ABSTRACT
Background and Aim
Endogenous erythropoietin (EPO) deficiency has been associated with anemia in patients with type 2 diabetes mellitus (T2DM) and renal injury. The purpose of this study was to characterize EPO levels and anemia in Vietnamese patients with T2DM and chronic kidney disease (CKD).
Methods
A prospective observational study was conducted with 219 patients who were selected: 129 with T2DM and CKD, 51 with T2DM without CKD, and 49 healthy controls. Kruskal–Wallis, Chi‐square tests, and univariate ROC curve analysis were conducted to evaluate the predictive value of variables for anemia and reduced EPO levels.
Results
Median EPO levels were significant lower in patients with T2DM and CKD (35.83 IU/L), compared with patients without CKD and healthy controls (p < 0.001). The reduction in EPO was 41.9% in patients with CKD, compared to 17.6% in those without CKD (p < 0.01). Patients with CKD exhibited significant decreases in red blood cell count, hemoglobin, hematocrit, mean corpuscular volume, and mean corpuscular hemoglobin concentration compared with patients without CKD (p < 0.001). Moderate and severe anemia were observed in 31.8% and 4.6% of CKD patients, respectively, whereas only 5.9% of patients without CKD had moderate anemia (p < 0.001). EPO levels and rates of reduction did not differ significantly between types of kidney injury but varied markly across stages of renal failure (p < 0.001). A decreased glomerular filtration rate (GFR) was the strongest predictor of lower EPO levels (AUC = 0.661), p < 0.01), while both albumin and GFR were the strongest predictors of anemia (AUC = 0.787 and 0.786, p < 0.001).
Conclusion
Vietnamese patients with T2DM and CKD exhibited a 3.16‐fold decrease in EPO concentration compared to healthy people. GFR was significantly associated with reduced EPO levels, while both albumin and GFR were the most reliable predictors of anemia in these patients.
Keywords: anemia, chronic kidney disease, erythropoietin, kidney injury, type 2 diabetes
1. Introduction
Type 2 diabetes mellitus (T2DM) is a metabolic condition characterized by insulin resistance and decreased beta‐cell activity [1, 2, 3]. By 2024, diabetes is estimated to affect roughly 589 million people aged 20–79, with 399.6 million living in urban and 189.1 million in rural areas. T2DM is responsible for more than 90% of all diabetes diagnoses [4, 5]. Kidney injury is a common consequence of type 2 diabetes [6, 7]. The prevalence of kidney impairment in T2DM patients varies by country: 35%–54% in the United States, 36% in Europe, 29%–35% in China, 16% in Ghana, 15% in Malaysia, 33%–37% in Spain, and 23%–35% in Thailand [8]. Renal complications in T2DM present clinically in varied degrees, depending on the level of structural damage or functional impairment of the kidneys [9, 10, 11]. Microalbuminuria is an early indicator of renal impairment; if ignored, it can escalate to macroalbuminuria [12, 13]. Diabetics with or without proteinuria can have a decreased glomerular filtration rate (GFR), as do those with T2DM and chronic kidney disease (CKD) [14, 15].
Even in patients with T2DM who do not have renal complications, anemia can emerge [16]. Previous studies have shown that anemia in T2DM patients without renal issues has a complex etiology, including several variables [17, 18, 19, 20]. The mechanisms that cause anemia in persons with type 2 diabetes and renal complications are complex, with erythropoietin (EPO) deficiency, iron deficiency, and chronic inflammation all playing important roles [16, 17, 21]. Several studies have examined anemia and EPO levels in persons with T2DM and CKD [16, 22, 23, 24]. These studies consistently demonstrated that the prevalence of anemia in diabetic patients with CKD and a poor GFR ranged between 21.6% and 41.6%. Furthermore, patients with diabetic nephropathy had lower EPO levels, which were associated favorably with GFR. Low EPO levels were prevalent in diabetic patients, especially those with CKD [16, 22, 23, 24].
We hypothesized that plasma EPO levels drop with CKD progression and independently predict anemia in Vietnamese patients with T2DM. There has been no thorough study on anemia, EPO levels, and their relationship with anemia‐related indicators in Vietnamese patients with T2DM and CKD. The purpose of this study was to report variations in EPO levels, characterize anemia, and investigate relationships with important markers in Vietnamese type 2 diabetic patients with CKD. The purpose was to understand the mechanisms that underpin anemia in this patient population and to highlight disparities in anemia characteristics between Vietnamese patients and other ethnic groups globally.
2. Materials and Methods
2.1. Patients and Study Design
The cross‐sectional observational study was conducted from January 2018 to December 2019. We employed a convenient sampling method, enrolling all eligible patients in the study from 2018 to 2019. A total of 219 qualified participants, divided into three groups: Group 1 contained 129 patients with T2DM and CKD; Group 2 included 51 patients with T2DM but no CKD; and Group 3 consisted of 49 healthy people who had routine health exams with normal results and were matched by age and gender to the patient groups. Clinical examinations, laboratory tests, therapy, and follow‐up were performed on all T2DM, including those with and without CKD, at Viet Tiep Hospital in Hai Phong. Meanwhile, the randomly selected healthy controls completed laboratory testing, which included a single measurement of EPO levels. The study protocols were followed in accordance with the Declaration of Helsinki. All participants agreed and signed the consent form.
2.1.1. Patient Selection Criteria
-
–
Patients with T2DM and CKD: These persons were diagnosed with type 2 diabetes using the Vietnamese Ministry of Health's diagnostic criteria, as well as CKD according to the KDIGO 2017 guidelines [25]. They were over 18 years old, agreed to participate in the study, and were followed at Viet Tiep Hospital, Hai Phong. Patients had both a full clinical evaluation and paraclinical tests. Furthermore, they had not been treated for anemia or given recombinant EPO.
-
–
Type 2 diabetes patients without CKD: These patients were diagnosed with type 2 diabetes using the Vietnamese Ministry of Health's diagnostic criteria, were over the age of 18, had no chronic or acute kidney disease, were not being treated for anemia, did not take recombinant EPO, and were agreed to participate in the study and signed the consent form.
