Metabolic diseases have been related to the body composition (BC) pattern, which is defined as the combination of variables that describe an individual’s distribution of fat and/or muscle, quantifying in vivo the body components, the quantitative relationships between the components, and their quantitative changes related to influential factors (1).

It is well known that due to the increase in the prevalence of obesity and diabetes mellitus (DM), other metabolic diseases have also increased; one of the most important is Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD), which nowadays is the most important chronic liver disease and one of the major indication for liver transplant worldwide, with an estimated prevalence of 30% (2, 3).

The inflammatory factors are one of the most important players in the relationship between metabolic diseases and BC (4). In MASLD patients, changes in BC are related to insulin resistance (IR), increase of lipolysis, and fatty acids accumulation in liver tissue (5); moreover, alterations in BC have been associated with an increased risk of presence and progression of MASLD (6–8).

In MASLD patients, BC assessment is important since it is associated with hepatic fat percentage and progression of liver disease (9, 10). Altered BC, characterized by high-fat tissue and low muscle mass, has been related to functional performance and metabolic comorbidities in patients with MASLD, especially in those with DM and cardiovascular diseases, which also get worse with fat tissue increase (10).

Whereas the Dual Energy X-ray Absorptiometry (DEXA) is the reference method for BC assessment (11), the Bioelectrical Impedance Analysis (BIA) has shown good concordance with DEXA, and it has been proposed as a good method of BC assessment in obese patients in whom the physiological and hydration conditions could interfere with measurement reliability (5, 12, 13). BIA is a noninvasive and relatively available method for BC analysis based on measuring resistance (R) and reactance (Xc), which allows to determine the fat and muscle percentage and hydration state through bioelectrical impedance vector analysis (BIVA) (14, 15).

While the relationship between changes in BC and the presence of liver steatosis has been established, there is little evidence about the characteristics of BC in patients with MASLD and differences related to body mass index (BMI); therefore, the aim of this study was to determine differences in BC according to BMI and metabolic comorbidities in patients with MASLD.

A total of 316 patients were included: BMI <25 kg/m² non-MASLD (n = 70), BMI <25 kg/m²-MASLD (n = 36), overweight-MASLD (n = 70), MHO-MASLD (n = 70), and MUO-MASLD (n = 70). 59.6% (n = 189) were male with a mean age of 49 ± 10 years; at the time of evaluation, 6.9% (n = 22) had a known diagnosis of DM and 16.7% (n = 53) had a known diagnosis of high blood pressure. Concerning metabolic risks, a decreased High-density Lipoprotein (HDL) level was the most prevalent (38.2%, n = 121), followed by abnormal triglycerides (34.1%, n = 108), and glucose impairments (28.1%, n = 89). The mean of dB/m was 287.9 ± 55.2; meanwhile, the mean of kPa was 3.7 ± 0.8. 1.2% (n = 4) of patients have significant fibrosis (8.0–11.9 kPa). General characteristics of patients are presented in Table 1.

Regarding the analysis of BC differences among all groups (Table 2), R and Xc show significant differences (p ≤ 0.0001), and PA did not show differences among groups. As expected, Fat% was significantly increased in terms of BMI (p ≤ 0.0001); however, in post-hoc analysis, Fat% was not different between the BMI <25 kg/m²-MASLD (34.9 ± 6.7%) and Overweight-MASLD (36.5 ± 6.3%) groups. SMM% was significantly lower in obesity groups with respect to overweight and normal weight groups (p < 0.05); no significant differences were observed among overweight and normal weight groups. Despite the differences among all groups (p ≤ 0.001), no significant differences were observed between BMI <25 kg/m² groups with (32.4 ± 6.2 kg) or without MASLD (33.5 ± 5.9 kg), and Overweight MASLD (38.1 ± 7.2 kg) compared to MHO-MASLD group (41.3 ± 8.4 kg) regarding water-related components. ECW was significantly higher according to BMI increase; however, no differences were observed according to metabolic health or unhealth in obesity groups. ECW/ICW ratio was significantly higher in the MHO-MASLD group and MUO-MASLD group compared with the BMI <25 kg/m² non-MASLD group (p = 0.001 and p = 0.02, respectively), and with the BMI <25 kg/m²-MASLD group (p = 0.001 and p = 0.01, respectively; Figure 1D).

Once again, only in MASLD groups (n = 246) all components showed differences in the One-way ANOVA test, except PA, where differences in BC were analyzed. Fat% was higher according to BMI; however, there was no difference between the BMI <25 kg/m²-MASLD group and the Overweight-MASLD group (34.9 ± 6.7% vs. 36.5 ± 6.3%, p = 1.00), nor between the MHO-MASLD group and the MUO-MASLD group (40.6 ± 8.1 vs. 40.3 ± 6.0%, p = 1.00) in post-hoc analysis. SMM% was significantly different among groups (p = 0.01), being higher in BMI <25 kg/m²-MASLD group and Overweight-MASLD group than in Obesity Groups (30% vs. 28%), but no significant differences were observed in post-hoc analysis (Figure 1A).

