Article Figures & Data
Figures
- Figure 1.
Associations between free vitamin D and colorectal cancer risk by selected participant characteristics. Odds ratios are presented as a per standard deviation increase in level of free vitamin D and result from conditional logistic regression models where cases and controls are matched on race, age, and sex. Odds ratios and 95% confidence intervals are adjusted for calendar week of sample collection, body mass index, education, smoking, physical activity, alcohol intake, history of colorectal cancer screening, and family history of colorectal cancer. Analyses include African American cases with greater than 3 years between blood draw and diagnosis and their matched controls. Free vitamin D is calculated as 25(OH)D:VDBP molar ratio (×103) and is a proxy for free 25-hydroxyvitamin D status. a, P values for interaction were calculated by inclusion of cross-product terms for free vitamin D and the variable of interest. We were unable to calculate a P value for interaction between sex and free vitamin D status because sex was used as a matching factor.
Tables
- Table 1.
Selected characteristics of African American participants at baseline by case status, SCCS, 2002–2009
Baseline characteristics, N (%) Controls (N = 440) Cases (N = 224) Age, years, median (IQR) 55 (13) 55 (13) Male sex 203 (46.1) 100 (44.6) Education <High school 163 (37.1) 104 (46.4) High school 147 (33.4) 68 (30.4) >High school 120 (27.3) 47 (21.0) Household income, $ <15,000 275 (62.5) 148 (66.1) 15,000–49,999 136 (30.9) 67 (29.9) ≥50,000 16 (3.6) 2 (0.9) Colorectal cancer screening, ever 121 (27.5) 55 (24.6) Family history of colorectal cancer 22 (5.0) 12 (5.4) BMI, kg/m2, median (IQR) 28.3 (9.9) 29.4 (10.0) Current smoker 184 (41.8) 75 (33.5) Non- and moderate alcohol consumers 337 (76.6) 175 (78.1) Physical activity, MET-h/d, median (IQR) 16.3 (22.7) 15.9 (21.0) Sedentary time, hours, median (IQR) 8.0 (6.2) 8.0 (6.3) Total energy, kcal, median (IQR) 2,212 (1748) 2,261 (1826) Meat, servings/day, median (IQR) 2 (2) 2 (1) Fruits and vegetable, servings/day, median (IQR) 3 (2) 3 (2) Folate, micrograms/day, median (IQR) 432 (373) 471 (388) Vitamin D supplement use 0 (0) 0 (0) 25-hydroxyvitamin D, ng/mL, median (IQR) 13.0 (10.2) 12.5 (10.7) Vitamin D–binding protein, μg/mL, median (IQR) 300.6 (56.9) 301.4 (62.2) NOTE: Subjects with data missing for the characteristic of interest are not included in this analysis.
Abbreviation: IQR, interquartile range.
- Table 2.
Associations between vitamin D biomarkers and colorectal cancer risk, African American participants of the SCCS
Biomarker Controls Cases ORa (95% CI) Ptrend ORb (95% CI) Ptrend 25(OH)D, ng/mL Tertile 1: ≤10.49 102 71 1 (Ref) 1 (Ref) Tertile 2: 10.50–16.08 103 47 0.65 (0.41–1.04) 0.62 (0.38–1.01) Tertile 3: >16.08 107 57 0.77 (0.48–1.21) 0.73 (0.45–1.17) Per standard deviation increase 0.87 (0.71–1.06) 0.17 0.82 (0.66–1.02) 0.07 VDBP, μg/mL Tertile 1: ≤280.20 146 74 1 (Ref) 1 (Ref) Tertile 2: 280.21–320.00 145 69 0.93 (0.62–1.40) 0.95 (0.63–1.45) Tertile 3: >320.00 149 81 1.10 (0.74–1.64) 1.08 (0.72–1.63) Per standard deviation increase 1.04 (0.88–1.23) 0.66 1.03 (0.87–1.23) 0.70 Free 25(OH)Dc Tertile 1: ≤5.02 103 60 1 (Ref) 1 (Ref) Tertile 2: 5.03–7.77 102 57 0.92 (0.57–1.46) 0.85 (0.53–1.39) Tertile 3: >7.77 107 58 0.90 (0.56–1.44) 0.83 (0.51–1.36) Per standard deviation increase 0.87 (0.71–1.06) 0.16 0.82 (0.66–1.01) 0.07 Abbreviations: 25(OH)D, 25-hydroxyvitamin D; VDBP, vitamin D–binding protein; Ref, reference.
↵aAnalyses result from conditional logistic regression models where cases and controls are matched on age, race, and sex. Analyses are adjusted for calendar week of sample collection, and body mass index.
↵bIn addition, adjusted for education, smoking, physical activity, alcohol intake, history of colorectal cancer screening, and family history of colorectal cancer.
↵cFree 25(OH)D biomarker is calculated as 25(OH)D:VDBP molar ratio (×103) and is a proxy for free 25-hydroxyvitamin D status.
- Table 3.
Associations between vitamin D biomarkers and colorectal cancer risk by time since blood draw, African American participants of the SCCS
Time between blood draw and cancer diagnosis ≤3 years >3 years Biomarker Cases ORa (95% CI) Ptrend Cases ORa (95% CI) Ptrend 25(OH)D, ng/mL Tertile 1: ≤10.49 22 1 (Ref) 49 1 (Ref) Tertile 2: 10.50–16.08 28 1.22 (0.62–2.38) 19 0.36 (0.19–0.69) Tertile 3: >16.08 32 1.45 (0.75–2.79) 25 0.45 (0.25–0.83) Per SD increase 0.98 (0.76–1.28) 0.90 0.69 (0.21–0.93) 0.02 VDBP, μg/mL Tertile 1: ≤280.20 24 1 (Ref) 50 1 (Ref) Tertile 2: 280.21–320.00 25 1.08 (0.57–2.03) 44 0.91 (0.56–1.48) Tertile 3: >320.00 38 1.55 (0.86–2.82) 43 0.85 (0.52–1.38) Per SD increase 1.26 (1.00–1.60) 0.05 0.89 (0.72–1.10) 0.27 Free 25(OH)Db Tertile 1: ≤5.02 20 1 (Ref) 40 1 (Ref) Tertile 2: 5.03–7.77 29 1.38 (0.71–2.71) 28 0.60 (0.33–1.09) Tertile 3: >7.77 33 1.64 (0.83–3.23) 25 0.51 (0.27–0.95) Per SD increase 0.96 (0.74–1.25) 0.76 0.71 (0.53–0.97) 0.03 Abbreviations: 25(OH)D, 25-hydroxyvitamin D; VDBP, vitamin D–binding protein; Ref, reference; SD, standard deviation.
↵aAnalyses result from conditional logistic regression models where cases and controls are matched on age, race, and sex. Analyses are adjusted for calendar week of sample collection, body mass index, education, smoking, physical activity, alcohol intake, history of colorectal cancer screening, and family history of colorectal cancer.
↵bFree vitamin D biomarker is calculated as 25(OH)D:VDBP molar ratio (x103) and is a proxy for free 25-hydroxyvitamin D status.
Volume 26, Issue 8
