0021-972X/04/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 89(9):4359 – 4363 Copyright © 2004 by The Endocrine Society doi: 10.1210/jc.2004-0678

Anemia in Patients with Type 1 Diabetes MERLIN C. THOMAS, RICHARD J. MACISAAC, CON TSALAMANDRIS, LYNDA MOLYNEAUX, INNA GOUBINA, GREG FULCHER, DENNIS YUE, AND GEORGE JERUMS Baker Medical Research Institute (M.C.T.), Melbourne, Victoria 8008, Australia; Endocrinology Unit and Department of Medicine (R.J.M., C.T., G.J.), University of Melbourne, Austin Health Medical Centre, Heidelberg, Victoria 3084, Australia; Diabetes Centre (L.M., D.Y.), Royal Prince Alfred Hospital and Discipline of Medicine, University of Sydney, New South Wales 2050, Australia; and Sydney Diabetes (I.G., G.F.), Department of Diabetes, Endocrinology, and Metabolism, Royal North Shore Hospital, St. Leonards, Sydney, New South Wales 2065, Australia Anemia is a common finding in diabetes, although most patients in these studies have type 2 disease. This study examines the prevalence and predictors of anemia in outpatients with type 1 diabetes. A full blood count was obtained in addition to routine testing in patients with type 1 diabetes at the Austin Medical Centre (n ⴝ 135), Melbourne, the Royal Prince Alfred Hospital (n ⴝ 42), and the Royal North Shore Hospital (n ⴝ 135), both in Sydney, Australia. One in seven patients had anemia (14%). Patients at greatest risk could be identified by the presence of diabetic kidney disease. More than half (52%) of patients with macroalbuminuria had anemia, compared with 24% of patients with mi-

I

T HAS BEEN recently recognized that anemia is a common complication of diabetes, particularly in patients with diabetic kidney disease. In a recent cross-sectional survey of patients with diabetes in a single clinic, we found that nearly a quarter of all outpatients had anemia (1). Anemia develops earlier than in patients with renal impairment from other causes (1– 4). In addition, a normochromic, normocytic anemia can also be observed in patients without overt renal disease (4). It is now established that reduced hemoglobin (Hb) levels, even to a limited degree, identify patients at increased risk of progressive renal disease (5, 6). However, the majority of patients in these studies had type 2 diabetes. As a consequence, anemia associated with ageing and/or renal impairment may be overrepresented in these populations. Against this, patients with type 2 diabetes may have a slower rate of progression of diabetic renal disease, and the cumulative incidence of nephropathy may be less than in patients with type 1 diabetes. The consequences of these difference on the prevalence and predictors of anemia in type 1 diabetes have not been established. As a follow-up to our initial study (1), we present a composite survey from three clinical centers to estimate the burden of anemia in outpatients with type 1 diabetes. Subjects and Methods Study participants

This study was designed as a clinical audit of outpatients with type 1 diabetes in current long-term follow-up in tertiary care specialist Abbreviations: ACE, Angiotensin-converting enzyme; AER, albumin excretion rate; CCr, creatinine clearance; Hb, hemoglobin; IHD, ischemic heart disease. JCEM is published monthly by The Endocrine Society (http://www. endo-society.org), the foremost professional society serving the endocrine community.

croalbuminuria and less than 8% of normoalbuminuric patients. Patients with diabetes and renal impairment were more than six times more likely to have anemia than those with normal renal function. Patients with anemia were more likely to have retinopathy and macrovascular complications than were patients with a normal hemoglobin level, independent of comorbid renal disease. Anemia is a prevalent finding in patients with type 1 diabetes and represents a significant unrecognized burden. Patients at greatest risk can be identified by the presence of renal disease, in the form of albuminuria and/or renal impairment. (J Clin Endocrinol Metab 89: 4359 – 4363, 2004)

diabetes clinics. The presence of type 1 diabetes was determined according to World Health Organization (WHO) criteria (7zrefx). Patient data were obtained from three Australian centers: the Austin Medical Centre (AMC) in Melbourne, Australia (n ⫽ 135), the Royal Prince Alfred (RPA) Hospital (n ⫽ 42) and the Royal North Shore (RNS) Hospital (n ⫽ 135), both located in Sydney, Australia. The majority of patients were referred to these clinics by general practitioners requiring assistance with surveillance and management of the long-term complications of diabetes. Approximately 20% of the patients were referred from other sources, including specialty units within each hospital. At the time of this audit, none of these clinics shared patient care with nephrology services. Data collected as part of these cross-sectional surveys were obtained with the approval of respective hospital ethics committees, and they meet National Health and Medical Research Council ethics guidelines.