2.1.2. Patient Exclusion Criteria
The study excluded people who had gout, systemic lupus erythematosus, pre‐existing hypertension, multiple myeloma, or thalassemia. During the study period, patients with acute conditions such as myocardial infarction, acute heart failure, pneumonia, stroke, and surgical conditions were excluded, as were menstruating or breastfeeding women, those undergoing hemodialysis or peritoneal dialysis, and those with concomitant hyperactive syndromes.
2.1.3. Criteria for Selecting Healthy People
They were healthy people over the age of 18 who had no history of kidney, urinary, hepatobiliary, or cardiovascular disease. They had no infections or musculoskeletal diseases, were not currently menstruating, pregnant, or breastfeeding, and agreed to participate in the study.
2.2. Clinical and Laboratory Diagnostic Criteria
Body mass index (BMI) (kg/m2): Thin (< 18.5), normal (18.5–22.9), overweight, and obese (> 23) [26].
Diabetes (mmol/L): normal (fasting blood glucose < 7.0, blood glucose at any time < 11.1) or elevated (fasting blood glucose ≥ 7.0, blood glucose at any time > 11.1) [27].
Anemia (hemoglobin g/L): normal (male ≥ 129, female ≥ 119) or anemic (male < 129, female < 119) [28].
GFR and renal failure levels: stage 1 renal failure (glomerular filtration rate ≥ 90 mL/min), stage 2 renal failure (GFR 60–89 mL/min), stage 3 renal failure (GFR 30–59 mL/min), stage 4 renal failure (GFR 15–29 mL/min), and stage 5 renal failure (GFR < 15 mL/min) [25].
Urea index (mmol/L): normal (2.5–7.5), elevated (> 7.5) [29].
Creatinine index (µmol/l). Normal (50–110), elevated (> 110) [30].
Protein index (g/L): ≥ 60 (normal), < 60 (decreased) [31].
Hemoglobin index (g/L): Female ≥ 120 (indicates no anemia), female < 120 (indicates anemia). Male ≥ 130 (no anemia) and male < 130 (anemia) [32].
HbA1C index (%): < 5.7 (normal), 5.7–6.4 (high risk of type 2 diabetes), and ≥ 6.5 (diabetes) [33].
Albumin index (g/L): ≥ 35 (normal), < 35 (decreased) [33].
RBC (T/L) index: Female 3.8–5.0 (normal), female < 3.8 (decreased), female > 5.0 (increased). Male 4.2–6.0 (normal), male < 4.2 (decreased), male > 6.0 (increased) [34].
HGB index (g/L): Female 120–160 (normal), male 130–180 (normal) [34].
HCT index (L/L): Female 0.37–0.42 (normal), male 0.4–0.47 (normal) [34].
MCV index (fL): < 80 (microcytic anemia), 80–100 (normocytic anemia), > (100 macrocytic anemia) [34].
MCHC index (g/L): < 320 (hypochromic anemia), 320–360 (normochromic anemia), > 360 (hyperchromic anemia) [34].
2.3. Quantitative Determination of Plasma EPO
Plasma EPO concentrations were measured using a commercially available Human EPO ELISA Kit (Invitrogen, Vienna, Austria). Venous blood samples (2 mL) were collected in EDTA tubes, centrifuged, plasma separated, and stored at −80°C before analysis. The assay was carried out according to the manufacturer's instructions, and EPO concentrations were determined using a standard calibration curve. All samples were measured twice. The detection limit for EPO using the Human EPO ELISA Kit was 0.14 mIU/mL, and there was no cross‐reactivity.
2.4. Data Analysis
The data were analyzed using SPSS version 20.0 software. Data normality was assessed using the Shapiro–Wilk test. As most continuous variables were not normally distributed, they are presented as median (interquartile range). Comparisons between two independent groups were performed using the Mann–Whitney U test, while comparisons among more than two groups were conducted using the Kruskal–Wallis test. Categorical variables were expressed as percentages and compared using the Chi‐square test or Fisher's exact test when appropriate. Correlations between variables were assessed using Spearman's rank correlation coefficient. Receiver Operating Characteristic (ROC) curves were constructed using individual variables to evaluate their ability to discriminate anemia. A p‐value < 0.05 was considered statistically significant. The simple (univariate) correlation coefficient (r) between two variables was calculated and interpreted as follows: |r | ≥ 0.7 (very strong correlation), 0.5 ≤ |r | < 0.7 (fairly strong correlation), 0.3 ≤ |r | < 0.5 (moderate correlation), and |r | < 0.3 (weak correlation). A positive r indicates a positive correlation, while a negative r indicates a negative correlation.
2.5. Ethics Statement
The study was conducted solely for the purpose of diagnosing and treating patients and was approved by the Military Medical Academy ethics council prior to implementation (No. 581/HVQY, December 25, 2017). All participants were fully informed about the purpose of blood sampling and did not incur any research‐related costs, participants read and signed the consent form to participate in the study. The information of all participants was kept confidential and will not be disclosed to any organization or individual without their consent. Personal information and medical records were encrypted, and medical documents were securely stored in accordance with the hospital's regulations.
3. Results
The study comprised 219 participants: 129 with T2DM and CKD (group 1), 51 with T2DM without CKD (group 2), and 49 healthy controls (group 3). The male‐to‐female ratios were 0.87 (group 1), 0.89 (group 2), and 1.13 (group 3) (Table 1). Among patients above the age of 60, 77.5% were in group 1, and 62.7% were in group 2. Group 1 and 2 had approximately comparable proportion of participants with a BMI > 23 was (46.5% and 47.1%, respectively). The prevalence of diabetes for > 10 years was 59.7% in group 1, compared to 37.3% in group 2. In terms kidney injury, group 1 had 69.0% renal failure, 20.2% MAU+, and 10.8% MAC+. The proportion of patients with stage 4 and stage 5 renal failure was 27.1% and 14.0%, respectively. The patients with renal failure had an average GFR of 41.6 mL/min/1.73 m2 (Table 1).
