Study of Offspring of Women With Type 1 Diabetes

Introduction

Health conditions result from a combination of genetic susceptibility and environmental influences. While the genotype is determined at conception the phenotype is modulated and influenced by environmental and epigenetic factors throughout life, not only postnatally but already in uteri and possibly even at preimplantation stages.

Recent studies have highlighted the possible role of intrauterine exposure to maternal diabetes in the pathogenesis of overweight, type 2 diabetes and cardiovascular disease. A hyperglycaemic intrauterine environment may also affect cognitive function, childhood growth and pubertal development in the offspring. During 1993-99 all pregnancies in women with pre-gestational type 1 diabetes were prospectively reported to a national registry in the Danish Diabetes Association. This nationwide registry contain detailed information of maternal demography, pregnancy outcome and diabetes status including glycated haemoglobin (HbA1c) in a prospective cohort of 900 women and their newborn offspring. These data are therefore ideal for a large-scale study on long-term effects of a diabetic intrauterine environment.

The study consists of 3 substudies.

Aims

Study I

To study long-term effects of a diabetic intrauterine environment in offspring of women with type 1 diabetes compared to matched controls from the background population with respect to:

A)

• Cognitive function
• Pubertal development
• Diabetes/pre-diabetes
• Overweight

1. To study if HbA1c-level in pregnancy is an independent predictor of these outcomes within the study group.

Study II

To study differences in offspring of women with type 1 diabetes compared to offspring from the background population with respect to:

• morbidity and congenital malformations
• medical treatment
• mortality

Study III

To study differences in offspring of women with type 1 diabetes compared to compared to matched controls from the background population without a family history of diabetes with respect to:

1. State-of-the-art metabolic characterization using euglycemic-hyperinsulinemic clamp
2. DNA methylation, RNA transcription and protein quantification in muscle and adipose-tissue

Design

The study is a prospective follow-up study. The study group includes the offspring of women with type 1 diabetes from the national diabetes birth registry (1993-99, n=900) with information of HbA1c prior to conception and/or 1st trimester HbA1c and a control group including offspring of women without diabetes who delivered during the same period matched with respect to gender and age of offspring and the family's postcode as an indirect marker of the socioeconomic background.

Material and Methods

Clinical characteristics at the time of exposure:

Study group (diabetes-exposed):

The following data were reported to the national diabetes birth registry:

Mothers:

• Demographics (age, parity, height, pre-gestational weight)
• Diabetes status (pregestational /1st trimester HbA1c, 3rd trimester HbA1c, pre-gestational /1st trimester Urine Albumin Excretion Rate, hypertension, proliferative retinopathy, diabetes duration, pregestational insulin requirements (IE/d), occurrence of severe hypoglycemic events)
• Pregnancy complications (Preeclampsia, preterm delivery)

Offspring:

• Demographics (gestational age, birth weight)
• Neonatal morbidity (congenital malformations, hypoglycaemia, respiratory distress syndrome, jaundice, systemic infection)

Control group Maternal age, parity, height, pre-gestational weight, information on preeclampsia, gestational age, birth weight and neonatal morbidity will be collected from medical records.

Examination program at follow-up:

Study I: Clinical study (450 diabetes-exposed and 450 controls aged 13-19 years) Clinical examination to determine the effect of intrauterine hyperglycaemia on growth, pubertal development, metabolism and cognitive function in childhood.

Examinations

The participants meet fasting in the morning. The following examinations will be performed:

Diabetes/prediabetes:

• Oral glucose tolerance test (OGTT) with glucose, insulin, C-peptide and proinsulin at 0, 30, 120 minutes
• Blood pressure
• HbA1c, cytokines (CRP, Interleukin-6), adiponectin, leptin, incretins
• GAD-antibodies
• Blood sample for DNA analysis

Obesity:

• Body mass index (BMI), waist/hip circumference
• Lipids (total, HDL-C, LDL-C, triglycerides)
• Dual energy X-ray absorptiometry (DEXA) scan to determine body composition

Growth and pubertal development:

