Evaluation and Management of Adult Hypoglycemic
Disorders: An Endocrine Society Clinical Practice
Guideline
Philip E. Cryer, Lloyd Axelrod, Ashley B. Grossman, Simon R. Heller, Victor M. Montori,
Elizabeth R. Seaquist, and F. John Service
Washington University School of Medicine (P.E.C.), St. Louis, Missouri 63110; Massachusetts General Hospital and
Harvard Medical School (L.A.), Boston, Massachusetts 02114; Barts and the London School of Medicine, Queen Mary
University of London (A.B.G.), London E1 2AD, United Kingdom; University of Sheffield (S.R.H.), Sheffield S10 2TN,
United Kingdom; University of Minnesota (E.R.S.), Minneapolis, Minnesota 55455; and Mayo Clinic (V.M.M., F.J.S.),
Rochester, Minnesota 55905
Objective: The aim is to provide guidelines for the evaluation and management of adults with
hypoglycemic disorders, including those with diabetes mellitus.
Evidence: Using the recommendations of the Grading of Recommendations, Assessment, Devel-
opment and Evaluation (GRADE) system, the quality of evidence is graded very low (QEEE), low
(QQEE), moderate (QQQE), or high (QQQQ).
Conclusions:We recommend evaluation and management of hypoglycemia only in patients in
whomWhipple’s triad—symptoms, signs, or both consistent with hypoglycemia, a low plasma
glucose concentration, and resolution of those symptoms or signs after the plasma glucose
concentration is raised—is documented. In patients with hypoglycemia without diabetes mel-
litus, we recommend the following strategy. First, pursue clinical clues to potential hypogly-
cemic etiologies—drugs, critical illnesses, hormone deficiencies, nonislet cell tumors. In the
absence of these causes, the differential diagnosis narrows to accidental, surreptitious, or even
malicious hypoglycemia or endogenous hyperinsulinism. In patients suspected of having en-
dogenous hyperinsulinism, measure plasma glucose, insulin, C-peptide, proinsulin, �-hydroxy-
butyrate, and circulating oral hypoglycemic agents during an episode of hypoglycemia and
measure insulin antibodies. Insulin or insulin secretagogue treatment of diabetes mellitus is
the most common cause of hypoglycemia. We recommend the practice of hypoglycemia risk
factor reduction—addressing the issue of hypoglycemia, applying the principles of intensive
glycemic therapy, and considering both the conventional risk factors and those indicative of
compromised defenses against falling plasma glucose concentrations—in persons with
diabetes. (J Clin Endocrinol Metab 94: 709–728, 2009)
Summary of Recommendations
1.0 Workup for a hypoglycemic disorder
1.1We recommend evaluation andmanagement of hypoglyce-
miaonly inpatients inwhomWhipple’s triad—symptoms, signs,or
both consistent with hypoglycemia, a low plasma glucose concen-
tration, and resolution of those symptoms or signs after the plasma
glucose concentration is raised—is documented (1QQQQ).
2.0 Evaluation and management of hypoglycemia in
persons without diabetes mellitus
2.1 Compared with a much less thorough workup, we rec-
ommend the following strategy in patients with hypoglycemia
without diabetes mellitus (1QQQE):
• Review the history, physical findings, and all available
laboratory data seeking clues to specific disorders—drugs,
ISSN Print 0021-972X ISSN Online 1945-7197
Printed in U.S.A.
Copyright © 2009 by The Endocrine Society
doi: 10.1210/jc.2008-1410 Received July 2, 2008. Accepted December 8, 2008.
First Published Online December 16, 2008
Abbreviations: CSII, Continuous sc insulin infusion; HAAF, hypoglycemia-associated au-
tonomic failure; HbA1C, hemoglobin A1C; MDI, multiple daily insulin injection; MEN-1,
multiple endocrine neoplasia, type 1; MRI, magnetic resonance imaging; NIPHS, nonin-
sulinoma pancreatogenous hypoglycemia syndrome; T1DM, type 1 diabetes mellitus;
T2DM, type 2 diabetes mellitus.
