Table 1.
Summary risk estimates and corresponding 95% CIs for prostate cancer associated with the highest versus the lowest coffee consumption
| No of studies | Summary RR | 95% CI | I2 (%) | P value* | |
| Overall | 15 | 0.91 | (0.84 to 0.98) | 53.2 | 0.008 |
| Prostate cancer category † | |||||
| Localised | 6 | 0.93 | (0.87 to 0.99) | 37.0 | 0.159 |
| Advanced | 8 | 0.88 | (0.71 to 1.09) | 52.7 | 0.039 |
| Fatal | 6 | 0.84 | (0.66 to 1.08) | 46.3 | 0.097 |
| Study location | |||||
| North America | 7 | 0.96 | (0.90 to 1.03) | 17.7 | 0.295 |
| Europe | 6 | 0.85 | (0.74 to 0.98) | 63.3 | 0.018 |
| Japan | 2 | 0.85 | (0.48 to 1.51) | 68.5 | 0.075 |
| NOS score | |||||
| 6 | 2 | 1.20 | (0.84 to 1.72) | 0 | 0.507 |
| 7 | 3 | 1.02 | (0.78 to 1.32) | 0 | 0.810 |
| 8 or 9 | 10 | 0.88 | (0.81 to 0.96) | 66.2 | 0.002 |
| Adjustment for confounders | |||||
| PSA testing ‡ | |||||
| Yes | 6 | 0.86 | (0.77 to 0.96) | 31.8 | 0.197 |
| No | 9 | 0.94 | (0.84 to 1.06) | 60.5 | 0.009 |
| Family history of prostate cancer | |||||
| Yes | 4 | 0.83 | (0.72 to 0.96) | 57.8 | 0.068 |
| No | 11 | 0.95 | (0.85 to 1.05) | 50.8 | 0.026 |
| Total energy intake | |||||
| Yes | 6 | 0.85 | (0.76 to 0.96) | 61.1 | 0.025 |
| No | 9 | 0.97 | (0.85 to 1.09) | 47.7 | 0.053 |
| Smoking status | |||||
| Yes | 10 | 0.86 | (0.79 to 0.94) | 52.0 | 0.027 |
| No | 5 | 1.03 | (0.95 to 1.11) | 0 | 0.805 |
| Alcohol consumption | |||||
| Yes | 6 | 0.87 | (0.84 to 0.98) | 49.2 | 0.008 |
| No | 9 | 0.93 | (0.84 to 1.03) | 57.2 | 0.017 |
| Physical activity | |||||
| Yes | 7 | 0.87 | (0.79 to 0.95) | 58.4 | 0.025 |
| No | 8 | 1.00 | (0.90 to 1.12) | 10.5 | 0.348 |
| BMI | |||||
| Yes | 9 | 0.86 | (0.78 to 0.94) | 56.9 | 0.017 |
| No | 6 | 1.03 | (0.95 to 1.11) | 0 | 0.897 |
| Diabetes | |||||
| Yes | 5 | 0.87 | (0.84 to 0.98) | 64.9 | 0.022 |
| No | 10 | 0.97 | (0.86 to 1.10) | 22.8 | 0.233 |
*P value for heterogeneity within each subgroup.
†Based on definition in each original study, the prostate cancer categories were classified as follows: (a) localised prostate cancer which included localised or non-aggressive cancers, (b) advanced prostate cancer which included advanced or aggressive cancers, (c) fatal prostate cancer which included fatal/lethal cancers or prostate cancer-specific deaths.
‡Since PSA testing was generally introduced after 1986, studies with follow-up periods that ended before 1986 were classified in the PSA-adjusted group.
BMI, body mass index; NOS, Newcastle–Ottawa Scale; PSA, prostate-specific antigen; RR, relative risk.