Type 1 Diabetes is a dangerous and life-long disease for which its prevalence is global. Research has shown that tight glycaemic control of this disease significantly reduces the risks of developing several life threatening diabetic complications. The Ets-Insulin-Bolus Calculator (EIBC), inspired by the Ets concept (Equivalent Teaspoon Sugar), was primarily designed to assist type I diabetics in improving their blood glucose control. The EIBC has shown to improve the average blood glucose level of type 1 diabetics. The need for this study however is to determine whether the ET!3C promotes tighter glycaemic control for type 1 diabetics based on a more-in-depth numerical analysis. With the use of the latest technology in blood glucose monitoring, the CGMS from Medtronic, mathematical models expressing and rating blood glucose control have been proposed and derived in this study. A clinical trial with type 1 diabetics has also been conducted. The use of the models together with the clinical trial results have shown that the EIBC does in fact promote tighter glycaemic control for type 1 diabetics. / Thesis (M.Ing. (Mechanical Engineering))--North-West University, Potchefstroom Campus, 2007.

Current methods of energy restriction are not successful for achieving long-term weight loss and maintenance for the majority of individuals. As a result, the prevalence of obesity and obesity related diseases continue to increase. This calls for the development of novel lifestyle interventions to combat the obesity epidemic. Hunger has been highlighted as a major factor influencing the long-term success of weight management methods and therefore how a given dietary intervention affects the appetite regulatory system may dictate the success of the diet by augmenting long-term adherence. In addition, the effect of a given dietary intervention on exercise may determine its suitability for exercising individuals and may influence the energy deficit that can be achieved by the diet. This thesis investigated the acute effects of two novel methods of dietary restriction; breakfast omission and severe energy restriction. The main aims for this thesis were to determine the effect of these methods of energy restriction on ad-libitum energy intake, subjective appetite sensations, and peripheral concentrations of hormones involved in appetite regulation. In addition, this thesis also investigated the effects of these methods of energy restriction on metabolism and glycaemic control at rest, and performance and perceived exertion during exercise. This work found that moderate and severe energy deficits induced by breakfast omission and 24 h of severe energy restriction, respectively, resulted in either no (Chapter VIII) or partial (Chapters IV and VII) energy intake compensation over the subsequent 24-48 h. Subjective appetite was increased during (Chapters IV, V, VII and VIII) and shortly after (Chapter VII) energy restriction, but this effect was transient and was offset after an ad-libitum (Chapters IV and VII) or standardised (Chapters V and VIII) meal. In addition, none of the work presented in this thesis demonstrated an appetite hormone response to energy restriction that was indicative of compensatory eating behaviour. Compared to breakfast omission, breakfast consumption resulted in an increased in resting energy expenditure and carbohydrate oxidation, with a concurrent reduction in fat oxidation during the morning. However, there were no differences after lunch (Chapter V). In response to a standardised breakfast, resting energy expenditure was suppressed (Chapter VII) or not different (Chapter VIII) the following morning, after 24 h severe energy restriction compared to energy balance. Plasma NEFA and fat oxidation was greater, carbohydrate oxidation was reduced, and postprandial insulin sensitivity was impaired in the after 24 h severe energy restriction (Chapter VI, VII and VIII). In Chapter IV, omission of breakfast in the morning was shown to reduce exercise performance in evening, even after provision of an ad-libitum lunch 4 h before. However, there was no difference in perception of effort during steady state exercise, independent of breakfast consumption or omission in the morning (Chapters IV and V). Collectively, breakfast omission and 24 h severe energy restriction reduce energy intake and promote an appetite regulatory response conducive to maintenance of a negative energy balance. Chronic intervention studies are now required to confirm whether these effects persist after long-term practice of these dietary interventions.

