Child Neuropsychology

2004, Vol. 10, No. 1, pp. 3652

Neurocognitive Correlates of Type 1 Diabetes

Mellitus in Childhood
Mary Desrocher1 and Joanne Rovet2,3
1

Department of Psychology, York University, Toronto, Ont., Canada,

Department of Pediatrics, University of Toronto, Toronto, Ont., Canada,
and 3Brain and Behavior Program, The Hospital for Sick Children, Toronto, Ont., Canada
2

ABSTRACT
Type 1 Diabetes Mellitus (T1DM) is one of the most prevalent chronic health conditions in children under the
age of 18 years. Complications of the disease include hypo- and hyperglycemia, which can have an impact on
childrens performance in assessment situations, in the clinic, and in school. Because there is no cure for this
disease, there is a need to understand the cognitive deficits associated with some of its complications, as this
knowledge will impact on the choice of treatment regimens as well as educational interventions. This paper
provides a comprehensive review of the relevant literature on the neurocognitive outcome of T1DM. In
particular, disease- and treatment-related variables that are associated with poor performance on cognitive
domains will be reviewed. Specifically, age of onset, duration, pubertal effects, and presence of hypoglycemia or hyperglycemia will be examined. These findings are not without controversy, and limitations
to conclusions will also be presented. Where relevant, recommendations for future research directions
will be provided.

NEUROCOGNITIVE CORRELATES
OF T1DM IN CHILDHOOD
Type 1 Diabetes Mellitus (T1DM) is one of the
most common chronic diseases of childhood,
affecting anywhere from 1 to 600 (Traisman,
1980) to 1 in 60 children under the age of 18 years
(Health Canada, 1999). T1DM is caused by
autoimmune destruction of pancreatic beta cells
where the hormone insulin, which is essential for
glucose metabolism, is produced. If insulin is
lacking, glucose accumulates in the bloodstream
and urine while the body literally starves. Organs
that require glucose as a primary fuel are
vulnerable to this perturbation of glucose metabolism. The brain in particular is susceptible to
damage, as it cannot store glucose in its cells, and

thus requires a continuous supply from the

vascular system.
Treatment of T1DM requires multiple daily
injections of exogenous insulin. The main goal of
diabetes therapy is to prevent the long-term
micro- and macro-vascular complications associated with long-standing hyperglycemia, such as
cardiovascular disease, nephropathy, retinopathy
and neuropathy. However, because this therapy
imperfectly mimics normal physiological insulin
production, children with diabetes are constantly
exposed to abnormal levels of both insulin and
glucose, each of which can have different manifestations on brain function. Furthermore, children with diabetes can also experience episodes
of severe hypoglycemia (from too much insulin)
or severe hyperglycemia (from not enough

Address correspondence to: Mary Desrocher, Ph.D., Assistant Professor, Department of Psychology, York
University, 4700 Keele St., Toronto, Ont., Canada M3J 1P3. E-mail: mdesroch@yorku.ca
Accepted for publication: May 22, 2003.
0929-7049/04/1001-036$16.00 # Taylor & Francis Ltd.

NEUROCOGNITIVE CORRELATES OF DIABETES

insulin), both of which can cause brain damage.

While intensive therapy has been effective for
adults and older adolescents, its efficacy and
safety have not been proven in children (Rovet,
1999). This is because the increased risk of
hypoglycemia associated with intensive therapy
is hypothesized to cause damage in children
whose brains are still developing and so may be
more vulnerable to the effects of severe hypoglycemia than adults (McCall, 1992).
Thus patients with diabetes, especially children, are at risk of impaired cognitive functioning.
This potentially is due to the subtle brain damage
associated with chronic (typically mild) aberrations in levels of blood glucose as well as to the
severe acute events these patients sometimes
experience. Even in children who are well controlled (i.e. whose overall blood glucose levels are
relatively stable over time through intensive disease management), there are periodic fluctuations in blood glucose levels. Therefore, it is not
surprising that children with diabetes are potentially at risk for neurocognitive impairments,
which can be transient or permanent depending
on the type of insult.
This paper serves to provide an overview of the
characteristics of T1DM and its management that
predispose children to different kinds of neuropsychological impairments. Although studies on
adult functioning are included, one must be cautious in drawing conclusions from research on
adults to children partly because of the effect of
historical variables, such as response to treatment
and management issues, and how these interact
with maturation. An exploration of the literature
on cognitive function in children with diabetes is
necessary to determine how neurocognitive deficits associated with T1DM in childhood are
related to disease variables. It will be argued
that, based on the extant literature, four key
relationships between T1DM and outcome exist:
motor and visuospatial deficits with an early age
of diabetes onset; attention and memory deficits
with hypoglycemia, particularly if this is either
severe and acute or mild and chronic; verbal and
executive function deficits with hyperglycemia;
and puberty with executive function. In addition,
this paper will also serve to identify areas in need
of future research.

37

TYPE 1 DIABETES MELLITUS:

DISEASE VARIABLES AND
ASSOCIATED COGNITIVE DEFICITS
The literature on cognitive deficits in children
with T1DM has been lacking a comprehensive
review of the findings to date. Although an
attempt has been made to relate diabetes variables
to cognitive outcome, and to suggested brain
regions that may be affected, it must be noted that
the research is anything but clear-cut. There are
several difficulties with drawing conclusions
about cognitive outcome in this population,
including a lack of hypothesis driven research,
little consistency in sample characteristics, lack of
consensus on measures to be used to measure
cognitive dysfunction, and lack of clarity on the
constructs to be studied. What follows is a review
of the literature on the relationship of disease
variables to cognitive outcome, particularly complications, age of onset, duration, and findings
associated with hormonal changes at puberty.
Hypoglycemia
Hypoglycemia results when blood glucose levels
fall below a critical threshold. Hypoglycemia
typically begins when concentrations of blood
glucose reach 3 mM, and symptoms such as
tremor, weakness, confusion, and lack of concentration become evident. In situations of severe
hypoglycemia, blood glucose concentrations may
reach much lower levels, and unconsciousness,
convulsions, coma, and even death can occur.
Hypoglycemia arises as a consequence of excessive insulin intake, excessive exercise, or eating
the wrong foods. Because of the difficulties in
balancing insulin injections with activity and diet
in children, as many as 31% of all children with
T1DM experience one or more episodes of severe hypoglycemia (Daneman, Frank, Perlman,
Tamm, & Ehrlich, 1989), and this appears to be
more common in individuals with good rather
than poorly controlled diabetes (Davis, Keating,
Byrne, Russell, & Jones, 1998).
Among patients with diabetes, there is a subpopulation that experiences a condition known
as hypoglycemia unawareness, which involves
blunted sensations due to frequent episodes of
hypoglycemia. Because these individuals fail to

38

MARY DESROCHER & JOANNE ROVET

experience the physiological warning signs of

hypoglycemia, they do not take measures to
correct it and so are predisposed to severe episodes. In addition, patients with diabetes may
also experience nocturnal hypoglycemia, which
occurs during sleep and may have an impact on
brain and cognitive functioning (Matyka, FordAdams, & Dunger, 2002).
The treatment of hypoglycemia requires supplementation of glucose via ingestion of candy or
juice while severe hypoglycemia requires immediate medical intervention. Very young children
(especially below age 5) are thought to be at
increased risk for severe hypoglycemic episodes
and associated brain damage (Eeg-Olofsson &
Petersen, 1966) because of the increased difficulty
in managing diabetes in the very young. These
children are also less likely to express verbally the
sensations of hypoglycemia and experience it at a
time when their brains are more vulnerable to its
effects.
Studies with animals have shown damage to
cerebral neurons from a hypoglycemic episode
that occurs within 45 min of its onset (Auer &
Siesjo, 1988). Furthermore, neurological damage
from hypoglycemia is though to be limited to
brain structures with the densest distribution of
insulin receptors, such as the hippocampus (Auer
& Siesjo, 1988). Because the hippocampal formation is relevant for memory and learning, these
abilities are hypothesized to be especially vulnerable to hypoglycemia (Hershey, Craft, Bhargava,
& White, 1997). There is some suggestion that the
acute effects of hypoglycemia may be transitory,
whereas it is only the recurrent episodes of severe
hypoglycemia that lead to impairments in cognitive functioning (Holmes, 1990). It stands
to reason, therefore, that those children who
have experienced recurrent severe hypoglycemia
should have more pronounced deficits in cognitive function than those with nonrecurrent episodes of lowered blood glucose.
Cognitive Deficits Associated
With Hypoglycemia
Hypoglycemic events have been associated with a
number of deficits in various cognitive domains,
including motor, visuospatial, attention, memory,
and executive function deficits. In terms of motor

