The intrauterine period is a vulnerable period of development. Any adverse environment can permanently change the body's organ structure and function, expressed as an increased disease risk later in life. Studies show that variability in growth patterns in early life is associated with obesity and other cardiovascular diseases in adulthood, but the genetic and environmental determinants of these processes are largely unknown. The main objectives of this study were to identify genetic and environmental pre- and postnatal factors associated with early growth in infancy and childhood and later metabolic outcomes in adulthood from the Northern Finland Birth Cohorts (NFBCs). Several maternal and paternal factors, such as height, smoking, parity and pre-eclampsia, had direct association with faster postnatal height growth, some of which had their association mediated by size-at-birth variables. It was observed that an obesogenic environment in utero and during a child's growth exerts a 'programming' effect on the glucose-insulin axis as well as other cardio-vascular risk factors in adolescence. Moreover, the study shows that Leukocyte Telomere Length (LTL) at 31 years, a marker for aging, is inversely associated with multiple measures of adiposity in both men and women, and that a BMI increase in women from childhood to adulthood is associated with shorter telomeres at age 31. Two new genetic variants in/near SBNO1 and HMGA2 genes are associated with infant head circumference, which may indicate influence of brain growth and neurodevelopment via early life. Variants in/near LEPR-LEPROT, FTO, TFAP2B and GNPDA2 showed an age-dependent association with adiposity in early childhood, while three loci (FTO, TFAP2B and GNPDA2) had their effect on adult adiposity mediated by early growth phenotypes. This study emphasises the clinical importance of early growth markers as they may inform public health policy aimed at improving the pre-pregnancy environment and to monitor childhood growth during the first few years of development.

Batrachochytrium dendrobatidis (Bd) is a recently emerged multihost fungal pathogen, which has rapidly spread globally. It has been detected in 52 different countries, and in over 500 amphibian species. In susceptible hosts, it causes the disease chytridiomycosis. Although globally, Bd causes amphibian declines (and in some parts of the world, has resulted in a 40% loss of amphibian species), host responses are inconsistent and Bd appears able to coexist with some amphibian species or assemblages of species in a state of endemism. Batrachochytrium dendrobatidis was first detected in wild amphibians in the UK in 2004, in an introduced population of North American bullfrogs (Lithobates catesbeianus). Apparent spillover of infection into native species, including deaths in wild common toad (Bufo bufo) metamorphs at the index site, and in natterjack toads (Epidalea calamita) bred for reintroduction, has led to considerable concern for the conservation status of our native amphibian species. This thesis reports the results of cross-sectional, longitudinal and experimental investigations into the epidemiology of Bd in the UK. National cross-sectional surveys were conducted in 2008 and 2011, during which skin swabs were collected from almost 9000 amphibians and tested for the presence of Bd DNA using real-time PCR. Infection was detected at 30 sites (16% total), in all six native amphibian species tested, and in three of four non-native amphibian species tested. There was no evidence of change in either the distribution or prevalence of infected sites between the two survey years. There was no obvious spatial clustering and intra-site prevalence was almost uniformly low (median 10-12%). The results showed a strong positive association between Bd occurrence and the presence of non-native amphibian species, suggesting that co-introduction with non-native amphibians may have played a significant role in the current distribution of infection in the UK. Infection was also associated with sites occupied by natterjack toads, a species that has been the subject of long-term conservation management (including translocation events), providing further evidence that human-assisted movement of amphibians has contributed to the current distribution of Bd in the UK. A longitudinal study conducted in 2010 found a strong seasonal pattern of infection in newt species, with a higher prevalence of infection during the later, warmer months of the breeding season. Seasonal variation in Bd infection (and chytridiomycosis) has been commonly reported. However previous efforts have typically been conducted in tropical rather than temperate climates, where this pattern is reversed. This study also found, in terrestrially sampled common toads, an apparent increase in the prevalence of infection post-breeding suggesting that the breeding period is associated with an increased exposure or susceptibility to Bd (or both). Finally, two large-scale experiments were conducted to investigate infection in smooth newts (Lissotriton vulgaris) and palmate newts (Lissotriton helveticus). There was no evidence that experimental exposure to Bd was associated with increased mortality, or morbidity. In addition, even at high levels of exposure, infection was both rare and short lived, suggesting that these two native species are unlikely to provide a reservoir for infection in the UK. Overall, these results show a widespread but patchy distribution of infection, consistent with multiple point-introductions. At infected sites, the prevalence of Bd was low, and infected individuals had low infectious burdens, suggesting that the full expression of chytridiomycosis is not present in the UK at this time. A strong correlation with non-native amphibians and no evidence of broad-scale range expansion between 2008 and 2011 suggest that natural colonisation of ponds may not play a substantial role in the epidemiology of infection in this country. As a result, a combination of strict biosecurity protocols for fieldworkers, avoidance of long-distance translocations of native species and the prevention of future release of non-native species may be sufficient to control further spread of infection in the UK.

