This was an observational, prospective study conducted at an Indian tertiary care center from January to March, 2010. Consecutive patients admitted with diagnosis of ACS to Intensive Coronary Care Unit (ICCU) or Acute Medical Care Unit (AMCU) were included. Patients were excluded only if they refused to give informed consent.

An ACS was diagnosed by treating cardiologist on the basis of the history, electrocardiogram (ECG) and elevated cardiac biomarkers. Patients who had been admitted to ICCU or AMCU were then reassessed and diagnosis was further verified by investigator, who completed a protocol-driven case record form.

We collected data on demographic and clinical information that was relevant to ACS and anaemia (e.g., past history of Coronary Artery Disease (CAD), previous revascularisation, hypothyroidism, hypertension, diabetes, chronic kidney disease, drug history and dietary habits). Chronic kidney disease was defined according to National Kidney Foundation Disease Outcomes Quality Initiative definition i.e., either kidney damage or a Glomerular Filtration Rate (GFR) of less than 60mL/min/1.73 m² of body surface area lasting for longer than 3 months [19]. A broad range of clinical attributes were included to describe the potential associations between these clinical attributes of ACS patients and anaemia.

Haematology and blood chemistry were obtained from the first admission blood sample, which was obtained in Emergency Medicine Ward, ICCU or AMCU.

Patients identified by the cardiology team as having high-risk features, such as ongoing ischemia, were triaged for additional investigations, which usually involved Coronary Angiography (CAG). When required these patients had undergone Percutaneous Coronary Intervention (PCI) or Coronary Artery Bypass Graft surgery (CABG) and such details were also recorded including the number coronary vessels affected. The severity of CAD was assessed by number of abnormal epicardial coronaries demonstrated by CAG. Duration of hospital stay and in-hospital death was used as a surrogate measure of hospital outcome.

Anaemia was defined according to World Health Organization criteria as Haemoglobin (Hb) <120 g/l in women and <130 g/l in men [20,21]. Its severity was graded as mild when Hb was 100-129 g/l in men and 100-119 g/l in women, and either moderate or severe when Hb was 70-99 g/l and less than 70 g/l, respectively in women. In men, it was graded either moderate or severe when Hb was 80-109 g/l and less than 80 g/l, respectively [20,21]. Anaemic status, severity and its possible causes were identified from admission blood samples with haematinic parameters like serum ferritin, vitamin B12 and folate levels in most of the patients. In patients with normal haematinic parameters relevant additional tests were done to evaluate the aetiology. Anaemia was classified as normocytic if Mean Corpuscular Volume (MCV) was 83-101 fl [22]. Any value above this was considered as macrocytic and any value below this as microcytic.

Automated cell counter (Coulter LH-500, Beckman Coulter) was used to measure haemoglobin and red cell indices on admission.

The Chairman of the Institutional Ethical Committee of the study centre approved this study. The study protocol conforms to the ethical guidelines of the Declaration of Helsinki as reflected in a priori approval by the institution’s clinical research and education committee. The informed consent was taken in their local language from all patients.

Continuous variables were presented as mean and standard deviation and categorical variables as counts and percentages. Significance was assessed at p<0.05 level of significance. Student’s t-test (two tailed, independent) had been used to find the significance of study parameters on continuous scale (between two groups Inter group analysis). Chi-square/ Fisher-Exact test had been used to find the significance of study parameters on categorical scale between two or more groups. Levene’s test was performed to assess the homogeneity of variance.

Haemoglobin values were categorized as anaemia or no anaemia and potential associations with haemoglobin levels were identified using descriptive analyses. Associations with admission duration were also studied. Variables significant (p<0.05) in a simple univariate regression analysis were entered into a forward step-wise regression model, with entry and removal criteria of p < 0.2 and p < 0.4, respectively. A multivariate association significant at p <0.05 was taken as predictor of duration of hospital stay. The STATA (StataCorp LP, Texas, USA) program, version 10.1 and Epi info (CDC, Georgia, USA) program, version 2.3 were used for the analysis of the data.

Of 130 (87 males) consecutive patients with ACS admitted during study period, 62 (47.7%) patients had unstable Angina, 13 (10%) had NSTEMI and 55 (42.3%) had STEMI. Overall prevalence of anaemia was 51.5% of the study population. Prevalence of anaemia was related to gender, it was more common in women (n=30, 69.8%) compared to men (n=37, 42.5%; p<0.01). Mean haemoglobin in the first admission blood sample was higher in men (132 ± 28 g/l) than in women (108 ± 23 g/l).