Determination of variables Standard indices were collected from the most recent routine clinical visit including creatinine, urea, fasting lipid profile, and HbA1c determined by HPLC. Urinary creatinine, urea, and albumin excretion were estimated from a 24-h urine collection. Clinical data including anthropometric measurements, age, gender, body mass index, duration of diabetes, length of follow-up, and the presence or absence of macrovascular disease were obtained from patient records on all patients. In addition, an Hb concentration was obtained for all patients. Results obtained outside the outpatient setting (e.g. patients in an emergency situation or hospitalized) were excluded. Hb was primarily handled both as a continuous variable and recoded as a binary outcome for estimating the prevalence of anemia. The presence of anemia was defined by an Hb level less than 13 g/dl (130 g/liter) in men and less than 12 g/dl (120 g/liter) in women, which is a gender-specific definition established by the WHO (8). The level of albuminuria was defined categorically from the most recent urinalysis according to standard guidelines. Albumin excretion rate (AER), derived from 24-h urinary albumin measurement, was categorically defined from the three most recent AER measurements. Macroalbuminuria was defined as two of three AER measurements more than 200 ␮g/min. Microalbuminuria was defined as two of three AER measurements between 20 –200 ␮g/min. Normoalbuminuria was defined by two of three AER measurements less than 20 ␮g/min. The term

4359

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 00:54 For personal use only. No other uses without permission. . All rights reserved.

4360

J Clin Endocrinol Metab, September 2004, 89(9):4359 – 4363

Thomas et al. • Anemia in Type 1 Diabetes

elevated albuminuria was used to denote patients with either micro- or macroalbuminuria (AER ⬎ 20 ␮g/min). Creatinine clearance (CCr) was determined using the Cockroft-Gault formula expressed per 1.73-m2 of body surface area. For categorical analysis, moderate renal impairment was defined by a CCr of less than 60 ml/min䡠1.73 m2.

Statistical methods Because clinical data were obtained separately, observations pertaining to each population are provided separately in addition to pooled results. Continuous data are expressed as mean ⫾ the sem. Differences in continuous variables were compared using Student’s t tests (two groups) or one-way ANOVA (three or more groups, with subgroups compared using Fisher’s partial least squares post hoc test). Differences in categorical variables were compared using ␹2 analysis. Pearson correlation was used to analyze univariate associations between continuous variables. Logistic regression was used to analyze associations between independent predictors.

Results Outpatient population

The clinical characteristics of each of the outpatient populations are shown in Table 1. In all three centers there were approximately the same number of men and women with a third of patients older than 50 yr. The median duration of diabetes was 20 yr, with most (⬎75%) patients being followed in each clinic for more than 10 yr. The prevalence of elevated albuminuria (micro- or macroalbuminuria) varied between 13–27% in the three centers. Approximately 11% of patients in each center had evidence of moderately impaired renal function (CCr ⬍ 60 ml/min䡠1.73 m2). Overall, over one quarter of the patients in each center had some evidence of kidney disease defined either as moderately impaired renal function and/or elevated albuminuria. Prevalence of anemia

The overall distribution of Hb levels was similar in all three centers (Fig. 1). Hb levels in each population were TABLE 1. Patient characteristics from three outpatient centers

Age (yr) Age ⬎50 yr (%) Gender (male) Median duration of diabetes (yr) Body mass index (kg/m2) Microalbuminuria Macroalbuminuria HbA1c Creatinine clearance (ml/min䡠1.73 m2) Moderate renal impairment Elevated albuminuria and/or renal impairment

AMC (n ⫽ 138)

RPA (n ⫽ 42)

RNS (n ⫽ 135)

45 ⫾ 1 38% 56% 20

42 ⫾ 1 30% 52% 20

43 ⫾ 1 27% 51% 20

26.2 ⫾ 0.4

26.5 ⫾ 0.6

26.0 ⫾ 0.4

19% 8% 8.3 ⫾ 0.2% 102 ⫾ 3

11% 11% 8.2 ⫾ 0.3% 101 ⫾ 3

8% 5% 7.4 ⫾ 0.2% 102 ⫾ 2

11%

14%

9%

29%

30%

18%

Data expressed as mean ⫾ SEM unless otherwise indicated. Prevalence data are expressed as a percentage of the center population unless otherwise indicated. Moderate renal impairment was defined by a CCr ⬍ 60 ml/min䡠1.73 m2. Elevated albuminuria was defined by an AER ⬎ 20 ␮g/min (i.e. micro- and macroalbuminuria). AMC, Austin Medical Center; RPA, Royal Prince Alfred Hospital; RNS, Royal North Shore Hospital.

