Association between bioelectrical impedance analysis parameters and muscle-related measurements obtained using a non-diagnostic ultrasound-based technique

Abstract

Background

An in-depth analysis of body composition (BC) is fundamental for the assessment of nutritional status. A multifactorial approach is important for this assessment, integrating parameters and measurements obtained through different techniques, most notably bioelectrical impedance analysis (BIA) and ultrasound. In this study, we aimed to evaluate associations between BIA parameters and ultrasound measurements.

Methods

The BC of 993 adults (539 women and 454 men) was evaluated using BIA (BIA 101 RJL, Akern Bioresearch, Florence, Italy) and a non-diagnostic ultrasound-based technique at the thigh, employing a 2.5 MHz, A-mode probe (BodyMetrixTM, IntelaMetrix, Inc., Livermore, CA, USA). Ultrasound scans were analysed using Adipometria v 1.8.8 software (Hosand Technologies s.r.l., Verbania, Italy) and the muscle echogenicity (EG) was obtained. Ultrasound measurements were correlated with BIA parameters.

Results

In the total population, muscle EG was directly correlated with the BIA parameter resistance (r = 0.347) and inversely correlated with several BIA parameters, including intracellular water (r = −0.479 and − 0.386, % and L, respectively) and phase angle (r = −0.487). These associations were maintained even after stratification of subjects by sex.

Conclusion

The observation of significant associations between BIA parameters and ultrasound measurements suggests the importance of not considering the two methods as exclusive, but of integrating them into a holistic approach for the in-depth assessment of BC and nutritional status. Such an approach provides detailed information for the identification of optimal nutritional treatment, its monitoring over time, and possible optimisation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-025-06963-9.

Background

It is well known that careful assessment of body composition (BC) is fundamental for the clinical framing of the subject at baseline, as well as follow-up during nutritional interventions [1, 2]. The rapid detection of simple parameters such as weight or body mass index (BMI) provides useful information in nutritional screening, but the analysis of BC is indispensable, in order to estimate quantitative (and possibly qualitative) body components and nutritional status [2]. BC data can be used to optimize nutritional treatment. For this reason, bioelectrical impedance analysis (BIA) has been used for several decades as a valid and reliable method for the assessment of BC. Measuring resistance (Rz) and reactance (Xc) allows for the assessment of the hydration status of the subject (total body water, TBW) and intra- and extracellular fluid distribution (intracellular water [ICW] and extracellular water [ECW], respectively), considering that Rz is inversely proportional to TBW [3, 4]. Through the use of predictive multiple regression models, accurate estimates of BC are obtained [3, 5, 6].

An accurate assessment of BC and nutritional status, is crucial for all classes of subjects, but it is indispensable in specific settings, such as certain pathological states (e.g. obesity and endocrine-metabolic alterations) or sport practice, in which BC may be both compromised and varied. Under specific conditions, such as obesity or exercise, BIA parameters are altered, mostly due to changes in body fluids. Excess adiposity in subjects with obesity is responsible for an expansion of ECW, resulting in an increase in the ECW/ICW ratio [7, 8]. This alteration in fluid distribution, in addition to reflecting a state of cellular dehydration, is indicative of both oedema/water retention (mainly evident in the lower limbs) [9, 10] and subclinical systemic inflammation [2]. This complex scenario, from a BIA perspective, is reflected in increased Rz values and reduced Xc, as well as the Rz-Xc derived phase angle (PhA) [2]. With exercise training, on the other hand, diametrically opposite conditions occur. Sport, by leading to improvements in the function and integrity of cell membranes, and also to positive changes in intracellular components, is responsible for an increase in metabolically active muscle mass and strength [11–13]. Muscle is an excellent conductor, given the abundance of water and electrolytes. Consequently, the increase in muscle mass, concomitant with sport training, reduces Rz values [14, 15] and increases Xc [15, 16] and PhA [17]. In athletes, PhA can exceed the value of 8.5° [2]. The presence of alterations in BIA parameters in conceptually opposite conditions, such as those described above, with respect to conditions of normal weight and normohydration, implies the need not to consider them exclusively, but to integrate them with further parameters and measurements for a more correct interpretation.

In this context, although different techniques and parameters have been used for assessing and monitoring BC and nutritional status, the need to adopt a holistic and integrative approach, capable of providing a global, morphological and functional view of the subject, is becoming increasingly consolidated [18, 19].

