Bioelectrical phase angle predicts sarcopenia in older adults with type 2 diabetes mellitus Authors: Methavee Poochanasri, Kasidid Lawongsa, Thanapat Limpaarayakul, Jiranon Pirapattanapong, Thawee Songpatanasilp, Nattapol Sathavarodom, Apussanee Boonyavarakul, Parinya Samakkarnthai Published: 2026 ABSTRACT AbstractSarcopenia is common among older adults with type 2 diabetes mellitus (T2DM) and is associated with increased risks of functional decline and adverse health outcomes. Phase angle (PhA), derived from bioelectrical impedance analysis, has emerged as a potential marker of muscle health. This cross-sectional study aimed to evaluate the predictive value of PhA for detecting sarcopenia in older adults with T2DM and to determine sex-specific cutoff values for sarcopenia screening. A total of 390 patients aged ≥60 yr with T2DM were included. Sarcopenia was defined according to the Asian Working Group for Sarcopenia 2019 criteria. Lower PhA values were significantly associated with sarcopenia, and PhA showed good discriminatory performance for detecting sarcopenia. Sex-specific cutoff values improved the diagnostic accuracy of PhA for sarcopenia screening. These findings suggest that PhA may serve as a practical and noninvasive tool for identifying sarcopenia in older adults with T2DM. Further longitudinal studies are warranted to confirm its prognostic value.Keywords: sarcopenia, phase angle, type 2 diabetes mellitus, bioelectrical impedance analysis, BIA, cutoff value Introduction Sarcopenia is a progressive muscle-wasting disorder characterized by declines in muscle mass, strength, and function, significantly affecting functional independence and overall health, particularly among older adults.1 This condition increases the risk of falls, functional impairment, and mortality.1,2 Type 2 diabetes mellitus (T2DM) is a recognized risk factor for sarcopenia, and growing evidence suggests that individuals with sarcopenia and type 2 diabetes face higher risks of dementia, cardiovascular disease (CVD), and mortality.3,4 Thus, early identification of sarcopenia in patients with diabetes is essential to reduce potential long-term complications. Phase angle (PhA), derived from bioelectrical impedance analysis (BIA), has recently emerged as a promising biomarker for detecting and monitoring sarcopenia.5 It reflects key physiological properties, including cell membrane integrity, muscle mass, and intracellular water distribution.6,7 A lower PhA indicates muscle deterioration and correlates positively with reduced appendicular skeletal muscle mass and decreased handgrip strength.8 Previous studies have examined the role of PhA in various clinical conditions. Curvello-Silva et al., in a cross-sectional study of obese individuals, reported a significant association between low PhA and hyperuricemia.9 Similarly, Barrea et al. found a positive correlation between vitamin D levels and PhA in adults with obesity.10 Additionally, a recent meta-analysis highlighted PhA as a potential marker for the early detection of sarcopenia in older adults and cancer patients, including those with gastric, colorectal, and prostate cancers, further supporting its role in sarcopenia screening.5 The pathophysiological link between type 2 diabetes and sarcopenia is complex, involving insulin resistance, chronic inflammation, and oxidative stress.11,12 Insulin resistance in skeletal muscle disrupts protein synthesis and accelerates protein breakdown, leading to progressive muscle loss.13 The chronic inflammatory state associated with diabetes further impairs muscle anabolism and promotes inflammatory cell infiltration, exacerbating muscle degeneration.14 Additionally, the accumulation of advanced glycation end products significantly contributes to mitochondrial dysfunction and oxidative stress, further accelerating muscle deterioration.12,15 Collectively, these factors increase the prevalence of sarcopenia among individuals with diabetes.16 Recent evidence indicates that individuals with type 2 diabetes have lower PhA values than age- and BMI-matched non-diabetic counterparts.17,18 Given the established link between low PhA and muscle deterioration, PhA may serve as an early indicator of