John C. Peek

Childhood outcomes of assisted reproductive technology

There is a large population of children conceived via assisted reproductive technology (ART), which continues to increase worldwide, without a clear understanding of associated long-term outcomes. ART children are more likely to be the result of multiple pregnancies, and thus to be born prematurely or low birthweight. There is growing evidence that ART children are phenotypically and biochemically different from naturally conceived children, but the mechanism(s) leading to these changes have not been elucidated. There is a possible increased risk of rare imprinted gene disorders in these children. However, it remains unclear whether more subtle changes in DNA methylation occur commonly, leading to differences in gene expression and phenotype in ART children. Although an increased risk of cancer among ART children has been reported, the role of ART in the development of cancer has not been demonstrated. Further research and ongoing surveillance of ART children is essential to better understand the possible effects of ART on the long-term health of this population.

Many women undergoing fertility treatment make poor lifestyle choices that may affect treatment outcome

STUDY QUESTION What are the lifestyle choices and dietary aspects of women about to undergo fertility treatment in New Zealand? SUMMARY ANSWER A considerable proportion of women about to undergo fertility treatment make poor lifestyle choices, including the consumption of alcohol and caffeine. WHAT IS KNOWN ALREADY Women undergoing fertility treatment are highly motivated to achieve pregnancy, but there are relatively few published data on their lifestyle, lifestyle changes or dietary aspects. STUDY DESIGN, SIZE, DURATION This was a cross-sectional study of 250 women aged 20-43 years, taking place between March 2010 and August 2011. PARTICIPANTS/MATERIALS, SETTING, METHODS Women about to undergo IVF or ICSI treatment in two large fertility clinics in Auckland and Hamilton, New Zealand. Lifestyle and dietary intake questionnaires were individually administered once to each participant 35 days (SD = 22) prior to fertility treatment initiation. Outcome measures included incidence of smoking, consumption of alcohol and caffeinated beverages, BMI, detailed intake of dietary supplements and fertility treatment success. Consumption of certain nutrient supplements was compared with the general female New Zealand population. MAIN RESULTS AND THE ROLE OF CHANCE There were high rates of alcohol (50.8%) and caffeine (86.8%) consumption. Most women (82.8%) reported at least one lifestyle change in preparation for fertility treatment, but less than half of women who consumed alcohol regularly reduced their intake and 60% did not change consumption of caffeinated beverages. Similarly, the majority of women did not change their exercise levels (64.4%) or BMI (83.6%) ahead of fertility treatment. Coffee intake appeared unrelated to treatment outcome, but women who consumed caffeinated herbal tea (36.4% of the study population consumed green tea) had lower odds of becoming pregnant (odds ratio, OR 0.52; P = 0.041 versus those not consuming caffeinated herbal tea). Women who abstained from drinking or reduced alcohol intake had twice the odds of becoming pregnant than those who maintained their drinking habits prior to fertility treatment (OR 2.27; P = 0.049). While 93.2% of women took a folic acid supplement, 16.8% had an inadequate intake compared with the current New Zealand prenatal recommendation of 800 mcg/day. Women who held a university degree or higher qualification had twice the odds of becoming pregnant as women with lower levels of education (OR 2.08; P = 0.017), though this finding appeared to be unrelated to lifestyle or dietary habits. LIMITATIONS, REASONS FOR CAUTION The study involved self-reported behaviours that might have been misrepresented by respondents. In addition, our questionnaires covered the period following the first clinical assessment but ∼5 weeks prior to fertility treatment initiation, so that we cannot ascertain whether dietary intakes and lifestyle choices persisted over the course of treatment itself. WIDER IMPLICATIONS OF THE FINDINGS Many women about to undergo fertility treatment make poor lifestyle choices that may negatively affect their chances of becoming pregnant. These findings may be more widely applicable to other women attempting to become pregnant. Specific advice for women regarding healthy lifestyle choices while undergoing fertility treatment is warranted. STUDY FUNDING/COMPETING INTERESTS A.A.G. received financial support from Abbott Nutrition Research & Development Asia-Pacific Center; J.C.P. is a shareholder of Fertility Associates; the other authors have no financial or non-financial conflicts of interest to disclose.

