Nigerian Journal of Medicine

: 2021  |  Volume : 30  |  Issue : 5  |  Page : 494--500

Predictors of anemia in Type 2 diabetes mellitus without renal disease

Ernest Yorke1, Nana Ama Barnes1, Josephine Akpalu1, Eunice Boafo2, Albert G B Amoah1,  
1 Department of Medicine and Therapeutics, School of Medicine and Dentistry, College of Health Sciences, University of Ghana, Accra, Ghana
2 National Diabetes Management and Research and Center, Korle-Bu Teaching Hospital, Accra, Ghana

Correspondence Address:
Dr. Ernest Yorke
Department of Medicine and Therapeutics, School of Medicine and Dentistry, University of Ghana, P.O. Box 4236, Accra


Background: Among type 2 diabetes mellitus (T2DM) patients without renal disease, hemoglobin levels tend to be lower than their counterparts without diabetes mellitus with a similar estimated glomerular filtration rate. Low hemoglobin levels have been associated with increased morbidity. Objective: We sought to determine the predictors of anemia among T2DM patients without renal disease attending the diabetes center at a tertiary teaching hospital in Accra, Ghana. Materials and Methods: One hundred and ninety-five type 2 diabetes patients aged 35 years and above without overt nephropathy and 184 controls without diabetes were recruited from the National Diabetes Management and Research Center and Orthopedic Outpatients Clinic of the KBTH, respectively. Both groups had full blood count, red cell indices, serum creatinine, and glycated hemoglobin determination as well as medical history and anthropometric measurements. Results: Seventy-nine percent of cases and 75.5% controls were females. The prevalence of anemia among participants with diabetes mellitus was 53.1% compared with 20.5% among controls (P = 0.00). In both groups, more males than females were anemic (68.3% cases, 51.1% controls). The mean duration of diabetes was 9.6 ± 6.22 years among the cases. Logistic regression revealed male gender and diabetes duration >10 years as significant predictors of anemia without renal disease in individuals with T2DM. Conclusion: The prevalence of anemia among T2DM participants without overt renal disease was two and half times the rate in control participants without diabetes mellitus. Physicians would be aware and take appropriate steps to identify and manage it to reduce associated morbidity.

How to cite this article:
Yorke E, Barnes NA, Akpalu J, Boafo E, Amoah AG. Predictors of anemia in Type 2 diabetes mellitus without renal disease.Niger J Med 2021;30:494-500

How to cite this URL:
Yorke E, Barnes NA, Akpalu J, Boafo E, Amoah AG. Predictors of anemia in Type 2 diabetes mellitus without renal disease. Niger J Med [serial online] 2021 [cited 2021 Nov 29 ];30:494-500
Available from:

Full Text


Globally, diabetic kidney disease occurs in approximately one-third of all persons affected with diabetes mellitus.[1],[2]

Among those with diabetes mellitus, compared with persons without diabetes with a similar estimated glomerular filtration rate (eGFR) (normal or impaired), anemia occurs earlier and is more severe.[3],[4],[5],[6],[7],[8]

In the presence of nephropathy, erythropoietin (Epo) deficiency is thought to be the underlying mechanism for anemia,[8],[9],[10] whereas Epo resistance is responsible for the anemia in the absence of nephropathy.[11],[12],[13]

The effects of anemia include increased risk and severity of retinopathy, especially when hemoglobin (Hb) is <12 g%,[14] development of left ventricular hypertrophy and heart failure,[15],[16],[17] impaired wound healing,[18] and diminished general well-being.[19]

Very few studies to date have been done on anemia in the absence of renal disease,[20] among subjects with diabetes mellitus, and data on the prevalence and risk factors of anemia in diabetes mellitus in Ghana and sub-Saharan Africa as a whole are scanty.

We sought to determine the predictors of anemia among type 2 diabetes mellitus (T2DM) patients without renal disease who attended clinic at the National Diabetes Management and Research Center (NDMRC), Korle-Bu Teaching Hospital in Accra.

