Besides maternal hyperglycemia, a strong link between maternal pregravid weight/maternal triglyceride(Tg) levels and fetal growth/childhood obesity/metabolic syndrome in offspring has been demonstrated in largely Caucasian populations. This study aims to explore the link between maternal hypertriglyceridemia and offspring health in Asians.

The prevalence of gestational diabetes mellitus (GDM) worldwide is increasing, with accelerated rates of obesity in women of child-bearing age, increasing maternal age and improved survival of very low and high birth weight (BW) female offspring in modern times. GDM prevalence rates are known to be higher in some Asian populations compared with Western populations. The incidence rate amongst South Indians in India, and Malaysians in an urban tertiary centre in Kuala Lumpur are 14 and 11.4% respectively while in Western populations GDM only accounts for 2-9% of all pregnancies.

It is now well-established that exposure to an intrauterine diabetic milieu, in both pregnant women with type 2 diabetes (T2DM) and those with GDM, imparts increased future risk of diabetes, obesity, metabolic syndrome and cardiovascular disease in offspring. Besides maternal hyperglycemia, a strong link between maternal pregravid weight/maternal triglyceride (Tg) levels and fetal growth/childhood obesity/metabolic syndrome in offspring has been demonstrated in largely Caucasian populations.

The mechanisms of the developmental origins of adult disease however are poorly understood. The in utero environment via changes to the epigenome that do not involve alterations in DNA sequence can exert stimulatory or inhibitory effects on fetal growth and adiposity6. It has been proposed that a pathological metabolic milieu in utero results in fetal epigenetic changes such as DNA methylation and histone modification, thus modulating biological processes related to intrauterine development, such as gene expression, chromatin accessibility, DNA replication, imprinting and human disease patterns.

Lipotoxicity in adults is well-established to result in the development of insulin resistance and pancreatic beta-cell dysfunction. Much emphasis has already been given to management of maternal hyperglycemia during pregnancy. It is possible however that maternal hypertriglyceridemia on the other hand will soon become a therapeutic target in the fight against childhood diabetes and obesity. Normal pregnancy is associated with a physiological increase in Tg from 10 weeks gestation onwards; with a 2-3 fold increase in serum Tg levels by the 3rd trimester. These elevations can be attributed to enhanced hepatic production of VLDL, increased intestinal absorption of dietary lipids, reduced clearance of Tg due to decreased extra-hepatic lipoprotein lipase activity and reduced insulin sensitivity during pregnancy. Even so, Tg levels during pregnancy are known to be higher in women with T2DM and GDM when compared with healthy pregnant controls. Maternal hypertriglyceridemia independent of maternal glucose has been linked with increased BW and fetal adiposity in offspring of Caucasian women. This association has been shown to hold true in women with GDM as well as those with Normal Glucose Tolerance (NGT) but risk factors for GDM. Maternal hypertriglyceridemia has even been linked with pre-eclampsia. Indeed, the fact that Langer et al, has demonstrated that insulin therapy results in lower macrosomia rates in obese GDM women than diet therapy despite equivalent glycemic control seems to indicate that some other metabolic target besides glucose which is modulated by insulin therapy is in play. It is highly likely that this 'invisible' metabolic variable is maternal triglyceride.

Determining early predictors of development of metabolic syndrome (MS) in offspring will enable targeted childhood intervention programs. In Western populations it has been established that pregravid maternal BMI is a strong independent predictor of childhood obesity/MS whereas birth weight and GDM alone are poor correlates of offspring obesity and risk of metabolic syndrome. In Boney et al's cohort of Caucasian offspring who were evaluated from age 6, for example, combination of large for gestational age (LGA) status and maternal GDM was associated with OR of 10.4 of insulin resistance in offspring at age 11yrs and increased prevalence of MS at any age. However the prevalence of MS in offspring was not significantly different between offspring of mothers with NGT whether LGA/average for gestational age (AGA), and AGA offspring of GDM mothers. Neither were maternal GDM alone and LGA at birth independently associated with insulin resistance. Maternal obesity alone however independently conferred an approximate increase in hazard of MS of 2-fold by 11 years of age whereas GDM status alone did not. The risk of developing MS in GDM offspring diverged between those with LGA and AGA status at age 7, so that by the time these children were 11 years old the risk was 3.6 fold greater in LGA compared with AGA offspring of GDM mothers. Alarmingly 50% of GDM offspring who had LGA status at birth had at least 2 out of 4 criteria of MS at any age-point between the ages of 6-11 years. These associations and relationships however are less well-delineated in Asian populations. In addition, much of these data, including those of Boney et al's are based on diagnostic definitions of GDM that have altered since the evidence obtained from the HAPO study. It is possible that Malaysian GDM offspring may develop a higher risk of metabolic syndrome at an age younger than 7 years and therefore require intervention programs before the age of school attendance. Hence the need for studies evaluating metabolic health in GDM offspring in the local Malaysian setting based on current diagnostic criteria of GDM that will enable more effective intervention programs tailored to the local setting and based on local data. There is also evidence in Caucasian and Hong Kong Chinese offspring that in utero hyperinsulinemia is an independent predictor of abnormal glucose tolerance in childhood. In Hong Kong Chinese offspring, cord blood insulin was very predictive of abnormal glucose tolerance in children aged 8 years (OR 6.12) whereas BW > 4kg was not, indicating perhaps that in Asians as well, BW is a poor early predictor of risk of metabolic dysfunction. Evidence indicates that dyslipidemia precedes development of dysglycemia in offspring of GDM mothers but that there are subtle differences between Caucasians and Asians with regards to which component of the lipid profile is impaired. Studies have shown that hypertriglyceridemia is more prevalent than dysglycemia in Caucasian GDM offspring aged 7-11 years , whereas low HDL is more common in Hong Kong Chinese GDM offspring aged 8.

