A case of persistent anemia and alcohol abuse
Gemma Lewis, Matthew P Wise, Christopher Poynton and Andrew Godkin* About the authors
Correspondence *Department of Integrated Medicine, University Hospital of Wales, Heath Park, Cardiff CF14 4XW, UK
Email andrew.godkin@cardiffandvale.wales.nhs.uk
Summary
Background A 56-year-old male with a history of excess alcohol consumption for over 10 years and type 2 diabetes mellitus (diagnosed 14 years previously) presented to the accident and emergency department with severe anemia and a 1-day history of nausea and 'coffee ground' vomiting. He had been admitted to hospital on several occasions in the previous 18 months, primarily because of anemia, and had received 30 units of transfused blood during this period. Previous extensive investigations included six esophagogastroduodenoscopies and a colonoscopy, a barium follow-through study, and a radionucleotide Meckel's scan. The prior working diagnosis was anemia secondary to blood loss.
Investigations Laboratory investigations (full blood count [including reticulocytes], microscopic blood film examination, hematinics, liver function tests with direct and indirect bilirubin measurement, prothrombin time, and lactate dehydrogenase level), transjugular liver biopsy and bone-marrow biopsy.
Diagnosis Alcohol-related anemia caused by acute hemolysis, sideroblastic anemia and cirrhosis.
Management Correction of anemia by blood transfusion (6 units), and prevention of recurrence by strict abstinence from alcohol.
The case
A 56-year-old male was admitted to hospital via the accident and emergency department with a 1-day history of nausea and 'coffee ground' vomiting. In the previous 18 months the patient had undergone extensive investigations for anemia that had included six esophagogastroduodenoscopies and a colonoscopy. A barium follow-through study (which demonstrated Meckel's diverticulum) and a normal radioisotope Meckel's scan had also been performed. The working diagnosis was anemia secondary to blood loss and, because no notable source of bleeding was identified in the upper gastrointestinal tract or colon, a capsule endoscopy was planned to look for a cause of bleeding in the small intestine. The patient had been admitted to hospital on several occasions during this 18-month period and had received 30 units of transfused blood. His medical history included colonic polyps (cleared during three colonoscopies in a 5-year period before the onset of anemia) and diet-controlled type 2 diabetes mellitus, which had been diagnosed 14 years previously when it had also been noted that his alcohol intake was high. At the time of his current admission, the patient stated that his alcohol intake during the past 2 years had been 1 unit a day.
Examination revealed a fully conscious and lucid individual who seemed to be well nourished and who had a normal BMI. The patient's breath smelt of alcohol, however, and recent high levels of alcohol consumption were confirmed with a measured blood alcohol level of 286 mg per 100 ml blood. Icterus, conjunctival pallor, bruising and multiple spider nevi were present. The patient's blood pressure was 110/50 mmHg with no postural drop and his pulse was 90 beats/min in sinus rhythm. The results of a chest examination were unremarkable. No organomegaly was palpable in the abdomen and the results of a rectal examination were normal.
A full blood count confirmed the presence of anemia; the patient had a hemoglobin level of 66 g/l (normal 130–165 g/l), mild macrocytosis (mean corpuscular volume <101 fl) and thrombocytopenia (66
109 platelets/l blood). A blood film was taken from the patient and was found to be markedly abnormal owing to the presence of spherocytes, acanthocytes and stomatocytes (Figure 1). Hematinics were normal, but the results of liver function tests were abnormal (Table 1). Splenomegaly was excluded by ultrasound examination of the abdomen; however, the texture of the liver appeared coarse. A sudden rise in the patients' serum bilirubin concentration to 148
mol/l (8.7 mg/dl) on the third day after admission combined with the presence of microspherocytosis (Figure 1) raised the possibility of hemolysis. This diagnosis was supported by the finding of an increased unconjugated bilirubin level (from approximately 20–30% in the patient to a peak of 74%), an elevated lactate dehydrogenase level of 557 U/l (normal values), and a low haptoglobin level of 50
mol/l (500 mg/dl) (normal values).
