Electronic Theses and Dissertations (PhDs)
Permanent URI for this collectionhttps://hdl.handle.net/10539/37932
Browse
Search Results
Item From the onset: Impact of Nutrition and Lifestyle during the Preconception period in Urban South African Young Adults(University of the Witwatersrand, Johannesburg, 2023-03) Mukoma, Gudani Goodman; Norris, Shane A.Background: In South Africa, 22% of adolescents are overweight or obese, the onset of tobacco smoking is shown to peak between the ages of 15 and 22, 1 in 3 adolescents watch more than 3 hours of television per day, and nearly half of all adults are insufficiently active. Physical inactivity, poor diet, risky alcohol use, illicit drug use are among the behavioural risk factors associated with obesity and mental health problems, all of which have morbidity and mortality implications for adult health. Risks in later life include premature death, long-term disability, childbirth complications, gestational diabetes, diabetes and cardiovascular diseases. However, there is data paucity showing the personal, social, and environmental factors that are determinants of health, especially diet, physical activity (PA), obesity, and associated NCDs in South African adolescents and young adults. Aim: To investigate the individual, household, and environmental factors that influence adolescents' dietary and physical activity habits and to identify ways in which these factors can be leveraged for interventions to better ensure the health of future generations, especially during the crucial preconception years. Methods: This thesis was purposely designed to use a sequential mixed-methods approach that integrates quantitative (Chapter 3 paper 1: cross-sectional and Chapter 6 paper 4: longitudinal) and qualitative (Chapter 4 paper 2: longitudinal and Chapter 5 paper 3: cross-sectional) analyses in order to meet the four specific objectives of my research. The methods selected for this series of investigations were primarily influenced by the substantive research questions that arose, as opposed to methodological and epistemological concerns alone. I utilized three pre-existing data sources, including the "Birth-to-Twenty Cohort," the "Determinants of Type 2 Diabetes Mellitus (T2D)" study, and the "Soweto household enumeration survey." I have gathered new prospective data that is quantitatively and qualitatively longitudinal and cross-sectional. Results: The findings of this thesis in the context of Soweto show that the relationship between dietary patterns and nutritional status (BMI) is independent of socioeconomic status (SES). Adolescents and young adults face a variety of intersecting barriers resulting from personal preferences and their living conditions, which influence their dietary and physical activity habits while occurring at the time; this is important to consider when designing interventions to promote healthy behaviour change. Unexpected stressors, such as the outbreak of the COVID-19 pandemic, contributed to exacerbating adolescents' and young adults' poor health conditions, and as a result, the prevalence of poor nutrition intake, a lack of physical activity, and mental health issues increased. Although the nutrient patterns of adolescents and adults were comparable over time, their associations with BMI were not. The associations with BMI of the "plant-driven nutrient pattern," "fat-driven nutrient pattern," and "animal-driven nutrient pattern" revealed sex differences. Conclusion: Adolescent diet and lifestyle continue to be important research areas in the intent to enhance preconception health and reducing maternal and infant mortality.Item Nutrition of ageing black South African women and correlates with anthropometry and cardiometabolic outcomes(University of the Witwatersrand, Johannesburg, 2023-09) Kankwende, Caroline Belinda Tsitsi; Gradidge, Philippe; Norris, Shane; Chikowore, TinasheBackground: Obesity is most prevalent among black women who reside in urban areas in South Africa yet the nutrient patterns of this cohort of women has never been investigated, nor have correlates of body composition indices such as adiposity and body mass index (BMI). These body composition indices are important to analyse as they have been shown to be positively associated with hypertension which is prevalent in this cohort of women. Aim: There were three main goals: 1) To determine the baseline nutrient patterns of middle-aged black South African women residing in Soweto and correlates to body composition indices 2) To evaluate the longitudinal association of nutrient patterns with adiposity in a cohort of middle-aged black South African women over a period of 5.5- years 3) To elucidate the longitudinal associations of nutrient patterns and blood pressure and to explore whether this is an indirect effect mediated by body mass index (BMI) using structural equation modelling Methods: A longitudinal