Leptin is an established candidate gene for the pathogenesis of obesity and diabetes. It is secreted by adipocytes and placental tissue and is involved in energy metabolism and insulin sensitivity control. Leptin levels are increased in obesity, diabetes and pregnancy. 2hr post-OGTT glucose levels have been shown to correlate with placental leptin gene methylation levels in pregnant women with impaired glucose tolerance and increased placental leptin gene methylation is associated with decreased fetal leptin gene expression. 2 hour plasma glucose however only explains between 19-28% of DNA methylation variability at the leptin gene locus. The relationship between maternal hypertriglyceridemia and placental leptin gene methylation/leptin gene expression however has not been as well studied. It is possible that maternal triglyceride levels are also associated with alterations in placental leptin gene methylation thus leading to increased fetal adiposity. These changes in leptin gene DNA methylation adaptations are potentially important as they may be part of the mechanism whereby the in utero diabetic milieu results in fetal programming i.e. the induction of permanent changes in cellular structure or function in offspring that result in adult metabolic dysfunction/obesity. Plasma leptin is known to play an important neurotrophic role in hypothalamic development. It has been hypothesized that alterations to leptin homeostasis in utero may disrupt neuronal connections between nuclei that control energy balance and appetite in the hypothalamus perhaps leading to obesity.

Insulin resistance is a key feature of T2DM and has been implicated in the early pathogenesis of the disease. Micro-RNA 29 has been implicated in the development of insulin resistance in type 2 diabetes in vitro. Micro-RNAs are small untranslated/non-coding RNAs that negatively regulate mRNA translation. MiRNAs are involved in biological functions such as cell growth and proliferation, development, differentiation, organogenesis and metabolism and therefore have been linked to the pathogenesis of diabetes. In vitro studies with adipocyte cell lines have shown that genetic upregulation and overexpression of the miRNA 29 (a,b,c) family occurs in in the presence of hyperglycemia and that this blocks insulin-stimulated glucose uptake by the adipocyte by inhibiting insulin signalling via the Akt pathway (and hence expression of GLUT4 on the surface of the cell membrane). MiRNA 29 is also upregulated in skeletal muscle in a rat model of insulin resistance. There is also evidence that miRNA 29 is elevated in the presence of high NEFA levels and that this leads to skeletal muscle insulin resistance (unpublished data, personal communication, Zierath IDF 2013). Skeletal muscle is the primary site of glucose uptake postprandially, accounting for 75% of insulin-mediated glucose removal from plasma. Elevated miR29 is also associated with reduced insulin secretion. Human in vivo evidence linking miRNA 29 to insulin resistance and elevated NEFA is however lacking. Pregnancy is a meta-inflammatory state that is associated with markedly elevated levels of triglyceride, cytokines and NEFA in obese and dysglycemic women. It is possible that fetal exposure to high NEFA levels in women with gestational diabetes will result in elevated fetal miRNA 29 at birth which would be positively correlated with measures of insulin resistance such as HOMA2-%S. Removal from this glucolipotoxic milieu with parturition could result in a reduction in miRNA 29 with time and reduced insulin resistance.