 Materials and Methods

Study design

The study design was a cross-sectional, hospital-based study.

Study site/population

After written informed consent, persons aged 35 years and above with type 2 diabetes attending the clinic at the NDMRC, Korle-Bu Teaching Hospital were recruited as cases by the systematic random sampling.

The NDMRC is a national resource center for diabetes care, training, and research and also offers outpatient services.

We excluded participants with type 1 diabetes mellitus (young patients who have been exclusively treated with insulin from diagnosis), known hemoglobinopathies or sickle cell anemia, patients with bleeding disorders or bleeding tendencies, patients on blood boosters or hematinics, known malignancy, participants with proteinuria (macroalbuminuria) and or eGFR <89 ml/min/1.73 m2,[21] very ill patients, and women participants who were pregnant.

We recruited individuals without diabetes (known or after screening) from the orthopedic clinic as controls, and the inclusion and exclusion criteria were similar to those for the cases.

The minimum sample size (N) calculated was 139 after an assumed prevalence (P) of 10% for anemia in diabetes without renal disease,[20] Z score at 95% confidence level (1.96) and level of significance of 0.05 (d). Ten percent nonrespondent rate was assumed, and a minimum of 153 each of cases and controls age-matched were targeted to be recruited. After screening 400 individuals in each group, we recruited 195 cases and 184 controls through systematic random after the application of the inclusion and exclusion criteria.

The data abstraction tool (questionnaire) was pretested and administered by trained research assistants who also did the body measurements.

Data collection

Data on gender, age, occupation, employment status, educational status, smoking history and alcohol use as well as a personal history of diabetes mellitus, hypertension, dyslipidemia, bleeding disorders, known malignancies, etc., were obtained.

The weight of the participants who wore only light-clothing, without shoes or hair ornaments, was measured (nearest 0.1 kg) using a calibrated electronic heavy duty Seca 770-floor digital scale (Seca Hamburg, Germany), whereas height was measured using a height measure (nearest 0.1 cm).

The weight in kilograms divided by the square of height in meters was used to determine the body mass index (BMI).

With the clothes removed, the waist circumference was measured using a measuring tape (non-elastic). Duplicate measures were taken to the nearest 0.1 cm at the end of expiration midpoint between the lower rib margin and the iliac crest. The mean of the two was used. Similarly, hip circumference was measured at the maximal circumference of the buttocks, horizontally all-around the body to the nearest 0.1 cm.

On two occasions, blood pressure was measured at an interval of 1 min with participants seated using a professional HEM 907 blood pressure monitor after at least a 5–10 min rest. The mean of the two measurements was used.

The sickling status of patients was determined (using sodium metabisulfite test) and persons found to be sickling positive were subjected to hemoglobin electrophoresis (using ion-exchange chromatography). Spot early morning midstream urine sample was obtained, and proteinuria was assessed using a dipstick.

Diabetes was diagnosed by the World Health Organization criteria of fasting plasma glucose ≥7.0 mmol/l (126 mg/dl) and/or 2-h plasma glucose 11.1 mmol/l (200 mg/dl) or regular medication for diabetes.[22]

Cutoff levels of anemia were defined by hemoglobin of <12 g/dl for women and <13 g/dl for men.[23]

We used the simplified modification of diet in renal disease formula (186.3/[serum creatinine, mg/dL−1.154]/[age, years− 0.203]) was used to calculate the eGFR; calculated values were multiplied by. 742 for women, and by 1.21 as for African Americans and normal values defined as >89 mL/min/1.73 m2.[24]

Controlled diabetes was defined as glycated hemoglobin level of 7% (53 mmol/mol) or less and also preprandial plasma glucose 3.9–7.2 mmol/l (70–130 mg/dl).[25] Obesity, overweight, normal, and underweight BMIs were defined as 30.0 or more, 25–30, 25–18.5 and <18.5, respectively[26] whereas waist circumference of 94 cm or more in men and 80 cm or more in women was considered increased.[27]

Waist-hip ratio (WHR) >0.9 for men and >0.85 for women was indicated as increased WHR.[26] For the analysis, the duration of diabetes was grouped into 0–5, 5–10, and >10 years after the diagnosis.

