Rao ZC, Xu CY, Hua LY, Fang ZR, Zhu CY, Zhang YM. Visceral fat area and red-blood-cell-related indices in adults with type 2 diabetes: Cross-sectional evidence of nonlinear associations. World J Diabetes 2026; 17(8): 121923 [DOI: 10.4239/wjd.121923]
Corresponding Author of This Article
Yi-Ming Zhang, MD, Doctor, Department of Endocrinology, The Quzhou Affiliated Hospital of Wenzhou Medical University, Quzhou People’s Hospital, No. 100 Minjiang Road, Kecheng District, Quzhou 324000, Zhejiang Province, China. zhangym1611@wmu.edu.cn
Research Domain of This Article
Endocrinology & Metabolism
Article-Type of This Article
research-article
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Rao ZC, Xu CY, Hua LY, Fang ZR, Zhu CY, Zhang YM. Visceral fat area and red-blood-cell-related indices in adults with type 2 diabetes: Cross-sectional evidence of nonlinear associations. World J Diabetes 2026; 17(8): 121923 [DOI: 10.4239/wjd.121923]
Zi-Chen Rao, Liang-Yan Hua, Zi-Ru Fang, Chun-Yan Zhu, Yi-Ming Zhang, Department of Endocrinology, The Quzhou Affiliated Hospital of Wenzhou Medical University, Quzhou People’s Hospital, Quzhou 324000, Zhejiang Province, China
Chun-Yi Xu, Department of Endocrinology and Metabolism, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou 310000, Zhejiang Province, China
Author contributions: Rao ZC and Xu CY contributed equally to this work and share first authorship; Zhang YM conceived the study, critically revised the manuscript and supervised the work; Rao ZC, Xu CY, and Hua LY curated the data; Rao ZC and Xu CY performed the statistical analyses; Fang ZR and Zhu CY contributed to data interpretation and manuscript revision; Rao ZC drafted the manuscript; all authors have read and approved the final manuscript.
AI contribution statement: No AI tool was involved in the generation of research data, interpretation of results, or formulation of conclusions. During the preparation of the point-by-point response, we used AI-assisted tools only for limited language support, mainly to improve English wording and sentence clarity. The reviewers’ comments were understood and answered by the authors ourselves. No AI tool was used to generate or change the study data, statistical results, tables, figures, interpretation of results, or conclusions. All revisions in the manuscript were made and checked by the authors. We have also reviewed the response document again and confirm that the scientific content and the replies to the reviewers reflect the authors’ own work and judgment.
Institutional review board statement: The study protocol complied with the Declaration of Helsinki and was approved by the Ethics Committee of Quzhou People’s Hospital (Quzhou Affiliated Hospital of Wenzhou Medical University) (approval No. 2022-110).
Informed consent statement: The requirement for informed consent was waived because this study was a cross-sectional analysis of de-identified routine-care data and posed minimal risk to participants.
Conflict-of-interest statement: The authors declare that they have no conflict of interest.
STROBE statement: The authors have read the STROBE Statement—a checklist of items, and the manuscript was prepared and revised according to the STROBE Statement-a checklist of items.
Data sharing statement: The datasets used and/or analyzed during the current study are not publicly available due to institutional regulations and data privacy requirements, but are available from the corresponding author on reasonable request.
Corresponding author: Yi-Ming Zhang, MD, Doctor, Department of Endocrinology, The Quzhou Affiliated Hospital of Wenzhou Medical University, Quzhou People’s Hospital, No. 100 Minjiang Road, Kecheng District, Quzhou 324000, Zhejiang Province, China. zhangym1611@wmu.edu.cn
Received: April 7, 2026 Revised: May 4, 2026 Accepted: June 12, 2026 Published online: August 15, 2026 Processing time: 122 Days and 22.8 Hours
Abstract
BACKGROUND
Visceral adiposity is common in type 2 diabetes (T2D) and is linked to cardiometabolic risk. Complete blood counts are routinely measured, but their relationship with visceral fat in T2D is not well described.
AIM
To evaluate the association between visceral fat area (VFA) and red blood cell (RBC)-related indices in Chinese adults with T2D, and to explore potential nonlinear relationships and threshold effects across different patient subgroups.