Maternal age and ovarian stimulation independently affect oocyte mtDNA copy number and cumulus cell gene expression in bovine clones

STUDY QUESTION Does maternal ageing and ovarian stimulation alter mitochondrial DNA (mtDNA) copy number and gene expression of oocytes and cumulus cells from a novel bovine model for human IVF? SUMMARY ANSWER Oocytes collected from females with identical nuclear genetics show decreased mtDNA copy number and increased expression of an endoplasmic reticulum (ER) stress gene with repect to ovarian stimulation, whilst differences in the expression of genes involved in mitochondrial function, antioxidant protection and apoptosis were evident in relation to maternal ageing and the degree of ovarian stimulation in cumulus cells. WHAT IS KNOWN ALREADY Oocyte quality declines with advancing maternal age; however, the underlying mechanism, as well as the effects of ovarian stimulation are poorly understood. Human studies investigating these effects are often limited by differences in age and ovarian stimulation regimens within a patient cohort, as well as genetic and environmental variability. STUDY DESIGN, SIZE, DURATION A novel bovine cross-sectional maternal age model for human IVF was undertaken. Follicles were aspirated from young (3 years of age; n = 7 females) and old (10 years of age; n = 5 females) Holstein Freisian clones following multiple unstimulated, mild and standard ovarian stimulation cycles. These bovine cloned females were generated by the process of somatic cell nuclear transfer (SCNT) from the same founder and represent a homogeneous population with reduced genetic and environmental variability. Maternal age and ovarian stimulation effects were investigated in relation to mtDNA copy number, and the expression of 19 genes involved in mitochondrial function, antioxidant protection, oocyte-cumulus cell signalling and follicle development in both oocytes and cumulus cells. MATERIALS, SETTING, METHODS Young (3 years of age; n = 7 females) and old (10 years of age; n = 5 females) Holstein Freisian bovine clones were maintained as one herd. Stimulation cycles were based on the long GnRH agonist down-regulation regimen used in human fertility clinics. Follicle growth rates, numbers and diameters were monitored by ultrasonography and aspirated when the lead follicles were >14 mm in diameter. Follicle characteristics were analysed using a mixed model procedure. Quantitative PCR (qPCR) was used to determine mtDNA copy number and reverse transcriptase-qPCR (RT-qPCR) was used to measure gene expression in oocytes and cumulus cells. MAIN RESULTS AND THE ROLE OF CHANCE Method of ovarian stimulation (P = 0.04), but not maternal age (P > 0.1), was associated with a lower mtDNA copy number in oocytes. Neither factor affected mtDNA copy number in cumulus cells. In oocytes, maternal age had no effect on gene expression; however, ovarian stimulation in older females increased the expression of GRP78 (P = 0.02), a gene involved in ER stress. In cumulus cells, increasing maternal age was associated with the higher expression of genes involved in mitochondrial maintenance (TXN2 P = 0.008 and TFAM P = 0.03), whereas ovarian stimulation decreased the expression of genes involved in mitochondrial oxidative stress and apoptosis (TXN2 P = 0.002, PRDX3 P = 0.03 and BAX P = 0.03). LIMITATIONS, REASON FOR CAUTION The low number of oocyte and cumulus cell samples collected from the unstimulated cycles limited the analysis. Fertilization and developmental potential of the oocytes was not assessed because these were used for mtDNA and gene expression quantification. WIDER IMPLICATIONS OF THE FINDINGS Delineation of the independent effects of maternal age and ovarian stimulation regimen on mtDNA copy number gene expression in oocytes and cumulus cells was enabled by the removal of genetic and environmental variability in this bovine model for human IVF. Therefore, these extend upon previous knowledge and findings provide relevant insights that are applicable for improving human ovarian stimulation regimens. STUDY FUNDING/COMPETING INTERESTS Funding was provided by Fertility Associates and the University of Auckland. J.C.P. is a shareholder of Fertility Associates and M.P.G. received a fellowship from Fertility Associates. The other authors of this manuscript declare no conflict of interest that could be perceived as prejudicing the impartiality of the reported research.

