From The Archives – Deciphering Fact from Fiction in Hypoglycemia

January 6, 2011

Please enjoy this post from the Clinical Correlations archives first posted March 26, 2009

By: Melissa Price, MD

Faculty Peer Reviewed

A 42 year-old male phlebotomist with a history of anxiety presented to the emergency room complaining of four hours of dizziness and diaphoresis. He denied taking any medications. His vitals were stable, his physical exam was significant for a lethargic, diaphoretic young man without focal findings, and his fingerstick value was 43mg/dL. His chest X-ray, EKG, and laboratory results, with the exception of plasma glucose, were within normal limits.

Why is this previously healthy patient suddenly hypoglycemic? Occasionally, a patient presents himself with such a curious constellation of signs and symptoms that even the most experienced and devoted physician is left in an inquisitive stupor. It is from these patients that we often learn the most, not only of the intricacy of clinical medicine, but of the human mind.

This article will address ways to determine causes of hypoglycemia with a focus on factitious disorder. Initial evaluation of a hypoglycemic patient requires distinguishing patients who are presenting with known diabetes from those without, as this will drastically alter the differential diagnosis. For example, diabetic patients can suffer from hypoglycemia due to recent medication dose adjustments, dosing errors, brittle diabetes, dietary or activity changes, and changes in renal function affecting drug metabolism. A basic differential diagnosis should include insulinoma (isolated or MEN), brittle diabetes, autoimmune hypoglycemia, unintentional medication misuse (think of an elderly nondiabetic patient with a diabetic spouse), artifactual or lab error (if pt is asymptomatic), suicide attempt, and factitious hypoglycemia. Additionally, rare disorders can result in adult hypoglycemia such as noninsulinoma pancreatogenous hypoglycemia and nonislet cell tumor-induced hypoglycemia amongst others.

Symptoms of hypoglycemia such as shakiness, diaphoresis, palpitations, and paresthesias for example, should correlate with low blood glucose. One can seek out Whipple’s triad in a hypoglycemic patient who suffers symptoms of low blood sugar with a correlated low glucose and improvement of those symptoms with glucose correction. Careful attention must be paid to the patient’s history; a sudden hypoglycemic occurrence should raise concern for a possible factitious disorder. Purposeful induction of hypoglycemia classically occurs in middle-aged women, often with a psychiatric history, employed in healthcare fields, who use or misuse insulin and sulfonylureas. Though our patient is a male, he seems to fit into this demographic but, one must not assume factitious disorder without more objective evidence. Various laboratory data must be obtained in addition to a thorough history, exam, and clinical judgment.

The patient’s glucose level was 43mg/dL on initial fingerstick, and plasma glucose was 41 mg/dl. His glucose normalized readily with IV dextrose and orange juice. He was observed in the ER and discharged after an uneventful stay.

One week later, the patient presented similarly and was admitted to the hospital for further work-up. His glucose levels remained within normal limits overnight. The patient expressed significant fear of repeated hypoglycemic episodes if discharged and a 72-hr fast was carried out for completeness. Laboratory results from a single sample of blood included a glucose of 38mg/dL with elevated levels of both insulin and C-peptide. A serum assay screen for sulfonylureas was positive.

Pancreatic beta-cells create insulin by the conversion of pro-insulin to equimolar amounts of insulin and C-peptide. Plasma or serum glucose should be tested first and measured at the time of a symptomatic hypoglycemic episode in order to document true hypoglycemia. If glucose concentrations are normal, your workup may end here. If abnormal, a serum insulin level and C-peptide level should be checked with repeat glucose testing at the time of a hypoglycemic event. C-peptide is helpful to distinguish endogenous from exogenous insulin sources. Since insulin and C-peptide are secreted together in an equimolar fashion, an elevated C-peptide level in hypoglycemia suggests endogenous rather than exogenous hyperinsulinemia. However, as illustrated in our case study, factitious hypoglycemia secondary to sulfonylurea-stimulated insulin secretion biochemically mimics an insulinoma. Thus, both insulinoma and sulfonylurea-induced hypoglycemia will result in elevated insulin (but in an insulinoma, serum insulin levels are usually < 100 uU/ml) and C-peptide levels. Here, clinical history and sulfonylurea assays, will be essential.

With exogenous insulin use in Type 2 Diabetes Mellitus and nondiabetics, the C-peptide level is appropriately suppressed. Patients with Type 1 Diabetes Mellitus are severely insulin deficient with almost undetectable C-peptide levels, but if C-peptide levels remain low during the hypoglycemia lab draw, endogenous hyperinsulinism can be ruled out nonetheless. To complicate the diagnosis, insulin antibodies can bind to the insulin receptor, inappropriately releasing insulin. Diabetic patients treated with insulin frequently develop anti-insulin antibodies and in autoimmune hypoglycemia syndrome, insulin antibodies are spontaneously generated. These antibodies interfere with serum insulin and C-peptide measurements and one can see a wide spectrum of values.

As in our case study, a prolonged supervised fast (up to 72 hours) in a controlled setting is a classic, reliable test used to evaluate hypoglycemia in warranted cases. However, these fasts are costly and complicated so they should be avoided as a first-line diagnostic aid. Imaging can be pursued if biochemical evidence of an insulin- secreting tumor is present. If any data collected provide evidence of surreptitious hypoglycemia, the patient should be confronted in a nonaccusatory manner and an appropriate psychiatric referral should be made. The notion that a patient would intentionally assume a sick role first appeared in the medical literature as early as second century A.D.. The current DSM-IV-TR divides factitious disorder into three major subtypes based on whether the presenting signs and symptoms are psychological, physical, or a combination of the two. One of the most outstanding consequences of factitious disorder is the estimated forty-million dollars it costs the United States to care for these patients annually. Thus, it is imperative that physicians be aware of this disease, its various presentations, and efficient ways to delineate fact from fiction using a scientific, methodical approach.

Faculty Peer Reviewed by Julie Probst MD

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders.4th rev edition, Amer. Psychiatric Assoc. Washington, DC: 2000.

Feldman, M., Ford C. V., Reinhold, T. Patient or Pretender: Inside the Strange World of Factitious Disorders. John Wiley & Sons Inc, 1993.

Grunberger, G, Weiner, JL, Silverman, R. et al. Factitious hypoglycemia due to surreptitious administration of insulin. Ann Intern Med. 1988; 108: 252. Hypoglycemia [internet], [updated 2008 Apr 22; cited 2008 Oct 17.

Hirshberg, B et al. Clinical case seminar: repaglinide-induced factitious hypoglycemia. J Clin Endocrinol Metab. 2001; 86 (2): 475- 477.

Service, John F.. Diagnostic approach to hypoglycemia in adults [internet]. 2007 Aug 20 [updated 2008 May 31; cited 2008 Oct 17] . Available from

Service, John, F. Etiology and diagnosis of factitious hypoglycemia [internet]. 2004 Aug 18 [updated 2008 May 31; cited 2008 Oct 17] . Available from