Table 4.
Meta-analyses of the risk of oral cancer in smokeless tobacco users compared to non-users (based on the estimates given in Table 3).
Heterogeneity | |||||||
Type of smokeless tobacco | No. of estimates combined | Adjustment factorsa | Fixed-effect OR/RR (95%CI) |
Random-effects OR/RR (95%CI) |
χ2 | I2 | p(χ2) |
Allb | 38 | any | 1.54 (1.40–1.69) | 1.87 (1.40–2.48) | 283.3 | 86.9 | < 0.0001 |
19 | none | 1.58 (1.39–1.78) | 2.18 (1.35–3.53) | 232.1 | 92.2 | < 0.0001 | |
15 | sm | 1.31 (1.13–1.53) | 1.35 (1.04–1.76) | 28.7 | 51.3 | 0.0114 | |
7 | sm, al | 1.02 (0.82–1.28) | 1.9 | 0.0 | 0.9321 | ||
7 | sm, al, age | 1.02 (0.82–1.28) | 1.9 | 0.0 | 0.9321 | ||
Chewing | 14 | any | 1.38 (1.21–1.58) | 1.36 (0.98–1.89) | 61.2 | 78.8 | < 0.0001 |
8 | none | 1.37 (1.13–1.68) | 1.28 (0.71–2.30) | 46.4 | 84.9 | < 0.0001 | |
6 | sm | 1.39 (1.16–1.66) | 1.42 (0.99–2.03) | 14.9 | 66.3 | 0.0110 | |
1 | sm, al, age | 1.00 (0.71–1.41) | -- | -- | -- | ||
Snuff | 15 | any | 2.00 (1.67–2.39) | 2.39 (1.17–4.89) | 182.4 | 92.3 | < 0.0001 |
7 | none | 5.36 (3.73–7.69) | 5.15 (0.88–30.2) | 116.7 | 94.9 | < 0.0001 | |
7 | sm | 1.35 (1.09–1.69) | 1.28 (0.76–2.14) | 24.8 | 75.8 | 0.0004 | |
3 | sm, al | 0.88 (0.65–1.18) | 0.4 | 0.0 | 0.7998 | ||
2 | sm, al, age | 0.93 (0.64–1.36) | 0.2 | 0.0 | 0.6555 | ||
Smokelessc | 24 | any | 1.27 (1.13–1.42) | 1.46 (1.17–1.83) | 65.6 | 65.0 | < 0.0001 |
13 | none | 1.18 (1.03–1.35) | 1.26 (0.98–1.61) | 31.5 | 61.9 | 0.0017 | |
7 | sm | 1.09 (0.85–1.42) | 5.4 | 0.0 | 0.4967 | ||
6 | sm, al, age | 1.04 (0.80–1.35) | 1.8 | 0.0 | 0.8751 | ||
4 | sm, al, age, ss | 1.32 (0.77–2.25) | 0.8 | 0.0 | 0.8467 |
a Smoking (sm), alcohol (al), age, social status (ss)
b For each study/sex, the OR/RR for smokeless tobacco from Table 3 was included if available, otherwise that for chewing tobacco or snuff was used. For study 25 where estimates for snuff and chewing tobacco, but not smokeless tobacco, were available, that for chewing tobacco was included as it was for smokers and nonsmokers combined
c This includes all the OR/RR estimates from Table 3 where the exposure was smokeless, both from studies where only estimates for unspecified exposure were available and from studies where it had been estimated from the separate data for chewing tobacco and snuff