Abstract
Objective
Anemia is reportedly associated with all-cause mortality in elderly individuals. Kidney dysfunction is a major cause of anemia in the elderly. Kidney dysfunction is characterized by glomerular and renal tubular damage (RTD). We aimed to examine the association between RTD and anemia, and the impact of anemia on all-cause mortality in the general population.
Methods
This prospective cohort study included 3,464 participants aged ≥40 years who participated in a community-based health checkup. The estimated glomerular filtration rate (eGFR), urinary albumin-to-creatinine ratio (UACR), and urinary β2-microglobulin to creatinine ratio (UBCR) were measured. The primary endpoint was all-cause mortality rate.
Results
The prevalence of anemia, but not albuminuria, increased with a reduced eGFR and advanced RTD (UBCR ≥300 μg/gCr). A multivariate logistic analysis demonstrated that UBCR was significantly associated with the presence of anemia in the general population, but not with eGFR or UACR. RTD anemia was defined as comorbid anemia and RTD. Anemia in the RTD group showed a lower eGFR and higher brain natriuretic peptide and heart-type fatty acid-binding protein levels than those without. During the median follow-up period of 6,794 days, 772 all-cause deaths occurred. Univariate and multivariate Cox proportional hazards regression analyses demonstrated that anemia of the RTD was independently associated with all-cause mortality in the general population. Anemia of the RTD significantly improved the C-statistic with a significant net reclassification index and integrated discrimination improvement.
Conclusion
RTD was closely associated with anemia. RTD anemia could be a novel risk factor for all-cause mortality in the general population.
Keywords: anemia, renal tubular damage, glomerular damage, general population, mortality
Introduction
Anemia, defined as hemoglobin (Hb) levels ＜13 g/dL in males and ＜12 g/dL in females, is a common public health problem worldwide (1). In developed countries, the prevalence of anemia increases with age. The prevalence of anemia begins to increase at 50 years of age and reaches approximately 20-25% in individuals aged 75 years (2,3). Anemia in elderly individuals is attracting increasing attention because it leads to low physical activity and a high all-cause mortality rate (4-6). Anemia is also associated with an increased risk of chronic kidney disease (CKD) (7).
CKD, generally defined as glomerular damage (GD) such as reduced estimated glomerular filtration rate (eGFR) and albuminuria, is also an important health problem associated with high all-cause mortality (8). CKD contributes to the development of anemia or anemia through insufficient erythropoietin production, uremic toxin-induced bone marrow suppression, iron deficiency, vitamin B12 and folate acid deficiency, reduced red blood cell survival, and inflammation (9). Consequently, the prevalence of anemia increases with advancing stages of CKD progression (10). Anemia of CKD is known to reduce the quality of life and increase mortality (9,11). As anemia and CKD interact, the simultaneous assessment and management of these conditions are required.
Renal tubular damage (RTD) is another type of renal dysfunction. Renal tubular cells exhibit diverse regulatory and endocrine functions. Renal tubules play multifunctional roles such as acid-base balance modulation, active vitamin D synthesis, and reabsorption of sodium, water, bicarbonate, and amino acids (12). Urinary β2 microglobulin, a low-molecular-weight protein produced by all cells expressing the major histocompatibility complex class I antigen, is readily filtered through the glomerulus and almost completely reabsorbed by the proximal tubules (13). Because an increase in the excretion of urinary β2 microglobulin results from impaired reabsorption in the proximal tubules, the urinary β2-microglobulin to creatinine ratio (UBCR) is a reliable marker of RTD (14). We previously reported that RTD, defined as the UBCR, is closely associated with anemia and subsequent poor clinical outcomes in patients with heart failure (15,16). A recent report indicated that erythropoietin is produced by renal erythropoietin-producing cells located around renal tubule cells (17). Whether RTD is associated with anemia in the general population remains unclear.
We aimed to examine the association between anemia and RTD, and the impact of anemia on all-cause mortality in the general population.
Materials and Methods
Ethics statement
This study was part of a Molecular Epidemiological study utilizing the Regional Characteristics of the 21st Century Centers of Excellence Program in Japan. The details of the study methodology have been previously described (18). The Ethical Review Committee of Yamagata University Faculty of Medicine approved the study (Updated IRB number, 2023-263), and all participants provided written informed consent. All the procedures were performed in accordance with the principles of the Declaration of Helsinki.
Study cohort
This analysis was part of a community-based health study of inhabitants of the town of Takahata in northern Japan (total population 26,026). We enrolled 3,509 participants aged ≥40 years (1,575 males and 1,934 females) who participated in a community-based health checkup from 2004 to 2005. Of these, 45 were excluded because they had not undergone urinalysis. Finally, we analyzed 3,464 subjects.
Follow up period and endpoint
All subjects were prospectively monitored during a median follow-up period of 6,260 (5,592-6,341) days. The primary endpoint was all-cause mortality rate. The secondary endpoint was a composite of kidney disease-related and cardiovascular and cerebrovascular deaths. The cause of death was determined by reviewing death certificates at the end of 2021. Death codes (International Classification of Diseases, 10th revision) were reviewed.
