Introduction

Bioelectrical impedance analysis (BIA) is a widely employed method for estimating body composition in both epidemiological and clinical settings. Its popularity stems from its non-invasive nature, cost-effectiveness, and ease of application and interpretation [1]. BIA is a doubly indirect method, and its effectiveness in estimating body composition is impacted by factors such as the selection of equipment, as well as population-specific variables like age, sex, and the equations employed [2,3,4]. By utilizing the raw parameters of resistance (R) and reactance (Xc) provided by BIA, it becomes possible to calculate phase angle (PhA) [5]. PhA is a parameter that provides insights into cellular membrane health, hydration status, and overall well-being, serving as a prognostic indicator across various clinical conditions [5, 6]. Furthermore, PhA has garnered recognition as a valuable marker for assessing muscle mass and muscle function in recent times [7].

PhA holds both clinical and scientific significance; however, its relationship with markers of metabolic and cardiovascular health remains underexplored. Given PhA’s association with cell membrane health [5], it holds promise as a valuable marker for monitoring cellular health concerning adiposity and cardiovascular assessment. Given the influence of such conditions on inflammation status [8], we hypothesize that understanding the behavior of PhA in this perspective may provide insights into overall health at the cellular level. Considering the high prevalence and mortality associated with cardiovascular diseases [9], and the practicality of measuring PhA compared to other imaging methods such as ultrasound (US), computed tomography, and magnetic resonance imaging, it is reasonable to conduct research to explore this connection. Therefore, our study endeavors to investigate the link between PhA and indicators of cardiometabolic health among patients with heart diseases receiving outpatient care.

Subjects and methods

Study design

This cross-sectional hospital-based study was conducted at a public referral unit in cardiology in Pernambuco, Brazil, from September 2021 to October 2022. The study sample consisted of adult outpatients of both sexes, aged between 26 and 59 years. Participants with physical limitations that hindered physical-functional evaluation, individuals with pacemakers or metallic implants, those with conditions such as edema, ascites, hepatopathies, or splenomegaly, individuals undergoing hormone therapy (excluding insulin therapy), those in the post-abdominal surgical recovery phase, individuals who had undergone weight loss surgery, pregnant women, and women who had given birth within 6 months prior to the study’s commencement were excluded from the study.

Sample size was calculated considering a 5% alpha error, a 20% beta error, an estimated correlation between phase angle and visceral fat of 0.5 (ρ), and a variability of 0.15 (d²), totaling a minimum sample size of 88 patients. To cover potential losses, 20% was added to the minimum sample size, resulting in a final sample size of 106 patients. Patients willing to participate were conveniently sampled in this study if they met the eligibility criteria.

Clinical and covariates

Sociodemographic information, encompassing age and sex, was collected. Regarding lifestyle factors, variables such as alcohol consumption and smoking were documented. Alcohol consumption was evaluated as a dichotomous response (yes or no) [10]. In the case of smoking, individuals smoking at least one cigarette per day were categorized as smokers, those who never smoked were classified as non-smokers, and individuals who had smoked at some point in their lives but were not currently smoking at the time of the research were identified as former smokers [11].

The presence of the following comorbidities, namely systemic arterial hypertension (SAH) and type 2 diabetes mellitus (2DM), was assessed. Self-reported participant diagnoses were initially obtained, and subsequent validation was conducted by cross-referencing with medical records. This validation encompassed confirming the utilization of antihypertensive and hypoglycemic medications, as well as obtaining medical confirmation of comorbidities. Furthermore, measurements of systolic blood pressure (SBP) and diastolic blood pressure (DBP) were recorded.

Cardiometabolic parameters

Framingham Risk Score [12] was utilized to evaluate the risk of cardiovascular diseases. The risk score was computed based on categorized values of age, sex, total cholesterol, HDL-C, SBP, smoking status, and 2DM. Smoking status was ascertained through self-reported information and categorized as “current smokers” (encompassing both current smokers and recent quitters) or “non-smokers” (encompassing individuals who had never smoked or had quit smoking long ago). For analytical purposes, the results of the Framingham score were stratified into two categories: lower risk for individuals with scores between low and intermediate (5–20%), and higher risk for those with high scores (>20%) [12], indicating an elevated risk of cardiovascular diseases.