Figure 1 Blood film from the case patient
Spherocytes (spherical erythrocytes lacking the normal area of central pallor), acanthocytes (erythrocytes with irregularly spaced projections) and stomatocytes (erythrocytes with an oval or rectangular area of central pallor) and microspherocytosis can be seen. Spherocytes are typically found in patients with hemolytic anemia, whereas acanthocytes and stomatocytes can be present in patients with liver disease. The neutrophil present shows vacuolation characteristic of acute alcohol ingestion (May-Giemsa stain, original magnification
1,000).
Table 1 The patient's laboratory test results
Full tableFigures & Tables indexDownload Power Point slide (70K)
Further examination of the patient's blood indices revealed a marked increase in both the percentage of reticulocytes and the absolute reticulocyte count (Table 2). A direct Coombs test was negative and the fasting serum lipid profile was normal. All these results suggested that hemolysis might have been contributing substantially to the anemia. The results of hemoglobin and bilirubin measurements in the patient recorded during the previous 18 months were then reviewed and revealed similar episodes of a rapidly falling hemoglobin level with a sudden increase in serum bilirubin concentration. These and other serially recorded laboratory indices are shown in Figure 2.
Figure 2 Laboratory variables for the 18 months preceding the current admission of the case patient
The patient's ferritin level was initially low, which might have reflected chronic blood loss and/or poor dietary intake. A steady increase in the ferritin level thereafter to >225 pmol/l (>100 ng/ml) reflects the replenishment of iron stores by the multiple blood transfusions received by the patient. The arrows indicate examples where a sudden fall in hemoglobin concentration is mirrored by a rise in bilirubin, reflecting bouts of hemolysis.
Full figure and legend (31K)Figures & Tables indexDownload Power Point slide (78K)Table 2 Comparison of hemoglobin concentration and red blood cell counts with the reticulocyte response in the case patient.a
Full tableFigures & Tables indexDownload Power Point slide (85K)
Underlying liver disease was diagnosed from a transjugular liver biopsy sample, which revealed micronodular cirrhosis with low-grade steatohepatitis. Bone-marrow aspiration was also undertaken and showed mild erythroid hyperplasia (30% of cells were erythroblasts), in keeping with a bone-marrow response to blood loss. A Perl's iron stain of the bone marrow revealed the conspicuous presence of ringed sideroblasts and no evidence of iron deficiency (Figure 3). Both the liver pathology and the sideroblastic bone-marrow changes were consistent with a high alcohol intake.
Figure 3 Histological image of a bone-marrow aspirate from the case patient that demonstrates the presence of ringed sideroblasts
Abnormal accumulation of hemosiderin in the mitochondria surrounding the nucleus can be seen (Perl's iron stain, original magnification
1,000).
The patient's admission was complicated by another unwitnessed hematemesis after vomiting; no obvious melena was present but a stool sample gave a positive fecal occult blood test. A repeat esophagogastroduodenoscopy (that revealed mild portal gastropathy) and a radioactively labeled red blood cell scan (to identify intraluminal aggregations of labeled blood cells that would suggest gastrointestinal hemorrhage) were undertaken. Neither investigation revealed substantive intestinal blood loss. The patient was transfused with a total of 6 units of blood during his 10-day admission. A subsequent pre-arranged capsule endoscopy performed shortly after the patient was discharged revealed no abnormalities. The patient completely abstained from alcohol for a 6-month follow-up period as an outpatient and his hemoglobin level was maintained at over 100 g/l for the first time in 2 years. After 6 months the patient traveled abroad and was no longer seen in the clinic.
The anemia in this case was prompted by an acute alcohol binge of a patient with a history of chronic alcohol consumption, and could be attributed to acute hemolysis, sideroblastic anemia and cirrhosis.
Discussion of diagnosis
Anemia is defined as a reduction in blood hemoglobin concentration (and hence also a reduced hematocrit), to <135 g/l in adult males and <115 g/l in adult females. The main causes of acquired anemia can be divided into loss of blood (hemolysis or hemorrhage) or inadequate bone-marrow function (e.g. ineffective erythropoiesis, sideroblastic changes, hematinic deficiency, chronic inflammation or malignancy). A simple, pragmatic algorithm that can be used for the investigation of anemia in alcoholics is shown in Figure 4. It is important to remember that alcohol can complicate the interpretation of the results of certain common laboratory tests that are used to investigate the causes of anemia. For example, although the mean corpuscular volume is increased by recognized causes that include vitamin B12 deficiency or folic acid deficiency, it is also raised by high alcohol intake alone. In addition, the ferritin level is raised by inflammation of the liver (as seen after high alcohol intake), which can mask the decrease in ferritin that occurs with iron deficiency.
