Article Text


Endocrinology and diabetes

Statistics from


A. Hewitson, J. A. Sharp, J. J. Plumb, J. J. Buck. Department of Paediatrics, West Suffolk Hospital NHS Trust, Hardwick Lane, Bury St Edmunds, Suffolk

Introduction: It is recommended that children with diabetes should have access to advice 24 hours a day.

Aim: To determine 1) the response to a typical out of hours query from a diabetic patient, training offered for “sick day rules”; and 2) how well this matched the expectations of the PDSNs in the Anglia region.

Methods: Each Anglia hospital was rung 1–2 times by one researcher between 5pm and 11pm. Advice was requested for a 12 your old diabetic patient with a sore throat, pre tea blood sugar 26, alert, moderate ketonuria, and no vomiting. The time for response and status of health professional was noted. The respondent was informed of the telephone survey and asked what advice he/she would give. The availability of training, written guidelines for these calls, and feedback to the diabetes team was ascertained. PDSNs for the hospitals were contacted between 9am and 4pm and questioned regarding their opinion of the out of hours service.

Results: 25 calls were made to 16 hospitals. Average time for a response via switchboard was 4 min (range 1–10). Calls were transferred to children’s ward (4), patient assessment unit (4), casualty (3), and direct to on call doctor (14). Calls were taken by nurses (10), junior doctors (14), and consultant (1). 7/25 (40%) respondents (all SHOs) suggested immediate admission of the patient, 11/25 (44%) suggested an extra dose of insulin, and of these 9 specified type and amount and 2 suggested a repeat blood sugar. Three respondents suggested a call to NHS direct or GP. 5/25 (20%) reported training on sick day rules, 10/25 (40%) reported written guidelines available for staff taking calls. 20/25 (80%) would give feedback to the PDSN. 12/16 hospitals had a PDSN who could be contacted. Average patient load was 150 per WTE DSN. 9/12 (75%) reported training on sick day rules, 7/12 (58%) reported written guidelines for those taking calls. 10/12 (83%) said feedback was given to them. Many suggested deficiencies in the out of hours advice.

Conclusion: Response to a typical out of hours query from a child with diabetes was very variable in the Anglia Region and could result in inappropriate management. Provision of training in sick day rules, written guidelines for staff taking calls, and feedback to the diabetes team is patchy. Many offering advice lack confidence and the PDSNs perceive inadequacies in the service.


W. Sultan1, S. M. Ng1, D. M. Broadbent2, A. Ghatak1, S. F. Dixon1, C. S. Smith1, M. Didi1. 1Royal Liverpool Children’s Hospital Alder Hey; 2Royal Liverpool University Hospital

Aims: To evaluate the risk factors associated with presence of diabetic eye disease in children.

Methods: All children with type 1 diabetes aged 5–16 years who attended Alder Hey Hospital were invited in 1998 and 2002 for mydriatic 3 field 45 degree 35 mm photography with modified Wisconsin grading. Disease positive cases underwent slit-lamp biomicroscopy. The 70 patients who had assessments on both occasions were evaluated to identify risk factors for presence (E+) or absence (E-) of diabetic eye disease. The risk factors assessed were duration of diabetes (years), age at examination (years), age at diagnosis (years), pubertal status, mean HbA1c for the preceding 12 months, body mass index (BMI) SDS, early morning urine albumin creatinine ratio (ACR), and mean blood pressure.

Results: Prevalence of diabetic eye disease was 8/190 (4.2%) in 1998 and 7/164 (4.3%) in 2002. Seventy children from 1998 were re-examined in 2002. Six patients who had no diabetic eye changes in 1998 developed diabetic eye disease by 2002. Five of these six patients had background retinopathy. One patient had pre-proliferative retinopathy. One patient who had background retinopathy in 1998 had a normal assessment in 2002 (HbA1c 10.1% v 9.1%). Significant risk factors for eye disease in the population studied longitudinally using univariate analyses were duration of diabetes (9.8 years (SD 1.3) in E+ group v 6.7 years (2.0) in E-group, p 0.005) and age at assessment (13.7 years (1.0) in E+ group v 11.8 years (2.0) in E-group, p 0.005). Pubertal status, age at diagnosis, HbA1c, mean blood pressure, urine ACR, and BMI (SDS) were not found to be significant. Multivariate analysis found duration of diabetes as the only significant independent risk factor for presence of diabetic eye disease (p 0.01).