For the study, various metabolic parameters were deemed elevated, in accordance with the criteria established by Faludi et al. in 2017 and Cobas et al. in 2022 [national guidelines] [13, 14]. These parameters included fasting blood glucose (FBG) (>100 mg/dL), HbA1C (>5.7%), total cholesterol (TC) (>190 mg/dL), HDL-c (<40 mg/dL), LDL-c (>130 mg/dL), and triglycerides (>150 mg/dL). All pertinent blood samples results were retrieved from the medical records.

Ultrasound (US)

The evaluation of visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) involved abdominal ultrasound conducted by a single trained observer, with participants undergoing a fasting period of at least 4 h. The Apogee 3500 color digital ultrasound imaging system (SIUI, Shantou, China), equipped with a 4.0 MHz convex transducer, was employed for this procedure. The measurement of visceral fat adhered to the protocol outlined by Mauad et al. in 2017 [15].

The thickness of the fat measured in the subcutaneous layer was performed with a linear transducer at a frequency of 10.0 MHz. All individuals were evaluated in the dorsal decubitus position, with the right arm elevated. The measurement of subcutaneous fat was performed with the transducer positioned transversely 1.0 cm above the umbilical scar, on the xiphopubic line, without exerting pressure on the abdomen, in order not to underestimate the measurement. The anatomical limits for measuring subcutaneous thickness were the skin and the external (superficial) surface of the rectus abdominis muscle, quantified in centimeters.

The measurement of visceral fat was performed with a convex transducer at a frequency of 4.0 MHz, positioned transversely 1.0 cm above the umbilical scar, on the xiphopubic line, without exerting pressure on the abdomen, in order not to underestimate the measurement. The anatomical limits for measuring visceral fat thickness had as a reference point the internal (deep) surface of the rectus abdominis muscle and the anterior wall of the aorta, with the individual exhaling, quantified in centimeters. The measurements of VAT were quantified in centimeters (cm). In line with the recommendations of Leite et al. [16], a cutoff point of ≥9 cm for men and ≥8 cm for women was employed to identify the presence of visceral obesity.

Figure 1 shows an example of VAT image by US obtained during our study. A prior assessment of the reproducibility of intra-evaluator US measurements was conducted on 10% of the sample. Intra-evaluator reproducibility was notably high, with an Intraclass Correlation Coefficient of 0.99 for SAT evaluation and 0.97 for VAT.

Fig. 1
figure 1

Visceral adipose tissue (VAT) using ultrasound imaging.

Bioelectrical impedance analysis (BIA)

A single-frequency (50 kHz) Biodynamic Body Composition Analyzer, model 310e, from Biodynamics Corporation in Seattle, WA, USA, was employed to assess R, Xc, and calculate PhA. Patients adhered to a 4-h fasting period, refrained from alcohol consumption for the past 48 h, and avoided intense physical exercise the day before the test. They received advanced notice of these requirements at least 3 days prior. During the assessment, patients laid in a dorsal decubitus position on a bed, free of any metallic objects. The test involved the use of four electrodes placed on the right wrist and ankle.

The manufacturer’s specifications highlight that these meters are modern, digital, and capable of precisely measuring impedance, independent of stray capacitance. Designed with a solid-state, digital architecture, these meters remain stable without drifting and do not necessitate recalibration. BIA 310 meter accurately captures resistance and reactance across the entire range found in the human body. Detailed specifications for R and Xc include a 0–1500 ohms range for R, and 0–300 ohms for Xc, with 1 ohm resolution, 0.001 ± 0.1 ohms accuracy for R, 0.002 ± 0.1 ohms accuracy for Xc, all measured at a frequency of 50 kHz.