functioning, it has been shown that motor speed

can be affected by transient or recurrent hypoglycemia. Several studies have directly manipulated hypoglycemia using a stepped glucose
clamp technique (Holmes, Hayford, Gonzalez,
& Weydert, 1983; Holmes, Koepke, Thompson,
Gyves, & Weydert, 1984). Patients with diabetes,
as well as individuals without diabetes who
undergo similar procedures, both show transient
reduction in motor speed when hypoglycemic
(Gschwend, Ryan, Atchison, Arslanian, & Becker,
1995; Lobmann et al., 2000; Ryan et al., 1990)
while their performance does not return to normal
until 4090 min after euglycemia is restored. In
another study exploring direct manipulations of
blood glucose levels, Lobmann et al. compared
adults with T1DM and normal controls on an
event-related potential (ERP) paradigm involving
a timed motor task. In addition, the insulin-clamp
technique was used to induce hypoglycemia
experimentally in both groups. Stimulus selection, response selection, and reaction time components of the brain waves were recorded from
various scalp locations. In both the diabetes and
nondiabetes groups, hypoglycemia was found to
slow the latency, but not the amplitude component,
which is thought to be indicative of abnormal
stimulus processing.
Acute naturally occurring hypoglycemia has
also been shown to affect motor functioning.
Reich et al. (1990), compared the pegboard performance of 6- to 14-year-old children at a diabetes camp, while euglycemic (baseline) and right
after an acute episode of hypoglycemia. These
children showed slower motor speeds following
the hypoglycemic event. Poor fine motor control
appears to be related to nonacute events, or a
history of recurrent hypoglycemia, as shown in a
study by Golden et al. (1989), in which they found
that preschool children experiencing frequent
mild and even asymptomatic hypoglycemia performed poorly on copying tasks. Frier (2001) has
similarly reported that children and adolescents
with a longstanding history of chronic, mild
hypoglycemic episodes exhibit deficits on motor
tasks, particularly under timed conditions. Severe
hypoglycemia has also been shown to affect
motor speed, with a significant decrement on
timed tasks in children with T1DM relative to

NEUROCOGNITIVE CORRELATES OF DIABETES

controls (Hershey, Bhargava, Sadler, White, &

Craft, 1999).
Visuospatial deficits have been found to be
associated with recurrent hypoglycemic events
in a number of studies. Rovet, Ehrlich, and
Hoppe (1988), in a study of 6- to 14-year-old
children with diabetes, observed a greater likelihood of deficits on a variety of visuospatial
tasks (e.g., Block Design, Object Assembly, Primary Mental Abilities Spatial Relations subtest,
and Beery-Buktenica Test of Visual-Motor Integration) in children with early-onset diabetes and
a history of hypoglycemic seizures. These findings were subsequently replicated in a study by
Rovet, Ehrlich, Czuchta, and Akler (1993). Golden
et al. (1989), using the Stanford-Binet Pattern
Analysis subtest, found that young children with
recurrent episodes of mild hypoglycemia were
most at risk for impairments on this task. Severe
hypoglycemia has also been implicated in deficits
of pattern recognition (Hershey et al., 1999).
Deficits in attention have long been recognized
in children with T1DM and a history of severe
hypoglycemia, especially on tasks requiring attention to detail, or maintaining focus (Bjorgass,
Gimse, Vik, & Sand, 1997; Drash et al., 1983;
Gschwend et al., 1995; Ryan, Vega, & Drash, 1985;
Ryan et al., 1990; Ryan, Williams, Finegold, &
Orchard, 1993). A variety of attention tests have
been used to study patients with severe recurrent
hypoglycemic events including cancellation tests
(Bjorgass et al., 1997; Ryan et al., 1985), simple
and choice-reaction time tests (Gschwend et al.,
1995; Holmes et al., 1983; Ryan et al., 1985, 1990),
go-no go tests (Ryan et al., 1993), distraction
measures (Hagen et al., 1990), interference tasks
(Mitrakou et al., 1991), as well as behavioral
questionnaires (Northam, Bowden, Anderson, &
Court, 1992). However, since most studies used
only single instruments, the multicomponential
nature of attention has not been fully appreciated.
One exception is the study by Rovet and Alvarez
(1997) in which a theoretically motivated multicomponential approach was used to show that
different aspects of attention are sensitive to different diabetes-related factors. In this study, 9- to
18-year-old children with T1DM and matched
controls were evaluated using a variety of clinical
attention instruments (e.g., Matching Familiar Fig-

39

ures Test, Continuous Performance Test, Trail

Making, Wisconsin Card Sorting, Visual Search)
that yielded information about five attention
components: select, focus, shift, inhibit, and sustain. Results revealed that while the diabetes group
as a whole did not differ from controls on any of
these components, when children with diabetes
were subgrouped according to whether or not
they had experienced hypoglycemic seizures,
those with seizures, regardless of age of onset,
had the greatest difficulty with selective, focused
and inhibitory aspects of attention.
Memory difficulties have often been described
in children with T1DM and these also appear to
be associated with hypoglycemia. This variable
has been associated with a weakness in both
visuospatial (Rovet et al., 1988; Ryan et al., 1985)
and verbal memory (Hagen et al., 1990), especially if the child experienced seizures from hypoglycemia (Reich et al., 1990; Rovet et al., 1993).
Kaufman, Epport, Engilman, and Halvorson
(1999) assessed children with T1DM on verbal
selective reminding tests. Their findings revealed that children who had experienced prior
hypoglycemic seizures did more poorly than
those who did not. One key study that illustrates
memory deficits in this population is that of
Hershey et al. (1997). These researchers evaluated
patients with diabetes who ranged in age from 15
to 42 years and had diabetes from at least age 14.
All participants were assessed on tasks of declarative memory (e.g., California Verbal Learning
Test, Paragraph and Pattern Recall, and Recognition tasks), and tests of nondeclarative memory
abilities (e.g., Picture Priming, Word-Stem Priming). Results revealed that both declarative and
nondeclarative verbal memory were poorer in
patients who had at least one severe hypoglycemic episode than those without seizures, while
nonverbal nondeclarative memory was affected in
both diabetic groups. Later findings by this group
(Hershey et al., 1999) have shown similar effects
on a spatial memory task, suggesting that severe
hypoglycemia can have effects on nonverbal
memory.
A few studies have shown that children with a
history of hypoglycemia may perform more
poorly on tasks of executive function. Bjorgass
et al. (1997), using the Trail Making Test, found

