Search tips
Search criteria 


Logo of archdischArchives of Disease in ChildhoodVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Arch Dis Child. 2007 April; 92(Suppl1): A13–A15.
PMCID: PMC2066105


Endocrinology and diabetes

G/TUES/END1 A decade of trends in childhood diabetes in Wessex 1996 to 2005

M. Riddle, presenting, A. McAulay, S. Payne, P. Thomas. Poole Hospital NHS Trust, Poole, Dorset, UK

AimTo analyse data collected for children with diabetes <16 years in the Wessex region and assess significant trends from 1996 to 2005.

MethodData were collected annually by questionnaire from each of the 10 districts on: occurrence rates, length of stay (LOS) and incidence of diabetic ketoacidosis (DKA) at presentation, re‐admission rates with DKA, clinic staff numbers, and mean clinic HbA1c.

ResultsThere is a significant trend of increasing incidence (p = 0.001) and prevalence (p<0.001) of diabetes over the study period. The LOS for new patients (not in DKA) has shown a decrease (from 3.3 to 1.6 days), and varies between districts. This variation may be related to Paediatric Diabetic Nurse Specialist (PDNS) cover per patient (r = −0.54, p = 0.11). The incidence of DKA at presentation has remained constant (average = 23%). The re‐admission rates of DKA show a downward trend (6% in 1996 to 3.7% in 2004), but were not statistically significant (p<0.2). The incidence of DKA re‐admissions varies between districts from a mean of 2.1% to 9.6% in 2005. The incidence of DKA re‐admissions is not directly related to PDNS cover per patient (r = −0.06, p = 0.88). The number of consultants per 100 patients varied between 5.5 and 0.9 across the districts in 1996 and between 2.7 and 0.65 in 2005. The PDNS ratio per 100 patients varied between 1.5 and 0.5 in 2005. Mean HbA1c levels varied between districts but different assays hinder comparison.

ConclusionsThis is a large regional study providing 10 years of data. Both incidence and prevalence of childhood diabetes are increasing in the Wessex region. Fewer consultants are looking after more children. PDNS cover remains poor in some districts and may explain variation in LOS of new patients but not re‐admission with DKA rate. Collection of these data has allowed comparisons between districts and has informed changes in service provision.

G/TUES/END2 Mortality among 0–29 year olds with Type 1 diabetes in Yorkshire, UK

R. Feltbower1, presenting, R. Parslow1, H. J. Bodansky2, C. Stephenson1, F. Campbell3, P. McKinney1. 1University of Leeds, Leeds, UK; 2The General Infirmary at Leeds, Leeds, UK; 3St James's Hospital, Leeds, UK

AimsTo investigate mortality in a population‐based cohort of children and young people with Type 1 diabetes (T1D) using a high quality specialist register in Yorkshire.

MethodsDetails of all patients diagnosed with T1D aged 0–14 years (1978–2004) and aged 15–29 years (1991–2004) from West Yorkshire were included. Linkage to the NHS Central Register ascertained follow‐up status for each individual. Person‐years at risk analysis calculated expected numbers of deaths using UK mortality rates. Standardised mortality ratios (SMR) with 95% confidence intervals were derived using the Poisson distribution.

Results3349 and 895 patients aged 0–14 years and 15–29 years represented 52 173 person years of follow‐up. 10 patients (0.2%) were untraceable. Overall 110 (2.6%) patients died, of whom 79 were male (72%). 76 (2.3%) and 34 (3.8%) deaths occurred in those diagnosed aged 0–14 and 15–29 years. The overall SMR was 3.3 (95% CI 2.7–4.0); males demonstrated a significantly higher SMR (4.4; 3.5–5.5) than females (2.0; 1.3–2.8).

ConclusionPatients under 30 years with T1D have a threefold excess mortality risk compared to the background population, an effect significantly more pronounced in males. Cause of death and variation in mortality by period and age at diagnosis and duration of disease are being investigated.

G/TUES/END3 Young age at diabetes diagnosis associated with an increased risk of related serious adverse events

E. C. Ikazoboh1, presenting, A. Hutchison2, N. Hopper1, D. R. Jones3. 1Department of Paediatrics, Royal Surrey County Hospital, Guildford, Surrey, UK; 2Department of Pharmacoepidemiology, Postgraduate Medical School, University of Surrey, Guildford, Surrey, UK; 3Centre for Endocrinology Diabetes and Research (CEDAR), Royal Surrey County Hospital, Guildford, Surrey, UK

AimTo determine whether age at diabetes diagnosis is associated with the risk of subsequent related serious adverse events (SAEs) in children in the UK.