Biochemical markers
For all subjects, a blood sample for the measurement of the serum heart-type fatty acid-binding protein (H-FABP) level was drawn and centrifuged at 2,500× g for 15 min at 4°C within 30 min of collection, and the serum was stored at -70°C until analysis. H-FABP levels were measured using a two-step sandwich enzyme-linked immunosorbent assay kit (MARKIT-M H-FABP; Dainippon Pharmaceutical, Tokyo, Japan). The detection limit, measurement range, and reference interval for H-FABP were 0-250 ng/mL, 1.0-37, ＜6.2 ng/mL, respectively. Blood samples were also obtained to measure brain natriuretic peptide (BNP) levels. The sample was transferred to a chilled tube containing 4.5 mg ethylene diamine tetra acetic acid disodium salt and aprotinin (500 U/mL) and centrifuged at 1,000× g for 15 min at 4°C. The clarified plasma samples were frozen, stored at -70°C, and thawed immediately before the assay. The BNP levels were measured using a commercially available radioimmunoassay specific for human BNP (Shiono RIA BNP Assay Kit; Shionogi, Tokyo, Japan). The detection limit, measurement range, and reference interval of BNP were 0-2,902, 3.4-1,392, ＜18.4 pg/mL, respectively (19).
The total cholesterol levels were measured using the cholesterol oxidase enzyme method. Triglyceride levels were measured by an enzymatic method. The high-density lipoprotein cholesterol levels were measured using a chemically modified enzyme method. Fasting blood sugar levels were measured using an enzyme electrode method. glycosylated hemoglobin A1c level were measured using high performance liquid chromatography (HPLC).
Kidney function and anemia
Urine and venous blood samples were collected in the early morning. The urine β2-microglobulin levels were measured using the latex agglutination method (BML, Tokyo, Japan). RTD was defined as UBCR ≥300 μg/gCr, as previously reported (16,20). The urine albumin levels were quantitatively measured by immunoturbidimetry of a single-spot urine specimen collected early in the morning. The urine albumin level was corrected for the urine creatinine level to obtain the urine microalbumin/creatinine ratio (UACR). The glomerular filtration rate was estimated using the diet modification in the renal disease equation with the Japanese coefficient as previously reported (21). According to the K/DOQI clinical guidelines, reduced eGFR and albuminuria were defined as eGFR ＜60 mL/min/1.73 m2 and UACR ≥30 mg/gCr, respectively (8). GD was defined as a reduced eGFR and/or albuminuria. The Hb levels were measured simultaneously. Anemia was defined as Hb ＜13 g/dL in males and ＜12 g/dL in females, according to the World Health Organization guidelines (22).
Definition of cardiovascular risks
A history of smoking, cardiovascular disease, and cancer was collected using a health checkup questionnaire. Obesity was defined as a body mass index (BMI) ≥25 kg/m2. The other definitions of cardiovascular disease risk in this study were as follows: hypertension, systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or current use of antihypertensive medication; diabetes mellitus, fasting blood sugar ≥126 mg/dL, glycosylated hemoglobin A1c (National Glycohemoglobin Standardization Program) ≥6.5%, or current use of anti-diabetic medications; hyperlipidemia, total cholesterol ≥220 mg/dL, triglycerides ≥150 mg/dL, or current use of anti-dyslipidemia medication.
Statistical analysis
The sample size was calculated using the sample size formula of the proportional hazard regression model. All values are expressed as the mean ± standard deviation. Owing to the non-normal distributions of UACR, UBCR, BNP, and H-FABP, we utilized log-transformed UACR, UBCR, BNP, and H-FABP for all analyses. Continuous variables were analyzed using Student's t-test or the Mann-Whitney U test. Categorical data were analyzed using the chi-squared test. Associations between anemia and kidney dysfunction, such as reduced eGFR, albuminuria, and RTD, were assessed using the chi-squared test. Univariate and multivariate logistic analyses were performed to determine the predisposing factors for anemia. Survival curves were constructed using the Kaplan-Meier method and compared using the log-rank test. Cox proportional hazard analysis was used to determine the independent predictors of all-cause mortality and the composite of kidney disease-related, cardiovascular, and cerebrovascular deaths. Significant predictors in the univariate Cox proportional hazards regression analysis (age, sex, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, and smoking) were entered into multivariate Cox proportional hazards regression analysis. We calculated the receiver operating characteristic (ROC) curves, net reclassification index (NRI), and integrated discrimination index (IDI) to determine the quality of improvement in the corrected reclassification following the addition of anemia to the baseline model. Statistical significance was set at p＜0.05. Statistical analyses were performed using a standard statistical software package (JMP version 14; SAS Institute, Cary, USA; EZR, Saitama Medical Center, Jichi Medical University, Shimotsuke, Japan).
Results
Comparison of the clinical characteristics between the subjects with and without anemia.
A flowchart of the study selection process is shown in Fig. 1. The baseline characteristics of the subjects are presented in Table 1. The mean patient age was 63 years. A total of 400 (11.5%) subjects had anemia. Obesity, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, smoking, reduced eGFR, albuminuria, and RTD were identified in 858 (25%), 1,264 (36%), 240 (7%), 476 (14%), 50 (1.5%), 1,122 (32%), 237 (7%), 670 (19%), and 440 (13%) patients, respectively.
Figure 1.
Flowchart of the study selection process. GD: glomerular damage, RTD: renal tubular damage
Table 1.