Statistical analysis

The analysis of the data from the abstraction form was done using the Statistical Package for the Social Sciences software version 20, IBM SPSS Statistics, New York, USA.

Descriptive statistics for the outcome measures of interest including the prevalence as well as the frequency of anemia among age and sex-specific groups were also calculated both among cases and controls.

Inferential statistics chi-square were used to determine the significant associations between anemia and age, sex, duration of diabetes, HBA1C, fasting plasma glucose, WHR, BMI, and employment status in both cases and controls at 95% significant level.

Logistic regression analysis was performed to determine the significant predictors of anemia in diabetics (at the 95% significant level). The independent variables used included age, gender, duration of diabetes, HBA1C, WHR, and BMI.

All P values for this work were two-tailed with P < 0.05 as significant.

Ethical approval

All patients provided written informed consent, with those unable to understand or sign the informed consent excluded. The study presented minimal risk to participants and approval was sought from the University of Ghana Medical School Ethics and Protocol Review Committee before the study was commenced. The Protocol Identification Number was MS-Et/M.4-P. 4.7/2009/10. All participants also provided consent for the publication.

It complied with the Helsinki Declaration of 1975 on human experimentation (Revised 2020).


A total of 379 participants participated in the study, of which 195 (51.5%) were cases and 184 (48.5%) were controls. Among the cases, 41 (21%) were male and 154 (79%) were female, whereas 45 (24.5%) were male and 139 (75.5%) were female among controls.

The mean duration of diabetes among the cases was 9.64 ± 6.22 years with a mean HBA1c of 8.0% ±1.84 (64 mmol/mol). Mean fasting blood sugar was 8.72 ± 3.35 and 5.99 ± 3.40 mmol/l among cases and controls, respectively (P < 0.05).

Baseline characteristics of study participants are shown in [Table 1].{Table 1}

Medications and other comorbid medical conditions

Most of the diabetes participants (72.8%) used metformin-based medications (metformin alone or in combination with other medications). Ninety out of 195 diabetes participants (46.1%) used a metformin/sulphonylurea combination, 22 (11.3%) cases were on metformin alone whiles those on sulphonylurea alone were 15 (7.7%). The percentage of respondents with anemia was generally higher than those with normal hemoglobin among those on metformin-containing medications.

Antihypertensives represented the most widely used medication among both cases (130 out of 195) and controls (95 out of 184), that is, 66.7% versus 51.6%; followed by soluble aspirin and then statins.

Only one person smoked among the cases while two controls were smokers. None of the respondents had cancer. While 11 controls and 8 cases were sickling positive, only one control had hemoglobin SS on electrophoresis who was excluded from the analysis.

Prevalence of anemia

Data for 3 cases and 2 controls with respect to anemia were incomplete and were excluded from the analyses of anemia, One-hundred and two out of 192 (53.1%) cases compared with 37 out of 182 (20.3%) controls had anemia (P = 0.000). A higher proportion of males than females within both cases (68.3%) and controls (51.1%) groups were diagnosed with anemia (P = 0.028 vs. 0.000) [Table 2].{Table 2}

Red cell indices of participants with anemia: Hemoglobin, mean cell volume, and mean corpuscular hemoglobin

The mean hemoglobin (Hb) concentration was 11.08 ± 0.98 and 11.99 ± 0.79 (P = 0.000) for cases and controls, respectively. Mean MCV and MCH for both cases and controls were within normal limits; however, in terms of proportions, 63% and 60% were normocytic for cases and controls whereas 68.6% and 59.5% of cases and controls were normochromic, respectively. Only, 1 case had macrocytic anemia [Table 3].{Table 3}