There is however a dearth of prospective large studies examining the relationship between pregravid BMI/maternal hypertriglyceridemia in Asian mothers with GDM, and the metabolic health of offspring at birth and during childhood. Limited data on Korean women with GDM and Japanese pregnant women with NGT but positive diabetic screens is available but little research has been carried out in Southeast Asia. In addition these Asian studies only examined the link between maternal triglyceride and BW but not offspring body composition. Maternal hypertriglyceridemia has been associated with increased BW and adiposity at birth, but very few studies have looked at the association between maternal hypertriglyceridemia and body weight/ adiposity during childhood in either Asian or Caucasian populations. There is no consensus, even in western populations, on normal ranges of maternal triglyceride levels at different stages of pregnancy. Given the well-established fact that adult adiposity is greater in Asians compared with Caucasians of the same BMI, one could hypothesize that the association between maternal hyperlipidemia and offspring adiposity and metabolic health will be stronger in Asians. Asian mothers are known to be twice as likely to have GDM at a lower BMI. Indeed there is already some evidence that the correlation between maternal triglyceride levels and birth weight is stronger in Asian-born GDM women compared with Australian and European-born mothers. It is also possible Malaysian mothers with BMI considered to be in the non-obese range for Caucasian mothers i.e. 23-30 kg/m2 might have offspring with metabolic dysfunction. These associations have not been studied in Malaysian GDM offspring of Malay, Chinese and Indian descent. In recent times the trans-generational impact of maternal nutrition, via fetal epigenetic changes, on the metabolic health of offspring has been recognized. Given the differences, both current and historical, in socioeconomic development that impact on population health/wealth between Western Europe/North America and countries such as India, China and Malaysia it would be reasonable to postulate that maternal hypertriglyceridemia may have a larger impact on the health of offspring in developing nations with the sudden transition from poverty to relative wealth. It is worth evaluating the relationship between maternal metabolic characteristics during pregnancy, and offspring metabolic health, growth and development in Chinese and Indian Malaysians as these relationships may differ from those derived from studies done in Korea, Japan, Hong Kong, China and India given the markedly different historical trajectories of socio-economic development amongst even Asian countries. Very little work has also been done amongst mothers with GDM of Malay descent as this ethnic group is mainly only found in South-east Asia i.e. Malaysia and Indonesia.

With this in mind, we designed a prospective observational longitudinal cohort study to examine the associations of pregravid BMI and maternal hypertriglyceridemia in Asian GDM mothers with placental epigenetic changes to the placental leptin gene and offspring adiposity/metabolic function at birth/during childhood. We also aimed to evaluate cord blood miRNA29 at birth and 6 months of age. This study will also evaluate 2nd and 3 rd trimester serum triglyceride ranges in lean and obese Asian pregnant women with Normal Glucose Tolerance. The study will require participation of researchers from 5 disciplines: obstetrics, genetics, neonatology, paediatric endocrinology and adult endocrinology.

Inclusion Criteria:

1. >18 years of age
2. Malaysian Indian Chinese or Malay descent
3. singleton pregnancy
4. gestation 12- 32 weeks

Exclusion Criteria:

1. Multiple pregnancy
2. Assisted conception
3. pregestational type 1 or type 2 diabetes
4. endocrine disorders that affect weight/insulin resistance i.e. hypo- or hyper-thyroidism, Cushing's syndrome , acromegaly and phaeochromocytoma
5. overt diabetes diagnosed during pregnancy(FPG >7.0/ A1c >6.5%/ RPG> 11.1)
6. Smokers
7. Chronic systemic disease
8. Infection(HIV/Hepatitis B/C, TB, TORCHES)
9. diseases requiring treatment with exogenous steroids such as bronchial asthma, SLE, rheumatoid arthritis
10. Fetal anomalies
11. IUGR
12. premature delivery (<37 weeks by clinical assessment or ultrasound) likely because of maternal disease other than GDM/PIH
13. significant meconium or fetal distress on cardiotocography at delivery
14. Infants unable to undergo assessment of body composition within 48 hrs of birth

This study is implemented in 2 phases.

In Phase 1 , N=400, pregnant women (lean with NGT 150, obese with NGT 100, GDM 150) of Asian descent will be recruited from the UMMC Antenatal Clinic.

In Phase 2, the offspring of 150 GDM mothers, 40 lean NGT controls and 40 obese NGT mothers from Phase 1 form a cohort that will continue to be followed up at intervals (6 months, 1 year, 3 years, 5 years, 7 years and 10 years) till age 10 years with evaluation of body composition, growth, pubertal development and metabolic health. Equal numbers of girls and boys will be recruited. Data will be collected on breastfeeding practices and diet/physical activity.