MethodsFor the years 1992 to 2006, all children with newly‐diagnosed diabetes were identified from the UK General Practice Research Database. This database has been widely used and validated for diabetes epidemiology and comprises anonymised medical and prescribing records from over 750 general practices in the UK. Diabetes related SAEs—diabetic ketoacidosis (DKA) and hypoglycaemia (including comas)—experienced by these patients were identified from the database, and using Cox proportional hazards, survival to first SAE in the 3 years following diagnosis was compared between 0–5, 6–10, and 11–18‐year‐olds.

Results237 out of 1564 children experienced SAEs. Event‐free survival was lowest in the youngest age group: after 3 years, 21.9% of 0–5‐year‐olds had experienced an SAE compared with 14% and 13.5% of the 6–10 and the 11–18‐year‐olds, respectively. Children aged 0–5 years were 1.8 times more at risk of experiencing an SAE than 11–18‐year‐olds, while there was no difference in risk between 6–10 and 11–18‐year‐olds. Corresponding hazard ratios (HR) were: 1.82 (95% CI 1.33–2.49; p<0.001) for 0–5 year olds and HR 1.07 (95% CI 0.79–1.44; p 0.68) for 6–10 year olds (reference: 11–18‐year‐olds). Analysis by type of SAE showed a trend of increasing episodes of DKA (from 2.9 to 5.7%) and decreasing hypoglycaemic episodes (from 19.1 to 7.7%) with an increase in age. There was no difference in risk between boys and girls (HR 1.02; 95% CI 0.79–1.31).

ConclusionChildren aged under 6 at diagnosis are approximately at a twofold increased risk of diabetes‐related SAEs, especially of hypoglycaemic episodes than older children.

G/TUES/END4 Body composition in children shortly after diagnosis of type 1 diabetes and after the first six weeks of insulin treatment

N. Davis1, presenting, J. Bursell1, J. Warner2, J. Gregory3. 1Bristol Children's Hospital, Bristol, UK; 2University Hospital of Wales, Cardiff, UK; 3Wales College of Medicine, Cardiff, UK

IntroductionRapid weight gain is often observed following initiation of insulin therapy in children with T1DM. Much of this is usually attributed to changes in hydration but little information exists on the precise body composition changes. We aimed to examine body composition changes of children with T1DM shortly after diagnosis and after 6 weeks of insulin treatment.

MethodsBody composition was measured by whole body DEXA scan and by BMI in 30 patients (18 male) with T1DM from 3 to 10 days after diagnosis and six weeks later, and on one occasion in 14 controls (8 male).

ResultsThe BMI SDS (−1.6 (1.0)) of the female patients was lower than male patients (0.0 (1.2)) despite similar pubertal development but there was no difference in percentage body fat between genders. Body composition normalised and there were no gender differences at six weeks. HbA1c at diagnosis was associated with the change in body fat in grams and the change in BMI SDS (r = 0.60, p<0.05 and r = 0.57, p<0.005 respectively).

Table thumbnail
Abstract G/TUES/END4

DiscussionInsulin deficiency at diagnosis of T1DM causes a catabolic state which is predominantly lipolytic with normalisation of body composition within 6 weeks of treatment despite disproportionate gains in fat (42%) compared to lean tissue (58%). Girls have a lower BMI SDS than boys shortly after diagnosis and our data would suggest the body composition changes over 6 weeks reflect the severity of the metabolic prodrome.

G/TUES/END5 Changes in body composition on stimulant medication in 13 children (12 boys) with attention deficit hyperactivity disorder

A. Poulton1, presenting, E. Melzer2. 1University of Sydney, Sydney, New South Wales, Australia; 2Nepean Hospital, Penrith, New South Wales, Australia

AimsGrowth attenuation is a common side effect when stimulant medication is used to treat children with attention deficit hyperactivity disorder (ADHD). The aim of this prospective study was to investigate changes in body composition using dual‐energy x ray absorptiometry (DXA) in children starting treatment with stimulant medication.

MethodsChildren newly diagnosed with ADHD were recruited from paediatric private practice. Stimulant medication was started when clinically indicated, with DXA taken pre‐treatment and repeated after an average of 7 months of treatment.

ResultsTo date 13 children (12 boys) aged 4.7–8.7 years (mean 7.1 (1.1) years) have had baseline and repeat DXA scans after 3.9–11.7 months (mean 7.0 (2.2) months) of treatment. There was a significant relationship between change in total tissue mass and change in bone mineral density (BMD) (r = 0.55, p = 0.05), with 2 of the 3 boys who lost more than 1 kg in weight recording a transient reduction in BMD.

Table thumbnail
Abstract G/TUES/END5

ConclusionIn this pilot sample, stimulant medication was associated with loss of percentage and total body fat and increases in lean tissue and bone mass. The correlation between change in weight and change in BMD suggests that bone mineralisation may occur more slowly in children who are losing weight. Weight loss on stimulant medication may be due to loss of fat; increases in lean tissue and bone mass may be related to increases in height.