Baseline Characteristics in All Subjects and Comparisons of Baseline Clinical Characteristics between Subjects with and without Anemia in the General Population.
| Variables | All subjects n=3,464 | Anemia (-) n=3,064 | Anemia (+) n=400 | p value |
|---|---|---|---|---|
| Age, years | 63±10 | 62±10 | 64±12 | 0.0009 |
| Men/women, n | 1,562/1,902 | 1,447/1,617 | 115/285 | <0.0001 |
| Obese, n (%) | 858 (25) | 800 (26) | 58 (15) | <0.0001 |
| BMI, kg/m2 | 23.6±4.4 | 23.7±4.5 | 22.5±3.1 | <0.0001 |
| Hypertension, n (%) | 1,264 (36) | 1,122 (37) | 142 (36) | 0.6615 |
| Systolic BP, mmHg | 134±16 | 135±16 | 131±17 | 0.0002 |
| Diastolic BP, mmHg | 80±10 | 80±10 | 76±10 | <0.0001 |
| Diabetes mellitus, n (%) | 240 (7) | 215 (7) | 25 (6) | 0.5647 |
| HbA1c, % | 5.7±0.7 | 5.7±0.7 | 5.7±0.7 | 0.7946 |
| Fasting blood sugar, mg/dL | 95±17 | 95±17 | 91±16 | <0.0001 |
| Dyslipidemia, n (%) | 476 (14) | 426 (14) | 50 (13) | 0.4380 |
| Atrial fibrillation, n (%) | 50 (1.5) | 47 (1.5) | 3 (0.8) | 0.1783 |
| Smoking, n (%) | 1,122 (32) | 1,038 (34) | 84 (21) | <0.0001 |
| Previous cardiovascular disease, n (%) | 460 (13) | 399 (13) | 61 (15) | 0.2248 |
| Previous cancer, n (%) | 74 (2) | 66 (2) | 8 (2) | 0.8412 |
| Biochemical data | ||||
| Blood urea nitrogen, mg/dL | 16.2±4.5 | 16.1±4.1 | 17.0±7.2 | 0.0003 |
| eGFR, mL/min/1.73m2 | 81±16 | 82±16 | 80±20 | 0.1295 |
| Reduced eGFR, n (%) | 237 (7) | 191 (6) | 46 (12) | 0.0003 |
| G1/G2/G3a/G3b-G5, n | 1,010/2,216/204/34 | 894/1,978/174/18 | 116/238/30/16 | <0.0001 |
| Log10-transformed UACR, mg/gCr | 1.08±0.54 | 1.07±0.54 | 1.13±0.55 | 0.0305 |
| Albuminuria, n (%) | 670 (19) | 578 (19) | 92 (23) | 0.0534 |
| A1/A2/A3, n | 2,794/602/68 | 2,486/520/58 | 308/82/10 | 0.1501 |
| Log10-transformed UBCR, μg/gCr | 2.05±0.42 | 2.04±0.42 | 2.14±0.49 | <0.0001 |
| RTD, n (%) | 440 (13) | 370 (12) | 70 (18) | 0.0033 |
| Log10-transformed BNP, pg/mL | 1.30±0.36 | 1.27±0.36 | 1.49±0.37 | <0.0001 |
| Loge-transformed H-FABP, ng/mL | 1.25±0.44 | 1.24±0.42 | 1.33±0.53 | 0.0002 |
| Total protein, g/dL | 7.5±0.4 | 7.5±0.4 | 7.3±0.4 | <0.0001 |
| Serum albumin, g/dL | 4.5±0.3 | 4.5±0.2 | 4.4±0.3 | <0.0001 |
| Total cholesterol, mg/dL | 201±32 | 202±32 | 188±33 | <0.0001 |
| Triglyceride, mg/dL | 107±64 | 110±65 | 87±49 | <0.0001 |
| HDL-C, mg/dL | 59±15 | 59±15 | 61±15 | 0.0029 |
| LDL-C, mg/dL | 124±30 | 126±30 | 112±28 | <0.0001 |
| Uric acid, mg/dL | 5.1±1.3 | 5.1±1.3 | 4.5±1.5 | <0.0001 |
Data are expressed as mean±SD, number (percentage).
BMI: body mass index, BNP: brain natriuretic peptide, BP: blood pressure, eGFR: estimated glomerular filtration rate, HbA1c: glycosylated hemoglobin A1c, HDL-C: high-density lipoprotein cholesterol, H-FABP: heart-type fatty acid binding protein, LDL-C: low-density lipoprotein cholesterol, RTD: renal tubular damage, UACR: urinary albumin-creatinine ratio, UBCR: urinary β2-microglobulin-creatinine ratio
Continuous variables were analyzed using Student’s t-test or the Mann-Whitney U test. Categorical data were analyzed using the chi-square test.
Participants with anemia were older, more likely to be female, had a higher prevalence of reduced eGFR and RTD, and a lower prevalence of obesity and smoking than those without anemia. Subjects with anemia had lower BMI, systolic and diastolic blood pressure, fasting blood sugar, total protein, serum albumin, total cholesterol, triglyceride, low-density lipoprotein cholesterol, and uric acid levels, and higher blood urea nitrogen, log10-transformed UACR, log10-transformed UBCR, log10-transformed BNP, and loge-transformed H-FABP levels than those without anemia.
Association of anemia with kidney dysfunction
The prevalence of anemia increased with decreasing eGFR but was not altered with increasing albuminuria (Fig. 2A, B). The prevalence of anemia was higher in subjects with RTD than in those without RTD (Fig. 2C). Additionally, it increased with increasing UBCR levels (Supplementary material 1).