As far as morphological and functional assessment are concerned, several authors suggest that the combination of BIA and tissue ultrasound contribute to a more detailed analysis of nutritional status and cardiometabolic risk with diagnostic and prognostic value in certain disease states [19]. Specifically, among BIA data, these authors recommend the evaluation of raw parameters, most notably the PhA [19], which has previously been indicated as a surrogate marker of cell damage and function [20, 21], inflammation [22], muscle abnormalities and function [18], muscle strength [23] and, by extension, overreaching and overtraining [24]. The rationale behind this predictive role of PhA lies in the changes in the resistive (Rz) and capacitive (Xc) components, resulting from the hydro-electrolyte imbalance that occurs between intra- and extracellular spaces in response to inflammatory and oxidative states [25]. This condition, deriving from an increase in the release of inflammatory cytokines and oxidising substances, is reflected in a reduction in the values of BIA raw parameters - Rz and Xc - resulting in a mathematical reduction in PhA [24]. If these physical principles of PhA are translated into the context of tissue assessment, their further value as indicator of muscle quality becomes clear [26].

Although the concept of muscle quality has not yet been precisely clarified, as pointed out by a European consensus on the diagnosis and definition of sarcopenia, it refers to changes in the architecture and composition of muscle [26]. Interestingly, this document suggests PhA [26] and ultrasound [26, 27] as tools for assessing muscle quality. With regard to ultrasound, particular emphasis is placed on measurements of both muscle thickness and echogenicity (EG) [27]. Muscle EG values are higher in response to increased tissue sound reflexes, which in turn are due to alterations in muscle architecture, fibrosis, intramuscular fat infiltration or inflammation [28].

The integrative approach in the use of BIA and ultrasound finds further application due to the relative limitations of both methods. In the case of BIA, the estimation of BC is based on the assumption of a constant hydration of the fat-free mass (FFM). Further, this method primarily provides global information on body components, without offering direct indications about their specific characteristics (e.g. hydration level) or their distribution (e.g. fat and muscle thickness) across individual body districts. Conversely, the ultrasound method provides site-specific BC data. However, although ultrasound assessment of EG may offer indirect indications of tissue hydration status, it does not allow for quantitative information about it, as BIA. This highlight once again the importance of integrating BIA and ultrasound data, in order to enhance and contextualise their interpretation.

Noteworthy, previous evidence has shown correlations between muscle EG and BIA parameters, such as body cell mass, Rz and PhA [18, 29–32], corroborating the association between BIA and ultrasound. In particular, it has been reported that PhA is negatively correlated with traditional ultrasound-measured muscle EG in patients with disease-related malnutrition [29], and in women with obesity [32]. On the other hand, in patients with amyotrophic lateral sclerosis, muscle EG positively correlated with Rz, although no significant correlations have been found with PhA [31]. Collectively, thus, this evidence suggests that integrating ultrasound with BIA may provide insights for BC and nutritional status assessments. However, these studies were conducted using ultrasound devices intended for medical-diagnostic use, which are often expensive, not easily transportable, and unsuitable for use by non-medical health professionals, such as nutritionists, dietitians or trainers. To overcome these device-related limitations, non-diagnostic ultrasound-based devices designed for the qualitative and quantitative, non-clinical assessment of subcutaneous tissues are currently used in research and clinical practice [33]. Ultrasound measurements obtained with this device have been validated against conventional ultrasound in cadaver [34] and living humans [35] studies, and average values have been recently reported in a large cohort of subjects stratified by sex, age and BMI [36]. Furthermore, a recent study demonstrated the intra- and inter-operator reliability of fat and muscle measurements performed with this portable ultrasound-based device [37].

The main aim of the present study is to identify for the first time, in a large cohort of women and men, possible associations between BIA parameters and muscle-related ultrasound measurements obtained by means of non-diagnostic ultrasound-based technique. Compared with the data already available in the literature, the results of our study broaden the understanding of the association between BIA parameters and ultrasound measurements through the evaluation of muscle EG. Given the influence of hydration status in the assessment of BC, the associations were adjusted for the TBW/FFM ratio, allowing for cleaner information that is not tainted by possible hydro-electrolyte imbalances. Overall, the results of this study could pave the way for the development of predictive models that integrate BIA parameters and ultrasound measurements, enabling a more detailed analysis and monitoring of BC and nutritional status, and providing an additional tool to nutritionists and athletic trainers.

Material and methods

Design and setting

In the present retrospective observational study, anthropometric measurements and BC data from 1,010 adults (aged 18–79 years and with a BMI ranging from 15.98 to 55.36 kg/m²) were analysed, in compliance with the study protocol approved by the Local Ethics Committee of Università Telematica Pegaso, resolution no. PROT./E 005082 of 19/07/2024. All participants provided written informed consent in accordance with current legislation. Among the study participants, 17 subjects with BMI < 18.50 kg/m² (14 women and 3 men) were excluded, thus a final sample of 993 subjects was included in the statistical analysis (Fig. 1).