Table 1.
Characteristics of type 2 diabetic patients with chronic kidney disease and healthy people in the study.
| Characteristics | Patients with CKD (n = 129) (n, %) | Patients without CKD (n = 51) (n, %) | Healthy people (n = 49) (n, %) |
|---|---|---|---|
| Gender | |||
| Male | 60 (46.5) | 24 (47.1) | 26 (53.1) |
| Female | 69 (53.5) | 27 (52.9) | 23 (46.9) |
| Age (year) | |||
| < 50 | 6 (4.7) | 8 (15.7) | 4 (8.2) |
| 50–59 | 23 (17.8) | 11 (21.6) | 12 (24.5) |
| 60–69 | 48 (37.2) | 22 (43.1) | 17 (34.7) |
| ≥ 70 | 52 (40.3) | 10 (19.6) | 16 (32.6) |
| Mean ± SD | 67.06 ± 9.46 | 61.96 ± 10.51 | 66.44 ± 12.65 |
| BMI | |||
| < 18.5 | 12 (9.3) | 5 (9.8) | 0 (0.0) |
| 18.5–22.9 | 57 (44.2) | 22 (43.1) | 29 (74.4) |
| ≥ 23 | 60 (46.5) | 24 (47.1) | 10 (25.6) |
| Mean ± SD | 22.50 ± 2.95 | 22.85 ± 3.43 | 22.32 ± 1.88 |
| Diabetes duration (year) | |||
| < 1 | 6 (4.6) | 5 (9.8) | — |
| 1 –< 5 | 16 (12.4) | 12 (23.5) | — |
| 5– < 10 | 30 (23.3) | 15 (29.4) | — |
| ≥ 10 | 77 (59.7) | 19 (37.3) | — |
| Median (quartile) | 10.0 (6.5–17.0) | 7.0 (3.0–10.0) | — |
| Types of kidney injury | |||
| MAU (+) | 26 (20.2) | — | — |
| MAC (+) | 14 (10.8) | — | — |
| Kidney failure | 89 (69.0) | — | |
| CKD stages | |||
| 1 | 8 (6.2) | — | — |
| 2 | 32 (24.8) | — | — |
| 3 | 36 (27.9) | — | — |
| 4 | 35 (27.1) | — | — |
| 5 | 18 (14.0) | — | — |
| GFR | |||
| GFR (mL/min/1.73 m2) | 41.6 | — | — |
| Median (quartile) | (22.4–65.3) | ||
| Min–Max (mL/min/1.73 m2) | 6.6–117.2 | — | — |
Plasma EPO concentrations varied significantly between the three groups, with group 1 exhibiting the lowest amount (35.83 IU/L), followed by group 2 (44.72 IU/L), and group 3 having the highest (113.31 IU/L). EPO status also differed across groups 1 and 2, with group 1 reporting a 41.9% decline in EPO concentration against a 17.6% decrease in group 2 (Table 2). In groups 1 and 2, patients had lower RBC counts (58.1% vs. 23.5%), lower HGB levels (65.1% vs. 21.6%), and lower HCT values (69.0% vs. 43.1%). However, MCV and MCHC levels did not differ significantly between the two groups. In terms of anemia severity, group 1 exhibited moderate and severe HGB reduction rates of 31.8% and 4.6%, respectively, whereas group 2 had only 5.9% of patients with moderate HGB reduction and none with severe reduction (Table 2).
Table 2.
Plasma EPO concentrations and red blood cell indices in type 2 diabetic patients in and healthy controls.
| Characteristics | Patients with CKD (n = 129) | Patients without CKD (n = 51) | Healthy people (n = 49) | p |
|---|---|---|---|---|
| EPO concentration | ||||
| Median (IU/L) | 35.83 | 44.72 | 113.31 | < 0.001a |
| Quartile (IU/L) | (21.55–56.34) | (37.25–63.90) | (30.5–378.42) | |
| Mean ± SD | 47.35 ± 43.03 | 60.50 ± 58.03 | 218.01 ± 208.96 | |
| Min | 6.22 | 15.47 | 15.95 | |
| Max | 261.33 | 387.27 | 612.75 | |
| EPO status | ||||
| Normal | 75 (58.1) | 41 (80.4) | — | < 0.01b |
| Median (IU/L), (quartile) | 49.31 (39.92–74.78) | 47.41 (40.09–65.53) | 113.31 (30.55–378.42) | |
| Decrease | 54 (41.9) | 9 (17.6) | — | |
| Median (IU/L), (quartile) | 20.03 (15.90–24.63) | 26.83 (21.54–30.17) | — | |
| Increase | 0 | 1 (2.0) | — | |
| Median (IU/L), (quartile) | — | 387.27‐ | — | |
| Red blood cell indices | ||||
| Decrease RBC (T/L) | 75 (58.1) | 12 (23.5) | −4.62 ± 0.63 | < 0.001c |
| (X ± SD) | 3.88 ± 0.68 | 4.47 ± 0.53 | ||
| Decrease HGB (g/L) | 84 (65.1) | 11 (21.6) | −137.59 ± 11.76 | < 0.001c |
| (X ± SD) | 115.79 ± 22.45 | 133.55 ± 13.16 | ||
| Decrease HCT (L/L) | 89 (69.0) | 22 (43.1) | — | < 0.001c |
| (X ± SD) | 0.34 ± 0.06 | 0.39 ± 0.04 | ||
| MCV (fL) (X ± SD) | 88.98 ± 5.70 | 88.83 ± 9.04 | 90.54 ± 5.56 | > 0.21c |
| MCHC (g/L) (X ± SD) | 334.64 ± 10.45 | 338.78 ± 11.85 | 330.45 ± 9.11 | < 0.05c |
| Anemia levels | ||||
| No (n, %) | 45 (34.9) | 40 (78.4) | — | > 0.15b |
| HGB(g/L) (X ± SD) | 157.31 ± 5.24 | 162.47 ± 4.36 | ||
| Mild (n, %) | 37 (28.7) | 8 (15.7) | — | |
| HGB(g/L) (X ± SD) | 116.29 ± 4.59 | 119.12 ± 7.79 | ||
| Moderate | 41 (31.8) | 3 (5.9) | — | |
| HGB (g/L) (X ± SD) | 95.41 ± 8.93 | 105.0 ± 2.64 | ||
| Severe | 6 (4.6) | 0 (0.0)‐ | — | |
| HGB (g/L) (X ± SD) | 71.33 ± 6.25 |
Kruskal – Wallis test.