40

MARY DESROCHER & JOANNE ROVET

that children with early-onset diabetes and a

history of hypoglycemia had difficulties shifting
between numbers and letters (Trail B). Brady
(2000) showed that preschool-aged children with
diabetes with a history of hypoglycemia had
deficient nonverbal reasoning abilities on the
Differential Abilities Scale. Similarly, Hershey
et al. (1997) found that adult and adolescent
patients with diabetes with a past history of
hypoglycemic seizures showed the greatest deterioration after an interference trial on the California Verbal Learning Test. Ryan et al. (1990) have
reported deficits in decision-making and planning
in children who suffered acute experimentallyinduced episodes of hypoglycemia. Whether findings of acute effects are similar to those under
recurrent hypoglycemia is a question for future
research.
Neuropathology of Hypoglycemia
Given the widespread effects of hypoglycemia on
a number of cognitive domains, it is difficult to
pinpoint any one area of the brain that is
particularly susceptible to low blood glucose
levels. Deficits in motor function may implicate
a complex network including the primary motor
cortex, supplementary motor areas, premotor
areas, thalamus, cerebellum, and brainstem.
Visuospatial deficits implicate the parietal lobes.
Attentional networks that may be affected include
the thalamus, parietal, and frontal lobes. Memory
deficits implicate the temporal (declarative memory) and frontal lobes (nondeclarative memory).
Executive function deficits implicate the frontal
lobes. One must be cautious in attempting to infer
neurological damage from results on neuropsychological tests, as many of these tasks tap into
multiple domains of function, and thus may
implicate multiple centers of the brain. It is
therefore necessary to base our knowledge of the
neurological bases of cognitive dysfunction on
studies of brain pathology.
Neuropathological changes that have been
associated with hypoglycemia, particularly seizures, include neuronal atrophy in the cortex,
caudate, and hippocampus (McCall, 1992).
Chronic hypoglycemia, that does not cause seizure activity, has been associated with alterations
of neuronal transport in the same brain regions

noted above (McCall, 1992). In patients with diabetes subjected to studies of regional cerebral
blood flow via single-photon emission computed tomography, repeated episodes of severe
hypoglycemia were associated with greater tracer
uptake in the prefrontal cortex and reduced uptake
in the calcarine cortex (Macleod et al., 1994;
Tallroth, Ryding, & Agardh, 1992). MRI findings
showed a patient who developed severe amnesia
after hypoglycemic coma had a marked lesion in
the left temporal lobe (Chalmers et al., 1991).
As yet, there have been few studies that directly explore the relationship of hypoglycemia to
alterations in brain systems underlying particular
cognitive functions. One study by Perros, Deary,
and Sellar (1997) used MRI scans to explore brain
regions in adults with hypoglycemic seizures
dating from childhood. Their main finding was
diffuse atrophy across the cortex of this group
relative to age-matched controls, with relatively
large decreases in volume in the hippocampal
formation, on both the right and left side. The
neuropsychological findings suggest associated
memory deficits. Ferguson et al. (2003), in an
exploration of young adults with a history of
T1DM, found no effects of a history of severe
hypoglycemia on brain structures. Their findings
must be interpreted with caution as the authors
provided no information on when the hypoglycemic events occurred during development. As
noted above, it may be the case that early episodes
of hypoglycemia are more detrimental than episodes occurring in adolescence or adulthood. It
must be noted that, to date, there are no studies
exploring the brains of children with a history of
severe hypoglycemic events.
Hyperglycemia
Children with T1DM are also at risk for frequent
episodes of hyperglycemia due to poor management and regimen non-compliance, particularly
lack of adherence to dietary and exercise restrictions, as well as illness, and unusual levels of stress.
Hyperglycemia occurs when blood glucose levels
reach 14 mM or higher. In the individual who does
not have diabetes, hyperglycemia is normally
counteracted by the excretion of insulin from the
pancreas. However, because the individual with
T1DM lacks an endogenous supply of insulin, this

NEUROCOGNITIVE CORRELATES OF DIABETES

mechanism is not readily available to him or her.

Not only do hyperglycemic episodes contribute to
the well-known complications of diabetes (e.g.,
nephropathy, retinopathy, neuropathy), they may
also be associated with cognitive impairment. In
the extreme form, prolonged and severe hyperglycemia can lead to a serious and debilitating
condition known as diabetic ketoacidosis (DKA),
which is a chemical imbalance that causes acute
illness, unconsciousness, and coma, possibly
leading to death. Individuals who survive DKA
may suffer CNS damage and thus it is important
to explore the cognitive outcomes associated
with hyperglycemic events (Tsalikian, Daneman,
Becker, Crumrine, & Drash, 1980).
Cognitive Deficits Associated
With Hyperglycemia
With the exception of DKA, it is thought
that acute episodes of hyperglycemia are less
generally problematic than episodes of hypoglycemia in children (Gonder-Frederick, Cox,
Driesen, Ryan, & Clarke, 1994). While chronic
hyperglycemia has not been hypothesized to
lead to any specific cognitive impairment in
children, recent findings suggest some complications, particularly retinopathic alterations, may be
evident following the hormonal changes of
puberty (Donaghue et al., 1993; Lawson, Sochett,
Chait, Balfe, & Daneman, 1996).
Findings in the neuropsychological literature
suggest a number of cognitive domains that may
be affected by hyperglycemia. Elevations in blood
glucose have been associated with poor performance on visuospatial tasks (Rovet, Ehrlich, &
Czuchta, 1990). Information ascertained from
parent-completed diabetic diaries about intervening hospitalizations and episodes of hypo- and
hyperglycemia allowed the researchers to determine that a history of ketonuria, due to hyperglycemia, was associated with poor spatial abilities 1
year after diagnosis, suggesting a disruptive influence of ketones on posterior brain structures.
However, this finding was not replicated in the
children at a later age (e.g., Rovet et al., 1993), or
in work by other researchers who examined
visuospatial performance under conditions of
induced hyperglycemia (Gschwend et al., 1995).
In terms of visual attention, Rovet and Alvarez

41

(1997) have shown that higher ambient blood

glucose levels at time of testing were associated
with increased difficulty on the inhibit component
of attention.
Deficits associated with hyperglycemia have
also been found in memory and executive function. Kaufman et al. (1999) assessed children with
T1DM on verbal selective reminding tests. Their
findings revealed that children who were prone to
hyperglycemia (as indicated by elevations in the
HbA1c levels in the 2 years prior to testing) performed poorly on this task. Adults with T1DM
showed difficulties on the Porteus Maze Test,
particularly if they had a history of poor metabolic control, contributing to frequent hyperglycemia (Hills-Briggs & Echemendia, 2001).
Neuropathology of Hyperglycemia
Given the findings of deficits on visuospatial,
attention, memory, and executive function tasks,
one might expect to find alterations in the parietal,
temporal, and frontal lobes of persons suffering
chronic hyperglycemia. As yet, there have been
no studies utilizing neuroimaging techniques to
explore brain alterations due to this variable.
Studies with both animals and humans (adults)
have shown the main types of neuropathology
associated with hyperglycemia are degeneration of
ganglion cells, demyelinization, neuronal damage,
and biochemical changes (DeJong, 1977; McCall
& Figlewicz, 1997; Mooradian, 1988), and that
these changes are fairly diffuse. Some researchers
have argued that a young childs brain is protected
to a greater degree than an adults because of its
greater flexibility in using nonglucose fuels, such
as lactate, for energy production (Amiel, 1996;
McCall & Figlewicz, 1997). Ferguson et al. (2003)
found a statistically nonsignificant trend (p < .10)
towards small white matter lesions in the right
temporal lobes of adults who experienced chronic
hyperglycemia associated with retinopathy. Few
studies have directly explored the relationship of
brain damage associated with hyperglycemia to
tasks that have been selected based on their known
relationship to various brain pathologies.
Age at Disease Onset
An early age of onset for T1DM, particularly
prior to 5 years of age, is associated with different

42

MARY DESROCHER & JOANNE ROVET

neurocognitive deficits than if diabetes develops

later in childhood (Holmes, 1990; Kaufman et al.,
1999; Rovet, 1999; Ryan, Vega, Longstreet, &
Drash, 1984). This early-onset effect has been
attributed to the sensitivity of the young developing brain to alterations in glucose levels, as well
as to the added difficulty in controlling blood
glucose levels in very small children (Ryan,
1990). It is difficult to separate out the effects of
early-onset of diabetes from hypoglycemia, as
young children are more susceptible to hypoglycemia unawareness, and are unable to verbalize
their symptoms. The literature presented below
often considers early-onset with the risk of
hypoglycemia, and it must be kept in mind that
the effects of these two variables are very difficult
to dissociate in retrospective research.
Cognitive Deficits Associated
With Early Age of Onset
Differential effects on cognitive function have
been demonstrated according to age of onset, with
early-onset of T1DM gaining the most attention.
Motor deficits have been associated with earlyonset of T1DM. Patients with early-onset diabetes
appear to demonstrate slower motor speeds than
those who develop it later or for a shorter period
of time (Ryan et al., 1985). A pertinent issue
regarding motor deficits in children with T1DM is
the role of a speed-accuracy tradeoff (see Ryan
et al., 1984, 1985). According to Ryan (1990),
children with later-onset T1DM appear to develop
a cautious response style as a consequence of the
additional careful attention they must pay to detail
in treating this disorder, while this style leads to
slower decision-making and responding. Others,
however, have argued that a speed-accuracy
tradeoff on motor tasks involving complex and
unfamiliar stimuli may be characteristic of
children with T1DM regardless of age of onset
(Frier, 2001; Holmes et al., 1983, 1984). Alternatively, children with early-onset may be slow
for neurological reasons due to chronic insults of
diabetes prior to age 5 at a time when the brain is
particularly vulnerable.
Visuospatial deficits have been noted on construction and puzzle tasks (Rovet, 1999; Rovet
et al., 1993). Among children tracked from time
of diagnosis for a period of 8 years, those diag-