The characteristics of patients with anaemia and without anaemia were described in [Table/Fig-1]. Mean age of anaemic patients was higher than non anaemic (62.69±10.36 vs. 53.40±13.74 years; p<0.01). Occurrence of anaemia was higher in patients with chronic kidney disease (13.4%; p<0.01), hypothyroidism (13.4%; p=0.04), and diabetes (35.8%; p=0.04). The proportion of unstable angina (58.2%; p=0.013) was more in anaemic patients whereas the proportion of STEMI (29.9%; p<0.01) was more in non anaemic patients.

Normocytic, microcytic or macrocytic anaemia at the time of admission was seen in 41 (61.2%), 19 (28.4%) and 7 (10.4%) patients, respectively. Microcytic anaemia was more commonly associated with female gender (p<0.01). In 46 (68.6%) patients, a single identifiable cause was found and in 5 (7.5%) patients multiple causes were seen whereas in 16 (23.9%) no cause could be identified. A low ferritin <20 μ/l) concentration was found in 20 patients (29.9%), low plasma vitamin B12 concentration (<190 ng/l) in 9 (13.5%), a low serum folate (<4 μg/l) concentration in 4 (6%) and an increased plasma Thyroid Stimulating Hormone (TSH) concentration (>5.5mU/l) in 3 (4.5%). The other causes were anaemia of chronic disease (3.0%), sickle cell trait (1.5%), chemotherapy induced bone marrow suppression (1.5%) and autoimmune haemolytic anaemia (1.5%).

The most common cause of anaemia was iron deficiency (29.9%). Out of 20 patients with iron deficiency, three patients had gastric erosions seen in upper gastrointestinal endoscopy. It was performed to evaluate the cause of upper GI bleed, one patient presented with haematemesis and other two patients had malena.

Five (7.5%) of the study patients had multiple cause of anaemia. Out of these five patients three (4.5%) had chronic kidney disease and vitamin B12 deficiency, one (1.5%) had combine Iron and vitamin B12 deficiency and other had folate deficiency and chronic kidney disease (1.5%).

Either aspirin or clopidogrel therapy were more often in anaemic patients than non anaemic patients at hospital presentation. Six (66.6%) out of nine patients with vitamin B12 deficiency were diabetics on metformin therapy.

Mean serum albumin were lower in patients with anaemia (36 ± 3 g/l) than non anaemic patients (39 ± 4 g/l) (p<0.01). Raised serum creatinine (mean) was seen in anaemic patients compared to non anaemic patients (143 vs. 99 μ mol/l; p<0.01).

Data of 94 (72.3%) patients who had undergone the CAG during the study was available. Single Vessel Disease (SVD) was more often found in non anaemic group than anaemic (39.3% vs. 26.3%; p=0.102). Triple Vessel Disease (TVD) was associated with anaemia (p =0.03).

Total 45 (34.6%) patients underwent revascularisation (PCI or CABG), 43 (95.6%) underwent PCI and 2 (4.4%) underwent CABG. Among 43 PCI patients, 18 (41.9%), 19 (44.2%) and 6 (14%) had SVD, Double Vessel Disease (DVD) and TVD, respectively. Two third of the patients with TVD had anaemia whereas only one third of patients with SVD and DVD had anaemia. CABG was done in two patients, one of them had TVD whereas other patient required intracardiac repair of septal rupture.

Of 130 study patients, 28.5% patients had congestive heart failure. Congestive heart failure was common in anaemic group compared to non anaemic group (p<0.01). Seven patients died (5.4%), five of whom had undergone non invasive management. Proportion (71.4%, n=5) of patients with anaemia was more in patients who died till discharge but it was not statistically significant (p=0.4).

Eight (6%) acute coronary syndrome patients underwent blood transfusion, of whom 7 (87%) were anaemic on admission to hospital. Out of the 45 patients who had undergone interventional management, 3 (6.7%) and 1 (2.2%) patients were transfused pre and post catheterization, respectively. Compared to non anaemic patients, anaemic patients had a longer mean duration of hospital stay (12.7 vs. 9 days; p<0.01). Admission characteristics that were univariate and multivariate predictors of duration of hospital stay (days) are shown in [Table/Fig-2,3], respectively. Haemoglobin and coronary revascularisation after admission were multivariate predictors of admission duration (p = 0.02 and 0.016).