Chi – square test.
Mann ‐ Whitney test, CKD, chronic kidney failure.
There were no variations in EPO concentration or reduction between type 2 diabetics with MAU+, MAC+, and chronic kidney failure. Red blood cell indices (RBC, HGB, and HCT) differed considerably between the patient groups (p < 0.001). MCV and MCHC did not exhibit any significant differences (p > 0.05). Anemia prevalence varied substantially amongst patients with chronic kidney failure (76.4%), MAC+ (57.1%), and MAU+ (30.8%) (p < 0.001) (Table 3).
Table 3.
Association between EPO concentration and red blood cell indices with characteristics of renal injury in patients with type 2 diabetes.
| Characteristics | MAU + (n = 26) | MAC + (n = 14) | CKD (n = 89) | p |
|---|---|---|---|---|
| EPO concentration (IU/L) | ||||
| Median, (quartile) | 37.21 (26.97–77.22) | 37.35 (21.71–46.14) | 32.53 (21.0–49.56) | > 0.17a |
| EPO status | ||||
| Normal | 9 (34.6) | 5 (35.7) | 40 (44.9) | > 0.09b |
| Decrease | 17 (65.4) | 9 (64.3) | 49 (55.1) | |
| Red blood cell indices | ||||
| RBC (T/l), (X ± SD) | 4.33 ± 0.59 | 3.96 ± 0.43 | 3.74 ± 0.68 | < 0.001a |
| HGB (g/L), (X ± SD) | 131.73 ± 21.70 | 121.07 ± 17.58 | 110.31 ± 21.07 | < 0.001a |
| HCT (L/L), (X ± SD) | 0.38 ± 0.06 | 0.36 ± 0.04 | 0.33 ± 0.06 | < 0.001a |
| MCV (fL), (X ± SD) | 89.23 ± 6.03 | 91.0 ± 7.5 | 88.58 ± 5.27 | > 0.12a |
| MCHC (g/L), (X ± SD) | 338.65 ± 9.19 | 334.57 ± 10.95 | 333.48 ± 10.54 | > 0.26a |
| Anemia status | ||||
| No anemia | 18 (69.2) | 6 (42.7) | 21 (23.6) | < 0.001a |
| Yes | 8 (30.8) | 8 (57.1) | 68 (76.4) | < 0.001a |
Kruskal–Wallis test.
Chi–square test.
EPO concentrations were substantially different (p < 0.001) among type 2 diabetes patients with renal failure stages 1 + 2 (37.35 IU/L), stages 3 + 4 (42.09 IU/L), and stage 5 (22.57 U/L). Patients with stage 5 renal failure had the highest proportion of EPO reduction (83.3%), followed by stages 1 + 2 (35.0%) and stages 3 + 4 (32.5%). Red blood cell parameters change amongst patients at different stages of renal failure, with the exception of MCV, which revealed no significant variation (p > 0.05). Anemia was detected in 88.9% of type 2 diabetic patients with stage 5 renal failure, compared to 73.2% in stages 3 + 4% and 40% in stages 1 + 2 (Table 4).
Table 4.
Association between EPO concentration and red blood cell indices with chronic kidney disease stages in type 2 diabetes patients.
| Characteristics | Stages of chronic kidney failure | p | ||
|---|---|---|---|---|
| Stages 1 + 2 (n = 40) | Stages 3 + 4 (n = 71) | Stage 5 (n = 18) | ||
| EPO concentration (IU/L) | ||||
| Median, (quartile) | 37.35 (25.87–65.0) | 42.09 (22.98– 57.88) | 22.57 (15.39– 27.03) | < 0.001 a |
| EPO status | ||||
| Normal | 26 (65.0) | 46 (64.8) | 3 (16.7) | < 0.01 b |
| Decrease | 14 (35.0) | 25 (32.5) | 15 (83.3) | |
| Red blood cell indices | ||||
| HC (T/l), (X ± SD) | 4.20 ± 0.56 | 3.82 ± 0.67 | 3.43 ± 0.68 | < 0.001 a |
| HST (g/L), (X ± SD) | 128.0 ± 20.77 | 113.09 ± 20.7 | 99.33 ± 19.34 | < 0.001 a |
| HCT (L/L), (X ± SD) | 0.37 ± 0.05 | 0.33 ± 0.06 | 0.30 ± 0.05 | < 0.001 a |
| MCV (fL), (X ± SD) | 89.85 ± 6.54 | 88.71 ± 4.8 | 88.1 ± 7.0 | > 0.14 a |
| MCHC (g/L), (X ± SD) | 337.22 ± 9.9 | 334.45 ± 9.97 | 329.66 ± 11.09 | < 0.001 a |
| Anemia status | ||||
| No anemia | 24 (60.0) | 19 (26.8) | 2 (11.1) | < 0.001b |
| Yes | 16 (40.0) | 52 (73.2) | 16 (88.9) | < 0.001 b |
Kruskal–Wallis test.
Chi‐square test.
In patients with T2DM and CKD, several factors predict anemia, including blood GFR and albumin, which show a high multivariate linear connection with peripheral blood hemoglobin (r = 0.649). The ROC curve analysis showed that albumin and GFR were the best predictors, with p = 0.000, area under the curve (AUC) values of 0.828 for albumin and 0.776 for GFR, and cut‐off values of 37.5 g/L for albumin and 36.5 mL/min/1.73 m2 for GFR (Figure 1).
Figure 1.
ROC curve predicting factors associated with anemia in type 2 diabetes patients with chronic kidney disease.
