Rovet and Ehrlich, 199959816 children from 2 studies in Canada
Mean age at diagnosis 4.5 ± 3.0 years, 10 males
Controls: matched for socio-economic status, sex and age
9 children had had hypoglycaemic seizures
Evaluations at diagnosis, 1, 3 and 7 years
Verbal intelligence quotient
Perceptual, fine motor, visuomotor, visual memory, attention
Those with hypoglycaemic seizures more likely to decline than without (67% vs. 14%, p < 0.05)
Children with diabetes with history of seizure scored significantly lower (p < 0.01)
Methodology and analyses unclearProspective and retrospective data collection, nested case–control studyIIb–III
Hannonen et al, 200359921 children with type 1 diabetes and 10 healthy children in Finland
Aged 5–11years, diabetes duration 1–10 years
3 groups:
Children with type 1 diabetes with least 1 episode of severe hypoglycaemia (n = 11)
Children with type 1 diabetes and no history of hypoglycaemia (n = 10)
Controls (n = 10)
Wechsler Intelligence Scale for Children-Revised (WISC-R)
NEPSY (developmental neuropsychological assessment)
Attention: type 1 diabetes without history of severe hypoglycaemia had test scores that were significantly lower than controls (p < 0.05)
Phonological processes in type 1 diabetes with history of severe hypoglycaemia scored lower than controls (p < 0.05)
Memory: digit span forward both type 1 diabetes with and without scored lower compared with controls (p < 0.01)
Severe hypoglycaemia defined as unconsciousness/convulsions with blood glucose concentration < 3 mmol/l
Controls were volunteers from a local sports club or children of hospital staff
Significant (p < 0.05) differences between the people with diabetes with and without hypoglycaemia regarding: duration of type 1 diabetes, age at diagnosis and neurological impairment
Case–control, interviewed 2–3 sessions within 2 weeks;, examiner blinded to type 1 diabetes groupIII
Matyka et al, 199960329 prepubertal (over age 5 years) recruited from paediatric clinic in Oxford
Mean age 9.5 years, mean diabetes duration 3.4 years
Controls: 15 healthy children (siblings or friends of cases), mean age 9.5 years
Overall episodes of hypoglycaemia = 24
Study night 1: n = 13
Study night 2: n = 7
  1. Cognitive function tests such as: apiral maze, pegboard, immediate story recall, digit span
  2. Mood: Children’s Depression Inventory
  1. No significant differences found between 17 children with nocturnal hypoglycaemia compared with a control night
  2. Higher after a night of hypoglycaemia: median score 5 (range 2–8.5) vs. 3 (1.5–6.5) on control night, p = 0.03
Severe hypoglycaemia defined as blood glucose concentration <3.5 mmol/l on 2 successive 15-min measurements
Significant difference between those with hypoglycaemia and without for higher insulin dose, p < 0.0006
Case–control, 2 overnight glucose measurements at home with cognitive and mood assessment the next dayIII
Austin and Deary, 19996041441 patients with type 1 diabetes
Aged 13–39 years
Patients who had 5 or more hypoglycaemic incidents
patients who had no hypoglycaemic incidents
Cognitive scoreGeneral ability (mean ± SEM (n)):
Year 2: −3.32 ± 3.64 (8) vs. −0.04 ± 0.23 (1121)
Year 5: 1.06 ± 1.51 (50) vs. −0.10 ± 0.28 (943)
Year 7: 4.50 ± 2.41 (30) vs. 0.91 ± 0.56 (371)
Year 9: 1.11 ± 2.16 (18) vs. 1.00 ± 0.90 (139)
No significant difference in the cognitive scores when split into the topics spatial ability, processing speed, verbal, memory and finger tapping
Further analysis from DCCT subgroup onlyControlled studyIIa
Holmes and Richman, 198559642 children with type 1 diabetes from a university paediatric department, Iowa
Aged 6–16 years
Inclusion criteria:
IQ = 85–129
Study population split into groups (onset/duration):
early (age < 7 years)/long (≥ 5 years) n = 12
late (≥ 7 years)/short n = 9
early/short (< 5 years) n = 10
late/long n = 11
Verbal subtests
Median performance IQ scores significantly lower in early onset, longer duration group (p < 0.05)
Higher rates of reading and memory impairment in early onset long duration group
Word recognition ability, visual-motor skill, auditory verbal learning test and digit span were NSCross-sectionalIII
Rovet, 199060963 children with newly diagnosed type 1 diabetes at Hospital for Sick Children in Toronto
33 boys, mean age 7.32 ± 4.3 years
Controls: 40 siblings without diabetes under age 12 years
Neurocognitive function examined at diabetes onset (n = 63) and 1 year later (n = 60), cases compared with controls only at T0Cognitive function including general IQ, verbal and spatial abilities, memory, academic achievementNo significant difference in outcomes except one test (WISC-R vocabulary) differed between diagnosis and 1-year follow-up in the children with diabetes (p < 0.05)
Number of monthly low blood sugar readings (< 4.0 mmol/dl) without symptoms, with symptoms and episodes of unconsciousness, convulsions correlated positively with improved outcome over time
Number of subjects differed with almost every comparisonCohort and nested case–controlIIb–III
Davis et al, 199661112 children with type 1 diabetes from a paediatric clinic in Australia
6 boys, mean age 12.4 ± 2.6 years
Effects of hyperglycaemiaCognitive function:
8/12 had a decrease in performance IQ score when hyperglycaemic (p < 0.05) (as did the performance centile score)Subjects randomised to a euglycaemic state on one occasion and a hyperglycaemic state on anotherCrossover RCT, assessed on 2 occasions, 6 months apartIb
Sansbury, 199759728 children with type 1 diabetes > 1 year duration from a university medical centre in Atlanta, USA
