Skip to main content
. 2011 Nov 9;97(1):279–285. doi: 10.1210/jc.2011-1507

Table 3.

Variables predicting serum 25(OH)D in obese subjects

Variable n Simple linear regression
Multiple stepwise regression
β r2 P β r2 P
Age 411 −2.18 0.094 <0.001 −1.89 0.098 <0.001
Soda intake 395 −3.49 0.041 <0.001 −3.14 0.132
Skip breakfast 411 −8.66 0.053 <0.001 −5.62 0.158
Juice intake 395 −2.40 0.022 <0.001 −2.00 0.172
Milk intake 395 3.28 0.023 0.002
Fruit/vegetable intake 405 3.82 0.021 0.003
BMI Z score 411 −3.00 0.003 0.273

Skip breakfast, Yes/no; soda/juice/milk intake, number of 8-ounce glasses/day; fruit/vegetable intake, less than one, one or two, or at least three servings per day; β, regression coefficient.

The simple linear regressions are for the individual independent variables indicated. For the multiple stepwise regression analysis, only the four independent variables incorporated into the model are listed, and the r2 values displayed are cumulative [i.e. the four variables together account for ∼17% of the total variance in 25(OH)D levels]. The r2 for the three dietary variables (with age ignored) is 0.106.