Figure 2.
Association of anemia with reduced eGFR (A), albuminuria (B), and renal tubular damage (C). eGFR: estimated glomerular filtration rate, RTD: renal tubular damage, UACR: urinary albumin to creatinine ratio
Predisposing factors for anemia in the general population
To identify the predisposing factors for anemia, we performed univariate and multivariate logistic analyses for the presence of anemia. In the univariate logistic analysis, log10-transfomred UBCR was significantly associated with the presence of anemia (Table 2). Age, sex, obesity, smoking, and log10-transfomred UACR were also associated with the presence of anemia. A multivariate logistic analysis demonstrated that log10-transformed UBCR was significantly associated with the presence of anemia after adjusting for age, sex, obesity, hypertension, diabetes mellitus, dyslipidemia, smoking, eGFR, and log10-transformed UACR (Odds ratio, 1.244; 95% confidence interval, 1.099-1.412; p=0.0005; Table 2).
Table 2.
Univariate and Multivariate Logistic Analyses for Anemia in the General Population.
| Variables | Univariate analysis | Multivariate analysis | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p value | OR | 95% CI | p value | |
| Age, per 1-year increase | 1.017 | 1.007-1.028 | 0.0008 | 1.021 | 1.009-1.034 | 0.0007 |
| Sex (women vs. men) | 2.217 | 1.771-2.794 | <0.0001 | 2.299 | 1.722-3.095 | <0.0001 |
| Obesity | 0.483 | 0.359-0.642 | <0.0001 | 0.497 | 0.366-0.664 | <0.0001 |
| Hypertension | 0.952 | 0.765-1.182 | 0.6615 | 0.866 | 0.679-1.102 | 0.2435 |
| Diabetes mellitus | 0.901 | 0.556-1.388 | 0.6476 | 0.970 | 0.598-1.506 | 0.2435 |
| Dyslipidemia | 0.885 | 0.640-1.199 | 0.4380 | 0.767 | 0.551-1.051 | 0.1001 |
| Smoking | 1.927 | 1.505-2.493 | <0.0001 | 1.205 | 0.879-1.659 | 0.2478 |
| eGFR* | 0.921 | 0.828-1.024 | 0.1280 | 1.016 | 0.904-1.141 | 0.7818 |
| UACR* | 1.116 | 1.001-1.231 | 0.0329 | 0.907 | 0.791-1.036 | 0.1524 |
| UBCR* | 1.268 | 1.144-1.405 | <0.0001 | 1.244 | 1.099-1.412 | 0.0005 |
CI: confidence interval, eGFR: estimated glomerular filtration rate, OR: odds ratio, UACR: urinary albumin-creatinine ratio, UBCR: urinary β2-microglobulin-creatinine ratio
*Per 1-SD increase.
Univariate and multivariate logistic analyses were performed to determine the predisposing factors for anemia.
Anemia of RTD in the general population
Since we found a significant association between anemia and RTD, we divided the subjects with anemia into three groups based on kidney dysfunction: simple anemia group (normal kidney function, n=260), anemia in the GD group (reduced eGFR and/or albuminuria, n=69), and anemia in the RTD group (RTD, n=71).
As shown in Table 3, subjects in the anemia group were older, included more males, and had a higher prevalence of diabetes mellitus and reduced eGFR than those in the control and simple anemia groups. They also had higher levels of blood urea nitrogen, UACR, BNP, and H-FABP than the control and simple anemia groups. Notably, the subjects in the anemia group had a lower serum albumin level than the control and anemia of the GD groups.
Table 3.
Comparisons of Baseline Characteristics between Subjects with and without Anemia of RTD.
| Variables | Control n=3,064 | Simple anemia n=260 | Anemia of GD n=69 | Anemia of RTD (+) n=71 | p value |
|---|---|---|---|---|---|
| Age, years | 62±10 | 61±12 | 68±10*† | 71±8*† | <0.0001 |
| Men/women, n | 1,447/1,617 | 60/200 | 21/48 | 34/37 | <0.0001 |
| Obese, n (%) | 800 (26) | 33 (13) | 15 (22) | 10 (15) | <0.0001 |
| BMI, kg/m2 | 23.7±4.5 | 22.3±3.0* | 22.9±3.2 | 22.6±3.1 | <0.0001 |
| Hypertension, n (%) | 1,122 (37) | 72 (28) | 37 (54) | 32 (45) | 0.0003 |
| Systolic BP, mmHg | 135±16 | 129±17* | 135±16† | 139±16† | <0.0001 |
| Diastolic BP, mmHg | 80±10 | 75±10* | 77±11* | 78±10 | <0.0001 |
| Diabetes mellitus, n (%) | 186 (6) | 8 (3) | 5 (7) | 9 (13) | 0.0258 |
| HbA1c, % | 5.7±0.7 | 5.6±0.5 | 5.8±0.7 | 5.8±1.0† | 0.0174 |
| Fasting blood sugar, mg/dL | 95±17 | 90±13* | 94±18 | 94±24 | 0.0002 |
| Dyslipidemia, n (%) | 426 (14) | 31 (12) | 11 (16) | 8 (11) | 0.6871 |
| Atrial fibrillation, n (%) | 47 (1.5) | 1 (0.4) | 2 (2.9) | 0 (0) | 0.1154 |
| Smoking, n (%) | 1,038 (34) | 51 (19) | 15 (22) | 18 (25) | <0.0001 |
| Previous cardiovascular disease, n (%) | 399 (13) | 32 (12) | 16 (23) | 13 (18) | 0.0789 |
| Previous cancer, n (%) | 66 (2) | 5 (2) | 0 (0) | 3 (4) | 0.2406 |
