Islet autoantibodies may emerge during the first year of life, suggesting that early life exposures may be pivotal. Consequently, infant and childhood dietary factors have been implicated as the vehicle for environmental triggers in the pathogenesis of the disease.

Exposure to cow’s milk in early neonatal life has received considerable attention in the pathogenesis of T1D. In the “Childhood Diabetes in Finland (DiMe)” study, high consumption of cow’s milk protein was strongly associated with the emergence of β-cell autoantibodies and progression to clinical T1D in initially unaffected siblings of children with T1D, potentially secondary to a misdirected immune response to bovine serum albumin due to a shared structure with islet protein p69. Alternatively, this process could be due to a higher titer of antibodies to bovine insulin, resulting in immunization to bovine insulin, a molecule that differs structurally from human insulin in only three amino acid positions. However, more recent large-scale studies have been inconclusive. The “Trial to Reduce IDDM in Genetically at Risk (TRIGR)” randomized infants at increased risk of T1D to receive either an extensively hydrolyzed formula or a conventional cow milk–based formula and found that weaning to a hydrolyzed formula did not reduce the cumulative incidence of T1D after a median follow-up of 11.5 years. In contrast, early reports have suggested that children who were exclusively breastfed for prolonged periods as infants are at lower risk of developing T1D due to decreased gut permeability and enterovirus infection protection. More recent studies have suggested that nonbreastfed children have a higher risk of developing T1D than infants who were breastfed; however, this association seems to be independent from breastfeeding duration.

Gluten has also been incriminated as an important diabetes promoting agent. Two prospective studies have shown an association between introduction of cereals in infancy, particularly within the first 3 months of life, and early β-cell autoimmunity, possibly secondary to effects of gut and immune system immaturity in at-risk individuals. Yet, delaying gluten exposure until 12 months of age in the “Primary Prevention of Type 1 Diabetes in Relatives at Increased Genetic Risk (BABYDIET)” study was safe but did not cause a substantial reduction in the risk for islet autoimmunity in children at risk.

Vitamin D deficiency has also been implicated as a risk factor for T1D. Results reported by a Finnish study found that regular, high-dose vitamin D supplements in infancy were associated with a decreased risk compared with no supplementation. Yet, this could be influenced by additional genetic factors as there is also a striking difference in the annual incidence rate of T1D in the neighboring populations of Finland and Russian Karelia (42/100,000 vs. 7.8/100,000) with no difference reported in the circulating vitamin D concentrations in pregnant women and schoolchildren. The “Diabetes Autoimmunity Study in the Young (DAISY)” found no association between vitamin D intake or 25-hydroxyvitamin D concentrations and islet autoimmunity and T1D in childhood, although multiple vitamin D metabolism polymorphisms have been implicated in the risk of islet autoimmunity and progression to T1D.

Omega-3 fatty acids have also been reported to play a protective role in the etiology of T1D and studies have shown that a higher omega-3 fatty acid intake is associated with lower risk of β-cell autoimmunity, particularly in young children.

Viruses such as herpesviruses, mumps, rubella, rotavirus, retroviruses, and, in particular, enteroviruses have long been implicated in the etiology of T1D. Viruses can initiate autoimmunity by at least four mechanisms: (1) molecular mimicry between viral proteins and autoantigens; (2) release of autoantigens following β-cell cytokine-induced injury; (3) cytokine-induced upregulation of major histocompatibility complexes and costimulatory molecules on antigen-presenting cells, enabling them to present self-peptides in immunogenic form to T-cells; and (4) interference with central and peripheral self-tolerance. Notably, in animal models, viral infections can both promote and diminish autoimmunity which may have potential as a future prevention strategy.

Viral infections during childhood may also play a role in the development of immunoregulatory mechanisms which protect against diabetes. The hygiene hypothesis proposes that improved hygiene in the Western world has led to a decline in immunity to common infections and increased incidence of autoimmunity. Put another way, with early infectious exposures, young children build appropriate immune responses to pathogens. This idea is also supported by the findings that daycare attendance in early infancy confers protection against the development of childhood diabetes. The relationship between viral infection and autoimmunity appears to be temporal, as studies have shown that enteroviral infections before weaning are beneficial and infections after are associated with susceptibility to T1D development. Vaccinations have not been associated with the development of T1D.

There are many other environmental factors that have been proposed to be involved in the pathogenesis of T1D. Increased weight gain in infancy has been reported to be an important risk factor but has not been confirmed in larger studies. Children who develop T1D have been shown to be both heavier and taller in infancy than their peers without diabetes. Psychological stress may also constitute a trigger in the etiology of T1D. Children with T1D have been shown to be more frequently exposed to stressful situations than children without diabetes. Psychological stress is particularly frequent in the 2 years preceding T1D diagnosis, and stress-related changes in hormonal and neuronal signals may contribute to the development of T1D in genetically susceptible individuals.