METHODS
We conducted a cross-sectional study among adults with T2D at the Metabolic Management Center of Quzhou People’s Hospital. A formal sample size calculation was performed to ensure sufficient statistical power. VFA was measured by multifrequency bioelectrical impedance analysis. We examined RBC count, hemoglobin (HGB), hematocrit (HCT) and other blood cell indices. To address potential selection bias and maximize statistical power, multiple imputation by chained equations was used to handle missing glycated HGB (HbA1c) data (n = 1180). We fitted multivariable linear regression models with each hematological index as the outcome and VFA as the exposure. We used generalized additive models and two-piecewise linear regression to assess nonlinearity and threshold effects, where nonlinearity refers to associations that deviate from a straight line, such as those exhibiting plateau or threshold patterns. We also performed prespecified subgroup analyses by age, sex, body mass index (BMI) and HbA1c. Sensitivity analyses were conducted to compare the results between the complete-case (n = 590) and the imputed datasets to ensure the robustness of the findings.
RESULTS
We included 1180 adults with T2D. In fully adjusted models, using multiple imputation to account for missing glycemic data, higher VFA was independently associated with higher HCT, HGB, and RBC and white blood cell (WBC) counts. Smoothing curves revealed nonlinear patterns for HCT, HGB and RBCs: The increases were steeper at lower VFA levels and reached a plateau at higher levels. In threshold models, the inflection point was approximately 45 cm2 for HCT, HGB and RBCs (all P < 0.001 for the nonlinear trend). Subgroup analyses showed that these associations were more robust in men and in participants with BMI < 28 kg/m2. Sensitivity analyses confirmed that the results were consistent across HbA1c categories, and the inclusion of imputed data significantly enhanced the statistical power for the association between VFA and WBC count.
CONCLUSION
In adults with T2D, greater visceral adiposity was independently and robustly associated with higher HCT, HGB, and RBC and WBC counts, with the relationships appearing nonlinear and characterized by a threshold effect. These findings, which remained consistent after addressing missing data through multiple imputation, suggest that routine hematological indices are sensitive indicators of visceral fat burden, particularly in specific clinical subgroups. Longitudinal studies are needed to clarify the temporal sequence and the potential clinical utility of integrating these indices into metabolic risk stratification.
Core Tip: This study revealed a nonlinear relationship between visceral fat area (VFA) and red blood cell-related indices, specifically hematocrit, hemoglobin and red blood cell count, in adults with type 2 diabetes (T2D). The association was more pronounced at lower VFA levels (inflection point approximately 45 cm2) and in nonobese individuals. These findings suggest that routine hematological markers serve as accessible indicators of visceral adiposity burden. This emphasizes the importance of integrated metabolic and hematological monitoring to better assess cardiometabolic risk in T2D management.
Citation: Rao ZC, Xu CY, Hua LY, Fang ZR, Zhu CY, Zhang YM. Visceral fat area and red-blood-cell-related indices in adults with type 2 diabetes: Cross-sectional evidence of nonlinear associations. World J Diabetes 2026; 17(8): 121923
Type 2 diabetes mellitus (T2D) is one of the most common chronic diseases worldwide and is associated with a high burden of cardiovascular and microvascular complications[1]. Even when glucose control is improved, many patients with T2D still have substantial residual cardiovascular risk. In this context, obesity especially central or visceral obesity is regarded as a key, but potentially modifiable, risk factor associated with adverse outcomes in T2D[2,3].
Visceral adipose tissue, which accumulates around the abdominal organs, is more metabolically active than subcutaneous fat and has stronger effects on insulin resistance, dyslipidemia and systemic inflammation[4,5]. Patients with T2D often have increased visceral fat even when their body mass index (BMI) is not markedly elevated. Visceral fat area (VFA), usually measured by computed tomography, is therefore considered a more precise marker of harmful adiposity than BMI or waist circumference (WC), particularly in people with T2D[6,7].
Red blood cell (RBC)-related indices such as RBC count, hemoglobin (HGB) and hematocrit (HCT) are routinely measured in clinical practice[8,9]. These indices reflect oxygen-carrying capacity and blood viscosity[9]. Previous studies have shown that elevated HGB and HCT are associated with increased blood viscosity, endothelial dysfunction, and a higher risk of cardiovascular and thrombotic events[10,11]. In patients with T2D, abnormal RBC indices may worsen tissue hypoxia, impair microcirculation and promote a prothrombotic state, thereby contributing to vascular complications[12,13]. Several studies have suggested that people with obesity tend to have higher RBC indices than those with normal weight, but data specifically focusing on patients with T2D are limited[14,15].
