Umeå University's logo

Type 2 diabetes (T2D) is one of the most rapidly increasing metabolic diseases worldwide, affecting more than 500 million individuals. The condition is characterized not only by chronic hyperglycemia but also bya long preceding phase of metabolic dysregulation. Before glucose levels reach diagnostic thresholds, most individuals experience a period of prediabetes with insulin resistance and compensatory elevations in both insulin secretion and circulating insulin levels. This stage is far from benign: prediabetes is associated with an increased risk of cardiovascular disease, neuropathy, and renal impairment. Understanding the mechanisms that drive these early metabolic changes is therefore important for improving prevention and treatment strategies. Central components in the development of prediabetes are insulin secretion, insulin clearance, and insulin sensitivity. Despite their importance, these processes are challenging to measure accurately in large populations and their respective contributions to early cardiometabolic disease remain incompletely understood. This thesis aims to deepen our understanding of insulin dynamics in prediabetes through four complementary studies.

The first study addressed how to measure first-phase insulin secretion reliably in large-scale research. By conducting meta-analyses of published studies comparing gold-standard methods of insulin secretion with commonly used surrogate measures, we evaluated the validity of these surrogate measures and provide guidance for their use in epidemiological settings. In total, 33 studies comprising 5362 participants with normal glucose tolerance, prediabetes, or T2D were included. Among surrogate measures evaluated, those derived from the first 30 min of the oral glucose tolerance test (OGTT) most closely approximated gold-standard measures of first-phase insulin secretion.

The second study examined how insulin secretion is reflected in the metabolic signature in individuals with prediabetes. By comparing individuals with prediabetes and high insulin secretion, prediabetes withlow secretion, normal glucose tolerance, and T2D, we investigated metabolic differences in plasma samples collected during an OGTT. Plasma samples from 100 participants were obtained at 0 and 120 min during the OGTT, and 280 metabolites and 218 putative lipids were quantified using ultra-high performance liquid chromatography quadrupole time-of-flight mass spectrometry. Classification models identified key metabolites and lipids distinguishing the four groups, including ceramides. Notably, the ratio of Cer(d18:1/24:1) toCer(d18:1/24:0), which has been associated with cardiovascular risk, was highest in the prediabetes group with high insulin secretion.

The third study investigated the role of insulin dynamics in the development of atherosclerosis in a population without diabetes. By examining correlations between insulin secretion, insulin clearance, and insulin sensitivity on the one hand, and subclinical atherosclerosis in coronary and carotid arteries on the other, we explored how early metabolic disturbances contribute to atherosclerotic cardiovascular disease. A total of 2054 individuals 50–64 years of age were included. Insulin dynamics were assessed using an OGTT and atherosclerosis using computed tomography of the coronary arteries and ultrasound of both carotid arteries. Our findings indicate that increased insulin resistance is associated with early atherosclerotic changes, even after adjustment for cardiovascular risk factors.

The fourth study shifts focus from pathophysiology to clinical management, exploring how digital health interventions can support individuals with T2D in achieving remission. Diabetes remission can be achieved through substantial weight loss. Participants in this study followed a low-calorie diet for 3 months, after which they gradually reintroduced food to support weight maintenance. They were monitored through digital meetings and self-reported body weight and blood glucose on a digital platform. In this study, we examined participants’ experiences of this eHealth-supported intervention using interviews analyzed with qualitative content analysis. Ten participants with more than 1 year’s experience of eHealth support for weight loss were interviewed. The results highlighted both empowering aspects, such as increased autonomy, and challenges, such as feelings of isolation and communication barriers.

The use of validated surrogate measures enables the estimation of insulin secretion in large populations. Looking ahead, these measures, combined with lipid profiling may contribute to more person-centeredcare by helping to identify individuals at elevated risk of future cardiovascular disease. Furthermore, our finding that insulin resistance is associated with early atherosclerosis underscores the need for early risk identification to provide preventive interventions. To enable broad implementations of such strategies, resource-efficient solutions are essential, which we explored in our final study.