Serum Hemoglobin Concentration and Risk of Renal Function Decline in Early Stages of Diabetic Kidney Disease

A Nationwide, Biopsy-Based Cohort Study

Masayuki Yamanouchi; Kengo Furuichi; Miho Shimizu; Tadashi Toyama; Yuta Yamamura; Megumi Oshima; Shinji Kitajima; Akinori Hara; Yasunori Iwata; Norihiko Sakai; Yuki Oba; Shusaku Matsuoka; Daisuke Ikuma; Hiroki Mizuno; Tatsuya Suwabe; Junichi Hoshino; Naoki Sawa; Yukio Yuzawa; Hiroshi Kitamura; Yoshiki Suzuki; Hiroshi Sato; Noriko Uesugi; Yoshihiko Ueda; Shinichi Nishi; Hitoshi Yokoyama; Tomoya Nishino; Kenichi Samejima; Kentaro Kohagura; Yugo Shibagaki; Hirofumi Makino; Seiichi Matsuo; Yoshifumi Ubara; Takashi Wada

Disclosures

Nephrol Dial Transplant. 2022;37(3):489-497.

Abstract and Introduction

Abstract

Graphical Abstract

(Enlarge Image)

Background: Prognosticating disease progression in patients with diabetic kidney disease (DKD) is challenging, especially in the early stages of kidney disease. Anemia can occur in the early stages of kidney disease in diabetes. We therefore postulated that serum hemoglobin (Hb) concentration, as a reflection of incipient renal tubulointerstitial impairment, can be used as a marker to predict DKD progression.

Methods: Drawing on nationally representative data of patients with biopsy-proven DKD, 246 patients who had an estimated glomerular filtration rate (eGFR) ≥60 mL/min/1.73 m² at renal biopsy were identified: age 56 (45–63) years; 62.6% men; Hb 13.3 (12.0–14.5) g/dL; eGFR 76.2 (66.6–88.6) mL/min/1.73 m²; urine albumin-to-creatinine ratio 534 (100–1480) mg/g Crea. Serum Hb concentration was divided into quartiles: ≤12, 12.1–13.3, 13.4–14.5 and ≥14.6 g/dL. The association between serum Hb concentration and the severity of renal pathological lesions was explored. A multivariable Cox regression model was used to estimate the risk of DKD progression (new onset of end-stage kidney disease, 50% reduction of eGFR or doubling of serum creatinine). The incremental prognostic value of DKD progression by adding serum Hb concentration to the known risk factors of DKD was assessed.

Results: Serum Hb levels negatively correlated with all renal pathological features, especially with the severity of interstitial fibrosis (ρ = −0.52; P < 0.001). During a median follow-up of 4.1 years, 95 developed DKD progression. Adjusting for known risk factors of DKD progression, the hazard ratio in the first, second and third quartile (the fourth quartile was reference) were 2.74 [95% confidence interval (CI) 1.26–5.97], 2.33 (95% CI 1.07–5.75) and 1.46 (95% CI 0.71–3.64), respectively. Addition of the serum Hb concentration to the known risk factors of DKD progression improved the prognostic value of DKD progression (the global Chi-statistics increased from 55.1 to 60.8; P < 0.001).

Conclusions: Serum Hb concentration, which reflects incipient renal fibrosis, can be useful for predicting DKD progression in the early stages of kidney disease.

Introduction

Diabetic kidney disease (DKD) is not just the most prevalent form of chronic kidney disease (CKD) but also the leading cause of end-stage kidney disease (ESKD) worldwide.^[1–3] It also accounts for an increase in cardiovascular disease (CVD) and mortality.^[4] Since morbidity and mortality increase as the renal function declines,^[5] the early identification of patients at high risk of CKD progression is important so that clinicians can provide intensive treatment that may ultimately alter the prognosis of DKD. Although some novel biomarkers have shown to be of prognostic value in patients with diabetes in the early stages of CKD,^[6] it is too laborious for routine use. This predicament highlights the need for widely available, routinely measured markers that reflect early kidney damage for prognosticating the clinical course of DKD, especially in the earlier stages of CKD.

