Many diseases can be prevented or mitigated through behaviour change, but we lack a quantitative model that can accurately predict these changes and inform policies designed to promote them. Here we introduce a quantitative model of health behaviour that takes into account individual-level barriers, the health system, and spread between individuals. We investigate limits of the model where each of these determining factors is dominant, and use them to predict behaviour from data. We apply the model to individual-level geographic barriers to mothers giving birth in a health facility, and find evidence that ease-of-access is a major determinant of delivery location. The geographic barriers allow us to explain the observed spatial distribution of this behaviour, and to accurately predict low prevalence regions. We then apply the model to the role of the health system in determining health facility usage by mothers of sick children. We show that local health facility quality does predict usage, but that this predictive power is significantly less than that gained by including unaccounted-for spatial correlation such as social influence. We also show evidence that results-based funding, rather than traditional input-based funding, increases usage. We develop a psychologically-motivated ‘complex contagion’ model for social influence and incorporate it into a general model of behaviour spread. We apply this model to short-lived behavioural fads, and show that ‘nudges’ can be very effective in systems with social influence. We successfully fit the model to data for the online spread of real-world behaviour, and use it to predict the peak time and duration of a fad before the peak occurred. Finally, we discuss ways to incorporate disease state into the model, and to relax the limits used in the rest of the thesis. We consider a model which links health behaviour to disease, and show that complex contagion leads to a feature that is not present in traditional models of disease: the survival of an epidemic depends non-trivially on the initial fraction of the population that is infected. We then introduce two possible models that include both social influence and an inhomogeneous population, and discuss the type of data that might be required to use them predictively. The model introduced here can be used to understand and predict health behaviours, and we therefore believe that it provides a valuable tool for informing policies to combat disease.

The main areas considered within this qualitative study are the extent to which the Adult Support and Protection (Scotland) Act (2007) (ASPA) impacts upon the civil and human rights of adults’ by exploring the “problem” it was developed to resolve, the reality of implementation and the construction of thresholds for intervention in practice. Despite a level of clarity about the need for this legislation inconsistencies of understanding about where the ASPA should be targeted created challenges for implementation, particularly around the issue of capacity. The scope of the population for whom the ASPA was intended remains sizeable and broadly unformed. The vision of the framers that the ASPA would provide support and protection for a range of adults at risk of harm without being overly intrusive in their lives appears, at least partly, to have been realised. Challenges to implementation have largely focused on; the parameters of the ASPA and the population it aims to protect, the conceptualising of what an adult protection referral might consist of and the impact of this understanding on thresholds for intervention. Procedural challenges identified were specifically related to the involvement of health and the understanding of adult protection of other stakeholders, for example the police, inconsistent recording of data and information sharing. The interaction between formal and informal knowledge and consideration of a range of key concepts drawn upon by practitioners to determine thresholds for intervention creates a built in inconsistency of approach with a clear element of subjectivity. The rights based approach integral to all intervention under the ASPA, was well applied by the practitioners in the study and could be considered to have protected the citizenship of the adults to some extent. Perhaps more accurately it could be said that the already conditional citizenship experienced by many of the adults was not further eroded.

Street children are a population highly at risk of HIV/AIDS/STIs, which is becoming an overriding concern. Due to the critical importance of the problem under investigation, this study focuses on the causes and consequences of risks involved in the dynamics of HIV/STIs transmission and the occurrence of AIDS. The study utilised a qualitative paradigm, with two methods of data collection from children and young people in the street; these were observation and in-depth interviews, which emerged as the most appropriate methods for investigating the HIV/AIDS risks and vulnerability of street children. The study was guided throughout by a public health theoretical framework. The study revealed that children leave home due to parental mistreatment; they engage in risky sexual behaviour living in the street, they have little or no understanding of HIV, AIDS and STIs or of the respective relationship between these, and they have negative attitudes towards HIV/ STIs treatment and people affected by HIV/AIDS. Four domains of HIV/STIs and AIDS risks and vulnerability of street children were identified: parental mistreatment (causing vulnerability to exposure and thus the likelihood of acquiring HIV and STIs); high risk-taking sexual behaviour (creating vulnerability to infection); lack of knowledge regarding HIV, AIDS and STIs (vulnerability to re-infection); negative attitudes towards HIV/STIs treatment and people affected by HIV/AIDS (resulting in denial, failure to seek treatment and contributing to the perpetuation of the problem); and the effects of living in the street (increasing vulnerability to progression from HIV to AIDS). By exploring the prime and subsequent root risk factors, these complex interlinking risks have been analytically conceptualised, providing a model which explicates the complete phenomenon of risks and vulnerability to HIV/STIs and AIDS for street children, as well as for broader society, in a cyclical manner. Hence, HIV/STIs and AIDS is not a health problem among street children only, it is a public health problem in the broader society in Nepal. Having identified these problems for street children, this study offers an intervention plan, the CAP model. This model extends previous public health approaches and argues for targeted action to prevent risk and vulnerability for children in the street, and suggestions for policy and legislation which would enable the implementation of the model are offered.

