Skip to main content
. 2014 Apr 3;25(10):2366–2375. doi: 10.1681/ASN.2013101085

Table 3.

Univariate and multivariate regression model identifies association of genotype and biochemical variables with bone disease

Variable Unit Bone Disease
Unadjusted Adjusted, All (for Genotype or Sex and Agea) Unadjusted, Comp/Hom Only
N OR 95% CI P Value N Adjusted OR 95% CI P Value N OR 95% CI P Value
Lower Upper Lower Upper Lower Upper
NPT2c mutation Alleles 122 20.8 7.84 55.09 <0.001 89 33.9a 9.52 120.9 <0.001
Women Yes 133 0.71 0.33 1.55 0.39 43 0.43 0.08 2.41 0.34
Age Year 94 0.96 0.92 0.99 0.02 41 1.02 0.97 1.08 0.45
Serum 25(OH)D 50 nmol/L 71 0.74 0.31 1.79 0.51 71 1.18 0.40 3.49 0.77 33 1.22 0.22 6.79 0.82
Serum calcium 1 SD 79 1.32 0.80 2.18 0.28 77 1.41 0.87 2.28 0.16 33 1.19 0.75 1.88 0.45
Serum phosphate 1 SD 100 0.39 0.24 0.63 <0.001 98 0.50 0.30 0.83 <0.01 42 0.45 0.19 1.03 0.06
Serum intact PTH 10% UL 78 0.73 0.62 0.87 <0.001 76 0.94 0.78 1.14 0.51 32 1.16 0.71 1.88 0.56
Serum 1,25(OH)2D 50 pmol/L 86 1.62 1.20 2.20 0.002 85 1.32 1.04 1.69 0.02 39 1.43 1.01 2.02 0.04
uCa/uCrea 50% UL 88 2.62 1.36 5.05 0.004 85 1.45 0.66 3.21 0.35 38 1.38 0.51 3.72 0.52
TRP(%) 10% 68 0.23 0.13 0.40 <0.001 66 0.21 0.08 0.56 0.002 27 0.02 0.00 0.94 0.05
Serum ALP 1 SD 73 2.46 1.73 3.50 <0.001 72 2.00 1.24 3.20 0.004 29 1.45 1.04 2.02 0.03
Renal calcifications Yes 133 2.90 1.24 6.77 0.01 122 1.39 0.36 5.33 0.63 43 0.43 0.13 1.40 0.16

A univariate logistic regression model (GENMOD; SAS Institute, Inc.) was applied to estimate the unadjusted odds ratios (ORs) and those ORs were further adjusted for genotype (adjusted OR) as well as sex and age using a multivariate logistic regression model. An OR of one indicates that renal or bone involvement is equally likely to occur in individuals with a given variable. An OR greater than one indicates that renal or bone involvement is more likely to occur in individuals with high measurements of a given variable at the indicated units. An OR less than one indicates that renal or bone involvement is less likely to occur with high measurements of a given variable at the indicated units. 95% CI, 95% confidence interval.

a

Odds ratios were further adjusted for sex and age by using a multivariate logistic regression model.