Health in Childhood Following Assisted Reproductive Technology

The investigators aim to conduct a clinical follow-up study of three groups of children: Exposed children conceived by ART(assisted reproductive technology) after frozen embryo transfer (FET) and two unexposed control groups; (1) Children conceived by ART after fresh embryo transfer and (2) Naturally conceived children. All children will be identified via their mother's CPR number in the in-vitro fertilization (IVF) and medical birth registry. Only singletons 6-8 years of age (birth year 2009-2011) born in hospitals in the Capital Region will be included. Access to maternal postal addresses will be retrieved through the CPR registry. The investigators will contact the mothers by an invitation letter by post including the patient information and the information for the child. The invitation letter and patient information will be similar for study and control groups. The mothers can reply via sms, email or telephone. If they do not reply within 2 weeks, the investigators will send them one postal reminder. In case they respond, the investigators will schedule the visit. The investigators will start contacting and examining the oldest children. Children of mothers with preexisting diabetes (type 1 or type 2) or gestational diabetes will be excluded the study.

Subjects:

• Singletons conceived by ART after FET (N=200)
• Singletons conceived by ART after fresh embryo transfer (N=200). (N~1000 mothers will be contacted, 40% of eligible ART subjects are expected to participate).
• Singletons born after natural conception (N=200). A random sample of naturally conceived children matched with FET children on month and year of birth and child sex (N~ 670, 30% of eligible subjects are expected to participate).

Clinical examinations:

The clinical examinations of the children will be carried out in the Clinic of Growth and Reproduction and Clinic of Cardiology at Rigshospitalet. The investigators aim that each child attends all examinations at the same day and the full examination will last no longer than 30-40 minutes. The participants meet fasting in the morning and a fasting blood sample will be collected. Local anesthetic cream will be sent to all participants prior to the examination. The fasting blood sample will be analyzed for: Glucose profile: glucose, insulin, c-peptid (calculation of HOMA-IR), HbA1c - Lipid profile: Total cholesterol, LDL, HDL, Triglycerides - Growth factors: IGF-I and IGFBP-3 - Androgen status and Anti Müllerian Hormone (AMH) - Lipids and cytokines (CRP, IL-6, adiponectin, leptin, inkretin) - Biobank sampling The clinical examination includes determination of: Height, weight, waist and hip circumference - Skinfold measurements (Harpenden Skinfold Caliper) - Pubertal staging (Tanner) - Blood pressure (systolic, diastolic) measured by FINA press device (valid continuous measurements)

In a subset of children n=50 in each group (in total 150 children) the investigators will perform a cardiovascular magnetic resonance imaging (CMR):

• Magnetic resonance imaging (MRI) determined diastolic and systolic function of cardiac chambers and large vessels of the thorax, and measurement of epi- and pericardial fat tissue

Scientific objective I:

Determination of risk of obesity and metabolic disease in children conceived by ART with FET compared to children conceived after fresh embryo transfer and after natural conception. The primary outcome is overweight and obesity as determined by BMI (SDS for age). Secondary outcome measures are fat percentage measured by skinfolds and abdominal circumference, visceral fat (MRI) and calculation of insulin sensitivity (HOMA-IR) as a risk marker of altered metabolism and epigenetic changes of metabolic pathways, growth factors and andrenal androgens.

Scientific objective II:

Determination of markers of cardiovascular risk in children whose mothers conceived by FET compared to children conceived after ART fresh embryo transfer and after natural conception. Cardio-vascular outcome measures will be left ventricular (LV) and right ventricular (RV) myocardial mass and function, pulmonary artery and aortic distensibility and estimation of epi- and pericardial fat tissue.

