Fig. 1Bone mineral density (BMD) values according to plasma macrophage migration inhibitory factor quartiles after adjusting for confounding variables in the (A) lumbar spine, (B) femoral neck, and (C) total femur. Confounding variables included age, body mass index, current smoking, alcohol intake (≥3 units/day), exercise status (≥30 min/day), parental history of osteoporotic fracture, years since menopause, and 25-hydroxyvitamin D3. The mean estimated BMDs (95% confidence interval) were calculated by analysis of covariance (ANCOVA). The P values for the trends were generated using multiple linear regression analysis. aStatistically significantly different from the lowest quartile (Q1) by ANCOVA.
Fig. 2Bone-specific biochemical turnover markers (BTMs), such as (A) osteocalcin and (B) C-terminal telopeptide of type I collagen (CTX) according to the plasma macrophage migration inhibitory factor quartiles after adjusting for confounding variables. Confounding variables included age, body mass index, current smoking, alcohol intake (≥3 units/day), exercise status (≥30 min/day), parental history of osteoporotic fracture, years since menopause, and 25-hydroxyvitamin D3. The mean estimated BTMs (95% confidence interval) were calculated using analysis of covariance (ANCOVA). The P values for the trends were generated using multiple linear regression analysis. aStatistically significantly different from the lowest quartile (Q1) by ANCOVA.
Table 1Baseline Characteristics of the Study Population
Variable |
Postmenopausal women (n=246) |
Age, yr |
62.7±6.2 |
Height, cm |
155.4±5.2 |
Weight, kg |
57.0±7.0 |
Body mass index, kg/m2
|
23.6±2.9 |
Current smoker |
7 (2.8) |
Alcohol intake (≥3 U/day) |
37 (15.0) |
Exercise (≥30 min/day) |
94 (38.2) |
Parental history of OF |
37 (15.0) |
Years since menopause, yr |
12.1±6.8 |
MIF, ng/mL |
25.0±20.4 |
Bone mineral density, g/cm2
|
|
Lumbar spine |
0.894±0.123 |
Femoral neck |
0.756±0.094 |
Total femur |
0.813±0.097 |
Corrected calcium level, mg/dLa
|
8.9±0.4 |
Phosphorus, mg/dL |
3.8±0.5 |
Bone turnover marker |
|
Osteocalcin, ng/mL |
30.5±14.3 |
CTX, ng/mL |
0.560±0.223 |
25-OH-D3, ng/mL |
30.5±15.5 |
Table 2Correlation of Plasma Macrophage Migration Inhibitory Factor Levels with Bone Mineral Density and Bone Turnover Markers
Variable |
Unadjusted |
Age and BMI adjusted |
Multivariable adjusted |
γ |
P valuea
|
γ |
P valueb
|
γ |
P valuec
|
Bone mineral density, g/cm2
|
|
|
|
|
|
|
Lumbar spine |
–0.327 |
<0.001 |
–0.335 |
<0.001 |
–0.317 |
<0.001 |
Femoral neck |
–0.224 |
<0.001 |
–0.236 |
<0.001 |
–0.223 |
0.001 |
Total femur |
–0.229 |
<0.001 |
–0.245 |
<0.001 |
–0.232 |
<0.001 |
Bone turnover markersd |
|
|
|
|
|
|
Osteocalcin |
0.186 |
0.015 |
0.189 |
0.014 |
0.187 |
0.017 |
CTX |
0.217 |
0.001 |
0.218 |
0.001 |
0.196 |
0.004 |
Table 3Differences in Plasma MIF Levels according to Osteoporosis Status
Adjusted model |
Estimated MIF levels, ng/mL (95% CI)a
|
Subjects with osteoporosis (n=94) |
Subjects without osteoporosis (n=152) |
P valueb
|
Unadjusted |
29.08 (24.98–33.19) |
22.42 (19.19–25.64) |
0.012 |
Age and BMI adjusted |
28.94 (24.82–33.05) |
22.51 (19.27–25.74) |
0.017 |
Multivariable adjustedc
|
28.22 (24.09–32.35) |
22.72 (19.50–25.94) |
0.041 |
Table 4Multiple Logistic Regression Analyses to Determine the Ability of Plasma MIF Levels to Predict Risk of Osteoporosis
|
OR (95% CI)a
|
P value |
Unadjusted |
1.387 (1.065–1.806) |
0.015 |
Age and BMI adjusted |
1.371 (1.051–1.788) |
0.020 |
Multivariable adjusted |
1.319 (1.007–1.729) |
0.044 |