ICW was significantly higher only among the BMI <25 kg/m²-MASLD and MHO-MASLD (18.5 ± 4.0 kg vs. 21.9 ± 5.2 kg, p = 0.003) and MUO-MASLD groups (18.5 ± 4.0 kg vs. 22.8 ± 5.0, p = 0.0002; Figure 1B). Instead, ECW was significantly different among all groups, increasing in terms of BMI, but once again, without difference in obesity groups (Figure 1C). When the ECW/ICW ratio was analyzed, we observed an increase according to BMI; however, the BMI <25 kg/m²-MASLD group only showed significant differences with MHO-MASLD and MUO-MASLD groups (0.80 ± 0.1 vs. 0.90 ± 0.1, p = 0.005, and 0.80 ± 0.1 vs. 0.89, p = 0.04, respectively), whereas the Overweight-MASLD group only showed differences with the MHO-MASLD group (0.85 ± 0.1 vs. 0.90 ± 0.1, p = 0.04; Figure 1D).

The differences in water components were confirmed with the BIVA qualitative analysis. According to the RXc point graphic and tolerance ellipses, with the increase of BMI, the points were situated in vectors that represent more fluids but not necessarily in those that represent fewer lean tissues (21, 22) (Figure 2). Regarding the BIVA tissue classification, normal tissue was majorly prevalent in the BMI <25 kg/m² non-MASLD (70.4%, n = 50/70) group, BMI <25 kg/m²-MASLD (83.3%, n = 30/36) group, and Overweight-MASDL (70%, n = 49/70) group; however, it was decreased in the MHO-MASLD (51.4%, n = 36/70) group and MUO-MASLD (45.7%, n = 32/70) group. The prevalence of sarcopenia-cachexia tissue was higher in BMI <25 kg/m² groups (20%), and lower in Overweight-MASLD (4.2%, n = 3) group, MHO-MASLD group, and MUO-MASLD group (2.9%, n = 2, both). Conversely, overhydration was higher in Overweight-MASLD (17.1%, n = 12) group, MHO-MASLD (35.7%, n = 25) group, and MUO-MASLD (37.1%, n = 26) group. In the BMI <25 kg/m² non-MASLD group, only 7% (n = 5) presented overhydration, and it was not present in the BMI <25 kg/m²-MASLD group (Figure 3).

Factors associated with abnormal BC were analyzed, according to the ECW/ICW ratio and the BIVA tissue classification. In all patients (n = 316), female sex, abnormal WC, abnormal HDL levels, and the presence of liver steatosis were independent factors associated with abnormal BC, according to ECW/ICW ratio (Table 3); as for BIVA, female sex [OR 3.3 CI95% (1.9–5.5), p ≤ 0.001] and abnormal WC [OR 2.2 CI95% (1.3–3.9), p ≤ 0.001] were independently associated with abnormal BC (Table 3). We performed the bivariate and multivariate analyses adjusting by sex, and then the abnormal WC and the presence of liver steatosis maintained the independent association with abnormal BC, according to ECW/ICW ratio (Table 4); as for BIVA, only abnormal WC [OR 2.7 CI95% (1.4–5.2), p = 0.003] was independently associated with abnormal BC (Table 4). When these analyses were carried out, only in MASLD groups, female sex, and abnormal WC were independent factors associated with abnormal BC in both criteria (data not shown); in adjusted analysis by sex, abnormal WC maintained the independent association with BIVA [OR 4.3 CI 95% (1.8–10.1), p = 0.001] as reference, but abnormal WC only showed association in bivariate analysis [OR 3.6 CI 95% (1.8–7.3), p = 0.001] with ECW/ICW ratio as reference.

The evidence of alterations in body composition in the MASLD scenario is scarce. In our study, changes in BC in patients with MASLD were observed, with significant differences compared to healthy patients (BMI < 25 kg/m² and non-MASLD). As expected, Fat%, TBW, and ECW/ICW ratio were increased according to BMI increase, and conversely, SMM was decreased. However, significant differences were not observed among all groups and in MASLD groups in post-hoc analysis.

Fat accumulation is now considered a major risk factor for mortality, independent of obesity (23). A significant increase in Fat% has been observed in patients with MASLD and BMI < 25 kg/m²; this has been observed in the United States population by Mainous III et al. (24) and in the Rotterdam cohort (OR 1.77, p ≤ 0.05) (25). In our population, we observed a significant difference in the increase of Fat% in patients with BMI <25 kg/m², being higher in those with MASLD. The prevalence of MASLD in patients with BMI <25 kg/m² is relatively low; in our population, we previously reported a prevalence of 7.9% (26), even though the BMI < 25 kg/m²-MASLD group is smaller than the other groups. One of the strengths of our study is the inclusion and comparison of this group of patients with other MASLD phenotypes, taking their low prevalence into account. BC assessment could be an early detection tool in these patients in whom MASLD is not an initial clinical suspicion. Another strength of this study is the BIVA analysis, which is the qualitative point of view of BC. As far as we know, it has not been evaluated in patients with MASLD. According to our results, changes observed in BC are consistent with BIVA in overhydration and lean mass tissue terms, according to the tolerance ellipses and BIVA tissue classification (Figures 2, 3).