Development of clinical priority access criteria for assisted reproduction and its evaluation on 1386 infertile couples in New Zealand

BACKGROUND In New Zealand ranking patients for elective, publicly funded procedures uses clinical priority access criteria (CPAC). A CPAC to prioritize patients seeking assisted reproductive technology (ART) was developed in 1997 and implemented nationwide in 2000. This study describes the development of the ART CPAC tool and its evaluation on 1386 couples referred to a single tertiary service from 1998 to 2005. METHODS A total of 48 health professionals and consumers assisted in criteria development. A score between 0 and 100 points was calculated for each couple and those who reached ≥65 points were eligible for publicly funded ART. Couples beneath the treatment threshold were placed on active review; the review being the date the score was calculated to reach the treatment threshold. Couples who would never be eligible or who were on active review were offered private treatment. Treatments and outcomes (spontaneous and treatment dependent live birth pregnancies) were used to evaluate the criteria. RESULTS Three social criteria (duration infertility, number of children and sterilization status) and two objective criteria (diagnosis and female age) formed the priority score. Of the evaluated couples, 643 (46%) were eligible within 1 year of referral (Group 1), 451 (33%) >1-5 years from referral (Group 2) and 292 (21%) couples were never eligible (Group 3). The predominant ART was IVF. A total of 480 couples had at least one IVF treatment with 404 (84%) having publicly funded treatment. A total of 762 (55%) women gave birth, 473 from treatment and 289 spontaneously. Group 1 had more pregnancies from treatment while Group 2 had most pregnancies overall being mainly from spontaneous pregnancies. Compared with Group 3 cases the hazard ratio using time to spontaneous live birth pregnancy for Group 1 couples was significantly lower, 0.51 (95% confidence interval 0.36-0.74) and for Group 2 cases significantly higher, 1.86, (1.35-2.58). Treatments using ART were evaluated for the three eligibility groups, with the never eligible divided into women age <40 (Group 3a) and woman age ≥40 at referral (Group 3b). Compared with Group 1 cases the hazard ratio to treatment dependent live birth pregnancy was similar for Groups 2 and 3a but significantly lower for Group 3b (0.37, 0.14-0.90). CONCLUSIONS The clinical priority access score was able to discriminate between the chance of pregnancy with and without treatment and those offered and not offered treatment. The CPAC is a useful model for informing the allocation of public funding for ART in other countries.

Assessing embryo quality by combining non-invasive markers: early time-lapse parameters reflect gene expression in associated cumulus cells

STUDY QUESTION Are there associations between early time-lapse parameters, expression of candidate embryo viability genes in cumulus cells and embryo quality on Day 5? SUMMARY ANSWER Early time-lapse parameters correlate to the expression levels of candidate embryo viability genes in cumulus cells but a combined analysis including both time-lapse and candidate gene expression did not identify significant predictors of embryo quality on Day 5. WHAT IS KNOWN ALREADY Recent evidence suggests that early time-lapse parameters are predictive of blastocyst development. Similarly, a number of candidate genes in cumulus cells have been identified as potential markers of embryo viability. Relationships between time-lapse parameters and candidate gene expression in cumulus cells have not been investigated, and a combined analysis of these markers has not been attempted in relation to embryo quality. STUDY DESIGN, SIZE, DURATION A total of 78 embryos obtained by ICSI from 22 patients were studied by time-lapse and measurement of cumulus cell gene expression of known markers of embryo viability. Time-lapse and cumulus cell gene expression data were assessed in relation to embryo quality on Day 5. PARTICIPANTS/MATERIALS, SETTING, METHODS All women, aged 32-40 years, underwent ICSI treatment for male infertility. Embryos with annotatable time to pronuclear breakdown (tPNB), division to two cells (t2C), three cells (t3C), four cells (t4C) and five cells (t5C) were included in the study. Expression levels of 27 candidate genes for embryo viability were measured in 78 associated cumulus cell masses using quantitative real-time PCR. MAIN RESULTS AND THE ROLE OF CHANCE Cumulus cell expression of 11 candidate genes involved in energy metabolism (ATPase, H+ transporting, lysosomal 70 kDa, V1 subunit A (ATP6V1A), NADH dehydrogenase (ubiquinone) 1 alpha subcomplex, 1, 7.5 kDa (NDUFA1), lactate dehydrogenase A (LDHA), phosphofructokinase platelet (PFKP) and solute carrier family 2 member 4 (SLC2A4), mitochondrial biogenesis (DNA directed RNA polymerase, mitochondrial (POLRMT) and transcription factor A, mitochondrial (TFAM), signalling (prostaglandin-endoperoxide synthase 2), steroidogenesis (cytochrome P450, family 11, subfamily A, polypeptide 1 (CYP11A1) and cell stress (heat shock 70 kDa protein 5 (HSPA5) and peroxiredoxin 3 (PRDX3)) correlated to time-lapse parameters of the developing embryo, largely for t3C onwards (all P < 0.05). Expression of ATP synthase, H+ transporting, mitochondrial Fo complex, subunit E (ATP51), HSPA5, PFKP, PRDX3 and versican (VCAN) and the parameter t4C were also related to embryo quality on Day 5 (all P < 0.05). Ordinal logistic regression, where gene expression and time-lapse parameters were combined, did not identify any significant predictors of embryo quality on Day 5. LIMITATIONS AND REASON FOR CAUTION Data are from a preliminary study, limited by a small sample size and using more than one ovarian stimulation protocol. A possible limitation is that each follicle was treated as an independent observation, although a considerable fraction of embryos were from the same patient. WIDER IMPLICATIONS OF THE FINDINGS Results presented in this study suggest that some of the variation of time-lapse parameters may be related to cumulus cell gene expression and thus the ovarian microenvironment in which the oocyte developed. Although the current study did not identify significant predictors of embryo quality on Day 5, investigation in a larger cohort may determine whether cumulus cell gene expression and time-lapse parameters can be combined to predict embryo quality. STUDY FUNDING/COMPETING INTERESTS Funding was provided by Fertility Associates Ltd, the Auckland Medical Research Foundation and the University of Auckland. J.C.P. has a 0.5% shareholding in Fertility Associates. All other authors of this manuscript have nothing to declare and no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Ovarian stimulation leads to shorter stature in childhood

