Prevalence and associated factors of anaemia in patients with type 2 diabetes mellitus: a cross-sectional study in a tertiary care medical unit, Sri Lanka

Abstract

Background

Anaemia is a global public health issue that impacts individuals of all ages in both developed and developing countries. Anaemia is common in patients with diabetes mellitus; however, it is often undiagnosed and untreated. The main aim of this study was to assess the prevalence and associated factors of anaemia in patients with type 2 diabetes mellitus admitting to a medical unit at National Hospital Kandy.

Methods

A descriptive, cross-sectional study was conducted in type 2 diabetes mellitus (T2DM) patients admitted to a medical ward at National Hospital Kandy (NHK). They were assessed with a pre-tested, interviewer-administered, structured questionnaire using consecutive sampling method. The data was entered and analyzed using SPSS 26.

Results

Total 252 patients with diabetes were included. The prevalence of anaemia in patients with T2DM was 31.3%. The corresponding values for males and females were 34.2% and 65.8% respectively. Independent predictors for anaemia among diabetic patients were older age, female gender, poor glycemic control, diabetes duration > 5 years, diabetic nephropathy, retinopathy, neuropathy, stage ≥ 3 chronic kidney disease (CKD), ischaemic heart disease (IHD), peripheral vascular disease (PVD), diabetic foot ulcers (DFU) and usage of aspirin. These were significantly associated with the prevalence anemia among patients with type 2 diabetes mellitus. Multivariate logistic regression analysis revealed that female gender, age ≥ 65 years, diabetic duration > 5 years, poor glycaemic control, stage ≥ 3 CKD, diabetic nephropathy and retinopathy were associated with greater odds for the presence of anaemia.

Conclusion

We found that 31.3% T2DM patients in a medical ward at NHK had previously undiagnosed anaemia. Anaemia screening during diabetes diagnosis, maintaining glycaemic control and raising patient awareness can reduce anaemia prevalence, improve patient quality of life and potentially reduce microvascular complications.

Introduction

Anaemia is a worldwide public health concern that affects people of all ages. The definition of anaemia is a haemoglobin level of less than 12 g/dL in women and less than 13 g/dL in men. Anaemia is caused by a decrease in the amount of haemoglobin in the blood, which lowers the ability of red blood cells to carry oxygen [1]. According to the World Health Organization (WHO), the prevalence of anaemia is estimated to be 24.8% worldwide. Preschool-aged children have an estimated prevalence of anaemia of 47.4%, pregnant women 41.8% and non-pregnant women 30.2%. Interestingly, as per WHO statistics the prevelance of anaemia in Sri Lanka is 29% [2].

Patients with diabetes mellitus frequently exhibit anaemia, yet it goes undiagnosed and untreated. Anaemia is frequently seen in people with diabetes who also have concurrent renal insufficiency [3]. Interestingly, a study discovered that anaemia was prevalent in diabetics even prior to renal impairment [4]. There is evidence in available literature that anaemia increases the severity and worsens the outcome of peripheral small vessel disease in patients with diabetes [5]. The development of diabetic retinopathy, diabetic neuropathy, and diabetic nephropathy is accelerated by anaemia. It also increases cardiovascular mortality in patients with diabetes [5]. Haemoglobin concentration and fasting blood glucose are significantly correlated. Individuals with poorly managed diabetes and those with both diabetes and renal insufficiency are at a higher risk of developing a high incidence of anaemia [5].

Multiple factors contribute to the pathophysiology of anaemia in diabetes [1]. It is partially caused by diabetic kidney disease which leads to renal tubulointerstitium modifications, affecting interstitial fibroblasts, tubular cells, and capillaries essential for hemopoietic function [6]. In the context of compromised renal compensation, anaemia is also caused by functional haematinic deficits, systemic inflammation, erythropoietin resistance, cytokine-induced suppression of erythropoietin production and decreased red cell survival [6]. Furthermore, it has been proposed that a decrease in haemoglobin levels may be caused by the widespread use of angiotensin-converting enzyme inhibitors in diabetes, especially in individuals with elevated urine albumin levels [7]. Splanchnic sympathetic activation of erythropoietin production can be inhibited by renal denervation caused by diabetic autonomic neuropathy [3]. Anaemia in diabetes is related to gender, age, the length of diabetes and glycaemic control [8]. Other causes of anaemia in patients with T2DM include upper gastrointestinal haemorrhage caused by antiplatelets and vitamin B12 insufficiency created by metformin [9].