| Biochemical data | |||||
| Blood urea nitrogen, mg/dL | 16.1±4.1 | 15.9±4.6 | 18.7±7.7*† | 19.4±11.9*† | <0.0001 |
| eGFR, mL/min/1.73m2 | 82±16 | 85±15* | 72±22*† | 72±28*† | <0.0001 |
| Reduced eGFR, n (%) | 191 (6) | 0 (0) | 26 (38) | 20 (28) | <0.0001 |
| G1/G2/G3a/G3b-G5, n | 894/1,978/174/18 | 88/172/0/0 | 14/29/20/6 | 14/37/10/10 | <0.0001 |
| Log10-transformed UACR, mg/gCr | 1.07±0.54 | 0.88±0.30* | 1.56±0.50*† | 1.65±0.70*† | <0.0001 |
| Albuminuria, n (%) | 578 (19) | 0 (0) | 53 (77) | 39 (54) | <0.0001 |
| A1/A2/A3, n | 2,486/520/58 | 260/0/0 | 16/51/2 | 32/31/8 | <0.0001 |
| Log10-transformed UBCR, μg/gCr | 2.04±0.41 | 1.99±0.31 | 1.96±0.40 | 2.87±0.49*†‡ | <0.0001 |
| RTD, n (%) | 370 (12) | 0 (0) | 0 (0) | 71 (100) | 0.0033 |
| Hemoglobin, g/dL | 13.8±1.4 | 11.2±1.3* | 11.5±0.9* | 11.5±1.0* | <0.0001 |
| Log10-transformed BNP, pg/mL | 1.27±0.37 | 1.42±0.32* | 1.58±0.41*† | 1.66±0.42*† | <0.0001 |
| Loge-transformed H-FABP, ng/mL | 1.24±0.42 | 1.23±0.45 | 1.43±0.52*† | 1.60±0.66*† | <0.0001 |
| Total protein, g/dL | 7.5±0.4 | 7.3±0.4* | 7.5±0.4† | 7.3±0.5* | <0.0001 |
| Serum albumin, g/dL | 4.5±0.2 | 4.4±0.3* | 4.4±0.2* | 4.3±0.3*‡ | <0.0001 |
| Total cholesterol, mg/dL | 202±32 | 188±31* | 193±29 | 184±41* | <0.0001 |
| Triglyceride, mg/dL | 110±66 | 82±40* | 95±49 | 96±8 | <0.0001 |
| HDL-C, mg/dL | 59±15 | 61±14* | 60±15 | 61±19 | 0.0255 |
| LDL-C, mg/dL | 126±30 | 113±28* | 116±27* | 105±28* | <0.0001 |
| Uric acid, mg/dL | 5.1±1.3 | 4.3±1.2* | 4.9±1.5† | 5.0±1.8† | 0.8875 |
Data are expressed as mean±SD, number (percentage), or median (interquartile range).
BMI: body mass index, BNP: brain natriuretic peptide, BP: blood pressure, eGFR: estimated glomerular filtration rate, GD: glomerular damage, HbA1c: glycosylated hemoglobin A1c, HDL-C: high-density lipoprotein cholesterol, H-FABP: heart-type fatty acid binding protein, LDL-C: low-density lipoprotein cholesterol, PVC: premature ventricular contraction, RTD: renal tubular damage, UACR: urinary albumin-creatinine ratio, UBCR: urinary β2-microglobulin-creatinine ratio
*p<0.05 vs. control, †p<0.05 vs. simple anemia, ‡p<0.05 vs. anemia of GD by analysis of variance Turkey’s post hoc test. Categorical data were analyzed using the chi-square test.
The subjects in the GD group were older and had a higher prevalence of reduced eGFR and albuminuria than those in the control and simple anemia groups. They also had higher levels of blood urea nitrogen, UACR, BNP, and H-FABP and a lower eGFR than the control and simple anemia groups.
Anemia of RTD and mortality in the general population
During the follow-up period, 719 all-cause and 219 composite cardiovascular, cerebrovascular, and kidney disease mortalities were observed. A Kaplan-Meier analysis demonstrated that subjects with anemia in the RTD group had the greatest risk of all-cause mortality in the general population (Fig. 3A). A Kaplan-Meier analysis demonstrated that subjects with anemia in the RTD group had the greatest risk of composite of kidney disease-related, cardiovascular, and cerebrovascular deaths in the general population (Fig. 3B). The mortality rates were similar between the control and simple anemia groups.
Figure 3.
(A) A Kaplan-Meier analysis for all-cause mortality among four groups based on the anemia and kidney dysfunction. (B) A Kaplan-Meier analysis for composite of kidney disease-related, cardiovascular, and cerebrovascular deaths among four groups based on the anemia and kidney dysfunction. GD: glomerular damage, RTD: renal tubular damage
Univariate and multivariate Cox proportional hazards regression analyses were performed to determine risk factors for all-cause mortality. Univariate Cox proportional hazard regression analysis showed that anemia with RTD was significantly associated with all-cause mortality (Fig. 4A). A multivariate Cox proportional hazard regression analysis demonstrated that anemia of the RTD was independently associated with all-cause mortality after adjusting for age, sex, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, and smoking (hazard ratio, 2.049; 95% confidence interval, 1.455-2.801; p＜0.0001; Fig. 4C). Similar results were obtained when all-cause mortality was substituted for premature death, defined as death occurring before the average age of death in the Japanese population (Supplementary material 2).
