Table 2.
The association of BMI with proximal femoral QCT measurements stratified by obesity status in men of the MrOS study
BMI < 30 (n=2397) | BMI>=30 (n=670) | P for Interaction * | |||
---|---|---|---|---|---|
| |||||
SD increase per unit of BMI | P | SD increase per unit of BMI | P | ||
Femoral neck | |||||
BMD (g/cm3) | |||||
Integral | 0.05 ± 0.008 | <0.0001 | −0.012 ± 0.014 | 0.39 | 0.0003 |
Cortical | 0.042 ± 0.008 | <0.0001 | −0.008 ± 0.013 | 0.54 | 0.002 |
Trabecular | 0.002 ± 0.0004 | <0.0001 | −0.0008 ± 0.0006 | 0.18 | 0.0005 |
Volume (cm3) | |||||
Integral | 0.027 ± 0.008 | 0.001 | −0.015 ± 0.014 | 0.28 | 0.014 |
Medullary | 0.006 ± 0.009 | 0.47 | −0.011 ± 0.014 | 0.45 | 0.33 |
Percent cortical volume | 0.042 ± 0.008 | <0.0001 | −0.009 ± 0.013 | 0.49 | 0.002 |
Cross-sectional area (cm2) | 0.03 ± 0.009 | 0.0005 | −0.02 ± 0.014 | 0.14 | 0.003 |
DXA BMD | 0.105 ± 0.008 | < 0.0001 | 0.035 ± 0.014 | 0.011 | <0.0001 |
P-value was adjusted for age, site, race and physical activity.
The interaction was to test the difference of point estimates between the non-obese and the obese.