Several variables, such as creatinine, HbA1c, and GFR, predict lower plasma EPO levels in patients with T2DM and CKD (p < 0.05). GFR had the highest predictive value, with p < 0.01, AUC of 0.661, and a cutoff value of 29.25 ml/min/1.73 m2 (Figure 2).
Figure 2.
ROC curve predicting factors associated with reduced EPO in patients with type 2 diabetes and chronic kidney disease.
In multivariate logistic regression analysis, both GFR and albumin were independently and negatively associated with anemia. Specifically, lower eGFR (OR = 0.968, 95% CI: 0.949–0.988, p = 0.002) and lower albumin levels (OR = 0.702, 95% CI: 0.599–0.824, p < 0.001) were significantly associated with increased odds of anemia. Interestingly, HbA1c was also inversely associated with anemia (OR = 0.804, 95% CI: 0.663–0.976, p = 0.027) (Table 5).
Table 5.
Military multivariate logistic regression analysis of factors associated with anemia among patients in the study.
| Variable | β coefficient | OR | 95% CI | p value |
|---|---|---|---|---|
| GFR | −0.032 | 0.968 | 0.949–0.988 | 0.002 |
| Albumin | −0.354 | 0.702 | 0.599–0.824 | < 0.001 |
| HbA1c | −0.218 | 0.804 | 0.663–0.976 | 0.027 |
The scatter plot reveals a positive association between EPO and GFR. As GFR increases, so does EPO concentration; however, this relationship is weak and spread out. The data points were spread out and not firmly grouped around the regression line, showing a weak association. This is consistent with prior Spearman analysis findings (r = 0.295). The R 2 value of 0.059 indicates that GFR only accounts for 5.9% of the variation in EPO concentration, showing that other factors have a significant role. The plot contains outliers with high EPO levels, potentially affecting correlation strength and dispersion (Figure 3).
Figure 3.
Scatter plot between GFR and EPO concentration of patients in the study.
4. Discussion
4.1. Plasma EPO Concentration in Patients With Type 2 Diabetes and CKD
Patients with T2DM and CKD exhibited considerably reduced plasma EPO concentrations. Renal tubular damage and hypoxia‐inducible factor dysfunction, both of which reduce EPO production, are possible reasons of these patients' low EPO levels. Fujita et al. in Japan found that around 73.0% of diabetic patients had relatively low EPO levels [35], Mojiminiy in Kuwait had 49.7% [36], and Lin et al. in China had 68.0% [37]. Furthermore, EPO deficiency is thought to contribute to unexplained anemia in diabetic patients, even in the absence of renal injury. Hemoglobin and iron status are the two most important factors determining EPO levels in T2DM patients. In contrast, receptor and transport protein inhibitors are frequently involved in the pathways that contribute to reduced EPO levels in patients with T2DM and CKD [38]. Previous studies have shown that EPO levels gradually fall with each stage of renal failure, reaching their lowest value at stage 5 [35, 39]. This drop occurs because mildly injured kidneys can still create EPO to stimulate bone marrow erythropoiesis; but, significant kidney impairment (stages 4–5 renal failure) cause EPO production to cease, resulting in overt anemia. This explains why patients with T2DM without CKD frequently have higher EPO levels than those with CKD, and how low plasma EPO levels are related with anemia in this patient population.
4.2. Anemia Features in Patients With Type 2 Diabetes and CKD
The study found that patients with T2DM and CKD presented lower mean HC, HST, HCT, and MCHC levels in their peripheral blood than those without CKD. The frequency of anemia was 65%, with only 7.1% having severe anemia (HGB < 80 g/L). Previous studies have also found that the rate of anemia in patients with T2DM increased proportionally with the severity of renal failure. Diabetes is considered as an indipendent factor affecting hemoglobin saturation levels, hence anemia can rise in diabetic patients even if no renal impairment has occurred. Diabetes is associated to anemia through a variety of processes, including decreased EPO production, altered hypoxia‐sensing mechanisms, chronic systemic inflammation, iron deficiency, and increased urine excretion of transferrin and EPO. Anemia is more common in patients with T2DM and CKD, as serum hemoglobin levels correlate linearly with the GFR. Anemia in T2DM and CKD patients can be caused by a lack of EPO, iron deficiency, or both acute and chronic inflammation. The anemia reported in these patients is often normochromic with normal red blood cell volume (88% of patients in this study), whereas hypochromic anemia (related to iron deficiency) is rare. Therefore, if anemia in T2DM and CKD patients is not diagnosed and treated according to its pathophysiology, it might proceed to hypochromic anemia, which is predominantly caused byiron shortage. This emphasizes the necessity of evaluating iron status in patients with T2DM and CKD.
4.3. The Association Between Anemia, EPO Concentrations, and Kidney Injury in Patients With Type 2 Diabetes
Anemia is frequently related with the clinical symptoms of renal injury in patients with T2DM and CKD. Specifically, patients with chronic renal failure had lower hemoglobin levels than those with MAU+ or MAC+. In contrast, the HCT, MCV, and MCHC indices show no significantly difference between these patient groups. Furthermore, the severity of anemia is closely related to the course of CKD. Diabetic nephropathy is a complex pathologic disorder defined by a wide range of clinical signs. Anemia in diabetic patients is caused by a number of underlying disease processes and is affected by numerous diabetes treatments. Diabetic patients may develop anemia prior to renal failure, even if their GFR is higher than 90 mL/min/1.73 m2. Diabetic patients with kidney injury should be treated with angiotensin‐converting enzyme (ACE) inhibitors or angiotensin II type 1 (AT1) receptor blockers to manage blood pressure, minimize proteinuria, and safeguard renal and cardiovascular function. This syndrome contributes to anemia in diabetic patients with kidney injury, whether they are in the early or late stages of the disease.