Figure 4.
Univariate and multivariate Cox proportional hazard regression analyses for all-cause mortality (A, C) and composite of kidney disease-related, cardiovascular, and cerebrovascular deaths (B, D) in the general population. Unadjusted (A, B) and adjusted (C, D) hazard ratios. A multivariate analysis was adjusted for age, sex, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, and smoking. GD: glomerular damage, RTD: renal tubular damage. *p<0.05 vs. control, †p<0.05 vs. simple anemia, ‡p<0.05 vs. anemia of GD.
Univariate and multivariate Cox proportional hazard regression analyses were performed to determine the risk factors for the composite of kidney disease-related, cardiovascular, and cerebrovascular deaths. A univariate Cox proportional hazard regression analysis showed that anemia of GD and anemia of RTD were significantly associated with a composite of kidney disease-related, cardiovascular, and cerebrovascular deaths (Fig. 4B). A multivariate Cox proportional hazard regression analysis demonstrated that anemia of RTD was independently associated with a composite of kidney disease-related, cardiovascular, and cerebrovascular deaths after adjustment for age, sex, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, and smoking (hazard ratio, 3.370; 95% confidence interval, 1.978-5.404; p＜0.0001; Fig. 4D).
Improvement of reclassification by addition of anemia of RTD to predict all-cause mortality
To examine whether the model fit and discrimination improved with the addition of anemia of RTD to the basic predictors, such as age, sex, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, and smoking, we evaluated the improvement in the C index, NRI, and IDI. The ROC curve analysis demonstrated that the C index of the baseline model for all-cause mortality was significantly improved by the addition of anemia to the RTD (Fig. 5). The NRI and IDI were also significantly improved by the addition of anemia to the RTD (Fig. 5).
Figure 5.
Upper, Receiver operating characteristics curves of baseline model with and without anemia of renal tubular damage for all-cause mortality in the general population. Lower, comparison of C index, NRI, and IDI for all-cause mortality. Baseline model includes age, sex, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, and smoking. IDI: integrated discrimination index, NRI: net reclassification index, RTD: renal tubular damage
Severity of RTD and mortality in subjects with anemia of RTD
Finally, we performed a subgroup analysis of the subjects with anemia of the RTD. As shown in Supplementary material 3, univariate and multivariate Cox proportional hazard regression analyses demonstrated that UBCR was an independent predictor of all-cause mortality in subjects with anemia of RTD, thus suggesting that the severity of RTD was associated with mortality even in these subjects.
Discussion
The novel findings from this study are as follows: (1) RTD is significantly associated with anemia; (2) anemia of RTD and GD were identified in 17.8% and 17.3% of patients with anemia, respectively; (3) a Kaplan-Meier analysis showed that anemia of RTD had the highest all-cause mortality among the four groups; (4) a multivariate Cox proportional hazard regression analysis demonstrated that anemia of RTD is an independent predictor for mortality in the general population; and (5) anemia of RTD improves the capacity to predict mortality in the general population.
RTD and anemia
There have been no reports on the relationship between RTD and anemia in the general population. Among older adults with anemia, approximately one-third have evidence of kidney dysfunction and/or chronic inflammation, another third have malnutrition, and the remaining third have unexplained anemia (3). Importantly, anemia due to chronic inflammation mainly results from CKD anemia in older adults (5), thus suggesting the importance of CKD in the development of anemia in older adults. Unexpectedly, we found that RTD was more closely associated with anemia than GD in the general population. The current understanding of the mechanism underlying anemia in CKD is that erythropoietin production is insufficient because of the transformation of renal erythropoietin-producing cells into myofibroblasts around the renal proximal tubules (17). As UBCR is a marker of reabsorption in the renal proximal tubules, our data may underestimate the impact of RTD on anemia. Further studies are required to accurately diagnose anemia associated with RTD.
Anemia of RTD and mortality
The Chronic Renal Insufficiency Cohort study demonstrated that several RTD markers, such as kidney injury molecule-1 and neutrophil gelatinase-associated lipocalin, are risk factors for certain cardiovascular diseases and all-cause death, even after adjusting for eGFR and UACR in patients with CKD (23). In the cadmium-polluted region of Japan, UBCR is reportedly associated with mortality due to kidney and renal tract diseases (24). Therefore, RTD, which is defined as UBCR, may be a predictor of cardiac and renal disease-associated mortality in the general population.
RTD is reportedly associated with a rapid decline in eGFR in the general population (25), suggesting that RTD is an early marker of GD. Similar to RTD, despite a slight decrease in serum albumin, a reduced serum albumin level is reportedly associated with a rapid decline in eGFR in the general Japanese population (26). We previously reported a close association between RTD, hypoalbuminemia, and objective malnutrition in patients with heart failure (27,28). In the present study, the anemia of the RTD group had a lower serum albumin level, even when compared with that of the GD group. Thus, we could not exclude the possibility that anemia of RTD is affected by anemia due to GD. Furthermore, anemia in the RTD group had higher levels of BNP and H-FABP, which are prognostic markers of cardiovascular events (29). The serum albumin and BNP levels are associated with exceptional survival at the highest age (30). In addition, anemia with myocardial damage is associated with higher all-cause and cardiovascular mortality in the general population (31). These reports support our finding that anemia in RTD is a risk factor for mortality in the general population.