• Height, sitting height, weight, head circumference
• Pubertal stage (Breast development/testicular size and pubic hair)
• Hirsutism in girls?
• Insulin-like Growth Factor-1 (IGF-1), free EGF-1, Insulin-like growth factor-binding protein 3 (IGFBP-3), Insulin-like growth factor-binding protein 1 (IGFBP-1)
• Testosterone, estradiol, Sex hormone-binding globulin (SHBG), Follicle- stimulating hormone (FSH), Luteinizing hormone (LH), inhibin A, inhibin B, Anti-Müllerian hormone (AMH)
• Adrenal androgens: dehydroepiandrosterone sulphate (DHEA-S), and Δ4 androgen androstenedione (ADION)
• Thyroid hormones: Thyroid-stimulating hormone (TSH), Thyroxine (T4), free Thyroxine (fT4)

Cognitive examination including:

• Assessment of global intelligence: Reynolds Intellectual Assessment Scales (RIAS)
• Assessment of specific cognitive functions Attention Learning and Memory Psychomotor speed and reaction time

The children and their parents will be asked to fill in a questionnaire addressing psychosocial aspects, demographics (height, weight) and history of menstrual cycle. Furthermore, the children will be asked to fill in a questionnaire on Self-reported physical activity (IPAQ-questionnaire).

Primary endpoints

• Cognitive function
• Pubertal development
• Diabetes/pre-diabetes
• Overweight (> 85-percentile for age and gender)

Secondary endpoints

• BMI
• Body composition (% body fat)
• Blood pressure
• Dyslipidaemia
• Insulin levels
• Markers of endothelial function
• Markers of autoimmunity (GAD-antibodies)
• Polycystic ovary syndrome (PCOS)

Study II: Register-based study (n=900 diabetes-exposed; for every case we will sex and age match 100 controls, resulting in n=90,000 controls aged 12-18 years) A register-based study will be performed to determine the effect of intrauterine hyperglycaemia on congenital malformations, morbidity and mortality in childhood.

Study population In the cohort of 900 children from the national diabetes birth cohort, we will study all subjects utilizing register data concerning morbidity, mortality and use of prescription medicine. For every diabetes-exposed person Statistics Denmark will identify 100 age, sex and calendar-time matched controls from the background population from the Central Person Register (Statistics Denmark).

Morbidity and mortality - Operational strategy

1. In the National Registry of Patients (NRP) and the National Cancer Registry all diagnoses from identified diabetes-exposed and controls will be identified.
2. In the National Drug Prescription Database all information concerning prescription drugs on identified diabetes-exposed and controls will be identified.
3. In the National Registry of Death all diagnoses from identified diabetes-exposed and controls will be identified.

Endpoints Morbidity: diabetes-exposed vs. controls from the background population Medical treatment: diabetes-exposed vs. controls from the background population Mortality: diabetes-exposed vs. controls from the background population

Statistics Incidence rates will be calculated as new cases per 100.000 per year and analyzed by Poisson regression. Via Statistics Denmark controls will be identified and matched appropriately.

Morbidity and mortality will be analyzed by the Kaplan-Meier statistic, with log-rank test and Cox regression with relevant covariates.

Study III: Sub-study (n=50 diabetes-exposed and n=50 un-exposed controls aged 18-19 years) Offspring of mothers with type 1 diabetes, born before 1 January 1994 (over 18 years of age at the date of invitation), and who participated in the study A will be invited to participate in study B. Study B is anticipated to involve approximately 50 diabetes-exposed and approximately 50 un-exposed controls 18 - 19 years of age at time of examination. The study groups will be matched according to BMI, gender, date of birth and level of physical activity. All participants should be drug-naive and healthy with a BMI between 20 and 30, and controls should have no family history of diabetes. Since all participants will be over the age of 18, they will be able to provide informed written consent.

Exclusion criteria:

1. Any unknown disease or need for medication that occurs after inclusion,
2. Abnormal ECG, screening blood tests and/or severe hypertension, and
3. Impaired glucose tolerance in non-diabetic subjects.