nosed before age 4 were more likely than those

diagnosed past this age to show visuospatial
processing deficits. In terms of visual attention,
children with early-onset diabetes have difficulty
with selective attention on measures such as
the Matching Familiar Figures Test (Rovet &
Alvarez, 1997).
Memory deficits have also been associated
with age of onset, with early onset associated
with a weakness in both visuospatial (Rovet et al.,
1988; Ryan et al., 1985) and verbal memory
(Hagen et al., 1990). Wolters, Yu, Hagen, and
Kail (1996) have shown that children who are
diagnosed with T1DM prior to the age of 5 years,
were more likely to show poor recall of words in
an assessment of immediate memory (no delay),
although their strategies for rehearsal were no
different from those of children with later onset
diabetes or controls. Executive function deficits
seem to also be affected by onset of T1DM prior
to the age of 5. Bjorgass et al. (1997), using the
Trail Making Test, found that children with earlyonset diabetes had difficulties shifting between
numbers and letters (Trail B). Rovet and Alvarez
(1997), examining adolescents with diabetes and
matched controls on the Stroop Color-Word Interference task, found those with early-onset diabetes were more susceptible to interference
effects than controls or adolescents with lateonset diabetes. Brady (2000) similarly showed
that preschool-aged children with diabetes with
a history of hypoglycemia had deficient nonverbal
reasoning abilities on the Differential Abilities
Scale. Verbal fluency may also be affected by
T1DM. In Northam et al.s (2001) follow-up study
of children after 2 years of diabetes, the earlyonset diabetes group generated fewer words
belonging to a specific category or starting with
a designated letter on the Controlled Oral Word
Association Test (COWAT) than did the other
study participants.
Ack, Miller, and Weil (1961) were the first to
report significantly and substantially reduced IQ
scores (in the order of 1020 IQ points) in
children who developed diabetes prior to the age
of 5, in contrast to children developed it at a later
age who did not differ from nondiabetic sibling
controls. A series of subsequent studies have
corroborated this finding, showing further that

NEUROCOGNITIVE CORRELATES OF DIABETES

the deficit is not global but rather selective for

Performance IQ subtests (Holmes & Richman,
1985; Northam, Anderson, Werther, Adler, &
Andrewes, 1995; Northam, Anderson, Werther,
Warne, & Andrewes, 1999; Rovet et al., 1988;
Ryan et al., 1985). Moreover, this deficit appears
mainly to reflect motor slowing on timed tests of
the Performance subscale.
Not all domains of functioning have been
found to be impaired in children with earlyonset T1DM. Brady (2000) compared 20 children
with early-onset diabetes (mean age 4.5 years)
to healthy controls on the Peabody Picture Vocabulary Test and the Expressive One-Word Picture
Vocabulary Test, which evaluate receptive and
expressive language skills, respectively. Results
revealed no difference between these groups on
either aspect of language functioning.
Neuropathology of Early Onset T1DM
As with the literature on hypoglycemia, the effects
of an early age of onset seem to involve multiple
areas of the brain, including motor and executive
function circuits of the frontal cortex, the parietal
lobes, and the temporal lobes. Although it has been
suggested that early age of onset may differentially
affect posterior brain regions (Rovet, 1999), as
shown by visuospatial deficits upon early testing,
these results have not been corroborated by
findings in the area of neuroimaging. One study
by Davis, Trundle, Ives, Robind, and Jones (2002)
found that children (815 years) with early-onset
diabetes (less than 6 years of age) were more likely
to show structural abnormalities on MRI scans,
across the cortex. In particular, there was evidence
of mesial temporal sclerosis (atrophy) in the
dominant hemisphere of 15% of their participants.
Given the widespread nature of the brain abnormalities in this study, the question then becomes
one of whether these same findings would be seen
soon after diagnosis, prior to the age of 5. No
studies have yet been conducted to image the
brains of children under the age of 8 years.
Duration of Diabetes
Generally, neurocognitive deficits seen in children with T1DM are not evident until the child
has had the disease for at least 13 years (Holmes &
Richman, 1985; Northam et al., 2001; Rovet et al.,

43

1993; Sansbury, Brown, & Meacham, 1997). This

finding suggests that there is some mechanism
whereby alterations in glucose levels may have a
cumulative, rather than a singular, effect.
Cognitive Deficits Associated
With Longer Durations of T1DM
Children with T1DM, when tested after a long
duration of illness, show reduced motor strength,
slower motor speeds, and poorer fine motor
control. Rovet (1999), for example, found that
adolescents with diabetes of more than 8 years
duration evinced significant reductions in motor
strength. Decrements in motor speed have been
observed over longer durations of the disorder
using both psychometric tasks (such as the
Performance subtests of the Weschler Intelligence Scales for Children) (Meuter et al., 1980;
Ryan et al., 1984; Thomas, Gruneklee, Gries, &
Lohmann, 1980), as well as pegboards, tests of
finger tapping, and computerized tests of motor
responding. Patients with longer disease durations
appear to demonstrate slower motor speeds than
those who develop it later or are tested soon after
diagnosis (Holmes & Richman, 1985).
Visuospatial deficits have been found to be
more pronounced with longer durations of T1DM
(Rovet, 1999; Rovet et al., 1993). These deficits
were not present at 1 or 3 years post diagnosis,
suggesting that a factor associated with having
diabetes for more than a few years is critical.
Visual and verbal memory deficits seem to
become more pronounced if diabetes is longstanding (Rovet et al., 1993). In a study following
a large cohort of 314-year-old children from the
time of diagnosis over the first 6 years of illness,
Northam et al. (2001) found changes in verbal and
nonverbal learning and memory after 2 years of
diabetes on tests such as Digit Span, Rey Auditory
Verbal Learning Test, LHermitte Board (a test of
spatial memory), and the Rey-Osterrieth Complex
Figure Test (Northam et al., 1995, 1999). After 6
years of illness, there was a general finding of
poorer memory performance relative to controls,
suggesting a decline in memory functioning
(Northam et al., 2001).
Deficits in executive functioning may be associated with longer duration of the disease. In
Northam et al.s (2001) follow-up study of children

