Amid the global obesity and type 2 diabetes epidemic, insulin resistance among women of child-bearing age continues to increase. Pregnancy is a natural state of insulin resistance and increased insulin secretion. When the physiological changes of pregnancy are superimposed on pre-existing insulin resistance, metabolic changes associated with pregnancy can become exaggerated. This can lead to an environment characterised by nutrient excess and inflammation, which is associated with adverse pregnancy conditions such as gestational diabetes mellitus (GDM) and foetal overgrowth. In the field of diabetes research, i.e. as a context separate from pregnancy, there has been growing interest in chronically elevated insulin, termed hyperinsulinaemia, and the role it plays in the aetiology of various cardiometabolic conditions. In this context, according to The Kraft Model of Hyperinsulinaemia, an elevated fasting and/or post-glucose load (100-gram glucose) plasma insulin response may be one of the earliest indicators for the development of type 2 diabetes later in life. While these concepts cannot be directly applied to pregnancy, it is likely that many women of child-bearing age will enter pregnancy with pre-existing insulin resistance and hyperinsulinaemia. As such, this may lead to further exaggerated changes in maternal metabolism (i.e. in the form of glucose, lipids, amino acids, and inflammatory markers). Moreover, since women who experience adverse pregnancy outcomes such as GDM appear to exist on a more adverse track toward later life chronic metabolic diseases, it is hypothesised that identifying hyperinsulinaemia (according to The Kraft Model) during pregnancy may aid in identifying these at-risk women sooner in pregnancy. Such early identification could provide an opportunity to enhance lifestyle management to mitigate the exaggeration of metabolic changes associated with pregnancy.

The overarching aim of this body of work was to investigate the relationship between insulin response patterns during pregnancy and adverse pregnancy outcomes. A further aim was to explore current dietary advice for the management of GDM. To investigate this aim, the first study was a systematic review (15 articles from 11 randomised controlled trials (RCTs), n = 3,614 participants) to extend knowledge on dietary interventions during pregnancy that may be effective for reducing the risk of adverse perinatal outcomes linked to maternal metabolic dysfunction (Chapter 3). The primary outcomes were neonatal cord blood insulin, c-peptide, and adiposity as markers of foetal hyperinsulinaemia and overnutrition. Findings demonstrated an overall protective effect of lifestyle-based interventions (including nutrition, physical activity, behaviour change, and monitoring) with elements of personalisation and dietary advice based on low glycaemic load (GL) approaches for reducing neonatal adiposity.

Maternal insulin response patterns and the prevalence of hyperinsulinaemia (applying the Kraft algorithm) according to varying degrees of glucose tolerance were then investigated in an observational study (Chapter 4). For this study, a historical dataset, dated mid-1970s to early-1990s, of results from presumed pregnant women (n = 926) who underwent three-hour oral glucose tolerance tests (OGTTs, 100-gram glucose load) with insulin assays was examined. Demographic information available included only participant age, weight, height, and date of testing. Further medical information and gestational age at time of testing was not available. Almost a third (31%) of the women with normal glucose tolerance demonstrated a delayed peak hyperinsulinaemia pattern (Kraft pattern III). Among the women with GDM (2010 American Diabetes Association criteria), the prevalence of Kraft pattern III hyperinsulinaemia was 75%. Kraft pattern IV hyperinsulinaemia was only observed in 3% of participants, the majority of which were categorised as GDM (GDM vs. normal: n = 13, 6% vs. n = 9, 1.5%; p<0.001).

To further explore the relationship between early pregnancy insulin response patterns and changes later in gestation, a case series of four pregnant women living in New Zealand was conducted (Chapter 5). Participants underwent an OGTT (75-gram glucose, 2-hour) at ≤ 14 weeks’ gestation. Insulin response patterns were defined using a modified version of the Kraft algorithm to include four plasma insulin and glucose samples from the OGTT (fasting and 30-, 60-, and 120-minutes post oral glucose). Glycaemic measures, weight gain, foetal growth, dietary and lifestyle changes, and obstetric outcomes were examined throughout pregnancy. Two cases presented with Kraft pattern IIB hyperinsulinaemia early in pregnancy. For both of these women, difficulty during labour was observed (shoulder dystocia and surgically assisted delivery for Case #1, and emergency caesarean section for Case #4). A protective effect of healthy lifestyle choices to limit gestational weight gain (GWG) was postulated.

The final component of this research explored the current lifestyle management of GDM pregnancies by maternity-based health professionals in New Zealand in a qualitative study (Chapter 6). Key barriers to providing dietary advice included limited accessibility of health services, service delivery issues, fragmentation and conflicting information across health providers and alternative information sources (e.g. internet-based), and gaps in nutrition guidelines and clinical resources.

Collectively, this body of work demonstrated the potential significance of using insulin response patterns to recognise hyperinsulinaemia early in the gestational period. Several gaps in care were highlighted around the dietary management of women during early gestation at risk of GDM and those later in pregnancy once diagnosed. Recognising a spectrum of insulin response patterns may help to identify metabolic dysfunction much sooner in pregnancy than current screening programmes. Earlier identification of women with such an increased risk could further enhance dietary and lifestyle management to be more proactive and personalised. Further research into this area could help inform the development of early screening tools for women with a higher risk of adverse outcomes and guide treatment decisions. In practice, a better understanding of metabolic phenotypes may aid in progressing therapeutic dietary advice to become more targeted and personalised for better maternal and offspring outcomes.