S P E C I A L F E A T U R E
C l i n i c a l P r a c t i c e G u i d e l i n e
J Clin Endocrinol Metab, March 2009, 94(3):709–728 jcem.endojournals.org 709
by Xiaohui Yu on March 30, 2009 jcem.endojournals.orgDownloaded from
critical illnesses, hormone deficiencies, nonislet cell
tumors.
• When the cause of the hypoglycemic disorder is not evident,
i.e. in a seemingly well individual, measure plasma glucose,
insulin, C-peptide, proinsulin, and �-hydroxybutyrate con-
centrations and screen for oral hypoglycemic agents, during
an episode of spontaneous hypoglycemia, and observe the
plasma glucose response to iv injection of 1.0 mg glucagon.
These steps will distinguish hypoglycemia caused by endog-
enous (or exogenous) insulin from that caused by othermech-
anisms. Also, measure insulin antibodies.
• When a spontaneous hypoglycemic episode cannot be ob-
served, formally recreate the circumstances in which symp-
tomatic hypoglycemia is likely to occur, i.e. during a fast of up
to72hor after amixedmeal.The findings of symptoms, signs,
or both with plasma concentrations of glucose less than 55
mg/dl (3.0mmol/liter), insulin of at least 3.0�U/ml (18 pmol/
liter), C-peptide of at least 0.6 ng/ml (0.2 nmol/liter), and
proinsulin of at least 5.0 pmol/liter document endogenous
hyperinsulinism; �-hydroxybutyrate levels of 2.7 mmol/liter
or less and an increase in plasma glucose of at least 25 mg/dl
(1.4 mmol/liter) after iv glucagon indicate mediation of the
hypoglycemia by insulin (or by an IGF).
• In a patient with documented fasting or postprandial endog-
enous hyperinsulinemic hypoglycemia, negative screening for
oral hypoglycemic agents, and no circulating insulin antibod-
ies, conduct procedures for localizing an insulinoma. These
may include computed tomography or magnetic resonance
imaging (MRI), transabdominal and endoscopic ultrasonog-
raphy, and, if necessary, selective pancreatic arterial calcium
injectionswithmeasurements of hepatic venous insulin levels.
• Tailor treatment to the specific hypoglycemic disorder, taking
into account the burden of hypoglycemia on patient well-
being and patient preferences.
3.0 Evaluation and management of hypoglycemia in
persons with diabetes mellitus
3.1We suggest that persons with diabetes become concerned
about the possibility of developing hypoglycemia when the self-
monitored blood glucose concentration is falling rapidly or is no
greater than 70 mg/dl (3.9 mmol/liter) (2QEEE).
3.2Given the established long-termmicrovascular benefits
of glycemic control, we recommend that the therapeutic gly-
cemic goal be the lowest mean glycemia [e.g. hemoglobin A1c
(HbA1C)] that can be accomplished safely in a given patient at
a given point in the progression of that individual patient’s
diabetes (1QQQQ).
3.3 We recommend that the prevention of hypoglycemia in
diabetes involve addressing the issue in each patient contact and,
if hypoglycemia is a problem,making adjustments in the regimen
based on review and application of the principles of intensive
glycemic therapy—diabetes self-management (supported by ed-
ucation and empowerment), frequent self-monitoring of blood
glucose, flexible and appropriate insulin or insulin secretagogue
regimens, individualized glycemic goals, and ongoing profes-
sional guidance and support—and consideration of each of the
known risk factors for hypoglycemia (1QQQE).