Most previous studies examining obesity and blood cell indices have used BMI or WC as surrogate measures of adiposity[16,17]. These anthropometric indices do not distinguish between visceral and subcutaneous fat and may not fully capture the metabolic risk associated with central obesity in T2D[18]. Evidence directly relating VFA to RBC indices in patients with T2D is still scarce, and it is unclear whether the association is linear across the whole VFA range or whether there are threshold effects[19].
Against this background, we conducted a cross-sectional study in adults with T2D from a real-world metabolic management center (MMC) to examine the relationship between VFA and RBC-related indices. Our primary aim was to evaluate the association of VFA with RBC count, HGB and HCT after adjustment for potential confounders. We also used smooth curve fitting and piecewise linear models to characterize the shape of these associations and to explore potential threshold effects. Finally, we performed subgroup analyses to assess whether the association between VFA and hematological indices differed by key clinical characteristics in patients with T2D.
MATERIALS AND METHODS
Study design and population
This cross-sectional study was conducted at the MMC of the Department of Endocrinology, Quzhou People’s Hospital, The Quzhou Affiliated Hospital of Wenzhou Medical University (Quzhou, Zhejiang Province, China). The MMC provides standardized assessment and follow-up for adults with T2D.
Consecutive patients who attended the MMC between December 2022 and June 2025 were screened. Eligible participants met the following criteria: (1) Age ≥ 18 years; (2) Diagnosis of T2D according to the American Diabetes Association criteria; and (3) Completion of the baseline MMC evaluation with available data on VFA, complete blood count and key covariates.
Patients were excluded if they had: (1) Acute diabetic complications at the time of evaluation (e.g., diabetic ketoacidosis or hyperosmolar hyperglycemic state); (2) Type 1 diabetes, gestational diabetes or other specific types of diabetes; (3) Known severe hepatic or renal insufficiency (such as decompensated cirrhosis, end-stage renal disease or maintenance dialysis); (4) Clinically evident acute infection or other acute inflammatory disease; (5) Active malignancy, hematological disorders, recent blood transfusion or other severe systemic illness that could markedly affect blood cell indices; or (6) Pregnancy, lactation or missing data on VFA, hematological indices or major covariates required for the analysis. After applying these criteria, 1180 adults with T2D were included in the final analysis.
The sample size was determined based on the primary objective of detecting a significant association between VFA and RBC indices. Assuming a small effect size (correlation coefficient r = 0.1) with an alpha of 0.05 and power of 0.90, a minimum sample size of approximately 1050 participants was required. Our final sample of 1180 participants provided sufficient statistical power.
Clinical and anthropometric measurements
Demographic information (age and sex), and lifestyle factors (current smoking and current drinking) were collected at baseline by trained nurses using standardized questionnaires. Body weight and height were measured with participants wearing light clothing and no shoes. WC was measured with a nonelastic tape at the midpoint between the lowest rib and the iliac crest after a normal expiration. Blood pressure was measured on the right arm in the seated position after at least 5 minutes of rest using an automated sphygmomanometer. Two consecutive readings were obtained at 1-2 minutes intervals, and the average values of systolic blood pressure (SBP) and diastolic blood pressure (DBP) were used in the analyses. VFA and subcutaneous fat area were assessed using a standardized multifrequency bioelectrical impedance body composition analyzer in the MMC.
Laboratory measurements and hematological indices
Fasting venous blood samples were drawn in the morning after an overnight fast of at least 8 hours. Glycated HGB (HbA1c) was measured by high-performance liquid chromatography. Serum total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C), uric acid (UA), liver enzymes [aspartate aminotransferase (AST) and alanine aminotransferase (ALT)] and serum creatinine were determined using standard enzymatic methods on automated analyzers under strict internal and external quality control procedures. C-reactive protein (CRP) was measured using a standard immunoassay on automated analyzers with internal and external quality control. Estimated glomerular filtration rate (eGFR) was calculated from serum creatinine, age and sex using the chronic kidney disease epidemiology collaboration equation.