The role that dental team members can play in the smoking cessation field is largely unevaluated. The work of this PhD thesis encompasses two phases, the first was to develop a means of determining smoking status, using analysis of continine, a nicotine metabolite. The second phase looked at the efficacy of dental hygienist-delivered smoking cessation advice in a dental setting. The aim of the first study was to compare continue levels in different biological fluids collected from both smokers and non-smokers, and to relate the findings to self-reported smoking status. Patients recruited to the study were asked to provide samples of urine, blood and saliva (both stimulated and unstimulated). Data collected from patients by questionnaire included information on smoking behaviour, such as daily number of cigarettes smoked, and environmental exposure to smoke. Following sample collection, patients were asked to rate the acceptability of each sampling method. Samples were analysed using enzyme immunoassay (EIA) kits. In total, 80 patients participated, with 49 smokers and 31 non-smokers. There was clear differentiation between smokers and non-smokers (p<0.001) for all the different samples in terms of cotinine concentration. A significant relationship was seen between cotinine levels and daily number of cigarettes for both salivas and urine (all p < 0.001) but not for serum. Participants found serum and urine collection methodologies 'very acceptable' (67% and 66%, respectively) whereas 9% found collection of stimulated saliva 'not at all acceptable'. Thus, continine, as analysed by EIA kits, whatever the collection method, shows good differentiation between smokers and non-smokers. Salivary samples have the advantage of being non-invasive. However, collection methodology is important, as continine levels may vary.

Previous research has suggested the existence of socio-economic and gender variation in uptake of hospital cardiology services but it is not known at which stage of the care pathway these variations arise. This thesis aims to address this gap in the literature by exploring the perceptions of an behaviour in response to chest pain of men and women living in two socio-economically contrasting areas of Glasgow. It has 3 objectives: to describe respondents' perceptions of and actions in response to chest pain; to ascertain whether there are differences in perceptions and behaviour associated with socio-economic status or gender; and, to relate any observed differences to the known variations in uptake of cardiology services. Respondents' interpretations of chest pain were related to their perceived vulnerability to heart disease and their previous illness experience. Their decisions of whether to present to a general practitioner were influenced by the quality of previous encounters with doctors and by their lay care and self-care. Women generally felt less vulnerable to heart disease than men, and often believed that their health problems were secondary to those of family members. These factors led women to report a reluctance to present with chest pain. Compared with more affluent people, those from the deprived area generally felt more vulnerable to heart disease but the perception that they were to blame for their ill-health often led to reluctance to present with chest pain. Differences in perceptions and behaviour might partly explain gender inequities in access to hospital cardiology services but socio-economic inequities in uptake of services have proved more difficult to explain.

Background: The incidence and prevalence of diabetes mellitus (diabetes) is increasing globally. The majority of people have type 2 diabetes, initially managed by lifestyle changes and the gradual introduction of oral, followed by subcutaneous, medications. Due to the increasing numbers of people with type 2 diabetes, different models of care management have been utilised within the United Kingdom. One Health Board within Scotland received funding to undertake a service redesign whereby the responsibility of care for people with type 2 diabetes was transferred from secondary care to primary care, with referral to specialist services based on clinical need. Aim: The aim of this thesis is to present the body of published work on an evaluation research study on the change in service delivery for people with type 2 diabetes. The published works have arisen from two funded research studies resulting in one peer reviewed report and five peer-reviewed papers published in international journals.

