Table 3.
Multivariable linear regression for square root transformed 25(OH)D weighted for the household population aged 12 to 79 years, Canada, 2007 to 2009
|
Variable |
β* ± SE |
Mean 25(OH)D** |
P |
Partial |
|---|---|---|---|---|
| (95% CI) | r2*** | |||
| 91-Day UV-B |
0.04 ± 0.01 |
0.5 (0.3, 0.8) |
0.0001 |
0.03 |
| Sex |
|
|
|
0.01 |
| Female |
(Reference) |
54.2 (49.6, 58.9) |
|
|
| Male |
−0.31 ± 0.05 |
49.8 (48.3, 51.3) |
< 0.0001 |
|
| Age |
|
|
|
0.01 |
| 12-19 years |
−0.13 ± 0.10 |
52.3 (49.6, 55.0) |
0.16 |
|
| 20-59 years |
(Reference) |
54.2 (49.6, 58.9) |
|
|
| 60-79 years |
0.36 ± 0.09 |
59.6 (56.8, 62.4) |
0.0001 |
|
| Ethnicity |
|
|
|
0.06 |
| White |
(Reference) |
54.2 (49.6, 58.9) |
|
|
| Non-White |
−1.10 ± 0.12 |
39.3 (36.5, 42.1) |
< 0.0001 |
|
| Sun Exposure |
|
|
|
0.01 |
| < 30 min/d |
(Reference) |
54.2 (49.6, 58.9) |
|
|
| ≥ 30 min/d and < 2 h/d |
0.22 ± 0.07* |
57.5 (55.4, 59.6) |
0.002 |
|
| ≥ 2 h/d |
0.42 ± 0.09 |
60.5 (57.7, 63.4) |
< 0.0001 |
|
| Vitamin D Supplementation |
|
|
|
0.01 |
| No |
(Reference) |
54.2 (49.6, 58.9) |
|
|
| Yes |
0.44 ± 0.06 |
60.8 (59.1, 62.6) |
< 0.0001 |
|
| Milk Consumption |
|
|
|
0.02 |
| < Once/d |
(Reference) |
54.2 (49.6, 58.9) |
|
|
| Once/d |
0.29 ± 0.05** |
58.6 (57.0, 60.2) |
< 0.0001 |
|
| > Once/d |
0.58 ± 0.06 |
63.1 (61.3, 64.9) |
< 0.0001 |
|
| Body Mass Index |
|
|
|
0.02 |
| < 25 kg/m2 |
(Reference) |
54.2 (49.6, 58.9) |
|
|
| ≥ 25 and < 30 kg/m2 |
−0.13 ± 0.07# |
52.2 (50.3, 54.2) |
0.05 |
|
| ≥ 30 kg/m2 |
−0.63 ± 0.05 |
45.3 (44.0, 46.7) |
< 0.0001 |
|
| Physical Activity Index |
|
|
|
0.01 |
| Inactive |
(Reference) |
54.2 (49.6, 58.9) |
|
|
| Moderately Active |
0.30 ± 0.08 |
57.1 (55.4, 58.8) |
0.0009 |
|
| Active | 0.19 ± 0.06 | 58.6 (56.3, 61.1) | 0.0002 |
Model r2 = 0.21.
*For 91-day UV-B, the estimated mean increase in square root transformed 25(OH)D for a 1 kJ/m2 increase in solar UV-B irradiance between 11 am and 4 pm at the CHMS collection sites averaged over the 91-days prior to blood draw; for other predictors, the estimated mean change in square root transformed 25(OH)D compared to the reference category.
**For 9-day UV-B, the estimated mean increase in back transformed 25(OH)D (nmol/L) for a 1 kJ/m2 increase in solar UV-B irradiance between 11 am and 4 pm at the CHMS collection sites averaged over the 91-days prior to blood draw; for other predictors, the estimated mean change in back transformed 25(OH)D (nmol/L) compared to the reference category.
***Total variance in square root transformed 25(OH)D explained by each predictor, controlling for all other variables.
*Significantly different than ≥ 2 h/d (P = 0.003).
**Significantly different than > once/d (P<0.0001).
#Significantly different than ≥ 30 kg/m2 (P<0.0001).