EPO concentrations had no significant relationship with the clinical signs of kidney injury such as chronic kidney failure, MAC+, and MAU+. Although the incidence of anemia and decrease in EPO levels were directly related to the severity of renal injury, this relationship was not statistically significant (p > 0.05). This lack of significance could be attributed to the absence of MAU+ and MAC+ patients in the sample, as well as the unusual distribution of EPO values. In constrast, the findings revealed a significant and relevant correlation between GFR, EPO, and hemoglobin in the etiology of renal disease, as well as a correlation between EPO concentration and CKDstage. The study found that a drop in GFR was associated with a decrease in EPO levels; however, why EPO levels did not increase in response to anemia is unknown. Thus, in patients with T2DM and CKD, EPO levels are reduced due to both diabetes and kidney injury. Therefore, EPO therapy may be considered as a potential therapeutic option for anemia in this patient population, in accordance with current clinical guidelines. However, additional studies are required to determine its effect in delaying the progression of renal disease.
4.4. Factors Associated With Anemia in Patients With Type 2 Diabetes and CKD
Data analysis revealed that patients with T2DM and CKD, and anemia exhibited lower levels of protein, albumin, HbA1c, GFR, hemoglobin, and EPO than those without anemia. Multivariate analysis and ROC curves revealed that these indicators are independent risk factors for anemia. Albumin and GFR were the best predictors of anemia among diabetes patients who also had CKD. Our data revealed a substantial correlation between the severity of hypoalbuminemia and kidney injury, regardless of clinical or histopathological variables. Patients with low albumin levels had a 7.37‐fold higher chance of having end‐stage renal disease, demonstrating that albumin is an important prognostic factor for type 2 diabetes patients. Zhang et al. found that low serum albumin levels are prevalent in patients with end‐stage kidney disease and are linked with anemia. This conclusion is consistent with our results, where albumin was substantially related with hemoglobin levels and had moderate predictive value for anemia [40]. People with hypoalbuminemia are more prone to anemia, which can promote hypoxia and accelerate kidney injury in patients with T2DM, regardless of prior renal impairment. Adane et al. found that people with hypoalbuminemia are more vulnerable to anemia, which can promote hypoxia and accelerate kidney damage in T2DM patients, regardless of existing renal impairment. Furthermore, previous research have revealed the prognostic importance of GFR in anemia among T2DM patients [41]. The current study demonstrates the significance of blood albumin levels and renal function in assessing and predicting anemia in patients with T2DM and CKD. This tendency is consistent with data from studies conducted in Europe, the United States, and China, where anemia rates rise dramatically as kidney function deteriorate. Our findings suggest that routine hemoglobin, serum albumin, and renal function (GFR) testing should be explored in individuals with T2DM, particularly those at risk of CKD. Although EPO measurements are not regularly advised in clinical practice, they may be useful in certain situations of unexplained anemia. Furthermore, early detection of anemia may provide timely treatment interventions, such as iron supplements and erythropoiesis‐stimulating drugs, to enhance patient outcomes.
4.5. Factors Associated With Reduced EPO Levels in Patients With Type 2 Diabetes and CKD
Our findings showed that patients with low EPO levels exhibited significantly low GFR, HbA1c, and hemoglobin levels than those without low EPO. Multivariate and ROC curve analysis revealed that GFR is the most reliable predictor of decreasing plasma EPO. In patients with T2DM, a GFR < 30 mL/min/1.73 m2 is particularly problematic as it correlates with a significant reduction in EPO levels. In such circumstances, recombinant EPO is required for replacement therapy. Persistently low EPO levels can lead to significant impairment of kidney function, lowering patients' quality of life and lifespan. Previous study indicates that while EPO insufficiency is common in early‐stage CKD, patients with GFR values > 30 mL/min/1.73 m2 can still respond physiologically to anemia. Notably, Fujita's study highlighted EPO's critical role in maintaining kidney function and found that EPO insufficiency frequently precedes the start of CKD in diabetic patients [35]. Our data show that low EPO levels are associated with a rapid fall in GFR, particularly among patients with iron deficiency. This show a bidirectional correlation in which deteriorating kidney function reduced EPO production, which exacerbates renal function decline.
4.6. Study Limitations
This study has several limitations. First, its cross‐sectional design prevents causal inferences between EPO levels, anemia, and kidney function. Second, the study was conducted in a single center with a small sample size, which may limit the findings' generalizability. Third, critical factors impacting anemia, such as iron status and inflammatory indicators, were not thoroughly examined. Finally, any confounding variables may not have been fully accounted for in the multivariate analysis.
5. Conclusion
EPO concentrations in Vietnamese patients with T2DM and CKD were found to be considerably lower than those in patients without CKD and healthy control. Concurrently, patients with CKD had a significantly higher frequency of anemia than other groups, with anemia prevalence positively related to both the severity of CKD stages and the specific forms of renal injury. Furthermore, albumin levels and GFR proved to be reliable predictors of anemia in T2DM and CKD patients. A decrease in plasma EPO concentrations was directly associated with the progression of renal failure, although there was no significant relationship seen between EPO levels and the type of kidney injury. Additionally, GFR was strongly related with lower EPO levels in this patient population.
Author Contributions
Nguyen Ngoc Anh: conceptualization, resources, software, methodology, validation. Đo Thi Thanh Loan: resources, software. Le Viet Thang: conceptualization, investigation, resources, methodology. Nguyen Thi Thuy: conceptualization, investigation, funding acquisition, writing – original draft, methodology, validation, visualization, writing – review and editing, software, formal analysis, project administration, data curation, supervision, resources.
Funding
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
Transparency Statement
All of the authors affirm that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.
Acknowledgments
We thank the cementless total hip arthroplasty patients who agreed to participate in this study. We would also like to thank the Military Hospital 103, Hai Phong University Medical Center, Hai Phong University of Medicine and Pharmacy for their assistance in performing this study.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
The dataset analyzed for this study is available from the corresponding author on reasonable request.