study of children and their families, originally called the Birth to Twenty Plus (Bt20+) cohort and now referred to as the Middle-aged Soweto Cohort (MASC), was used to as the original dataset for this thesis. This study also drew on another embedded study, the Study of Women Entering and in Endocrine Transition (SWEET) study of older women transitioning through menopause. Data on (i) dietary information; (ii) body composition and anthropometry measurements; (iii) blood pressure; (iv) lifestyle behaviours (physical activity, tobacco use, and alcohol use); (v) psychosocial factors; (vi) socioeconomic status; and (vii) educational status were used. A total of 498 Women aged between 40 and 60 years old were included in the study. Principle component analysis (PCA) was applied on the dietary data both at baseline and at 5- years follow-up. This was conducted to extract nutrient patterns from 25 nutrients derived from the food frequency questionnaire (FFQ) and the resulting nutrient patterns are detailed in results chapter 3 (nutrient patterns derived from the baseline FFQ) and results chapter 4 (comparison between both baseline and follow-up nutrient patterns from the FFQ). Simple and complex body composition were recorded for each participant with complex body measurements taken using (DXA). Chapter 3 details the results of the 3 baseline nutrient patterns and correlates with body composition parameters. Using generalized estimating equations, associations between both baseline and 5- years follow-up nutrient patterns and adiposity were evaluated. The results are discussed in results chapter 4. Lastly, the results chapter 5 examined associations between both baseline and follow-up nutrient patterns and blood pressure were examined and furthermore, investigated whether BMI mediates the relationship between repeated measures of nutrient patterns and blood pressure. Results: The majority of the research participants (88%) were classified as individuals having obese status defined by their BMI. The fat mass index (FMI), lean mass index (LMI), gynoid fat, hip and waist circumference, and visceral and subcutaneous adipose tissue (VAT and SAT respectively) measurements were all substantially larger in the group with predominantly individuals having overweight and obese classification compared to woman in the lean group (p <0.001). Protein consumption was greater in the group with individuals having overweight/obese classification, while fat and carbohydrate consumption were matched. At baseline, the "Plant driven nutrient pattern," characterized by higher factor loadings of plant protein, starch, and B vitamins, explained 25% of the total nutritional variance; the "Animal protein driven nutrient pattern," characterized by animal protein and saturated fat, explained 23% of the variance; and the third pattern was the "Vitamin C, sugar and potassium driven nutrient pattern," which had higher factor loadings of vitamin C, sugar and potassium. At baseline, increased consumption of the animal protein driven nutrient pattern resulted in a 1.19 kg/m2 (p = 0.002) increase in BMI, 10.17 cm2 for VAT, 24.43 cm2 for SAT, 0.01 (p = 0.009) increase for VAT/SAT ratio, 0.69 kg/m2 (p = 0.005) increase for FMI, and 0.48 kg/m2 (p = 0.002) increase for LMI. Furthermore at baseline, statistically significant associations were found for the animal protein driven nutrient pattern with all body composition indicators. Subcutaneous adipose tissue increased in the presence of a plant-driven nutrition pattern (p = 0.045). At 5-year follow-up, although the value of the factor loadings of the individual nutrients changed between baseline and follow-up, the nutrients with the highest loadings for each principal component (PC) did not change therefore the overall nutrient patterns remained the same. Only DXA-derived measurements of fat mass, FMI, VAT, and gynoid fat mass (FM) increased with time, while lean mass considerably reduced. Repeated measures of the animal protein driven nutrient pattern was associated with significant increases in FMI, LMI and VAT and repeated measures of the vitamin C, sugar, and potassium driven nutrient pattern was significantly associated with an increase in FMI and LMI. For the purposes of this study, repeated measures of animal-driven nutrient patterns were shown to be significantly related with repeated measures of systolic blood pressure (SBP) only. When structural equation modelling (SEM) was applied, only significant relationships were observed between age and SBP. This relationship was not mediated by BMI but may involve other factors that were not included in this analysis. Conclusions: This thesis explored the nutrient patterns linked to obesity and cardiometabolic complications, namely blood pressure, in a cohort of black middle-aged African females. It has been previously demonstrated that this cohort has been has a high prevalence