Complete blood counts were measured on automated hematology analyzers in the hospital laboratory. The following indices were analyzed: RBC count, HGB, HCT, mean corpuscular volume (MCV), mean corpuscular HGB (MCH), mean corpuscular HGB concentration (MCHC) and white blood cell (WBC) count. To facilitate the presentation and interpretation of regression coefficients, RBC, HCT, MCV, MCH and MCHC were multiplied by 100 and expressed as RBC100, HCT100, MCV100, MCH100 and MCHC100, respectively; this rescaling did not affect the direction or statistical significance of the associations.
Definitions of exposure, outcome and covariates
The primary exposure of interest was VFA, analyzed as a continuous variable (per 1 cm2 increment). For descriptive comparisons, participants were also classified into two groups according to VFA (< 80 cm2 vs ≥ 80 cm2). The main outcomes were the hematological indices described above. Each blood cell parameter was first analyzed individually in relation to VFA. Potential confounders were selected a priori based on clinical relevance and previous literature. The final adjustment set included age, sex, eGFR, current smoking, current drinking, TC, TG and HbA1c. BMI and WC were considered descriptive indicators of general and abdominal adiposity and were additionally used in subgroup analyses.
Statistical analysis
Cross-sectional evidence of nonlinear associations refers to findings from a single-point-in-time assessment demonstrating that the relationship between an exposure (VFA) and an outcome (RBC indices) does not follow a constant straight line. Instead, the effect changes at different levels or thresholds, such as the observed plateau patterns.
Continuous variables were expressed as median (interquartile range) because of skewed distributions, and categorical variables as n (%). Baseline characteristics were compared between VFA groups (< 80 cm2 vs ≥ 80 cm2) using the Wilcoxon rank-sum test (Mann-Whitney U test) for continuous variables and the c2 test (or Fisher’s exact test when appropriate) for categorical variables. The associations between blood cell indices and VFA were evaluated using linear regression models. VFA was entered as the dependent variable, and each hematological index as the independent variable, modeled both continuously and in quartiles. Three models were fitted: (1) A non-adjusted model with no covariates; (2) Model 1, adjusted for age and sex; and (3) Model 2, additionally adjusted for eGFR, current smoking, current drinking, TC, TG and HbA1c. Regression coefficients (β) and their 95% confidence intervals were reported. Linear trends across quartiles of each hematological index were assessed by entering the quartile variable as an ordinal term in the models.
To further characterize potential nonlinear relationships, generalized additive models with penalized smoothing splines were used to fit smoothing curves between VFA and those blood cell indices that showed robust linear associations in the primary analyses (HCT100, HGB and RBC100). When a nonlinear pattern was suggested, two-piecewise linear regression models with a single inflection point were applied to estimate threshold effects. The optimal inflection point was determined using a likelihood-based recursive algorithm that searched over a grid of candidate values. A likelihood ratio test was used to compare the two-piecewise model with a simple linear model.
Prespecified subgroup analyses were conducted for HCT100, HGB and RBC100 to examine whether their associations with VFA differed by age (< 60 years vs ≥ 60 years), sex (male vs female), BMI (< 28 kg/m2 vs ≥ 28 kg/m2) and glycemic control (HbA1c < 7.0% vs ≥ 7.0%). Within each stratum, we fitted the fully adjusted model (model 2), excluding the stratifying variable itself. P values for interaction were obtained by including a multiplicative cross-product term between the hematological index and the subgroup indicator in the corresponding model.
To address potential selection bias and maximize statistical power, multiple imputation by chained equations was used to handle missing values of HbA1c (missing in 39.7% of the cohort). Five imputed datasets were generated, and the results were pooled according to Rubin’s rules. All baseline variables and outcomes were included in the imputation model. Sensitivity analyses were performed by comparing the results from the imputed dataset (n = 1180) with those from the complete-case analysis (n = 590 for the fully adjusted model) to ensure the robustness of the findings. All analyses were based on available data. A two-sided P value < 0.05 was considered statistically significant. Statistical analyses were performed using EmpowerStats (www.empowerstats.com) and R software (version 4.2.2).
Ethics statement
The study protocol complied with the Declaration of Helsinki and was approved by the Ethics Committee of Quzhou People’s Hospital, The Quzhou Affiliated Hospital of Wenzhou Medical University (approval No. 2022-110). The requirement for informed consent was waived because this study was a cross-sectional analysis of de-identified routine-care data and posed minimal risk to participants.