12 boys, mean age 12.6 years
Age at diabetes onset
Duration of disease
Metabolic control (HbA1c)
Cognitive function:
Matching Familiar Figures Test
Child Behaviour Checklist
Increase in chronological age associated with decreased full-scale IQ (p < 0.004), arithmetic (p < 0.007), verbal fluency (p < 0.005), and block design (p < 0.01)
Increased type 1 diabetes duration associated with lower MFFT scores (p < 0.01)
Poor metabolic control associated with lower vocabulary subtest scores (p < 0.03)
Participants were more likely to be from a higher socio-economic status than non-participantsObservationalIII
Northam et al, 1998610116 patients from children’s hospital in Melbourne with newly diagnosed type 1 diabetes from 1990–1992
Aged 3–14 years, 55 boys
112 ‘well’ controls from schools, 54 boys
  1. General intelligence: WPPSI-R for children under 7 years old and WISC-R for children 7 years old and over
  2. Vocabulary
    Block design
    Speed of processing
  3. Child Behaviour Checklist
  1. At baseline no differences between 2 groups
  2. Significant differences between two years after baseline and baseline:
    Vocabulary p < 0.01
    Block design p < 0.05
    Learning p < 0.01
    Speed of processing p < 0.05
  3. No significant difference found
Analyses performed ANOVA and MANCOVA;, smaller developmental gains in type 1 diabetes compared with controls
Type 1 diabetes showed a less positive change in general intelligence scores
Case–control, 2 assessments: T1 = 3 months after type 1 diabetes onset and T2 = 2 years post-baselineIIb–III
Rovet and Alvarez, 1997600103 children with diabetes from a clinic in Toronto
Aged 9.3–18.3 years, diabetes duration > 2 years
100 controls
Attention (mean composite scores)
Intelligence (WISC-R and MMFFT)
Compared with controls select attention scores lower (p = 0.05)
Poorer performance by patients with early onset (< 6 years old, n = 51) (p = 0.01)
Seizure history (S+): S+ had lower verbal IQs than controls (p < 0.01), no difference found compared with children with diabetes and S−, S+ associated with poorer inhibit and focus aspects of attention
Control source: friends, siblings, cousins of study group, friends/children of hospital employees
Retrospective diabetes history data collection
Observational 3-year study period, subjects randomly assigned to 1 or 4 testsIII
McCarthy et al, 2002601 and 2003602Children with type 1 diabetes (n = 244)
Mean age 14.8 ± 3.2 years
Sibling control group (n = 110) and matched classmate control group (n = 209)
Academic performanceThe study found that current academic performance by children and young people with type 1 diabetes was not lower than the sibling control group or the matched classmate control group
The children and young people with type 1 diabetes performed better than their siblings on maths (mean stand score 115.0 vs. 111.1, p < 0.02) and core total (mean stand score 113. vs. 110.5, p < 0.04) and better than their matched classmates on reading (mean stand score 108.9. vs. 106.8, p < 0.04)
The study found lower achievement in children and young people with type 1 diabetes who had poor metabolic control than those with average control
Socio-economic status and parent ratings of behaviour problems were significantly correlated with academic achievement, medical variables added only slightly to predictive precision
Case–control studyIII
Wysocki et al, 2003605142 children and young people with type 1 diabetes
Aged 6 to 15 years
No association between occurrence or frequency of severe hypoglycaemia and cognitive function (intelligence quotient)Case–control studyIII
Kaufman et al, 199960655 children with type 1 diabetes
Aged 5 to 10 years
Association between neurocognitive test scores and hypoglycaemiaNo association found
Subjects with a history of hypoglycaemic seizures had lower scores on tests assessing memory skills, including short-term memory (p < 0.03)
Case–control studyIII
Bjorgaas et al, 199760715 children and young people with type 1 diabetes compared with healthy children/young people matched for age, gender and social background
Aged 9 to 16 years
Cognitive performanceNo difference found
Children/young people with type 1 diabetes who had experienced an episode of severe hypoglycaemia: those with onset of diabetes before the age of 5 years had lower psychomotor efficiency scores than those with onset of diabetes after the age of 5 years
Case–control studyIII
Golden et al, 198960823 children with type 1 diabetes
Aged 5.9 ± 1.8 years
Association between hypoglycaemia and results of the Stanford–Binet Intelligence ScaleNo association found
The relative frequency of asymptomatic hypoglycaemia correlated with scores on the abstract/visual reasoning scale
Case–control studyIII

From: Evidence tables

Cover of Type 1 Diabetes
Type 1 Diabetes: Diagnosis and Management of Type 1 Diabetes in Children and Young People.
NICE Clinical Guidelines, No. 15.2.
National Collaborating Centre for Women's and Children's Health (UK).
London: RCOG Press; 2004 Sep.
Copyright © 2004, National Collaborating Centre for Women’s and Children’s Health.

Apart from any fair dealing for the purposes of research or private study, criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, no part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

The rights of National Collaborating Centre for Women’s and Children’s Health to be identified as Author of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act, 1988.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.