Few research on the prevalence of anaemia in diabetes were found in available literature. A 2003 study at Kent & Canterbury Hospital revealed that anaemia significantly impacts the quality of life in diabetic patients, yet their awareness of testing is low [10]. In 2006, a study in Dublin found that 11% of males and 15% of females were anaemic, with 74% having a serum creatinine level below 110µmol/l and 72% having a clearance of over 60 ml/min [11]. A study at James Cook University hospital found 15% of patients with diabetes had previously undiagnosed anaemia, with 36% having eGFR < 60 ml/min per 1.73m2, compared to population-based cohorts [12]. A 2013 study in China found that Chinese people with T2DM are more susceptible to anaemia, especially those with pre-existing cardiovascular or renal issues [13]. A 2014 Pakistan study found females have a higher incidence and risk of anaemia, with 49.5% in poorly controlled diabetics, highlighting the need for anemia screening [14]. In 2016, study in Gujarat, India, revealed anemia in 18% of T2DM patients and suggests treating it can help manage the disease’s complications [15]. A study conducted in Kuwait in 2017 found that anaemia is more common in female diabetics (38.5%) and those with poorly managed blood sugar (33.46%) and increases with age [16]. A research in Western Australia found anaemia complicating diabetes mellitus is linked to higher death risk and doubles T2DM risk, affecting 11.5% of subjects [17]. Research indicates a link between vitamin B12 deficiency and diabetes, with long-term metformin use linked to increased risk of insufficiency, emphasizing the need for B12 testing [9].

A 2019 study by Awofisoye et al. found that anemia is prevalent in T2DM patients, including those with normal eGFR and is primarily associated with longer diabetes duration [18]. Research in Delhi, India, found high prevalence of iron deficiency and anaemia in T2DM patients, even without nephropathy, necessitating further research [19]. In 2020 study at Alkhair Medical Centre in Sudan found 7% prevalence of anemia in T2DM patients, with higher rates in females and males [20]. A meta-analysis and systematic review conducted in Africa in 2021 revealed that anaemia in diabetes patients hastens long-term consequences, with higher prevalence in diabetic foot lesions [21]. Another study was carried out in 2021 in Ethiopia to determine the prevalence of anaemia and the factors that are linked to it in individuals with T2DM where they found that sex, treatment mode, microvascular problems, and hypertension are linked to anaemia prevalence, suggesting screening could prevent complications [22]. A 2021 study at Worabe Comprehensive Specialized Hospital in Southern Ethiopia found 25.3% of patients with metabolic syndrome had anemia, significantly linked to alcohol use, dyslipidemia, obesity and diabetes [23].

Interestingly, a study conducted in Sri Lanka among patients with T2DM attending the Family Practice Centre, University of Sri Jayewardenepura, in 2022, reported that the prevalence was 35.29% with female preponderance [24]. A 2023 study in Iran found that obesity, T2DM, albuminuria, CKD and hypertriglyceridemia are significantly associated with prevalent anemia in T2DM patients. [25]. In the same year in Saudi Arabia they found that anaemia is more common among female diabetics (53.3%) than among diabetic men (42.4%) and smokers (2.78 times higher risk) [26]. A study in Chennai, India, found 60% prevalence of anaemia, notably higher in females, those with poor glycaemic control and those with at least one microvascular complication [27].

