Skip to main content
. 2017 Dec 13;8:344. doi: 10.3389/fendo.2017.00344

Table 3.

Multivariate-adjusted mean WB-BMD across tertiles of dietary patterns- general linear model test.

Dietary pattern Tertile 1 (low level) (N = 59) Tertile 2 (medium level) (N = 59) Tertile 3 (high level) (N = 59) P for trend P-post hoc analysis
Food pattern 1

WB-BMD (g/cm2) 1.023 ± 0.12 1.022 ± 0.13 1.085 ± 0.10 0.002 I vs. III 0.021
II vs. III 0.004
Model 1 1.010 ± 0.01 1.006 ± 0.01 1.086 ± 0.01 <0.001 I vs. III 0.001
II vs. III <0.001
Model 2 1.021 ± 0.01 1.013 ± 0.01 1.070 ± 0.01 0.043 I vs. III 0.040
II vs. III 0.019

Food pattern 2

WB-BMD (g/cm2) 1.039 ± 0.10 1.020 ± 0.14 1.069 ± 0.11 0.031 II vs. III 0.031
Model 1 1.023 ± 0.01 1.013 ± 0.01 1.065 ± 0.01 0.054 II vs. III 0.022
Model 2 1.023 ± 0.01 1.000 ± 0.01 1.081 ± 0.01 0.003 I vs. III 0.018
II vs. III 0.001

Food pattern 5

WB-BMD (g/cm2) 1.030 ± 0.11 1.030 ± 0.12 1.057 ± 0.13 0.34
Model 1 1.014 ± 0.01 1.032 ± 0.01 1.056 ± 0.01 0.19
Model 2 1.021 ± 0.01 1.039 ± 0.01 1.044 ± 0.01 0.64

Model 1: BMI, glucose, and creatinine adjusted. Model 2: further adjustment for gender, current smoking and medications (antiresorptive agents/calcium/vitamin D supplementation; statins; antihypertensives; anticoagulants; oral antidiabetics).

WB-BMD, whole-body bone mineral density; BMI, body mass index.