3.4 We recommend that both the conventional risk factors
and those indicative of compromised defenses against hypogly-
cemia be considered in a patient with recurrent treatment-in-
duced hypoglycemia (1QQQQ). The conventional risk factors
are excessive or ill-timed dosing of, or wrong type of, insulin or
insulin secretagogue and conditions underwhich exogenous glu-
cose delivery or endogenous glucose production is decreased,
glucose utilization is increased, sensitivity to insulin is increased,
or insulin clearance is decreased. Compromised defenses against
hypoglycemia are indicated by the degree of endogenous insulin
deficiency, a history of severe hypoglycemia, hypoglycemia un-
awareness, or both as well as recent antecedent hypoglycemia,
prior exercise or sleep, and lower glycemic goals per se.
3.5 With a history of hypoglycemia unawareness (i.e. recur-
rent hypoglycemia without symptoms), we recommend a 2- to
3-wk period of scrupulous avoidance of hypoglycemia, with the
anticipation that awareness of hypoglycemiawill return inmany
patients (1QQEE).
3.6Unless the cause is easily remediable,we recommend that an
episode of severe hypoglycemia should lead to a fundamental re-
view of the treatment regimen and the glycemic goals (1QQQQ).
3.7 We recommend that urgent treatment of hypoglycemia
should be accomplished by ingestion of carbohydrates if that is
feasible, or by parenteral glucagon or glucose if it is not feasible
(1QQQQ).
Method of Development of Evidence-Based
Recommendations
The Task Force followed the approach recommended by the
Grading of Recommendations, Assessment, Development, and
Evaluation system (1) with guidance from the methodologist
(V.M.M.).Adetaileddescriptionof this grading schemehasbeen
published (2). In brief, strong recommendations use the phrase
“we recommend” and the number 1, and weak recommenda-
tions use the phrase “we suggest” and the number 2. The Task
Force has confidence that patients who receive care according to
the recommendations will derive, on average, more good than
harm. Suggestions require more careful consideration of the pa-
tient’s circumstances, values, andpreferences.Cross-filled circles
(Q) indicate the quality of the evidence:QEEE denotes very low
quality evidence;QQEE, low quality;QQQE, moderate quality;
and QQQQ, high quality. The quality of the evidence indicates
the panel’s confidence that the estimates of risks and benefits
associated with the recommended course of action compared
with an alternative course of action are correct and unlikely to
change importantly with new research.
Linked to each recommendation is a description of the Evi-
dence, the Values that panelists considered in making the rec-
ommendation (when making these explicit was necessary), and
Remarks, a section inwhich panelists offer technical suggestions
for testing conditions. The latter come from the unsystematic
observations of the panelists and should, therefore, be consid-
ered suggestions.
710 Cryer et al. Evaluation and Management of Adult Hypoglycemia J Clin Endocrinol Metab, March 2009, 94(3):709–728
by Xiaohui Yu on March 30, 2009 jcem.endojournals.orgDownloaded from
1.0 Workup for a Hypoglycemic Disorder
Recommendation
1.1We recommend evaluation and management of hypogly-
cemia only in patients in whom Whipple’s triad—symptoms,
signs, or both consistent with hypoglycemia, a low plasma glu-
cose concentration, and resolution of those symptoms or signs
after the plasma glucose concentration is raised—is documented
(1QQQQ).
1.1 Evidence
Clinical hypoglycemia is a plasma (or serum) glucose con-
centration low enough to cause symptoms and/or signs, includ-
ing impairment of brain function. The clinical manifestations of
hypoglycemia are nonspecific, it is not possible to state a single
plasma glucose concentration that categorically defines hypo-
glycemia, and a lowmeasured plasma glucose concentration can
be artifactual. Therefore, hypoglycemia is confirmed by docu-
mentation of Whipple’s triad (3): symptoms, signs, or both con-
sistent with hypoglycemia, a low plasma glucose concentration,
and resolution of those symptoms or signs after the plasma glucose
concentrationisraised. IntheabsenceofWhipple’s triad, thepatient
may be exposed to unnecessary evaluation, costs, and potential
harms, without expectation of benefit. This very large potentially
beneficial effect of documenting Whipple’s triad upgrades the evi-
dence (based on consistent clinical observations), thus supporting a
rating of high quality. (A rare exception would be a patient who is
physically unable to communicate symptoms.)