Abdominal fat accumulation seems to be a better indicator of MASLD than BMI or the presence of obesity (27). In our study, we observed a significant increase in Fat% in both BMI <25 kg/m² and overweight/obese patients. On the other hand, WC was an independent factor associated with altered BC in all patients and also when only MASLD patients were analyzed, without differences among the number of comorbidities in obese patients.

Sarcopenia increases in MASLD and is considered a progression factor independent of obesity and IR. Muscle strength was not evaluated in our study, so we cannot use the sarcopenic obesity concept (decreased muscle mass, increased fat, and decreased muscle strength) (28). We refer to myopenia instead, which exclusively refers to low muscle mass (23); in obese patients, it will be myopenic obesity.

In patients with MASLD, a decrease in SMM has been associated with BMI, Fat Mass Index, and WC, with significant differences regarding sex, majorly attributed to hormones. Onishi et al. (10) evaluated the associated factors to SMM decrease in patients with MASLD, finding that BMI, Fat-free mass Index, and WC were independent associated factors. However, the study was conducted in an Asian population with a different BMI cut-off to determine overweight. Despite this, our results confirm that SMM is significantly lower in patients with MASLD according to BMI increase; however, despite detecting a trend, no significant differences were observed in terms of BMI classification or the presence of comorbidities in obese patients in a post-hoc analysis. Statistical significance could be lost since our study universe corresponds to an open-apparently healthy population that attended a check-up unit with an overall mean age (49.2 ± 10.5) and stage of liver steatosis (287.9 ± 55.2) in which significant muscle loss is not expected. However, this observed trend is clinically significant for early recognition of decreased SMM.

As for water measurements, we observed a significant decrease in the TBW percentage, according to BMI increase (Table 2). TBW percentage has been observed to reflect higher levels of adiposity, and this could affect the reliability of measurements of fat-free mass. However, this could produce a clinical underestimation of obesity if only TBW is considered for body composition assessment (20). Therefore, the evaluation of TBW components is a more reliable measurement, especially in obese patients, since one of the characteristics of obesity is an alteration in fluid regulation; changes in ECW and ICW have been attributed to the high proportion of ECW in adipose tissue, the relationship of ECW with chronic inflammation (29), obesity-related edema, and hormonal responses to fat tissue, leading to a primary deficiency in hemodynamic fluid regulation that could not be reversible in morbid obesity (30, 31).

The increase in water components of BC at the expense of increased fat could be the explanation for the difficulty of muscle mass recovery, even in lean patients; moreover, this fluid alteration seems to persist after weight loss becomes irreversible (30, 32).

Although it is interesting to highlight that the ECW/ICW ratio has demonstrated to be an overload water marker and, even more, a mortality marker in populations different than ours and in cardiovascular risk populations (15, 33–37), there is no evidence of this ratio in MASLD patients. However, different studies (33, 36, 37) show that this ratio could be an early marker of muscle mass and function loss. In our study, this ratio was higher in patients with BMI >25 kg/m²; therefore, if we evaluate it with SMM percentage, even if no statistical difference was observed, it could be considered an early marker of sarcopenia.

Regarding PA, Chen et al. (38) observed that it is lower in patients with MASLD compared to non-MASLD patients. When the analysis was adjusted by BMI, sex, and comorbidities, PA was associated with MASLD risk; however, this association was not observed in patients with BMI >30 kg/m². The authors concluded that PA could be an indicator in MASLD management limited to overweight patients.

Abnormal BC has been established as a risk factor and as an indicator for the presence of liver steatosis (8, 39). According to our results, liver steatosis is an independent factor associated with altered BC when it is defined by the ECW/ICW ratio in multivariate analysis. Therefore, changes in BC seem to be one more factor affected by MASLD development.

From the anthropometric point of view, the assessment of MASLD patients would need deeper indicators than BMI, including BC analysis, which seems to be a tool for patients’ diagnosis, classification, muscle mass measurement, and follow-up. Improvement of BC has been related to a decrease in liver fat content in patients with MASLD (40). Currently, there is insufficient evidence to assess whether newest treatments that have demonstrated to reverse steatosis or fibrosis also impact BC. Although weight loss is the cornerstone of MASLD treatment, it is important to evaluate whether treatment schemes could have a “negative” impact on BC, especially in those patients with increased Fat% and decreased SMM in whom weight loss without improvement or maintenance of SMM could remain a risk for metabolic and cardiovascular mortality, despite weight loss.

BC in MASLD patients varies according to BMI increase; changes could be explained by loss of SMM and not necessarily by the presence of metabolic abnormalities. High WC and the presence of steatosis are independent factors associated with altered BC.