normally distributed with only a small degree of skewness seen at the lower end of the Hb range. Approximately one in seven patients (n ⫽ 45 of 314) had anemia as defined by WHO (gender-specific) guidelines, including 15% of female and 13% of male patients. This prevalence rate was not significantly different among each of the three centers. Approximately 5% of men and 8% of women had Hb levels less than 11 g/dl (110 g/liter) denoting the threshold for intervention in patients with renal disease (9). The effect of diabetic kidney disease

On multivariate logistic regression analysis, the presence of elevated albuminuria was the strongest independent predictor of anemia. More than half of all patients with macroalbuminuria (⬎200 ␮g/min) had WHO-defined anemia (52%), compared with 24% of patients with microalbuminuria (20 –200 ␮g/min) and less than 8% of patients with normoalbuminuria (AER ⬍ 20 ␮g/d). This trend was observed even after adjusting for the lower CCr seen in patients with greater levels of albuminuria (Fig. 2). Renal impairment was also independently associated with the prevalence of anemia in patients with type 1 diabetes. More than half (56%) of all patients with anemia had moderate renal impairment (CCr ⬍ 60 ml/min䡠1.73 m2), compared with less than 10% of patients with an Hb level within the normal range. In addition, patients with type 1 diabetes and moderate renal impairment were more than 5 times as likely to have anemia as patients with normal renal function (CCr ⬎ 90 ml/min䡠1.73 m2). However, the relationship between CCr and anemia was not confined to patients with moderate renal impairment. A significant association between CCr and the prevalence of anemia (P ⬍ 0.001) was also seen in patients with a serum creatinine level within the normal range [in women, ⬍1.1 mg/dl (100 ␮mol/liter) and in men, ⬍1.5 mg/dl (130 ␮mol/liter). Overall, 69% of all patients with anemia (n ⫽ 31 of 45) had either moderate renal impairment and/or elevated albuminuria. Of the remaining anemia patients without diabetic kidney disease (n ⫽ 14), most were premenopausal women (n ⫽ 11; 80%). The effect of other variables

In patients with type 1 diabetes, the Hb level was approximately 1 g/dl (10 g/liter) lower in women than in men (Fig. 1); this difference was incorporated into WHO guidelines for the diagnosis of anemia, meaning that gender is eliminated as a predictor of anemia, although it remained a strong determinant of Hb levels (9). In addition, anemia was not associated with age, duration of diabetes, or body mass index in patients with type 1 diabetes. Although anemia has been previously associated with a spurious reduction in HbA1c levels (10), there was no association between HbA1c and Hb levels in patients with type 1 diabetes. Anemia and vascular complications of diabetes

Patients with type 1 diabetes and anemia were more than twice as likely to have established macrovascular disease (25%) compared with patients without anemia (12%). This

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 00:54 For personal use only. No other uses without permission. . All rights reserved.

Thomas et al. • Anemia in Type 1 Diabetes

J Clin Endocrinol Metab, September 2004, 89(9):4359 – 4363 4361

FIG. 1. Distribution of Hb concentrations in patients with diabetes from three centers, stratified for gender (female, unfilled symbols; male, filled symbols).

FIG. 2. The prevalence of anemia stratified according to the levels of CCr and albuminuria.

was largely determined by the increased risk of ischemic heart disease (IHD) in anemic patients. This effect was independent of the presence of renal disease, as twice as many patients with anemia had IHD with or without renal impairment (Fig. 3, top). Similarly, twice as many patients with anemia had IHD with or without elevated albuminuria (Fig. 3, bottom). Although the rates of peripheral and cerebrovascular disease were increased in patients with anemia, the small numbers of patients meant that this did not reach significance. However, patients with anemia were also twice as likely to have proliferative retinopathy (20%) compared with those without anemia (10%; P ⬍ 0.05). Discussion