Fig. 1.

Flowchart of the study

Body composition assessment

Anthropometric measurements

Anthropometric measurements were assessed by qualified nutritionists. All participants wore only underwear, without shoes. Body weight and height were measured to the nearest 0.1 kg and 0.5 cm, respectively, using a calibrated beam scale and wall stadiometer. BMI was calculated according to the formula: weight (kg)/height² (m²). Based on the obtained values, subjects were stratified into the following BMI categories, according to the World Health Organization classification [38]: 18.50–24.99 kg/m² (normal weight), 25.0–29.99 kg/m² (overweight) and greater than or equal to 30.00 kg/m² (obesity).

Bioelectrical impedance analysis (BIA) parameters

The BIA was conducted using an 800A current, 50 kHz frequency device (BIA 101 RJL, Akern Bioresearch, Florence, Italy). The examination was conducted by qualified nutritionists, following the guidelines of the European Society of Parenteral and Enteral Nutrition (ESPEN) [39], after checking the device with resistors and capacitors of known value. Specifically, participants (fasting, resting from exercise for at least 6 h and with no alcohol intake in the 24 h prior to testing) were supine, with both the lower and upper limbs slightly apart from the body. After cleansing the skin with alcohol, electrodes (BIATRODES Akern Srl; Florence, Italy) were placed on the hand (one at the level of the phalangeal-metacarpal joint and the other at the level of the midpoint between the distal projection of the radius and ulna) and on the foot (one at the level of the transverse arch and the other at the level of the midpoint between the medial and lateral malleoli). All electrodes were placed on the right limbs. The values which were directly measured by the test were resistance (Rz) and reactance (Xc), from which the PhA value, expressed in degrees (°), was obtained, as the arctangent of the Xc/R ratio multiplied by 180/π. The following hydration status parameters, were measured using the manufacturer's proprietary algorithms with Bodygram Plus software (Akern, Florence, Italy): TBW, TBW/FFM, ECW, and ICW.

Non-diagnostic ultrasound-based technique

Ultrasound measurements of muscle were performed using the non-diagnostic, 2.5 MHz, A-mode portable ultrasound BodyMetrixTM device (IntelaMetrix, Inc., Livermore, CA, USA) connected to a computer or tablet. Ultrasound scans were conducted on the left thigh, following the previously described measurement protocol [36]. Once acquired, the ultrasound scans were analysed by the same operator using the software Adipometria v 1.8.8 (Hosand Technologies s.r.l., Verbania, Italy). Supplementary Figures S1-S3 show examples of thigh ultrasound scans and their analysis. The scan in supplementary Figure S1 was obtained from a male, athletic subject; scans in supplementary Figures S2 and S3 were obtained from male subjects with obesity. The images show the scan analysis mode, in which individual tissues (subcutaneous adipose tissue and muscle) can be selected. For subcutaneous adipose tissue (SAT), the red line traces the separation between superficial SAT (sSAT) and deep SAT (dSAT), which are automatically identified by the analysis software with green and red colouring, respectively. In the muscle section, the upper red line marks the boundary between SAT and muscle, while the lower line marks the lower limit of the muscle. Within the muscle section, the individual muscle heads of the quadriceps highlighted in this ultrasound pathway (rectus femoris and vastus intermedius) are selected with a ‘freehand’ function, thus without the insertion of an additional red line at the level of the connective tissue separating the two heads. The ‘freehand’ function was preferred for muscle analysis, in order to be able to exclude continuous hyper-echogenic connective tissue structures from the selection of individual muscle areas. In this way, tissue analysis was performed exclusively at the muscle level. Both the red lines delimiting the individual tissue areas and their colouring are drawn by the operator during ultrasound scan analysis, using specific functions of the analysis software. Once the tissue areas have been identified (by inserting the red lines) and selected, the programme provides the relevant quantitative (thickness) and qualitative (echogenicity and derived indices) analyses.

Muscle EGs was directly measured by the analysis program: rectus femoris EG, vastus intermedius EG, and total muscle EG (rectus femoris + vastus intermedius), expressed as percentage values (defined as muscle-related parameters).