After all, it is obvious that there is relatively little information in the literature currently under publication about the prevalence and correlates of anaemia in Asians, particularly in Sri Lankan patients with T2DM. Patients with diabetes will function better and have a higher quality of life if their anaemia is identified early and treated. Hence the purpose of this study was to ascertain the prevalence of anaemia and the factors that are linked to it in T2DM patients at a tertiary care hospital in Sri Lanka.

Methods

Study design and population

The study was a descriptive cross-sectional study with a survey-based strategy conducted at NHK. The study population included all patients age ≥ 18 years diagnosed to have T2DM, admitted to a medical ward at NHK. Patients less than 18 years, known patients with anemia including patients with primary haematological diseases like thalassemia, haemolytic anaemia, aplastic anaemia and haematological malignancies, pregnant women, patients with chronic kidney disease secondary to causes other than diabetic nephropathy, patients who are on any form of treatment for anaemia at the time of admission, patients with type 1 diabetes, gestational diabetes mellitus and diabetes due to other causes other than T2DM were excluded from the study. Ethical clearance for the study was obtained by the Ethics Review Committee at NHK.

Sample size calculation and sampling method

The sample size was calculated using the following formula (a single population proportion formula), in which, n is the sample size, Zα/2 is the confidence coefficient, p is the expected prevalence of anaemia among patients with type 2 diabetes mellitus and d is the precision level.

$$\:\text{n}=\frac{{(\text{Z}{\upalpha\:}/2)}^{2}\text{p}(1-\text{p})}{{d}^{2}}$$

Due to the unavailability of any large-scale studies published in Sri Lanka, with regard to prevalence and associations of anaemia among patients with T2DM, the expected proportion of patients with anaemia among patients with T2DM was adopted from a cross-sectional study performed in Gujarat, India [15]. Therefore, the expected prevalence was taken as 18%. A 5% precision level, 95% confidence interval and a confidence coefficient of 1.96 were used for the calculation. Calculated sample size was 227. A non-responder rate of 10% was added which gives a final sample size of 252. All patients fulfilling the eligibility criteria and consenting to participate were selected for the study. Data of the study subjects were collected from September 2022 to December 2022 using consecutive sampling method.

Data collection and analysis

Face-to-face interviews were used to gather data. Additionally, some data were gathered from bed head tickets. A structured, closed-ended questionnaire was used, with some adjustments made to fit the local context (Annexure A). The questionnaire was divided into three sections: questions about the subject’s background and demographics; questions about diabetes and its complications; and questions about anaemia and its contributing variables. From each case record, information was collected under the following headings: age, gender, ethnicity, duration of diabetes, presence or absence of diabetes related macrovascular and microvascular complications, Hb value, mean corpuscular volume (MCV) and medication history (metformin, aspirin, clopidogrel and anticoagulants). Informed written consent was obtained from all participants. No patient-identifiable information was collected and each research participant was assigned a unique identification number. Confidentiality of all the gathered data was strictly maintained.

Patients were classified as diabetics if their glycated haemoglobin (HbA1c) was ≥ 6.5%, fasting blood sugar (FBS) was ≥ 126 mg/dl) or random blood sugar (RBS) ≥ 200 mg/dl. Additionally, anaemia was defined as having a haemoglobin level of less than 13.0 g/dl for males and less than 12.0 g/dl for females as per WHO definition. Patients with HbA1c levels of < 7% were considered as having well controlled diabetes group, whereas levels of ≥ 7% were considered as having poorly controlled diabetes [28]. According to mean corpuscular volume (MCV), anaemia was classified as normocytic (MCV 80–100 fl.), macrocytic (MCV > 100 fl.), or microcytic (MCV < 80 fl.) based on the size of the red blood cells [29].

Patients with eGFR of < 60 ml/min/ 1.73m² for more than 3 months were considered as having stage ≥ 3 CKD stage and albuminuria of ≥ 30 mg/24 h or an albumin-to-creatinine ratio of ≥ 30 mg/g on ≥2 occasions, checked at least three months apart was defined as diabetic nephropathy as per Kidney Disease Improving Global Outcomes (KDIGO) guidelines. This information was obtained from clinic records. Diagnosis of peripheral neuropathy was based on American Diabetes Association diagnostic criteria. A combination of typical symptoms with typical signs were used in diagnosis including 10 g monofilament. PVD was diagnosed based on significant arterial narrowing on doppler ultrasonography of the lower limbs.

