Table 4 

Results from multiple linear regression analyses with BMD at different measuring sites as the outcome. 

AP lumbar aBMD 
Lateral lumbar aBMD 
Lumbar vBMD 
Total hip aBMD 
Femoral neck aBMD 
Total radius aBMD 



Constant all variables 
0.829 
0.539 
0.2390 
0.783 
0.654 
0.625 

R^{2 }all variables 
0.290 
0.245 
0.266 
0.402 
0.343 
0.591 

R^{2 }for demographic variables 
0.254 
0.163 
0.171 
0.338 
0.279 
0.569 

R^{2 }for disease related variables 
0.079 
0.181 
0.276 
0.174 
0.198 
0.253 



Demographic variables 
B 
Pvalue 
B 
Pvalue 
B 
Pvalue 
B 
Pvalue 
B 
Pvalue 
B 
Pvalue 


Sex 
NA 
NA 
NA 
0.056 
0.009 
NA 
0.082 
< 0.001 

Age, years 
NA 
NA 
NA 
NA 
NA 
0.001 
0.002 

Menopause 
0.156 
< 0.001 
0.104 
< 0.001 
NA 
0.053 
0.048 
0.076 
< 0.001 
0.035 
0.009 

BMI, kg/m^{2} 
0.014 
< 0.001 
NA 
NA 
0.014 
< 0.001 
0.011 
< 0.001 
0.004 
< 0.001 



Diseaserelated variables 
B 
P value 
B 
P value 
B 
P value 
B 
P value 
B 
P value 
B 
P value 


Years since diagnosis 
NA 
0.003 
< 0.001 
NA 
NA 
NA 
NA 

Years since symptom debut 
NA 
NA 
0.001 
< 0.001 
0.002 
0.019 
0.002 
0.028 
NA 

BASMI 
NA 
NA 
0.006 
< 0.001 
0.023 
< 0.001 
0.029 
< 0.001 
0.008 
0.016 

Hemoglobin 
NA 
0.002 
0.008 
NA 
NA 
NA 
NA 

WBC 
NA 
NA 
NA 
0.007 
0.05 
NA 
NA 

PLT 
0.000 
0.003 
NA 
NA 
NA 
NA 
NA 



All demographic and diseaserelated variables that were significantly correlated with BMD in the first analyses were entered as covariates. The table shows only covariates that remained significantly associated with BMD in the stepwise multiple linear regression models. Coding for categorical variables, 1 = yes, 0 = no; for sex, 1 = woman, 2 = man. Beta values are unstandardized regression coefficients. AP, anteroposterior; BASMI, Bath Ankylosing Spondylitis Metrology Index; aBMD, areal bone mineral density; vBMD, volumetric BMD; BMI, body mass index; mSASSS, modified Stoke Ankylosing Spondylitis Spine Score; PLT, platelet count; WBC, white blood cell count. 

Klingberg et al. Arthritis Research & Therapy 2012 14:R108 doi:10.1186/ar3833 