Fetal exposure to excess glucocorticoids has been proposed as a key determinant of pregnancy outcome, as well as a predictor of long term health of the offspring through a phenomenon known as ‘developmental programming’. Obesity and ‘stress’ during pregnancy are two potential sources of altered fetal exposure to glucocorticoids. One in five pregnant women is obese at antenatal booking, and maternal obesity increases risk of offspring complications including higher birth weight, potentially leading to long-term programming effects on the offspring. Likewise, maternal anxiety during pregnancy has been identified as a programming factor, increasing the risk of psychopathology in the offspring. This thesis tests the hypothesis that in humans this association is mediated by altered action of glucocorticoids, by examining circulating levels of maternal glucocorticoids during pregnancy and through measurement of key genes in the placenta regulating fetal glucocorticoid exposure. Serum cortisol levels were measured at 16, 28 and 36 weeks gestation in n=173 class III obese (BMI 44.0±4.5kg/m2) and n=107 lean (BMI 22.8±1.6kg/m2) pregnant women. Serial corticosteroid binding globulin (CBG) concentrations were measured in a subset (n=39 lean, 26 obese) and free cortisol levels calculated using Coolen’s equation. CRH concentrations were measured at the same time points in obese (n=20) and lean (n=22) pregnant women Salivary cortisol was measured in samples collected at bed-time, waking and 30 minutes after waking. mRNA levels of candidate genes regulating glucocorticoids and fetal/placental growth including 11-beta hydroxysteroid dehydrogenase type 2 (11βHSD2), which inactivates cortisol, insulin-like growth factor 2 (IGF2) and glucocorticoid receptor (GR) were measured in first trimester (n=32), second trimester (n=15) and term (n=60) placental samples. DNA methylation of key regions controlling the expression of the IGF2, GR and 11βHSD2 genes was measured by pyrosequencing in first trimester and term samples. Levels of mRNAs encoding 11βHSD1, 11βHSD2, GR and MR were measured in term placentas collected from women from Helsinki, Finland in whom anxiety during pregnancy had been prospectively assessed using validated questionnaires. Term placental samples from a subset of the obese and lean women who had also completed stress questionnaires during pregnancy were used to examine replication of findings. Cortisol levels rose similarly during pregnancy in obese and lean but were significantly lower throughout pregnancy in obese women (p<0.05). The diurnal rhythm of cortisol was maintained. CBG levels also increased, though this change was lower in obese (1.21-fold (±0.9) vs 1.56-fold (±0.07), p<0.01). In obese women, lower calculated free cortisol at 16 weeks gestation was associated with higher birth weight after adjustment for other factors (r=-0.46, p<0.05). Placental mRNA encoding 11βHSD2 increased in association with increasing obesity in early pregnancy (r=0.44, p<0.01) and was highest in term placenta in obese women with macrosomic (>4000g) offspring (p<0.05). Placental transcript abundance of GR also increased in association with increasing obesity in early pregnancy (r=0.38, p<0.05), but was lowest in term placenta from obese with macrosomic offspring (p<0.05). IGF2 mRNA abundance was lower in the placentas of obese women with macrosomic offspring at term compared to both lean women and obese women with normal weight offspring (p<0.01). Methylation results are reported. Placental mRNA levels encoding 11βHSD1 (which converts inactive cortisone to active cortisol) at term was found to positively associate with maternal anxiety measured in the first trimester of pregnancy in a group of pregnant Finnish women (β=0.3, p<0.05). Findings were similar in the replication sample in lean women only (β=4.6, p<0.05). Lower circulating and bioavailable cortisol levels in early pregnancy, together with a greater placental ‘barrier’ to maternal glucocorticoids represent key mechanisms contributing to higher birth weight in offspring of obese women. Regeneration of active glucocorticoids in placenta and increasing placental sensitivity to glucocorticoids increases fetal glucocorticoid exposure and offers insight into the biological mechanisms underlying adverse offspring effects of maternal prenatal anxiety.