Anemia represents a significant and underrecognized burden in patients with type 1 diabetes. We have previously shown in a cross-sectional survey of all patients with diabetes at the AMC (including ⬃20% with type 1 diabetes), that one in five patients have Hb levels below the gender-specific normal range (1). However, these results were driven to a large extent by elderly patients with type 2 diabetes and renal

impairment. To better characterize the role of diabetes in the pathogenesis of anemia, the data presented in this article examine only patients with type 1 diabetes and combine results with two additional similar tertiary referral centers. Notably, we found that 15% of female and 13% of male patients with type 1 diabetes have anemia. In comparison, the prevalence of anemia in a Caucasian population of a similar age is less than 2% in men and less than 6% in women (11). Our survey is limited in that it has not attempted to define the (the main cause of anemia in the general population) is uncommon in patients with diabetes at the AMC and more so in patients with type 1 diabetes (12). In addition, we have not excluded blood loss, which may be associated with autoimmune gastritis in type 1 diabetes (13). Nonetheless, the fact that patients at greatest risk could be identified by the presence of renal disease, manifested as impaired renal function and/or elevated albuminuria, suggests that the predominant cause of anemia may be renal in origin. Fifty-two percent of patients with macroalbuminuria had anemia, compared with 28% of patients with microalbuminuria and less than 8% of normoalbuminuric patients. These results are

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 00:54 For personal use only. No other uses without permission. . All rights reserved.

4362

J Clin Endocrinol Metab, September 2004, 89(9):4359 – 4363

Thomas et al. • Anemia in Type 1 Diabetes

FIG. 3. The prevalence of IHD in patients with anemia, stratified for the level of renal function (top) and AER (bottom).

consistent with a recent small survey that demonstrated that 13 of 27 patients with type 1 diabetes and persistent proteinuria were anemic. Notably, this survey also demonstrated that these patients had inappropriately low circulating erythropoietin levels, again implying a renal etiology to this disorder. However, patients with diabetes are still able to mount an appropriate response to acute hypoxia (14), suggesting that the renal cells that produce erythropoietin are not simply lost in the process of interstitial damage that characterize diabetic renal disease. It seems likely that the anemia-sensing (rather than secretory) mechanisms are dysfunctional in the anemia of diabetes. Although the mechanism of this remains to be established, it is conceivable that thickening of the endothelial basement membrane and changes in regional blood flow mediated through up-regulation of the local renin angiotensin system may directly contribute to anemia. It has also been suggested that autonomic degeneration as a result of diabetes may diminish erythropoietin release (15). Because autonomic neuropathy is closely correlated with renal injury, it is difficult to assess its independent influence. However, denervated kidneys used for transplantation appear to release erythropoietin normally (16). This study also demonstrates that renal function is a key predictor of Hb levels in patients with type 1 diabetes. As renal function falls, the prevalence of anemia increases exponentially in all three centers. By the time that patients developed significant renal impairment (CCr ⬍ 60 ml/ min䡠1.73 m2), more than half of all patients had anemia. This compares with less than 15% of non-Hispanic Americans in the general population with a similar degree of renal im-

pairment (16), again consistent with the finding that the rate of anemia in patients with diabetes are at least double that seen in patients with renal disease of other causes (3). It is important to note that many patients with renal impairment in our survey had a serum creatinine in the normal range, particularly in women in whom reduced muscle mass can lead to pseudonormalization of serum creatinine levels. Notably, in patients with a normal serum creatinine, CCr was still associated with the prevalence of anemia (P ⬍ 0.01 in all centers). This finding underlines the importance of an estimated CCr or glomerular filtration rate in the management of patients with diabetes. In addition, the finding of anemia in a patient with diabetes should direct the treating physician to investigate the possibility of unrecognized impairment. Patients with type 2 diabetes and an Hb level less than 12 g/dl (120 g/liter) are said to have a 2-fold increased risk of background diabetic retinopathy and a 5-fold increased risk of preproliferative or proliferative retinopathy, compared with those with higher Hb levels (17). Similarly, we find that the rate of vision-threatening retinopathy is significantly increased in patients with anemia. This is thought to be due to retinal hypoxia leading to the up-regulation of growth factor such as vascular endothelial growth factor and other genes involved in neoangiogenesis, capillary permeability, and apoptosis (18). Furthermore, in this survey, patients with type 1 diabetes and anemia were more than twice as likely to have IHD, irrespective of renal impairment or albuminuria. Notably, anemia in these patients is associated with increased rates of hospitalization and premature mortality (5, 19). Although anemia is clearly associated with both micro-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 00:54 For personal use only. No other uses without permission. . All rights reserved.