As defined by the manufacturer, the percentage value of EG is calculated by the analysis software, which directly measures the signal intensity point by point within the tissue area selected in the ultrasound scan. From this acquisition, the calculation of the EG value, being not-diagnostic, is simplified with regard to the grey gradation (0–255) by defining specific values ranges. The values obtained, through manufacturer’s proprietary formulae/algorithms, are normalised on a 0–100 scale, yielding a percentage value. The %EG used in the statistical analysis of this study was automatically provided by the research version of the analysis software; the commercially available version provides a muscular %EG value expressed on an inverse scale.

Physical activity

Subjects who habitually performed at least 30 min of aerobic physical activity daily (YES/NO) were classified as physically active, as already widely reported in previous studies [40, 41]. No participants were engaged in hard exercise, and there were no athletes. As it was not the main purpose of this study (which focused on the evaluation of possible relationships between parameters obtained from two-body composition assessment instruments, irrespective of subject classification), the exercise data were not considered in the statistical analysis.

Statistical analysis

The main results are reported as mean ± standard deviation (SD). Differences between women and men were analysed by unpaired Student’s t-test. The Pearson r correlation coefficient was used to assess correlations between the single monitored parameters. A multiple regression analysis model (stepwise method) was performed, with results expressed as Beta (β), t, and R², to estimate the influence of: (i) BIA-related parameters on ultrasound muscle-related parameters (EG), (ii) rectus- and vastus-related ultrasound parameters (EG) on PhA, and (iii) rectus- and vastus-related ultrasound parameters (EG) on ICW (expressed as %). Only parameters that showed a significant correlation were included in the analysis. Variables with a variance inflation factor (VIF) > 10 were excluded to avoid multicollinearity. Values ≤ 5% were considered statistically significant. Statistical analyses were performed and graphs produced using the IBM SPSS Statistics Software (PASW Version 21.0, SPSS Inc., Chicago, IL, USA) and GraphPad Prism (version 6.01, La Jolla, CA, USA).

Results

Data from 993 subjects (539 women and 454 men) were included in the statistical analysis (Table 1). Overall, participants were adults (31.95 ± 12.70 and 35.20 ± 14.19 years, p < 0.001, women and men, respectively) and with mean BMI of 25.40 ± 5.29 and 26.86 ± 4.85 kg/m², women and men, respectively (p < 0.001). With regard to BIA parameters, as reported in Table 1, women presented lower PhA value (Δ = −1.04°, p < 0.001), TBW (Δ = −7.54% and −15.8 L, p < 0.001 for both), ECW expressed in L (Δ = −4.84 L, p < 0.001) and ICW expressed as % and L (Δ = −4.16% and −10.95 L, p < 0.001 for both) compared to men. On the other hand, women presented higher Rz and Xc values (Δ =  + 104.41 and + 3.9 Ohm, respectively, p < 0.001 for both) and %ECW (Δ =  + 4.19%, p < 0.001) compared to men. With regard to ultrasound parameters, women exhibited higher EG values of total muscle (Δ =  + 4.06%, p < 0.001), rectus (Δ =  + 4.42, p < 0.001), and vastus (Δ =  + 3.83, p < 0.001), compared to men (Table 1).

Table 1.

Descriptive analysis of study populations

Parameters	Total (n = 993)	Women (n = 539)	Men (n = 454)	p-value
	Mean ± SD	Mean ± SD	Mean ± SD
Age (years)	33.71 ± 13.62	31.95 ± 12.70	35.20 ± 14.19	< 0.001
Weight (kg)	76.07 ± 17.17	68.25 ± 13.80	85.35 ± 16.12	< 0.001
Height (cm)	170.58 ± 10.04	164.15 ± 6.80	178.23 ± 7.61	< 0.001
BMI (kg/m2)	26.07 ± 5.14	25.40 ± 5.29	26.86 ± 4.85	< 0.001
Rz (Ohm, Ω)	470.37 ± 81.28	518.11 ± 69.46	413.70 ± 52.98	< 0.001
Xc (Ohm, Ω)	55.04 ± 8.68	56.82 ± 9.08	52.92 ± 7.67	< 0.001
PhA (°)	6.78 ± 1.05	6.30 ± 0.90	7.34 ± 0.92	< 0.001
TBW/FFM (%)	73.48 ± 1.28	73.34 ± 1.12	73.60 ± 1.44	0.006
TBW (%)	57.59 ± 7.47	54.14 ± 6.95	61.68 ± 5.83	< 0.001
TBW (L)	43.53 ± 10.08	36.31 ± 4.72	52.11 ± 7.76	< 0.001
ECW (%)	42.78 ± 4.29	44.70 ± 3.94	40.51 ± 3.51	< 0.001
ECW (L)	18.42 ± 3.70	16.20 ± 2.33	21.04 ± 3.28	< 0.001
ICW (%)	57.22 ± 4.29	55.32 ± 3.98	59.48 ± 3.48	< 0.001
ICW (L)	25.19 ± 6.95	20.11 ± 3.22	31.06 ± 5.33	< 0.001
EG muscle total (%)	32.77 ± 8.99	34.63 ± 8.82	30.57 ± 8.69	< 0.001
EG rectus (%)	44.58 ± 11.34	46.60 ± 11.08	42.18 ± 11.18	< 0.001
EG vastus (%)	20.87 ± 8.80	22.62 ± 8.51	18.79 ± 8.70	< 0.001