RESULTS
Baseline characteristics according to VFA
A total of 1180 participants were included in this study; 772 (65.4%) had a VFA ≥ 80 cm2 and 408 (34.6%) had VFA < 80 cm2. Age was similar between the two groups, but participants with VFA ≥ 80 cm2 were more likely to be men and to report current drinking and current smoking (all P < 0.001). Compared with those with VFA < 80 cm2, participants with VFA ≥ 80 cm2 also had higher SBP and DBP, BMI, TG, UA, AST and ALT, CRP and WBC count, while HDL-C levels were lower (all P < 0.001). TC, LDL-C, HbA1c and eGFR were broadly similar between the groups. With respect to blood cell indices, higher VFA was associated with higher RBC count, HGB and HCT (all P < 0.001), whereas MCV, MCH, MCHC, platelet count and platelet indices (mean platelet volume and platelet distribution width) did not differ significantly between the VFA categories (Table 1).
Table 1 Baseline characteristics of the study population according to visceral fat area, n (%)/median (interquartile range).
In multivariable linear models after multiple imputation (n = 1180), higher VFA was positively associated with primary hematological indices. When VFA was modeled as a continuous variable, it showed significant positive associations with RBC, HGB, HCT and WBC count across all three models. These associations remained robust after adjustment for age, sex, eGFR, smoking, drinking, TG, TC and HbA1c (Table 2). Compared with the lowest VFA quartile (Q1), participants in higher quartiles had progressively higher RBC, HGB, HCT and WBC count, with significant linear trends observed in all models (all P for trend < 0.0001). Sensitivity analysis using the complete-case dataset (n = 590) yielded highly consistent result. In contrast, mean corpuscular indices (MCV, MCH and MCHC) showed no consistent or clinically significant linear associations with VFA in the fully adjusted models (Supplementary Tables 1 and 2).
Table 2 Association between visceral fat area and hematological indices after multiple imputation (n = 1180).
To facilitate clinical interpretation, we provide a worked example based on the fully adjusted model (adjust II). The regression coefficient (β) for HGB was 0.05, indicating that for every 10 cm2 increase in VFA, HGB levels increase by 0.5 g/L. For rescaled indices such as RBC100 (β = 0.20) and HCT100 (β = 0.02), the coefficients translate to an actual increase of 0.02 × 1012/L and 0.002 L/L (approximately 0.2%), respectively, for every 10 cm2 increment in VFA. This conversion logic applies to all rescaled data presented in the study.
Smooth curve analysis between VFA and RBC indices
In fully adjusted generalized additive models, smooth curves were used to visualize the association between VFA and RBC indices, including HGB, RBC count (RBC100) and HCT (HCT100) (Figure 1). All three curves showed a broadly increasing pattern: HGB, RBC100 and HCT100 rose steeply at lower VFA levels and then gradually levelled off as VFA entered the middle and higher ranges, indicating a modest nonlinear relationship rather than a simple straight line. At high VFA, HGB tended to plateau or even decline slightly, whereas RBC100 and HCT100 remained stable with only mild further increases. In line with these patterns, threshold (two-piecewise) models identified an inflection point in the mid-range of VFA for all three indices, with a much steeper positive slope below the inflection point and a clearly attenuated association above it (Table 3). Conversely, WBC count showed a consistent linear increase across the VFA range without a significant threshold effect (P for likelihood ratio test = 0.549). Overall, both the smooth curves and the threshold models consistently suggested that increasing VFA is associated with higher levels of HGB, RBC100 and HCT100, and that the strength of this association diminishes once VFA exceeds a certain level.
Figure 1 Smoothed dose-response relationships between visceral fat area and hematologic indices.
A: Red blood cell count (× 100); B: Hematocrit (× 100); C: Hemoglobin; D: White blood cell. The solid red lines represent the estimated smooth functions, and the dashed blue lines indicate 95% confidence intervals. Rug plots along the X-axis denote the distribution of visceral fat area (VFA). While red blood cell, hematocrit, and hemoglobin (Figure 1A-C) exhibit a rapid initial increase followed by a plateau at higher VFA levels (suggesting non-linear associations), the white blood cell count (Figure 1D) shows a consistent linear increase across the entire observed range of visceral adiposity.
Table 3 Threshold effect analysis of the association between visceral fat area and red blood cell indices.