Background: Obesity has emerged as a major public health problem across the globe. Unfortunately, all the efforts, to date, to treat obesity have limited success. Despite the increased publicity on health and economic consequences, its prevalence continues to rise, even in countries that were previously battling under-nutrition. The annual weight gain which may lead to obesity if it remains uncontrolled, is small, averaging 0.7-1.0kg/year. Aiming to halt this small weight gain could be a solution towards combating the obesity epidemic and thus reduce its prevalence. One of the critical life stages where weight gain occurs is the transition between adolescence and young adulthood and especially evident in those attending higher education in the US. There is very limited data from the UK on the magnitude of weight gain in this population. This thesis aimed to explore the weight changes occurring during that life-stage, identify factors affecting those weight changes, explore weight gain prevention data and design and test interventions aiming to prevent any weight gain. Methods: Several studies were carried out during this PhD programme in order to answer all the research questions. Three systematic reviews were employed to examine 1) weight changes in young adulthood, 2) weight gain prevention studies in young adulthood and 3) the effect of calorie-labelling on calories purchased as an anti-obesity measure. A pilot study was carried out to test and explore methodologies for collecting data from young adults on lifestyle and lifestyle changes. A qualitative study was carried out to supplement the data from the pilot study on the importance of any weight changes for young adults. A prospective cross-sectional study was carried out to examine the weight changes and lifestyle changes occurring during the first year of studies in young adults. An interrupted time-series study was carried out to test the hypothesis that calorie-labelling might have an effect on preventing weight gain in young adults. A cross-sectional study was carried out to test the effect of calorie-labelling on sales and choices in independent catering facilities where young adults represent a significant proportion of the customers. Lastly, a randomised trial was carried out to test the hypothesis than on-line programmes based on two different behavioural theories could help young adults to avoid any unwanted weight gain. Results: The systematic literature review of weight changes in young adults identified 27 studies reporting a mean weight increase of 0.7-3.75kg in those attending higher education. The pilot study examining methodologies and weight changes in young adults attending higher education in the UK found a weight increase between 0.5-5.5kg by 56% of the participants and the best recruitment method to be the on-line method compared to mail or in-person recruitment. The prospective study that looked at weight changes among first-year students attending a large university in the UK found a weight change of 1.8kg in a 9-month period. Baseline weight explained 48% of the variation observed in weight changes. Despite the belief that physical activity or consumption of fruit and vegetables is linked to weight management, neither of these protected against weight gain. The literature review on weight gain prevention studies among young adults, identified twelve studies (five of those conducted in higher education settings). Six of the studies found an effect on preventing weight gain or maintaining weight. No specific techniques were identified to be more effective as the studies that found an effect followed similar techniques with those that did not. The systematic review and meta-analysis on the effect of calorie-labelling on calories chosen/purchased identified seven studies. Overall, there was no effect of calorie-labelling on calories chosen/purchased, however customers noticing the calorie-labels, reduced the calories chosen/purchased by 124.5kcal. Students are generally supportive of the presence of calorie information in a range of products and settings including alcohol products. In the time-interrupted study which was conducted over two years, young adults that were exposed to calorie information did not gain the expected weight observed in young adults in the same setting in the year prior to the implementation of calorie-labelling. The cross-sectional study conducted in an independent catering outlet examined the effect of calorie-labelling on sales of products. Prominent calorie-labelling led to substantial reduction in sales of all labelled products but mostly among those that were high calorie products. The randomised controlled trial led to weight loss among those who were randomised to the intervention groups while those in the control group gained the anticipated weight over a 9-month period. Conclusion: Young adults in the UK gain weight when starting higher education. The weight-gain is similar to that observed in young adults in higher education in other countries but higher than the weight gain observed in the general population. Interventions based on different behavioural models were all successful at abolishing this weight–gain. Applying these interventions in a larger scale or making them part of future public health policies could be a significant step towards halting the obesity epidemic.