BACKGROUND We aimed to determine whether children conceived with ovarian stimulation alone (OS(A)) would differ phenotypically and biochemically from naturally conceived children of fertile and subfertile parents. METHODS Healthy pre-pubertal children aged 3-10 years, born at term, after singleton pregnancies were recruited in Auckland (New Zealand) and were allocated into three groups: (i) children conceived following OS(A) and naturally conceived children of (ii) subfertile and (iii) fertile parents. Anthropometric, endocrine and metabolic parameters were recorded. Childrens heights and body mass index (BMI) were expressed as standard deviation scores (SDS) and corrected for genetic potential (i.e. parental height or BMI). RESULTS Three hundred fifty-two children were studied: 84 OS(A) subjects and 268 naturally conceived controls consisting of 54 children of subfertile parents and 214 children of fertile parents. Children of subfertile and fertile parents did not differ in measured outcomes. Overall, OS(A) children were shorter than children of both subfertile (SDS: -0.08 ± 0.09 versus 0.32 ± 0.07; P= 0.001) and fertile (SDS: -0.08 ± 0.09 versus 0.45 ± 0.10; P= 0.004) parents when corrected for genetic height potential. OS(A) boys were shorter than boys of subfertile (SDS:-0.18 ± 0.14 versus 0.42 ± 0.16; P= 0.03) and fertile (SDS: -0.18 ± 0.14 versus 0.35 ± 0.08; P= 0.01) parents. There was also a trend towards OS(A) girls being shorter than girls of subfertile parents (P= 0.06), but not significantly shorter than those of fertile parents (P= 0.17). OS(A) children also had a lower corrected BMISDS than children of subfertile (SDS-0.90 ± 0.15 versus -0.37 ± 0.17; P= 0.06) and fertile (-0.90 ± 0.15 versus -0.34 ± 0.10; P= 0.008) parents. Among metabolic parameters, fasting glucose was lower in OS(A) children than that in children of fertile parents (4.62 ± 0.07 versus 4.81 ± 0.04; P= 0.006). CONCLUSIONS Conception after OS(A) was associated with shorter stature, particularly in boys, compared with naturally conceived children of fertile and subfertile parents.

Estimation of fertility and fecundity in women receiving artificial insemination by donor semen and in normal fertile women

Summary. The pregnancy rates after artificial insemination by donor semen (AID) have been compared with pregnancy rates in normal fertile women to assess the efficiency of AID. To do this, the curve y=a(1−(1−b)x) was fitted to life‐tabled cumulative pregnancy rates. The equation describes a model in which the parameter a is the proportion of women who are potentially fertile under the conditions of treatment, and in which the parameter b is the pregnancy rate per cycle (or fecundity) of these fertile women. For 259 AID patients with no previous pregnancy a was 65% while for 57 AID patients with a previous pregnancy after AID ‘a’ was 99.9%. The values of b were similar for the two groups of patients, being 20% and 22% respectively. Women without fertility problems who had become pregnant after discontinuing oral contraception provided the reference group. Since only pregnant women were selected, a was 100% by definition. The values of b for the reference group were 22% for 100 primigravid women and 20% for 100 multigravid women. Only 65% of the AID patients were potentially fertile with AID, but those that were fertile became pregnant at the same rate as normal women who discontinued oral contraception.