pre-sarcopenia, facilitating timely interventions.17 However, research specifically evaluating the predictive value of PhA for sarcopenia in diabetic populations remains limited. Identifying a clinically relevant PhA cutoff for sarcopenia screening in patients with type 2 diabetes could significantly enhance early detection and inform targeted preventive strategies. Therefore, the present study aims to investigate the relationship between PhA and sarcopenia in individuals with type 2 diabetes and to establish an optimal, clinically meaningful PhA cutoff value for sarcopenia screening. Materials and methods Study design and participants This retrospective cross-sectional study was conducted at Phramongkutklao Hospital in Bangkok, Thailand, between December 2023 and November 2024. The study included patients aged 60 yr or older with a confirmed diagnosis of T2DM. Because this was a retrospective cross-sectional study based on available clinical data; a formal sample size calculation was not performed prior to data collection. All eligible patients during the study period were included in the analysis. Exclusion criteria included individuals with active cancer, neurological impairments causing weakness or limited mobility, uncontrolled underlying medical conditions, clinically significant edema or fluid retention, severe diabetic foot complications (eg, active foot ulcers, major foot deformities, or lower-limb amputation), or an inability to undergo a physical assessment. Patients with metallic implants or pacemakers were also excluded because of potential interference with BIA. After applying the inclusion and exclusion criteria, 390 patients were enrolled in the study. Ethics approval and data collection The study was approved by the Royal Thai Army’s Institutional Review Board (Approved No. R080h/66) in accordance with international standards, including the Belmont Report, the Declaration of Helsinki, the CIOMS Guidelines, and the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use-Good Clinical Practice (ICH-GCP). Patient data were collected through face-to-face interviews and electronic medical records. Demographic and baseline characteristics, including age, sex, underlying diseases, family history of diabetes, educational level, smoking status, alcohol consumption, and diabetes duration, were obtained. Information on diabetic complications, including results from the monofilament test and diabetic retinopathy screening, was also collected. Additionally, laboratory test results and current medications were documented using the hospital’s electronic medical record system. BIA measurement and PhA Body composition was assessed using multifrequency BIA. To ensure standardized measurements, participants were required to fast overnight, avoid physical activity for at least 8 h, and empty their bladder prior to testing. During the BIA measurement, participants stood barefoot on the analyzer’s footplate, holding the handgrips with their arms extended downward in a neutral standing position. The hand electrodes were placed in contact with all fingers of both hands, and the foot electrodes were positioned under the heels and forefeet. The device measured body fat mass, skeletal muscle mass (SMM), and BMI. The PhA was derived using the following equation: PhA (°) = arctan where Xc represents reactance and R represents resistance. The skeletal muscle index (SMI) is calculated using the following equation: SMI (kg/m2) = . Muscle strength and physical performance measurements Muscle strength and physical performance were assessed using standardized procedures. Calf circumference (CC) was measured at the widest part of the non-dominant leg while the participant sat with flat feet on the floor and the knee at 90° flexion. A CC of <34 cm in men and <33 cm in women was considered indicative of low muscle mass.19 Neck circumference (NC) was measured at the midpoint of the neck, with cutoffs of <32.8 cm in women and <38.0 cm in men used as screening criteria for sarcopenia.20 The SARC-F questionnaire assessed functional limitations related to strength, walking assistance, rising from a chair, climbing stairs, and falls. Each domain was scored from 0 to 2, with a total score ranging from 0 