This thesis examines the changes in health governance at both global and country levels brought by the Global Fund to Fight AIDS, Tuberculosis and Malaria (the Global Fund), a self-described public/private partnership intended as a financing mechanism to achieve Millennium Development Goal (MDG) 6. Since the G8 announced the Global Fund's creation in 2001, it has succeeded in mobilising over US$30 bn in commitments, primarily from donor governments. This thesis is rooted in the 'high politics' of International Relations (IR), and in particular its literature on globalisation, governance and international institutions. Where this literature has failures or gaps, it draws from the Development Studies and International Political Economy (IPE) literatures. It also relies on key informant interviews undertaken in Geneva, Lilongwe and Zomba with executives of international institutions, and those involved in Malawi's HIV/AIDS response including government representatives and staff from the National AIDS Commission, donors, NGOs and those working on the front line. This thesis relies on a descriptive, single case study to create a 'thick' narrative. Rather than deriving generalisations, it provides a basis for further research into the nature and effects of systemic change in how health is governed that the Global Fund signals. This thesis makes three contributions to knowledge: 1) It provides a basis to evolve the IR literature on globalisation, governance and international institutions to consider the nature, significance and effects of the Global Fund as a form of institutional innovation which is disrupting the traditional multilateral order, particularly for international institutions working in health; 2) It challenges the use of the term 'country ownership' to mean 'putting the country in the driver's seat', and instead notes the double deficit in external accountability that arises when global politics and country evidence collide in a Global Fund convened elite, mediated space for country ownership; and 3) It synthesises observations from field work in Malawi on the exercise of the Global Fund's authority and its dislocation from external accountability when failures occur. The IR literature is silent on the rise of the Global Fund's authority. It fails to contend with the notion that country ownership is as much about the burden of responsibility as it is about agenda setting. This highlights the dislocation between the loci for authority and accountability despite the Global Fund's growing authoritative territorial claims.

Food recipes developed from commonly consumed local traditional foods were packaged as 100 g powdered soup products at an average cost per 100 g of 0.26 USD. This randomised controlled feeding trial involved 120 eligible pregnant women aged between 21-35 years (median gestational age = 18 weeks) assigned to intervention (100 g/d FMM soup + daily diet) or control (100 g/d local commercially sold soup + daily diet, placebo) group from enrolment to term (average, 20 weeks of feeding). Maternal energy uptake increased from 6.67 MJ/day (66% of DRI) to 7.96 MJ (79% of DRI) in the intervention group compared to 7.11 MJ (70% of DRI) in the control group. Intake of FMM + daily diet also provided 19.51 g/d additional protein and total intake of 1085 of DRI. Additional Fe of 11.78 mg/d increased intake from 7.91 mg/d (27% of DRI) to 19.51 mg/d (73% of DRI). Calcium intake increased to 52% of DRI, Mg from 79% to 118% of DRI, Cu to 85% of DRI and Zn from 74 to 103% of DRI. Folate intake increased from 33 to 56% of DRI, B12 to 118% of DRI, vitamin C to 49% of DRI and vitamin A to 164% of DRI. Haematological changes from baseline to the end of the 3rd trimester included a non-significant increase in red cell mass in the intervention group. Significant improvements in Hb levels were observed in both groups. The intervention group showed marked improvements in mean cardiovascular volume and transferring saturation compared to the control group. Serum Fe increased from 11.18 to 11.68 μmol/L in the intervention group but dropped from 12.21 to 10.78 μmol/L in the control group. Average pregnancy weight gain in the intervention group was 11.5 kg compared to 10.40 kg in control subjects. Average birth weight in the intervention group was 3.02 kg compared to 2.71 kg in controls. Incidence of low birth weight (LBW) was 8.8% in the intervention group and 12.5% in the whole group compared to 16.2% LBW in the control group.