Based on sex and age, participants were divided into groups according to their derived SF-BIA PhA. PhA values were calculated using the formula PhA = arctangent (Xc/R) × 180/π, using data acquired from a 50 kHz frequency. Low PhA was classified according to the cutoff threshold of <5th percentile for the respective age group and sex within the Brazilian population, as proposed by Barbosa-Silva et al [17], despite being validated using a different BIA device.

Muscle strength and anthropometry

Muscle strength was assessed using Handgrip Strength (HGS), measured with a JAMAR® digital dynamometer, following established techniques recommended by the American Society of Hand Therapists were followed [18]. Anthropometric measurements [19] were collected, including weight, height, waist circumference (WC), waist-to-height ratio (WtHR), and Body Mass Index (BMI). WC was measured 2 cm above the umbilical scar and the measurement was taken at the end of a normal expiration. The WtHR was calculated by dividing the WC by height.

Statistical analyses

Continuous variable normality was examined using the Kolmogorov-Smirnov test. Results for continuous variables were reported as mean ± standard deviation (SD) or median with interquartile range (IQR). To compare means or medians, the independent student t-test or Mann-Whitney U test was applied as appropriate. Quade’s non-parametric ANCOVA test was utilized to adjust the comparisons of nutritional characteristics based on sex. The correlation between PhA and variables of interest was evaluated using Spearman’s rho correlation. The strength of correlation analyses was categorized based on the magnitude of the correlation coefficient. (rho/ρ values) as follows: very high (ρ = 0.90–1.00), high (ρ = 0.70–0.90), moderate (ρ = 0.50–0.70), low (ρ = 0.30–0.50), or negligible (ρ = 0.00–0.30) [20].

Categorical variables were expressed as absolute (n) and relative (%) frequencies. Pearson’s chi-square and Fisher’s exact tests were utilized for categorical variables. Crude and adjusted logistic regression analyses were conducted to assess the association between PhA and cardiometabolic parameters. Receiver operating characteristic (ROC) curve analysis was conducted to assess the strength’s association of PhA with visceral adiposity and cardiovascular risk. Performance was classified based on the area under the receiver operating characteristic curve (AUC) values: excellent (0.90–1.00), good (0.80–0.90), fair (0.70–0.80), poor (0.60–0.70), and failed (0.50–0.60) [21]. The optimal cutoff values for PhA associated with such variables were determined using the Youden index within the AUC.

Data were analyzed using IBM SPSS Statistics version 20 (SPSS Inc., Chicago, IL, United States), and MedCalc version 22.0.0.9 software (MedCalc, Mariakerke, Belgium). The level of statistical significance was set at p < 0.05 (two-sided).

Results

Patients’ characteristics

The study enrolled one hundred and five patients, with a median age of 47 years. Among these participants, 61.9% (n = 65) were women. The frequency of high cardiovascular risk was 25.7% (n = 25), according to the Framingham score. Low PhA was found in 29.5% (n = 31) of our population. Table 1 provides details on the sociodemographic, lifestyle, clinical, and nutritional characteristics of our population. Patients with low PhA had higher SBP, higher levels of triglycerides, a higher BMI, WC, and WtHR. Individuals with low PhA also demonstrated higher values of VAT, lower values of HGS, and higher scores on the cardiovascular risk (Framingham) scale (p < 0.05), even after adjustment for sex (Table 1).

Table 1 Characteristics of the patients based on the phase angle classification (n = 105).

Correlation and association analyses

PhA (°) showed a negligible to low negative correlation with BFG, HbA1C%, TC, TG, BMI, WC, VAT, and the Framingham score. PhA exhibited a moderate positive correlation with HGS (ρ = 0.62 p < 0.001) (Table 2). Higher values of PhA were inversely associated with higher BMI and WtHR, even after adjusting for confounders, such as age, sex, alcoholism, smoking, diagnosis of SAH, and 2DM, Moreover, higher values of PhA were inversely associated with elevated VAT [adjusted OR = 0.79 (95% CI 0.69;0.91)] and higher cardiovascular risk scores [adjusted OR = 0.74 (95% CI 0.60;0.89)]. PhA were independently associated with greater HGS [adjusted OR = 1.98 (95% CI 1.40;2.80)] (Table 3).

