Table 3 Subgroup analysis

Study design

Cross-sectional

9 [19, 22–24, 26–28, 30, 34]

1.73 (1.53–1.95)

48%

Cohort

8 [20, 21, 25, 29, 31–33, 35]

1.25 (1.16–1.35)

43%

Effect measure

RR

2 [25, 31]

1.33 (1.16–1.51)

32.2%

HR

3 [29, 32, 35]

1.09 (1.05–1.54)

55.5%

OR

12 [19–24, 26–28, 30, 33, 34]

1.30 (1.01–1.76)

86.7%

Average age at baseline

 < 50 years old

6 [21, 25, 26, 31–33]

1.29 (1.10–1.52)

52%

 ≥ 50 years old

11 [19, 20, 22–24, 27–30, 34, 35]

1.50 (1.31–1.72)

62%

Follow-up period (longitudinal only)

 < 10 years

3 [25, 31, 35]

1.29 (1.05–1.58)

54%

 ≥ 10 years

5 [20, 21, 29, 32, 33]

1.23 (1.12–1.35)

45%

Sample size

 < 10,000

6 [20, 24, 25, 29, 34, 35]

1.30 (1.20–1.41)

40%

 ≥ 10,000

11 [19, 21–23, 26–28, 30–33]

1.63 (1.35–1.96)

66%

Diagnostic method

CES-D

7 [20, 21, 26, 30, 31, 33, 35]

1.19 (1.07–1.33)

9.6%

PHQ-9

3 [22, 24, 34]

1.62 (1.05–2.48)

84.7%

HADS

2[27, 28]

1.64 (1.30–2.08)

0.0%

Treated by antidepressants

3 [21, 25, 29]

1.40 (1.24–1.57)

0.0%

Other methods (Zung Self-Rating Depression Scale, GHQ, DASS-21)

4 [19, 23, 32]

1.05 (0.54–2.04)

53%