Anaemia presents as a cardiovascular risk factor not only in the general population, but also found to be associated with a considerably elevated rate of cardiovascular events in patients with ACS, heart failure, and in patients undergoing PCI and CABG [23]. Anaemia at hospital admission in patients presenting with ACS ranges between 15% and 43% according to some studies [24]. Prevalence of anaemia in previous studies by Bindra et al., in United Kingdom (UK), Arant et al., in United States (US), Anker et al., in Europe and Sulaiman et al., in Middle East was 18%, 21%, 28% and 28%, respectively [25–28]. All of them used same definition for anaemia as in our study. Anaemia was present in about half (51.53%) of the population of present study. The higher prevalence of anaemia in our study may be explained by the presence higher overall prevalence of anaemia (i.e., 24% in men and 55% in women) in our country where nutritional deficiency is common [7]. However, the previous studies mentioned were from developed nations like UK, US, Middle East and other European countries.

The gender wise difference in prevalence of anaemia and mean haemoglobin may be partly explained by the presence of significant difference in the overall prevalence of anaemia in both sexes. Moderate to severe anaemia was significant in women compared to men (34.9% vs. 20.8%). Mean haemoglobin level was also lower in women when compared to men (108 vs. 132 g/l) even in patients with normal haemoglobin likely due to hormonal influences [29].

The mean haemoglobin of patients with anaemia decreased as age increased from 4^th decade to 8^th decade. Even though such observation was not statistically significant (p=0.2) in our study, the similar trend was reported previously by Bindra et al., Mean age of the patients with anaemia was higher compared to patients without anaemia (62.7 vs. 53.4 years; p <0.01) [25]. The relationship of anaemia and age is not completely understood, it may be a feature of normal ageing process, or it may be pathological due to chronic conditions resulting anaemia of chronic disease, or chronic inflammation in elderly [30].

In present study, the type of ACS was also related to anaemia. Unstable angina (58.2%) was more common in anaemic patients and STEMI (55.6%) was more common in non anaemic patients (p=0.013 and <0.01, respectively). Melissa et al., proved that anaemia at presentation was independently associated with recurrent ischemia by using continuous ECG monitoring in non ST elevation ACS (i.e., unstable angina and NSTEMI) patients [29]. They suggested that mismatches in oxygen supply and demand and bleeding related complications are the potential mechanisms by which anaemia leads to recurrent ischemia. This may explain why unstable angina tended to be more common in anaemic patients of our study. In a previously reported study by Bindra et al., STEMI was more often observed in non anaemic patients similar to our study but the accurate mechanisms could not be explained [25].

In present study, anaemia was associated with chronic kidney disease and also the mean creatinine (143 vs. 99 μmol/L; p<0.01) was higher in anaemic patients. Anaemia has been a common associate of advanced renal disease but may also occur in lesser degrees of renal impairment. The reduced erythropoietin production and impaired haemopoietic response to this hormone are the probable reasons [31]. In studies by Reinecke et al., and Lee et al., renal dysfunction was found to be an adverse prognostic marker in ischemic heart disease patients particularly who were anaemic [17,18]. Morici N et al., also reported a weak relation of anaemia with renal dysfunction, age and body mass index [32]. Furthermore, the combination of anaemia and renal dysfunction resulted in poor prognosis in severe forms of ischemic heart disease with heart failure [33].

Diabetes was common in patients with anaemia compared to non anaemic patient and it was statistically significant. This association of anaemia and diabetics was explained due to presence subclinical renal dysfunction in them affecting the production and impaired haemopoietic response of erythropoietin [34,35]. In a large prospective study, David H Shu et al., showed that presence of both diabetes and anaemia in patients with myocardial infarction have a considerably raised mortality than those with either diabetes or anaemia alone [4].

In the study population, anaemia was strongly associated with antiplatelet drug intake, both aspirin (p<0.01) and clopidogrel (p<0.01). The presence of occult or overt blood loss with use of these anti platelet drugs is the possible explanation, similar observations were reported Bindra et al., Rao SV et al., and Berry C et al., [25,36,37]. Anaemia, particularly pre-procedural, has significant affect on the outcomes and mortality hence needs to be addressed but the antiplatelets can be continued, as a previous study by Pilgrim T et al., showed that there has been no increase in mortality or major bleeding rates even in moderate to severe anaemia patients with dual antiplatelet therapy (DAPT) [38]. Hence anti platelet deprivation is not recommended after ACS, rather another study by Wang H et al., reported increased stent thrombosis in severe anaemia which required anti platelet therapy [39].