Figure 4 An algorithm giving a straightforward approach to the investigation of anemia in alcoholics
Abbreviations: AST, aspartate transaminase; EGD, esophagogastroduodenoscopy; LDH, lactate dehydrogenase.
Full figure and legend (58K)Figures & Tables indexDownload Power Point slide (106K)The etiology of anemia in alcoholics is complex and often multifactorial, causes include a combination of poor nutrition, chronic inflammation, blood loss, liver dysfunction and ineffective erythropoiesis.1 Effective erythropoiesis can be indicated by the reticulocyte count.2 It is important to be aware of both the absolute number of reticulocytes and the percentage of reticulocytes, as marked anemia can misleadingly increase the relative percentage. Thus, if the lifespan of the red blood cells is short, the percentage of reticulocytes will seem to be higher than the absolute numbers would suggest (reflecting the reduced number of mature red blood cells). Clearly, sequential reticulocyte counting is useful for obtaining information on the lifespan of RBCs in these situations.
The accurate assessment of anemia in alcoholics requires identification of the relative contributions of the direct toxic effects of alcohol and poor nutrition on the synthesis, function and survival of red blood cells, and of the role of underlying chronic disease such as cirrhosis or chronic sepsis. The presence or absence of chronic liver disease can be obvious from the patient's medical history and examination, but a liver biopsy sample is often useful for typing and staging of disease. Confirmation of chronic liver disease can then prompt the search for other factors that might be contributing to anemia. These contributing factors include abnormalities of red blood cells (e.g. target cells, spur cells) associated with perturbations in lipid metabolism, hemolysis, anemia of chronic disease, acute or chronic blood loss from varices and/or gastropathy (which is often exacerbated by coexistent coagulopathy and hypersplenism).
Alcohol can be directly toxic to the bone marrow, as suggested by pre-erythroblast vacuolation or sideroblastic changes. Alcoholism and the debilitating factors with which it is often associated (e.g. chronic sepsis and poor nutrition) frequently causes marked anemia even if liver disease is absent (e.g. sideroblastic anemia, megaloblastic anemia and the anemia of chronic inflammation). Thus, a proportion of alcoholics have both reduced erythrocyte production in the bone marrow and accompanying reduced red blood cell survival time. Studies have shown that bone-marrow biopsy samples from alcoholics are abnormal in the majority of cases: approximately two-thirds reveal megaloblastic and/or sideroblastic changes and fewer than a fifth are suggestive of any underlying iron deficiency.3, 4 If doubt remains about the diagnosis after an initial careful inspection of the peripheral blood film, a bone-marrow aspirate can provide useful information and, equally importantly, exclude more-sinister causes of the anemia (e.g. malignancy or other blood dyscrasias).
Some diagnoses are pertinent to both excess alcohol intake and liver disease, for example the marked hemolysis of spur cell anemia that can be seen in patients with jaundice and severe alcoholic cirrhosis5 and the mild hemolysis found with steatosis, hypertriglyceridemia and acute alcohol ingestion (Zieve syndrome).6 The hemolytic crises prompted by alcoholic binges are poorly understood. The case described here exhibited features similar to those of Zieve syndrome, even though his serum triglyceride levels were normal; these features might, therefore, have reflected the effect of alcohol directly on the erythrocyte.7
The present case demonstrates the often complex mechanism of anemia in alcoholics. The etiology included both reduced red blood survival and reduced red blood cell production (e.g. bone-marrow dysfunction) and hence the anemia was the end result of more than one process. Table 3 summarizes some of the different pathological processes that can contribute to anemia in alcoholics, with or without accompanying chronic liver disease.