SPSS version 26 was used for inferential analysis. P values less than 0.05 were considered statistically significant. Chi squared testing was utilized for categorical data, and the student’s t test was used to analyze continuous variables. Multivariable logistic regression was used to determine the association between anaemia and different variables.

Results

Fig. 1

Gender distribution of study participants

The study population consisted of 252 participants with T2DM with a mean (SD) age of 56.85 years (95% CI: 55.47–58.22). As shown in Fig. 1, among them, 51.2% were females and 48.8% were males. Among patients with diabetes the average age of patients with anemia was found to be 63.65 years (95% CI: 61.60–65.70). The average age of T2DM patients without anaemia was found to be 53.74 years (95% CI: 52.16–55.32). In patients with anaemia compared to those without anaemia, there was a statistically significant difference in average age (P < 0.001). This suggests that the likelihood of anaemia rises with advancing age.

Fig. 2

Relationship between anaemia and gender, duration of diabetes and glycaemic control

Among these 252 patients, 79 patients were found to be anemic by the WHO criteria accounting for 31.1% (95% CI: 26 to 31). Among anaemic patients, 27 were males (34.2%) and 52 were females (65.8%). As illustrated in Fig. 2, diabetic females are more likely than diabetic males to have anaemia (p = 0.02).

The mean Hb level in the total diabetic population was 12.66 g/dL (95% CI: 12.43 to 12.89). The mean Hb level among the diabetic patients with anemia was 10.47 g/dl (95% CI: 10.11 to 10.82) and 10.51 g/dl (95% CI: 9.73 to 11.29) in females and males respectively. The average MCV of the total anemic patients in the diabetic group was 81.54 fl. (95% CI: 78.47 to 84.60). The MCV for the 27 anemic males was 84.29 fl. (95% CI: 81.68 to 86.91) and the average MCV for the 52 anemic females was 80.10 fl. (95% CI: 75.63 to 84.58). Therefore, the majority of patients had a normocytic anemia. Out of the 79 anaemic patients with diabetes, 21 had microcytic anaemia (26.58%), 55 had normochromic normocytic anaemic (69.62%) and 3 had macrocytic anaemia (3.80%).

Among patients with anaemia 63.3% had poor glycaemic control whereas 36.7% had good control. Among patients without anaemia 64.7% had well controlled diabetes and 35.3% had poorly controlled diabetes. Patient with anaemia had poor glycaemic control which was statistically significant (p < 0.001). Regarding the duration of diabetes, out of the 79 patients with diabetes and anaemia, 82.3% patients had diabetes mellitus for > 5 years duration. Among diabetic patients without anaemia only 59.5% had diabetes > 5 years. This difference was statistically significant (p = 0.001). Patients who have had diabetes for a longer period of time therefore have a higher risk of developing anaemia.

Fig. 3

Relationship between anemia and microvascular complications, macrovascular complications and diabetic foot ulcers

As shown in Fig. 3, out of the diabetic patients with anaemia 64.7% had nephropathy and 26.6% had stage ≥ 3 CKD. Both of these values were statistically significant when comparing with diabetic patients without anaemia where only 51.9% had nephropathy and 6.4% had CKD (p = 0.001 and p = < 0.001 respectively). Among the diabetic patients with anaemia 68.4% had retinopathy (proliferative- 1.30% and non-proliferative- 82.3%) while in patients without anaemia only 52% had retinopathy. This was statistically significant (p = 0.015). 56.7% of anaemic patients had neuropathy while 41.6% patients without anaemia had neuropathy. This was a statistically significant difference (p = 0.023). Therefore, the risk of anaemia was increased in patients with nephropathy, ≥ 3 stage CKD, retinopathy, and neuropathy.