Thomas et al. • Anemia in Type 1 Diabetes

and macrovascular complications in patients with type 1 diabetes, it remains to be established what role anemia may have in the development or progression of these complications. Because there is a direct relationship between anemia and diabetic kidney disease, anemia such as albuminuria may be a marker of more potent microvascular disease rather than being directly pathogenic. Indeed, a number of studies (6), including the recent Reduction on Endpoints in NIDDM with Angiotensin II Antagonist Losartan (RENAAL) trial (5), have suggested that reduced Hb levels, even within the normal range, identify patients with type 2 diabetes at increased risk for progressive renal disease. However, there is good reason to suggest that anemia may play a direct role in this process. Certainly, anemia has direct mitogenic and fibrogenic effects on the kidney and the heart, associated with expression of growth factors, hormones, and vasoactive reagents, many of which are also implicated in the diabetic microvascular disease (20). Anemia is also correlated with oxidative stress, because erythrocytes represent a major antioxidant component of the blood (21). It is yet to be established that correction of anemia in patients with type 1 diabetes will have any benefit. Tiredness, a common complaint in many patients with diabetes, may significantly improve after correction of anemia. Similarly, cognitive function, sexual function, and reduced tolerance for exercise and ability to work may all be favorably influenced after correction of anemia (22). Treatment of erythropoietin in diabetic patients with heart failure may also have beneficial effects on end organ function (23). In a few small clinical studies, supplementation with erythropoietin in patients with diabetes has been associated with a reduction in macular hard exudates and edema (24). However, such benefits must be carefully balanced against the significant financial cost involved in treating these patients, as well as the potential for deleterious effects of erythropoietin including raised blood pressure and pure red cell aplasia. Correcting anemia by other means such as transfusion also carries associated risks, including human leukocyte antigen sensitization that may inadvertently render a young patient unable to be subsequently transplanted. It is hoped that upcoming trials will help clarify the potential role of correction of anemia in patients with diabetes. In the meanwhile, this survey should encourage heightened awareness of the unrecognized burden of anemia in patients with type 1 diabetes. Acknowledgments Received April 8, 2004. Accepted May 25, 2004. Address all correspondence and requests for reprints to: Dr. Merlin C. Thomas, Baker Medical Research Institute, P. O. Box 6492, Melbourne, Victoria 8008, Australia. E-mail: [email protected]. M.C.T. was supported by an Australian National Health and Medical Research Council (NHMRC) biomedical scholarship. This material was presented as part of the 2003 International Society of Nephrology satellite meeting (Diabetes and the Kidney) in Dusseldorf, Germany. Results pertaining to the prevalence and predictors of anemia in the AMC cohort alone have been previously published as part of a smaller cross-sectional survey of patients with both type 1 and type 2 diabetes in that center (Thomas MC, MacIsaac RJ, Tsalamandris C, Power D,

J Clin Endocrinol Metab, September 2004, 89(9):4359 – 4363 4363

Jerums G. Unrecognized anemia in patients with diabetes. Diabetes Care 2002;24:1164 –1169). All authors attest no conflict of interest in the submission of this manuscript.