Zusammenfassung Arbeitshypothese und Fragestellung: Kontinuierliche Glukosemonitoringsysteme erlauben die Messung über einen Zeitraum von mehreren Tagen und geben so eine zuverlässige Erfassung der mittleren interstitiellen Glukosebelastung und ihrer Dynamik. Eigene Voruntersuchungen erbrachten dabei relativ stabile individuelle 'Glukosemuster' im Tagesverlauf, die sich bei stabiler antiglykämischer Therapie auch Jahre später noch nachweisen lassen. Gleichzeitig demonstrierten prospektive Studien nur eine geringe Assoziation zwischen dem HbA1c als Langzeitmarker der durchschnittlichen Glukosebelastung und kardiovaskulären Ereignisse bzw. Mortalität. Vorrangiges Ziel der Arbeit ist es deshalb zu testen, ob weitere Parameter der Glykämiekontrolle mit zukünftigen Mikroangiopathien, Mortalität und kardiovaskulären Ereignissen assoziiert sind. Methoden: Es handelt sich um eine retrospektive Kohortenstudie mit insgesamt 315 eingeschlossenen Probanden. Die Studienpopulation basiert auf drei Forschungsprojekten, welche am Studienzentrum für metabolisch-vaskuläre Medizin, GWT TU-Dresden GmbH in Dresden im Zeitraum 2010-2013 durchgeführt wurden. Die gesammelten anamnestischen, klinischen und klinisch-labordiagnostischen Parameter sowie die Blutzuckermessungen des CGM wurden aus den Archiven zu einer Datenbank zusammengefasst und zur Erstellung von Fallgruppen verwendet. Zusätzlich wurden durch telefonische Interviews und routinemäßig erhobene Daten im Rahmen von Folgeuntersuchungen am Studienzentrum klar definierte kardiovaskuläre Ereignisse ermittelt, die sich nach Durchführung des CGM manifestiert haben. Resultate: Die Ergebnisse zeigen in einer univariaten Cox-Regression in der Gesamtkohorte eine Signifikanz (p < 0.001) der glukosebezogenen Parameter interstitielle Glukosefluktuation, HbA1c und Diabetesdauer auf die Ausprägung von kardiovaskulären Endpunkten. In einer multivariaten Cox-Regression mit Einbezug der Glukoseparameter interstitielle Glukosefluktuation, HbA1c und Diabetesdauer in der Subgruppe der Diabetiker zeigt sich jedoch nur die Diabetesdauer (p < 0.05) mit signifikanter Assoziation auf die Ausprägung von kardiovaskulären Endpunkten. Die Parameter interstitielle Glukosefluktuation (p=0.44) und HbA1c (p=0.35) verlieren hier ihre signifikante Assoziation. Dies spiegelt sich auch in den ROC-Analysen wieder. Diabetesdauer weist einen AUC (Area under the curve) von 0.66 in den ROC-Analysen auf. Bei einem theoretisch optimalen Cut-Off von 9 Jahren kann hier eine Sensitivität von 61% und Spezifität von 70% beobachtet werden. Innerhalb der glukosebezogenen Parameter zeigt sich die Diabetesdauer als stärkster Klassifikator für die Prognose von kardiovaskulären Erkrankungen. HbA1c (p<0.05) zeigt in einer multivariaten Cox-Regression in der Gesamtpopulation eine stärkere Assoziation auf die Ausprägung von kardiovaskulären Erkrankungen als die interstitielle Glukosefluktuation (p=0.25). Jedoch demonstrieren beide ähnliche Ergebnisse in der ROC-Analyse. HbA1c weist bei einem Cut-Off von >=6.8% eine Sensitivität von 37% und eine Spezifität von 81% auf. Die AUC des Parameters HbA1c beträgt hier in der ROC-Analyse 0.59. Für die interstitielle Glukosefluktuation konnte ein optimaler Cut-Off von 1.6mmol/L errechnet werden und weist hier eine Sensitivität von 48% und eine Spezifität von 66% auf. Die AUC der interstitiellen Glukosevariabilität hat einen Wert von 0.57. Weiterhin erscheint der systolische Blutdruck in einer multivariaten Cox-Regression, in der alle kardiovaskulär relevanten Parameter aus