The selected subgroup of offspring of mothers with type 1 diabetes and matched controls will be investigated by state-of-the-art metabolic characterization using euglycemic-hyperinsulinemic clamp with tracer glucose combined with indirect calorimetry allowing reliable estimates of whole-body glucose disposal rates, endogenous glucose production and glucose and lipid oxidation. Moreover assessment of physical activity level using the IPAQ-questionnaire will be complemented by studies of maximal oxygen consumption (VO2max). During the clamp studies, tissue biopsies from subcutaneous abdominal fat and thigh muscle (m. vastus lateralis) will be obtained for studies of potential long-term molecular consequences of intrauterine exposure to hyperglycemia - metabolic memory of birth. This will include application of several discovery-mode (hypothesis free), global approaches such as transcriptional profiling using microarray-based technologies, quantitative proteomics, bioinformatics including pathway analysis and subsequent validation of observed abnormalities using qRT-PCR, immunoblotting and other more classical protein technologies.

Examinations:

Physical activity (Day 1)

The children and their parents will be asked to fill in a questionnaire addressing psychosocial and demographic aspects and self-reported physical activity (IPAQ-questionnaire). VO2-max is determined by a graded maximal test (VO2-peak) on a cycle ergometer using indirect calorimetry.

State-of-the-art metabolic characterization (Day 2):

Euglycemic-hyperinsulinemic clamp Subjects are admitted after a 12-h overnight fast to the centre at Odense University Hospital. They are instructed to consume a standardized diet, and to refrain from physical activity for 48-h before the experiments. The study subjects are examined by a euglycemic-hyperinsulinemic clamp (insulin 40mU/min/m2 for 4-h) using tracer technology as described55. The studies are combined with indirect calorimetry allowing estimates of glucose disposal rates, endogenous glucose production, glucose and lipid oxidation, and non-oxidative glucose metabolism as described. Blood samples are drawn every 20 min during the clamp for assessment of plasma glucose, FFA, adipokines, and serum insulin and C-peptide. During the basal and insulin-stimulated states, tissue biopsies are taken from m. vastus lateralis and subcutaneous abdominal fat using a modified Bergstrom needle with suction under local anesthesia. Each biopsy is rapidly frozen in liquid nitrogen within 30 s and stored at -130°C for later analysis as described below. Body fat (%) is determined by the bioimpedance method.

Transcriptomics and proteomic analysis of skeletal muscle and adipose tissue include:

Microarray-based transcriptional profiling and biological pathway analysis:

Discovery-mode quantitative proteomics of tissue biopsies:

Targeted quantitative proteomics:

Quantitative RT-PCR

Immunoblotting

Genetic analysis:

Analysis of DNA methylation

Whole genome methylation analysis

Methylation-Sensitive High Resolution Melting analysis (MS-HRM)

Gene Expression Analysis

Sequencing

Statistical Analysis Mann-Whitney U test will be used to assess for each CpG, whether two groups had the same distribution of methylation. To eliminate probes that did not work and other diverging values, beta values in each group will be trimmed for extremes prior to the calculation of the average beta value. Next, only genes with the highest 5% tail of the absolute average beta value differences between the two groups will be selected. Thus CpGs with a p-value below 0.005 and with the highest beta value differences between the two groups will be used for identification of genes with statistically significant diabetes-specific changes in methylation. Comparison of array methylation results with MS-HRM results will be performed with a Mann-Whitney U test and a Chi2 test for trend. Spearman coefficients will be calculated to assess the correlation between methylation and gene expression.

Novelty and importance

The rapidly increasing burden of overweight and cardiovascular disease is becoming a threat to both the individual and global economy. It is therefore essential to identify risk groups to target preventive strategies. Exposure to intrauterine hyperglycemia contributes to the epidemic through a vicious-cycle passing increased susceptibility on to the next generation via pathways that are only sparsely understood.

This study includes a very large cohort of diabetes-exposed offspring of well-characterized prospectively studied women with type 1 diabetes. It includes information on maternal glycemia during pregnancy, which is imperative to assess possible associations with estimates of offspring metabolism and cognitive function. The register-based study is unique and enables collection of information on morbidity and mortality in the whole cohort. The specific molecular signature of intrauterine exposure to hyperglycemia in human tissues has to our knowledge not been examined before, and although the above-mentioned studies suggest the possible involvement of certain pathways in skeletal muscle and adipose tissue, it is important to investigate the metabolic memory of birth using hypothesis-free, discovery-mode global approaches to discover potential perturbations in both expected and unexpected pathways.