Data are expressed as mean and SD. Difference between women and men were analysed by paired Student’s t test. A p value in bold type denotes a significant difference (p < 0.05)

BMI; body mass index, Rz; resistance, Xc; reactance, PhA; phase angle, TBW; total body water, FFM; fat-free mass, ECW; extracellular water, ICW; intracellular water, EG; echogenicity, SD; standard deviation

Correlation study

Total muscle-related ultrasound parameters (EG) were correlated with BIA-related parameters (Rz, Xc, PhA, TBW, ECW, and ICW) (Table 2). In the total population and after stratification by sex, total muscle EG correlated positively with Rz (p < 0.001), and negatively with Xc (p = 0.001, 0.004, and < 0.001 for total, women, and men, respectively), PhA, TBW (% and L), and ICW (% and L) (p < 0.001 for all). Only ECW showed a different trend, correlating significantly and positively when expressed in % in total population and after stratification by sex (p < 0.001 for all); when expressed in L, ECW correlated significantly and negatively only in the total population (p < 0.001) (Table 2). After adjusting for BMI and TBW/FFM, the same correlations were maintained, except for ECW in L with EG (negative and significant in women p = 0.003, not significant in men) (Table 2).

Table 2.

Correlations between total muscle EG and BIA parameters in total population and stratified by sex before and after adjustment for BMI and TBW/FFM

Parameters	Before adjustments		After adjustements
	r	p value	r	p value
	Total
Rz (Ohm, Ω)	0.347	< 0.001	0.424	< 0.001
Xc (Ohm, Ω)	–0.102	0.001	–0.074	0.019
PhA (°)	–0.487	< 0.001	–0.489	< 0.001
TBW (%)	–0.341	< 0.001	–0.396	< 0.001
TBW (L)	–0.308	< 0.001	–0.379	< 0.001
ECW (%)	0.485	< 0.001	0.489	< 0.001
ECW (L)	–0.112	< 0.001	–0.207	< 0.001
ICW (%)	–0.479	< 0.001	–0.484	< 0.001
ICW (L)	–0.386	< 0.001	–0.429	< 0.001

	r	p value	r	p value
	Women
Rz (Ohm, Ω)	0.307	< 0.001	0.407	< 0.001
Xc (Ohm, Ω)	–0.123	0.004	–0.059	0.172
PhA (°)	–0.420	< 0.001	–0.426	< 0.001
TBW (%)	–0.283	< 0.001	–0.349	< 0.001
TBW (L)	–0.290	< 0.001	–0.396	< 0.001
ECW (%)	0.417	< 0.001	0.427	< 0.001
ECW (L)	0.003	0.950	–0.130	0.003
ICW (%)	–0.408	< 0.001	–0.419	< 0.001
ICW (L)	–0.425	< 0.001	–0.453	< 0.001

	r	p value	r	p value
	Men
Rz (Ohm, Ω)	0.220	< 0.001	0.309	< 0.001
Xc (Ohm, Ω)	–0.216	< 0.001	–0.204	< 0.001
PhA (°)	–0.472	< 0.001	–0.467	< 0.001
TBW (%)	–0.254	< 0.001	–0.323	< 0.001
TBW (L)	–0.175	< 0.001	–0.294	< 0.001
ECW (%)	0.475	< 0.001	0.471	< 0.001
ECW (L)	0.086	0.067	0.031	0.507
ICW (%)	–0.471	< 0.001	–0.469	< 0.001
ICW (L)	–0.308	< 0.001	–0.391	< 0.001

A p-value in bold type denotes a significance difference (p < 0.05)

BIA; bioelectrical impedance analysis, Rz; resistance, Xc; reactance, PhA; phase angle, TBW; total body water, ECW; extracellular water, ICW; intracellular water, EG; echogenicity

The influence of BIA-related parameters on total muscle-related ultrasound parameters were assessed using multiple regression analyses models with EG as dependent variable of each single model (Table 3). In particular, regarding the influence of BIA-parameters on total muscle EG, ICW (L) entered at the first step (p < 0.001), followed by PhA and TBW (%) in women, whereas ECW (%) entered at the first step (p < 0.001) in men (Table 3).