der univariaten Analyse eingeschlossen wurden, als signifikant (p<0.05) auf die Ausprägung von kardiovaskulären Ereignissen. Bei einem Cut-off von 152mmHg systolischen Blutdruck wurden in der ROC-Analyse eine Sensitivität von 43% und Spezifität von 72% ermittelt. Die AUC des systolischen Blutdruckes ergab hier einen Wert von 0.63. Schlussfolgerung: Anders als in den vergangenen Studien zu Blutglukosevariabilität wurde neben einer möglichen Assoziation der interstitiellen Glukosevariabilität hier zusätzlich der Versuch unternommen einen definierten Grenzwert von 1.6mmol/L zu gestalten ab dem ein erhöhtes kardiovaskuläres Risiko wahrscheinlich ist. Dieser Grenzwert soll helfen der aktuellen uneinheitlichen Studienlage bezüglich der interstitiellen Glukosevariabilität zu vereinheitlichen, um dauerhaft vergleichbare Ergebnisse in diesem Bereich zu liefern. Es zeigt sich die interstitielle Glukosevariabilität als schwacher Prädiktor für die Ausprägung von kardiovaskulären Erkrankungen Jedoch zeigt der etablierte HbA1c- Wert eine ebenfalls schwache prädiktive Aussagekraft, welche mit der der interstitiellen Glukosevariabilität vergleichbar ist. Weitere Studien, vor allem prospektive Studien sollten den Zusammenhang der interstitiellen Glukosevariabilität prüfen, um eine frühzeitige Risikoabschätzung auf die Ausprägung von kardiovaskulären Erkrankungen zu ermöglichen und zu verbessern.:Inhaltsverzeichnis 3 Abbildungsverzeichnis 5 Tabellenverzeichnis 7 Abkürzungsverzeichnis 9 1. Einführung 10 1.1 Beschreibung des Krankheitsbildes 10 1.2. Pathomechanismus des metabolischen Gedächtnisses 11 1.3 Parameter der Glykämiekontrolle 12 1.4 CG- Monitoring 13 1.5 Arbeitshypothese und Fragestellung 15 2. Material und Methoden 15 2.1 Datenmanagement 15 2.2 Ethische, rechtliche und regulatorische Aspekte 16 2.3 Studiendesign 16 2.4 Studienpopulation 17 2.5 Auswahlkriterien 18 2.5.1 Ein- und Ausschlusskriterien der Studie ORIGIN 18 2.5.2 Ein- und Ausschlusskriterien der Augenstudie Novartis 19 2.5.3 Ein- und Ausschlusskriterien der Studie GLORY 20 2.6 CGM- System 21 2.7 Parameterauflistung 23 2.7.1 Ermittelte Parameter 23 2.7.2 Primäre Endpunkte 25 2.8 Statistische Analyse 26 2.8.1 Fallzahlberechnung: 26 2.8.2 Datenmanagement und statistische Analyse 27 3. Ergebnisse 28 3.1 Deskriptive Analysen der gesamten Studienkohorte 28 3.2 Darstellung der Studienkohorte und der ermittelten Endpunkte 33 3.3 Explorative Analyse der Glukoseparameter aus dem CG-Monitoring 34 3.3.1 Univariate Cox-regression aller Parameter des CG-Monitorings in der gesamten Studienkohorte 34 3.3.2 Korrellation der Blutglukoseparameter mit Signifikanz auf Endpunktausprägung in der univariaten Cox-Regression 35 3.3.3 Multivariate Regression der relevanten, signifikanten Blutglukoseparameter des CG-Monitorings 36 3.3.4 Kaplan-Meier Analysen der gesamten Studienkohorte 37 3.4 Grenzwertbestimmungen der SD der interstitiellen Glukosefluktuation zur prognostischen Voraussage von kardiovaskulären Erkrankungen 44 3.4.1 ROC- Analyse der SD der interstitiellen Glukosefluktuation 44 3.4.2 Kaplan Meier-Statistik mit dem ermittelten Cut-Off bei 1.6mmol/L der SD der interstitiellen Glukosefluktuation 47 3.4.3 Gruppenvergleich von anamnestischen und klinisch-labordiagnostischen Parametern zwischen den Gruppen gebildet mit Cut-off 1.6mmol/L 48 3.4.4 Univariate Coxregression zwischen den SD-Gruppen mit Cut-off von 1.6mmol/L 49 3.5 Vergleich der SD der interstitiellen Glukosefluktuation mit HbA1c-Wert, klinisch-labordiagnostischen und anamnestischen Parametern 49 