Anemia and risk factors

Patients with uncontrolled diabetes (using fasting blood glucose and HBA1c), males, overweight/obese individuals, high waist circumference, long duration of diabetes (>10 years), and those who were employed had a numerically higher proportion of anemia. However, risk factors with significant association with anemia were only diabetes duration (P = 0.048) and gender (male) (P = 0.028) [Table 4].{Table 4}

After stepwise logistic regression, male gender and duration of diabetes > 10 years emerged as significant predictors for the development of anemia. The likelihood of developing anemia in a male compared to a female is 1.5 times more (odds ratio [OR] =1.5). Having diabetes for more than 10 years confers two times the increased risk (OR = 2.3) of developing anemia compared with those with diabetes <5 years. Poor glycemic control appeared to be protective in the development of anemia (P = 0.033; OR = 0.08) [Table 5].{Table 5}


Among the cases, the combined effect of a longer mean duration of diabetes, poor control of diabetes, high mean fasting blood glucose coupled with a high prevalence of hypertension and dyslipidemia portends bad prognosis in terms of microvascular and macrovascular complications. This is unsurprising since these are well-known risk factors for both microvascular and macrovascular complications of diabetes.[25],[28]

Risk factors for anemia

Our study revealed that a significantly higher proportion of cases with anemia compared to controls. Very few studies have looked at anemia without renal disease. In a primary care clinic in Israel, Goldhaber et al. found anemia in 32% of unselected patients with diabetes but normal serum creatinine levels.[29] Salah et al. in a Cairo pediatric diabetic clinic found over a two months screening period for anemia and other complications among 200 types 1 patients that 35% had anemia, and none of the screened patients had renal failure.[30]

In a large cross-sectional survey of 17 general practices in the United Kingdom by New et al. in 2007,[31] compared to the general population, 5072 identified diabetics were more likely to have Hb <11 g/dl (5.9% vs. 1.4%, P < 001).

Craig et al. also showed in the United Kingdom when they followed 62 patients with diabetes without nephropathy (no proteinuria and normal renal function) over a mean period of seven years that 10 patients (16.1%) developed anemia.[20] Fifty-eight percent of participants had continued deterioration in Hb since presentation though there was no iron deficiency anemia demonstrated. Further tests revealed the expected increase in Epo production in response to low Hb but without the expected reticulocyte count rise.

Because our study did not measure serum iron, ferritin, transferrin, folate, and Vitamin B12 levels, it would be difficult to truly ascribe the level of anemia found in this study solely to the effect of diabetes. However, our findings showed that among those on metformin-containing medications, a higher proportion had anemia generally. Vitamin B-12 malabsorption may result from long-term metformin use thus increasing the risk of developing Vitamin B-12 deficiency.[32],[33] In addition, although the mechanism unpinning it is not clear, folate concentration may also reduce from metformin treatment.[34] Finally, Vitamin B-12 and folate deficiencies may cause increases in homocysteine concentrations, which is an emerging cardiovascular risk factor.[35],[36],[37] In a multicenter randomized placebo controlled trial involving 380 patients (followed over 52 months),[38] metformin treatment compared with placebo, was associated with a significant mean decrease in Vitamin B-12 concentration. It was found out that a decrease in Vitamin B12 concentration grows over time.

Thiazolidinediones as a class appear to cause small decreases in hemoglobin/hematocrit as well. The observed reduction in Hb is thought to be due to bone marrow suppression rather than an increase in total body water.[39]

Furthermore, angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs), by causing renin-angiotensin system blockade may reduce EPO levels[40] and possibly cause anemia. Jones et al. showed that people with diabetes and anemia are more likely to be taking ACE inhibitors or ARBs;[41] however, a direct role for renin-angiotensin system blockade in the cause of anemia in patients with diabetes is unknown at this present time.[41],[42]

The higher proportion of anemia in males than females (68.3% vs. 51.1%, P = 0.028) agreed with the finding of Craig et al. that 17.8% males as compared with 11.8% females with diabetes without renal disease developed anemia when followed over seven years.[20]

Logistic regression revealed male gender and long duration (>10 years) of diabetes as significant predictors of anemia in persons with diabetes without renal disease.