Subgroup analyses of the associations between VFA and RBC indices
In the prespecified subgroup analyses, the positive associations of HCT100, HGB and RBC100 with VFA were broadly consistent, but their magnitudes varied across clinical strata. Overall, these hematological indices showed clearer and stronger relationships with VFA in men than in women, in whom the associations were generally weak or absent. For HGB and RBC100, the associations were also more pronounced among participants with BMI < 28 kg/m2, whereas they were attenuated and no longer significant in those with BMI ≥ 28 kg/m2. In addition, the link between RBC100 and VFA was evident in individuals aged < 60 years but not in those aged ≥ 60 years. By contrast, none of the three indices demonstrated a meaningful interaction with HbA1c categories (Table 4).
Table 4 Subgroup analyses of the associations of hematocrit 100, hemoglobin and red blood cell 100 with visceral fat area.
In this cross-sectional study of adults with T2D, VFA was positively associated with RBC-related indices. Higher VFA was consistently linked to higher RBC count, HGB and HCT in all multivariable models, and these associations persisted after adjustment for renal function, lipids, glycemic control and lifestyle factors. Smoothing and threshold analyses suggested nonlinear relationships, with steeper increases at lower VFA levels and a plateau at higher levels. By contrast, the association with WBC count was modest, and mean corpuscular indices showed no robust linear relationships. In subgroup analyses, the positive associations between VFA and RBC count, HGB and HCT were stronger in men, in participants with BMI < 28 kg/m2 and, RBC count, in those aged < 60 years, while they were broadly similar across HbA1c strata.
Several clinical studies have shown that obesity is accompanied by higher RBC-related indices[20]. Lee and Kim[21] reviewed the pathophysiological role of visceral adipose tissue in cardiometabolic diseases and emphasized that visceral adiposity is closely associated with metabolic dysfunction, chronic inflammation and cardiometabolic risk. Broadly similar patterns have been observed in adults[22]. Christakoudi et al[23] found that, among adults without severe comorbidities, obese individuals had consistently higher HGB, HCT and RBC counts than normal-weight subjects and proposed the concept of obesity-related enhancement of erythroid activity. Oliveira et al[24] showed that adiposity-related measures, including body fat percentage, fat mass index and BMI, differed in their performance for detecting cardiometabolic outcomes in adults. A systematic review and meta-analysis including 26 studies and > 10000 overweight or obese participants confirmed a stable positive association of overweight and obesity with higher HGB, HCT and RBC count[25]. Overall, these data support the notion that excess adiposity, particularly abdominal obesity, is closely linked to higher RBC-related indices[26]. In contrast, evidence in patients with T2D is limited: Most studies have relied on BMI or WC as anthropometric markers to examine their relationships with HGB or HCT, or have treated blood cell indices as covariates in models of cardiovascular outcomes, with few directly assessing the association between RBC-related indices and visceral fat[27].
Multiple lines of evidence support a link between visceral fat and higher RBC indices[28]. In obese animal models, expanding white adipose tissue becomes markedly hypoxic and produces more inflammatory cytokines and adipokines[29]. Takikawa et al[30] demonstrated evidence that expanding adipose tissue in obesity becomes hypoxic, with activation of hypoxia-inducible factor-1α (HIF-1α)-related pathways, inflammatory infiltration and adipokine dysregulation. Anatomically and metabolically, visceral fat is the most active and hypoxia-prone depot within white adipose tissue[31]. Consistently, human biopsies of visceral adipose tissue from obese individuals demonstrate reduced oxygen tension and upregulation of HIF-1α and inflammatory genes, whereas subcutaneous fat is less affected, indicating that expanding visceral fat operates in a state of hypoxia and low-grade inflammation[32,33]. On this background, accumulating experimental and clinical evidence suggests that adipose hypoxia and inflammation may enhance erythropoietic activity through activation of the erythropoietin (EPO) bone marrow axis[34]. Under hypoxic conditions, adipose-derived stromal cells have been shown to upregulate EPO expression[35]. In obese animal models, increased renal HIF signaling and EPO expression are accompanied by modest rises in HGB and HCT[36]. Consistently, clinical studies report higher circulating EPO levels in severely obese individuals than in normal-weight controls. These levels correlate positively with fat mass and WC, whereas HGB and HCT usually remain within the high-normal range[37-39]. Notably, substantial weight loss is associated with partial improvement in EPO levels and hemorheological abnormalities[40].