Conclusion: 1) There is evidence that progression of diabetic eye disease can occur during childhood years; and 2) the only independent risk factor for diabetic eye disease is the duration of diabetes.


S. A. Sukthankar, M. A. Lewis, N. J. Webb, N. D. Plant, D. A. Price, C. M. Hall. Departments of Paediatric Nephrology and Paediatric Endocrinology, Royal Manchester Children’s Hospital, Hospital Road, Pendlebury M27 4HA

Introduction: PTDM (particularly type II) following renal transplantation has a reported incidence of 4–41% in adults and children. This may be due to the introduction of newer immunosuppressive agents like tacrolimus.

Aim: We discuss our experience of PTDM from January 2002 to October 2003.

Results: Since 2002, 32 children underwent renal transplantation at our centre. Four (3 male and 1 female, three Caucasian, and one Asian, median age 12 years) developed PTDM. None had pre-transplant hyperglycaemia, one had family history of type II diabetes, and two were on growth hormone (mean 42 μg/kg/day) for 6 months pretransplant. Standard immunosuppression regimen included prednisolone (60 mg/m2/day, tapering to 10 mg/m2/day by 6 weeks), azathioprine (2 mg/kg/day), and tacrolimus (0.3 mg/kg/day, to maintain trough levels 10–15 μg/L in the first 6 weeks, and 5–10 μg/L thereafter). All had transient hyperglycaemia (maximum 20–45 mmol/L) within 12 h after transplantation, which settled after reduction or elimination of the glucose content in replacement intravenous fluids. Three developed persistent hyperglycaemia (13–26 mMol/L) after a mean of 5 weeks post-transplantation (10 days–10 weeks). One presented with diabetic ketoacidosis 5 month after transplantation. Serum insulin levels were low (6.2 mU/L, range 3.7–8.9), and anti Islet cell and anti GAD antibodies were undetectable. All four required insulin replacement therapy (0.5 to 1 U/kg/day of Mixtard30), in addition to reduction in the dose of prednisolone, and maintenance of tacrolimus trough levels in the lower range (3–5 μg/L). One stopped insulin after 10 weeks, one changed to oral metformin after 2 months (stopped after a further 3 weeks); and two are still on insulin (5 and 8 weeks).

Conclusion: PTDM (incidence 12.5%) in our population had a variable onset, duration, and symptoms. It seems to be transient, and responds to insulin replacement therapy. This indicates the need for increased awareness, close surveillance, and collaboration between nephrology and endocrinology for management of PTDM.


M. G. Gnanalingham1, A. Mostyn1, M. E. Symonds1, D. A. Giussani2, T. Stephenson1, D. Gardner1,2. 1Centre for Reproduction and Early Life, Institute of Clinical Research, Nottingham University; 2Department of Physiology, University of Cambridge

Introduction: Umbilical cord occlusion (UO) sufficient to restrict fetal blood supply by 30% results in a range of fetal endocrine adaptations, including increased fetal plasma cortisol, which would be predicted to promote organ maturation. The present study therefore aimed to determine if UO results in upregulation of specific mitochondrial proteins involved in energy regulation (that is, uncoupling protein (UCP) 1 or 2, voltage dependent anion channel (VDAC), and cytochrome c) within the fetal lung and brown adipose tissue (BAT).