Table 3 Summary of the different factors that can contribute to anemia and/or a low hematocrit in alcoholic patients
Full tableFigures & Tables indexDownload Power Point slide (75K)
Treatment and management
Abstention from alcohol is key to halting the progression of alcohol-related anemia. Acute withdrawal symptoms can be managed by a short course (e.g. 3–7 days) of gradually reduced doses of benzodiazepines, such as chlordiazepoxide. Abstention is easily maintained in some patients, especially when physical illness emerges, whereas other individuals might require more-specific psychological counseling to help manage their addiction. Medication to reduce the urge to drink has been shown to be superior to placebo for the maintenance of abstinence. For example, acamprosate and/or coadministered naltrexone for 3 months reduced relapse rates from 75% (placebo) to 50–60% (either drug alone) and 25% (combination treatment).8 Disulfiram, which causes extremely unpleasant adverse effects if alcohol is ingested, is occasionally still used to encourage abstinence in selected patients. Anemia is corrected in alcoholics not only by stopping alcohol intake, but also by treating other underlying conditions (e.g. nutritional deficits). Blood transfusion is reserved for patients with symptomatic anemia, or for patients with severe hemorrhage with cardiovascular compromise. If the anemia results from a true iron deficiency and blood loss from the gastrointestinal tract, this needs to be investigated appropriately. In the present case, the patient managed to stop drinking with close monitoring and counseling on an outpatient basis and without the need for pharmacological adjuncts.
Conclusions
Alcohol abuse, in the form of both binge drinking and high chronic intake, is becoming increasingly prevalent in many populations and poses a mounting challenge to health-care professionals because of the wide range of physical and psychological illnesses it can cause.9
This case demonstrates several important learning points regarding the management of alcohol-related disease. Firstly, it highlights how alcoholism can be complicated by both the problems associated with chronic alcohol ingestion (e.g. cirrhosis, bone-marrow dysfunction, psychological problems) and the acute problem of binge drinking.
Secondly, the case illustrates how misconceptions about the causes of alcohol-related anemia can lead physicians down an erroneous diagnostic pathway. A common misconception is that iron deficiency and blood loss are the major factors contributing to the anemia. Admitting physicians often ask for endoscopic studies to be performed at an early stage in cases of alcohol-related anemia. This decision can be driven in part by the fear of varices (which are often a consequence of chronic liver disease) and the associated morbidity and mortality of variceal hemorrhage. The numerous endoscopic investigations previously undertaken in this case demonstrate this point succinctly and emphasize the difficulties that can be encountered when trying to obtain a correct diagnosis.
Finally, this case highlights how anemia in alcoholics is often a complicated multifactorial disease process. After a complete medical history of the patient has been taken and a full examination performed, essential first-line investigations after the presence of anemia is established should include hematinics and a blood film with reticulocyte count; this information can then be used to guide further investigations and management.
References
- Colman N and Herbert V (1980) Hematologic complications of alcoholism: overview. Semin Haematol 17: 164–176 | ChemPort |
- Rowan RM et al. (1996) The reticulocyte count: progress towards the resurrection of a useful clinical test. Clin Lab Haematol 18 (Suppl 1): S3–S8
- Eichner ER and Hillmann RS (1971) The evolution of anaemia in alcoholic patients. Am J Med 50: 218–232 | Article | PubMed | ISI | ChemPort |
- Savage D and Lindenbaum J (1986) Anaemia in alcoholics. Medicine (Baltimore) 65: 322–338 | PubMed | ChemPort |
- Doll D and Doll N (1982) Spur cell anaemia. South Med J 75: 1205–1210 | PubMed | ISI | ChemPort |
- Zieve L (1958) Jaundice, hyperlipidaemia and haemolytic anaemia: a heretofore unrecognised syndrome associated with alcoholic fatty liver and cirrhosis. Ann Intern Med 48: 471–496 | PubMed | ISI | ChemPort |
- Goebel KM and Schneider J (1981) Erythrocyte membrane fluidity, lipid peroxidation and lysis in alcoholic liver disease. Acta Biol Med Ger 40: 571–576 | PubMed | ISI | ChemPort |
- Kiefer F et al. (2003) Comparing and combining naltrexone and acamprosate in relapse prevention of alcoholism: a double-blind placebo controlled study. Arch Gen Psychiatry 60: 92–99 | PubMed | ISI | ChemPort |
- Piromohamed M and Gilmore I (2000) Alcohol abuse and the burden on the NHS—time for action. J R Coll Physicians Lond 34: 161–162 | PubMed |
Competing interests
The authors declared no competing interests.
Contact the journal about this article
Subject areas under which this article appears: Liver | Gastrointestinal bleeding