to 10. A total score of ≥4 indicated a risk of sarcopenia.21 The SARC-CalF tool, which combines the SARC-F questionnaire with CC measurement, was also applied. A total score of ≥11 was classified as high risk for sarcopenia.22 To assess lower extremity function, the chair stand test recorded the time participants needed to complete 5 repetitions of standing up and sitting down from a chair. A completion time of ≥12 s indicated low physical performance.23 Handgrip strength was measured using a digital hand dynamometer. Participants performed 3 grip tests with their dominant hand while seated, maintaining a 90° elbow flexion. The highest recorded value (in kg) was used for analysis. According to the Asian Sarcopenia Working Group (AWGS) 2019 guidelines, low handgrip strength was defined as <28.0 kg in men and <18.0 kg in women.19 The gait speed test assessed walking ability by measuring the time required for participants to walk 6 m at a standard pace. Two trials were conducted, and the average was calculated in meters per second (m/s). A gait speed of <1.0 m/s was considered the cutoff for low physical performance.19 Sarcopenia diagnosis Sarcopenia was diagnosed using the AWGS 2019 criteria. According to these guidelines, low muscle mass is defined as <7.0 kg/m2 in men and <5.7 kg/m2 in women, as measured by BIA. Low muscle strength is defined as handgrip strength <28.0 kg in men and <18.0 kg in women. Low physical performance is defined as gait speed <1.0 m/s or chair stand time ≥12 s. A diagnosis of sarcopenia is confirmed when a participant exhibits low muscle mass in combination with either low muscle strength or low physical performance.19 After diagnosis, participants are categorized into sarcopenic and non-sarcopenic groups for further analysis. Statistical analysis Statistical analyses were conducted using STATA software (StataCorp. 2023. Stata Statistical Software: Release 18, StataCorp LLC). Categorical variables were presented as frequencies and percentages, and continuous variables were summarized using the mean and SD for normally distributed data and the median with IQR for non-normally distributed data. Comparisons between the sarcopenia and non-sarcopenia groups were conducted using the independent t-test or the Mann–Whitney U test for continuous variables and the Chi-square test or Fisher’s exact test for categorical variables, as appropriate. The association between PhA and sarcopenia was analyzed using binary logistic regression, with results reported as odds ratios (OR) and 95% CI. Additionally, linear regression analysis was performed to evaluate the association between PhA and basic clinical parameters, including age and body weight. A multivariable logistic regression model was constructed to adjust for potential confounders. The goodness-of-fit of the models was assessed using the Hosmer–Lemeshow test. To determine the diagnostic performance of PhA in predicting sarcopenia, receiver operating characteristic (ROC) curve analysis was performed. The area under the curve (AUC) with 95% CI was calculated to assess discrimination ability. The optimal PhA cutoff points for sarcopenia were determined using Youden’s index separately for men and women. Diagnostic test parameters, including sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), likelihood ratios (LR+ and LR−), and posterior probabilities, were calculated. Additionally, Pearson correlation analysis was performed to examine the relationships between PhA and sarcopenia-related components, including skeletal muscle mass index (SMI), handgrip strength, gait speed, and 5-time chair stand test performance. A scatter plot was used to illustrate these correlations. All statistical tests were 2-sided, and p-values <.05 were considered statistically significant. Data and resource availability The data supporting the findings of this study are not publicly available because participant consent obtained at the time of data collection does not permit open data sharing. Access to the data may be granted for collaborative research upon reasonable request and subject to institutional and ethical approvals. Interested researchers should contact the corresponding author to discuss potential collaboration