Table 1.
Subgroup analysis of the pooled association of body mass index with prostate cancer specific mortality and all-cause mortality.
| Subgroup | Prostate Cancer Specific Mortality | All-Cause Mortality | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Obesity (BMI ≥ 30) Compared with Normal Weight (BMI < 25) | BMI Continuous Per 5 kg/m2 | Obesity (BMI > 30) Compared with Normal Weight (BMI < 25) | BMI Continuous Per 5 kg/m2 | |||||||||
| N | HR (95% CI) | I2 | N | HR (95% CI) | I2 | N | HR (95% CI) | I2 | N | HR (95% CI) | I2 | |
| All Studies (Total) | 37 | 1.19 (1.10–1.28) | 44.3 | 31 | 1.09 (1.05–1.12) | 44.3 | 23 | 1.09 (1.00–1.18) | 43.9 | 10 | 1.03 (1.01–1.05) | 24.3 |
| Population | ||||||||||||
| Population-Based Incident PC | 22 | 1.24 (1.18–1.31) | 0.0 | 18 | 1.10 (1.07–1.14) | 31.3 | 6 | 1.10 (0.92–1.28) | 25.5 | 0 | - | - |
| Industry Workers Incident PC | 2 | 1.12 (0.55–1.68) | 78.5 | 3 | 1.01 (0.74–1.27) | 65.9 | 0 | - | - | 0 | - | - |
| Radical Prostatectomy | 2 | 1.58 (1.06–2.10) | 0.0 | 2 | 1.04 (0.91–1.17) | 0.0 | 5 | 1.29 (1.11–1.48) | 0.0 | 1 | 1.00 (0.89–1.11) | - |
| Localized PC | 8 | 1.04 (0.97–1.10) | 0.0 | 6 | 1.05 (0.97–1.12) | 55.3 | 5 | 1.20 (1.09–1.30) | 0.0 | 4 | 1.04 (1.01–1.07) | 39.3 |
| Advanced PC | 2 | 1.12 (0.33–1.91) | 73.3 | 1 | 1.28 (0.97–1.59) | - | 6 | 0.91 (0.83–1.00) | 0.0 | 4 | 1.02 (0.99–1.05) | 0.0 |
| African-Caribbean Ancestry | 1 | 0.94 (0.44–2.33) | - | 1 | 0.76 (0.29–1.23) | - | 1 | 0.94 (0.16–1.72) | - | 1 | 0.80 (0.55–1.05) | - |
| Country | ||||||||||||
| USA | 17 | 1.13 (1.00–1.28) | 61.6 | 15 | 1.09 (1.04–1.14) | 36.8 | 12 | 1.05 (0.93–1.17) | 63.8 | 7 | 1.04 (1.02–1.05) | 0.0 |
| European Nordic Countries | 9 | 1.22 (1.12–1.32) | 0.0 | 9 | 1.08 (1.01–1.14) | 45.5 | 2 | 1.22 (1.04–1.40) | 0.0 | 1 | 1.02 (0.98–1.05) | - |
| European Central Countries | 7 | 1.33 (1.14–1.53) | 0.0 | 4 | 1.09 (1.01–1.17) | 25.2 | 6 | 1.26 (1.02–1.50) | 0.0 | 1 | 0.89 (0.70–1.08) | - |
| Other | 4 | 1.51 (1.11–1.92) | 0.0 | 3 | 1.25 (0.80–1.70) | 68.9 | 3 | 1.04 (0.67–1.41) | 0.0 | 1 | 0.80 (0.55–1.05) | - |
| Country status | ||||||||||||
| Developed Countries | 35 | 1.18 (1.10–1.27) | 45.2 | 29 | 1.09 (1.05–1.12) | 36.6 | 22 | 1.09 (1.00–1.19) | 46.3 | 9 | 1.03 (1.01–1.05) | 8.4 |
| Developing Countries | 2 | 1.47 (0.98–1.97) | 0.0 | 2 | 1.13 (0.43–1.82) | 80.3 | 1 | 0.94 (0.44–2.00) | - | 1 | 0.80 (0.55–1.05) | - |
| Exposure Measurement 1 | ||||||||||||