39.2% of anaemic patients had a past history of IHD, 13.9% had CVD and 21.5% had PVD. In contrast, among non-anaemic diabetic patients only 25.4% had IHD, 12.1% had CVD and 11.0% had PVD. Further, there was a statistically significant difference in IHD and PVD but not in CVD. 13.9% anaemic patients had DFU while only 4% non-anaemic patients had DFU (p = 0.005). This difference was statistically significant. Therefore, anaemia is seen more in patients with diabetic wounds.

Fig. 4

Relationship of anaemia with medications

As depicted in Figs. 4 and 87.3%of anaemic patients were on metformin. 49.4% were on aspirin and 25.3% were taking clopidogrel. Among the non-anaemic patients 78% were on metformin, 23.1% were on aspirin and 23.7% were on clopidogrel. There was a statistically significant difference in anaemia with aspirin use but not with clopidogrel or anticoagulation use. Therefore, anaemia was seen more in patients who were taking aspirin.

As per Tables 1 and 2 multivariable logistic regression analysis that considered factors associated with odds of anaemia revealed that increasing age, female sex, poor glycemic control, diabetic duration > 5years, diabetic nephropathy, diabetic retinopathy and stage ≥ 3 CKD were associated with greater likelihood of anaemia.

Table 1 Comparison of age, duration of diabetes, diabetic control, complications and medications in diabetic patients

Table 2 Multivariable logistic regression analysis between anaemia with age, gender, glycaemic control, complications, DFU and medication

Discussion

Anaemia is a prevalent public health issue globally as well as in Sri Lanka [30]. It is estimated that 24.8% of the world’s population are affected from anaemia [2]. Interestingly there aren’t enough studies on the prevalence of anaemia in Sri Lanka. The majority of research focused on the prevalence of anaemia among young people and expectant mothers. Each study had limited its scope to evaluate the risk factors for anaemia in certain age groups [31]. However, the WHO reports that 29% of Sri Lankans suffer from anaemia [2].

The main aim of this study was to assess the prevalence of anaemia in patients with T2DM and to assess the risk of anaemia according to gender, age, glycaemic control, microvascular and macrovascular complications and medications in patients admitted to a medical ward a NHK which is a tertiary care hospital in Sri Lanka.

As mentioned in the literature survey few studies across developing nations including India, Pakistan, Iran, Kuwait Hongkong and Ethopia have assessed the prevalence of anaemia in diabetes. Further there was one study conducted in Sri Lanka, at an outpatient setting in a tertiary care hospital. However, interestingly results of this study showed some differences from the above studies.

The prevalence of anaemia in this diabetic population was found to be 31.1% which was close to the prevalence conducted at University of Sri Jayewardenepura, Sri Lanka [35.29%]. However, our study was conducted in an inpatient setting at a medical ward in contrast to latter which was conducted in an outpatient family practice centre. Our prevalence was also similar to that of Kuwait and Africa which were 29.7% and 35% respectively [16, 21]. The prevalence in our study was higher than that of some other developing countries including in Sudan [20], Ethiopia [22, 23] and Iran [25]. In addition, the prevalence was higher than is few developed countries including Hongkong [13] and Western Australia [17]. Is it also observed that our prevalence is lower than that of India [27], Pakistan [14] and Nigeria [18], which are developing countries with similar socio-economic background. These discrepancies in the prevalence of anaemia among patients with diabetes may be caused by variations in the cutoff value used to measure it, the sample size and changes in the general features of the research area and population.

Mean MCV was 81.54 fl. (95% CI: 78.47 to 84.60) among anemic patients. The majority of patients had normocytic normochromic anemia which was similar to studies conducted in India [15] and Nigeria [18]. Interestingly this was in contrast to few studies including studies in Ethiopia [8] and Sudan [20] where the commonest type of anaemia was hypochromic microcytic anaemia. Causes for normocytic normochromic anaemia in patients with diabetes is multifactorial.