Table 3.

Multiple regression analysis model (Stepwise method) with total muscle EG as dependent variable to estimate the major influencer between BIA parameters in women and men

Parameters	Multiple regression analysis
	R2	β	t	p Value
Women
Model 1
ICW (L)	0.179	−0.425	−10.807	<0.001
Variables excluded: Rz, PhA, TBW (%), TBW (L), ECW (%), ECW (L), ICW (%)
Model 2
ICW (L)	0.222	−0.272	−5.733	<0.001
PhA (°)		−0.260	−5.668	<0.001
Variables excluded: Rz, TBW (%), TBW (L), ECW (%), ECW (L), ICW (%)
Model 3
ICW (L)	0.228	−0.291	−3.497	<0.001
PhA (°)		−0.186	−5.990	0.001
TBW (%)		−0.108	−2.121	0.023
Variables excluded: Rz, TBW (L), ECW (%), ECW (L), ICW (%)

Parameters	Multiple regression analysis
	R2	β	t	p Value
Men
Model 1
ECW (%)	0.226	0.475	11.477	<0.001
Variables excluded: Rz, Xc, PhA, TBW (%), TBW (L), ICW (%)

Only parameters that showed a significant correlation with total muscle EG were included in the analysis

A p-value in bold type denotes a significance difference (p < 0.05)

EG; echogenicity, BIA; bioelectrical impedance analysis, Rz; resistance, Xc; reactance, PhA; phase angle, TBW; total body water, ECW; extracellular water, ICW; intracellular water

Interestingly, significant differences were observed in the mean values of muscle-related ultrasound parameters in women and men with PhA and ICW (%) values above and below the median (median of PhA = 6.2° and 7.3°, women and men, respectively; median ICW (%) = 55.15 and 59.70%, women and men, respectively). In particular, in subjects with PhA and/or ICW (%) above the median, muscle EG was significantly lower (p < 0.001 for all) than subjects with PhA and/or ICW (%) below the median (Figs. 2 and 3).

Fig. 2.

Difference of muscle-related ultrasound parameters above and below the median of PhA values in women and men. Pink graphs refer to women; blue graphs refer to men. Abbreviations: phase angle, PhA; echogenicity, EG

Fig. 3.

Difference of muscle-related ultrasound parameters above and below the median of %ICW values in women and men. Pink graphs refer to women; blue graphs refer to men. Abbreviations: intracellular water, ICW; echogenicity, EG

Discussion

With a view to a comprehensive and integrative approach to BC assessment, this study focused on analysing the relationship between BIA parameters and non-diagnostic ultrasound-based measurements intended for non-clinical use in a large cohort of women and men. To the best of our knowledge, these data demonstrate for the first time correlations between BIA parameters and non-diagnostic/non-clinical ultrasound measurements at thigh level, emphasising the importance of combining the two methods to obtain more detailed and complete information on nutritional status. Ultrasound assessments were performed at the level of the quadriceps femoris as one of the largest districts of the body and, therefore, representative of the skeletal musculature, as well as being considered one of the most referenced sites due to its correlation with strength and physical performance [42–44].

