Baseline characteristics of the study population are presented in Table 1. The mean age was 61.6 years. The majority of participants where non-Hispanic White (42 %), followed by Hispanic/Latino (27 %), African American (17 %) and Chinese American (14 %). On average, men were overweight with a mean BMI of 27.6 kg/m2. The participants reported an average of 12 hours a week of physical activity. Moreover, 42% of participants were hypertensive, 13 % stated active cigarette smoking, 15 % had diabetes mellitus, and 24% were taking a cholesterol-lowering medication. The mean total testosterone level was 15 nmol/L.
Association between Sex Hormones and Abdominal Muscle Area
Total testosterone was significantly associated with total abdominal muscle area in the first model (B=1.39, 95 % CI 0.0 - 2.8, p = 0.05), which was accentuated with further adjustment (Model 2: 1.81, 0.2 - 3.5, p = 0.03; Model 3: 1.79, 0.1 - 3.4 p < 0.01) (Table 2). No significant associations were found between total testosterone and abdominal stabilizing muscle area (Table 3), while the associations were significant in all models for abdominal locomotor muscle area (Table 4).
Levels of estradiol were significantly associated with total abdominal muscle area in all three models: Model 1 (2.14, 0.8 - 3.6, p < 0.01), Model 2 (1.97, 0.6 - 3.4, p < 0.01), Model 3 (1.79, 0.4 - 3.2, p = 0.01) with similar results for abdominal locomotor and abdominal stabilizing muscle area (Table 3 and Table 4).
No significant associations were found between levels of free testosterone and total abdominal muscle areas (Table 2), as well as stabilizing. However, significant associations were presented between free testosterone and locomotor area in model 1 & model 2(0.38, 0.0 – 0.7, p = 0.04, 0.37, 0.0 – 0.7, p = 0.04) with borderline significance in model 3 (0.37, -0.0, 0.7, p = 0.05), respectively (Table 2, 3 & 4).
Associations between Sex Hormones and Abdominal Muscle Radiodensities
Total testosterone was significantly associated with total abdominal muscle radiodensity in Models 2 and 3, but not in Model 1 (Model 1: 0.04, -0.2 - 0.3, p = 0.79; Model 2; 0.32, 0.1 - 0.7, p = 0.04; Model 3: 0.3, 0.0 - 0.6, p = 0.04) (Table 2). Similar results were found for radiodensity of stabilizing muscles (Table 3), but not for locomotor muscle (Table 4).
No significant associations were found between free testosterone and abdominal muscle radiodensities in fully adjusted models (Table 2, 3 & 4).
No significant associations were found between levels of estradiol and total abdominal and stabilizing muscle radiodensity (Table 2 & 3), but there was a borderline significant association between estradiol and abdominal locomotor muscle radiodensity ((Model 1: -0.27, -0.0 - 0.6, p = 0.09: Model 2: 0.28, -0.0 - 0.6, p = 0.07; Model 3: 0.26, - 0.0 - 0.6, p = 0.09) (Table 4).
Higher SHBG levels were associated with a lower radiodensity of abdominal muscle in all models (Model 1: -0.35, -0.6 - -0.1, p = 0.02: Model 2: -0.35, -0.6 - -0.1, p = 0.02; Model 3: -0.34, -0.6 - -0.1, p = 0.02) (Table 2). The results were similar for abdominal stabilizing and locomotor muscles (Table 3).
Associations between Sex Hormones and Abdominal Muscle Area Indexes
A significant association was found in all models for total testosterone and TAMAi (Model 1: B= 0.10, 0,0 - 0.2, p < 0.01, Model 2: 0.11, 0.1 - 0.2, p < 0.01, Model 3: 0.10, 0.1 - 0.2, p <0.01) (Table 2). That is, in fully adjusted models, one SD increase in testosterone levels resulted with an increase of 0.10 cm2/(weight/height2) in abdominal muscle area index. Similar relationships were observed between total testosterone and abdominal locomotor and stabilizing muscle area indices (Table 3 & 4).
Estradiol was found to be significantly associated with total abdominal muscle index (TAMAi) in model 1 (B = 0.10, 0.0 - 0.1, p= 0.03), which was borderline significant in Models 2 and 3 (Model 2: 0.05, -0.0 - 0.1, p= 0.06, Model 3: 0.05, -0.0 - 0.1, p= 0.06). Significant associations were found for estradiol and abdominal locomotor muscle area index in all three models but not for abdominal stabilizing muscle area index (Table 3 and 4).
Significant associations were shown for free testosterone with total, stabilizing and abdominal muscle area index in fully adjusted models (0.08, 0.0, 0.1, p = 0.008, 0.05, 0.0 – 0.1, p = 0.03, 0.02, 0.0 - 0.04, p = 0.02), respectively. Inverse non-significant associations were found between levels of SHBG and abdominal muscle areas, and muscle area indexes in both model 2 and model 3.