In this study it was found that diabetic females were at higher risk of anaemia than diabetic males. This finding is in consistent with most of the studies in available literature [14,15,16, 20, 24]. However, there were few studies, were males had a higher risk of anaemia than females [13, 21, 22]. In our study, 5.77% of anaemic females had menstrual cycles > 3days whereas 5.2% of non-anaemic females had menstrual cycles > 3 days. This was not statistically significant. A higher incidence of anaemia in women may be caused by inadequate nutrition or a lack of empowerment, which leads to a decreased emphasis on their own health. Education initiatives including health awareness campaigns, the distribution of foods high in iron, the prescription of vitamin and iron supplements and awareness of the difficulties associated with diabetes can all help to improve this [32].

According to our study the mean age of individuals suffering from anaemia is considerably greater than the mean age of those without the condition. These findings are in agreement with studies in China [13], Kuwait [16] and Ethiopia [8]. Moreover, this was similar to the study conducted at Sri Jayawardanapura University Sri Lanka [24]. Age-related increases in this incidence could be caused by vitamin deficiencies (such as folate, cyanocobalamin), undiagnosed bone marrow diseases or a higher number of comorbidities [16].

Based on our results, those with poorly managed diabetes had a noticeably greater prevalence of anaemia. This was consistent with findings of many studies including studies conducted in Pakistan [14], Kuwait [16] and Africa [21]. Patients with poorly controlled diabetes have erythrocyte precursors in the bone marrow that are susceptible to long-term direct toxicity from glucose. Further, mature erythrocytes may be impacted by oxidative stress, which can cause abnormalities in erythrocyte function [13]. Anemia-related diabetes patients had higher levels of fasting blood sugar and HbA1c. Nonetheless, some research revealed that anaemic patients had reduced HbA1c levels [33]. They explained that the glycation process could be slowed down by decreased haemoglobin concentration and increased red blood cell turnover in chronic illness anaemia, which would therefore cause erroneous reports of lower HbA1c values [34].

The correlation between the length of diabetes and the frequency of anaemia was another noteworthy discovery. In keeping with some earlier studies, the current analysis demonstrated that T2DM duration > 5 years had a strong independent correlation with anaemia, independent of glycemic indices and nephropathy. These findings were similar to studies conducted in Nigeria [18], Ethiopia [8], Sri Lanka [24] and Iran [25] where patients with diabetes for > 5 years had a high risk of anemia. These observations suggest that patients with diabetes should be screened for anemia in routine management in routine practice and should be treated to minimize the risk of microvascular complications such as nephropathy and retinopathy.

The findings of this study indicate a substantial correlation between the incidence of anaemia and diabetes-related microvascular complications. Anaemia-related tissue hypoxia is linked to a number of mitogenic and fibrogenic effects that are implicated in the development of microvascular complications, and it is well recognized that these effects play a major part in diabetes-associated organ damage [35]. We discovered that among individuals with type 2 diabetes, anaemia was significantly correlated with two key indicators of diabetic kidney disease (DKD): albuminuria > 30 mg/24 hours and e-GFR ≤ 60 ml/min/1.73 m2. Our results were consistent with previous research indicating a greater incidence of anaemia in T2DM and nephropathy patients. Interestingly there was a significant association between retinopathy and neuropathy in contrast to most of the studies. The study conducted in Sri Lanka in 2022 stated that there was no association of anaemia with retinopathy. However, a study in Ethiopia concluded that anaemia was significantly associated with all microvascular complications including retinopathy and nephropathy [8]. In terms of how anaemia affects diabetic retinopathy, it appears that hypoxia brought on by anaemia increases the release of vaso-proliferative factors (X factor), which in turn results in progression of diabetic retinopathy.

