Table 3. Pooled RRs of thyroid cancer risk and alcohol consumption (drinkers versus nondrinkers), in strata of selected covariates.
| No. of studies/reports | RR (95% CI) | P for heterogeneity within subgroup | P for heterogeneity between subgroup | |
|---|---|---|---|---|
| Overall | 24/27 | 0.80 (0.71, 0.90) | ||
| Gender a | 0.256 | |||
| Women and Men | 13 | 0.77 (0.64, 0.94) | 0.000 | |
| Women | 13 | 0.81 (0.74, 0.89) | 0.886 | |
| Men | 5 | 0.77 (0.58, 1.02) | 0.599 | |
| Study design | 0.937 | |||
| Cohort | 7/8 | 0.87 (0.78, 0.96) | 0.809 | |
| Case-control | 17/19 | 0.75 (0.63, 0.89) | 0.000 | |
| Control source b | 0.000 | |||
| Population based | 13/15 | 0.74 (0.62, 0.88) | 0.005 | |
| Hospital based | 4 | 0.76 (0.52, 1.11) | 0.000 | |
| Geographic area | 0.411 | |||
| America | 12/14 | 0.79 (0.67, 0.93) | 0.000 | |
| Europe | 8/9 | 0.77 (0.65, 0.92) | 0.021 | |
| Asia | 4 | 1.01 (0.58, 1.75) | 0.026 | |
| Publication year | 0.447 | |||
| ≥ 2005 | 15/16 | 0.79 (0.67, 0.93) | 0.000 | |
| < 2005 | 8/9 | 0.83 (0.73, 0.94) | 0.771 | |
| No. of cases | 0.758 | |||
| ≥ 200 | 16/16 | 0.80 (0.71, 0.90) | 0.000 | |
| < 200 | 10/11 | 0.81 (0.60, 1.11) | 0.003 | |
| Tumor subtype | 0.081 | |||
| TC | 14/15 | 0.83 (0.72, 0.96) | 0.000 | |
| DTC | 9/11 | 0.75 (0.61, 0.93) | 0.016 | |
| PTC | 1 | 0.70 (0.50, 0.90) | -- | |
| Quality score | 0.014 | |||
| High (NOS≥ 7) | 18/21 | 0.80 (0.71, 0.91) | 0.007 | |
| Low (NOS< 7) | 6 | 0.78 (0.61, 1.01) | 0.000 | |
| Adjusted for multivariate | 0.217 | |||
| Yes | 18/21 | 0.82 (0.72, 0.93) | 0.000 | |
| No | 6 | 0.73 (0.57, 0.95) | 0.013 |
a
The sum of studies exceeds 24 because five studies reported results for both men and women, separately.
b
Among case–control studies only.
RR, relative risk; CI, confidence interval; TC, thyroid cancer; DTC, differentiated thyroid cancer; PTC, papillary thyroid cancer; NOS, Newcastle-Ottawa Scale.