Methods: Nine ewes were entered into the study, which were all chronically instrumented with fetal vascular catheters. Five fetuses were then subjected to 3 days UO beginning at 125 days gestation by automated compression of the umbilical cord, while the remaining four acted as controls (C). At 137 (SD 2) days gestation, all fetuses were humanely euthanased. UCP 1 and 2 mRNA abundance was measured by RT-PCR using oligonucleotide primers designed specifically to ovine UCP 1 and 2. The abundance of VDAC and cytochrome c mitochondrial proteins was determined by immunoblotting. Results are given as means (SD) in arbitrary units, as a ratio of a reference sample included on all gels. Statistical differences between groups was analysed by Mann-Whitney U test.

Results: There were no significant differences in body, BAT, or lung weights between groups, although UO lungs had a tendency to be smaller. UO resulted in enhanced UCP 2 mRNA in lungs (C 72.3 (3.0); UO 114.8 (5.4), p<0.01) and UCP 1 mRNA in BAT (C 70.0 (2.4); UO 118.6 (5.2), p<0.01). The abundance of VDAC and cytochrome c were also raised in both the lung and BAT (for example VDAC lung: C 25.8 (0.9); UO 35.8 (1.7), p<0.05).

Conclusions: Chronic UO results in precocious maturation of lung and BAT mitochondria. These changes may be important in promoting effective adaptation to the extra-uterine environment after birth.


S. M. Ng, S. C. Wong, M. Didi. Department of Endocrinology, Alder Hey Children’s Hospital, Liverpool, UK

Aims: To determine the head circumference and linear growth during the first 3 years in infants with congenital hypothyroidism (CH) in relation to the aetiology of CH and initial biochemical severity of thyroid function.

Methods: Head circumference and linear growth of 125 patients with CH was evaluated from diagnosis to 3 years of age. All infants had radionuclide scans prior to treatment. Patients were categorised into athyreosis, dysgenesis, and dyshormonogenesis. OFC SDS, length SDS, initial thyroid hormones, and age for normalisation of plasma TSH were compared between the groups.

Results: There were 125 children in the study: athyreosis (n = 34), dysgenesis (n = 73), and dyshormonogenesis (n = 18). No difference was found in gestation, birthweight, age starting L-T4, and initial dose of L-T4 (μg/kg/day) between groups. At diagnosis, confirmatory total thyroxine (T4) was significantly lower for athyreosis when compared with dysgenesis and dyshormonogenesis (p<0.05). Confirmatory TSH were significantly lower in dyshormonogenesis compared with the other two groups (p<0.0001). Normalisation of TSH in the dyshormonogenesis group occurred at 6 weeks compared with 12 months for athyreosis and dysgenesis based on our laboratory normal reference range (p<0.0001). At diagnosis, OFC SDS were similar in all three groups. Children with athyreosis showed significantly larger OFC compared with dysgenesis and dyshormonogenesis from 1 to 3 years (p<0.01). Length SDS were within one SDS of normal population standards and did not differ between the three groups. Pearson’s correlation of OFC SDS at 3 years of age showed a significant negative correlation with confirmatory total T4 at diagnosis (R −0.40, p 0.004). Multivariate analysis for OFC SDS at 3 years of age showed confirmatory T4 as the only independent risk factor.

Conclusion: Children with athyreosis show significantly larger OFC from 1 to 3 years of age compared with dysgenesis and dyshormonogenesis, independent of linear growth difference. Our data show that initial confirmatory T4 at diagnosis is an independent factor influencing head growth in the 3rd year of life.


S. Kalkan, W. F. Paterson, M. D. C. Donaldson. Department of Child Health, Royal Hospital for Sick Children, Glasgow, Scotland

Introduction: Tall stature, a major characteristic of Marfan’s syndrome, may be of concern to the family, particularly if the patient is a girl. Experience with treatment options—sex steroid or somatostatin analogue—for height reduction in girls is limited. We have evaluated our experience of oestrogen treatment in girls with Marfan’s syndrome attending the paediatric endocrine clinic in Glasgow between 1989 and 2003.

Methods: Retrospective casenote analysis. Outcome in terms of final/near final height data was compared in treated v untreated girls. Cardiovascular health was assessed by measuring aortic root diameter and blood pressure.