Table 2 Correlation of the phase angle with markers related to metabolism, adiposity, muscle function, and cardiovascular risk (n = 105).
Table 3 Association between phase angle (independent variable) and parameters of metabolism, adiposity, muscle function, and cardiovascular risk (dependent variables).

ROC curve analysis

For the entire sample, the PhA showed a good association with elevated VAT, (AUC 0.82 (95% bootstrap CI 0.67;0.91) and the cutoff associated was ≤5.59. Upon stratified analysis by sex, stronger associations were observed among women (Fig. 2), [AUC 0.92 (95% bootstrap CI 0.82;0.97)], with the same cutoff. Among males, values of PhA ≤ 4.82 were associated with high VAT. PhA (for the entire sample) presented with fair association with higher cardiovascular risk [AUC 0.70 (95% bootstrap CI 0.55;0.82)] and the cutoff associated was ≤5.06. This association was also stronger among females [AUC 0.86 (95% bootstrap CI 0.72;0.94)] (see Fig. 3). The cutoff values associated with higher cardiovascular risk were PhA ≤ 3.55 for females, and ≤6.78 for males.

Fig. 2
figure 2

ROC curve analysis: strength of the association between phase angle and visceral adipose tissue (US-derived), both for the total sample and when stratified by sex.

Fig. 3
figure 3

ROC curve analysis: strength of the association between phase angle and high cardiovascular risk (Framingham’s score), both for the total sample and when stratified by sex.

Discussion

Our study investigated the relationship between PhA and several cardiometabolic markers among individuals under the care of a cardiology outpatient unit. To the best of our knowledge, no previous study has investigated the values of that could be associated with elevated VAT and higher cardiovascular risk in outpatient adults. Our main findings revealed that PhA exhibited independent and inverse associations with markers of cardiometabolic risk, notably both high VAT and Framingham’s scores. Such results underscore the potential utility of PhA as a valuable indicator in assessing and managing cardiometabolic health conditions.

A prior cohort [22] yielded divergent results in relation to metabolic parameters. This cohort reported a positive correlation between PhA and BMI, TC, and TG. These differences can be attributed to several factors, including the different BIA device for calculate PhA, a distinct population (Iranian) with potentially unique characteristics, and variations in blood test protocols. Metabolic diseases, such as metabolic syndrome, diabetes, cardiovascular disease, and obesity, share a common pathogenic pathway characterized by low-grade inflammation that impacts tissue electrical properties. This, in turn, leads to cellular oxidative stress and, consequently, a compromise in PhA [23]. Despite these connections, there has been limited research investigating the relationship between PhA and oxidative stress markers [23].

Other cohort [24] demonstrated that low PhA in women was independently associated with the development of cardiovascular diseases over a 24-year follow-up period. The authors underscore the significance of incorporating PhA as an additional predictor of cardiovascular risk. Additionally, a cross-sectional study [25] revealed that higher PhA was linked to a reduced risk of the first cardiovascular event in men, including those categorized as higher risk [25]. These results underscore that monitor PhA changes may serve as a valuable indicator of improvements in overall health status [26].

Prior studies [26, 27] indicate that the PhA may be influenced by various factors, including sex, age, nutritional status, and excess adiposity, highlighting lower cellular integrity in individuals with higher body mass. Previous cohorts [28,29,30] have shown that an increase in %BF, particularly VAT, is associated with a decrease in PhA degrees. [28,29,30]. A recent review [31] suggests that PhA could serve as a clinical prognostic marker in conditions associated with obesity, such as inflammatory states and glycemic/metabolic control. This is attributed to the fact that PhA is influenced by BMI, regional fat distribution, and obesity-related comorbidities [31]. Consequently, PhA holds potential as a marker of cellular health during weight management programs, with potential relevance to VAT and cardiovascular risk.

