1.54)

63.0

< 0.01

Recency of MHT

Past

6

1.07 (0.97–1.19)

31.4

0.20

Current

6

1.52 (1.45–1.60)

0.0

0.66

Study design

Cohort

6

1.25 (1.01–1.55)

38.9

0.15

Case–control

5

1.43 (1.07–1.91)

80.8

< 0.01

Nested case–control

1

1.34 (1.03–1.73)

Study quality

Good and fair

10

1.28 (1.08–1.51)

65.7

< 0.01

Poor

2

1.60 (1.15–2.22)

0.0

0.35

MI

Regimen

E only

9

0.85 (0.79–0.91)

0.0

0.67

Combined EP

8

0.77 (0.71–0.84)

20.4

0.27

Duration

< 5 years

3

0.91 (0.73–1.12)

0.0

0.54

≥ 5 years

2

0.51 (0.34–0.76)

0.2

0.32

Timing of initiation

Early

3

0.78 (0.62–0.98)

0.0

0.80

Late

4

0.79 (0.73–0.84)

0.0

0.68

Routes of administration

Oral

2

0.87 (0.57–1.32)

83.3

0.01

Non-oral

3

0.75 (0.60–0.93)

0.0

0.45

Underlying diseases

With

1

0.84 (0.72–0.98)

Without

9

0.79 (0.74–0.84)

0.0

0.88

Recency of MHT

Past

4

0.84 (0.75–0.95)

0.0

0.68

Current

4

0.81 (0.59–1.10)

76.7

< 0.01

Study design

Cohort

5

0.85 (0.76–0.95)

0.0

0.79

Case–control

5

0.77 (0.72–0.83)

0.0

0.94

Nested case–control

0

Study quality

Good and fair

7

0.78 (0.73–0.84)

0.0

0.78

Poor

3

0.84 (0.74–0.95)

0.0

0.89

  1. CI, confidence interval; MI, myocardial infarction; E, estrogen; EP, estrogen and progesterone; VTE, venous thromboembolism. *Summary estimates (95% CI) were measured by fixed-effect models if the I2 was < 30% and P for heterogeneity was > 0.05; otherwise, the summary estimates (95% CI) were measured by random-effect models.