Despite the effectiveness of brief lifestyle interventions delivered in primary health care (PHC), implementation in routine practice remains suboptimal. Previous research suggests that there are many barriers to PHC clinicians addressing lifestyle risk factors, however few studies have identified the importance of various factors and how they shape practices. This thesis aimed to develop and describe a theoretical model to explain the lifestyle risk factor management practices of PHC clinicians and to identify critical leverage points for intervention. The study analysed data collected as part of a larger feasibility project of risk factor management in three community health teams in NSW, Australia, involving 48 PHC providers working outside of general practice. Grounded theory principles were used to inductively develop a model, involving three main stages of analysis: 1) an initial model was developed based on quantitative analysis of clinician survey and audit data, and qualitative analysis of a purposeful sample of participant interviews (n=18) and journal notes; 2) the model was then refined through additional qualitative analysis of participant interviews (n=30) and journal notes; and 3) the usefulness of the model was examined through a mixed methods and case study analysis. The model suggests that implementation of lifestyle risk factor management reflects clinicians??? beliefs about commitment and capacity. Commitment represents the priority placed on risk factor management and reflects beliefs about role congruence, client receptiveness and the likely impact of intervening. Capacity beliefs reflect clinician views about self efficacy, role support and the fit between risk factor management and ways of working. The model suggests that clinicians formulate different intervention expectations based on these beliefs and their philosophical views about appropriate ways to intervene. These expectations then provide a cognitive framework guiding their risk factor management practices. Finally, clinicians??? appraisal of the overall benefits and costs of addressing lifestyle issues acts to positively reinforce or to diminish their commitment to implementing these practices. The model extends previous research by outlining a process by which clinicians??? perceptions shape implementation of lifestyle risk factor management in routine practice. This provides new insights to inform the development of effective strategies to improve such practices.

Globally, the prevalence of diabetes has seen a significant increase, rising from 211 million in 1990 (3.96% of the global population at that time) to 476 million in 2017 (6.31% of the global population in 2017). Extensive research has been conducted to study the causes of diabetes from a data-driven approach, leading to the development of prospective models for predicting future diabetes risks. These studies have highlighted the strong correlation between diabetes and various biomarker factors, such as BMI, age, and certain blood test measures. However, there is a lack of research that focuses on building prospective models to predict future diabetes risks based on lifestyle factors. Therefore, this thesis aims to employ popular machine learning methods to investigate whether it is possible to predict future diabetes using prospective models that incorporate self-controllable lifestyle factors. Our analysis produced remarkable results, with the biomarker model achieving an average validation AUC score of 0.78, while the lifestyle model reached 0.70. Notably, lifestyle features demonstrate a greater predictive capacity for short-term new-onset diabetes when compared to the long-term endpoint. The biomarker model identified visceral fat as the most significant risk factor, whereas income level and employment emerged as the top risk factors in the lifestyle model. This thesis represents an innovative approach to diabetes prediction by leveraging lifestyle factors, providing valuable data-driven insights into the root causes of diabetes. It addresses a critical research gap by highlighting the significant role of lifestyle factors in predicting the future onset of diabetes, particularly within the context of parametric modeling. / Thesis / Master of Science (MSc)

BACKGROUND: Pregnancy-induced hypertension (PIH) and gestational diabetes mellitus (GDM) are among the leading causes of disability and death for women and their babies. Identifying risk factors for these pregnancy-related complications is essential to their prevention. Studies identifying preventive models for PIH and GDM are few. OBJECTIVES: This study was designed to evaluate lifestyle and personal predictors of PIH and GDM in a cohort of nearly 20,000 pregnant women. METHODS: The exposure data for the study were derived from a combination of a telephone interview and a questionnaire completed approximately 2 months after conception during the period from 1984 to 1987. The initial questionnaires asked for information on three periods: 3 months before conception, at conception, and 2 months after conception. Subjects included 19,312 women, aged 18-<45 years, who did not have excessive intakes of alcohol or food, were neither underweight (BMI >18.5) nor extremely overweight (BMI <40), and did not use illegal drugs during the first trimester of pregnancy. Outcome data on the mother and baby were collected approximately one year after the expected data of delivery. Logistic regression models were used to estimate the odds ratios (OR), and 95% confidence intervals (CI), as well as receiver operating characteristic (ROC) curves predicting PIH and GDM. Akaike Information Criteria (AIC) were used to select the best predictors of these two outcomes. Factors found not to affect PIH or GDM (based on a two-unit decrease in the AIC) were excluded from the final models. RESULTS: Based on the outcome data collected, there were 204 PIH cases, 358 GDM cases, and 538 who had PIH and/or GDM. After selecting the outcome predictors using AIC values, we identified three predictive models—one each for PIH, GDM, and either PIH or GDM. Factors found to predict PIH included age, previous hypertension or type 1 or 2 diabetes, pre-pregnancy BMI, parity, exercise, red meat consumption, margarine consumption, cigarette smoking, and weight change at 2 months. The final AIC value for PIH was 2084.12 and the AUC value was 0.76. GDM was predicted by age, previous GDM (in an earlier pregnancy), pre-pregnant BMI, height, exercise, race, dairy consumption, and cigarette smoking, with an AIC value of 3288.74 and an AUC value of 0.70. The combined model (predicting either PIH or GDM) was best predicted by age, history of GDM in a previous pregnancy, pre-pregnant BMI, previous history of hypertension, height, exercise, dairy consumption, red meat consumption, parity numbers, cigarette smoking, and weight change at 2 months with an AIC value of 3288.74 and an AUC value of 0.71. CONCLUSIONS: In these analyses, separate models predicting PIH and GDM were better than a combined model predicting PIH or GDM. These final models indicate that we can reasonably identify women who are at increased risk for adverse maternal outcomes associated with hypertensive disorders or diabetes during pregnancy.