Among many behaviour changes associated with recent declines in HIV prevalence in Zimbabwe, the number of men reporting buying sex halved between 1999 and 2005 possibly due to economic deterioration reducing disposable income. As the economy recovers, a key question is how this is reflected in sexual behaviour trends. A recent modelling analysis of the modes of transmission of HIV in Zimbabwe suggested diminished role for sex work in driving new infections. A central conclusion was that the analysis was hampered by a lack of contemporary data on sex work in Zimbabwe. As Zimbabwe enters a new strategic phase of policy towards interventions, the data of the women at risk study are timely. Sex workers - defined broadly as women > 17 years old who received money, goods or any form of material benefit in exchange for sexual intercourse - were enrolled into a 12 month cohort across four socio-economic areas in rural Zimbabwe. Biological and behavioural data were collected. Two methods were used to recruit women: a venue-based strategy; and a modified snowball. Women were asked questions for 20-30 minutes in a face-to-face interview delivered either onsite in venues or at an arranged meeting time. The population size from the census was triangulated with a capture-recapture survey and with the prevalence of reporting on a general population survey. Depending on the defintions SW make up between 5.9% and 7.2% of the general female population according to WR study data. This is slightly higher than, but comparable to a general population estimate of 4.7%. The venue-based population was estimated as 1.36%. Research into the payments for sex revealed a 27% price reduction for sex when a client requests a condom. Client preferences determine whether a condom will be used or not. This has implications for interventions targeted at sex work.

Background The majority of unintended pregnancies are attributed to contraceptive failures from incorrect and inconsistent use. Long-Acting Reversible Contraception (LARCs) are highly effective in reducing unintended pregnancies, however, use of LARC is low in the United Kingdom (UK), and the intrauterine device (IUD) is the least used method in the UK. This thesis examines the impact of the pay for performance incentive on prescribing of long-acting reversible contraception in UK primary care, and also investigates the effectiveness of interventions to improve uptake of copper-IUD devices. Methods (1) Mixed-level logistic regression analysis of data from the UK Office of National Statistics ONS (2) Longitudinal data analysis of long-acting reversible contraception before and after the introduction of pay for performance in a sample of general practices across the UK (3) Cochrane systematic review on the effectiveness of interventions to improve the use of copper-IUD internationally. Results (1) Women who accessed their contraception from family planning clinics were more likely to use LARCs than those who accessed it from general practices. (2) Since pay for performance indicators on offering contraceptive information were implemented in 2009, the prescribing of LARCs has increased by 4% annually. This has resulted in 8700 more women being prescribed LARC at the practices included in this study. (3) Provision of contraceptive counselling and appointment and recall systems for IUD insertions was found to be effective in increasing uptake of copper-IUD, and the timing of counselling on pre- or post-natal patients was found to be important. Conclusions Uptake of more effective contraceptive methods such as LARCs was associated with financial incentives for contraceptive information provision in general practices. Provision of consistent and structured contraceptive counselling in primary care for women of reproductive age may improve the use of LARCs and should in turn reduce unintended pregnancies in the long-run.

Tackling obesity is a major public health challenge. Obesity strategy and policy in England has been heavily criticised for being ineffective in tackling the problem of obesity. Although many studies have evaluated interventions for obesity, there is no clear evidence about which interventions are effective in the prevention and management of obesity in primary care and community settings. The main purpose of the study was to evaluate the effectiveness of current public health approaches for obesity prevention and management, in particular focusing on behaviour change models as intervention and also policy actions implemented at local level (primary care trusts) in England. It also intended to identify obesity strategies relevant to Brunei as lessons learned from England. The methods used were systematic review, policy analysis and validation. The main finding of this study had shown that the 'Transtheoretical model stages of change' used in combination with diet and physical activity has limited impact on weight loss (about 2 kg or less) and there was no conclusive evidence for sustainable weight loss beyond 12 months. In addition, there were significant variations found in obesity strategies implemented by primary care trusts in north-west London based on analysis using the Imperial College Obesity Strategy Assessment Framework. The key lesson learned for Brunei from England's experiences was the development and application of Brunei-IC-OSAF that contributed to the formulation of 'comprehensive obesity policy' for Brunei. The tool had identified strengths and weaknesses of the existing obesity strategies implemented in the country. The outcomes of the study contributed to the existing evidence about what works in tackling obesity in public health settings, informing better pathway development and commissioning decisions.

