Relation of vitamin D intake, leisure-time physical activity, and season to plasma 25(OH)D in premenopausal women
| Vitamin D sources | β* | P | ||
|---|---|---|---|---|
| Vitamin D intake (IU/d) | ||||
| From food | 0.028 | <0.001 | ||
| From supplements† | 0.014 | <0.001 | ||
| Leisure-time physical activity (MET-hour/wk) | 0.168 | <0.001 | ||
| Season† | ||||
| Spring | 4.012 | 0.054 | ||
| Summer | 20.988 | <0.001 | ||
| Fall | 4.736 | 0.026 | ||
| Winter (reference) | — | |||
↵* β values are estimated from linear regression analysis [25(OH)D is treated as a continuous variable]. β values represent absolute mean difference in nmol/L of plasma 25(OH)D for increments of 1 IU vitamin D intake, 1 MET-hour/wk of physical activity, or for blood collected in spring, summer, or fall compared with winter (adjusted mean in winter, the season of reference, is 57.0 nmol/L). β values are estimated from a model that included age (y), body mass index (kg/m2), and all variables in the table. Pseason < 0.0001 (3 degrees of freedom).
↵† The number of women who took vitamin D supplements was 190. The number recruited in spring, summer, fall, and winter was 249, 184, 218, and 90, respectively.