Basic laboratory and clinical characteristics
Basic laboratory and clinical characteristics of the 578 participants (92 men and 486 women) with and without NAFLD enrolled in this study are shown in Table 1. Thirty (32.6%) men and 124 (25.5%) women participants had NAFLD. Among the men, the participants with NAFLD were younger and had higher body weight, BMI, serum TG, ALT compared with those without NAFLD, but there were no differences in height, serum TC, HDL-C, LDL-C, AST, GGT, ALB, Cr, FPG and UA between the two groups. Among the women, the participants with NAFLD were older and had higher weight, BMI, TG, serum AST, ALT, GGT, ALB, FPG, and UA, and lower HDL-C compared with those without NAFLD. However, there were no differences in height, serum TC, LDL-C and Cr between the two groups. As illustrated in Table 1, the levels of GS were lower in NAFLD patients, compared with NAFLD-free participants both in men and women (men: 1.1 ± 0.3 vs. 1.3 ± 0.3 kg/m2, P = 0.008; women: 1.1 ± 0.3 vs. 1.2 ± 0.2 kg/m2, P = 0.008). The levels of muscle mass were also significantly lower in NAFLD patients, compared with NAFLD-free participants (men: 6.6 ± 1.2 vs. 7.2 ± 1.3 kg/m2, P = 0.032; women: 5.2 ± 1.3 vs. 5.6 ± 1.1 kg/m2, P = 0.011). Whereas, there was no statistical difference in the levels of HGS between the two groups (men: 35.8 ± 8.6 vs. 37.6 ± 9.7, P = 0.404, women: 23.4 ± 5.8 vs. 24.5 ± 5.1 kg, P = 0.053). The prevalence of sarcopenia was higher among the participants with NAFLD (men: 20.0% vs. 9.7%, P = 0.295, women: 15.3% vs. 8.0%, P = 0.019). Moreover, the prevalence of sarcopenic obesity was also higher in the NAFLD group both in men and women, but without statistical difference (men: 10.0% vs. 3.2%, P = 0.394, women: 3.2% vs. 0.8%, P = 0.134).
Associations of muscle mass and anthropometric, biochemical variables of NAFLD
We performed Pearson correlation analysis to determine the correlations between muscle mass and anthropometric, biochemical variables of NAFLD. The correlation analyses between anthropometric, biochemical variables and muscle mass in participants were shown in Table 2. We found that muscle mass was positively correlated with body weight (𝑟 = 0.292, P = 0.005), BMI (𝑟 = 0.291, P = 0.005), and HGS (𝑟 = 0.315, P = 0.002), while negatively correlated with age (𝑟 = − 0.244, P = 0.019) and FPG (𝑟 = − 0.251, P = 0.016) among men participants. Meanwhile, muscle mass was positively correlated with age (𝑟 = 0.111, P = 0.015), body weight (𝑟 = 0.295, P<0.001), BMI (𝑟 = 0.326, P<0.001), serum ALT (𝑟 = 0.139, P = 0.002), UA (𝑟 = 0.142, P = 0.002), and FPG (𝑟 = 0.111, P = 0.015) among women participants (Table 2).
Associations of HGS and anthropometric, biochemical variables of NAFLD
We found that HGS was positively correlated with body weight (𝑟 = 0.255, P = 0.014), height (𝑟 = 0.51, P<0.001), GS (𝑟 = 0.408, P<0.001) and muscle mass (𝑟 = 0.315, P = 0.002), while negatively correlated with age (𝑟 = − 0.504, P<0.001) among men participants. Meanwhile, HGS was positively correlated with body weight (𝑟 = 0.206, P<0.001), GS (𝑟 = 0.538, P<0.001) and height (𝑟 = 0.315, P<0.001), while negatively correlated with age (𝑟 = − 0.533, P<0.001), serum AST (𝑟 = − 0.391, P<0.001), TG (𝑟 = − 0.144, P = 0.002), UA (𝑟 = − 0.193, P<0.001), and FPG (𝑟 = − 0.177, P<0.001) among women participants. (Table 3)
Independent impact of muscle mass and muscle strength on the presence of NAFLD
A logistic regression model was conducted to evaluate the gender-specific relationship between the components of sarcopenia (LMM or LMS) and NAFLD risk (models 1–2, Table 4). The relationship between LMM and NAFLD was statistically significant in both models. In model 1, the OR with 95% CI for NAFLD were 2.91 (95%CI, 1.58–5.35) and 1.89 (95%CI 1.51–2.38) in men and women, respectively. Further, the fully adjusted model (model 2) showed LMM was still associated with an increased risk of NAFLD with OR of 2.88 (95%CI 1.52–5.46) in men and 2.08 (95%CI 1.63–2.67) in women. Participants with LMS showed significantly high odds of NAFLD with OR of 1.15 (95% CI, 1.04–1.26) and 1.15 (95%CI 1.02–1.28) in men after adjusting in Model 1 and Model 2. However, there was no statistical significance in muscle strength in women after adjusting in Model 1 and Model 2, with OR of 1.05 (95%CI 0.99–1.11) and 1.01 (95%CI 0.95–1.07), respectively.
ROC curve of muscle mass
The ROC curves of muscle mass plotted for the diagnoses of NAFLD by gender were shown in Fig. 2. The cut-off value of muscle mass was 8.0 kg/m2 in men and 4.9 kg/m2 in women, with the sensitivity of 33.9% and 68.8%, and the specificity of 90.0% and 47.6%, respectively. The areas under ROC for NAFLD were 0.624 (95% CI 0.501–0.748, P < 0.063) and 0.592 (95% CI 0.531–0.653, P < 0.031) in men and women, respectively.