Screening of karyotype and semen quality in an artificial insemination program: acceptance and rejection criteria.

Semen quality and karyotype were screened in all men offering to be donors for an artificial insemination (AID) program. The criteria for accepting or rejecting semen have now been set with respect to this sample of the population. There was no evidence of differences between the pregnancy rates of the accepted donors. One of 172 potential donors with a clear medical history had a chromosomal abnormality, 4 had pericentric inversions of chromosome 9, and 14 had other heterochromatic variants. Of the recipients of AID, 5 of 196 women had chromosomal abnormalities, and 12 had heterochromatic variants.

Oocyte mitochondrial deletions and heteroplasmy in a bovine model of ageing and ovarian stimulation

STUDY HYPOTHESIS Maternal ageing and ovarian stimulation result in the accumulation of mitochondrial DNA (mtDNA) deletions and heteroplasmy in individual oocytes from a novel bovine model for human assisted reproductive technology (ART). STUDY FINDING The levels of mtDNA deletions detected in oocytes increased with ovarian ageing. Low levels of mtDNA heteroplasmy were apparent across oocytes and no relationship was identified with respect to ovarian ageing or ovarian stimulation. WHAT IS KNOWN ALREADY Oocyte quality decreases with ovarian ageing and it is postulated that the mtDNA may have a role in this decline. The impact of ovarian stimulation on oocyte quality is poorly understood. Human studies investigating these effects are often limited by the use of low quality oocytes and embryos, variation in age and ovarian stimulation regimens within the patients studied, as well as genetic and environmental variability. Further, no study has investigated mtDNA heteroplasmy in individual oocytes using next-generation sequencing (NGS), and little is known about whether the oocyte accumulates heteroplasmic mtDNA mutations following ageing or ovarian stimulation. STUDY DESIGN, SAMPLES/MATERIALS, METHODS A novel bovine model for the effect of stimulation and age in human ART was undertaken using cows generated by somatic cell nuclear transfer (SCNT) from one founder, to produce a homogeneous population with reduced genetic and environmental variability. Oocytes and somatic tissues were collected from young (3 years of age; n = 4 females) and old (10 years of age; n = 5 females) cow clones following multiple natural ovarian cycles, as well as oocytes following multiple mild (FSH only) and standard (based on human a long GnRH agonist protocol) ovarian stimulation cycles. In addition, oocytes were recovered in a natural cycle from naturally conceived cows aged 4-13.5 years (n = 10) to provide a heterogeneous cohort for mtDNA deletion studies. The presence or absence of mtDNA deletions were investigated using long-range PCR in individual oocytes (n = 62). To determine the detection threshold for mtDNA heteroplasmy levels in individual oocytes, a novel NGS methodology was validated; artificial mixtures of the Bos taurus and Bos indicus mitochondrial genome were generated at 1, 2, 5, 15 and 50% ratios to experimentally mimic different levels of heteroplasmy. This NGS methodology was then employed to determine mtDNA heteroplasmy levels in single oocytes (n = 24). Oocyte mtDNA deletion and heteroplasmy data were analysed by binary logistic regression with respect to the effects of ovarian ageing and ovarian stimulation regimens. MAIN RESULTS AND THE ROLE OF CHANCE Ovarian ageing, but not ovarian stimulation, increased the number of oocytes exhibiting mtDNA deletions (P = 0.04). A minimum mtDNA heteroplasmy level of 2% was validated as a sensitive (97-100%) threshold for variant detection in individual oocytes using NGS. Few mtDNA heteroplasmies were detected across the individual oocytes, with only 15 oocyte-specific variants confined to two of the 24 oocytes studied. There was no relationship (P > 0.05) evident between ovarian ageing or ovarian stimulation and the presence of mtDNA heteroplasmies. LIMITATIONS, REASON FOR CAUTION The low number of oocytes collected from the natural ovarian cycles limited the analysis. Fertilization and developmental potential of the oocytes was not assessed as the oocytes were destroyed for mtDNA deletion and heteroplasmy analysis. WIDER IMPLICATIONS OF THE FINDINGS If the findings of this model apply to the human, this study suggests that the incidence of mtDNA deletions increases with age, but not with degree of ovarian stimulation, while the frequency of mtDNA heteroplasmies may be low regardless of ovarian ageing or level of ovarian stimulation. STUDY FUNDING AND COMPETING INTERESTS Funding was provided by Fertility Associates, the Nurture Foundation for Reproductive Research, the Fertility Society of Australia, and the Auckland Medical Research Foundation. J.C.P. is a shareholder of Fertility Associates and M.P.G. received a fellowship from Fertility Associates. The other authors of this manuscript declare no conflict of interest that could be perceived as prejudicing the impartiality of the reported research.

