Table 3.
Multivariable-adjusted Associations of Study- and Month-specific Quartiles of Circulating 25(OH)D3 Concentrations With Colorectal Adenoma by Demographic and Lifestyle Characteristics in the Pooled CPRU and MAP Studies, United States, 1991–2002
| Characteristic and Quartile of 25(OH)D3 | No. of Cases | No. of Controls | ORa | 95% CIa | Pinteraction |
| Ageb | 0.35 | ||||
| <Median | |||||
| 1 | 42 | 69 | 1.00 | ||
| 2 | 38 | 96 | 0.58 | 0.32, 1.06 | |
| 3 | 54 | 93 | 0.85 | 0.48, 1.53 | |
| 4 | 44 | 98 | 0.67 | 0.36, 1.22 | |
| Ptrendc | 0.44 | ||||
| ≥Median | |||||
| 1 | 112 | 101 | 1.00 | ||
| 2 | 121 | 108 | 0.84 | 0.55, 1.28 | |
| 3 | 113 | 91 | 0.79 | 0.51, 1.21 | |
| 4 | 91 | 114 | 0.47 | 0.30, 0.73 | |
| Ptrendc | 0.001 | ||||
| Obesed | 0.14 | ||||
| No | |||||
| 1 | 108 | 108 | 1.00 | ||
| 2 | 102 | 151 | 0.60 | 0.40, 0.91 | |
| 3 | 121 | 143 | 0.73 | 0.48, 1.09 | |
| 4 | 116 | 177 | 0.55 | 0.36, 0.82 | |
| Ptrendc | 0.02 | ||||
| Yes | |||||
| 1 | 41 | 58 | 1.00 | ||
| 2 | 54 | 49 | 1.16 | 0.60, 2.23 | |
| 3 | 44 | 36 | 1.10 | 0.54, 2.23 | |
| 4 | 16 | 30 | 0.46 | 0.20, 1.06 | |
| Ptrendc | 0.14 | ||||
| Physical activitye | 0.06 | ||||
| <Median | |||||
| 1 | 87 | 110 | 1.00 | ||
| 2 | 84 | 99 | 1.00 | 0.63, 1.58 | |
| 3 | 76 | 81 | 1.14 | 0.71, 1.84 | |
| 4 | 51 | 93 | 0.55 | 0.33, 0.91 | |
| Ptrendc | 0.06 | ||||
| ≥Median | |||||
| 1 | 67 | 59 | 1.00 | ||
| 2 | 73 | 103 | 0.50 | 0.30, 0.86 | |
| 3 | 90 | 102 | 0.58 | 0.34, 0.97 | |
| 4 | 84 | 118 | 0.52 | 0.30, 0.88 | |
| Ptrendc | 0.05 | ||||
| NSAID or aspirin use | 0.04 | ||||
| <Once/week | |||||
| 1 | 77 | 84 | 1.00 | ||
| 2 | 90 | 111 | 0.86 | 0.54, 1.38 | |
| 3 | 109 | 94 | 1.33 | 0.83, 2.14 | |
| 4 | 86 | 105 | 0.92 | 0.56, 1.50 | |
| Ptrendc | 0.80 | ||||
| ≥Once/week | |||||
| 1 | 77 | 85 | 1.00 | ||
| 2 | 68 | 92 | 0.68 | 0.41, 1.13 | |
| 3 | 56 | 88 | 0.47 | 0.27, 0.79 | |
| 4 | 49 | 106 | 0.33 | 0.19, 0.56 | |
| Ptrendc | 0.0001 |
Abbreviations: CI, confidence interval; CPRU, Cancer Prevention Research Unit; MAP, Markers of Adenomatous Polyps; NSAID, nonsteroidal antiinflammatory drug; 25(OH)D, 25-hydroxyvitamin D3; OR, odds ratio.
Odds ratios with 95% confidence intervals were adjusted for age (continuous), sex, family history of colorectal cancer in a first-degree relative, regular use of aspirin or NSAIDs, study (CPRU vs. MAP), smoking (current, ever, or never), physical activity (continuous), body mass index (continuous), total red and processed meat intake (continuous), alcohol intake (continuous), calcium intake (continuous), retinol intake (continuous), and folate intake (continuous). A stratification variable was not included in the model.
Study-specific cutpoints were calculated on the basis of the median distribution in controls and were defined as follows for age: CPRU: <52 versus ≥52 years; MAP: <55 versus ≥55 years.
Ptrend values (2-sided) were calculated by including the median of each quartile of blood 25(OH) D3 as a continuous variable in addition to all above-mentioned covariates in the multivariable models.
Obesity was defined as a body mass index of ≥30 kg/m2.
Study-specific cutpoints were calculated on the basis of the median distributions in controls and were defined as follows for physical activity: CPRU: <166 versus ≥166 metabolic equivalent task-hours/week; MAP: <165 versus ≥165 metabolic equivalent task-hours/week.