3.5.1 Vergleich der SD der interstitiellen Glukosefluktuation mit HbA1c 49 3.5.2 ROC-Analyse zu Bestimmung des Cut-Offs von HbA1c in der Studienpopulation 50 3.5.3 Univariate Cox-Regression klinisch-labordiagnostischer und anamnestischer Parameter auf die Ausprägung diabetischer Endpunkte 54 3.5.4 Multivariates Cox-Regressionsmodell mit allen signifikanten Variablen aus der univariaten Cox-Regressionsanalyse 55 3.6 Analysen in der Subgruppe Diabetiker 56 3.6.1 Deskriptive Analyse der Subgruppe der Diabetiker 56 3.6.2 Vergleich der Blutglukoseparameter HbA1c, SD der interstitiellen Glukosefluktuation und Diabetesdauer innerhalb der Diabetiker 57 3.6.3 ROC Analyse der Erkrankungsdauer von Diabetes innerhalb der Subgruppe der Diabetiker 59 3.6.4 Multivariate Cox-Regression relevanter diabetesbezogener Parameter in der Subgruppe der Diabetiker 62 3.6.5 ROC- Analyse des systolischen Blutdruckes innerhalb der Diabetiker 62 3.7 Analysen in der Subgruppe der Nicht-Diabetiker 65 3.7.1 Deskriptive Analyse der Subgruppe der Diabetiker 65 4. Diskussion und Zusammenfassung 66 4.1 Übersicht 66 4.2 Validität des Forschungskonzeptes 67 4.3 Explorative Statistik 69 4.4.1 Analysen in der Gesamtpopulation 69 4.4.2 Analysen in der Subgruppe der Diabetiker 71 4.4.3 Analysen in der Subgruppe der Nicht-Diabetiker 74 4.5 Begrenzungen der Studie 75 4.6 Empfehlungen für weiterführende Forschungen 77 / Working hypothesis and question: Continuous glucose monitoring systems allow measurement over a period of several days and thus provide a reliable detection of the mean interstitial glucose load and its dynamics. Our own preliminary examinations showed relatively stable individual 'glucose patterns' during the course of the day, which can still be demonstrated years later with stable antiglycaemic therapy. At the same time, prospective studies demonstrated only a slight association between HbA1c as a long-term marker of average glucose load and cardiovascular events and mortality. The primary aim of the work is therefore to test whether other parameters of glycemia control are associated with future microangiopathies, mortality and cardiovascular events. Methods: It is a retrospective cohort study with a total of 315 included subjects. The study population is based on three research projects that were carried out at the study center for metabolic-vascular medicine, GWT TU-Dresden GmbH in Dresden in the period 2010-2013. The collected anamnestic, clinical and clinical laboratory diagnostic parameters as well as the blood sugar measurements of the CGM were combined from the archives into a database and used to create case groups. In addition, clearly defined cardiovascular events were identified through telephone interviews and routinely collected data as part of follow-up examinations at the study center, which manifested themselves after the CGM was carried out. Results: In a univariate Cox regression in the overall cohort, the results show a significance (p <0.001) of the glucose-related parameters of interstitial glucose fluctuation, HbA1c and duration of diabetes on the expression of cardiovascular endpoints. In a multivariate Cox regression including the glucose parameters interstitial glucose fluctuation, HbA1c and diabetes duration in the subgroup of diabetics, however, only the diabetes duration (p <0.05) with a significant association with the expression of cardiovascular endpoints was shown. The parameters interstitial glucose fluctuation (p = 0.44) and HbA1c (p = 0.35) lose their significant association here. This