Background
Chronic kidney disease (CKD) is known to cause anemia mainly due to inappropriate erythropoietin (EPO) secretion [
1,
2], especially in patients with end-stage renal failure. However, anemia can occur early in the development of CKD, defined on the basis of the estimated glomerular filtration rate (eGFR). This indicates that the progression of renal anemia is not governed solely by glomerular function [
3]. In recent studies, EPO-producing cells were identified in the renal tubulointerstitium surrounding the central vessels, not in the glomeruli [
4,
5]. Thus, unfavorable changes in the interstitial microenvironment, such as inflammation, oxidative stress and ischemia, can impair the function of EPO-producing cells. Recently, Souma et al. demonstrated that the phenotypic transition of EPO-producing cells to non-EPO-producing myofibroblasts is modulated by inflammatory molecules, and suggested the connection between anemia and renal fibrosis in CKD in a mouse model [
6].
Hypertension is one of the leading causes of CKD associated with arteriosclerosis, interstitial inflammation and fibrosis, as well as tubular insufficiency secondary to endothelial dysfunction and progressive infiltration of macrophages and T cells to the perivascular interstitium induced by chronic exposure to high blood pressure [
7-
10]. Therefore, the development of renal damage can contribute to the pathogenesis of anemia, even in earlier stages of CKD in patients with hypertension.
Renal resistive index (RI) in the intra-renal arteries at the level of the corticomedullary junction using pulsed Doppler ultrasonography is a simple index of renal vascular resistance [
11,
12], and high renal RI is known to be associated with severe interstitial fibrosis, arteriosclerosis and renal function decline [
13-
15].
Accordingly, we examined whether renal RI is associated with the presence and severity of anemia, and whether high renal RI predicts the future development of anemia in patients with hypertension.
Clinical outcome
Medical records were retrospectively reviewed for each patient. The study end-point was 1) new anemia (<12.0 g/dl for women and <13.0 g/dl for men) for non-anemic patients, and 2) decreased hemoglobin levels greater than 1 g/dl and/or initiation of treatment including iron supplementation and EPO-stimulating agents for anemic patients. Anemic events within 3 months after minor surgery or cardiac catheterization and within 6 months after major surgery were excluded from the analysis.
Statistical analysis
All continuous variables were approximately normally distributed and presented as the mean ± standard deviation. In order to assess differences between the two groups, Pearson’s chi-square test was used for nominal scales and Student’s t test was used for all other scales. To make comparisons among three groups, we used Bonferroni’s multivariate comparison test. The linear correlations between the variables were parametrically evaluated using Pearson’s product moment correlation coefficient. We performed stepwise multivariate linear regression analysis to determine independent predictors of hemoglobin levels. Moreover, we examined the predictive value of potential clinical parameters for the study end-point, namely, the future development of anemia. Cumulative event rates were assessed using the Kaplan–Meier method and compared using the log-rank test. The impact of clinical predictors on the future development of anemia was assessed by univariate and multivariate Cox proportional hazards regression. Hazard ratios are given with 95% confidence intervals (CI). Reliability was assessed by calculating intraobserver and interobserver intraclass correlation coefficients. For all tests, a p value <0.05 was considered statistically significant. Data were analyzed using SPSS Statistics for Windows (Version 20.0, IBM Corp., Armonk, NY, US).
Discussion
We demonstrated for the first time that renal RI was independently associated with hemoglobin level in patients with hypertension. Furthermore, high renal RI and the presence of proteinuria predicted the future development of anemia after correcting for confounding factors including age and diabetes mellitus in the present study.
According to Japan’s 2011 National Nutrition Survey, anemia was diagnosed in 8.9% of men and 8.5% of women aged 60-69 in the general population [
24]. In the present study, however, the prevalence of anemia was as high as 37% in men and 34% in women in this study population. Renal damage secondary to hypertension is characterized histologically by interstitial fibrosis, arteriosclerosis and glomerular sclerosis [
10]. Chronic exposure to high blood pressure initially causes endothelial dysfunction and progressive infiltration of macrophages and T cells to the perivascular interstitium. Interactions of these cells and their cytokines with parenchymal cells cause interstitial fibrosis and tubular insufficiency [
7-
10]. In recent studies, EPO-producing cells were identified in the renal tubulointerstitium surrounding the central vessels [
4,
5]. Unfavorable changes in the interstitial microenvironment such as inflammation, oxidative stress and ischemia resulting from hypertension can impair the function of EPO-producing cells. Recently, Souma et al. demonstrated that the phenotypic transition of EPO-producing cells to non-EPO-producing myofibroblasts is modulated by inflammatory molecules, and suggested the connection between anemia and renal fibrosis in CKD in a mouse model [
6]. Therefore, the development of CKD secondary to hypertension ought to affect hematopoiesis even in the earlier stages of CKD defined by eGFR. Indeed, there was a significant difference in the hemoglobin levels between patients with renal RI higher than the median value (0.70) and those with renal RI ≤0.7 only in the subgroup of 30 ≤ eGFR <60 ml/min/1.73 m
2.
We further demonstrated that renal RI and urine protein predicted the future development of anemia. Renal RI reflects the severity of vascular and tubulointerstitial lesions, and has been shown to correlate with an inflammatory state [
13-
15,
25]. We also found that there was no relationship between the change in hemoglobin level and change in eGFR during the follow-up period. These findings may suggest that baseline renal RI and urine protein, rather than worsening glomerular function defined by eGFR, had a greater impact on the development of anemia. Although it would be difficult to identify the underlying mechanisms responsible for the development of anemia precisely, damage of renal interstitial compartments and arteriosclerosis may contribute to the disease process independently of glomerular function, especially in the early stages of CKD defined by eGFR. The causal interrelationship between the progression of CKD and the development of anemia warrants further investigation.
