Skip to main content

Table 1 Subgroup analyses on the association of depression and fracture risk

From: Association between depression and the risk for fracture: a meta-analysis and systematic review

  No. of reporta HR (95% CI) Q-Statistic P value for heterogeneity I2 (%) P value between groups
Fracture Type
 Any fracture 9 1.16 (1.05–1.29) 15.74 0.073 42.8 0.834
 Hip fracture 10 1.31 (1.08–1.59) 19.75 0.011 59.5  
 Nonvertebral fracture 4 1.51 (1.23–1.85) 3.28 0.351 8.5  
 vertebral fracture 2 1.16 (1.00–1.36) 2.7 0.100 63.5  
Sex
 Male 6 1.59 (1.11–2.29) 6.36 0.273 21.4 0.429
 Female 12 1.23 (1.09–1.39) 29.87 0.002 63.2  
 Mixed 7 1.22 (1.08–1.37) 12.83 0.046 53.2  
Mean age at baseline
  < 65 11 1.15 (1.05–1.26) 27.07 0.003 63.1 0.050
  > =65 14 1.36 (1.22–1.50) 13.81 0.387 5.9  
Study region 0.036
 America 14 1.13 (1.05–1.22) 22.95 0.042 43.4  
 Europe, Australia 9 1.76 (1.44–2.17) 4.67 0.792 0  
 Asia 2 1.36 (1.06–1.73) 2.53 0.112 60.5  
Duration of follow-up
  < 10 17 1.23 (1.10–1.38) 35.61 0.003 55.1 0.679
  > =10 8 1.28 (1.11–1.47) 17.89 0.012 60.9  
Type of depression measure
 Self-reported scales 21 1.29 (1.16–1.44) 45.44 0.001 56.0 0.469
 physician diagnoses 4 1.18 (1.03–1.36) 9.59 0.023 68.6  
Sample size
  < 5000 9 1.44 (1.25–1.66) 5.76 0.674 0.0 0.049
  > =5000 16 1.18 (1.08–1.29) 38.78 0.001 61.3  
Control BMD in models
 Yes 10 1.15 (1.03–1.29) 19.87 0.019 54.7 0.109
 No 15 1.36 (1.20–1.54) 29.9 0.008 53.2  
Control for antidepressants use
 Yes 7 1.19 (1.03–1.39) 12.52 0.051 52.1 0.513
 No 18 1.28 (1.15–1.42) 42.31 0.001 59.8  
Control for smoking
 Yes 16 1.22 (1.08–1.38) 30.01 0.012 50 0.637
 No 9 1.28 (1.14–1.44) 20.23 0.009 60.5  
Control for physical activity
 Yes   1.19 (1.01–1.41) 18.03 0.012 61.2 0.404
 No   1.24 (1.44–1.35) 28.79 0.025 44.4  
  1. Note: aFour articles reported their results by sex group and four articles by type of fracture; there are 25 reports from 16 articles;BMD bone mineral density; CI confidence interval; HR Hazard ratio; Q-Statistic, Cochrane Q statistic; I2, the percentage of total variation due to heterogeneity among studies