Anemia is common among patients with CKD^[7] and it is associated with the risk of progression of kidney disease, CVD and mortality.^[8–11] Although possible causes of anemia in CKD include iron and vitamin deficiency, the major cause of anemia in CKD is renal tubulointerstitial impairment that interferes with erythropoietin production.^[12,13] Therefore, anemia is more common and severe in the later stages of CKD.^[7] Meanwhile, anemia occurs in the earlier stages of CKD in patients with diabetes.^[14] However, data on the association between anemia, tubulointerstitial lesions and renal prognosis are lacking in the early stages of CKD in patients with diabetes due to the fact that kidney biopsy is not always applicable for them.

Fortunately, the Ministry of Health, Labor and Welfare in Japan has currently established a nationwide biopsy-proven cohort of DKD, including patients in the early stages of CKD.^[15,16] We therefore postulate that anemia in the early stages of CKD reflects incipient kidney injury due to tubulointerstitial damage, and thus serum hemoglobin (Hb) concentration in the early stages of CKD could be useful for prognosticating the clinical course of DKD. The objectives of our study were to: (i) evaluate the association between serum Hb concentration and tubulointerstitial lesions in kidney biopsy specimens and (ii) quantify the risk for progression of CKD, according to the serum Hb concentration, in patients with Type 2 diabetes and biopsy-proven DKD in the early stages of CKD.

1 2 3 4

Next Section

Table 1. Baseline clinical and pathological characteristics of the patients, overall and stratified by the quartiles of serum Hb concentration
Patient characteristics	All	Serum Hb concentration				P for trend
Patient characteristics	All	Quartile 1 ≤12.0 g/dL	Quartile 2 12.1–13.3 g/dL	Quartile 3 13.4–14.5 g/dL	Quartile 4 ≥14.6 g/dL	P for trend
Clinical characteristics at baseline
Serum Hb concentration, g/dL	13.3 (12.0–14.5)	11.1 (9.8–11.7)	12.8 (12.3–13.1)	13.8 (13.6–14.2)	15.4 (15.1–16.1)	<0.001
Number of patients	246	64	61	61	60
Age, years	56 (45–63)	57 (49–66)	58 (49–64)	57 (46–63)	51 (41–58)	<0.001
Male, %	62.6	42.2	52.5	63.9	93.3	<0.001
Diabetes duration, years	10 (5–17)	13 (8–18)	10 (5–20)	11 (6–17)	7 (3–14)	0.027
BMI, kg/m²	22.9 (20.9–25.6)	21.8 (20.3–24.0)	22.8 (20.3–25.4)	22.8 (20.8–25.4)	25.0 (22.1–27.7)	<0.001
HbA1c, %	7.9 (6.6–9.6)	7.4 (6.2–8.3)	8.8 (7.2–10.2)	8.5 (6.9–10.2)	7.7 (6.5–8.9)	0.718
eGFR, mL/min/1.73 m²	76.2 (66.6–88.6)	70.3 (65.8–79.4)	76.3 (65.9–86.9)	77.4 (67.6–86.6)	81.9 (68.4–95.7)	0.002
UACR, mg/g creatinine	534 (100–1480)	1510 (241–2700)	693 (153–1545)	360 (50–1121)	252 (120–650)	<0.001
Systolic BP, mmHg	136 (124–150)	133 (125–153)	138 (126–156)	140 (127–152)	130 (117–140)	0.046
Diastolic BP, mmHg	78 (70–87)	76 (68–84)	77 (70–86)	80 (70–89)	78 (70–88)	0.391
Total cholesterol, mg/dL	210 (177–238)	212 (185–252)	209 (170–243)	202 (175–232)	212 (181–235)	0.236
Low-density lipoprotein cholesterol, mg/dL	131 (106–158)	121 (105–172)	129 (92–161)	129 (109–150)	136 (109–158)	0.743
High-density lipoprotein cholesterol, mg/dL	45 (37–56)	50 (34–63)	52 (40–70)	42 (34–51)	45 (37–49)	0.239
Triglycerides, mg/dL	141 (97–200)	134 (85–189)	107 (90–158)	155 (103–192)	168 (112–253)	0.024
Medication usage, %
RAS blocker	50.7	56.4	50.0	56.7	38.7	0.110
Glucose-lowering medication	52.2	46.2	61.1	66.7	35.5	0.577
Lipid-lowering medication	16.8	20.7	17.2	26.1	0.0	0.166
Ever having smoked, %	53.8	33.3	47.4	59.1	66.7	0.033
History of CVD, %	16.1	7.9	17.1	21.4	17.1	0.232
Pathological characteristics at baseline
Renal Pathology Society Diabetic Nephropathy Classification
Glomerular lesions, %						<0.001