Since this was a prospective observational study, we could not determine the mechanism by which anemia of RTD exacerbated the outcomes in the general population. However, we showed that the prediction model for all-cause mortality improved significantly after adding anemia to the RTD, as evidenced by improvements in the C-index, NRI, and IDI. This result confirms the clinical importance of anemia in RTD for predicting mortality, and raises the possibility that it could be an additional clinical factor to consider in the general population.
Limitation
This study was conducted with a relatively large number of participants and a long follow-up period, suggesting that the results were reliable. However, this study was associated with several limitations. First, the baseline information was collected at a single time point. Subsequent medical intervention may have affected RTD anemia. Second, non-fatal diseases were not assessed. Thus, we may have underestimated the impact of anemia on the clinical outcomes of RTDs. Finally, since we had no precise data regarding the iron status and erythropoietin and hepcidin levels, we could not determine the mechanism of anemia in RTD. Further studies are thus required to determine their clinical application.
Conclusion
For the first time, we demonstrated a close association between RTD and anemia. Notably, RTD anemia independently predicted mortality and improved the predictive capacity for all-cause deaths in the general population. RTD anemia could therefore be a feasible marker for all-cause mortality in the general population.
The authors state that they have no Conflict of Interest (COI).
Supplementary Material
Univariate and multivariate analyses for all-cause mortality in subjects with anemia of renal tubular damage. CI, confidence interval; eGFR, estimated glomerular filtration rate; HR, hazard ratio; UACR, urinary albumin-creatinine ratio; UBCR, urinary Beta2-microglobulin-creatinine ratio.*Per 1-SD increase.
The association between prevalence of anemia and the quintiles based on the urinary β2-microglobulin to creatinine ratio in the general population.
(A) Kaplan-Meier analysis for premature deaths among four groups based on the anemia and kidney dysfunction. GD, glomerular damage; RTD, renal tubular damage. (B, C) Univariate and multivariate Cox proportional hazard regression analyses for premature deaths in the general population. Unadjusted (B) and adjusted (C) hazard ratios. Multivariate analysis was adjusted for age, sex, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, and smoking. GD, glomerular damage; RTD, renal tubular damage. *P<0.05 vs. control, †P<0.05 vs. simple anemia.
References
- 1.Stevens GA, Finucane MM, De-Regil LM, et al. Global, regional, and national trends in haemoglobin concentration and prevalence of total and severe anaemia in children and pregnant and non-pregnant women for 1995-2011: a systematic analysis of population-representative data. Lancet Glob Health 1: e16-e25, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Stauder R, Valent P, Theurl I. Anemia at older age: etiologies, clinical implications, and management. Blood 131: 505-514, 2018. [DOI] [PubMed] [Google Scholar]
- 3.Patel KV. Epidemiology of anemia in older adults. Semin Hematol 45: 210-217, 2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zakai NA, French B, Arnold AM, et al. Hemoglobin decline, function, and mortality in the elderly: the cardiovascular health study. Am J Hematol 88: 5-9, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lanier JB, Park JJ, Callahan RC. Anemia in older adults. Am Fam Physician 98: 437-442, 2018. [PubMed] [Google Scholar]
- 6.Riva E, Colombo R, Moreo G, et al.; REPOSI Investigators . Prognostic value of degree and types of anaemia on clinical outcomes for hospitalised older patients. Arch Gerontol Geriatr 69: 21-30, 2017. [DOI] [PubMed] [Google Scholar]
- 7.Hanna RM, Streja E, Kalantar-Zadeh K. Burden of anemia in chronic kidney disease: beyond erythropoietin. Adv Ther 38: 52-75, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kidney Disease: Improving Global Outcomes (KDIGO) Diabetes Work Group . KDIGO 2020 Clinical Practice Guideline for Diabetes Management in Chronic Kidney Disease. Kidney Int 98: S1-S115, 2020. [DOI] [PubMed] [Google Scholar]
- 9.Babitt JL, Lin HY. Mechanisms of anemia in CKD. J Am Soc Nephrol 23: 1631-1634, 2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Astor BC, Muntner P, Levin A, Eustace JA, Coresh J. Association of kidney function with anemia: the Third National Health and Nutrition Examination Survey (1988-1994). Arch Intern Med 162: 1401-1408, 2002. [DOI] [PubMed] [Google Scholar]
- 11.Gluba-Brzózka A, Franczyk B, Olszewski R, Rysz J. The Influence of Inflammation on Anemia in CKD Patients. Int J Mol Sci 21: 725, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Nakhoul N, Batuman V. Role of proximal tubules in the pathogenesis of kidney disease. Contrib Nephrol 169: 37-50, 2011. [DOI] [PubMed] [Google Scholar]
- 13.Bernier GM, Conrad ME. Catabolsm of human β2-microglobulin by the rat kidney. Am J Physiol 217: 1359-1362, 1969. [DOI] [PubMed] [Google Scholar]