One potential protective factor is the microbiome, as a study in nonobese mice with diabetes demonstrated associations between T1D prevention and probiotic administration. In another study, BioBreeding diabetes-prone rats were given lactobacillus strains isolated from BioBreeding diabetes-resistant rats and showed reduced rates of diabetes development.

Yet, with intensive treatment of hyperglycemia also comes an increased risk of hypoglycemia which may play an equally important role in the development of cardiovascular disease in T1D. Due to the potential for severe neurological compromise (i.e., impaired cognition, seizures, coma) in the setting of severe hypoglycemia, the autonomic system activates multiple counterregulatory responses including stimulated release of glucagon and epinephrine and increased gluconeogenesis at the level of the liver to acutely raise blood glucose concentrations and improve blood flow to the brain to prevent neuroglycopenia. Recurrent and severe hypoglycemia can lead to severe damage of cerebral tissues and a fine balance exists between prevention of longstanding hyperglycemia and recurrent hypoglycemia. In animal studies of hypoglycemia, rats exposed to recurrent hypoglycemic episodes exhibited a 44% increase in neuronal death when compared to rats exposed to insulin without hypoglycemia, demonstrating the potential for severe cerebral ischemia in the setting of low blood sugars. Recurrent hypoglycemia has also been directly associated with an increased frequency of brain lesions on MRI in individuals with T1D, as well as detrimental effects on cognitive function which can lead to an increased risk of cardiovascular events and mortality. Further, there is also evidence of structural white and gray matter changes in the brain in adults with T1D who exhibit impaired hypoglycemia awareness.

Euglycemia is also an essential feature of normal cardiac function and hypoglycemic episodes can result in multiple features of cardiac dysfunction, such as a widened pulse pressure with reduction of myocardial perfusion in diastole and ECG changes including QT prolongation with ST changes and arrhythmias. Catecholamine release in the setting of hypoglycemia, as well as an imbalance between insulin exposure and available carbohydrates, results in hypokalemia which may further result in ECG changes. These effects are most dangerous overnight in individuals with T1D as the altered mental status of sleep makes it difficult to assess for the effects of hypoglycemia and ensure swift treatment. Cardiac arrhythmias that occur secondary to severe hypoglycemia have resulted in a well-known complication entitled sudden cardiac death during hypoglycemia, a condition that may be preventable with the more widespread use of continuous glucose monitors to alert people with T1D of impending hypoglycemia. Recent studies suggest diurnal differences in individual susceptibility to cardiac arrhythmias, with daytime episodes of acute hypo- and hyperglycemia presenting a greater risk for arrhythmia than nighttime episodes in people with T1D. Lastly, individuals with recurrent hypoglycemia demonstrate greater endothelial dysfunction, as evidenced by flow-mediated dilatation and increased carotid intima-media thickness, an indicator of subclinical atherosclerosis. This effect has been theorized to be more significant in the setting of poorly controlled diabetes of long duration as there is a greater risk of baseline vascular damage.

The incidence of obesity is rapidly increasing in both youth and adults with T1D. Central adiposity, an important risk factor that is worsened by intensive insulin therapy, is increasingly recognized in T1D. The incidence of obesity was recently reported to be 37% in one cohort of adults with newly diagnosed T1D, and 78% of men with T1D in the “Urologic EDIC (UroEDIC)” study were overweight or obese. A similar trend has been observed worldwide and also in young persons with T1D. This trend is likely secondary to general population factors such as decreased physical activity and sleep and increased high-calorie food consumption, as well as T1D-specific factors such as an overemphasis on carbohydrate counting and/or aggressive insulin management strategies to target euglycemia, the need to treat hypoglycemia with food, disrupted sleep from device alarms and overnight hypoglycemia, and fear of hypoglycemia reducing physical activity. Indeed, the DCCT found that intensive insulin management improved glycemia and complications in T1D; however, it also resulted in weight gain, central adiposity, hypertension, dyslipidemia, and generalized inflammation, all factors associated with diabetic kidney disease and cardiovascular disease.

T1D is characterized by a distinct lipid profile including elevated triglycerides and a low HDL-C concentration which is exacerbated by elevated insulin resistance, a finding that is shared with T2D. Adults with poorly controlled T1D exhibit increased triglyceride-rich lipoproteins and low-density lipoprotein cholesterol (LDL-C) particles, with a shift towards the more atherogenic small, dense LDL-C, a finding that is also exhibited in youth with T1D. Impaired cholesterol metabolism is also an important feature of T1D and youth with T1D demonstrate impaired fatty acid oxidation, increased free fatty acid uptake, and increased lipid accumulation in a variety of organ systems. Insulin therapy and good glycemic control can help lower LDL-C and raise HDL-C concentrations, improve lipoprotein concentrations, reduce the rate of LDL-C production, and improve lipid transfers.

