In this study, we have demonstrated that serum Hb concentration, which is a widely available, routinely measured marker, is associated with renal pathological changes (especially with interstitial fibrosis) and DKD progression in patients with Type 2 diabetes in the early stages of CKD. The association between serum Hb concentration and DKD progression remained significant when adjusting for the known risk factors for DKD progression, including eGFR and UACR. Addition of the serum Hb concentration to the known risk factors of DKD progression improved the prognostic value of DKD progression, which suggests that serum Hb concentration can be useful for predicting DKD progression.
Biomarkers to predict DKD progression currently rely exclusively on eGFR and UACR. However, using these markers for prognosticating DKD progression is challenging in early stages of CKD, where eGFR has not yet declined and UACR has not severely increased. In addition, previous morphological and autopsy studies have demonstrated that some patients with diabetes and preserved renal function have already undergone glomerular and tubulointerstitial damage,[23–25] which highlights the need for biomarkers that relate to both incipient pathological changes and renal prognosis in patients with early stages of CKD. Recently, a number of novel biomarkers have shown to be of prognostic value in patients with diabetes in the early stages of CKD; however, they are too laborious for routine use. We therefore performed a thorough search of widely available, routinely measured, clinical applicable markers that were strongly associated with pathological changes and found that serum Hb concentration negatively correlated with all scores of renal lesions, especially with glomerular damage and scores of interstitial lesions, all of which are major predictors of DKD progression.[16,26–28] Among pathological lesions, it is well known that interstitial lesions rather than glomerular lesions correlate better with renal function loss both in DKD and in other glomerular diseases.[29,30] Recent studies proposed potential biomarkers that predict DKD progression including serum and urinary biomarkers of tubulointerstitial injury such as kidney injury molecule-1, liver-type fatty acid-binding protein and N-acetyle-β-D-glucosaminidase. Again, these biomarkers are favorable markers, but are not ready for clinical use.
We found that serum Hb concentration is a candidate marker that is associated with tubulointerstitial injury. Though many factors may explain the presence of anemia in patients with CKD, accumulating evidence suggests that renal fibrosis, which interferes with erythropoietin production, is the main reason for anemia.[12,13] However, studies investigating the association between tubulointerstitial injury and anemia in humans are scarce, since kidney biopsy is not always applicable for patients with DKD, especially with early stages of CKD. Our data showed a concrete relationship between tubulointerstitial fibrosis and serum Hb concentration.
It is known that there is a gender difference in serum Hb concentration, and females have been reported to have mean serum Hb concentration lower than males in both healthy adults and adults with CKD.[33,34] Since the mean circulating erythropoietin level does not differ between males and females, or between pre- and post-menopausal females, the difference in serum Hb concentration has been reported to be mainly due to estrogen hormone, which inhibits the production of red blood cells in the bone marrow. Therefore, we conducted the subgroup analysis separately for males and females to see whether there is a difference in results between genders. We found that despite the fact that females have a median serum Hb concentration lower than that of males (12.3 g/dL for females versus 13.8 g/dL for males), both the quartiles of serum Hb concentration for males and females stratified the cumulative incidence of DKD progression (Supplementary data, Figure S2A and B). However, it appeared that serum Hb concentration is much more discriminatory between the quartiles in males than females (P for trend: <0.001 for males versus 0.015 for females). This finding may be attributed to the fact that the age of the female population ranged from 30 to 82 years, which indicates that the study population includes both pre- and post-menopausal females, suggesting that miscellaneous estrogen levels may have somehow distorted the discriminatory power of serum Hb concentration on the incidence of DKD progression. In addition, iron deplete caused by menopause may also contribute to the change of serum Hb concentration. Unfortunately, data on estrogen and iron levels were not available in this study.
The strengths of our study are the use of study population from a real-world cohort of DKD across Japan with a long follow-up time that allows us to observe the relationship between serum Hb concentration and DKD progression, well evaluated pathological lesions and the precise ascertainment of renal outcome, all of which enabled robust analysis of the association between serum Hb concentration and renal pathological findings and DKD progression.
We acknowledge that our study has limitations. First, because of the nature of retrospective studies, we cannot conclude the causation between serum Hb concentration and DKD progression. For example, we cannot assure that the correction of serum Hb concentration to normal level reduces the rates of DKD progression, even though we found a strong association between serum Hb concentration and the risk of DKD progression. Second, we only measured serum Hb concentration at a single time point of kidney biopsy. Data on serum Hb concentration and interventions during the follow-up period were not available. Although considering them may have some advantages, we believe that the approach we adopted minimizes the risk of reverse causation in the interpretation of the results. Third, even though the study population is from a real-world cohort of biopsy-proven DKD, there may be a selection bias. There is a possibility that the population was biopsied because they were suspected to have any kind of kidney disease rather than DKD. In contrast, however, we believe that the use of biopsy-proven rather than inaccurate clinical diagnosis of DKD provides a clear picture of the clinical course of DKD. Fourth, data on erythropoietin level, ferritin level, iron level and vitamin deficiency were not available. Also, data on comorbid diseases that affect serum Hb level, such as gastrointestinal bleeding and malignancy, were not available. Fifth, the study population comprised Asian, mostly Japanese patients, and our findings may not apply to populations from different geographic origins. Sixth, we noticed a striking increase in cumulative incidence of DKD progression after Year 1 for patients in the lowest quartile. This originated from the lack of eGFR data for several patients before Year 1; these patients only had the second eGFR measurement at Year 1 since the first measurement of eGFR at renal biopsy (the reason for this is unknown). There is a possibility that they might have developed DKD progression before Year 1 if they had had eGFR measurements before Year 1. Conversely, this highlights the need for an intensive eGFR monitoring in patients with severe anemia even if they are in the early stages of DKD.
In conclusion, we have demonstrated that serum Hb concentration, which is a widely available, routinely measured, clinical applicable marker, is associated with renal pathological changes (especially with interstitial fibrosis) and DKD progression in patients with Type 2 diabetes in the early stages of CKD. Addition of serum Hb concentration to the known risk factors of DKD progression improved the prognostic value of DKD progression. These findings suggest that serum Hb concentration, as a reflection of interstitial damage, can be useful for predicting DKD progression.
The authors thank various people for their contribution to this project; Dr Yoko Yoshida, Ms Yuki Inoue, The Institute for Adult Disease, Asahi Life Foundation, Ms Yurina Takaishi, Ms Keiko Sahara and Ms Tokiko Hoshikawa, Toranomon Hospital, for their help in collecting the data.
This study was supported in part by a Ministry of Health, Labour and Welfare Grant-in-Aid for Diabetic Nephropathy and Nephrosclerosis Research (JP17ek0310003) and a grant for medical research from the Okinaka Memorial Institute for Medical Research, Tokyo, Japan. The funding source had no role in study design or execution, data analysis, manuscript writing or manuscript submission.
Nephrol Dial Transplant. 2022;37(3):489-497. © 2022 Oxford University Press