44

MARY DESROCHER & JOANNE ROVET

after 6 years of diabetes, the early-onset diabetes

group generated fewer words belonging to a
specific category or starting with a designated
letter on the Controlled Oral Word Association
Test (COWAT) than did the other study participants. Since this effect was not observed when
the children were tested soon after the time of
diagnosis, this finding suggests that a longer
duration of diabetes is necessary for the effects
on executive functioning to be observed, as has
also been shown in adults with early-onset T1DM
(Hershey et al., 1997). Deficits in Verbal IQ are
generally not seen in children with T1DM
(Northam et al., 1995, 1999; Rovet et al., 1988),
until some time after 6 years of illness, when
Verbal IQ levels appear to decline relative to
controls (Northam et al., 2001; Rovet & Ehrlich,
1988).
Neuropathological Changes Associated
With Disease Duration
Longer duration of T1DM has been associated
with a number of deficits in motor, visuospatial,
memory, and executive function, and is hypothesized to affect the frontal, temporal, and parietal
lobes. Added difficulties in verbal functioning
with age have been proposed to be associated with
the cumulative impact of mild hyperglycemia on
anterior language structures, such as Brocas area
in the frontal cortex (Kovacs, Ryan, & Obrosky,
1994). However, this is an untested hypothesis,
and neuroimaging studies are needed to provide
support for the idea of longitudinal changes in the
anterior cortex.
Puberty
Puberty is an independent risk factor for the
physical complications of diabetes due to
increased insulin resistance associated with gonadal and adrenal hormone changes of puberty
(Bloch, Clemons, & Sperling, 1987; Caprio et al.,
1989). Few studies have explored the relationship
of hormonal changes with the onset of cognitive
dysfunction.
Cognitive Deficits Associated
With Pubertal Changes
Recent studies have reported that certain selective
cognitive deficits may emerge only after the onset

of puberty (Rovet, 1999). Executive functioning

deficits have been noted in adolescents with
T1DM, particularly on interference tasks (Rovet,
1999). Note in this study, the effects were shown
to occur regardless of the age of diabetes onset,
suggesting something occurring at puberty that
could affect susceptibility to interference. Children who develop diabetes during later childhood
and adolescence exhibit differences from controls on the WISC Vocabulary, Similarities, and
Information subtests (Ryan et al., 1984) but these
may be due to difficulties with semantic memory
since these subtests tap into the long-term
memory stores of general knowledge. Ryan et al.
have alternatively posited that these children
acquire less information because they are more
often absent from school (Ryan et al., 1985).
Nevertheless, the impact of school absenteeism
when directly evaluated in recent studies was
shown to impact minimally on outcome
(McCarthy, Lindgren, Mengeling, Tsalikian, &
Engvall, 2002).
Neuropathology Related to Pubertal Changes
It seems that hormones not only have activational
effects, but also may trigger the onset of
hyperglycemia at puberty. It has been shown that
hyperglycemia directly affects timing of myelination (Vlassara, Brownlee, & Cerami, 1983). Since
the frontal lobes and parts of the reticular
formation have a protracted period of myelination
that extends well into the second and third
decades of life (Yakovlev & Lecours, 1975),
these structures may be especially vulnerable to
high levels of blood glucose during the pubertal
period. There are no studies which have explored
the brains of children with T1DM around the time
of puberty and correlated these findings with
serum hormone levels.
Summary
This review of the neurocognitive outcome in
children with T1DM shows a complex pattern of
deficits associated with this disease. The key diabetes variables that impact on outcome are age of
disease onset, presence of hypoglycemia, chronicity of effects (duration), and hyperglycemia
around the time of puberty, as summarized in
Table 1.

45

NEUROCOGNITIVE CORRELATES OF DIABETES

Table 1. Variables Related to Type 1 Diabetes Mellitus and Their Associated Neurocognitive Deficits.
Neurocognitive
domain

Diabetes-related variables
Early-onset Later-onset Recurrent
Acute
Hyperglycemia Longer Puberty
(<5 years) (>5 years) hypoglycemia hypoglycemia
duration

Motor deficits
Strength
Speed
Fine motor control
Visual deficits
Visuomotor
Visuospatial

X
X
X
X
X

X
X

X
X

?
X

Attention deficits
Selective
Focused
Shifting
Inhibition
Sustained
Memory deficits
Visual
Verbal
Set shifting
Susceptibility
to interference
Verbal fluency
Decision making
Planning
Route following

X
X
X

X
X

X
X

X
X

X
X

X
X

X
X
X

X
X
X
X

Language deficits
Intelligence deficits
FSIQ
VIQ
PIQ

X
X

MODEL OF NEUROCOGNITIVE
OUTCOME
Because of the complexity of the variables that
can affect outcome in children with T1DM, it is
necessary to create a model of how disease
variables may be related to neuropathology and
cognitive outcome. Such a model must be
strongly grounded in knowledge of neurocognitive development, and the extant literature on
neuropathological changes associated with
T1DM. Figure 1 is an attempt to provide a
hypothetical model illustrating these relation-

ships, which could be used to guide hypothesis

selection for future research. Timing of the
development of brain myelination and associated
cognitive outcome was obtained from seminal
reviews by Dodge, Prensky, and Feigin (1975)
and Golden (1981).
The key variables that seem to be related to
poor cognitive outcome are age of onset, hypoglycemia, hyperglycemia, puberty, and duration.
It is suggested that the critical period of earlyonset is prior to Age 5, which reflects a consensus
in the literature as to when the key effects of the
disease occur. The deficits that one would see to

46

MARY DESROCHER & JOANNE ROVET

Fig. 1. Model of the relationship between diabetes variables and neurocognitive development.

be related to this variable are motor and visuospatial in nature. Critical periods for hypoglycemia
and hyperglycemia are also indicated. Hypoglycemia is proposed to have detrimental effects to
neurocognitive function between birth and 12
years, when the brain is undergoing its most
rapid myelination. Because of this, and because
of the overlap with early-onset, one might expect
early deficits to include difficulties with motor
and visuospatial function. If hypoglycemia occurs
during key periods of myelination of circuits
involved in attention and memory, then deficits
in these areas may occur. Hyperglycemia seems to
be related to hormonal events present at puberty,
and the key domain of vulnerability would seem
to be in executive function, dependant on integrity
of the prefrontal cortex. With longer durations of
the disease, the likelihood for complications may
arise, either in very well-controlled situations,
which may increase the likelihood of hypoglycemia, or in cases of poor metabolic control,

increasing the incidence of hyperglycemia. This

model is preliminary and is in need of testing in
paradigms that explore neuropathology and cognitive outcome together with hypothesized key
periods of disease variable effects.

CONTROVERSIES
Despite evidence of cognitive deficits in children
with T1DM, there are studies that have not
corroborated these findings. One of key studies
which did not find any deficits came from the
Diabetes Control and Complications Trial
Research Group (1996). The results were based
on a 9-year study of 1441 adolescents and adults
(age 1339 years) with T1DM in 29 clinical
centers across the United States and Canada. Two
hundred and seventy eight patients were tested
using an extensive neuropsychological battery,
including measures of intelligence, memory,

NEUROCOGNITIVE CORRELATES OF DIABETES

attention, executive function, motor function, and

visuospatial ability. Independent variables of
interest were the treatment regimen, intensive
(three or more injections per day or continuous
subcutaneous infusion from an external pump) or
conventional (one or two injections per day)
treatment groups. During the period of study,
participants were to report to their center if they
experienced severe hypoglycemia, defined as an
episode in which the patient was in need of
intervention from another person, their glucose
level was below 2.78 mmol/l or the clinical manifestations were reversed by the administration of
oral, subcutaneous, or intravenous glucose. Those
who suffered two or more episodes of severe
hypoglycemia during the test period were separated out as a group from those who did not.
Testing was conducted at baseline, years 2, 5, and
7, and at the end of the 9-year period. The
main neuropsychological measure of interest
was worsening of test scores over the period of
testing, with decrements of performance from
the baseline state the key dependent variable of
interest. The main findings were that there was no
evidence that intensive treatment increases the
likelihood of cognitive worsening over time, and
that there was no evidence to show repeated
episodes of severe hypoglycemia were associated
with the development of clinically significant or
subtle cognitive impairment over the period of
testing.
Many diabetologists have taken these findings
as conclusive evidence that hypoglycemia is not
harmful to cognitive function in patients with
T1DM. However, there are some limitations to
the DCCT study that must be noted. First, the use
of a score of worsening is not likely to find change
in adolescents or young adults who are still
undergoing development and increasing in cognitive ability over time. Second, without a control
group it is difficult to make conclusions about the
scores at any one particular test period with regard
to how hypoglycemia affects cognitive function in
a normative sense. Third, the researchers acknowledge that they could not directly address
a causal relation between brain damage (as
reflected in poor neuropsychological outcome)
and hypoglycemia because the neurobehavioral
performance and hypoglycemic status were out-