is also reflected in the ROC analyzes. Diabetes duration shows an AUC (Area under the curve) of 0.66 in the ROC analyzes. With a theoretically optimal cut-off of 9 years, a sensitivity of 61% and specificity of 70% can be observed. Within the glucose-related parameters, the duration of diabetes is the strongest classifier for the prognosis of cardiovascular diseases. In a multivariate Cox regression in the total population, HbA1c (p <0.05) shows a stronger association with the manifestation of cardiovascular diseases than with interstitial glucose fluctuation (p = 0.25). However, both demonstrate similar results in the ROC analysis. With a cut-off of> = 6.8%, HbA1c has a sensitivity of 37% and a specificity of 81%. The AUC of the parameter HbA1c is 0.59 in the ROC analysis. For the interstitial glucose fluctuation, an optimal cut-off of 1.6mmol / L could be calculated and shows a sensitivity of 48% and a specificity of 66%. The interstitial glucose variability AUC is 0.57. In a multivariate Cox regression, in which all cardiovascular parameters from the univariate analysis were included, systolic blood pressure also appears to be significant (p <0.05) on the manifestation of cardiovascular events. With a cut-off of 152mmHg systolic blood pressure, a sensitivity of 43% and specificity of 72% were determined in the ROC analysis. The AUC of the systolic blood pressure resulted in a value of 0.63. Conclusion: In contrast to the previous studies on blood glucose variability, in addition to a possible association of interstitial glucose variability, an attempt was also made to set a defined limit value of 1.6mmol / L from which an increased cardiovascular risk is likely. This limit value is intended to help standardize the current inconsistent study situation regarding interstitial glucose variability in order to provide permanently comparable results in this area. Interstitial glucose variability is shown as a weak predictor of the development of cardiovascular diseases. However, the established HbA1c value also shows a weak predictive value, which is comparable to that of interstitial glucose variability. Further studies, especially prospective studies, should examine the relationship between interstitial glucose variability in order to enable and improve an early risk assessment of the severity of cardiovascular diseases.:Inhaltsverzeichnis 3 Abbildungsverzeichnis 5 Tabellenverzeichnis 7 Abkürzungsverzeichnis 9 1. Einführung 10 1.1 Beschreibung des Krankheitsbildes 10 1.2. Pathomechanismus des metabolischen Gedächtnisses 11 1.3 Parameter der Glykämiekontrolle 12 1.4 CG- Monitoring 13 1.5 Arbeitshypothese und Fragestellung 15 2. Material und Methoden 15 2.1 Datenmanagement 15 2.2 Ethische, rechtliche und regulatorische Aspekte 16 2.3 Studiendesign 16 2.4 Studienpopulation 17 2.5 Auswahlkriterien 18 2.5.1 Ein- und Ausschlusskriterien der Studie ORIGIN 18 2.5.2 Ein- und Ausschlusskriterien der Augenstudie Novartis 19 2.5.3 Ein- und Ausschlusskriterien der Studie GLORY 20 2.6 CGM- System 21 2.7 Parameterauflistung 23 2.7.1 Ermittelte Parameter 23 2.7.2 Primäre Endpunkte 25 2.8 Statistische Analyse 26 2.8.1 Fallzahlberechnung: 26 2.8.2 Datenmanagement und statistische Analyse 27 3. Ergebnisse 28 3.1 Deskriptive Analysen der gesamten Studienkohorte 28 3.2 Darstellung der Studienkohorte und der ermittelten Endpunkte 33 3.3 Explorative Analyse der Glukoseparameter aus dem CG-Monitoring 34 3.3.1 Univariate Cox-regression aller Parameter des CG-Monitorings in der gesamten Studienkohorte 34 3.3.2 Korrellation