Statistics:

Continuous outcome measures will be compared between the three groups using multiple linear regression with adjustment for potential confounding factors and child sex. In case of pronounced skewness, logarithmic transformation will be applied prior to analysis. Binary outcomes will be compared between the three groups using multiple logistic regression with adjustment for potential confounders and child sex. Additional multiple regression analysis with further inclusion of potential mediators will be performed to identify maternal/fetal and neonatal predictive factors for the different outcomes. Background characteristics will be compared between the three groups using one-way ANOVA for normally distributed data, the Kruskal-Wallis test for non-normally distributed continuous data, and the chi-square test for categorical data. P-values from secondary and explorative data analyses will be adjusted for multiple testing by controlling the false discovery rate to 5% as recommended for health science studies.

Sample size calculation:

Scientific objective I:

BMI was chosen as our primary endpoint and the investigators calculated the age specific BMI based on the Tinggard cohort.30 Childhood and adolescent BMI show dose-dependent association with adult mortality, with an increase in 1 SD in BMI z-score at age 8 years showing a HR of 1.13 of a fatal cardiovascular event in adulthood. The minimal clinically important difference in BMI was set as 0.3 SD scores, corresponding to an increased HR of 3.4%. Thus, with equal size groups, a significance level of 5% and power 90% there should be 191 children in each group to detect the minimal relevant difference. With 200 children in each group the investigators anticipate that the sample size will be sufficiently large to compensate for degrees of freedom lost due to adjustment for confounding factors (as in particular the inclusion of the confounders will imply a reduction of the residual variance).

Scientific objective II:

Power calculation for the right heart study:

The previously found differences in hypoxic vasoconstriction were not evidenced by an increased systolic pressure of the right ventricle at rest perhaps because the right ventricle is thin-walled and accommodating. Considering that pre-pubertal children may vary slightly in absolute heart size from adults, the investigators use the ratio of RVEDV/LVEDV for the power calculation. Thus, while normally close to 1.0, with poor run-off for the RV into the pulmonary circulation the investigators expect an increase in the RVEDV/LVEDV ratio >1.2 corresponding to a >20% increase in RVEDV in comparison with the LVEDV (that now acts like an "internal comparator"). With a conservative estimate for the common SD for the difference of 0.35, with an alpha of 0.05 and power (1-beta) of 0.80 the investigators will be able to detect a difference between groups of RVEDV/LVEDV from 1.0 to 1.20 with 36 children in each group (double-sided). The investigators aim for 50 children in each group to allow for drop-outs including non-analysable scans.

Power calculation for the left heart study:

With impaired flow-mediated vasodilatation in IVF children, the investigators expect arterial distensibility to decrease. With age (and disease) arterial distensibility decreases. To be clinically meaningful the investigators have based our power calculations on children aged 10 years born after IVF having stiff aortas comparable to "being 5-years older". With a conservative estimate for the common SD for the difference of 1.5 mm2/mmHg, with an alpha of 0.05 and a beta of 0.80 the investigators will be able to detect a difference between groups of aortic distensibility from 8.9 to 8.0 mm2/mmHg (based on data from 10 and 15-year old boys, respectively with 44 children in each group (double-sided). The investigators aim for 50 children in each group to allow for drop-outs including non-analysable scans.

Power calculation for the epicardial and pericardial fat tissue study:

Recent MRI and echocardiographic studies have demonstrated that children with obesity and metabolic syndrome may have more than two-fold as much epicardial and pericardial fat tissue than lean children, and that this is related to cardiac remodelling including left ventricular hypertrophy and impaired left ventricle longitudinal strain. In this study, anticipating smaller changes, we expect a >20% higher epicardial as well as pericardial fat tissue to be present in children born after IVF. With a conservative estimate for the common SD for the difference of 30% of the mean, with an alpha of 0.05 and a beta of 0.80 the investigators will be able to detect a difference between groups of aortic distensibility of 20% with 36 children in each group (double-sided). The investigators aim for 50 children in each group to allow for drop-outs including non-analysable scans.

Approvals:

The project has been approved by the Scientific Ethics Committee in the Capital Region, Denmark and by the Danish Data Protection Agency (Datatilsynet). Data from patient files will not be used in the study. The study will be conducted in accordance with the Helsinki Declaration and according to the Danish law on person identifiable data.