| Prediagnosis BMI | 22 | 1.23 (1.17–1.30) | 0.0 | 21 | 1.09 (1.06–1.13) | 35.0 | 19 | 1.08 (0.98–1.18) | 47.4 | 0 | - | - |
| Postdiagnosis BMI | 15 | 1.10 (0.96–1.23) | 42.8 | 10 | 1.07 (1.00–1.14) | 48.5 | 4 | 1.20 (1.03–1.37) | 0.0 | 10 | 1.03 (1.01–1.05) | 24.3 |
| Quality of the Evidence 2 | ||||||||||||
| Level 2 (Prospective Cohort) | 26 | 1.19 (1.10–1.28) | 34.4 | 26 | 1.09 (1.05−1.13) | 35.1 | 12 | 1.04 (0.92−1.16) | 50.3 | 6 | 1.02 (0.98−1.07) | 20.5 |
| Level 3 | 11 | 1.27 (0.99−1.55) | 57.5 | 5 | 1.08 (0.99−1.17) | 61.5 | 11 | 1.17 (1.02−1.32) | 40.0 | 4 | 1.03 (1.00−1.06) | 46.2 |
| Risk of Bias | ||||||||||||
| NOS: 8−9 | 9 | 1.24 (1.14−1.35) | 0.0 | 12 | 1.11 (1.07−1.15) | 26.6 | 2 | 1.46 (1.01−1.91) | 7.3 | 1 | 1.05 (1.02−1.08) | - |
| NOS: 6−7 | 27 | 1.17 (1.07−1.27) | 49.2 | 18 | 1.07 (1.02−1.12) | 41.6 | 16 | 1.10 (1.01−1.20) | 36.9 | 6 | 1.02 (0.98−1.06) | 34.2 |
| NOS <6 | 1 | 0.69 (0.03−15.03) | - | 1 | 1.31 (0.55−3.15) | - | 5 | 0.90 (0.80−1.07) | 10.3 | 3 | 1.02 (0.99−1.05) | 0.0 |
| Design | ||||||||||||
| Cohort | 36 | 1.19 (1.11–1.28) | 45.9 | 30 | 1.09 (1.05–1.12) | 39.1 | 22 | 1.09 (1.00–1.18) | 43.9 | 9 | 1.03 (1.01–1.05) | 8.4 |
| Case-Control | 1 | 0.94 (0.44–2.33) | - | 1 | 0.79 (0.29–1.23) | - | 1 | 0.94 (0.16–1.72) | - | 1 | 0.80 (0.55–1.05) | - |
| Stage | ||||||||||||
| Adjustment for Stage | 13 | 1.11 (0.95–1.27) | 44.0 | 8 | 1.08 (1.01–1.16) | 52.1 | 14 | 1.08 (0.95–1.21) | 55.6 | 5 | 1.04 (1.01–1.07) | 39.8 |
| Not Adjustment for Stage | 24 | 1.22 (1.16–1.29) | 0.0 | 23 | 1.09 (1.05–1.13) | 35.8 | 9 | 1.12 (1.00–1.23) | 10.0 | 5 | 1.01 (0.99–1.04) | 0.0 |
| Year of Publication | ||||||||||||
| <2016 | 24 | 1.19 (1.10–1.28) | 50.1 | 21 | 1.09 (1.05–1.12) | 39.4 | 6 | 1.13 (0.99–1.27) | 22.1 | 2 | 0.97 (0.71–1.22) | 75.9 |
| ≥2016 | 13 | 1.15 (1.04–1.26) | 28.9 | 10 | 1.08 (1.03–1.14) | 35.4 | 17 | 1.08 (0.97–1.19) | 42.4 | 8 | 1.03 (1.01–1.04) | 0.0 |
BMI, body mass index; HR, hazard ratio; NOS, Newcastle-Ottawa scale; PC, prostate cancer. p-values of the table show the results from heterogeneity analyses of each subgroup. 1 Of the 28 studies evaluating prediagnosis BMI, 1 collected BMI one year before diagnosis [67], 1 measured BMI at 18 years old [14], and 26 collected BMI from retrospective sources or at recruitment and time from measurement to diagnosis was unreported. 2 Quality of the evidence according to the Quality Rating Schemes for Studies and Other Evidence, modified from the Oxford Centre for Evidence-based Medicine for ratings of individual studies [32].