of obesity. According to literature reviews, programs focusing on nutritional and behavioural changes could aid African women in their fight against the obesity and hypertension epidemic that we are facing today. The animal-driven nutrient pattern was found to be substantially associated with increases in body fat in this cohort at baseline. At 5-year follow-up, the nutrient patterns remained the same and repeated measurements of the vitamin C, sugar, and potassium-driven nutrient pattern were associated with significant increases in FMI and LMI and the animal-driven nutrient pattern remained significantly associated with LMI, FMI and VAT, a measure of visceral obesity which is a major risk factor for cardiometabolic conditions. This is problematic in a population that consists predominantly of individuals that are classified as having an obese and overweight status. As a result of a higher BMI, a greater likelihood of developing cardiometabolic multimorbidity exists which is defined as the co-occurrence of two or three cardiometabolic conditions. This may result in reduced quality of life and an increased burden on the already overstretched healthcare system in South Africa. Furthermore, this study found that only the animal protein driven nutrient pattern had a significant relationship with SBP which was significant. When SEM was applied, BMI did not mediate the relationship between blood pressure and any of the nutrient patterns. No other noteworthy direct relationships between blood pressure and the other nutrient patterns were found. Researchers can apply the findings of this study to improve nutritional policies and guidelines aimed at combating not just obesity, but high blood pressure among black women in Sub-Saharan Africa. It is necessary to conduct further extensive research to verify these findings.Item Role of novel biomarkers in predicting chronic kidney disease progression among black patients attending a tertiary hospital in Johannesburg, South Africa(University of the Witwatersrand, Johannesburg, 2023) Meremo, Alfred Jackson; Naicker, Saraladevi; Duarte, Raquel; Paget, Graham; Dickens, CarolineBackground: Chronic kidney disease (CKD) is a leading health issue and its magnitude has been increasing globally; where the developing countries are the most affected and they are the least equipped to deal with its associated consequences. Chronic kidney disease can rapidly and quietly progress to late CKD stages in impoverished environments. Early recognition of patients who are likely to develop end-stage kidney disease (ESKD) is important. Methodology: A prospective longitudinal study was conducted on CKD patients of black ethnicity attending at the Charlotte Maxeke Johannesburg Academic Hospital (CMJAH) renal outpatient clinic in South Africa, as from September 2019 to March 2022. Patients provided blood and urine samples for investigations in the laboratory at study enrolment (0) and at the 24 months follow up. The concentrations of the transforming growth factor isoforms [(TGF)-β1, TGF-β2 and TGF-β3) were determined in serum and urine at baseline using the Human TGF-β duoset ELISA. Data were descriptively and inferentially processed by the REDcap and analyzed using STATA version 17 and multivariable logistic regression analysis was applied to find out the predictors of CKD progression. Results: A total of 312 patients were recruited into the study; the median age was 58 (IQR 46 -67) years and 162 (51.9 %) were male. Hypertension was present in majority (96.7 %) of the patients. Diabetes mellitus was present in 38.7 % of patients and 38.1 % of the study patients had both hypertension and diabetes mellitus. A total of 297 (95.2%) patients completed the study. Death was reported in 5 (1.6%) patients and 10 (3.2%) of patients were lost to follow up. The prevalence of CKD progression was 49.5%, 33% had CKD remission and 17.5% had CKD regression while the prevalence of CKD progression by change in uPCR > 30% was 51.9%. Almost half (47.8 %) had a sustained decline in eGFR of > 4 ml/min/1.73 m2 /year or more, 35.0% of the patients moved to a more severe stage of CKD and 19.9% had more than 30% 6 decline in eGFR in two years. For patients with CKD progression, 54.9% patients were men and at baseline, their median age was 59 (46 - 67) years, urine protein creatinine ratio (uPCR) increased at 0.039 (0.015-0.085) g/mmol, eGFR was 37 (32 -51) mL/min/1.73 m2; the median serum TGF-β1 was 21210 (15915 – 25745) ng/L and the median urine TGF-β3 was 17.5 (5.4 –76.2) ng/L. For those who had CKD progression, hypertension was present in the majority (95.2%) of the patients. Diabetes mellitus was present in 59 (40.1%) patients and 58 (39.5%) patients had both hypertension and diabetes mellitus; 48.3% had severely increased proteinuria, 45.6% patients had anaemia, 34.0% had hyperuricemia and 17.7% had hypocalcaemia at baseline. For those