Table 3. Multiple regression analysis to demonstrate the relationships between BMI and potential confounding variables on serum vitamin B12.
Partial R is the adjusted correlation coefficient.
| Factors associated with serum vitamin B12* | Partial R | β | 95% CI for β | t | p |
|---|---|---|---|---|---|
| BMI (kg/m2)* | -0.25 | -0.59 | (-0.74, -0.44) | 7.75 | <0.001 |
| Social/demographic | |||||
| - age | 0.03 | 0.002 | (-0.003,0.007) | 0.77 | 0.4 |
| - Townsend score | -0.03 | -0.003 | (-0.01, 0.004) | -0.76 | 0.5 |
| - smoking | -0.05 | -0.06 | (-0.13, 0.01) | -1.59 | 0.1 |
| - vegetarianism | -0.03 | -0.04 | (-0.12, 0.04) | -0.93 | 0.4 |
| - Vitamin use | 0.14 | 0.11 | (0.06, 0.16) | 4.16 | <0.001 |
| Pregnancy related | |||||
| - Parity | 0.01 | 0.009 | (-0.04, 0.06) | 0.37 | 0.7 |
| Haemodilutional | |||||
| - Haematocrit (%) | 0.15 | 2.11 | (1.21, 3.01) | 4.59 | <0.001 |
*Denotes variable transformed using natural logs (allowing β coefficients to be interpreted in terms of percentage change)