Muscle Area and Hounsfield Unit of Muscle/Liver Measured By Quantitative Computed Tomography Are Significantly Associated with Insulin Resistance and Low Bone Mass in Postmenopausal Women

Presentation Number: SAT 316
Date of Presentation: April 1st, 2017

Hye-Sun Park*1 and Sung-Kil Lim2
1Yonsei University College of Medicine, Korea, Republic of (South), 2Yonsei University College of Medicine, Seoul, Korea, Republic of (South)


To evaluate whether muscle area and lipid accumulation in muscle/liver are related with insulin resistance and bone mineral density (BMD), we performed quantitative computed tomography (QCT). 149 postmenopausal women > 60 years old were enrolled. Serum fasting glucose and insulin were measured and QCT was performed. The cross sectional areas (CSA) of the muscle were obtained using free-hand drawn regions of interests (ROI). Areas of gluteus and quadriceps were measured at the level of symphysis pubis and at the level of 7cm from the lesser trochanter, respectively. Hounsfield units (HUs) of quadriceps and gluteus were measured at ROIs and those of liver were measured at the liver segment VII. Hand grip test (HGT) and short physical performance battery (SPPB) were performed to estimate muscle strength. The mean age of enrolled subjects was 72.37 years, and BMI was 23.64 kg/m2. Smaller muscle area and lower HU of muscle/liver were related with higher insulin resistance. Both gluteus and quadriceps area were inversely related with HOMA-IR (r=-0.499 and -0.247, respectively) and fasting glucose level (r=-0.409 and -0.271, respectively). HU of gluteus, quadriceps and liver were also inversely related with HOMA-IR (r=-0.265, -0.289 and -0.506, respectively), not with HOMA-β. They were all statistically significant (p < 0.05). In contrast, HGT and SPPB were not associated with insulin resistance (p = 0.562 and 0.801, respectively). Furthermore, we compared the parameters between osteoporosis group and non-osteoporosis group. The mean value of muscle area was significantly lower in osteoporosis group. The mean value of gluteus area of osteoporosis group was 3125.20 ± 609.35 mm2, whereas 3421.46 ± 534.38 mm2 in non-osteoporosis group (p=0.004). The same trend was found in quadriceps. The mean value of quadriceps area in osteoporosis group was 3301.87 ± 506.31 mm2 and in non-osteoporosis group, it was 3514.98 ± 607.69 mm2 (p=0.032). In addition, HU of liver was also statistically different between the two groups. In osteoporosis group, liver HU was 55.90 ± 8.89, whereas in non-osteoporosis group, it was 59.87 ± 6.17 (p=0.002), which means higher fat content of liver in osteoporosis group. In osteoporosis group, hand grip strength was lower than in non-osteoporosis group (p = 0.011), however, no statistical significance was found on SPPB. In conclusion, parameters from QCT, such as muscle area and HU of muscle/liver, are clearly associated with high insulin resistance and low bone mineral density. There is lack of study regarding interrelation between muscle, liver, and bone altogether. This study shows that QCT is a simple and easy tool in evaluating interaction between insulin target organs. However, further investigations are needed to elucidate their causal relationship.


Nothing to Disclose: HSP, SKL