This qualitative descriptive study investigated the unique and specific impact of being pregnant while one's husband was deployed to a combat zone. Three specific aims were used to address the study objectives 1) Describe the experience of being pregnant while one's husband is deployed 2) Describe the women's experiences with health-care providers 3) Describe the types of support women sought or had access to during pregnancy. Participants were 16 women who had been pregnant while their husbands were deployed for greater than 30 days from 2004-2014, with no prior personal history of being deployed. Participants completed demographic questionnaires on their pregnancy course, and their husband's military and deployment history. Semi-structured telephone interviews were conducted with each participant individually. Content and matrix analysis were utilized to explore the study aims. The Stress and Coping Model by Lazarus and Folkman (1984) provided the theoretical framework for this study. Communication was an important part of receiving support from the husband and daily communication (n=4, 25%) was associated with more problem-based coping (75%) and feelings of emotional support from the husband (38%). Six participants stated a history of preterm labor or birth, and most of those participants used emotion-based coping predominantly (57%). Participants who perceived that their husbands experienced significant danger were more likely to use emotion-based coping (56%). Participants overall had more positive interactions with certified nurse midwives (76%) and civilian obstetric physicians (77%), and reported more negative interactions with military obstetric providers (87%). Primiparous participants reported that 61% of all experiences with providers were positive while multiparous participants were more likely to have negative (66%) or mixed (6%) experiences. Support systems sought or accessed were different for officer and enlisted wives as well as for different ages. The wives of enlisted soldiers were more likely to not participate, or have an unfavorable view (52%) of the FRGs. Whereas, the wives of officers felt more support and involvement (69%). Wives who were 29-years-old or less sought out more support from friends/co-workers (33%) than the 30-years-old or older group (19%). The 30-years-old or older group was more likely to have sought support from family (50%) versus the 29-years-old or younger group (40%). The difference in support sought or accessed from the FRG between the two age groups was much less significant (28% for ≤ 29-years old versus 31% for ≥ 30-years old). The findings from this study could be adapted to create a screening tool that would alert providers to those pregnant women who might need specific resources or social support.

In recent years, worldwide obesity rates have risen dramatically, putting major strain on public health systems and the economy. Obesity is a multifaceted disease and its development can be influenced by a variety of factors including genetic, psychological and environmental influences. One area of current focus in obesity research is that of early life programming. It has been well-established that certain early life factors can impact the physiology and behaviour of the offspring. Because of this, early life programming has become increasingly well studied in order to develop a deeper understanding of how early life can influence obesity development. Another area of interest lies in positive mood. While there has been much research into the effects of negative states such as stress and anxiety on feeding behaviour, there is still very little known about how positive states can influence food choice. Using rat models of prenatal stress, neonatal overnutrition and positive affect, this thesis aimed to investigate the effects of early life and mood factors on feeding behaviour and food choice. Prenatal stress has been extensively studied and is characterised by an enhanced stress response in the offspring. Using two rat models of prenatal stress- social and restraint stress, the effects of prenatal stress on feeding behaviour and food choice in the offspring were examined. In both models, no effects of prenatal stress on either food intake or food choice were observed. However, in both cases the expected alterations to the offspring's stress responses when exposed to an acute stressor were not replicated. This may suggest that models of prenatal stress are not as robust as often cited in the literature. As well as the prenatal environment, the early postnatal environment is also able to influence physiology and behaviour. In terms of obesity, a well-studied model is that of small litter size. Rats from small litters are over-nourished as neonates and because of this illustrate an increased body weight that persists throughout life. While this increase in weight gain has been well-established, there is no evidence examining the impact of neonatal overnutrition on long-term food choice. Therefore, food intake and food choice were measured in small and control litter rats over a 10-week period. When placed on an ad lib diet of bland chow, sucrose and lard, small litter rats consumed significantly more chow than control litter rats, whilst maintaining similar consumption of lard and sucrose. However, when offered a high-fat high-sugar (HFHS) pellet for two hours a day alongside ad lib chow, small litter rats illustrated increased consumption of the HFHS pellet compared to controls. This suggests that small litter rats may be programmed to adjust their food choices to enable them to maintain their increased body weight in comparison to controls. To examine the effects of positive affect on feeding behaviour, ultrasonic vocalisations (USVs, specifically those at 50 kHz) were used as a measure of positive affect in rats. In order to examine whether access to a food reward could induce a positive affect (as measured by an increase in 50 kHz USVs), rats were schedule-fed sweetened condensed milk and USVs measured before, during and after consumption. No differences in 50 kHz USVs were observed suggesting that a palatable food, whilst rewarding, does not alter affective state in the rat. Using heterospecific social contact (a tickling interaction simulating rough and tumble play) to induce positive affect, rats were presented with an hour-long sucrose preference test following social contact in order to examine the impact of positive affect on food choice. While no differences in sucrose consumption were found, a reduced sucrose preference was observed in rats receiving social contact compared to controls, suggesting that positive affect may play a role in mediating food choice. Finally, the effects of fasting (a negative stimulus thought to reduce 50 kHz USVs) and a food reward on motivation for social contact were examined. Both fasting and access to a food reward resulted in no differences in conditioned place preference to receive social interaction. Overall, the results obtained in this thesis implicate both neonatal overnutrition and, for the first time, positive affect as possible mediators of food choice, although further studies are required to fully establish these effects. Importantly, these results also raise questions regarding the reproducibility of some early life models, such as prenatal stress, and highlights the importance of sharing precise experimental protocols across laboratories. Through further investigation of the effects of early life and affective states on food consumption and choice, and the mechanisms behind these, this may enable the development of therapeutic interventions and preventative measures that can help slow, or even reverse, the global obesity epidemic.