Our findings reveal a significant inverse relationship between PhA and VAT. This implies that individuals with higher PhA values may have lower amounts of visceral fat, correlating with a reduced risk of cardiovascular disease, or individuals with lower VAT may exhibit higher PhA values. This leads us to hypothesize that enhanced cellular health and metabolism, as indicated by improved PhA, may play a role in preventing the accumulation of visceral fat and vice versa.

Our clinometric performance (AUROC) demonstrated that lower PhA values in our sample were associated with higher VAT and cardiovascular risk. Given the absence of validated cutoff values specific to our population and the particular BIA device used, conducting such analysis is relevant. It can demonstrate that lower cell health, as indicated by PhA, may correlate with poorer cardiometabolic health. Therefore, clinicians could potentially utilize PhA as a tool for monitoring overall cardiometabolic health, akin to previous studies [30].

VAT accumulation closely correlates with an elevated pro-inflammatory state, driven by increased levels of cytokines such as interleukin-1beta and tumor necrosis factor alpha. These cytokines are produced as a result of lipid overload and by macrophages subsequent to infiltration and exposure to stressed adipocytes in VAT [32], consequently amplifying the cardiometabolic risk [33].

The association between lower values of PhA and higher VAT, as well as cardiovascular risk, can be elucidated through the pathway of inflammation. In a state of inflammation, reactive oxygen species compromise cell integrity, leading to disruptions in cellular fluid balance, which in turn affects the capacitive properties of cell membranes and impairs PhA [23, 31]. Consequently, PhA has been suggested as a low-cost alternative biomarker for detecting early inflammatory states [23].

Despite the found association between PhA and cardiometabolic parameters, particularly VAT and cardiovascular risk, it is important to acknowledge that PhA remains a biomarker derived from BIA that cannot be fully utilized as a diagnostic parameter [34]. However, it might be employed as an additional marker related to overall cardiovascular health.

Strengths and limitations

This study provides insights into the association between the PhA and cardiometabolic risk factors among cardiology outpatients. Employing precise measurement techniques, and robust statistical analysis, the study comprehensively examines the relationships between PhA and diverse health parameters, such as visceral adiposity and cardiovascular risk. Through meticulous adjustment for confounding variables, the study enhances the validity of its findings. The potential clinical applications and the contribution to the existing scientific literature underscore its significance as a valuable addition to our comprehension of PhA as a marker in cardiometabolic health assessment.

Our study has limitations that warrant acknowledgment. The observational, single-center, and cross-sectional design of the study precludes the establishment of causal relationships. Unmeasured or residual confounding cannot be entirely excluded from our regression models. It is crucial to recognize that population-specific factors, including genetics, lifestyle, and dietary habits, may influence the observed associations. The distinctive characteristics of the study population may contribute to the identified relationships. Furthermore, the absence of validated cutoffs for defining low PhA specific to our population and the BIA device used represents an additional limitation for making initial comparisons. To address this limitation, we conducted an analysis utilizing raw PhA values and employed ROC analysis to identify associated cutoff points. This approach allowed us to surmount the challenge of lacking established cutoff values, enabling a more robust evaluation of PhA’s association with cardiometabolic health markers.

Clinical implications and future research

Our findings unveil new avenues for research and emphasize the potential clinical relevance of PhA as a valuable complementary marker for assessing cardiometabolic health. Importantly, our study contributes to the expanding body of evidence supporting the potential utility of PhA in both clinical practice and research settings. In light of the limited existing research on this topic, our study highlights the need for continued exploration of PhA’s value, especially in the context of multimodal interventions aimed at improving health outcomes.

By monitoring changes in PhA, we may gain valuable insights into the effectiveness of interventions and individuals’ recovery of cardiometabolic health status. We highlight the importance of future studies involving both similar and diverse clinical populations, utilizing our cutoff values and comparing findings. This approach will contribute to a more comprehensive understanding of the relationship between PhA and cardiometabolic assessments.