ACROCYANOSIS AS A CAUSE OF PSEUDOHYPOGLYCEMIA

To the Editor: Capillary blood glucose measurement is the main tool for the monitoring of diabetes mellitus. The fingertips are the preferred site for blood sampling, but this method may lead to erroneous estimations in several situations. An 84-year-old woman was referred to our geriatric hospital on September 23, 2008, for delirium. Acute renal failure was attributed to severe dehydration and probable overuse of nonsteroidal anti-inflammatory drugs. She was known for chronic renal failure, type 2 diabetes mellitus without drug treatment, and probable Alzheimer's disease. Her medication at home included lisinopril, hydrochlorothiazide, atorvastatin, and a multivitamin complex. On physical examination, her blood pressure was 120/60 mmHg, heart rate 86 beats per minute, and temperature 36.9°C. She presented with livedo reticularis of all four limbs and bilateral digital acrocyanosis. Laboratory test results indicated anemia (hemoglobin 9.6 g/dL, hematocrit 31.2%, mean corpuscular volume 100.3 fL), severe renal failure (creatinine 4.3 mg/dL), and hypoalbuminemia (2.6 g/dL). Liver function tests were normal. Venous plasma glucose was 96 mg/dL (at 10:30 a.m.). Pulse oximetry was unreadable because of signal failure. On arterial blood gas analysis, partial pressure of oxygen was 20.8 kPa, oxygen saturation was 99% (under oxygen 2 L/min), and metabolic acidosis was measured (pH=7.28, partial pressure of carbon dioxide=3.3 kPa, bicarbonate=11.7 mmol/L). At 1 p.m. (before the laboratory test results were available), a fingerstick glucose value of 25 mg/dL was obtained. The patient did not complain of adrenergic hypoglycemia symptoms, but the history was unreliable because of the delirium. She was immediately treated with oral and intravenous glucose (10% glucose, 250 mL/2 hours), without effect on delirium symptoms. We observed persisting low fingerstick glucose values (<60 mg/dL) over the subsequent 48 hours. Sepsis, renal failure, adrenal insufficiency, and malnutrition were considered, but the presence of acrocyanosis rapidly led to a suspicion of pseudohypoglycemia, and hypoglycemia was not confirmed on earlobe or venous blood (Table 1). The patient had no history of Raynaud's disease, but family members had noted acrocyanosis for the previous month. Nailfold capillary microscopy revealed no specific signs of a connective tissue disease. Antiphospholipid antibodies were negative. Peripheral arterial disease of the upper limbs was excluded using duplex ultrasonography. Over 6 days after admission, the acrocyanosis and livedo reticularis progressively disappeared, in parallel with the correction of the metabolic acidosis and the improvement in renal function. Glucose concentrations were measured simultaneously on blood from the fingertips and the earlobe and on venous plasma on day 3 and after disappearance of the acrocyanosis (Table 1). Capillary values were obtained using an Ascensia Contour Glucose monitor (Bayer HealthCare, Tarrytown, NY) using a plasma calibration. The initial fingertip values were lower than earlobe and venous values and were partly corrected by immersing the hand in warm water for 5 minutes. The values from all three sites were normal and nearly identical after correction of the acrocyanosis. Acrocyanosis induces pseudohypoglycemia, most likely through impaired blood flow in the digital microcirculation, leading to local increase in glucose consumption. Fingertip blood glucose has been shown to underestimate systemic blood glucose in several situations affecting the microcirculation, including severe hypotension (shock or severe dehydration),1 Raynaud's disease,2, 3 and upper extremity peripheral arterial disease.4 This problem must be distinguished from pseudohypoglycemia encountered in leukocytosis or polycythemia due to excessive glycolysis in vitro. This condition is associated with normal venous and capillary values when blood glucose is measured without delay after bloodletting.5 Fingertips are the preferred site for blood glucose monitoring. Measurement at less painful sites, such as the forearm or the thigh has been proposed, but because of lower blood flow, equilibration with the venous blood glucose is slower at these sites, possibly leading to late detection of hypo- or hyperglycemia6, 7 Thus, measurements at the fingertip are more accurate in states of rapid blood glucose fluctuations,8 but our observation emphasizes that this site can be unreliable in cases of suspected microcirculatory changes. Venous or earlobe capillary blood glucose testing should then be preferred. The artifactual character of hypoglycemia might have been suspected more rapidly by the absence of conventional, adrenergic hypoglycemia symptoms, but delirium made the history unreliable and was considered a possible hypoglycemia symptom in itself, leading to unnecessary treatment and investigations for hypoglycemia. An accurate and reliable biochemical diagnosis of hypoglycemia is particularly important in cognitively impaired patients. The presence of acrocyanosis or other microcirculatory changes should be considered when prescribing blood glucose monitoring for hospitalized patients or when confronted with unexpectedly low fingertip blood glucose values. Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this letter. Author Contributions: Each author participated in data collection, presentation of the results, and writing and correcting the letter. Sponsor's Role: No sponsor was involved.