and data access conditions. Results Baseline characteristics A total of 390 patients with type 2 diabetes were included in the analysis, comprising 199 men (51.0%) and 191 women (49.0%). All participants were of Thai origin and were recruited from patients attending the outpatient clinic at Phramongkutklao Hospital. Sarcopenia, defined by the AWGS 2019 criteria, was identified in 93 participants (23.8%). Sarcopenic individuals were significantly older than their non-sarcopenic counterparts (74.77 ± 8.48 vs 68.36 ± 6.48 yr, p < .001). Differences in educational background were also observed (p = .002). Sarcopenia was significantly associated with lower use of metformin (78.5% vs 88.2%, p = .019), higher use of DPP-4 inhibitors (40.9% vs 25.6%, p = .005), lower use of pioglitazone (8.6% vs 18.5%, p = .023), and higher use of calcium supplements (37.6% vs 15.5%, p < .001) (Table 1). Table 1Baseline characteristics of type 2 diabetes patients. Characteristics Overall (n = 390) Sarcopenia (n = 93) Non-sarcopenia (n = 297) p-value Sex, n (%) Men199 (51.0)37 (39.8)162 (54.6).013a Women191 (49.0)56 (60.2)135 (45.5) Age (yr) mean ± SD 69.89 ± 7.5274.77 ± 8.4868.36 ± 6.48<.001a Comorbidity, n (%) Hypertension312 (80.0)72 (77.4)240 (80.8).476 Dyslipidemia344 (88.2)83 (89.3)261 (87.9).721 MI/CAD42 (10.8)15 (16.1)27 (9.1).056 Cancer11 (2.8)3 (3.2)8 (2.7).728 Stroke25 (6.4)7 (7.5)18 (6.1).614 DM duration (year) median (IQR) 10 (5, 20)10 (5.5, 20)10 (5, 20).216 Family history of DM, n (%) 256 (65.6)57 (61.3)199 (67.0).311 Education, n (%) Primary school92 (23.6)32 (34.4)60 (20.2).002a Secondary school117 (30.0)16 (17.2)101 (34.0) Bachelor’s degrees181 (46.4)45 (48.4)136 (45.8) Smoking, n (%) 18 (4.6)2 (2.2)16 (5.4)1.000f Alcohol drinking, n (%) None331 (84.9)88 (94.6)243 (81.8) Sometime54 (13.8)5 (5.4)49 (16.5).007a Often5 (1.3)0 (0.0)5 (1.7) Medication, n (%) Metformin335 (85.9)73 (78.5)262 (88.2).019a Sulfonylurea137 (35.1)29 (31.2)108 (36.4).361 SGLT-2i122 (31.3)26 (27.7)96 (32.3).428 GLP1 RA52 (13.3)7 (7.5)45 (15.2).059 DPP-4i114 (29.2)38 (40.9)76 (25.6).005a Pioglitazone63 (16.2)8 (8.6)55 (18.5).023a Insulin118 (30.2)28 (30.1)90 (30.3).971 Statin351 (90.0)80 (86.0)271 (91.3).143 RAS inhibitor228 (58.5)46 (49.5)182 (61.3).044a Antiplatelet156 (40.0)39 (41.9)117 (39.4).662 CCB196 (50.2)37 (39.8)159 (53.5).021a Diuretic37 (9.5)9 (9.7)28 (9.4).943 Alpha blocker19 (4.9)1 (1.2)18 (6.1).055f Beta blocker68 (17.4)11 (11.8)57 (19.2).102 Vitamin D190 (48.7)46 (49.5)144 (48.5).869 Calcium81 (20.8)35 (37.6)46 (15.5)<.001a Steroid5 (1.3)2 (2.2)3 (1.0).343 Herbal30 (7.7)5 (5.4)25 (8.4).337 Polypharmacy151 (38.7)29 (31.2)122 (41.1).087 Diabetic neuropathy, n (%) 66 (16.9)20 (21.5)46 (15.5).177 Diabetic retinopathy, n (%) 48 (12.3)12 (12.9)36 (12.1).841 Body composition measurement Weight (kg) mean ± SD 67.62 ± 14.5753.44 ± 8.1172.07 ± 13.24<.001a Height (cm) mean ± SD 160.95 ± 8.89156.19 ± 8.15162.44 ± 8.60<.001a BMI (kg/m 2) mean ± SD25.99 ± 4.6721.89 ± 2.8627.28 ± 4.38<.001a Underweight12 (3.1)9 (9.7)3 (1.0) Normal82 (21.0)50 (53.8)32 (10.8)<.001a Overweight296 (75.9)34 (36.6)262 (88.2) Weight loss in 1 yr, n (%) 217 (55.6)56 (60.2)161 (54.2).309 Systolic blood pressure (mmHg) mean ± SD 134.39 ± 15.17132.51 ± 16.56134.97 ± 14.70.177 Diastolic blood pressure (mmHg) mean ± SD 70.42 ± 11.1766.42 ± 10.9471.54 ± 10.96<.001a Calf circumference (cm) mean ± SD 36.95 ± 4.1032.79 ± 2.4938.25 ± 3.61<.001a Waist circumference (cm) mean ± SD 94.50 ± 12.5684.55 ± 10.6597.61 ± 11.45<.001a Neck circumference (cm) mean ± SD 38.78 ± 4.1936.04 ± 4.9039.64 ± 3.54<.001a 5-Time chair stand (sec) median (IQR) 15.1 (12.3, 18.8)17.4 (13.7, 21.9)14.4 (11.7, 18.1)<.001a >12 s, n (%) 287 (73.6)79 (85.0)208 (70.0).004a Walk 4.5 m (s) median (IQR) 5.3 (4.6, 6.9)6.4 (4.9, 7.9)5.3 (4.5, 6.6)<.001a Gait speed (m/s) mean ± SD 0.82 ± 0.250.73 ± 0.270.85 ± 0.24<.001a <1 m/s, n (%) 299 (76.7)78 (83.9)221 (74.4).060 Hand grip strength (kg) mean ± SD 22.43 ± 8.1117.25 ± 6.9624.06 ± 7.77<.001a M < 28 kg; F < 18 kg, n (%) 208 (53.3)77 (82.8)131 (44.1)<.001a Body fat mass (kg) median (IQR) 22.8 (18.3, 28)18.2 (15.7, 21.2)24.7 (20.1, 29.6)<.001a Body fat percentage (%) mean ± SD 34.24 ± 7.9133.18 ± 7.5934.58 ± 7.99.138 M > 