Toxoplasma gondii affects a range of intermediate and secondary host species, including humans, with cats (Felidae) being the definitive host. The mechanism of action of T. gondii in its natural rat host can provide profound insights as to evolutionary trajectory of parasite manipulation. I used a range of non-invasive behavioural and physiological assays to assess the impact of infection on different aspects of innate rat behaviour. One key unique focus was an examination of the potential effect(s) of novel genetically modified parasite lines that overexpress tyrosine hydroxylase (TgTH) in comparison to wildtype T. gondii. The aim here was to test the hypothesis that TgTH, through its subsequent impact on neuromodulator levels, plays a significant mechanistic role in host behaviour modification. Consistent with previous studies, there was a clear effect of T. gondii on host behaviour, with effects seen in generalised anxiety and feline response tests. However contrary to predictions, there was little or no significant differences between wildtype and TgTH overexpressor infected rats in feline response tests, implying that TgTH does not play a major role in feline or olfactory related predation response. The implications of the role of TgTH in generalised anxiety and activity levels were more mixed. This suggests that other mechanisms of action, probably multiple, are involved. This work increases our understanding of the potential mechanisms this parasite has evolved to attain specific manipulation of the rodent intermediate host to increase transmission success to the feline definitive host. The results thus have both evolutionary implications for parasite-altered behaviour in intermediate host species, and applied implications in terms of the potential impact on health and behaviour of all infected hosts, including humans.

Infection with the Plasmodium falciparum malaria parasite results in an immune response which includes the production of antibodies against the blood-stage of infection. In recent years there has been an increase in the use of serological data to monitor malaria transmission intensity. Traditionally, EIR and parasite prevalence were the preferred tools for measuring malaria transmission intensity. Serology has been shown to be particularly useful in areas of low endemicity where traditional measures (EIR and parasite prevalence) are problematic. Transmission intensity in this case is usually described by the seroconversion rate obtained from fitting a catalytic model to age-stratified serological responses. The aim of my thesis was to better utilise the continuous measurements of antibody responses provided by serology studies to obtain improved estimates of transmission intensity. To do this, I developed a series of biologically motivated models to mimic the acquisition and decay in blood-stage antibody responses. In the first part of the thesis, I developed a discrete model as a direct extension of the catalytic model and fitted this to cross-sectional data from several sites in Cambodia to obtain an estimate of the exposure rate. In the second part of the thesis a series of continuous density models were developed to mimic antibody acquisition and loss for P. falciparum infections. These models were fitted to both the Cambodian data and separately validated by fitting to data from Tanzanian villages at a wider range of transmission intensities. In the final section I applied and extended the model to encompass a wider range of endemic transmission in Somalia, Bioko Island, Gambia and Uganda in order to assess the robustness of the method. My results show that estimates of the exposure rate obtained by fitting the density model are highly correlated with classic malariometric indices and that a key advantage of this approach is the increased precision in the estimates compared to estimates of the seroconversion rate, especially in areas of low transmission. This method could therefore be a useful alternative framework for quantifying transmission intensity which makes more complete use of serological data and shows potentials for detecting heterogeneity in malaria exposure.