In our study when the co-factors were individually assessed, there was a significant association between anaemia and IHD as well as PVD which was in contrast to many other studies in available literature. According to these results, anaemia may be linked to the onset of cardiovascular problems in diabetes patients; however, as age may be a significant confounding factor, interpretation of the data should be done with caution. Interestingly, in 2011 Chen CX et al. in China reported that anaemia was significantly associated with stroke and IHD. Moreover, a study conducted in Nigeria in 2019 revealed that anaemia was significantly associated with CVD and PVD [36]. Anaemia and PVD did not correlate in Chuan et al.‘s study [37]. In contrast to our findings, the authors in this study observed a strong correlation between anaemia and stroke in their patients.

When assessed as an individual co-factor, diabetic foot wounds were significantly associated with anaemia. Therefore, the risk of anaemia increases with DFU. This was similar to the result in the study conducted by Ritesh Kumar et al. in 2023 among hospitalized patients where they concluded that the prevalence of anemia was very high in DFU and more than 75% of the patients had moderate to severe anemia. Further, they discovered that the severity of anemia was associated with the severity of DFU. And that the most common cause of anemia was anemia of chronic disorder secondary to diabetic foot infection. They found that during the period of hospitalization, hemoglobin decreased despite improvement in DFU infection [38]. Interestingly, there were several other studies showing that diabetic foot ulcers are associated with anaemia. A study conducted in Nigeria revealed that 53.6% of patients with DFU had anemia [36] and this percentage was 85.67% [39] in a study conducted in Pakistan. These studies described that anaemia in DFU patients may also be brought on by malnourishment, diabetic nephropathy, and chronic inflammation. They explained that the ability of inflammatory cells, particularly macrophages, to absorb available iron and store it in ferritin, blocking the transfer of iron to the bone marrow, is one reason for persistent inflammation-induced anaemia.

Metformin, a diabetic drug, can potentially cause vitamin B12 insufficiency by interfering with the absorption of vitamin B12 [9]. However, in our study there was no association with metformin use and anaemia.

Interestingly, patients who were on aspirin had a statistically significant higher risk of anaemia.

Among them, 85% patient were on low-dose aspirin (LDA). Gaskell et al. reported that in the absence of obvious bleeding, it is unclear if LDA and anaemia are related, however in (a subset of) older people, there might be a connection between LDA and a decline in haemoglobin. Therefore, more research would help shed light on this important issue [40]. In our study clopidogrel and anticoagulants were not correlated with the risk of developing anaemia.

As shown in Table 2, multivariate logistic regression analysis revealed that female gender, age ≥ 65 years, poor glycaemic control, diabetes duration > 5 years, stage ≥ 3 CKD, diabetic nephropathy and retinopathy were associated with greater odds for presence of anaemia. Interestingly this was in contrast to the study conducted in China where male gender, old age, CKD, nephropathy, stroke and IDH were associated with greater odds for presence of anaemia [13].

Our study had several limitations. One such limitation is that details on study participant’s haematic level haven’t been included in the study. The reason for this was that there weren’t enough laboratory facilities available to measure ferritin, vitamin B12, and folate levels because this study was carried out during Sri Lanka’s economic crisis. Secondly, only diabetes patients who have undergone annual blood testing, nephropathy screening, retinopathy screening and patients who were compliant to medications were selected. This could have caused a selection bias. In addition, we were unable to identify the casualty since the study was cross-sectional; therefore, a longitudinal investigation is required to evaluate the association over time.

Conclusions and recommendations

In conclusion, we demonstrated a high prevalence of anemia among patients with T2DM a medical ward at NHK, which is a tertiary care hospital in Sri Lanka. This was associated with increasing age, female gender, duration of diabetes (> 5year), poor glycaemia control and microvascular complications (nephropathy, retinopathy, neuropathy). Screening and evaluation of anemia are essential in diabetic populations. Inco-operating anaemia screening in diabetic patients would enhance the quality of life in these patients. Patient’s need to be made aware of the possible risks and complications of anaemia as well as the significance of routine screening, particularly for those who have been identified as having related variables. Furthermore, further research is needed on anaemia in diabetics, particularly in the Sri Lankan population.