patients with CKD progression vs those without CKD progression, the baseline median serum TGF-β1 was 21210 (15915 – 25745) ng/L vs 24200 (17570 – 29560) ng/L, the baseline median urine TGF-β3 was 17.5 (5.4 – 76.2) ng/L vs 2.8 (1.8 – 15.3) ng/L; however, baseline serum and urine TGF-β isoforms did not predict progression of CKD on univariate and multivariable analyses. Regarding use of medications among patients with CKD progression, calcium channel blockers (amlodipine) were used by majority (85.2 %) of the patients. Diuretics were used by 63.4% of the patients and 31.7 % of the patients were using insulin. Variables associated with CKD progression after multivariable logistic regression analysis were moderately elevated proteinuria (OR 2.1, 95% CI (1.1 – 3.9), P= 0.019), severely elevated proteinuria (OR 6.1, 95 % CI (3.2 – 11.6), P = 0.001), hyponatraemia (OR 4.5, 95% CI 1.8 - 23.6, P= 0.042), hypocalcaemia (OR 3.8, 95 % CI 1.0 - 14.8, P = 0.047), anaemia (OR 2.1, 95% CI 1.0 - 4.3, P= 0.048), elevated HbA1c (OR 1.8, 95 % CI 1.2 - 2.8, P = 0.007), diabetes mellitus (OR 1.8, 95 % CI 1.9 - 3.6, P = 0.047), current smoking (OR 2.8, 95 % CI 1.9 - 8.6, P = 0.049), medications which were calcium channel blockers (OR 2.07, 95 % CI 1.04 – 4.12, P = 0.038), diuretics (OR 2.35, 95 % CI 1.37 – 4.00, P = 0.002), insulin (OR 1.96, 95 % CI 1.01 – 3.84, P = 0.048) and baseline serum calcium levels (OR 0.06, 95 % CI 0.01 -0.64, P = 0.019). An increase in uPCR > 30% at two years identified most patients with CKD progression; clinicians and nephrologists should utilize change in uPCR > 30% at two years to identify those patients with CKD who are likely to progress more rapidly, who require closer surveillance and monitoring with emphasis on slowing or stopping progression of the CKD. Conclusion: Our study has demonstrated a higher prevalence of CKD progression in a prospective longitudinal study among black patients than that reported in previous studies. CKD progression was associated with current smoking, hyponatremia, hypocalcemia, anaemia, elevated HbA1c, diabetes mellitus, and proteinuria. While patients with CKD progression had lower baseline concentrations of serum TGF-β1 and increased baseline urinary TGF-β3 concentrations, baseline serum and urine TGF-β isoforms did not predict progression of CKD. The roles of the various serum and urine TGF-β isoforms in CKD progression at baseline are still unclear and highlight the importance of further studies to determine their isoform specific effects.Item Predicting in-hospital mortality in heart failure patients using machine learning(University of the Witwatersrand, Johannesburg, 2023-05) Mpanya, Dineo; Ntsinjana, HopewellThe age of onset and causes of heart failure differ between high-income and low-and-middle-income countries (LMIC). Heart failure patients in LMIC also experience a higher mortality rate. Innovative ways that can risk stratify heart failure patients in this region are needed. The aim of this study was to demonstrate the utility of machine learning in predicting all-cause mortality in heart failure patients hospitalised in a tertiary academic centre. Six supervised machine learning algorithms were trained to predict in-hospital all-cause mortality using data from 500 consecutive heart failure patients with a left ventricular ejection fraction (LVEF) less than 50%. The mean age was 55.2 ± 16.8 years. There were 271 (54.2%) males, and the mean LVEF was 29 ± 9.2%. The median duration of hospitalisation was 7 days (interquartile range: 4–11), and it did not differ between patients discharged alive and those who died. After a prediction window of 4 years (interquartile range: 2–6), 84 (16.8%) patients died before discharge from the hospital. The area under the receiver operating characteristic curve was 0.82, 0.78, 0.77, 0.76, 0.75, and 0.62 for random forest, logistic regression, support vector machines (SVM), extreme gradient boosting, multilayer perceptron (MLP), and decision trees, and the accuracy during the test phase was 88, 87, 86, 82, 78, and 76% for random forest, MLP, SVM, extreme gradient boosting, decision trees, and logistic regression. The support vector machines were the best performing algorithm, and furosemide, beta-blockers, spironolactone, early diastolic murmur, and a parasternal heave had a positive coefficient with the target feature, whereas coronary artery disease, potassium, oedema grade, ischaemic cardiomyopathy, and right bundle branch block on electrocardiogram had negative coefficients. Despite a small sample size, supervised machine learning algorithms successfully predicted all-cause mortality with modest accuracy. The SVM model will be externally validated using data from multiple cardiology centres in South Africa before developing a uniquely African risk prediction tool that can potentially transform heart failure management through precision medicine.