der Blutglukoseparameter mit Signifikanz auf Endpunktausprägung in der univariaten Cox-Regression 35 3.3.3 Multivariate Regression der relevanten, signifikanten Blutglukoseparameter des CG-Monitorings 36 3.3.4 Kaplan-Meier Analysen der gesamten Studienkohorte 37 3.4 Grenzwertbestimmungen der SD der interstitiellen Glukosefluktuation zur prognostischen Voraussage von kardiovaskulären Erkrankungen 44 3.4.1 ROC- Analyse der SD der interstitiellen Glukosefluktuation 44 3.4.2 Kaplan Meier-Statistik mit dem ermittelten Cut-Off bei 1.6mmol/L der SD der interstitiellen Glukosefluktuation 47 3.4.3 Gruppenvergleich von anamnestischen und klinisch-labordiagnostischen Parametern zwischen den Gruppen gebildet mit Cut-off 1.6mmol/L 48 3.4.4 Univariate Coxregression zwischen den SD-Gruppen mit Cut-off von 1.6mmol/L 49 3.5 Vergleich der SD der interstitiellen Glukosefluktuation mit HbA1c-Wert, klinisch-labordiagnostischen und anamnestischen Parametern 49 3.5.1 Vergleich der SD der interstitiellen Glukosefluktuation mit HbA1c 49 3.5.2 ROC-Analyse zu Bestimmung des Cut-Offs von HbA1c in der Studienpopulation 50 3.5.3 Univariate Cox-Regression klinisch-labordiagnostischer und anamnestischer Parameter auf die Ausprägung diabetischer Endpunkte 54 3.5.4 Multivariates Cox-Regressionsmodell mit allen signifikanten Variablen aus der univariaten Cox-Regressionsanalyse 55 3.6 Analysen in der Subgruppe Diabetiker 56 3.6.1 Deskriptive Analyse der Subgruppe der Diabetiker 56 3.6.2 Vergleich der Blutglukoseparameter HbA1c, SD der interstitiellen Glukosefluktuation und Diabetesdauer innerhalb der Diabetiker 57 3.6.3 ROC Analyse der Erkrankungsdauer von Diabetes innerhalb der Subgruppe der Diabetiker 59 3.6.4 Multivariate Cox-Regression relevanter diabetesbezogener Parameter in der Subgruppe der Diabetiker 62 3.6.5 ROC- Analyse des systolischen Blutdruckes innerhalb der Diabetiker 62 3.7 Analysen in der Subgruppe der Nicht-Diabetiker 65 3.7.1 Deskriptive Analyse der Subgruppe der Diabetiker 65 4. Diskussion und Zusammenfassung 66 4.1 Übersicht 66 4.2 Validität des Forschungskonzeptes 67 4.3 Explorative Statistik 69 4.4.1 Analysen in der Gesamtpopulation 69 4.4.2 Analysen in der Subgruppe der Diabetiker 71 4.4.3 Analysen in der Subgruppe der Nicht-Diabetiker 74 4.5 Begrenzungen der Studie 75 4.6 Empfehlungen für weiterführende Forschungen 77

Metabolic syndrome (MetS) – a clustering of pathological conditions, including abdominal obesity, hypertension, dyslipidemia and hyperglycaemia – is closely associated with the development of type 2 diabetes mellitus (T2DM) and a high risk of cardiovascular disease. A combination of multigenetic predisposition and lifestyle choices accounts for the varying inter-individual risk to develop MetS and T2DM, as well as for the individual amount of the increase in cardiovascular risk in those patients. A physically active lifestyle can offset about half of the genetically mediated cardiovascular risk. Yet, the extent to which standardized exercise programmes can reduce cardiovascular risk differs between patients. Exercise parameters, such as frequency, intensity, type and duration or number of repetitions, differentially target metabolic function, vascular healthand physical fitness. In addition, exercise-induced molecular mechanisms are modulated by other patient-specific variables, such as age, diet and medication. This review discusses the molecular and cellular mechanisms underlying the effects of exercise training on cardiovascular risk specifically in patients with MetS and T2DM.