47

comes of the trial. As such, their research design

differed significantly from the majority of studies
in which cognitive dysfunction has been found in
this population, as most studies have not treated
hypoglycemia as an outcome variable, but as a
predictor. Fourth, the test measures used to assess
functions other than intelligence were normed on
adults, typically over the age of 16, which may
have led to floor effects in the adolescent group,
thus obscuring any differences between those
in recurrent versus nonrecurrent hypoglycemia
groups. Finally, in looking at individual cases in
this sample, there were some patients who did
show some clinical worsening by year 5 (n 23).
These results were not considered in the overall
discussion of group differences, as they were felt
to be due to neurological damage resulting from
mild repeated episodes of hypoglycemia prior to
baseline. There is an insufficient rationale given
for why these cases were not discussed, and
whether they represent natural variation in the
population from which patients were sampled.
In short, the methodological limitations of the
DCCT study are worth note. However, this study
is not atypical from others which have found
cognitive dysfunction in patients with T1DM, in
that it suffers from similar difficulties with patient
selection, operational definitions of disease variables, and selection of neuropsychological measures. Therefore, the results should be considered
on par with those of other studies which have
been conducted and which have found significant
differences.
A second controversy that has arisen is that
there is no consistency among studies in the size
of the effects found. Many researchers note that
when deficits occur, they are typically within 12
standard deviations of normative performance,
rather than in the clearly impaired range of
below 2 standard deviations. In terms of clinical
significance, there have been no studies which
have explored whether children with cognitive
deficits associated with T1DM are more likely to
come to the attention of neuropsychologists for
assessment, or whether they have more need for
rehabilitation of their deficits. This is an area in
need of further research.
Another controversy that has arisen in the
literature on cognitive deficits accompanying

48

MARY DESROCHER & JOANNE ROVET

T1DM is the suggestion that the effects of hypoglycemia and hyperglycemia are diffuse in terms
of their impact on brain structure, rather than
specifically affecting any one portion of the
brain. Certainly there is evidence, as reviewed
above, to support the suggestion that neuropsychological deficits, when present, may affect
diffuse areas of the brain. Further, there is also
evidence that there are widespread pathological
changes that can accompany complications of
T1DM, as noted in the sections on disease variables above. However, one must be aware that the
majority of the studies have been retrospective,
with several years passing between age of diagnosis and time of testing. Many studies have
focused on adults, largely due to the difficulty
with securing child patients for testing. The question then becomes one of when brain abnormalities arise, a question which cannot be answered
given the state of current research. More studies
are needed to explore the brains of children with
T1DM over the course of development. In addition, linking current knowledge of neurological
development and correlated function would be
useful in guiding future research into this area of
inquiry. It is hoped that the proposed model outlined above will provide a framework for exploring these questions in more detail.

LIMITATIONS AND FUTURE

DIRECTIONS
There are many methodological challenges in
conducting research on the neurocognitive outcome of T1DM. One difficulty is that the majority
of studies are retrospective in nature, leading to a
confound of diabetes variables, such as age of
onset, duration, and complications. This is
particularly problematic in that inferences about
childhood cognitive function are made by testing
adults, who have been diabetic for at least a
decade, if not more. The effects of one disease
variable cannot be separated out from that of
others explored in the same study. The two
variables which have been most confounded in
the extant research are age of onset and presence
of hypoglycemia. As noted above, children with
early onset T1DM are more prone to severe

hypoglycemia as compared to children with later

onset. As such, it is difficult to determine to what
extent a given cognitive function is impaired by
either variable. One means of conducting research
into the effects of diabetes variables on outcome
would be the use of prospective longitudinal
studies, such as those of Northam et al. (1995,
1999, 2001).
Another limitation of this area of research is
the selection of samples. The majority of children
who are selected for diabetes research come from
hospital settings, who may not be representative
of children in the community receiving care from
family physicians. These children may be receiving more intensive treatment, or may be those
who have suffered complications of T1DM and
must be followed to ensure that their status does
not worsen. Findings in this type of sample may
not be generalizable to noninstitutional settings.
Other sampling difficulties that are of note in the
literature on neurocognitive outcome are that
participants tend to vary widely in age (necessitating use of different test measures for different
age groups), the inclusion of improper or a lack of
control groups, and small sample sizes.
Related to sampling issues are issues of how
the characteristics of these groups are operationalized. In particular, researchers vary in their
definition of hypoglycemic episodes, with some
researchers taking a very liberal estimate of
defining severe hypoglycemia as any perturbation
below a blood glucose level of 3 mmol/l, with
others defining this variable as the presence of
convulsions or coma. In addition, researchers
differ as to whether they consider loss of consciousness as being indicative of severity, and
others considering three or more seizures to be
the standard. Many researchers rely on parental
self reports of seizures to quantify episodes of
hypoglycemia, which may be subject to bias, with
a few relying on information from medical charts.
Also, very few studies report ambient blood
glucose levels at time of testing, a variable
which must be considered to affect performance
and thus make group comparisons difficult.
A third limitation is inconsistencies in test
design, measures, and hypotheses. Ryan (1999)
notes that small cross-sectional studies show
more effects than prospective longitudinal studies,

NEUROCOGNITIVE CORRELATES OF DIABETES

possibly due to practice effects associated with

re-testing samples. Researchers tend to use measures to which they have access, rather than
selecting them with a theoretical basis in mind.
This makes it difficult to compare results across
studies, as the measures are not comparable.
Future research should aim for multicenter studies
which use tests that are age appropriate, and could
consist of clinical and experimental measures,
the latter to explore components of function.
The majority of studies on neurocognitive function in this population consist mainly of cataloging deficits in a selective, nonsystematic way.
This makes it difficult to link diabetes variables
to any particular outcome. Test batteries are
chosen without any particular hypothesis in
mind, and do not allow for full exploration of
the core cognitive deficits. Thus, in future research, test selection should be guided by both
knowledge of the developmental sequencing of
cognitive function and a firm grounding in developmental neurology.
A further limitation is related to the use of
clinical tests to explore cognitive dysfunction.
These tests are often developed atheoretically
and may tap into a number of domains. For
example, it is difficult to determine whether the
deficits seen in motor tasks involve slowed speed
of motor output alone, or whether they reflect an
impairment in higher order, psychomotor processing. This question may be addressed through use
of temporal imaging measures, such as ERPs or
MEGs, that are able to separate out earlier psychomotor task components from later motor components in tasks requiring a motor response.
Many tasks of visuospatial function rely not
only on constructive ability, identification, or
perceptual function, but may also be dependent
on visual attention or executive function. Using
basic experimental tasks of perception may be
useful in determining the level at which deficits
occur. Further studies using laboratory tasks are
warranted to examine how different aspects of
diabetes individually affect different attention
components, as only one study to date has
explored this issue. As yet, different subtypes of
memory deficits have not been mapped directly to
the specific diabetes-related variables. It would
also be of interest to determine if selective deficits

49

in episodic versus semantic memory are evident

in this population, as this would support the
suggestion that hippocampally based memory
functions are most affected by the neurological
effects of this disease (Auer & Siesjo, 1988).
Diabetes appears to contribute to deficits on
tasks such as the Wisconsin Card Sorting Test;
however, it is difficult to determine if this is a
deficit in executive functioning, working memory,
attention, or some combination thereof. Whether
such findings are truly due to a deficit of some
central executive function, or are secondary to
deficits of attention and memory, is a topic for
future research. While children with later-onset
diabetes show small but significant deficits in
VIQ, these may be secondary to other difficulties
such as problems with processing speed or semantic memory.
Much still remains to be researched in the area
of T1DM in childhood. To date, research has
provided only a basic understanding of the general extent of neurocognitive deficits in children
who have this disorder, while the structurefunction relationships that are affected by the
array of diabetes variables at specific times in
development have yet to be learned. Theoretically
motivated research will not only serve to further
our comprehension of the array of deficits that are
seen in this disorder, it may also provide physicians, psychologists, educators, children and
families with some answers to the question of
long-term outcome.
ACKNOWLEDGEMENTS
Several people were instrumental in the preparation of
this review. Drs. Kusiel Perlman and Denis Daneman
at the Hospital for Sick Children were helpful in
providing impressions of their patient populations and
the difficulties that they faced in terms of cognitive
functioning. Also appreciated are the contributions of
Lila Elkhadem, Mirisse Foroughe, Laura Gates, Miguel
Rivera, and Valorie Salimpoor in conducting literature
reviews. The authors thank the reviewers of the original
submitted version of this paper, and have incorporated
many of their suggestions into the final product. This
work has been supported by past grants from the
Canadian Diabetes Association (JR), the Banting and
Best Foundation (JR and MD), and by a current internal
grant from York University (MD).