We also demonstrated that the presence of proteinuria is a powerful independent predictor of the future development of anemia. Proteinuria is an important sign of CKD, which can result from hypertension, diabetes mellitus and diseases that cause inflammation in the kidneys. It is recognized that proteins abnormally filtered across the glomerular barrier have intrinsic renal toxicity linked to their over-reabsorption by proximal tubular cells and activation of tubular-dependent pathways of interstitial inflammation and fibrosis [
26,
27]. Protein overload causes a dose-dependent increase in nuclear factor kappa beta activation in proximal tubular cells that leads to myofibroblast transformation of EPO-producing cells [
6,
28,
29]. We demonstrated that patients with high renal RA and proteinuria had the greatest risk for the future development of anemia. The relationship between proteinuria and renal RI as causal mechanisms underlying the future development of anemia in CKD warrants further investigation. It has been reported that the risk of anemia in patients with diabetes mellitus is approximately two to three times that of the general population with the same level of eGFR due to tubulointerstitial injury secondary to proteinuria and dysglycemia [
30]. Diabetes mellitus is commonly found in patients with hypertension and vice versa. In the present study, 52% of patients had diabetes mellitus, and diabetic patients had a higher prevalence of proteinuria than non-diabetic ones. It is difficult to discriminate proteinuria secondary to diabetic nephropathy from that independent of diabetes mellitus per se. However, the presence of proteinuria was independently associated with the future development of anemia after correcting for diabetes mellitus in the present study.
The presence of anemia is known to result in a worse prognosis in terms of both morbidity and mortality. An earlier diagnosis and optimal treatment of anemia would reduce incidence rates of cardiovascular diseases, as well as slow the decline of renal function [
31]. Measuring renal RI in addition to conventional markers of renal damage including eGFR and urine protein will help clinicians to decide whether or not the anemia is secondary to renal damage, especially in the early stages of CKD. In addition, high renal RI may aid in alerting them to monitor hemoglobin levels even if patients have no sign of anemia or advanced renal failure on the basis of eGFR.
The limitations of this study include the small sample size and the retrospective nature of the data collection. Renal ultrasonography was performed for the screening of renal artery stenosis and the evaluation of renal arteriosclerosis at the physician’s discretion on the basis of the patients’ age, comorbidity and disease characteristics. Accordingly, the prevalence of coexisting atherosclerotic diseases in patients who were enrolled in the present study can be higher than that of hypertensive patients among the general population. This selection bias may influence the relationship between renal RI and anemia, and the clinical impact of renal RI on the future development of anemia. Renal RI reflects systemic vascular stiffness as well as renal arteriolosclerosis. Indeed, we found a statistically significant positive correlation between renal RI and pulse pressure, one of the surrogate measures of large artery stiffness. Risk factors for systemic atherosclerosis such as inflammation and oxidative stress may affect renal RI and anemia via different pathways, which can confound precise understanding of the causal mechanism underlying the anemia in patients with hypertension and renal damage. However, stepwise multivariate linear regression analyses revealed that renal RI but not pulse pressure was independently associated with hemoglobin level. We were unable to obtain sufficient information to examine precisely the etiology of anemia including iron metabolism, erythropoietin productivity and responsiveness, and inflammation. Iron deficiency is common in patients with CKD. Estrella et al. found that 42.6% of patients who had both anemia and CKD with eGFR between 15 and 59 ml/min/1.73 m
2 were classified as having functional or absolute iron-deficiency anemia (33.5% and 9.1%, respectively) [
32]. Therefore, patients in the anemic group and even in the non-anemic group can have varying degrees of iron deficiency. The causal relationships between iron status and anemia in the hypertensive population, especially those with renal damage, warrant further investigation. Artunc and Risler reported that the correlation between hemoglobin level and EPO concentration was gradually attenuated with increasing stages of CKD and was mostly lost in CKD stages 4 and 5 in an observational study, indicating the presence of more prominent EPO deficiency in advanced CKD [
2]. In other words, the pathogenesis of anemia in the early stages of CKD might be heterogeneous in terms of the severity of impairment of EPO productivity and responsiveness. Inflammation also influences EPO efficacy and production as well as renal atherosclerosis. Urine proteins were qualified by using a dipstick test in the present study. Therefore, the potential contribution of microalbuminuria to the pathogenesis of anemia, especially in patients with diabetes, was not investigated. The analysis of markers of tubulointerstitial damage such as beta-2 microglobulins or N-acetylglucosaminidase was also not carried out. Several factors including obesity-related obstructive sleep apnea, current smoking and COPD may contribute to persistent or intermittent hypoxia that lead to red cell production via EPO stimulation. Indeed, body mass index was positively related to hemoglobin level in the present study. However, the presence and severity of sleep apnea and their contribution to hemoglobin level were not assessed in the present study. Finally, follow-up data of renal RI, urine protein, and changes in nutritional status and medication were not included in our outcome evaluation.
Acknowledgements
The authors are grateful to GE Healthcare and Toshiba Medical Systems Corporation for their technical support, and to Yuko Sakurai, MT, CVT, and Harumi Fukuda, MT, CVT, for their technical assistance.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MT and KD designed the study, performed the data analysis and wrote the manuscript. MT, YS, ES, NK, SN and NF collected the data. MM undertook ultrasonography examination and collected the data. NF, TT, NY and MN provided comments on the manuscript. TY supervised the statistical analysis. MI supervised and approved the manuscript. All authors read and approved the final manuscript.