Dados prévios ja demostram que o controle intensivo da glicemia diminui o risco de eventos microvasculares em pacientes com diabetes mellitus. No entanto, seu efeito cardiovascular é incerto. Nós sumarizamos os dados de estudos das principais bases de dados. 2 revisores extraíram dados de estudos randomizados de pacientes com diabetes tipo 2, que visavam 2 níveis de intensidade da glicemia. Investigou-se as retinopatia, neuropatias, nefropatias, mortalidade cardiocascular e total, infarto do miocárdio (IAM), acidente vascular cerebral, amputação de membros e episódios de hipoglicemia. Realizamos a meta-análise para obter o risco relativo (RR). Foram incluídos 7 estudos com 27.814 pacientes. O controle intensivo reduziu o RR de IAM e amputação, além progressão da retinopatia, incidência de neuropatia periférica, incidência e progressão de nefropatia e microalbuminúria. Entretanto, dobrou o risco de episódios de hipoglicemia. Não houve diferenças quanto à mortalidade e outros resultados. Conclui-se que controle intensivo reduziu o risco de alguns eventos, sem reduzir a mortalidade, porém as custas do dobro da incidência de de hipoglicemia. / Previous data already show that intensive glucose control reduces the risk of microvascular events in patients with diabetes mellitus. However, its cardiovascular effect is uncertain. We summarize data from studies of the major databases. 2 reviewers extracted data from randomized studies of patients with type 2 diabetes, aimed at two intensity levels of blood glucose. We investigated the retinopathy, neuropathy, nephropathy, and total mortality cardiocascular, myocardial infarction (IAM), stroke, limb amputation and episodes of hypoglycemia. We conducted a meta-analysis to obtain the relative risk (RR). We included seven studies with 27.814 patients. The intensive control reduced the RR of IAM, and amputation, and progression of retinopathy, incidence of peripheral neuropathy, incidence and progression of nephropathy and microalbuminuria. However, it doubled the risk of hypoglycemia. There were no differences in mortality and other outcomes. We conclude that intensive control reduced the risk of some events without reducing mortality, but the expense of twice the incidence of hypoglycemia.

Magister Public Health - MPH (Public Health) / The prevalence of diabetes in Africa has been on the increase. A prevalence of 1%- 10% has been reported by different authors in different regions in Nigeria. The International Diabetes Federation estimates that 1.9% of Nigerians are diabetic and most of them have complications at the time of diagnosis. Laboratory measurement of Glycosylated hemoglobin (HbA1c) is the method of choice for monitoring glycaemic control but due to its cost and limited availability, most developing countries use fasting plasma glucose (FPG) measurement (which is less reliable) to assess glycaemic control. Most diabetic patients in Nigeria have poor glycaemic control and several factors have been implicated especially socio-economic, behavioral and treatment-related factors. Understanding the reasons for poor glycaemic control is essential in order to reduce the rate of diabetes complications.