- 14.Miyata T, Jadoul M, Kurokawa K, Van Ypersele de Strihou C. β-2 microglobulin in renal disease. J Am Soc Nephrol 9: 1723-1735, 1998. [DOI] [PubMed] [Google Scholar]
- 15.Otaki Y, Watanabe T, Takahashi H, et al. Association of renal tubular damage with cardio-renal anemia syndrome in patients with heart failure. Int J Cardiol 173: 222-228, 2014. [DOI] [PubMed] [Google Scholar]
- 16.Otaki Y, Watanabe T, Shishido T, et al. The impact of renal tubular damage, as assessed by urinary β2-microglobulin-creatinine ratio, on cardiac prognosis in patients with chronic heart failure. Circ Heart Fail 6: 662-668, 2013. [DOI] [PubMed] [Google Scholar]
- 17.Kragesteen BK, Giladi A, David E, et al. The transcriptional and regulatory identity of erythropoietin producing cells. Nat Med 29: 1191-1200, 2023. [DOI] [PubMed] [Google Scholar]
- 18.Konta T, Hao Z, Abiko H, et al. Prevalence and risk factor analysis of microalbuminuria in Japanese general population: the Takahata study. Kidney Int 70: 751-756, 2006. [DOI] [PubMed] [Google Scholar]
- 19.Yokoyama M, Watanabe T, Otaki Y, et al. Association of the aspartate aminotransferase to alanine aminotransferase ratio with BNP level and cardiovascular mortality in the general population: the Yamagata study 10-year follow-up. Dis Markers 2016: 4857917, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Dotare T, Ishiwata S, Matsue Y, et al. Prevalence and prognostic relevance of isolated tubular dysfunction in patients with acute heart failure. Circ J 86: 709-714, 2022. [DOI] [PubMed] [Google Scholar]
- 21.Matsuo S, Imai E, Horio M, et al.; Collaborators developing the Japanese equation for estimated GFR . Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 53: 982-992, 2009. [DOI] [PubMed] [Google Scholar]
- 22.Nutritional anaemias. Report of a WHO scientific group. World Health Organ Tech Rep Ser 405: 5-37, 1968. [PubMed] [Google Scholar]
- 23.Park M, Hsu CY, Go AS, et al.; Chronic Renal Insufficiency Cohort (CRIC) Study Investigators; CKD Biomarkers Consortium . Urine kidney injury biomarkers and risks of cardiovascular disease events and all-cause death: the CRIC study. Clin J Am Soc Nephrol 12: 761-771, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Sakurai M, Suwazono Y, Nishijo M, et al. Relationship between urinary β2-microglobulin concentration and mortality in a cadmium-polluted area in Japan: a 35-year follow-up study. J Appl Toxicol 41: 224-232, 2021. [DOI] [PubMed] [Google Scholar]
- 25.Kudo K, Konta T, Mashima Y, et al. The association between renal tubular damage and rapid renal deterioration in the Japanese population: the Takahata study. Clin Exp Nephrol 15: 235-241, 2011. [DOI] [PubMed] [Google Scholar]
- 26.Amano H, Yoshimura K, Iijima R, et al. A slight decrease in the serum albumin level is associated with the rapid progression of kidney dysfunction, even within the normal range. Intern Med 59: 2679-2685, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Otaki Y, Watanabe T, Shimizu M, et al. Association of malnutrition with renal dysfunction and clinical outcome in patients with heart failure. Sci Rep 12: 16673, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Otaki Y, Watanabe T, Takahashi H, et al. Comorbid renal tubular damage and hypoalbuminemia exacerbate cardiac prognosis in patients with chronic heart failure. Clin Res Cardiol 105: 162-171, 2016. [DOI] [PubMed] [Google Scholar]
- 29.Otaki Y, Watanabe T, Takahashi H, et al. Association of heart-type fatty acid-binding protein with cardiovascular risk factors and all-cause mortality in the general population: the Takahata study. PLoS One 9: e94834, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hirata T, Arai Y, Yuasa S, et al. Associations of cardiovascular biomarkers and plasma albumin with exceptional survival to the highest ages. Nat Commun 11: 3820, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Honda Y, Watanabe T, Otaki Y, et al. Gender differences in the impact of anemia on subclinical myocardial damage and cardiovascular mortality in the general population: the Yamagata (Takahata) study. Int J Cardiol 252: 207-212, 2018. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Univariate and multivariate analyses for all-cause mortality in subjects with anemia of renal tubular damage. CI, confidence interval; eGFR, estimated glomerular filtration rate; HR, hazard ratio; UACR, urinary albumin-creatinine ratio; UBCR, urinary Beta2-microglobulin-creatinine ratio.*Per 1-SD increase.
The association between prevalence of anemia and the quintiles based on the urinary β2-microglobulin to creatinine ratio in the general population.
(A) Kaplan-Meier analysis for premature deaths among four groups based on the anemia and kidney dysfunction. GD, glomerular damage; RTD, renal tubular damage. (B, C) Univariate and multivariate Cox proportional hazard regression analyses for premature deaths in the general population. Unadjusted (B) and adjusted (C) hazard ratios. Multivariate analysis was adjusted for age, sex, hypertension, diabetes mellitus, dyslipidemia, atrial fibrillation, and smoking. GD, glomerular damage; RTD, renal tubular damage. *P<0.05 vs. control, †P<0.05 vs. simple anemia.