In more advanced obesity and T2D, however, disturbances of iron homeostasis and obesity-related anemia become increasingly common[41]. Chronic inflammatory signaling is associated with the upregulation of hepatic hepcidin, which is consistent with reduced intestinal iron absorption and sequestration of iron within the reticuloendothelial system[42]. These changes progressively constrain erythropoietic capacity despite ongoing adipose tissue hypoxia[43,44]. These findings support a biological pathway whereby expansion of visceral fat is associated with adipose hypoxia and inflammation, which may be linked to the activation of the EPO erythropoiesis axis and subsequent elevations in RBC count, HGB and HCT[45,46]. In more advanced disease, iron-restricted erythropoiesis may partially blunt this response. The observed nonlinear pattern, which is characterized by a steep increase at lower VFA levels and a subsequent plateau, suggests a potential saturation effect of adipose-derived erythropoietic stimuli. At lower ranges of visceral adiposity, the bone marrow might exhibit heightened sensitivity to initial increases in inflammatory cytokines or adipose-derived EPO. However, as VFA exceeds a certain threshold, such as the 45 cm2 identified in our model, these stimulatory effects may be counterbalanced by systemic chronic inflammation and hepcidin-mediated iron sequestration, thereby preventing a further linear rise in RBC indices. This framework is consistent with the positive but nonlinear associations between VFA and RBC-related indices observed in our cohort of adults with T2D[47,48].
The identification of this nonlinear relationship, particularly the inflection point at approximately 45 cm2, provides a critical threshold for understanding the transition from physiological compensation to pathological dysregulation. This saturation effect likely reflects a complex biological trade-off: While initial VFA expansion triggers adaptive erythropoiesis to counter localized tissue hypoxia, excessive visceral adiposity may concurrently upregulate hepatic hepcidin, thereby constraining further iron-dependent RBC production. From a clinical perspective, this positive correlation especially within the lower VFA range carries significant hemorheological implications. Elevated RBC indices directly contribute to increased blood viscosity, which may exacerbate microvascular malperfusion in patients already compromised by T2D-related endothelial dysfunction. Consequently, the integration of routine hematological parameters with VFA assessment could refine risk stratification, helping to identify individuals who, despite having a relatively modest BMI, may be at a heightened risk for hyperviscosity-related vascular complications.
Our findings may have several implications for the management of patients with T2D. RBC count, HGB and HCT are inexpensive and routinely available laboratory parameters[8]. Beyond their traditional roles in identifying anemia or hemoconcentration, these indices show a significant association with underlying visceral fat burden in selected patients with T2D[49]. Although RBC-related indices cannot replace direct assessment of visceral adiposity, they may help clinicians identify individuals who simultaneously have excess visceral fat and an increased risk of an unfavorable hemorheological profile. In routine practice, considering these indices together with VFA and established cardiovascular risk factors may allow more refined risk stratification and more individualized management[50]. Prospective cohort and interventional studies are still needed to determine whether incorporating RBC-related indices in addition to VFA can improve prediction of vascular outcomes or guide therapeutic decision-making.
This study had several strengths. We focused on patients with T2D managed in a real-world MMC, a population in whom visceral fat accumulation and hemorheological abnormalities are both common and clinically relevant. Unlike studies relying solely on BMI or WC, we directly quantified VFA rather than using overall adiposity as a proxy, enabling a more accurate characterization of metabolically harmful fat in patients with T2D. Notably, the robustness of our findings was confirmed by multiple imputation (n = 1180). The high consistency in regression coefficients between the complete-case and imputed datasets suggests that missing HbA1c data did not introduce significant selection bias, while the enhanced significance of WBC count underscores the recovery of statistical power through rigorous statistical handling.
This study also had several limitations. First, despite the robust associations observed, its single-center, cross-sectional design precludes causal or temporal inference regarding the relationship between VFA and hematological indices, and VFA was estimated using multifrequency bioelectrical impedance rather than computed tomography or magnetic resonance imaging, which may limit generalizability. Second, although multiple potential confounders were adjusted for, residual confounding cannot be excluded, particularly from unmeasured chronic inflammatory or hypoxic conditions, iron status, medications influencing erythropoiesis and nutritional factors. Third, our stringent exclusion criteria, while ensuring a clean baseline to minimize confounding from severe systemic illnesses, may limit the generalizability of our findings to patients with advanced complications or acute inflammatory states. In addition, VFA and blood cell indices were measured only once, precluding assessment of intraindividual variability or longitudinal change. Finally, although subgroup and interaction analyses were prespecified, the number of comparisons was large and sample sizes within some strata were modest; these findings should therefore be interpreted as exploratory and warrant confirmation in independent cohorts and prospective studies.