25%; F > 35%, n (%) 301 (77.2)62 (66.7)239 (80.5).006a Skeletal muscle mass (kg) mean ± SD 23.66 ± 5.5918.45 ± 3.3925.29 ± 5.13<.001a Skeletal muscle mass index (kg/m 2) mean ± SD6.88 ± 1.225.55 ± 0.777.30 ± 1.03<.001a M < 7 kg/m 2; F < 5.7 kg/m 2, n (%)104 (26.7)93 (100.0)11 (3.7)<.001aaStatistically significant at the .05 level (α = .05).Statistical analyses were performed using the Chi-square test, Fisher’s exact testf, Independent t-test, and Mann–Whitney U test.Abbreviations: CAD, coronary artery disease; DM, diabetes millitus; MI, myocardial infarction; SMI, skeleton muscle index. Baseline characteristics of type 2 diabetes patients. Statistically significant at the .05 level (α = .05). Statistical analyses were performed using the Chi-square test, Fisher’s exact testf, Independent t-test, and Mann–Whitney U test. Abbreviations: CAD, coronary artery disease; DM, diabetes millitus; MI, myocardial infarction; SMI, skeleton muscle index. Body composition and functional characteristics Sarcopenic patients had significantly lower body weight (53.44 ± 8.11 vs 72.07 ± 13.24 kg, p < .001), height (156.19 ± 8.15 vs 162.44 ± 8.60 cm, p < .001), and BMI (21.89 ± 2.86 vs 27.28 ± 4.38 kg/m2, p < .001). Skeletal muscle mass and SMI were also lower in the sarcopenic group (18.45 ± 3.39 vs 25.29 ± 5.13 kg and 5.55 ± 0.77 vs 7.30 ± 1.03 kg/m2, respectively; both p < .001). Consistent with these findings, sarcopenic patients had lower calf, waist, and NCs. Functionally, sarcopenic patients had longer 5-time chair stand test times (17.4 vs 14.4 s, p < .001), slower gait speed (0.73 ± 0.27 vs 0.85 ± 0.24 m/s, p < .001), and lower handgrip strength (17.25 ± 6.96 vs 24.06 ± 7.77 kg, p < .001). The proportion of patients with low muscle strength (M < 28 kg; F < 18 kg) was significantly higher in the sarcopenia group (82.8% vs 44.1%, p < .001) (Table 1). Additional physical performance variables, including fatigue, frailty, SARC-F, and SARC-Calf scores, are provided in Table 2. Table 2Physical fitness measurement of T2DM patients. Characteristics Overall (n = 390) Sarcopenia (n = 93) Non-sarcopenia (n = 297) p-value Walking time (h/wk), mean ± SD15.56 ± 6.7214.11 ± 6.4916.02 ± 6.74.017aFrequency of exercise, n (%) None178 (45.6)45 (48.4)133 (44.8).530 1-2 day/wk29 (7.4)7 (7.5)22 (7.4) 3-4 day/wk75 (19.2)13 (13.9)62 (20.9) >5 day/wk108 (27.7)28 (30.1)80 (26.9)Fatigue, n (%)144 (36.9)41 (44.1)103 (34.7).101SARC-F score, median (IQR)1 (0, 3)2 (1, 4)1 (0, 2)<.001a >4 score, n (%)70 (18.0)29 (31.2)41 (13.9)<.001aSARC-Calf score, median (IQR)2 (1, 5)10 (2, 13)1 (0, 3)<.001a >11 score, n (%)62 (15.9)46 (49.5)16 (5.4)<.001aFrailty phenotype, n (%) Non-frail26 (6.7)0 (0.0)26 (8.8)<.001a Pre-frailty204 (52.3)40 (43.0)164 (55.2) Frailty160 (41.0)53 (57.0)107 (36.0)aStatistically significant at the .05 level (α = .05).Statistical analyses were performed using the Chi-square test, Fisher’s exact testf, Independent t-test, and Mann–Whitney U test. Physical fitness measurement of T2DM patients. Statistically significant at the .05 level (α = .05). Statistical analyses were performed using the Chi-square test, Fisher’s exact testf, Independent t-test, and Mann–Whitney U test. Laboratory parameters and PhA Sarcopenic individuals had a significantly lower mean PhA than non-sarcopenic patients (3.72 ± 0.63° vs 4.44 ± 0.65°, p < .001). Stratified by sex, PhA remained lower in sarcopenic patients (men: 3.92 ± 0.68° vs 4.65 ± 0.67°, p < .001; women: 3.58 ± 0.56° vs 4.18 ± 0.52°, p < .001). They also had higher BUN (18 vs 15.78 mg/dL, p = .017), lower eGFR (60.2 vs 70.68 mL/min/1.73 m2, p = .005), and lower hemoglobin and hematocrit levels (p < .001 for both). Alanine aminotransferase (ALT) levels were also significantly lower in sarcopenic patients (p = .017) (Table S1). Linear regression analysis demonstrated that PhA was significantly associated with age and body weight. Increasing age was associated with lower PhA values, whereas higher body weight was positively associated with PhA. These findings are consistent with the known relationship between body composition and bioelectrical impedance parameters. Diagnostic performance of PhA Receiver operating characteristic analysis showed that PhA had strong