Lower serum albumin was seen in anaemic than in non anaemic patients (36 vs. 39 g/l; p =0.02). Such observation was attributed to the impairment in both anabolic and haemopoietic functions in anaemic patients apart from nutritional deficiency [25].

Frequency of different types of anaemia according to mean corpuscular volume i.e., microcytic, normocytic or macrocytic were 28.4%, 61.2% and 10.4%, respectively. In the present study all women were postmenopausal and had significant microcytic anaemia compared to men which may be due to higher prevalence of malabsorption of iron in women leading to iron deficiency [40].

In study population, three patients with iron deficiency had evidence of gastric erosions and all of them had undergone PCI with stent placement. Bleeding following PCI or CABG is a known risk factor for causing anaemia as these patients were on multiple antiplatelet and anti fibrinolytic agents [14,15,18,36,37].

Other common causes identified were vitamin B12 deficiency (12%) and folate deficiency (6%). This may explained by the fact that daily intake of these micronutrients measured was less than Recommended Daily Allowance (RDA) in Indians, especially it was low in Andhra Pradesh state, where study was conducted, compared to other states in the country [41].

In patients who had undergone CAG, the SVD was common in non anaemic group than anaemic whereas multiple vessel disease was more in anaemic group. This finding was consistent with previous studies by Mc Kechnie et al., and Meneveau et al., but the appropriate explanation for such correlation cannot be adequately given with the current available knowledge [2,14]. Zeidman et al., had reported that patients with both CAD and anaemia tended to suffer from a more advanced degree of CAD compared to patients with CAD alone [5].

Congestive heart failure was common in anaemic group compared to non anaemic group and anaemia was associated with heart failure (p<0.01). Various prospective [3–5] and retrospective studies [8,16] published on relationship of anaemia and ACS had reported a higher prevalence of congestive heart failure. It has been stated in a recent study that haemoglobin level is an independent determinant of heart failure in ACS patients [42]. Presence of congestive heart failure is a predictor of in-hospital outcome following PCI or acute coronary events [14,18]. Proportion of anaemic patients were higher in patients who died than those survived till discharge although this was not statistically significant in our study, this may be due to small sample size and cross sectional nature of study, but it is a well proven fact that anaemia is an independent predictor of in-hospital and long term mortality [8,12,17,25,43].

In our study duration of hospital stay was used as a surrogate marker for hospital outcome. Mean duration was found to be longer in anaemic patients compared to non anaemic (12.7 vs. 9 days; p=0.03). Univariate and multivariate analyses were done to predict duration of hospital stay. Haemoglobin, diabetes, chronic kidney disease, serum creatinine, serum albumin, transfusion and coronary revascularisation (PCI or CABG) were found to be univariate predictors. But in multivariate analysis, lower haemoglobin was found to be an independent predictor of duration of hospital. The other parameter which was found as an independent predictor was coronary revascularisation during hospital stay. These findings were supported by earlier studies done by Wu et al., Bindra et al., Sabatine et al., Reincke et al., and Lee et al. They also reported that anaemic patients had longer duration of hospital stay and anaemia was independent predictor of in-hospital and 1-year mortality [8,16–18,25].

Main limitation of the study was duration of study which was short (3 months). With the present study, prognostic information could not be provided due to the limited sample size and lack of follow-up data after discharge. As our hospital is a multi specialty tertiary care centre receiving more severe and complicated cases, there may be referral bias due to which the study may not reflect the true prevalence. Aetiology of anaemia could not be assessed completely due to patient factors and limitation in the tests available.

Overall prevalence of anaemia was high in the study population. The most common cause of anaemia was nutritional deficiency followed by chronic kidney disease and blood loss due to antiplatelet agents. The presence of anaemia is a risk factor for increased morbidity and in-hospital mortality in these patients. As anaemia is a relatively frequent condition which remains often under-diagnosed and untreated, its prevention and earlier detection may lead to improved outcomes. Hence further research is needed to examine the role of correction of anaemia in reducing long-term morbidity and mortality in these patients.