References 1. Thomas MC, MacIsaac RJ, Tsalamandris C, Power D, Jerums G 2002 Unrecognized anemia in patients with diabetes. Diabetes Care 24:1164 –1169 2. Dikow R, Schwenger V, Scho¨mig M, Ritz E 2001 How should we manage anaemia in patients with diabetes? Nephrol Dial Transplant 17(Suppl 1):67–72 3. Astor BC, Muntner P, Levin A, Eustace JA, Coresh J 2002 Association of kidney function with anemia: The Third National Health and Nutrition Examination Survey (1988 –1994). Arch Intern Med 162:1401–1408 4. Bosman DR, Winker AS, Marsden JT, MacDougall IC, Watkins PJ 2002 Anemia with erythropoietin deficiency occurs early in diabetic nephropathy. Diabetes Care 24:495– 499 5. Keane WF, Lyle PA 2003 Recent advances in management of type 2 diabetes and nephropathy: lessons from the RENAAL study. Am J Kidney Dis 41(3 Suppl 1):S22–S25 6. Ueda H, Ishimura E, Shoji T, Emoto M, Morioka T, Matsumoko N, Fukumoto S, Miki T, Inaba M, Nishizawa Y 2003 Factors affecting progression of renal failures in patients with type 2 diabetes. Diabetes Care 26:1530 –1534 7. World Health Organization 1985 World Health Organization technical report series, no. 727. Diabetes mellitus: report of a WHO Study Group. Geneva: World Health Organization 8. World Health Organization 1968 Nutritional anemia: report of a World Health Organization Scientific Group. Geneva: World Health Organization 9. 2001 IV. NKF-K/DOQI Clinical Practice Guidelines for Anemia of Chronic Kidney Disease, 2000. Am J Kidney Dis [Erratum (2001) 38:442] 37(1 Suppl 1):S182–S238 10. Tarim O, Kucukerdogan A, Gunay U, Eralp O, Ercan I 1999 Effects of iron deficiency anemia on hemoglobin A1c in type 1 diabetes mellitus. Pediatr Int 41:357–362 11. Bennett SA, Magnus P 1994 Trends in cardiovascular risk factors in Australia. Results from the National Heart Foundation’s Risk Factor Prevalence Study, 1980 –1989. Med J Aust 161:519 –527 12. Thomas MC, MacIsaac R, Tsalamandris C, Jerums G 2004 Elevated iron indices in patients with diabetes. Diabet Med 21:798 – 802 13. De Block CEM, van CampenHout CM, De Leeuw IH, Keenoy BM, Martin M, Van Hoof V, Van Gaal LF 2000 Soluble transferrin receptor level: a new marker for iron deficiency anemia, common manifestation of gastric autoimmunity in type 1 diabetes. Diabetes Care 23:1384 –1388 14. Bosman DR, Osborne CA, Marsden JT, Macdougall IC, Gardner WN, Watkins PJ 2002 Erythropoietin response to hypoxia in patients with diabetic autonomic neuropathy and non-diabetic chronic renal failure. Diabet Med 19:65– 69 15. Jeffrey RF, Kendall RG, Prabhu P, Norfolk DR, Will EJ, Davison AM 1995 Re-establishment of erythropoietin responsiveness in end-stage renal failure following renal transplantation. Clin Nephrol 44:241–247 16. Hsu CY, McCulloch CE, Curhan GC 2002 Epidemiology of anemia associated with chronic renal insufficiency among adults in the United States: results from the Third National Health and Nutrition Examination Survey. J Am Soc Nephrol 13:504 –510 17. Qiao Q, Keinanen-Kiukaanniemi S, Laara E 1997 The relationship between hemoglobin levels and diabetic retinopathy. J Clin Epidemiol 50:153–158 18. Grimm C, Wenzel A, Groszer M, Mayser H, Seeliger M, Samardzija M, Bauer C, Gassmann M, Reme CE 2002 HIF-1-induced erythropoietin in the hypoxic retina protects against light-induced retinal degeneration. Nat Med 8:718 –724 19. McClellan WM, Flanders WD, Langston RD, Jurkovitz C, Presley R 2002 Anemia and renal insufficiency are independent risk factors for death among patients with congestive heart failure admitted to community hospitals: a population-based study. J Am Soc Nephrol 13:1928 –1936 20. Fine LG, Bandyopadhay D, Norman JT 2000 Is there a common mechanism for the progression of different types of renal diseases other than proteinuria? Towards the unifying theme of chronic hypoxia. Kidney Int 57(Suppl 75):S22–S26 21. Grune T, Sommerburg O, Siems WG 2000 Oxidative stress in anemia. Clin Nephrol 53(Suppl 1):S18 –S22 22. Cody J, Daly C, Campbell M, Donaldson C, Grant A, Khan I 2001 Recombinant human erythropoietin for chronic renal failure anemia in pre-dialysis patients. Cochrane Database Syst Rev 4:CD003 23. Silverberg DS, Wexler D, Blum M, Tchebiner JZ, Sheps D, Keren G, Schwartz D, Baruch R, Yachnin T, Shaked M, Schwartz I, Steinbruch S, Iaina A 2003 The effect of correction of anaemia in diabetics and non-diabetics with severe resistant congestive heart failure and chronic renal failure by subcutaneous erythropoietin and intravenous iron. Nephrol Dial Transplant 18:141–146 24. Friedman EA, L’Esperance FA, Brown CD, Berman DH 2003 Treating azotemia-induced anemia with erythropoietin improves diabetic eye disease. Kidney Int Suppl 87:S57–S63

JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 January 2017. at 00:54 For personal use only. No other uses without permission. . All rights reserved.