Breast cancer is the most common female cancer in Scotland, in common with many Western countries. This thesis aimed to analyse changes in the incidence and molecular epidemiology of breast cancer in Scotland. Part 1 concentrated on epidemiological research, with data derived from various agencies, and Part 2 on a laboratory project aimed at looking at changes in the molecular profile of breast cancers in two cohorts of patients in Glasgow. The period between 1987 and 1994 in which coverage of the country by the breast screening programme gradually increased was expected to raise incidence rates as seen in studies in Scandinavia and elsewhere; after 1994, incidence rates should have returned to normal in women aged 55-64, with incidence in women aged 50-54 remaining slightly above pre-existing rates. An observed/expected analysis of breast cancer incidence rates after 1994 was performed; this showed a 58% increase in rates in women aged 50-54 above that which would have been expected had the trends continued as expected in the absence of screening. In 55-59 year olds and 60-64 year olds there were 42% and 40% increases, respectively, above expected rates. Reproductive risk factors such as low parity and late age at first pregnancy are important risk factors in breast cancer. Reproductive risk factors are likely to affect the ‘birth-cohort’ incidence of breast cancer but the temporal effects of breast screening make this difficult to interpret. Breast cancer incidence in Scotland by year of birth was examined using a Lexis diagram. In women aged 50-54, 55-59 and 60-64, breast cancer incidence rates increased by birth cohort during the presence of the prevalent round of screening, a finding which is likely to have been due to detection of large numbers of asymptomatic tumours. However, in women who were offered screening after the prevalent round, incidence continued to rise with successive birth year, suggesting a contribution from risk factors. This is the first study of birth cohort incidence of breast cancer and its relation to screening (published in Breast Cancer Research and Treatment). The contribution of screening and risk factors to breast cancer incidence in Scotland was also assessed. A small rise in screening uptake between 1990 and 2001 and an increase in standardised detection ratio may indicate that screening improvements could be contributing to increasing incidence. The number of first pregnancies to women in Scotland aged 35-59 has risen from several hundred in 1976 to 2000 in 2001. A plot of completed family size in Scotland against maternal birth year shows that a steadily declining trend has been developing since the 1935 birth cohort. Based on data from the Scottish Health Surveys, the percentage of women with a BMI of over 25 has increased from 47.2% to 57.3% between 1995 and 2003. Mean BMI in women has increased from 25.7 to 26.9 over the same period. It is likely that the observed changes have contributed to changes in breast cancer incidence in Scotland. Using prescription and population data, the prevalence of HRT use in women aged 40-64 in Scotland was estimated; this estimated prevalence has increased from 13.8% in 1993 to 17% in 2001. It is difficult to know if this small increase in prevalence of HRT could have influenced breast cancer epidemiology. A study of breast cancer incidence by deprivation quintile showed that breast cancer incidence between 1991 and 2000 rose in all quintiles. Interaction analysis suggested that breast cancer incidence is rising to the same extent in deprived and affluent women. The risk factor analyses above were also applied to women of different socioeconomic standing (the results were published in Breast Cancer Research and Treatment). A laboratory project was carried out to assess whether increasing survival from breast cancer could be a result of changing molecular epidemiology. This project was an comparison of the prevalence of breast cancers which were ER, PR and Her2 positive and of different grades in two cohorts of Glasgow patients, from 1984-86 and 1996-1997. The application of current molecular techniques to stored tissue aimed to improve the quality of data compared to previous studies based on clinical databases using heterogeneous techniques. There were significant differences in grade distribution of tumours in the two cohorts (p=0.009) with fewer grade 1 and more grade 3 tumours in the second cohort. Further study showed the grade difference to be exerted by the tumours in screened women in the second cohort with there being no difference in grade between symptomatic patients in the two groups. 64.2% of the tumours in cohort 1 and 71.5% of the tumours in cohort 2 were ER positive (p=0.042); this is also likely to be a clinically significant difference. The difference between the cohorts appeared to be exerted by high percentage of screen-detected tumours in cohort 2 being ER positive; however this finding still supports a theory of changing biology. 44.9% of the tumours in cohort 1 and 49.9% of tumours in cohort 2 were PR positive (p=0.181). 21.5% of tumours in cohort 1 and 20.6% of tumours in cohort 2 were Her-2 positive; this was not a significant difference. An increase in ER positivity was seen in all age groups in the study, though multivariate analysis did suggest a contribution from a higher number of women over 60 in the more recent cohort. Kaplan-Meier analysis showed survival to be higher in the second cohort than the first. There was a significant difference in survival between ER positive and negative patients. Cox’s regression was performed; as expected this showed a multifactorial contribution to increases in survival in these cohorts rather than it being entirely due to changes in ER status. However the changes in ER status shown in a population of Glasgow patients over time may mean that the results of clinical trials carried out in many years ago need to be interpreted with caution when applying them to the women of today. The results of this project were published in the British Journal of Cancer. Overall, the epidemiological studies within this thesis shed an important new light on the factors contributing to breast cancer incidence in Scotland, with a major finding being a significant association between birth cohort and breast cancer incidence suggesting a significant impact being made by reproductive risk factors. This hypothesis is supported by the analysis of risk factor trends in Scotland undertaken within the thesis. The laboratory study has shown a significant lowering in grade and increase in ER positive status of tumours in a cohort of Glasgow women over time; while the changes are statistically explainable by known effects of a breast screening programme on tumour detection they could still represent a true change in biology. The results of all the studies contained in the thesis could have significant implications for future health service planning.