discriminative ability for sarcopenia, with an AUC of 0.792 (95% CI: 0.740-0.844, p < .001) in the overall population (Figure 1). When stratified by sex, the AUC was 0.785 for men and 0.786 for women (Figure 2). The optimal cutoff points, determined by Youden’s index, were <4.6° for men (Youden index = 0.472) and <4.0° for women (Youden index = 0.451), representing the best balance of sensitivity and specificity (Table 3). Detailed diagnostic metrics for each cutoff point, including sensitivity, specificity, likelihood ratios, and predictive values, are shown in Table 4. Receiver operating characteristic (ROC) curve analysis of phase angle in sarcopenia diagnosis for T2DM patient. Receiver operating characteristic (ROC) curve analysis of phase angle in sarcopenia diagnosis classified by sex. Table 3Performance of PhA of predict sarcopenia in type 2 diabetes patient. Phase angle Area under the curve (95% CI) p-value Optimal cut-off point Youden index Overall0.792 (0.740, 0.844)<.001aMen0.785 (0.705, 0.865)<.001a<4.4°0.361<4.5°0.461<4.6°0.472b<4.7°0.429Women0.786 (0.714, 0.858)<.001a<3.8°0.422<3.9°0.424<4.0°0.451b<4.1°0.415aThe model’s AUC is significantly from reference line (AUC = 0.5).bThe maximum of the Youden index.Abbreviation: PhA, phase angle. Performance of PhA of predict sarcopenia in type 2 diabetes patient. The model’s AUC is significantly from reference line (AUC = 0.5). The maximum of the Youden index. Abbreviation: PhA, phase angle. Table 4Diagnostic test evaluation of phase angle classified by sex. Diagnostic test Men (n = 199) Women (n = 191) Optimal cut-off point <4.6° <4.0° Sensitivity (95% CI)89.2% (74.6%, 97.0%)73.2% (59.7%, 84.2%)Specificity (95% CI)58.0% (50.0%, 65.7%)71.9% (63.5%, 79.2%)Positive likelihood ratio (95% CI)2.12 (1.72, 2.63)2.60 (1.90, 3.56)Negative likelihood ratio (95% CI)0.19 (0.07, 0.47)0.37 (0.24, 0.58)Positive predictive value (95% CI)32.7% (23.7%, 42.7%)51.9% (40.4%, 63.3%)Negative predictive value (95% CI)95.9% (89.9%, 98.9%)86.6% (78.9%, 92.3%)Odds ratio (95% CI)11.40 (4.01, 32.24)6.98 (3.48, 13.97)Posterior probability positive (95% CI)32.6% (28.2%, 37.5%)51.9% (44.1%, 59.6%)Posterior probability negative (95% CI)4.2% (1.6%, 9.7%)13.3% (9.0%, 19.4%) Diagnostic test evaluation of phase angle classified by sex. Using Fagan’s nomogram, the post-test probability of sarcopenia increased from 18.6% to 32.6% for men with PhA <4.6° and from 29.3% to 51.9% for women with PhA <4.0° (Figure 3). Correlation analysis also demonstrated significant associations between PhA and all sarcopenia components: SMI (Figure 4A), handgrip strength (Figure 4B), gait speed (Figure 4C), and chair stand test (Figure 4D). The full correlation coefficients across sexes are presented in Table 5. The Fagan’s nomogram of phase angle in men (blue line) and women (red line). Scatter plot showing correlation between phase angle and skeletal muscle mass index, hand grip strength, gait speed, and 5-time chair stand. Table 5Correlation between phase angle and components of sarcopenia. Components Total (n = 390) Men (n = 199) Women (n = 191) r p-value r p-value r p-value Skeletal muscle mass index0.576<.001a0.445<.001a0.535<.001aHand grip strength0.572<.001a0.513<.001a0.417<.001aGait speed0.314<.001a0.303<.001a0.229.002a5-Time chair stand−0.440<.001a−0.465<.001a−0.273<.001aaStatistically significant at the .05 level (α = .05).Statistical analyses were performed using the Pearson correlation (r). Correlation between phase angle and components of sarcopenia. Statistically significant at the .05 level (α = .05). Statistical analyses were performed using the Pearson correlation (r). In multivariable logistic regression, lower PhA remained independently associated with sarcopenia, low muscle mass, and low handgrip strength after adjusting for age, MI/CAD, and GFR Table 6Factors associated with the sarcopenia component of the phase angle. Outcomes Crude OR (95% CI) p-value Adjusted OR b (95% CI) p-value Men Sarcopenia4.81 (2.55, 9.08)<.001a2.75 (1.38, 5.48).004aLow SMI4.37 (2.42, 7.91)<.001a3.14 (1.65, 5.99).001aLow hand grip strength2.85 (1.78, 4.54)<.001a1.94 (1.16, 3.26).012aLow gait speed1.91 (1.19, 3.07).008a1.37 (0.81, 2.30).242Poor 5-Time chair stand2.02 (1.29, 