Class I	36.7	20.0	31.0	46.0	43.2
Class IIa	20.3	4.00	10.3	24.3	20.3
Class IIb	18.0	16.0	34.5	13.5	10.8
Class III	23.4	60.0	20.7	16.2	8.1
Class IV	1.6	0.0	3.5	0.0	2.7
Interstitial lesions
IFTA, %						<0.001
0	23.2	12.5	14.8	27.9	38.3
1	41.9	32.8	47.5	42.6	26.2
2	24.8	26.7	31.2	27.9	10.0
3	10.1	25.0	6.6	1.6	6.7
Interstitial inflammation, %						<0.001
0	31.7	15.6	27.9	36.1	48.3
1	53.7	60.9	59.0	50.8	43.3
2	14.6	23.5	13.1	13.1	8.4
Vascular lesions
Arteriolar hyalinosis, %						0.019
0	15.9	9.4	11.7	14.8	28.3
1	56.7	65.6	55.0	55.7	50.0
2	27.4	25.0	33.3	29.5	21.7
Large vessels arteriosclerosis, %)						0.059
0	27.7	28.1	16.7	31.0	35.0
1	47.5	42.2	56.7	40.0	51.7
2	26.8	26.7	26.6	29.0	13.3

Data are expressed as the median (25–75th percentiles) or percentage.

Table 2. Correlations between serum Hb concentration and baseline clinical and pathological characteristics
Patient characteristics	Clinical characteristics				Pathological characteristics
	eGFR	UACR	Systolic BP	HbA1c	RPS Diabetic Nephropathy Classification Glomerular classification	Interstitial lesions		Vascular lesions
	eGFR	UACR	Systolic BP	HbA1c		IFTA	Interstitial inflammation	Arteriolar hyalinosis	Large vessels arteriosclerosis
Hb	0.54***	−0.30***	−0.24***	0.31***	−0.47***	−0.52***	−0.26***	−0.26***	−0.21**
eGFR	–	−0.40***	−0.29***	0.31***	−0.49***	−0.60***	−0.27***	−0.20***	−0.39**
UACR	–	–	0.39***	−0.17***	0.39***	0.43***	0.13***	0.18***	0.20**
Systolic BP				−0.09*	0.33***	0.30***	0.13***	0.18**	0.16*
HbA1c	–	–	–	–	−0.14*	−0.20***	−0.02*	0.03	−0.05*
Glomerular classification		–	–	–	–	0.50***	0.20***	0.35***	0.31***
IFTA		–	–	–	–	–	0.50***	0.39***	0.39***
Interstitial inflammation		–	–	–	–	–	–	0.43***	0.11***
Arteriolar hyalinosis		–	–	–	–	–	–	–	0.13***

Spearman's correlation coefficients between serum Hb concentration and clinical and pathological characteristics were obtained using Spearman's correlation rank test.
*P < 0.05,
**P < 0.01,
***P < 0.001.
RPS, Renal Pathology Society.

Table 3. Crude and adjusted risk of DKD progression, according to the quartiles of serum Hb concentration
Models HR (95% CI)	Serum Hb concentration				P for trend
Models HR (95% CI)	Quartile 1 ≤12.0 g/dL	Quartile 2 12.1–13.3 g/dL	Quartile 3 13.4–14.5 g/dL	Quartile 4 ≥14.6 g/dL	P for trend
Model1	3.81 (2.14–6.80)	2.34 (1.29–4.24)	1.48 (0.78–2.80)	Ref.	<0.001
Model 2	4.37 (2.28–8.37)	2.64 (1.38–5.04)	1.61 (0.81–3.18)	Ref.	<0.001
Model 3	2.74 (1.26–5.97)	2.33 (1.07–5.75)	1.46 (0.71–3.64)	Ref.	<0.001

HR (95% CI) and P-values were determined for demographic and laboratory characteristics by univariable and multivariable Cox proportional models.
Model 1: univariable.
Model 2: adjusted for age and gender.
Model 3: adjusted for known risk factors of DKD progression (age, gender, BMI, diabetic duration, systolic BP, Hb A1c, eGFR, UACR and RAS blocker use).