Although total LDL-C concentrations are often within the normal range in people with T1D, LDL-C remains a significant predictor of cardiovascular disease risk. LDL-C particles in T1D are increasingly pathogenic as they are smaller, denser particles than those found in healthy controls and are more prone to entry, oxidation, and retention in the arterial wall, which may contribute to the development of early atherosclerosis. In fact, extensive evidence implicates the oxidation of LDL-C as a major player in atherosclerosis which may be compounded by the negative effects of hyperglycemia and other aspects inherent to diabetes that promote altered redox balance and increased oxidative stress. An intriguing more recent finding in this field has been the identification of autoantibodies to oxidized LDL-C which may be involved in atherosclerosis and coronary calcification, particularly in T1D.

Central adiposity may further promote dyslipidemia, including the development of secondary factors such as increased generalized inflammation and inflammation localized to the adipose tissue, as well as through the development of higher free fatty acid concentrations. Hypertriglyceridemia of diabetes involves changes in both the production and degradation of lipids, including a higher hepatic secretion of triglyceride-rich lipoproteins such as very low-density lipoproteins (VLDLs) and altered hydrolysis of these and other triglyceride-rich lipoproteins. Another potential component of hypertriglyceridemia may be postprandial triglyceride excursions, which may be more predictive of cardiovascular disease risk than fasting triglyceride concentrations usually obtained in clinic settings for dyslipidemia screening.

Lipoprotein lipase, a key enzyme involved in hydrolyzing fatty acids from triglycerides and delivering these fatty acids to tissues, may also be defective in T1D in a similar fashion as T2D. Notably, lipoprotein lipase-mediated hydrolysis of triglycerides has been identified as a mechanism for generating natural ligands for the nuclear receptor known as peroxisome proliferator-activated receptor alpha (PPAR-α), which, when activated by ligands, controls the expression of multiple genes involved in lipid metabolism, inflammation, and fatty acid oxidation.

Hepatic dysregulation, a consequence of hyperinsulinemia and hyperglycemia, may also impact lipid metabolism through the glycation of proteins and lipoproteins. In addition to altering the normal function of these entities, the breakdown of glycated proteins and lipoproteins, known as advanced glycation end products (AGEs), activates specific receptors for AGEs (RAGEs), resulting in responses closely linked to atherosclerotic complications, such as increases in matrix metalloproteinases thought to promote plaque destabilization and rupture.

HDL-C concentrations are inversely associated with coronary heart disease risk and significant effort has been applied to elucidating the mechanisms underlying the low HDL-C commonly observed in people with diabetes. Both abnormal production of HDL-C and remodeling of this lipid by plasma enzymes may contribute to the low concentration of circulating HDL-C observed in people with T1D. Expression and activity of endothelial lipase, a phospholipase that is synthesized in and expressed on the surface of vascular endothelium, catabolizes HDL-C, resulting in decreased concentrations of this putatively antiatherogenic lipoprotein. Elevated concentrations of endothelial lipase protein are significantly correlated with coronary artery calcification score as well as other features of metabolic syndrome including waist circumference, blood pressure, triglycerides, HDL-C, and fasting glucose in individuals with a family history of premature coronary heart disease. In addition, direct correlations have been observed between endothelial lipase levels and circulating markers of inflammation including high-sensitivity C-reactive protein, interleukin-6, and soluble intercellular adhesion molecule. Low-dose endotoxemia in 20 research participants increased endothelial lipase concentrations 12 to 16 hours after injection, and this increase in endothelial lipase correlated with reductions in plasma HDL-C.

Collectively these data suggest that low-intensity, generalized inflammation, a common feature of both T1D and T2D, controls HDL-C through effects on endothelial lipase, providing a possible mechanism for the low HDL-C in T1D and the exaggerated cardiovascular risk associated with insulin-resistant states such as T1D. Despite the clear epidemiologic inverse association between HDL-C and cardiovascular risk, the hypothesis that raising HDL-C can reduce cardiovascular events has not yet been proven. The recent failure of large randomized, placebo-controlled trials designed to test this hypothesis using cholesteryl ester transfer protein inhibitors and niacin, which both raise HDL-C concentrations, suggests that the biology of the athero-protective effects of HDL-C are likely very complex and cannot be ascribed exclusively to a single parameter such as HDL-C quantity, the current clinically monitored lipid parameter.