Introduction: Les personnes vivant avec le VIH (PVVIH) présentent des risques accrus d’ostéoporose et de lipodystrophie. Peu d’études se sont penchées sur l’association entre les déterminants du style de vie, le risque d’ostéoporose et le risque de lipodystrophie chez les PVVIH. Objectifs: L’objectif primaire était d’évaluer l’association entre l’ostéoporose, la lipodystrophie ainsi que différents déterminants du style de vie chez les PVVIH. Méthodologie: Tous les participants de la Cohorte canadienne VIH et vieillissement (CCVV) avec des données sur la densité minérale osseuse (DMO), mesurée par absortiométrie biphotonique à rayons X (DXA), ont été inclus dans cette étude transversale. Les déterminants du style de vie d’intérêt étaient : le revenu annuel, le niveau d’éducation, l’exercice physique ainsi que les consommations d’alcool, de tabac et de drogues illicites. Les covariables mesurées incluaient l’historique complet de la médication antirétrovirale, les comorbidités, les co-infections, la charge virale, le compte de CD4+ au recrutement et le compte de CD4+ nadir. L’ostéoporose a été définie par un score T de -2.5 ou moins. La lipodystrophie, évaluée par la composition corporelle via DXA, a été définie par un fat mass ratio (rapport des pourcentages de gras entre le tronc et les membres inférieurs) supérieur à 1.33 pour les femmes et 1.96 pour les hommes. Les rapports des cotes et les intervalles de confiance à 95% (IC95%) au recrutement ont été estimés en utilisant des régressions logistiques multivariées. Résultats: Nous avons inclus 547 PVVIH (âge médian 55 ans, 88% d’hommes) et 97 contrôles séronégatifs au VIH (âge médian 54 ans, 54% d’hommes). L’ostéoporose était présente chez 13% des PVVIH et 6% des contrôles (OR 2.21, IC 95% [0.96 – 6.06]). La lipodystrophie était présente chez 138 (28.3%, IC 95% 24.3 – 32.3%) des 487 PVVIH avec des données sur la disposition du gras corporel. Aucun des déterminants du style de vie était associé à l’ostéoporose ou à la lipodystrophie. Par contre, les covariables associées à un risque accru d’ostéoporose étaient l’âge avancé, un indice de masse corporelle (IMC) réduit et la co-infection à l’hépatite C. Les covariables associées au risque accru de lipodystrophie étaient l’âge avancé, l’hypertension, l’exposition prolongée aux antirétroviraux, ainsi que les expositions prolongées aux inhibiteurs nucléosidiques de la transcriptase inverse (INTI) et aux inhibiteurs de l’intégrase (INI). Conclusion: Aucune association n’a été décelée entre les déterminants du style de vie étudiés et l’ostéoporose ou la lipodystrophie. / Background: As a consequence of ART, people living with HIV (PLWH) are at higher risk for osteoporosis and lipodystrophy. However, the risk may also be influenced by lifestyle factors, but few studies have explored the association between modifiable lifestyle factors and the risk of osteoporosis or lipodystrophy in the PLWH population. Objectives: Our primary objective was to evaluate the lifestyle factors in relation to the risks of osteoporosis and lipodystrophy in a PLWH-based cohort. Methods: We conducted a cross-sectional analysis of data from the Canadian HIV and Aging Cohort Study (CHACS). We included all participants with available bone mineral density T-scores, which were measured by dual-energy X-ray absorptiometry (DXA) scans. Lifestyle risk factors of interest included annual income, education level, alcohol intake, tobacco use, illicit drug use and physical exercise. Other covariates considered were full antiretroviral medication history, medical comorbidities, coinfections, viral load, nadir CD4+ and current CD4+ count. Osteoporosis was defined by a T-score of -2.5 or lower at any of the measured sites. Lipodystrophy was assessed on whole body DXA and defined as a fat mass ratio (the ratio between trunk and lower limbs fat mass) greater than 1.33 for women and 1.96 for men. Baseline prevalence odds ratios (ORs) and 95% confidence intervals (95% CIs) were estimated by multivariate logistic regressions. Results: We included 547 PLWH (median age 55 years, 88% males) and 97 HIV-uninfected controls (median age 57 years, 64% males). Osteoporosis was present in 13.0% of PLWH (95% CI 10.2 – 15.8%) and 6% of controls (95% CI 1.4 – 11.0%); the OR of osteoporosis for HIV seropositivity was 2.21 (95% CI [0.96 – 6.06]). Lipodystrophy was found in 138 (28.3%, 95% CI 24.3 – 32.3%) of the 487 PLWH for which a fat mass ratio (FMR) was available. None of the lifestyle factors of interest were associated with osteoporosis or lipodystrophy. However, covariates associated with an increased risk of osteoporosis were increasing age, lower body mass index (BMI) and hepatitis C coinfection. Covariates associated with an increased risk of lipodystrophy were older age, hypertension, longer antiretroviral duration, and longer exposure to nucleoside reverse transcriptase inhibitors (NRTIs) and integrase strand inhibitors (INSTIs). Conclusion: No association was found between any of the lifestyle factors of interest and osteoporosis or lipodystrophy.