Table 3. Observational and instrumental variable analyses for the causal association of vitamin D binding protein with 25-hydroxy-vitamin D and common diseases and related traits in the CaMos cohort.
| Outcome | Observational Regression Analysis | IV Analysis | Test for Endogeneity p-Value* | ||
| Effect Estimate per 50 mg/l of DBP (95% CI) | p-Value | Effect Estimate per 50 mg/1 of DBP (95% CI) | p-Value | ||
| Trait (youth and adult cohorts; up to n = 2,254) | |||||
| 25OHD—nmol/l | 4.99 (3.91, 6.07) | 3.2×10−19 | 8.17 (5.42, 10.91) | 5.5×10−9 | 0.01 |
| Free 25OHD—pmol/l | −0.71 (−0.94, −0.47) | 3.1×10−9 | −0.11 (−0.71, 0.49) | 0.49 | 0.03 |
| Calcium—mmol/l | 0.01 (0.01, 0.02) | 1.5×10−5 | 0.01 (−0.003, 0.02) | 0.12 | 0.87 |
| PTH∥—ng/l | −0.01 (−0.03, 0.01) | 0.21 | 0.02 (−0.03, 0.07) | 0.34 | 0.11 |
| Fasting glucose—mmol/l | −0.001 (−0.002, −7.2×10−5) | 0.03 | −3.6×10−4 (−0.02, 0.02) | 0.73 | 0.59 |
| Fasting insulin∥—pmol/l | −0.02 (−0.05, 0.01) | 0.20 | 0.06 (−0.01, 0.13) | 0.11 | 0.02 |
| BMI—kg/m2 | −0.53 (−0.76, −0.29) | 9.1×10−6 | 0.15 (−0.43, 0.72) | 0.61 | 0.01 |
| BMD at femoral neck—g/cm2 | 0.002 (−0.003, 0.007) | 0.46 | −0.005 (−0.02, 0.01) | 0.43 | 0.23 |
| Disease (adult cohort only; up to n = 2,122) | |||||
| Stroke/TIA | 0.83 (0.68, 1.02) | 0.08 | 0.51 (0.32, 0.81) | 0.004 | — |
| Myocardial infarction | 0.90 (0.75, 1.08) | 0.25 | 0.80 (0.53, 1.21) | 0.29 | — |
| Diabetes | 0.79 (0.65, 0.96) | 0.02 | 1.00 (0.66, 1.53) | 0.98 | — |
| Hypertension | 0.91 (0.82, 1.01) | 0.09 | 0.82 (0.64, 1.05) | 0.11 | — |
| Osteoporosis | 0.96 (0.85, 1.09) | 0.53 | 1.32 (0.97, 1.79) | 0.07 | — |
Effect estimates are presented as absolute changes for continuous traits or ORs for disease per 1-SD increase in DBP. All observational regression and IV models were adjusted for age and sex only. Analyses of disease conditions were restricted to the adult cohort only. The null hypothesis is tested using α = 0.05. The Bonferroni method is used to maintain family-wise error rate when performing multiple testing through enforcing a more stringent α = 0.05/12 = 0.004 (12 comparisons).
*The Durbin-Wu-Hausman chi-square test (test for endogeneity) was computed for continuous outcomes.
The outcome was transformed using the natural logarithm, and effect estimates are reported accordingly.
Effect estimates with and without excluding participants with diabetes were similar for fasting glucose. Excluding participants with diabetes, effect estimate per 50 mg/l of DBP: −1.3×10−4 (95% CI −0.001, 3.9×10−4), p = 0.63, and 0.001 (95% CI −0.001, 0.002), p = 0.47, test for endogenicity, p = 0.33; they were also similar for fasting insulin. Excluding participants with diabetes, effect estimate per 50 mg/l of DBP: −2.4×10−4 (95% CI −0.001, 3.4×10−4), p = 0.42, and 0.001 (95% CI −1.9×10−4, 0.003), p = 0.09, test for endogenicity, p = 0.03.