The first noteworthy result is that the muscle EG showed a direct correlation with Rz and inverse correlation with the other BIA parameters, among which ICW and PhA were of particular interest. These observations appear plausible from both conceptual and physical perspectives. Rz is a raw BIA parameter that reflects the opposition of a conductor to current flow [3], and is therefore inversely proportional to TBW [3, 45]. A reduction in TBW and muscle electrolytes, hence, does not facilitate the flow of the alternating current applied in BIA, thereby resulting in an increase in Rz [3]. This increase in Rz correlates, at ultrasound level, with an increase in muscle EG which, in turn, may reflect a poorly (or worse) hydrated muscle. In addition to the relationship with Rz (and related TBW), an interesting finding emerges from the negative correlation between muscle EG and ICW. To understand this relationship, reference can be made to the concept of cell and muscle hydration. It is well known how an increase in muscle glycogen storages, binding water at a ratio of 1:3 [46], results in a corresponding increase in ICW [3] which, therefore, can be considered an indicator of muscle hydration. This concept explains why, as ICW (both % and L) increases, also indicating increased tissue hydration (at muscle level), a reduction in muscle EG is observed. Equally noteworthy is the inverse relationship observed between muscle EG and PhA, one of the most important BIA parameters. The importance of PhA lies as much in the fact that it is a direct parameter resulting from the ratio between Xc and Rz, as in its prognostic value, which provides information not only on the intracellular and extracellular distribution of fluids [47–50], but also on the integrity, permeability and thus functionality of cell membranes [48, 51]. PhA, therefore, reflecting body cell mass [48], can be considered as an index of cellularity, which justifies its inverse relationship with muscle EG. Although in this study muscle EG was evaluated with a non-diagnostic ultrasound-based technique (using the BodyMetrixTM device), data herein presented are consistent with previous observations obtained with conventional ultrasound, where a correlation with BIA parameters was demonstrated. Specifically, an inverse correlation was reported between muscle EG measured at the thigh level and PhA, in women with obesity [32], as well as in patients with disease-related malnutrition [29], whereas in patients with amyotrophic lateral sclerosis, muscle EG was found to correlate positively with Rz [31]. However, it is important to underline that the results obtained in the present study using BodyMetrix™ device cannot be directly compared with those derived from conventional diagnostic ultrasound systems. Although both technologies are based on the same physical principle (ultrasound), the BodyMetrix™ device employs a simplified and non-diagnostic analysis of muscle EG, expressed as a percentage, and, therefore, lacks the quantitative precision and imaging detail of clinical devices. Consequently, EG, numerical values obtained with BodyMetrix™ cannot be matched or interpreted against those derived from other devices, imaging platforms or software tools. At most, only qualitative trends over time (i.e., increases or decreases in EG) may be compared between methods, without assigning clinical or diagnostic significance to such changes.

In the context of its prognostic value, it should be taken into account that PhA is universally recognised as a surrogate marker of cell damage and function [20, 21], as well as inflammation [22], which represent the conceptual bases of our previous proposal of PhA as possible marker of overreaching and overtraining [24] in chronic and prolonged sport practice. In a broader sense, this confers on PhA a further value as an indicator of muscle quality [26]. The existence of a relationship with muscle EG, thus, confirms and reinforces its qualitative value.

As previously reported, muscular EG is also suggested as a parameter for the evaluation of muscle quality [26], since the onset of inflammatory or fibrotic processes, myosteatosis or alterations in muscle architecture (indicative of impaired tissue quality) result in a variation of EG, which tends to increase [28]. It is essential to clarify that such interpretations pertain to diagnostic ultrasound applications performed by trained medical professionals using standardised devices. The BodyMetrix™ device, as a non-diagnostic tool, does not allow definitive identification of such pathological changes. Therefore, in the context of the present study, increases in EG should not be interpreted as conclusive indicators of inflammation, fibrosis or myostatosis, but rather as non-specific changes only potentially associated with alterations in muscle composition or hydration status. This possible association should be confirmed in future comparative studies using diagnostic instrumentation/assessments against the device adopted in this study. Even if such associations were confirmed, their interpretation would remain purely speculative, given the non-diagnostic and non-clinical purpose of the instrument and analysis employed here. Nonetheless, this study interpretation is consistent with our previous study, that observed EG variations across age using non-diagnostic ultrasound in observational, non-clinical settings [36]. Appropriate caution should therefore be exercised when interpreting these findings, which should be regarded as exploratory and hypothesis-generating.

On the basis of these considerations, the integration of BIA parameters with those obtained through a non-diagnostic ultrasound-based technique may enable a differential assessment to be conducted for monitoring nutritional status. Specifically, it can be hypothesized that, in the absence of a different clinical diagnosis, the integration of muscle EG values (either at baseline or their variations over time) with:

1. ICW values (at baseline or their variations over time) may be indicative of a state of tissue (muscle) dehydration

2. PhA values (at baseline or their variations over time) may be indicative of potential inflammatory states, oxidative stress or fatigue or their exacerbations (i.e. overreaching/overtraining).

Nonetheless, a critical limitation to acknowledge is the mismatch between the whole-body nature of BIA parameters and the localised nature of ultrasound measurements. In particular, extrapolating EG data obtained at the thigh to whole-BC or hydration status should be done cautiously, as anatomical and functional variability in muscle structure and composition across different regions may influence both ultrasound and BIA measurements. However, it is worth noting that previous studies investigating correlations between ultrasound parameters and BIA metrics have also performed ultrasound evaluations at the quadriceps level [29, 31, 32], supporting the choice of this site as a representative and functionally relevant muscle group. Based on these findings and on the associations observed in our cohort, it is reasonable to hypothesise that quadriceps muscle EG (whether assessed with diagnostic or with non-diagnostic and non-clinical analysis) may serve as a proxy for global status. Neverthless, this remains a hypothesis that requires confirmation through future studies involving multiple anatomical sites and more refined reference standards. In particular, future longitudinal studies are needed to validate this hypothesis and consolidate the usefulness of integrating BIA parameters with non-diagnostic ultrasound-based technique for a more comprehensive qualitative assessment of BC and nutritional status.