The pathophysiological processes underlying the development of anemia are thought to be due to overt chronic inflammation leading to Epo resistance and therefore anemia.[11],[12],[13] Consequently, a longer duration of diabetes will afford the chronic inflammatory process to have an effect and subsequently influence the development of anemia in the absence of renal disease.

It must be said that in Africa, iron deficiency anemia, sickle cell anemia, chronic infections such as HIV, and tuberculosis remain the most important causes of anemia and this must be taken into consideration in evaluating persons for anemia.[43],[44],[45],[46]

The protective nature of poor glycemic control was a surprise finding (OR = 0.83). In general, poor glycemic control increases the risk of nephropathy.[25] It may well be that in the absence of nephropathy, glycemic control is not important in the development of anemia among type 2 diabetes patients.


This study has highlighted the high burden of anemia among a subset of T2DM without renal disease. It will help improve the awareness so that both patients and physicians would take appropriate steps to identify and manage it to reduce associated morbidity.


Parameters such as serum iron, transferrin saturation and ferritin (for iron deficiency) and red cell folate, and Vitamin B12 (for megaloblastic anemia) among anemic participants could not be measured due to inadequate funds. Urine microalbumin levels to completely exclude subjects with renal disease could also not be done for similar reasons. We excluded type 1 diabetes mellitus patients on clinical grounds by excluding young patients < 35 years who have been exclusively treated with insulin from diagnosis. This approach we believe will limit the possibility of falsely including type 1 patients without doing autoantibody testing. Notwithstanding, we believe our findings provide important information and clues for larger studies in future.


We wish to thank all the staff of the NDMRC and the Diabetes Research and Chronic Disease Reference Laboratory, University of Ghana Medical School, for their assistance. Furthermore, we are grateful to all the patients at the NDMRC and Orthopedic outpatient department who consented to take part in this study. We thank ROCHE pharmaceuticals for the part study grant.

Financial support and sponsorship

Research was partly funded by Roche Pharmaceuticals and by the authors.

Conflicts of interest

There are no conflicts of interest.