Authors and Disclosures

Masayuki Yamanouchi^1–4, Kengo Furuichi⁵, Miho Shimizu³, Tadashi Toyama³, Yuta Yamamura³, Megumi Oshima³, Shinji Kitajima³, Akinori Hara³, Yasunori Iwata^3,6, Norihiko Sakai³, Yuki Oba², Shusaku Matsuoka², Daisuke Ikuma², Hiroki Mizuno², Tatsuya Suwabe^1,2, Junichi Hoshino^1,2, Naoki Sawa^1,2, Yukio Yuzawa⁷, Hiroshi Kitamura⁸, Yoshiki Suzuki⁹, Hiroshi Sato¹⁰, Noriko Uesugi¹¹, Yoshihiko Ueda¹², Shinichi Nishi¹³, Hitoshi Yokoyama⁵, Tomoya Nishino¹⁴, Kenichi Samejima¹⁵, Kentaro Kohagura¹⁶, Yugo Shibagaki¹⁷, Hirofumi Makino¹⁸, Seiichi Matsuo¹⁹, Yoshifumi Ubara^1,2 and Takashi Wada³, on behalf of the Research Group of Diabetic Nephropathy, the Ministry of Health, Labour and Welfare, and the Japan Agency for Medical Research and Development

¹Nephrology Center, Toranomon Hospital, Tokyo, Japan, ²Nephrology Center, Toranomon Hospital Kajigaya, Kanagawa, Japan, ³Department of Nephrology and Laboratory Medicine, Kanazawa University, Ishikawa, Japan, ⁴Okinaka Memorial Institute for Medical Research, Tokyo, Japan, ⁵Department of Nephrology, Kanazawa Medical University School of Medicine, Ishikawa, Japan, ⁶Division of Infection Control, Kanazawa University, Ishikawa, Japan, ⁷Department of Nephrology, Fujita Health University School of Medicine, Aichi, Japan, ⁸Department of Pathology, National Hospital Organization Chibahigashi National Hospital, Chiba, Japan, ⁹Health Administration Center, Niigata University, Niigata, Japan, ¹⁰Department of Internal Medicine, JR Sendai Hospital, Miyagi, Japan, ¹¹Department of Pathology, Faculty of Medicine, Fukuoka University, Fukuoka, Japan, ¹²Department of Pathology, Dokkyo Medical University Saitama Medical Center, Saitama, Japan, ¹³Division of Nephrology and Kidney Center, Kobe University Graduate School of Medicine, Hyogo, Japan, ¹⁴Department of Nephrology, Nagasaki University Hospital, Nagasaki, Japan, ¹⁵Department of Nephrology, Nara Medical University, Nara, Japan, ¹⁶Department of Cardiovascular Medicine, Nephrology and Neurology, University of the Ryukyus School of Medicine, Okinawa, Japan, ¹⁷Department of Internal Medicine, Division of Nephrology, St Marianna University School of Medicine, Kanagawa, Japan, ¹⁸Okayama University, Okayama, Japan and ¹⁹Department of Internal Medicine, Division of Nephrology, Nagoya University Graduate School of Medicine, Nagoya, Japan

Correspondence to
Masayuki Yamanouchi; E-mail: m.yamanouchi@toranomon.gr.jp and Takashi Wada; E-mail: twada@m-kanazawa.jp

Authors' Contributions
M.Y., K.F., J.H. and T.W. designed the study protocol, researched data, contributed to the discussion, wrote the manuscript and reviewed and edited the manuscript. T.T., M.S., Y.Yuzawa, M.O., S.K., A.H., Y.I., N.Sawa, Y.O., S.Matsuo, D.I., H.Makino, T.S., N.Sakai, Y.Yamamura, H.K., Y.Suzuki, H.S., N.U., Y.Ubara, S.N., H.Y., T.N., K.S., K.K., Y.Shibagaki, H.Mizuno, S.Matsuoka and Y.Ueda contributed to the discussion and reviewed the manuscript. T.W. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Conflict of Interest Statement
No potential conflicts of interest relevant to this article were reported.