Table 5.
Risk Factor | No. of Cases | Total Controls (Population and Spouse/Friend) |
Population Controls |
Spouse/Friend Controls |
Sibling Controls |
Meta-Analysisa |
||||||||||
No. | Odds Ratiob | 95% Confidence Interval | No. | Odds Ratiob | 95% Confidence Interval | No. | Odds Ratiob | 95% Confidence Interval | No. of Casesc | No. of Controls | Odds Ratiod | 95% Confidence Interval | Odds Ratio | 95% Confidence Interval | ||
Hair color | ||||||||||||||||
Dark brown/black | 157 | 189 | 1.00 | 97 | 1.00 | 92 | 1.00 | 109 | 224 | 1.00 | 1.00 | |||||
Light brown | 243 | 207 | 1.26 | 0.94, 1.71 | 98 | 1.46 | 1.00, 2.13 | 109 | 1.11 | 0.76, 1.62 | 157 | 253 | 1.37 | 0.95, 1.97 | 1.62 | 1.11, 2.34 |
Fair or blonde | 140 | 63 | 2.34 | 1.59, 3.44 | 26 | 3.22 | 1.91, 5.43 | 37 | 1.71 | 1.06, 2.78 | 88 | 118 | 1.85 | 1.13, 3.01 | 1.96 | 1.41, 2.74 |
Red | 72 | 25 | 3.02 | 1.79, 5.11 | 12 | 3.20 | 1.60, 6.37 | 13 | 2.88 | 1.44, 5.74 | 42 | 67 | 1.61 | 0.91, 2.83 | 3.64 | 2.56, 5.37 |
Ptrend | <0.001 | <0.001 | <0.001 | 0.02 | ||||||||||||
Ability to tan (repeated exposure) | ||||||||||||||||
Very brown and deep tan | 84 | 119 | 1.00 | 63 | 1.00 | 56 | 1.00 | 50 | 123 | 1.00 | 1.00 | |||||
Moderate tan | 271 | 203 | 1.91 | 1.35, 2.72 | 92 | 2.11 | 1.37, 3.27 | 111 | 1.60 | 1.02, 2.49 | 179 | 293 | 1.69 | 1.09, 2.61 | 1.77 | 1.23, 2.56 |
Mild or occasional tan | 179 | 116 | 2.00 | 1.35, 2.94 | 55 | 2.20 | 1.36, 3.56 | 61 | 1.60 | 0.98, 2.63 | 123 | 189 | 2.01 | 1.25, 3.22 | 1.84 | 1.43, 2.36 |
No tan or only freckles | 75 | 43 | 2.11 | 1.29, 3.45 | 22 | 2.20 | 1.18, 4.09 | 21 | 1.91 | 1.00, 3.65 | 45 | 60 | 2.53 | 1.35, 4.74 | 2.09 | 1.67, 2.58 |
Ptrend | 0.002 | 0.007 | 0.06 | 0.002 | ||||||||||||
Mole count on back | ||||||||||||||||
None | 33 | 59 | 1.00 | 31 | 1.00 | 28 | 1.00 | 22 | 77 | 1.00 | 1.00a | |||||
1–4 | 88 | 103 | 1.37 | 0.80, 2.33 | 44 | 1.56 | 0.80, 3.03 | 59 | 1.25 | 0.65, 2.39 | 51 | 124 | 1.52 | 0.80, 2.87 | 1.44 | 1.29, 1.60 |
5–10 | 137 | 127 | 1.62 | 0.97, 2.71 | 68 | 1.62 | 0.87, 3.01 | 59 | 1.79 | 0.95, 3.36 | 90 | 167 | 2.22 | 1.17, 4.19 | 2.48 | 1.90, 3.23 |
11–15 | 67 | 43 | 2.39 | 1.31, 4.35 | 23 | 2.41 | 1.15, 5.01 | 20 | 2.25 | 1.05, 4.82 | 41 | 78 | 1.94 | 0.95, 3.96 | 4.82 | 3.05, 7.62 |
≥16 | 263 | 133 | 2.87 | 1.74, 4.71 | 60 | 3.33 | 1.81, 6.16 | 73 | 2.53 | 1.38, 4.65 | 172 | 171 | 4.72 | 2.47, 9.00 | ||
Ptrend | <0.001 | <0.001 | <0.001 | <0.001 | ||||||||||||
Reported family history of melanoma in first-degree relative | ||||||||||||||||
None | 525 | 442 | 1.00 | 214 | 1.00 | 228 | 1.00 | 1.00 | ||||||||
≥1 | 97 | 46 | 1.98 | 1.34, 2.92 | 21 | 1.91 | 1.13, 3.22 | 25 | 1.96 | 1.19, 3.23 | 1.74 | 1.41, 2.14 | ||||
P value | <0.001 | 0.02 | 0.008 |
Results from meta-analyses by Gandini et al. (24, 30). For mole count, the odds ratios from the meta-analysis are not directly comparable because they are based on mole count on arms (not on the back as in our study); however, we have used the same cutpoints, and a different pooled analysis (60) has shown similar risk estimates for mole counts on the trunk and upper limbs.
Analyzed by using unconditional logistic regression. Adjusted for age, sex, state, ethnicity (Caucasian, other), reported family history (first degree) of melanoma (none, any), education (junior high school, senior high school, vocational, university), and lifetime cumulative time spent outdoors (quartiles). Excludes spouse controls aged ≥45 years (n = 33) and cases and controls missing values for any of the covariates (6 cases, 4 population controls, 3 spouse controls).
Excludes 222 cases with no sibling controls.
Analyzed by using conditional logistic regression (matched by family). Adjusted for age, sex, education (junior high school, senior high school, vocational, university), and lifetime cumulative time spent outdoors (quartiles). Excludes sibling controls aged ≥45 years and cases and controls missing values for any of the covariates (5 cases; 11 sibling controls).