50

MARY DESROCHER & JOANNE ROVET

REFERENCES
Ack, M., Miller, I., & Weil, W. (1961). Intelligence
of children with diabetes mellitus. Pediatrics, 28,
764770.
Amiel, S.A. (1996). Studies in hypoglycaemia in
children with insulin-dependent diabetes mellitus.
Hormone Research, 45, 285290.
Auer, R.N., & Siesjo, B.K. (1988). Biological differences between ischemia, hypoglycemia, and epilepsy. Annals of Neurology, 24, 699707.
Bjorgass, M., Gimse, R., Vik, T., & Sand, T. (1997).
Cognitive Function in Type 1 children with diabetes
with and without episodes of severe hypoglycaemia.
Acta Paediatrica, 86, 148153.
Bloch, C.A., Clemons, P., & Sperling, M.A. (1987).
Puberty decreases insulin sensitivity. Journal of
Pediatrics, 110, 481487.
Brady, E. (2000). Cognitive and psychosocial functioning in preschool-aged children with T1DM.
Diabetes, 5, 49.
Caprio, S., Plewe, G., Diamond, M.P., Simonson, D.C.,
Boulwar, S.D., Sherwin, R.S., & Tamborlane, W.V.
(1989). Increased insulin secretion in puberty:
A compensatory response to reductions in insulin
sensitivity. Journal of Pediatrics, 114, 963967.
Chalmers, J., Risk, M.T., Kean, D.M., Grant, R.,
Ashworth, B., & Campbell, I.W. (1991). Severe
amnesia after hypoglycemia. Diabetes Care, 14,
922925.
Daneman, D., Frank, M., Perlman, K., Tamm, J., &
Ehrlich, R. (1989). Severe hypoglycemia in children
with insulin-dependent diabetes mellitus: Frequency
and predisposing factors. Journal of Pediatrics,
115(5, Pt. 1), 740742.
Davis, E.A., Keating, B., Byrne, G.C., Russell, M., &
Jones, T.W. (1998). Impact of improved glycaemic
control on rates of hypoglycaemia in insulin dependent diabetes mellitus. Archives of Diseases in
Childhood, 78, 111115.
Davis, E.A., Trundle, C.L., Ives, J., Robins, P.D., &
Jones, T.W. (2002). High prevalence of structural
CNS abnormalities in children with early onset
diabetes. Diabetes, 51, A3.
DeJong, R.N. (1977). CNS manifestation of diabetes mellitus. Postgraduate Medical Journal, 61,
101107.
Diabetes Control and Complications Trial Research
Group. (1996). Effects of intensive diabetes therapy
on neuropsychological function in adults in the
diabetes control and complications trial. Annals of
Internal Medicine, 124, 379388.
Dodge, P., Prensky, A., & Feigin, R. (1975). Nutrition
and the developing nervous system. St. Louis: C.V.
Mosby.

Donaghue, K.C., Bonney, M., Simpson, J.M.,

Schwingshandl, J., Fung, A.T., Howard, N.J., &
Silink, M. (1993). Autonomic and peripheral nerve
function in adolescents with and without diabetes.
Diabetes Medicine, 10, 664671.
Drash, A., LaPorte, R., Becker, D., Orchard, T., Wagener,
D., Rabin, B., & Kuller, L. (1983). Epidemiological
studies in children with diabetes mellitus and their
families: The Allegheny County Registry Experience. In G. Chiumello & M. Sperling (Eds.), Recent
progress in pediatric endocrinology (pp. 4, 125136).
New York: Raven Press.
Eeg-Olofsson, O., & Petersen, I. (1966). Childhood
diabetic neuropathy: A clinical and neuropsychological study. Acta Pediatrica Scandinavica, 55,
163176.
Ferguson, S.C., Blane, A., Perros, P., McCrimmon, R.J.,
Best, J.J.K., Wardlaw, J., Deary, I.J., & Frier, B.M.
(2003). Cognitive ability and brain structure in Type 1
diabetes: Relation to microangiopathy and preceding
severe hypoglycemia. Diabetes, 52, 149156.
Frier, B. (2001). Hypoglycaemia Clinical consequences and morbidity. International Journal of
Clinical Psychology: Supplement 11, 3, 5155.
Golden, C.J. (1981). The Luria-Nebraska Childrens
Battery: Theory and formulation. In G.W. Hynd &
J.E. Obrzut (Eds.), Neuropsychological assessment
and the school-age children: Issues and procedures.
New York: Grune and Stratton.
Golden, M.P., Ingersoll, G.M., Brack, C.J., Russell,
B.A., Wright, J.C., & Huberty, T.J. (1989). Longitudinal relationship of asymptomatic hypoglycemia
to cognitive function in IDDM. Diabetes Care, 12,
8993.
Gonder-Frederick, L.A., Cox, D.J., Driesen, N.R.,
Ryan, C.M., & Clarke, W.L. (1994). Individual
differences in neurobehavioral disruption during
mild and moderate hypoglycemia in adults with
IDDM. Diabetes, 43, 14071412.
Gschwend, S., Ryan, C., Atchison, J., Arslanian, S., &
Becker, D. (1995). Effects of acute hypoglycemia
on mental efficiency and counterregulatory
hormones in adolescents with insulin-dependent
diabetes mellitus. Journal of Pediatrics, 126,
178184.
Hagen, J.W., Barclay, C.R., Anderson, B.J., Freeman,
D.J., Segal, S.S., Bacon, G., & Goldstein, G.W.
(1990). Intellectual functioning and strategy use in
children with insulin-dependent diabetes mellitus.
Child Development, 61, 17141727.
Health Canada. (1999). Health care expenditures in
Canada from 19901995. Ottawa: Health Canada
Press.
Hershey, T., Bhargava, N., Sadler, M., White, N.H., &
Craft, S. (1999). Conventional versus intensive
diabetes therapy in children with Type 1 diabetes:

NEUROCOGNITIVE CORRELATES OF DIABETES

Effects on memory and motor speed. Diabetes Care,

22, 13181324.
Hershey, T., Craft, S., Bhargava, N., & White, N.H.
(1997). Memory and insulin-dependent diabetes
mellitus (IDDM): Effects of childhood onset and
severe hypoglycemia. Journal of the International
Neuropsychological Society, 3, 509520.
Hills-Briggs, F., & Echemendia, R.J. (2001). Association of metabolic control with problem-solving
skills. Diabetes Care, 24, 959.
Holmes, C.S. (1990). Neuropsychological sequelae of
acute and chronic blood glucose disruption in adults
with insulin-dependent diabetes mellitus. In C.
Holmes (Ed.), Neuropsychological and behavioral
aspects of diabetes (pp. 122154). New York:
Springer-Verlag.
Holmes, C.S., Hayford, J.T., Gonzalez, J.L., &
Weydert, J.A. (1983). A survey of cognitive
functioning at different glucose levels in diabetic
persons. Diabetes Care, 6, 180185.
Holmes, C.S., Koepke, K.M., Thompson, R.G., Gyves,
P.W., & Weydert, J.A. (1984). Verbal fluency
and naming performance in T1DM at different
blood glucose concentrations. Diabetes Care, 7,
454459.
Holmes, C.S., & Richman, L.C. (1985). Cognitive
profiles of children with insulin-dependent diabetes
mellitus. Journal of Developmental and Behavioral
Pediatrics, 6, 323326.
Kaufman, F.R., Epport, K., Engilman, R., & Halvorson,
M. (1999). Neurocognitive functioning in children
diagnosed with diabetes before age 10 years.
Journal of Diabetes and its Complications, 13,
3138.
Kovacs, M., Ryan, C., & Obrosky, D. (1994). Verbal
intellectual and verbal memory performance of
youths with childhood-onset insulin dependent
diabetes mellitus. Journal of Pediatric Psychology,
19, 475483.
Lawson, M.L., Sochett, E.B., Chait, P.G., Balfe, J.W., &
Daneman, D. (1996). Effect of puberty on markers
of glomerular hypertrophy and hypertension in
IDDM. Diabetes, 45, 5155.
Lobmann, R., Smid, H.G., Pottag, G., Wagner, K.,
Heinze, H.J., & Lehnert, H. (2000). Impairment and
recovery of elementary cognitive function induced
by hypoglycemia in type-1 patients with diabetes
and healthy controls. Journal of Clinical Endocrinology and Metabolism, 85, 27592766. Erratum in:
Journal of Clinical Endocrinology and Metabolism,
85, 4818.
Macleod, K.M., Hepburn, D.A., Deary, I.J., Goodwin,
G.M., Dougall, N., Ebmeier, K.P., & Frier, B.M.
(1994). Regional cerebral blood flow in IDDM
patients: Effects of diabetes and recurrent severe
hypoglycaemia. Diabetologia, 37, 257263.