Results: The study cohort comprised four treated and six untreated girls. Final height data were available in all the treated and in four of six untreated girls. Treatment was started in the four girls at a chronological age of 10.03 (SD 2.12) years, mean height 154.97 (9.78) cm, Tanner breast stage B1 in three and B2 in one. Ethinyl oestradiol was given in the mean (SD) dose of 40.65 (24.14) μg/day for 4.44 (1.81) years. Final height of the four treated girls was significantly less than that of the four untreated girls (mean (SD; range) 173.5 (1.99; 170.6–175.1) cm v 185.9 (2.99; 183.9–190.3) cm; 95% CI −2.65 to −0.51). No deaths occurred in the treated group whereas one untreated girl died from presumed arrhythmia aged 18 years. Aortic root diameter increased with age as expected, by 5.0 (2.2) and 5.8 (4.6) mm, in treated and untreated groups, respectively. There was no significant difference between the groups.

Conclusions: The oestrogen regimen used, which gives a lower dose than that traditionally used in the treatment of tall stature, has given encouraging results with no adverse events recorded. However, the number of patients treated in a large centre over a 14 years period is very small, indicating that further research must be multicentre in design.


T. Stephenson, J. Bispham, S. Pearce, J. Dandrea, M. E. Symonds. Centre for Reproduction and Early Life, Institute of Clinical Research, University Hospital, Nottingham, NG7 2UH

Aims: Maternal nutritional manipulation through gestation can promote fetal fat growth in conjunction with increased abundance of mRNA for the insulin like growth factor (IGF) I and II receptors (

). Maternal parity has a significant impact on fetal growth but the extent to which it may contribute to increased fat deposition after birth is not known.

Methods: Ten twin bearing adult sheep of similar body weight were entered into the study, of which 5 were primiparous (P) and 5 multiparous (M). All ewes were fed 100% of total metabolisable energy (ME) requirements throughout gestation and all ewes delivered normally at term. At 1 day of age one randomly selected lamb from each ewe was weighed and humanely euthanased. The remaining lamb was reared by its mother until 1 month of age when tissue sampling was performed. Total RNA was extracted from perirenal adipose tissue and RT-PCR analysis performed using oligonucleotide primers specific to ovine leptin, plus IGF I, and II receptors.

Results: In arbitrary units calculated as a ratio of an 18S rRNA internal control, are means (SD) and significant differences were analysed using Kruskal-Wallis and Mann-Whitney U tests.

Results: At 1 day lambs born to P ewes had significantly raised mRNA abundance for both IGF I (P 197 (14); M 123 (10) % ref (p<0.05)) and II receptors (P 121 (10); M 74 (7) % ref (p<0.05)). This difference was not maintained at 1 month of age as mRNA abundance decreased to very low levels in all animals. Leptin mRNA was greater in P offspring throughout (P 88.9 (11.3); M 47.7 (10.1) a.u. (p<0.05)). At 1 month, but not 1 day, of age the offspring of the P mothers possessed more fat (P 153 (10): M 108 (14) g (p<0.01)) although body weights were similar between groups.

Conclusion: Maternal parity is a major factor determining fat deposition. The higher mRNA abundance for IGF receptors together with increased leptin in the newborn of P mothers may explain the enhanced fat deposition over the first month of life.


S. C. Wong, S. M. Ng, M. Didi. Department of Endocrinology, Alder Hey Children’s Hospital, Liverpool

Aim: To study the timing of AR and body mass index (BMI) from infancy to late childhood in CH.

Methods: We performed a retrospective study of BMI and timing of AR in children with CH. Mean length/height and BMI were determined at 1, 4, 6, and 10 years. Timing of AR was determined by visual inspection of BMI v age graph plotted for each patient. The proportion of children with CH who reached AR by 37 months and 49 months were compared with two groups of children described in the literature: 889 healthy British children and 68 Scottish children treated for acute lymphoblastic leukaemia (ALL). Correlation of timing of AR with BMI at 10 years, initial severity of hypothyroidism, and age at normalisation of thyroid stimulating hormone (TSH) were examined.