The phenotype of an IVF child is associated with peri-conception measures of follicular characteristics and embryo quality

STUDY QUESTION Are childhood measures of phenotype associated with peri-conception parental, IVF treatment and/or embryonic characteristics of IVF children? SUMMARY ANSWER Birthweight, childhood body mass index (BMI) and height of pre-pubertal IVF children were strongly associated with peri-conception factors, including follicular and embryonic characteristics. WHAT IS KNOWN ALREADY A growing number of studies have identified a range of phenotypic differences between IVF and naturally conceived pre-pubertal children; for example, birthweights are lower following a fresh compared with a thawed embryo transfer. STUDY DESIGN, SIZE, DURATION This retrospective cohort study included IVF children (n = 96) born at term (>37 weeks) after a singleton pregnancy from the transfer of either fresh or thawed embryos in New Zealand. Between March 2004 and November 2008, these children were subjected to clinical assessment before puberty. PARTICIPANTS/MATERIALS, SETTING, METHODS Clinical assessment provided anthropometric measures of children aged 3.5-11 years old. Peri-conception factors (n = 36) derived retrospectively from parental, treatment, laboratory and embryonic variables (n = 69) were analysed using multiple stepwise regression with respect to standard deviation scores (SDSs) of the birthweight, mid-parental corrected BMI and height of the IVF children. Data from children conceived from fresh (n = 60) or thawed (n = 36) embryos, that met inclusion criteria and had high-quality data with >90% completeness, were analysed. MAIN RESULTS AND THE ROLE OF CHANCE Embryo treatment at transfer was identified as a predictor of birthweight with thawed embryos resulting in heavier birthweights than fresh embryos [P = 0.02, 95% confidence interval (CI) fresh minus thawed: -1.047 to -0.006]. Birthweight SDS was positively associated with mid-parental corrected BMI SDS (P = 0.003, slope 0.339 ± 0.100). Four factors were related (P < 0.05) to mid-parental corrected height SDS. In particular, child height was inversely associated with the diameter of lead follicles at oocyte retrieval (P < 0.0001, slope -0.144 ± 0.040) and with the quality score of embryos at transfer (P = 0.0008, slope -0.425 ± 0.157), and directly associated with the number of follicles retrieved (P = 0.05, slope 1.011 ± 0.497). Child height was also positively associated with the transfer of a fresh as opposed to thawed embryo (P < 0.001, 95% CI 0.275-0.750). LIMITATIONS, REASONS FOR CAUTION More than one embryo was transferred in most cycles so mean development and quality data were used. The large number of variables measured was on a relatively small sample size. Large cohorts from multiple clinics using a variety of treatment protocols and embryology methods are needed to confirm the associations identified and ultimately to test these factors as possible predictors of phenotype. WIDER IMPLICATIONS OF THE FINDINGS This is the first study to directly associate peri-conception measures of IVF treatment with a pre-pubertal childs phenotype. Demonstration that peri-conception measures relate to a pre-pubertal childs phenotype extends the range of factors that may influence growth and development. These findings, if corroborated by larger studies, would provide invaluable information for practitioners, who may want to consider the impact of ovarian stimulation protocols as well as the quality of the embryo transferred on a childs growth and development, in addition to their impact on pregnancy rate. STUDY FUNDING/COMPETING INTERESTS This work was supported by grants from the National Research Centre of Growth and Development New Zealand (grant 3682065) and the Australasian Paediatric Endocrine Group (APEG; grant 3621994), as well as a fellowship from Fertility Associates New Zealand awarded to M.P.G. In terms of competing interest, J.C.P is a shareholder of Fertility Associates. M.P.G. currently holds the position of Merck Serono Lecturer in Reproductive Biology. W.S.C. and P.L.H. have also received grants and non-financial support from Novo Nordisk, as well as personal fees from Pfizer that are unrelated to the current study. The other authors have no conflict of interest to declare.