Symptoms of hypoglycemia are categorized as neuroglyco-
penic (the result of brain glucose deprivation per se) and neuro-
genic or autonomic (largely the result of the perception of phys-
iological changes caused by the sympathoadrenal discharge
triggered by hypoglycemia) (4). Awareness of hypoglycemia is
mainly the result of the perception of neurogenic symptoms (4),
which are largely sympathetic neural, rather than adrenomed-
ullary, in origin (5). Some neurogenic symptoms, such as palpi-
tations, tremor, and arousal/anxiety, are adrenergic whereas
others, such as sweating, hunger, and paresthesias, are cholin-
ergic (4). Neuroglycopenic symptoms (4) range from behavioral
changes, fatigue, and confusion to seizure and loss of conscious-
ness, i.e. functional brain failure (6). Seemingly complete recov-
ery after the glucose level is raised is the rule, although on rare
occasions neurological recovery is delayed. Profound, prolonged
hypoglycemia can cause brain death (6). Signs of hypoglycemia,
such as diaphoresis and pallor, are often subtle, although neu-
roglycopenic manifestations are often observable.
In healthy individuals, symptoms of hypoglycemia develop at
ameanplasma glucose concentration of approximately 55mg/dl
(3.0 mmol/liter) (7). However, the glycemic thresholds for this
and other responses to hypoglycemia shift to lower plasma glu-
cose concentrations in patients with recurrent hypoglycemia (7–
10). Furthermore, whereas arteriovenous plasma glucose con-
centration differences are clinically negligible in the
postabsorptive state, antecubital venous plasma glucose concen-
trations are as much as one third lower than arterial glucose
concentrations (which are relevant to maintaining brain glucose
metabolism) when insulin secretion is increased substantially,
e.g. after a glucose load, causing glucose extraction across the
forearm (11). Finally, because of the provision of alternative
circulating fuels to the brain (specifically ketones), lower plasma
glucose concentrations occur in healthy individuals, particularly
in women and children, without symptoms or signs during ex-
tended fasting (7). For all of these reasons, it is not possible to
state a single plasma glucose concentration that categorically
defines hypoglycemia.
Plasma glucose concentrations used to document Whipple’s
triad, in the absence of insulin or insulin secretagogue treatment
of diabetes,must bemeasuredwith a reliable laboratorymethod,
not with self-monitoring of blood glucose. Although a distinctly
low, reliablymeasuredplasmaglucose concentrationobtained in
theabsenceof recognized symptomsor signs shouldnotbe ignored,
that finding raises the possibility of “pseudohypoglycemia”—an
artifact of continued glucosemetabolismby the formed elements of
thebloodafter thesample isdrawn.Thatmayoccurwhentheblood
sample is collected in a tube that does not contain an inhibitor of
glycolysis and separation of the plasma (or serum) from the formed
elements is delayed, particularly in the setting of erythrocytosis,
leukocytosis, or thrombocytosis (12).
Documentation of Whipple’s triad establishes that a hypo-
glycemic disorder exists. Its etiology may be apparent (e.g. in a
patient with insulin-treated diabetes) or a diagnostic challenge
(e.g. in a seemingly well individual with an insulinoma). On
the other hand, in a person who does not have diabetes mel-
litus an unequivocally normal plasma glucose concentration
[e.g. �70 mg/dl (3.9 mmol/liter) (7)] during a symptomatic
episode indicates that those symptoms are not the result of
hypoglycemia.