51

Matyka, K., Ford-Adams, M., & Dunger, D.B. (2002).

Hypoglycaemia and counterregulation during childhood. Hormone Research, 57(Suppl. 1), 8590.
McCall, A.L. (1992). The impact of diabetes on the
CNS. Diabetes, 31, 557570.
McCall, A.L., & Figlewicz, D.P. (1997). How does
diabetes mellitus produce brain dysfunction?
Diabetes Spectrum, 10, 2532.
McCarthy, A.M. Lindgren, S., Mengeling, M.A.,
Tsalikian, E., & Engvall, J.C. (2002). Effects
of diabetes on learning in children. Pediatrics,
109, E9.
Meuter, F., Thomas, W., Gruneklee, D., Gries, F., &
Lohmann, R. (1980). Psychometric evaluation of
performance in diabetes mellitus. Hormone and
Metabolic Research Supplement, 9, 917.
Mitrakou, A., Ryan, C., Veneman, T., Mokan, M.,
Jenssen, T., Kiss, I., Durrant, J., Cryer, P., & Gerich,
J. (1991). Hierarchy of glycemic thresholds for
counterregulatory hormone secretion, symptoms,
and cerebral dysfunction. American Journal of
Physiology, 260, E67E74.
Mooradian, A.D. (1988). Diabetic complications of the
nervous system. Endocrine Reviews, 9, 346356.
Northam, E., Anderson, P.J., Jacobs, R., Hughes, M.,
Warne, G.L., & Werther, G.A. (2001). Neuropsychological profiles of children with T1DM 6 years
after disease onset. Diabetes Care, 24, 17721787.
Northam, E., Anderson, P.J., Werther, G., Adler, R., &
Andrewes, D. (1995). Neuropsychological complications of insulin dependent diabetes mellitus in
children. Child Neuropsychology, 1, 7487.
Northam, E., Anderson, P.J., Werther, G.A., Warne,
G.L., & Andrewes, D. (1999). Predictors of change
in the neuropsychological profiles of children with
T1DM 2 years after disease onset. Diabetes Care,
22, 14381444.
Northam, E., Bowden, S., Anderson, V., & Court,
J. (1992). Neuropsychological functioning in
adolescents with diabetes. Journal of Clinical and
Experimental Neuropsychology, 14, 884900.
Perros, P., Deary, I.J., & Sellar, R.J. (1997). Brain
abnormalities demonstrated by magnetic resonance
imaging in adult IDDM patients with and without
history of recurrent severe hypoglycemia. Diabetes
Care, 20, 10131018.
Reich, J.N., Kaspar, J.C., Puczynski, M.S., Puczynski,
S., Cleland, J., DellAngela, K., & Emanuele, M.A.
(1990). Effect of a hypoglycemic episode on
neuropsychological functioning in children with
diabetes. Journal of Clinical and Experimental
Neuropsychology, 12, 613626.
Rovet, J.F. (1999). Neurological sequelae of pediatric
diabetes. In K.O. Yeates, M.D. Ris, & H.G. Taylor
(Eds.), Pediatric neuropsychology: Research, theory, and practice. New York: Guilford Press.

52

MARY DESROCHER & JOANNE ROVET

Rovet, J.F., & Alvarez, M. (1997). Attentional functioning in children and adolescents with IDDM.
Diabetes Care, 20, 803810.
Rovet, J.F., & Ehrlich, R.M. (1988). Effect of
temperament on metabolic control in children with
diabetes mellitus. Diabetes Care, 11, 7782.
Rovet, J.F., Ehrlich, R.M., & Czuchta, D. (1990).
Intellectual characteristics of children with diabetes
at diagnosis and one year later. Journal of Pediatric
Psychology, 15, 775788.
Rovet, J.F., Ehrlich, R.M., Czuchta, D., & Akler, M.
(1993). Psychoeducational characteristics of children and adolescents with insulin-dependent diabetes mellitus. Journal of Learning Disabilities, 26,
722.
Rovet, J.F., Ehrlich, R.M., & Hoppe, M. (1988). Specific
intellectual deficits in children with early onset
diabetes mellitus. Diabetes Care, 10, 510515.
Ryan, C.M. (1990). Neuropsychological consequences
and correlates of diabetes in childhood. In C.S.
Holmes (Ed.), Neuropsychological and behavioral aspects of diabetes (pp. 5884). New York:
Springer-Verlag.
Ryan, C.M. (1999). Memory and metabolic control in
children (editorial). Diabetes Care, 22, 12391241.
Ryan, C.M., Atchison, J., Puczynski, S., Puczynski, M.,
Arslanian, S., & Becker, D. (1990). Mild hypoglycemia associated with deterioration of mental
efficiency in children with insulin-dependent
diabetes mellitus. Journal of Pediatrics, 117, 3238.
Ryan, C.M., Longstreet, C., & Morrow, L. (1985). The
effects of diabetes mellitus on the school attendance
and school achievement of adolescents. Child:
Care, Health, and Development, 11, 229240.
Ryan, C.M., Vega, A., & Drash, A. (1985). Cognitive
deficits in adolescents who developed diabetes early
in life. Pediatrics, 75, 921927.

Ryan, C.M., Vega, A., Longstreet, C., & Drash, A.

(1984). Neuropsychological changes in adolescents
with insulin dependent diabetes. Journal of Consulting and Clinical Psychology, 52, 335342.
Ryan, C.M., Williams, T.M., Finegold, D.N., &
Orchard, T.J. (1993). Cognitive dysfunction in
adults with type 1 (insulin-dependent) diabetes
mellitus of long duration: Effects of recurrent
hypoglycaemia and other chronic complications.
Diabetologia, 36, 329334.
Sansbury, L., Brown, R.T., & Meacham, L. (1997).
Predictors of cognitive functioning in children with
insulin-dependent diabetes mellitus: A preliminary
investigation. Childrens Health Care, 26, 197210.
Tallroth, G., Ryding, E., & Agardh, C.D. (1992).
Regional cerebral blood flow in normal man during
insulin-induced hypoglycemia and in the recovery
period following glucose infusion. Metabolism, 41,
717721.
Traisman, H.S. (1980). Management of juvenile
diabetes (3rd ed.). St. Louis, MO: Mosby.
Tsalikian, E., Daneman, D., Becker, D.J., Crumrine,
P.K., & Drash, A.L. (1980). EEG changes during
therapy of diabetic ketoacidosis in children. Journal
of Pediatrics, 96, 11151116.
Vlassara, H., Brownlee, M., & Cerami, A. (1983).
Excessive nonenzymatic glycosylation of peripheral
and central nervous system myelin components in
diabetic rats. Diabetes, 32, 670674.
Wolters, C.A., Yu, S.L., Hagen, J.W., & Kail, R. (1996).
Short-term memory and strategy use in children with
insulin-dependent diabetes mellitus. Journal of Consulting and Clinical Psychology, 64, 13971405.
Yakovlev, P.I., & Lecours, A. (1975). The myelogenetic
cycles of regional maturation of the brain. In A.
Minkowski (Ed.), Regional development of the
brain in early life (pp. 364). Oxford: Blackwell.