3.16).002a1.33 (0.81, 2.18).259 Women Sarcopenia7.53 (3.74, 15.16)<.001a6.77 (3.07, 14.94)<.001aLow SMI7.02 (3.55, 13.89)<.001a6.55 (3.03, 14.17)<.001aLow hand grip strength3.84 (2.12, 6.96)<.001a2.86 (1.50, 5.47).001aLow gait speed1.43 (0.77, 2.66).2630.91 (0.43, 1.89).790Poor 5-Time chair stand0.63 (0.33, 1.21).1640.66 (0.31, 1.40).273aStatistically significant at the .05 level (α = .05).Low SMI: men <7 kg/m2 and women <5.7 kg/m2; low hand grip strength: men <28 kg and women <18 kg; low gait speed: <1 m/s; poor 5-time chair stand: >12 s.Statistical analyses were performed using the binary logistic regression adjusted for age, MI/CAD, and GFR level. Factors associated with the sarcopenia component of the phase angle. Statistically significant at the .05 level (α = .05). Low SMI: men <7 kg/m2 and women <5.7 kg/m2; low hand grip strength: men <28 kg and women <18 kg; low gait speed: <1 m/s; poor 5-time chair stand: >12 s. Statistical analyses were performed using the binary logistic regression adjusted for age, MI/CAD, and GFR level. Discussion The present study confirms the predictive value of PhA for detecting sarcopenia in individuals with type 2 diabetes. Sarcopenic individuals had significantly lower PhA values, and PhA was strongly correlated with muscle mass, strength, and physical performance. Using Youden’s index, we identified sex-specific cutoff points: 4.6° for men and 4.0° for women, with high diagnostic accuracy. Multivariable regression further established PhA as an independent predictor of sarcopenia after adjusting for key confounders. Our results align with previous studies in diabetic populations. Oyanagi et al. reported positive associations between PhA and markers of muscle mass and function in Japanese patients with type 2 diabetes but did not establish diagnostic thresholds.17 Hafızoğlu et al. demonstrated that lower PhA predicted sarcopenic obesity in older Turkish adults with type 2 diabetes, identifying a cutoff of 4.4°, though a sex-specific analysis was not conducted.24 Our study builds on these findings by defining sex-specific thresholds and extending the diagnostic application of PhA to sarcopenia beyond obesity phenotypes. Recent studies have further supported the role of PhA as a marker of sarcopenia. A recent study by Matsuyama et al. also reported that lower PhA values were significantly associated with reduced muscle mass and functional impairment in older adults, supporting the utility of PhA as a practical biomarker for identifying sarcopenia in clinical populations. These findings are consistent with our results and further strengthen the evidence that PhA derived from BIA may serve as a convenient screening tool for sarcopenia.25 Prior research in non-diabetic cohorts has identified similar PhA cutoffs. Yamada et al. reported thresholds of 4.05° in men and 3.55° in women in community-dwelling older adults using AWGS 2019 criteria.26 Hirose et al., in a study of patients with established CVD, identified PhA cutoffs of 4.55° for men and 4.25° for women to signify sarcopenia.27 A study in cirrhotic patients in the U.S. reported even higher sex-specific cutoffs (5.6° in men and 5.4° in women).28 These differences likely reflect population-specific variation in body composition and disease burden, though they consistently support PhA as a valid biomarker for sarcopenia. Although numerical cutoff values vary, a consistent trend across studies, including ours, is that women have lower optimal PhA thresholds than men.29,30 This difference likely reflects sex-based differences in body composition. Women typically have higher fat mass and lower skeletal muscle mass than men, both of which influence BIA-derived PhA values.5,31 Age-related declines in muscle quality may further affect PhA differently between sexes,32 reinforcing the need for sex-specific cutoff values.29,33 From a clinical perspective, our findings suggest that PhA may serve as a valuable screening tool for sarcopenia in patients with type 2 diabetes. Early detection of sarcopenia is crucial for implementing timely interventions, including resistance training, nutritional optimization, and pharmacological strategies to preserve muscle mass and function. Given its noninvasive