1Rahelić D. 7th edition of IDF diabetes atlas – Call for immediate action. Lijec Vjesn 2016;138:57-8.
2World Health O, World Health O: Diabetes Fact Sheet No. 312. Available from: http://wwwwhoint/mediacentre/factsheets/fs312/en/indexhtml. [Last accessed on 2018 Mar 01].
3Thomas MC, MacIsaac RJ, Tsalamandris C, Power D, Jerums G. Unrecognized anemia in patients with diabetes: A cross-sectional survey. Diabetes Care 2003;26:1164-9.
4Dikow R, Schwenger V, Schömig M, Ritz E. How should we manage anaemia in patients with diabetes? Nephrol Dial Transplant 2002;17 Suppl 1:67-72.
5Bosman DR, Winkler AS, Marsden JT, Macdougall IC, Watkins PJ. Anemia with erythropoietin deficiency occurs early in diabetic nephropathy. Diabetes Care 2001;24:495-9.
6Inomata S, Itoh M, Imai H, Sato T. Serum levels of erythropoietin as a novel marker reflecting the severity of diabetic nephropathy. Nephron 1997;75:426-30.
7Ishimura E, Nishizawa Y, Okuno S, Matsumoto N, Emoto M, Inaba M, et al. Diabetes mellitus increases the severity of anemia in non-dialyzed patients with renal failure. J Nephrol 1998;11:83-6.
8Kazmi WH, Kausz AT, Khan S, Abichandani R, Ruthazer R, Obrador GT, et al. Anemia: An early complication of chronic renal insufficiency. Am J Kidney Dis 2001;38:803-12.
9Winkler AS, Marsden J, Chaudhuri KR, Hambley H, Watkins PJ. Erythropoietin depletion and anaemia in diabetes mellitus. Diabet Med 1999;16:813-9.
10Rarick MU, Espina BM, Colley DT, Chrusoskie A, Gandara S, Feinstein DI. Treatment of a unique anemia in patients with IDDM with epoetin alfa. Diabetes Care 1998;21:423-6.
11Goicoechea M, Martin J, de Sequera P, Quiroga JA, Ortiz A, Carreño V, et al. Role of cytokines in the response to erythropoietin in hemodialysis patients. Kidney Int 1998;54:1337-43.
12Ricerca BM, Caduto S, Cotroneo P, Damiani P, Manto A, Pitocco D, et al. Blunted erythropoietin response to anemia in type 1 diabetic patients disease. Blood 1992;80:1639-47.
13Means RT Jr., Krantz SB. Progress in understanding the pathogenesis of the anemia of chronic disease. Blood 1992;80:1639-47.
14Qiao Q, Keinanen-Kiukaanniemi S, Laara E. The relationship between hemoglobin levels and diabetic retinopathy. J Clin Epidemiol 1997;50:153-8.
15Wexler D, Silverberg D, Blum M, Sheps D, Keren G, Wollman Y, et al. Anaemia as a contributor to morbidity and mortality in congestive heart failure. Nephrol Dial Transplant 2005;20 Suppl 7:i11-5.
16Levin A, Thompson CR, Ethier J, Carlisle EJ, Tobe S, Mendelssohn D, et al. Left ventricular mass index increase in early renal disease: Impact of decline in hemoglobin. Am J Kidney Dis 1999;34:125-34.
17Sandgren PE, Murray AM, Herzog CA, Solid CA, Gilbertson DT, Collins AJ, et al. Anemia and new-onset congestive heart failure in the general Medicare population. J Card Fail 2005;11:99-105.
18Roman RM, Lobo PI, Taylor RP, Goodkin DA, LaBrecque J, Powers KL, et al. Prospective study of the immune effects of normalizing the hemoglobin concentration in hemodialysis patients who receive recombinant human erythropoietin. J Am Soc Nephrol 2004;15:1339-46.
19Wolcott DL, Marsh JT, La Rue A, Carr C, Nissenson AR. Recombinant human erythropoietin treatment may improve quality of life and cognitive function in chronic hemodialysis patients. Am J Kidney Dis 1989;14:478-85.
20Craig KJ, Williams JD, Riley SG, Smith H, Owens DR, Worthing D, et al. Anemia and diabetes in the absence of nephropathy. Diabetes Care 2005;28:1118-23.
21National KF. K/DOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Am J Kidney Dis 2002;39 2 Suppl 1:S1.
22Definition WHO. Diagnosis of Diabetes Mellitus and Intermediate Hyperglycemia: Report of a WHO/IDF Consultation. Geneva: WHO; 2006. p. 50.
23World Health Organisation. Iron Deficiency Anaemia: Assessment, Prevention and Control. A Guide for Programme Managers. Geneva: World Health Organization; 2001. In.: WHO/NHD/01.3; 2015.