Observational studies have indicated that fruit and vegetables, and dietary antioxidants may play an important role in reducing the risk of chronic diseases, potentially by affecting pathogenic mechanisms such as oxidative stress and inflammation. Clinical trials investigating the effects of supplementation with single or a few antioxidants in high doses have, however, shown inconsistent results and thus have not been able to support the observational findings. It was therefore hypothesised that a supplement, containing a combination of antioxidants mainly extracted from fruit and vegetables, and supplied at moderate doses, might act more beneficially than single antioxidants given at pharmacological doses. The effects of such a supplement were investigated in two interventional studies described in this thesis. The effects on antioxidant status, metabolic control, oxidative stress and inflammation were investigated in overweight men and in patients with type 2 diabetes, subjects that could be expected to have elevated levels of oxidative stress and inflammatory activity. The results of the studies did not support the hypothesis that supplementation with antioxidants from fruit and vegetables may have beneficial effects by counteracting oxidative stress and inflammation, despite markedly increased plasma antioxidant concentrations. However, interesting associations were observed in diabetes patients at baseline between intake of antioxidant rich food as well as levels of antioxidants in plasma, and markers of oxidative stress and inflammation. These associations are compatible with the hypothesis that a high intake of fruit and vegetables and dietary antioxidants decrease oxidative stress levels, have anti-inflammatory effects and a beneficial influence on glycaemic control. The results also indicated that glycaemic control may affect the level of oxidative stress. The absence of beneficial effects from antioxidants might to some extent be explained by the initial levels of oxidative stress and inflammation and by the antioxidative status in the subjects included in the studies. Since the levels generally were comparable with those observed in healthy subjects, this might have decreased the ability to observe any beneficial effects of supplementation with additional antioxidants. Continued investigations are needed to characterise the individuals who potentially might benefit from antioxidant supplementation. In view of apparent positive effects from a high intake of fruit and vegetables found in observational studies and until more knowledge is available from interventional trials about possible benefits and potential risks of antioxidant supplementation it still seems reasonable to recommend a diet rich in fruit and vegetables.

Blood glucose management in Type 1 diabetes is crucial in preventing several diabetic complications. Blood glucose management is a complex task requiring diabetics too carefully administer the correct dosages of insulin by taking their blood glucose levels, food consumption, exercise, stress, illnesses and several other factors into account. Improved bolus calculation greatly aids in controlling blood glucose levels within a tight range. This study investigates how the ets-concept (Equivalent Teaspoons Sugar-concept) can be used to develop products to calculate insulin boluses. A cellular phone based software application was developed to calculate insulin boluses using the ets-concept. This product was tested in a clinical trial. A blood glucose characterization procedure was also developed to characterize the blood glucose response of a Type 1 diabetic to carbohydrate ingestion and insulin administration. The characterization procedure was used during the clinical trial to characterize patients in order to customize the bolus calculation products for the specific diabetic user. / Thesis (Ph.D. (Mechanical Engineering)--North-West University, Potchefstroom Campus, 2006

The correct administration of insulin is a constant challenge for type 1 diabetics. The correct insulin regime leads to fewer complications and an easier way of life. The amount of insulin administered must take into account the meals eaten, previous administered insulin, exercise etc. A rapid process for determining insulin regimes that is accessible to type 1 diabetics will greatly reduce diabetic complications later in life. This study researches such a process. Software is developed to use the ets-concept to simulate blood glucose levels. From these simulations blood glucose characterisation can be done to propose insulin regimes. Data gathered in previous studies is used to verify the results of this process. These results are compared to factors that describe the accuracy of a person's blood glucose control. The effects the new regimes will have are used to make recommendations to the end-user. Accurate characterisation leads to insulin regImes that will Improve the control performance of type 1 diabetes. / Thesis (M.Ing. (Electronical Engineering)--North-West University, Potchefstroom Campus, 2008.

Blood glucose management in Type 1 diabetes is crucial in preventing several diabetic complications. Blood glucose management is a complex task requiring diabetics too carefully administer the correct dosages of insulin by taking their blood glucose levels, food consumption, exercise, stress, illnesses and several other factors into account. Improved bolus calculation greatly aids in controlling blood glucose levels within a tight range. This study investigates how the ets-concept (Equivalent Teaspoons Sugar-concept) can be used to develop products to calculate insulin boluses. A cellular phone based software application was developed to calculate insulin boluses using the ets-concept. This product was tested in a clinical trial. A blood glucose characterization procedure was also developed to characterize the blood glucose response of a Type 1 diabetic to carbohydrate ingestion and insulin administration. The characterization procedure was used during the clinical trial to characterize patients in order to customize the bolus calculation products for the specific diabetic user. / Thesis (Ph.D. (Mechanical Engineering)--North-West University, Potchefstroom Campus, 2006