1.1 Values
Hypoglycemia is rare in persons who do not have drug-
treated diabetes mellitus (12–15). Furthermore, not requiring
Whipple’s triad to initiate a workup will very likely expose pa-
tients who do not have a specific pathology causing hypoglyce-
mia to unnecessary evaluations, costs, and potential harmswith-
out expectation of benefit. Therefore, we believe it is important
to document Whipple’s triad before concluding that a hypogly-
cemic disorder exists. On the other hand, hypoglycemia is com-
mon in persons with insulin- or insulin secretagogue-treated di-
abetes mellitus (12, 16). Confirmation of Whipple’s triad, e.g.
with self-monitoring of blood glucose, during an episode of sus-
pected hypoglycemia is advisable in such a patient. However, if
that is not practical, it is reasonable to assume the episode is
caused by hypoglycemia for therapeutic purposes because the
probability of that assumption is high and the potential negative
impact of an untreated episode is considerable.
2.0 Evaluation and Management of
Hypoglycemia in Persons without Diabetes
Mellitus
Background
Because of the effectiveness of the normal defenses against
falling plasma glucose concentrations (7), hypoglycemia is an
J Clin Endocrinol Metab, March 2009, 94(3):709–728 jcem.endojournals.org 711
by Xiaohui Yu on March 30, 2009 jcem.endojournals.orgDownloaded from
uncommon clinical event (12–15) except in persons who use
drugs that lower plasma glucose levels, particularly insulin or an
insulin secretagogue, to treat diabetes mellitus (12, 16). Hypo-
glycemia is a fact of life formost personswith type 1 diabetes and
many with type 2 diabetes. Although persons with diabetes are
not spared the risk for the same hypoglycemic disorders as those
without diabetes, the vast majority of their hypoglycemic epi-
sodes are the result of treatment of their diabetes. Furthermore,
the pathophysiology of hypoglycemia in diabetes is distinct,
and the diagnostic and management approaches are different
from those in individuals without diabetes (12, 16). Therefore,
we address hypoglycemia in persons without diabetes and in
those with diabetes separately.
Physiology and pathophysiology
Glucose is anobligatemetabolic fuel for thebrainunder phys-
iological conditions (6, 7). Because the brain cannot synthesize
glucose, use physiological circulating concentrations of alterna-
tive fuels effectively, or store more than a fewminutes’ supply as
glycogen, maintenance of brain function, and ultimately sur-
vival, requires a virtually continuous supply of glucose from the
circulation. That, in turn, requires maintenance of the plasma
glucose level within the physiological range because blood-to-
brain glucose transport is a direct function of the arterial plasma
glucose concentration. Redundant glucose counterregulatory
mechanisms normally effectively prevent or rapidly correct hy-
poglycemia (7). The critical physiological defenses include: 1) a
decrease in insulin secretion as glucose levels decline within the
physiological range; 2) an increase in glucagon secretion; or, in
its absence, 3) an increase in epinephrine secretion, both occur-
ring as glucose levels decline just below the physiological range.
Increased cortisol and GH secretion are involved in defense
against prolonged hypoglycemia. If these defenses fail to abort
the episode, plasma glucose levels will continue to fall. Symp-
toms, which prompt the behavioral defense of food ingestion,
normally develop at a mean plasma glucose concentration of
approximately 55 mg/dl (3.0 mmol/liter). At that and lower glu-
cose levels, insulin secretion is suppressed virtually completely
(7, 17); plasma insulin levels are below 3 �U/ml (18 pmol/liter),
C-peptide levels are below 0.6 ng/ml (0.2 nmol/liter), and pro-
insulin levels are below 5.0 pmol/liter (14).
Because external losses are normally negligible, hypoglyce-
mia develops when the sum of glucose utilization from the cir-
culation (largely by the brain but also by obligatory glycolytic
tissues, such as the renal medullae and erythrocytes, and insulin-
sensitive tissues, such as muscle) exceeds the sum of glucose de-
livery into the circulation (from ingested carbohydrates and he-
patic and renal glucose production) (12–15). Because of the
capacity to increase endogenous glucose production substan-
tially, hypoglycemia is typically the result of absolutely low rates
of glucose production or rates of glucose production that are low
relative to high rates of glucose utilization.
Recom