G/TUES/END6 Vitamin D deficiency: prevention is cheaper than treatment

C. Zipitis, presenting, G. Markides, I. Swann. Burnley General Hospital, Burnley, UK

IntroductionVitamin D deficiency is a chronic condition contributing to general ill health. Since cessation of funding for supplementation there is a widespread belief that vitamin D deficiency is making a come‐back in our catchment area. This study aims to verify this and also to assess the cost effectiveness of reintroduction of vitamin D supplementation in our Trust, Burnley Health Care NHS Trust.

MethodsVitamin D deficient patients presenting between January 1994 and May 2005 were identified and data retrospectively collected from their case notes. The cost of treatment and theoretical cost of primary prevention for the Trust population were calculated using the COMA and current DoH recommendations.

ResultsFourteen cases were identified. 86% presented in the last 5 years. 93% of the affected patients were of Asian origin. The estimated incidence of vitamin D deficiency is 1 in 923 children overall and 1 in 117 in children of Asian origin. The average cost of treatment for each child presenting with Vitamin D deficiency is £2500 while the theoretical cost of prevention of a new case of vitamin D deficiency in the Asian population through primary prevention according to COMA guidance is £2400.

ConclusionsIn our Trust vitamin D deficiency is re‐emerging. The majority of our patients are of Asian origin. The cost of primary prevention for this high risk population compares favourably both medically and financially to treatment of established disease. We suggest that Primary Care Trusts provide funds for vitamin D supplementation of Asian children for at least the first 2 years of life.

G/TUES/END7 An audit of 56 dynamic endocrine tests among children in a district general hospital: urgent need for national and international guidelines

M. Ogundele, presenting, M. Al‐Jubouri, I. Ahmad. Whiston Hospital, Prescot, UK

AimsAn audit of dynamic endocrine studies (including dexamethasone suppression (DS), synachten (cortrosyn) stimulation, growth hormone (GH), thyrothropin releasing hormone (TRH) and gonadotropin‐releasing hormone (GnRH) stimulation tests; water deprivation and water loading tests) was undertaken to assess adherence of the referring paediatricians to local clinical guidelines and protocols. We also evaluated the adequacy of documentation of the test procedures in the patients' notes and the proportion of positive and negative results.

Patients and MethodA retrospective analysis of children undertaking any of the endocrine dynamic studies (short synachten‐, GnRH‐, and GH‐stimulation, DS and water deprivation tests) over a 32‐month period (Nov 2002 to June 2005) was carried out at a district general hospital.

ResultsA total of 53 patients undertook 56 endocrine dynamic tests during the study period (equal sex distribution with a median age of 11 years, ranging from 1–20 years). The most common study was water deprivation test in 34% (n = 19). Others were clonidine GH stimulation and short synachten tests 20% (n = 11) respectively; GnRH stimulation 14% (n = 8) and DS tests 12% (n = 7). Only 7% (n = 4) of the tests were reported as positive (2 of 8 GnRH tests, 1 of 11 short synacthen and 1 of 7 DS tests), based on predefined diagnostic criteria. Two of them (50%) were proven to be falsely positive on repeat performance. No information leaflets or explicit informed consent were available for the patients and their carers in the district hospital and in any of the two nearest regional referral centres for the dynamic endocrine tests.

ConclusionWe estimated that more than half of the tests requested were avoidable if simpler screening tests were done and stricter selection criteria were used. We therefore recommend that well validated Regional/National/International guidelines for dynamic endocrine testing in children, simple information leaflets and explicit consenting procures are urgently needed.

G/TUES/END8 Response to growth hormone treatment in small for gestational age children with 3M syndrome

P. Murray1, presenting, L. Patel1, D. Hanson2, G. Black2, P. Clayton1. 1Endocrine Sciences Research Group, University of Manchester, UK; 2Academic Unit of Genetics, University of Manchester, UK

IntroductionThe 3M/Gloomy Face syndrome is a recessive disorder characterised by intra‐ and extra‐uterine growth retardation and associated with mutations in a component of the ubiquitin ligase E3 complex, Cullin 7. Being born small for gestational age with failure to catch‐up, these children are eligible for GH treatment.

AimTo assess the growth response to GH treatment in 3M syndrome.

MethodsRetrospective analysis of data from patients classified as 3M syndrome in the Kabi International Growth Study (KIGS) and from our regional growth clinic.