and accessible nature, BIA-derived PhA could be integrated into routine clinical practice as a cost-effective method for stratifying the risk of sarcopenia in diabetic patients. Despite its strengths, this study has several limitations. First, the cross-sectional design cannot establish causal relationships between PhA and the progression of sarcopenia. Longitudinal studies are needed to determine whether changes in PhA over time predict the development or worsening of sarcopenia. Second, because this study was conducted as a retrospective analysis of existing clinical data, a formal sample size calculation was not performed prior to data collection. Third, although BIA is widely used for body composition analysis, it may not offer the same precision as other methods, such as DXA or MRI. In addition, BIA measurements may be influenced by fluid imbalance or lower-limb conditions, such as peripheral edema or diabetic foot complications, which may affect impedance measurements. Furthermore, PhA is derived from BIA measurements, and sarcopenia in this study was also assessed using BIA-based muscle mass estimates. This methodological overlap may introduce potential measurement dependency between the predictor and outcome variables. Future studies using more precise body composition techniques are needed to further validate these findings. Lastly, the single-center design of our study may limit the generalizability of the findings, necessitating validation in larger, multicenter cohorts. In conclusion, this study establishes sex-specific PhA thresholds for sarcopenia detection in type 2 diabetes. PhA is a reliable, noninvasive marker of sarcopenia risk and may serve as a practical clinical tool to guide early detection and targeted intervention in this high-risk population. Supplementary Material Acknowledgments The authors would like to express their sincere gratitude to everyone who came forward to offer their assistance and useful ideas to ensure the successful completion of this study. Contributor Information Methavee Poochanasri, Department of Medicine, Bhumibol Adulyadej Hospital, Bangkok 10220, Thailand. Kasidid Lawongsa, Department of Family Medicine and Outpatient Department, Phramongkutklao Hospital, Bangkok 10400, Thailand. Thanapat Limpaarayakul, Division of Cardiology, Department of Medicine, Phramongkutklao Hospital, Bangkok 10400, Thailand. Jiranon Pirapattanapong, Division of Endocrinology, Department of Medicine, Phramongkutklao Hospital, Bangkok 10400, Thailand. Thawee Songpatanasilp, Department of Orthopedics, Phramongkutklao Hospital, Bangkok 10400, Thailand. Nattapol Sathavarodom, Division of Endocrinology, Department of Medicine, Phramongkutklao Hospital, Bangkok 10400, Thailand. Apussanee Boonyavarakul, Division of Endocrinology, Department of Medicine, Phramongkutklao Hospital, Bangkok 10400, Thailand. Parinya Samakkarnthai, Division of Endocrinology, Department of Medicine, Phramongkutklao Hospital, Bangkok 10400, Thailand. Author contributions Methavee Poochanasri and Kasidid Lawongsa contributed equally to this work. Methavee Poochanasri (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review & editing), Kasidid Lawongsa (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing—review & editing), Thanapat Limpaarayakul (Conceptualization, Data curation, Formal analysis, Investigation, Methodology), Jiranon Pirapattanapong (Conceptualization, Data curation, Formal analysis, Investigation, Methodology), Thawee Songpatanasilp (Conceptualization, Data curation, Methodology), Nattapol Sathavarodom (Conceptualization, Data curation, Methodology), Apussanee Boonyavarakul (Conceptualization, Data curation, Methodology), and Parinya Samakkarnthai (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Writing—review & editing) Funding This study was supported by the Thai Osteoporosis Foundation (TOPF). The funding did not influence the study design, data collection, analysis, manuscript preparation, or the decision to publish. Conflicts of interest There is no conflict of interest. Prior presentation Oral Presentation in Asian Federation of Osteoporosis Societies (AFOS) 2025, Bangkok, Thailand. References Associated Data Supplementary Materials