24Levey AS GT, Kusek JW, Beck GJ. MDRD Study Group: A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 2003;14:2573-80.
25American Diabetes A. Standards of medical care in diabetes – 2014. Diabetes Care 2014;37 Suppl 1:S14-80.
26Consultation WHOE. Waist Circumference and Waist-Hip Ratio. Report of a WHO Expert Consultation. Geneva: World Health Organization; 2008. p. 8-11.
27Alberti KG, Zimmet P, Shaw J. Metabolic syndrome – A new world-wide definition. A Consensus Statement from the International Diabetes Federation. Diabet Med 2006;23:469-80.
28Adler AI, Stratton IM, Neil HA, Yudkin JS, Matthews DR, Cull CA, et al. Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36): Prospective observational study. BMJ 2000;321:412-9.
29Goldhaber A, Ness-Abramof R, Ellis MH. Prevalence of anemia among unselected adults with diabetes mellitus and normal serum creatinine levels. Endocr Pract 2009;15:714-9.
30Salah N, El Hamid FA, Abdelghaffar S, El Sayem M. Prevalence and type of anaemia in young Egyptian patients with type 1 diabetes mellitus. East Mediterr Health J 2005;11:959-67.
31New JP, Middleton RJ, Klebe B, Farmer CK, de Lusignan S, Stevens PE, et al. Assessing the prevalence, monitoring and management of chronic kidney disease in patients with diabetes compared with those without diabetes in general practice. Diabet Med 2007;24:364-9.
32DeFronzo RA, Goodman AM. Efficacy of metformin in patients with non-insulin-dependent diabetes mellitus. The Multicenter Metformin Study Group. N Engl J Med 1995;333:541-9.
33Ting RZ, Szeto CC, Chan MH, Ma KK, Chow KM. Risk factors of vitamin B (12) deficiency in patients receiving metformin. Arch Intern Med 2006;166:1975-9.
34Carlsen SM, Følling I, Grill V, Bjerve KS, Schneede J, Refsum H. Metformin increases total serum homocysteine levels in non-diabetic male patients with coronary heart disease. Scand J Clin Lab Invest 1997;57:521-7.
35Boushey CJ, Beresford SA, Omenn GS, Motulsky AG. A quantitative assessment of plasma homocysteine as a risk factor for vascular disease. Probable benefits of increasing folic acid intakes. JAMA 1995;274:1049-57.
36Nygard O, Nordrehaug JE, Refsum H, Ueland PM, Farstad M, Vollset SE. Plasma homocysteine levels and mortality in patients with coronary artery disease. N Engl J Med 1997;337:230-6.
37Welch GN, Loscalzo J. Homocysteine and atherothrombosis. N Engl J Med 1998;338:1042-50.
38de Jager J, Kooy A, Lehert P, Wulffelé MG, van der Kolk J, Bets D, et al. Long term treatment with metformin in patients with type 2 diabetes and risk of vitamin B-12 deficiency: Randomised placebo controlled trial. BMJ 2010;340:c2181.
39Berria R, Glass L, Mahankali A, Miyazaki Y, Monroy A, De Filippis E, et al. Reduction in hematocrit and hemoglobin following pioglitazone treatment is not hemodilutional in Type II diabetes mellitus. Clin Pharmacol Ther 2007;82:275-81.
40Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension. Proc Natl Acad Sci U S A 1995;92:5510-4.
41Jones S, Smith D, Nag S, Bilous M, Winship S, Wood A, et al. Prevalence and nature of anaemia in a prospective, population-based sample of people with diabetes: Teesside anaemia in diabetes (TAD) study. Diabetic Med 2010;27:655-9.
42Stevens PE. Anaemia, diabetes and chronic kidney disease: Where are we now? J Ren Care 2012;38 Suppl 1:67-77.
43Iron Deficiency Anaemia. Assessment p, and Control. A Guide for Programme Managers. Geneva: World Health Organization; 2001 (WHO/NHD/01.3).
44McLean E, Cogswell M, Egli I, Wojdyla D, de Benoist B. Worldwide prevalence of anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993-2005. Public Health Nutr 2009;12:444-54.
45World Health Organisation. Iron Deficiency Anaemia: Assessment, Prevention, and Control. A Guide for Programme Managers. Geneva: WHO/NHD/01.3; 2001.
46World Health Organization. Worldwide Prevalence of Anemia 1993 – 2005: WHO Global Database on Anemia/Edited by Bruno de Benoist. Geneva, Switzerland: McLean E, Egli I, Cogswell M; 2008.