Moving forward, several avenues for future research are identified. First, longitudinal studies with repeated measurements of VFA and hematological indices are needed to clarify the temporal sequence and potential causal links. Second, interventional trials focusing on visceral fat reduction through intensive lifestyle modification, pharmacotherapy (e.g., glucagon-like peptide-1 receptor agonists), or metabolic surgery could help determine the reversibility of these hematological changes. Finally, incorporating biomarkers of iron metabolism (e.g., hepcidin) and systemic hypoxia would provide deeper mechanistic insights into the adipose-erythropoiesis axis in patients with T2D.
CONCLUSION
This study was conducted to address the limited evidence regarding the complex interplay between visceral fat and hematological parameters in the context of T2D. In adults with T2D, greater VFA was positively associated with higher RBC count, HGB and HCT, with nonlinear and phenotype-specific patterns. These findings provide novel insights into the hemorheological aspects of visceral adiposity, suggesting that routine hematological indices may serve as accessible markers for assessing metabolic health. Beyond simple clinical correlation, this evidence highlights the potential for integrating hematological monitoring into the risk stratification of patients with T2D. These associations may help explain how visceral adiposity is linked to hemorheological changes in T2D and warrant confirmation in prospective studies.
ACKNOWLEDGEMENTS
We thank Zheng-Ping Zhu for his valuable support and constructive input during the preparation of this manuscript.
He H, Pan L, Liu F, Ren X, Cui Z, Pa L, Wang D, Zhao J, Wang H, Wang X, Du J, Peng X, Shan G. Linear and non-linear relationships between red blood cell indices and cardiovascular risk factors: findings from the China National Health Survey.BMC Public Health. 2024;24:3451.
[RCA] [PubMed] [DOI] [Full Text][Reference Citation Analysis (0)]
Oliveira BR, Magalhães EIDS, Bragança MLBM, Coelho CCNDS, Lima NP, Bettiol H, Barbieri MA, Cardoso VC, Santos AMD, Horta BL, Silva AAMD. Performance of Body Fat Percentage, Fat Mass Index and Body Mass Index for Detecting Cardiometabolic Outcomes in Brazilian Adults.Nutrients. 2023;15:2974.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 2][Cited by in RCA: 12][Article Influence: 4.0][Reference Citation Analysis (0)]
Takikawa A, Mahmood A, Nawaz A, Kado T, Okabe K, Yamamoto S, Aminuddin A, Senda S, Tsuneyama K, Ikutani M, Watanabe Y, Igarashi Y, Nagai Y, Takatsu K, Koizumi K, Imura J, Goda N, Sasahara M, Matsumoto M, Saeki K, Nakagawa T, Fujisaka S, Usui I, Tobe K. HIF-1α in Myeloid Cells Promotes Adipose Tissue Remodeling Toward Insulin Resistance.Diabetes. 2016;65:3649-3659.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 69][Cited by in RCA: 98][Article Influence: 9.8][Reference Citation Analysis (2)]
Villaret A, Galitzky J, Decaunes P, Estève D, Marques MA, Sengenès C, Chiotasso P, Tchkonia T, Lafontan M, Kirkland JL, Bouloumié A. Adipose tissue endothelial cells from obese human subjects: differences among depots in angiogenic, metabolic, and inflammatory gene expression and cellular senescence.Diabetes. 2010;59:2755-2763.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 175][Cited by in RCA: 197][Article Influence: 12.3][Reference Citation Analysis (0)]
Krishnan J, Danzer C, Simka T, Ukropec J, Walter KM, Kumpf S, Mirtschink P, Ukropcova B, Gasperikova D, Pedrazzini T, Krek W. Dietary obesity-associated Hif1α activation in adipocytes restricts fatty acid oxidation and energy expenditure via suppression of the Sirt2-NAD+ system.Genes Dev. 2012;26:259-270.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 219][Cited by in RCA: 254][Article Influence: 18.1][Reference Citation Analysis (0)]