Limitations and strengths

As a major strength, this study reports for the first time correlations between parameters obtained by BIA and non-diagnostic ultrasound-based tissue analysis intended for non-clinical measurements in a large cohort of men and women. These results provide fundamental information for an in-depth assessment of nutritional status, with a particular focus on the evaluation of muscle quality. Furthermore, they highlight the importance of a multifactorial approach, integrating data obtained from two techniques. Nonetheless, this study also has certain limitations, mostly related to the retrospective nature of the statistical analysis. First, although the overall number of subjects is large, the distribution between men and women is not balanced. For the same reason, subjects of different ages (ranging from 18 to 79 years) and BMI (ranging from 18.63 to 55.36 kg/m²) were included in the entire cohort. Finally, with regard to the women, the lack of information regarding the phase of the reproductive cycle (e.g. fertile age, pre-menopause, and menopause) represents a further limitation. Similarly, for the entire study population, the absence of data on illnesses, drug therapies or types of sport practiced prevent relative adjustments in the analyses of muscle assessments. Further future studies conducted in specific settings (both clinical and non-clinical) are necessary.

Conclusion

On the basis of the results of this study, BIA and non-diagnostic ultrasound-based tissue analysis emerge as valid tools for nutritionists and trainers to achive an overall and detailed assessment of nutritional state and BC. Such assessment is not limited to the distribution of fat and muscle mass, but also includes the regional distribution and quality of subcutaneous tissues, providing useful information for a more accurate delineation of individual well-being. It is, therefore, evident that the integration of the two methods is fundamental in the subject evaluation. In particular, the integration of BIA parameters with ultrasound measurements may overcome the respective limitations of each method: (i) in total body BIA, the parameters obtained are not site-specific and (ii) in non-diagnostic tissue ultrasound, a single parameter can convey multiple information, which gain interpretative value if properly contextualised. Nevertheless, it is important to underline that this study is exploratory in nature, and the integration of these two techniques, although promising, requires further validation through longitudinal and interventional research. Such future investigations will be necessary to confirm the clinical and prognostic significance of the observed associations and to assess the reproducibility, sensitivity, and specificity of this combined approach across different populations and settings. The integration of these poly-instrumental parameters should, therefore, be interpreted with caution and currently considered as a potential tool for research and field-based monitoring, rather than as a definitive diagnostic method. Similarly, the interpolation of data obtained through these two methods into specific predictive formulae would allow the development of new indices and prognostic markers of health, a hypothesis that remains to be tested and validated in future controlled studies. In conclusion, the present work provides a conceptual and methodological basis for the integrative use of BIA and non-diagnostic ultrasound, laying the groundwork for future research aimed at improving the qualitative assessment of BC and nutritional status.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

BC

Body composition

BIA

Bioelectrical impedance analysis

BMI

Body mass index

dSAT

Deep subcutaneous adipose tissue

ECW

Extracellular water

EG

Echogenicity

FFM

Fat-free mass

ICW

Intracellular water

PhA

Phase angle

Rz

Resistance

SAT

Subcutaneous adipose tissue

SD

Standard deviation

sSAT

Superficial subcutaneous adipose tissue

TBW

Total body water

Xc

Reactance

Author contributions

Conceptualization: GA and LB; data curation: GA, ARAG, TS, MGS, NMVC, GR, LA; formal analysis: GA; methodology: GA and GM; supervision: LB; roles/writing—original draft: GA, LV; writing—review and editing: LB, GM, AC, DRW, MC, and MGT. All authors have read and agreed to the published version of the manuscript.

Funding

The article presented here won the Journal of Translational Medicine Award announced at the Congress ‘Clinical Nutrition Days of the Alto Tavoliere, 2nd edition—International Meeting’ and, accordingly, is exempted from publication fees.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethical approval and consent to participate

The study was conducted in accordance with the guidelines outlined in the Declaration of Helsinki, which provides ethical principles for medical research involving human subjects. Additionally, the Ethics Committee of the Università Telematica Pegaso granted a positive opinion on the study protocol (reference no. PROT./E 005082 of 19/07/2024).

Consent for publication

Not applicable.

Competing interests

Giuseppe Annunziata, Maria Grazia Tarsitano, and Luigi Barrea are currently members of the Editorial Board of the Journal of Translational Medicine.