References
- 1. Lu X., Xie Q., Pan X., et al., “Type 2 Diabetes Mellitus in Adults: Pathogenesis, Prevention and Therapy,” Signal Transduction and Targeted Therapy 9, no. 1 (2024): 262, 10.1038/s41392-024-01951-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Mlynarska E., Czarnik W., Dzieza N., et al., “Type 2 Diabetes Mellitus: New Pathogenetic Mechanisms, Treatment and the Most Important Complications,” International Journal of Molecular Sciences 26, no. 3 (2025): 1094, 10.3390/ijms26031094. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Ferrannini E. and Mari A., “β‐Cell Function in Type 2 Diabetes,” Metabolism: Clinical and Experimental 63, no. 10 (2014): 1217–1227, 10.1016/j.metabol.2014.05.012. [DOI] [PubMed] [Google Scholar]
- 4. Duncan B. B., Magliano D. J., and Boyko E. J., “IDF Diabetes Atlas 11th Edition 2025: Global Prevalence and Projections for 2050,” Nephrology Dialysis Transplantation 41, no. 1 (2025): 7–9, 10.1093/ndt/gfaf177. [DOI] [PubMed] [Google Scholar]
- 5. Genitsaridi I., Salpea P., Salim A., et al., “11th Edition of the IDF Diabetes Atlas: Global, Regional, and National Diabetes Prevalence Estimates for 2024 and Projections for 2050,” Lancet Diabetes and Endocrinology 14 (2025): 149–156, 10.1016/S2213-8587(25)00299-2. [DOI] [PubMed] [Google Scholar]
- 6. Jiang S., Zhang D., and Li W., “The Chronic Kidney Disease Epidemiology Collaboration Equations Perform Less Well in an Older Population With Type 2 Diabetes Than Their Non‐Diabetic Counterparts,” Frontiers in Public Health 10 (2022): 952899, 10.3389/fpubh.2022.952899. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Jha R., Lopez‐Trevino S., Kankanamalage H. R., and Jha J. C., “Diabetes and Renal Complications: An Overview on Pathophysiology, Biomarkers and Therapeutic Interventions,” Biomedicines 12, no. 5 (2024): 1098, 10.3390/biomedicines12051098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Fenta E. T., Eshetu H. B., Kebede N., et al., “Prevalence and Predictors of Chronic Kidney Disease Among Type 2 Diabetic Patients Worldwide, Systematic Review and Meta‐Analysis,” Diabetology and Metabolic Syndrome 15, no. 1 (2023): 245, 10.1186/s13098-023-01202-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Gao Z., Zuo M., Han F., et al., “Renal Impairment Markers in Type 2 Diabetes Patients With Different Types of Hyperuricemia,” Journal of Diabetes Investigation 10, no. 1 (2019): 118–123, 10.1111/jdi.12850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Di Vincenzo A., Bettini S., Russo L., Mazzocut S., Mauer M., and Fioretto P., “Renal Structure in Type 2 Diabetes: Facts and Misconceptions,” Journal of Nephrology 33, no. 5 (2020): 901–907, 10.1007/s40620-020-00797-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Pugliese G., Penno G., Natali A., et al., “Diabetic Kidney Disease: New Clinical and Therapeutic Issues. Joint Position Statement of the Italian Diabetes Society and the Italian Society of Nephrology on “The Natural History of Diabetic Kidney Disease and Treatment of Hyperglycemia in Patients With Type 2 Diabetes and Impaired Renal Function,” Journal of Nephrology 33, no. 1 (2020): 9–35, 10.1007/s40620-019-00650-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Reijrink M., Sluiter J. K. E., Te Velde‐Keyzer C. A., et al., “Severely Increased Albuminuria in Patients With Type 2 Diabetes Mellitus Is Associated With Increased Subclinical Atherosclerosis in Femoral Arteries With Na [(18)F]F Activity as a Proxy ‐ The DETERMINE Study,” Atherosclerosis 394 (2024): 117199, 10.1016/j.atherosclerosis.2023.117199. [DOI] [PubMed] [Google Scholar]
- 13. Aslan O., Demir M., and Koseoglu M., “Kidney Injury Molecule Levels in Type 2 Diabetes Mellitus,” Journal of Clinical Laboratory Analysis 30, no. 6 (2016): 1031–1036, 10.1002/jcla.21976. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Akazawa S., Sadashima E., Sera Y., and Koga N., “Decline in the Estimated Glomerular Filtration Rate (eGFR) Following Metabolic Control and Its Relationship With Baseline eGFR in Type 2 Diabetes With Microalbuminuria or Macroalbuminuria,” Diabetology International 13, no. 1 (2022): 148–159, 10.1007/s13340-021-00517-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Bakris G. L., “Recognition, Pathogenesis, and Treatment of Different Stages of Nephropathy in Patients With Type 2 Diabetes Mellitus,” Mayo Clinic Proceedings 86, no. 5 (2011): 444–456, 10.4065/mcp.2010.0713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Mehdi U. and Toto R. D., “Anemia, Diabetes, and Chronic Kidney Disease,” Diabetes Care 32, no. 7 (2009): 1320–1326, 10.2337/dc08-0779. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Antoniadou C., Gavriilidis E., Ritis K., and Tsilingiris D., “Anemia in Diabetes Mellitus: Pathogenetic Aspects and the Value of Early Erythropoietin Therapy,” Metabolism Open 25 (2025): 100344, 10.1016/j.metop.2024.100344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Chuang T. L., Koo M., and Wang Y. F., “The Impact of Diabetes, Anemia, and Renal Function in the Relationship Between Osteoporosis and Fasting Blood Glucose Among Taiwanese Women: A Cross‐Sectional Study,” BMC Women's Health 24, no. 1 (2024): 23, 10.1186/s12905-023-02851-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Barbieri J., Fontela P. C., Winkelmann E. R., et al., “Anemia in Patients With Type 2 Diabetes Mellitus,” Anemia 2015 (2015): 354737, 10.1155/2015/354737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. He B. B., Xu M., Wei L., et al., “Relationship Between Anemia and Chronic Complications in Chinese Patients With Type 2 Diabetes Mellitus,” Archives of Iranian Medicine 18, no. 5 (2015): 277–283. [PubMed] [Google Scholar]
