Association of the phase angle with type 2 diabetes and related traits: results from two prospective KORA studies

Nutrition & Diabetes ARTICLE www.nature.com/nutd OPEN Association of the phase angle with type 2 diabetes and related traits: results from two prospective KORA studies Feiling Ai1,2, Marie-Theres Huemer1, Wolfgang Rathmann3,4, Michael Roden 4,5,6, Christian Herder 4,5,6, Tanja Zeller7,8, ✉ Wolfgang Koenig 9,10,11, Jana Nano1,12, Michael Drey13, Annette Peters 1,2,10,14 and Barbara Thorand 1,2,14 1234567890();,: © The Author(s) 2026 OBJECTIVES: To investigate associations of the bioelectrical impedance analysis (BIA)-derived phase angle (PhA), an indicator of body cell mass, hydration status, and cell membrane integrity, with type 2 diabetes (T2D), prediabetes, and glycemic and insulinrelated traits. METHODS: Using data from the Cooperative Health Research in the Region of Augsburg (KORA) S3/S4 studies, we analyzed 7728 participants aged 25–74 years for prevalent T2D and 7006 participants who did not have diabetes at baseline for incident T2D. A subsample aged 55–74 years at S4 was followed to assess incident oral glucose tolerance test (OGTT)-defined prediabetes or T2D (prediabetes/T2D), and glycemic and insulin-related traits (S4/F4/FF4). The PhA was calculated from BIA 2000-S at 50 kHz. Logistic and Cox regressions were applied for binary outcomes, and two-level growth models for continuous traits. RESULTS: In S3/S4, 324 participants had prevalent T2D at baseline, and 707 developed T2D during a median 15.7-year follow-up. In S4/F4/FF4, during up to 14 years of follow-up, 251 out of 626 normoglycemic participants at S4 developed incident prediabetes/ T2D, and 792–804 participants without diabetes at S4 had three repeated measurements of continuous traits. The PhA (per 1-degree) was positively associated with incident T2D (hazard ratio [HR] and 95% confidence interval [CI] in S3/S4: 1.37 [1.21–1.54]) and incident prediabetes/T2D (HR [95% CI] in S4/F4/FF4: 1.33 [1.07–1.67]) without sex differences. The PhA (per 1-degree) was also positively associated with fasting glucose (beta [95% CI]: 1.2% [0.1–2.2%]) and insulin resistance (beta [95% CI]: 7.0% [2.3–11.7%]) cross-sectionally, and with changes in 2-h glucose longitudinally (beta [95% CI]: 4.5% [2.3–6.7%]) (S4/F4/FF4). In contrast, the PhA (per 1-degree) was inversely associated with prevalent T2D (odds ratio [95% CI] in S3/S4: 0.72 [0.56–0.93]) in men only. CONCLUSIONS: The PhA at 50 kHz had stage-dependent associations with glucose metabolism, with higher values observed during subclinical stages and lower values after diabetes manifestation. Nutrition and Diabetes (2026)16:11 ; https://doi.org/10.1038/s41387-026-00425-x INTRODUCTION Bioelectrical impedance analysis (BIA) is a non-invasive and relatively low-cost method for body composition assessment and has been implemented as an alternative to more invasive and costly techniques such as dual-energy X-ray absorptiometry, computerized tomography, and magnetic resonance imaging [1]. A key BIA-derived parameter is the phase angle (PhA), which is calculated from two raw BIA measurements of capacitive reactance (Xc) and resistance (R) [2]. The PhA serves as an indicator of body cell mass (BCM), cellular integrity, and tissue hydration status, particularly extracellular and intracellular water distribution [ECW/ICW] [3], supported by its associations with protein markers related to cell proliferation [4]. A higher PhA is mainly characterized by greater fat-free mass (FFM) and lower ECW/ICW ratios for both sexes [5]; while a lower PhA has been associated with detrimental cellular changes, such as reduced BCM, increased ECW/ICW ratios, and impaired cellular integrity [5, 6]. Throughout the lifespan, PhA values increase progressively from infancy to adolescence, stabilize during adulthood, and gradually decrease from around 50 years onwards [6]. Men tend to have higher values than women across life, which may result from their greater skeletal muscle mass [SMM] [6]. The PhA is positively 1 Institute of Epidemiology, Helmholtz Zentrum München, German Research Center for Environmental Health (GmbH), Neuherberg, Germany. 2Institute for Medical Information Processing, Biometry, and Epidemiology (IBE), Faculty of Medicine, LMU Munich, Pettenkofer School of Public Health, Munich, Germany. 3Institute for Biometrics and Epidemiology, German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University Düsseldorf, Düsseldorf, Germany. 4German Center for Diabetes Research (DZD), Partner Düsseldorf, Neuherberg, Germany. 5Department of Endocrinology and Diabetology, Medical Faculty and University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, Düsseldorf, Germany. 6Institute for Clinical Diabetology, German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich Heine University Düsseldorf, Düsseldorf, Germany. 7Institute for Cardiogenetics, University of Lübeck, University Hospital Schleswig-Holstein, University Heart Center Lübeck, Lübeck, Germany. 8 German Center for Cardiovascular Research (DZHK), Partner Site Nord, Lübeck, Germany. 9German Heart Center, TUM University Hospital, Munich, Germany. 10German Center for Cardiovascular Research (DZHK), Partner Site Munich Heart Alliance, Munich, Germany. 11Institute of Epidemiology and Medical Biometry, University of Ulm, Ulm, Germany. 12 Department of Radiation Oncology, Klinikum rechts der Isar, Technical University of Munich (TUM), Munich, Germany. 13Department of Medicine IV, LMU University Hospital, LMU Munich, Munich, Germany. 14German Center for Diabetes Research (DZD), Partner München-Neuherberg, Neuherberg, Germany. ✉email: Received: 26 August 2025 Revised: 9 April 2026 Accepted: 21 April 2026 F. Ai et al. 2 associated with body mass index (BMI) in individuals with normal or moderately elevated BMI; however, an inverse association has been observed when BMI exceeded 35 kg/m² [7] or 40 kg/m² [8]. Recently, the PhA has emerged as a promising biomarker for assessing inflammation, oxidative stress, muscle composition, cardiovascular risk, and nutritional status in metabolic diseases [9–11]. Few prior studies have indicated that the PhA may reflect underlying metabolic and cellular alterations in individuals with diabetes and related complications [12]. However, existing crosssectional studies have yielded inconsistent findings and were constrained by small sample sizes and minimal adjustments for potential confounders [13–18]. Moreover, no study to date has explored the longitudinal associations of the PhA with incident type 2 diabetes (T2D), incident prediabetes, or with changes in glycemic and insulin-related traits, leaving its potential role in early glucose dysregulation unclear. Therefore, using data from two population-based prospective cohorts, the present study aimed (1) to investigate the associations of the PhA with prevalent T2D and incident T2D; (2) to examine the longitudinal associations of the PhA with incident oral glucose tolerance test (OGTT)-defined prediabetes (...truncated)