Background: The NHS Health Check, the largest systematic cardiovascular disease (CVD) primary prevention programme globally, aims to reduce CVD burden and health inequalities by assessing and managing CVD risk among 40 to 74 year old individuals without existing vascular diseases. I evaluated the impact of the programme at local and national levels. Methods: Using electronic medical record data from general practices in Hammersmith and Fulham, I assessed CVD risk factor recording before the programme, the programme uptake in the first two years and the impact of the programme on CVD risk. National coverage of the programme in one financial year was assessed using data from Primary Care Trusts (PCTs). Results: There was good recording of smoking status (86.1%) and blood pressure (82.5%), with lower BMI (59.5%) and cholesterol (47.5%) recording among Health Check eligible patients before the programme in Hammersmith and Fulham. Uptake of the Health Check was lower than the national target (75%) at 39.2% among patients with an estimated high CVD risk, but matched the national required rate at 20.0% among all remaining eligible patients. There was significant reduction in mean global CVD risk score (28.2% to 26.2%) after one year among patients with estimated high risk that had a complete Health Check. The programme uptake was higher in patients living in more deprived areas among those not at estimated high risk (adjusted odds ratio = 0.88 (0.73-106)). Mean national coverage of the programme was lower (8.1%) than anticipated (18%), with large PCT-level variation (0% to 29.8%). Coverage was significantly greater in PCTs in more deprived areas (coefficient = -0.51 (-1.88-0.00), p-value: 0.035). Conclusions: Population-wide impact of the NHS Health Check may be limited by poor uptake of the programme. This and other limitations to the programme suggest that a targeted screening approach along with population-wide strategies may be a better option for more cost-effective prevention of CVD.

Background: Evidence of an association between marital status and well-being has been demonstrated, with married people reporting higher levels of well-being. However, the strength of this relationship in later life may be influenced by both societal context and gender. This thesis will examine the association between marital status and quality of life in older people and consider if this relationship varies by welfare arrangements and if gender moderates these associations. This research will also explore how quality of life is experienced within married couples. Methods: Quality of life was measured using CASP-12. The relationship with both current and past marital status was examined. Analyses were conducted using data from The Survey of Health, Ageing and Retirement in Europe (SHARE) and the English Longitudinal Study of Ageing (ELSA). A welfare state regime approach was used to examine societal context and due to feminist criticisms of these approaches, two further methods of comparison were used. A dyadic data technique was also used to examine the interdependence of quality of life for married people. Results: Current marital status was found to be a predictor of quality of life at older ages. However, variation in this association was observed across the welfare state regimes. When health and socio-economic circumstances were taken into account, the advantage of marriage for quality of life was often attenuated. Gender differences in this relationship were also observed; especially when gender focused methods were examined. Spousal interdependence of quality of life was also observed within married couples. Conclusions: This research expands our understanding of the association between marriage and well-being by suggesting that societal context is important and that marriage may offer men and women different kinds of protection as they age. It also suggests that for older married couples, quality of life is not just an individual experience.

Lymphogranuloma venereum (LGV) is a re-emerging sexually transmitted infection (STI) in men who have sex with men (MSM). The emergence of LGV has occurred in parallel to an increase in STIs in MSM. It represents a public health problem and an added burden in the control of STIs. This thesis aims to identify factors that have contributed to the persistent high levels of HIV prevalence among LGV cases. I use surveillance data to explore the re-emergence of LGV through statistical analyses. I then perform a literature review to better understand the social epidemiological context in which seroadaptive behaviours in HIV-positive MSM occur, a hypothesised cause for the re-emerging STIs. I finally look at the dynamics of HIV and LGV through a deterministic mathematical model where the effect of serosorting is investigated in more detail. Through statistical analyses I demonstrate that there is a strong behavioural component in explaining the association between the two infections. In the LGV Enhanced Surveillance HIV-positive LGV patients were more likely to report unprotected receptive anal intercourse compared to HIV-negative/unknown LGV patients. In a subsequent analysis I show that individuals with reported LGV re-infection were more likely to be HIV-positive, visit a clinic in London and have concurrent hepatitis C and gonorrhoea on their first recorded LGV episode. However, the data also suggests there is a diagnostic bias in favour of HIV-positive men who present with shorter duration of symptoms. I propose a conceptual framework of seroadaptive behaviours in HIV-positive MSM based on a literature review. The mathematical model demonstrates that infections with the same mode of transmission will be associated, but it provides further support to the hypothesis that serosorting can explain the particularly high HIV prevalence in those who acquire LGV. This work cannot exclude the possibility of a biological interaction but the current evidence points to behavioural, and likely network-level, differences between HIV-positive and -negative MSM as the main driver in LGV re-emergence. This has implications for surveillance and control of LGV. Rare STIs, such as LGV, may benefit from a more detailed and qualitative investigation of cases using a periodic sentinel model as well as from health promotion targeted specifically to the small subpopulation affected.