- 21. Portoles J., Martin L., Broseta J. J., and Cases A., “Anemia in Chronic Kidney Disease: From Pathophysiology and Current Treatments, to Future Agents,” Frontiers in Medicine (Lausanne) 8 (2021): 642296, 10.3389/fmed.2021.642296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Fathi A., Shahwan M., Hassan N., Jairoun A., and Shahwan M., “Prevalence of Anemia in Type 2 Diabetic Patients and Correlation With Body Mass Index and Kidney Function in Palestine,” Diabetes, Metabolic Syndrome and Obesity 17 (2024): 2293–2301, 10.2147/DMSO.S454916. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Badura K., Janc J., Wąsik J., et al., “Anemia of Chronic Kidney Disease—A Narrative Review of Its Pathophysiology, Diagnosis, and Management,” Biomedicines 12, no. 6 (2024): 1191, 10.3390/biomedicines12061191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Al‐Khoury S., Afzali B., Shah N., Covic A., Thomas S., and Goldsmith D. J., “Anaemia in Diabetic Patients With Chronic Kidney Disease—Prevalence and Predictors,” Diabetologia 49, no. 6 (2006): 1183–1189, 10.1007/s00125-006-0254-z. [DOI] [PubMed] [Google Scholar]
- 25. KDIGO , “Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease,” Kidney International Supplements 3 (2017): 1–150. [Google Scholar]
- 26. Haam J. H., Kim B. T., Kim E. M., et al., “Diagnosis of Obesity: 2022 Update of Clinical Practice Guidelines for Obesity by the Korean Society for the Study of Obesity,” Journal of Obesity and Metabolic Syndrome 32 (2023): 121, 10.7570/jome23031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Nelson M. and Dungan K. M., “Diagnostic Tests for Diabetes Mellitus,” in Endotext, eds. Feingold K. R., Ahmed S. F., Anawalt B., Blackman M. R., Boyce A., Chrousos G., et al. (MDText.com, Inc., 2022). [Google Scholar]
- 28. World Health Organization , Haemoglobin Cutoffs to Define Anaemia in Individuals and Populations (World Health Organization, 2024). [PubMed] [Google Scholar]
- 29. Laville S. M., Couturier A., Lambert O., et al., “Urea Levels and Cardiovascular Disease in Patients With Chronic Kidney Disease,” Nephrology Dialysis Transplantation 38, no. 1 (2022): 184–192, 10.1093/ndt/gfac045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Ávila M., Mora Sánchez M. G., Bernal Amador A. S., and Paniagua R., “The Metabolism of Creatinine and Its Usefulness to Evaluate Kidney Function and Body Composition in Clinical Practice,” Biomolecules 15, no. 1 (2025): 41, 10.3390/biom15010041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Thakkinstian A., Tran H., Reeves G., Murch S., and Attia J., “A Clinical Decision Rule to Aid Ordering of Serum and Urine Protein Electrophoresis for Case‐Finding of Paraproteins in Hospitalized Inpatients,” Journal of General Internal Medicine 23, no. 10 (2008): 1688–1692, 10.1007/s11606-008-0712-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. WHO , Nutritional Anemias: Tools for Effective Prevention and Control. (World Health Organization, 2017). 1,4,17. [Google Scholar]
- 33. Health Mo ., Guidelines for the Diagnosis and Treatment of Type 2 Diabetes Mellitus (Medical Publishing House, 2020), 8–10. [Google Scholar]
- 34. Phan DT ., “Physiology ‐ Biochemistry of Blood.” Postgraduate Lectures on Hematology and Blood Transfusion (Medical Publishing House, 2021), 83–90. [Google Scholar]
- 35. Fujita Y., Doi Y., Hamano T., et al., “Low Erythropoietin Levels Predict Faster Renal Function Decline in Diabetic Patients With Anemia: A Prospective Cohort Study,” Scientific Reports 9, no. 1 (2019): 14871, 10.1038/s41598-019-51207-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Mojiminiyi O. A., Abdella N. A., Zaki M. Y., El Gebely S. A., Mohamedi H. M., and Aldhahi W. A., “Prevalence and Associations of Low Plasma Erythropoietin in Patients With Type 2 Diabetes Mellitus,” Diabetic Medicine 23, no. 8 (2006): 839–844, 10.1111/j.1464-5491.2006.01893.x. [DOI] [PubMed] [Google Scholar]
- 37. Li Y., Shi H., Wang W. M., et al., “Prevalence, Awareness, and Treatment of Anemia in Chinese Patients With Nondialysis Chronic Kidney Disease: First Multicenter, Cross‐Sectional Study,” Medicine 95, no. 24 (2016): e3872, 10.1097/MD.0000000000003872. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Marathias K. P., Lambadiari V. A., Markakis K. P., et al., “Competing Effects of Renin Angiotensin System Blockade and Sodium‐Glucose Cotransporter‐2 Inhibitors on Erythropoietin Secretion in Diabetes,” American Journal of Nephrology 51, no. 5 (2020): 349–356, 10.1159/000507272. [DOI] [PubMed] [Google Scholar]
- 39. Panjeta M., Tahirović I., Sofić E., Ćorić J., and Dervišević A., “Interpretation of Erythropoietin and Haemoglobin Levels in Patients With Various Stages of Chronic Kidney Disease,” Journal of Medical Biochemistry 36, no. 2 (2017): 145–152, 10.1515/jomb-2017-0014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Junlin Zhang R. Z., Wang Y., Li H., et al., “The Level of Serum Albumin Is Associated With Renal Prognosis in Patients With Diabetic Nephropathy,” Journal of Diabetes Research 2019 (2019): 7825804, 10.1155/2019/7825804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Adane T. and Getawa S., “Anaemia and Its Associated Factors Among Diabetes Mellitus Patients in Ethiopia: A Systematic Review and Meta‐Analysis,” Endocrinology, Diabetes and Metabolism 4, no. 3 (2021): e00260, 10.1002/edm2.260. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
The dataset analyzed for this study is available from the corresponding author on reasonable request.