Search

Background: The current approach to prenatal care, which was established more than 80 years ago, is characterised by a high concentration of visits in the third-trimester of pregnancy which implies that firstly, most complications occur at this late stage of pregnancy and secondly, most adverse outcomes are unpredictable during the first or even the second trimester. Objectives: The objective of this thesis is to provide evidence that most pregnancy complications are predictable as early as 12 weeks’ gestation. The pregnancy complications examined include fetal aneuploidies, fetal structural defects, preeclampsia, preterm birth, gestational diabetes mellitus and fetal macrosomia. Methods: I have critically examined fourteen articles reporting on screening for pregnancy complications at 11-13 weeks’ gestation, where more than 90,000 singleton pregnancies were prospectively assessed at 11-13 weeks’ gestation as part of a routine prenatal visit for screening for trisomy 21. We recorded a series of maternal characteristics and history, measured maternal weight and height, performed a detailed ultrasound examination of the fetus, measured maternal uterine artery Doppler pulsatility index and maternal mean arterial pressure and collected blood for analysis of biomarkers for prospective or retrospective analysis. All data were prospectively entered into our data base as well as the pregnancy outcomes as soon as they became available. Ethical approval was obtained for these studies. Multivariate regression analysis was used to define the contribution of each maternal characteristic and history in predicting each adverse outcome and those with a significant contribution formed an algorithm to estimate the background risk (a priori risk) for each one of these complications. The potential value of biophysical and biochemical markers in improving the performance of the a priori risk in predicting adverse pregnancy outcomes, was evaluated. Results: First trimester effective screening for adverse pregnancy outcomes was provided by a combination of maternal factors and biophysical or biochemical markers. The developed predictive models could correctly identify the vast majority of aneuploidies, early preeclampsia and more than half of the cases of spontaneous preterm birth and gestational diabetes. First trimester prediction of fetal macrosomia was less effective compared with other complications. First trimester examination of fetal anatomy was feasible resulting in a high detection of fetal non-chromosomal defects, including more than half of fetal cardiac defects. Conclusions: Assessment of the mother and fetus at 11-13 weeks’ gestation can provide effective early identification of the high risk group of pregnancies with fetal and maternal adverse outcomes.

The aim of this thesis was to examine the role of biobehavioural and social variables in explaining adolescent antisocial behaviour. One study examined neuropsychological functioning in 115 young offenders. A more extensive second study was carried out on a sub-sample of the original young offender group, consisting of 48 participants. This second study used more detailed neuropsychological assessments and assessed participants' responses to emotional stimuli. Emotional functioning was assessed in 3 ways: by recording electrodermal responses during a fear conditioning task, by recording the eye-blink startle reflex while participants passively viewed different types of affective pictures, and by examining facial affect recognition. It was expected, first, that antisocial teenagers would be characterised by a sensation-seeking personality, neuropsychological impairments as evidenced by executive functioning tasks, low IQ, poor electrodermal fear conditioning, and reduced startle amplitudes, compared to age and sex matched controls. Second, it was expected that biobehavioural risk factors would interact with social risk factors in explaining ASB, and that social factors would moderate the biobehavioural - ASB relationship. We found that young offenders differed from matched controls in terms of personality traits, and neuropsychological and emotional functioning. With respect to the second hypothesis, it was found that biobehavioural risk factors did not interact with social variables in explaining different types of offending behaviour, contrary to previous studies. Specifically, the research findings indicated that young offenders were characterised by lower IQ and specific neuropsychological deficits in terms of working memory, planning and decision-making. Additionally, they had problems with the learning, processing, and recognition of emotions. Finally, we showed that different risk factors were associated with different types of offending, with both social and biobehavioural variables predicting prolific and persistent offending, and only biobehavioural factors predicting severe offending. The implications of these findings for policy and practitioners working with young offenders were discussed.

This thesis investigates three cutting edge issues in empirical finance. The first, examined in Chapter 2, is an investigation of the hypothesis that macroeconomic uncertainty is a significant risk factor in explaining deviations from the uncovered interest parity (UIP) condition (or time-varying risk premium) using data from the G7 countries. To analyze the relationship between the risk premium and macroeconomic risk factors, we employ VAR-GARCH-in-mean models. The results show that the currency risk premium may be due to macroeconomic volatility. The second issue (Chapter 3) concerns the causal linkages between monetary and financial market returns of the New Member States (NMS) with the euro zone following the introduction of the euro. To measure monetary convergence we employ UIP deviations and stock market integration is proxied by the differential of the risk-adjusted returns of the NMS countries versus a euro zone stock index. We look for a causal relationship between excess currency and excess stock market returns of the aforementioned countries. We employ Granger causality in mean and variance tests to assess this relationship. The main finding is that the excess currency return is a significant leading indicator of the excess stock market return, with stronger evidence of causality in the variance of the process. Finally, the third issue (Chapter 4) examines the time-varying co-movements of equity returns within major sectors, for a number of developed and developing countries. In particular, we ask the question if the conditional return correlations have increased excessively by historical standards during the recent financial crisis. We utilize asymmetric multivariate GARCH models. The main finding is that correlations have increased over the recent period, but the levels of co-movements are not excessive by historical standards. There is little difference between the co-movements of financial and non-financial stocks across countries, implying that this shock is largely undiversifiable.