Result: There were 53 children in total: 34 females and 19 males. Mean BMI (SD) was 0.87 (0.17) while length SDS was −0.02 (0.2) at 1 year of age. BMI had increased to 1.1 (0.23) and height was 0.02 (0.2) at 10 years of age. AR had occurred by 37 months in 37.7% children with CH, in 42.7% children treated for ALL (CH v ALL, p 0.58) and 4.5% healthy British children (CH v normal, p<0.0001). By the age of 49 months, 54.7% children with CH had reached AR compared with 21.4% of normal children (CH v normal, p<0.0001). Pearson’s correlation coefficient of timing of AR showed significant negative correlation with BMI at 10 years. (r −0.487, p 0.01). No correlation was found between timing of AR with initial thyroid function and age at normalisation of TSH.

Conclusion: Children with CH show significantly earlier AR than healthy British children and this showed significant negative correlation with BMI SDS at 10 years. AR does not appear to be directly related to initial severity of hypothyroidism or undertreatment.


S. Ehtisham1, N. Crabtree2, P. Clark2, N. Shaw1, T. Barrett1. 1Birmingham Children’s Hospital; 2University Hospital Birmingham

Introduction: Mortality from cardiovascular disease is four times higher in adults of South Asian (SA) origin than White Europeans (WE). This is related to ethnic differences in risk factors, including insulin sensitivity, which are present in childhood.

Aim: To test the hypothesis that sex and ethnic differences in insulin sensitivity are attributable to differences in total body fat in healthy teenagers.

Method: We recruited 135 teenagers aged 14–17 years (70 girls, 67 WE, 68 SA). Body composition was assessed by anthropometry and dual energy x ray absorptiometry (DXA). Insulin sensitivity was assessed by homeostasis model assessment (HOMA) of fasting glucose and insulin values. Statistical analysis was by univariate ANOVA on log transformed data.

Body mass index standard deviation score (BMI SDS) was not significantly different between the sex–ethnic groups. However, girls had more total body fat than boys (19970g v 14302 g, p<0.001) and more trunkal fat (31.0% v 20.2%, p<0.001). SA had more total body fat than WE (20140g v 15213g, p<0.05) and more trunkal body fat (30.4% v 22.8%, p 0.001). SA were less insulin sensitive than WE (HOMA-S 53% v 60%, p<0.05) and girls were less insulin sensitive than boys but this did not reach statistical significance. The sex–ethnic differences in insulin sensitivity were fully explained by sex–ethnic differences in total body fat measurements such that when total body fat was used as a covariate, the sex–ethnic differences were abolished.

Conclusion: BMI is a poor measure of body fat in teenagers. Sex and ethnic differences in insulin sensitivity in teenagers are accounted for by sex–ethnic differences in total body fat.


T. Y. Segal, R. M. Viner, P. C. Hindmarsh. London Centre for Paediatric Endocrinology & Metabolism, Royal Free & University College Medical School, London

Aims: Non-alcoholic steatohepatitis (NASH) is a recognised complication of obesity that in adults may progress to end stage liver disease. Raised ALT is the most sensitive marker of NASH. We examined the prevalence and predictors of abnormal liver function tests (LFT) in a clinical sample of obese children.

Methods: LFT were retrospectively examined in 124 subjects 3–18 years assessed for simple obesity (BMI centile >95%). Abnormal LFT defined as ALT >60 U/L or total bilirubin >20 μmol/L.

Results: LFT had been assessed in 81 (54%). Fifteen (19%) had raised ALT (seven (9%) were >100U/L). Two (3%) had raised bilirubin.

Conclusions: One fifth of obese children/adolescents have possible NASH. Risk is increased by insulin resistance, hypertriglyceridaemia, and hypertension. Further research is needed to evaluate hepatic fat accumulation and the natural history of abnormal LFT in obese children. Assessment of liver function should be part of routine assessment of obese children.

Abstract G156

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.