ResultsFifteen patients were identified (4 female and 11 male), of whom 5 were local. Median age at starting GH was 7.9 (range 2.9–11.1) years and duration of therapy was 2.5 (1.2–11.3) years at mean GH dose of 0.05 (0.026–0.073) mg/kg/day. Height velocity (HV) and HV SDS increased over the first year of treatment (4.1 v 5.9 cm/year, p = 0.007 and −1.6 v +0.3 SDS, p = 0.0007). However there was no significant change in height SDS over the first year of GH treatment (−4.4 v −3.95 SDS, p = 0.4, delta Ht SDS +0.36 (+0.03 to +1.7)). In addition, median observed HV for the first year of treatment was lower than the predicted HV (n = 9, 5.2 v 6.9 cm/year, p = 0.02) using the “Ranke” SGA model1 (fig). Median change in height SDS over 4 years GH treatment (n = 6) was +0.65 SD (+0.4 to +2.3 SD).

figure ac07abs4.f1
Abstract G/TUES/END8.

ConclusionIn general the response to GH in 3M syndrome is poor when compared to other SGA children. One mechanism for clearance of the GH receptor involves ubiquitination: these data would support an investigation of the possible inhibitory role of cullin 7 in GH signalling.

Rankeet alJ Clin Endocrinol Metab200388 pp 125

G/TUES/END9 Isolated menarche experience from a regional centre

N. Lipscomb, presenting, T. Barrett, N. Shaw, J. Kirk. Birmingham Children's Hospital, West Midlands, UK

AimTo review our experience of patients with isolated menarche.

MethodCase note review of patients diagnosed with isolated menarche (vaginal bleeding but absent pubertal changes) in the last decade.

ResultsSeven cases were identified—all were prepubertal, had a normal height velocity, and bone age appropriate for chronological age (n = 6). There was no evidence of foreign body, genital pathology or CSA. Androgen levels were not raised and thyroid function tests (TFTs) were normal (n = 5). Oestrogen levels were undetectable (n = 5). Luteinising hormone releasing hormone (LHRH) testing performed in all cases of recurrent bleeding showed a prepubertal LH response, although the follicle stimulating hormone (FSH) response was marked in two. Pelvic USS (n = 6) showed a prepubertal state in 5, and was “normal” in the other. Only two patients (1, 5) had an examination under anaesthesia (EUA); both were normal. One girl developed adrenarche after her bleeding had ceased.

Table thumbnail
Abstract G/TUES/END9

DiscussionIsolated menarche appears to be self‐limiting, and is followed by normal pubertal development and menarche. Only one patient had a single vaginal bleed; definitions vary as to whether bleeding has to be recurrent. The extent to which patients with this diagnosis need to be evaluated is unclear from the literature; TFTs, LHRH testing and pelvic USS (showing a pronounced FSH rise) are commonly performed, with earlier series also performing an EUA and screening for fibrous dysplasia. While foreign body and genital pathology should be excluded, routine EUA in a patient with cyclical bleeding and a normal pelvic USS appears unnecessary. Our findings confirm that the pelvic USS shows a prepubertal appearance, with the presence of follicles in some cases supporting the hypothesis that transient uterine stimulation is implicated in the pathophysiology.

G/TUES/END10 6q24 TNDM: exploring a genotype‐phenotype correlation

E. J. Davis1, presenting, I. K. Temple1, D. J. G. Mackay2, D. Robinson2, J. Shield3. 1Southampton University & Hospitals Trust, Southampton, Hants, UK, 2Wessex Regional Genetics Laboratories, Salisbury, Wilts, UK; 3Bristol Children's Hospital, Bristol, Avon, France

IntroductionTransient Neonatal Diabetes Mellitus (TNDM) is a rare condition affecting approximately 1: 400 000 live births. Infants present within the first 6 weeks of age and usually require insulin. The diabetes resolves spontaneously at a mean age of 3 months but may recur later in life. The commonest cause of TNDM are aberrations of the imprinted TNDM locus at 6q24.

MethodsWe report the clinical characteristics of patients from a twelve year series of 76 cases of TNDM referred to our unit for genetic testing, and explore a genotype‐phenotype correlation.

Results41% have paternal UPD6 (group 1), inheritance of both paternal homologues of chromosome 6 with no contribution from the mother. 41% have a paternally inherited duplication (group 2) and 18% are due to a maternal methylation mutation (group 3) resulting in hypomethylation at the locus. All 3 mechanisms result in overexpression of the paternally expressed genes ZAC and HYMA1 at 6q24. Provisional analysis has shown no difference between the 3 groups for age of onset, severity or relapse of diabetes. However there are differences between the 3 groups for minor associated features such as macroglossia and umbilical hernia. Initial results also suggest that learning difficulties are more common in children with UPD6 and methylation mutations.

ConclusionsThese results provide evidence to support a generalised hypomethylation syndrome in group 3 and the presence of as yet unidentified imprinted genes on chromosome 6.

Articles from Archives of Disease in Childhood are provided here courtesy of BMJ Publishing Group