The “Pittsburgh Epidemiology of Diabetes Complications Study,” a 10-year prospective study in adults with T1D of childhood onset (<17 years of age), further identified strong relationships between elevated LDL-C, elevated triglycerides, and low HDL-C with both mortality and coronary artery disease in T1D (RR 1.8–12.1 for LDL-C >100 mg/dL, 2.0–7.1 for triglycerides >150 mg/dL, 0.4–0.7 for HDL-C <45 mg/dL). Based on these results, it is recommended that the treatment goals for adults with T1D should be LDL-C <100 mg/dL (2.6 mmol/L), HDL cholesterol >45 mg/dL (1.1 mmol/L), and triglycerides <150 mg/dL (1.7 mmol/L) to reduce long-term risk of cardiovascular disease.

Insulin resistance is common in T1D and occurs independently from hyperglycemia. In studies using gold standard hyperinsulinemic-euglycemic clamps, both normal weight youth and adults with T1D demonstrate substantial whole body insulin resistance compared to their peers without diabetes of similar age and body mass index (BMI), and the degree of insulin resistance did not relate to hyperglycemia but instead related to hypoglycemia assessed via continuous glucose monitor, thus supporting the use of agents that may lessen hypoglycemia. Moreover, normal weight youth and adults with T1D also have adipose-specific insulin resistance, as evidenced by failure to suppress free fatty acids during hyperinsulinemia, causing chronic exposure to increased free fatty acids, which can induce muscle insulin resistance and vascular dysfunction, and also exhibit hepatic insulin resistance, contributing to hyperglycemia.

Insulin resistance is also a strong risk factor for cardiovascular disease in T1D. In adolescents with T1D, insulin resistance is associated with increased free fatty acids, a more atherogenic lipoprotein profile, as well as vascular dysfunction, decreased maximal exercise capacity, and increased waist circumference. In the “Coronary Artery Calcification in Type 1 Diabetes (CACTI)” study, skeletal muscle insulin resistance was also associated with a higher burden of coronary artery calcification in adults with T1D. Orchard et al. have also demonstrated that insulin resistance, but not glycemia, was associated with higher rates of hard coronary artery disease and mortality in T1D. Additionally, insulin resistance has been closely linked with markers of increased proximal vessel stiffness in the thoracic descending aorta on aortic MRI in adolescents with T1D. In the “SEARCH for Diabetes in Youth” study, elevated estimated insulin resistance associated with arterial stiffness. Thus insulin resistance may be a distinct metabolic complication of T1D that contributes to multiple aspects of cardiovascular disease risk, independent of glycemic control.

Inflammation plays a key role in the progression of cardiovascular disease and T1D is closely associated with chronic systemic inflammation and macrophage reprogramming. Atherosclerosis is also strongly driven by chronic inflammation. Significant inflammatory cell infiltrates have been found in atherosclerotic plaques and proinflammatory cytokines released by macrophages lead to endothelial dysfunction and microcalcifications. Systemic markers of inflammation (e.g., highly sensitive C-reactive protein) are strong predictors of future cardiovascular events. Two of the greatest contributors to cardiovascular disease are the development of vascular endothelial dysfunction, most commonly assessed as brachial artery flow-mediated dilation (FMD _BA ), and stiffening of the aorta and carotid arteries.

Considerable evidence now points to chronic, low-grade inflammation as a factor for initiation and perpetuation of atherothrombosis and insulin resistance. In the “Physicians’ Health Study” ( n = 22,000 men) and the “Women’s Health Study” ( n = 38,000 women), the relative risk of future myocardial ischemia, stroke, and cardiovascular death in these otherwise healthy individuals at baseline was linearly associated with highly sensitive C-reactive protein across the normal range (≤3 mg/L). Furthermore, this relationship was evident even after controlling for other risk factors. Similar findings have since been validated in other large cohorts, with only a few controversial exceptions. In a cross-sectional study of 48 people with T1D and 66 people without T1D from the DCCT, significantly higher concentrations of acute-phase proteins including alpha ₁ -acid glycoprotein and highly sensitive C-reactive protein were found in those with diabetes. Other circulating inflammatory biomarkers including interleukin-6, matrix metalloproteinase-9, pentraxin-3, lipoprotein-associated phospholipase A ₂ , and soluble adhesion molecules have demonstrated similar results for predicting cardiovascular disease risk, albeit with different magnitudes and variable usefulness as clinical tools. These results suggest that the clinical observations regarding highly sensitive C-reactive protein may reflect inflammatory responses in the vasculature.

Inflammatory changes have also been found in both adipose tissue and pancreatic β-cells, which may relate to subsequent development of insulin resistance and potential β-cell failure. Notably, infiltration of visceral adipose tissue by macrophages and other leukocytes has been shown to contribute to the systemic proinflammatory state observed in insulin resistance. Moreover, studies with salicylates, thiazolidinediones, and other agents have raised the question of whether treating systemic inflammation can improve insulin resistance and ultimately the course of diabetes itself. Hence, treatments that target chronic inflammation may have a significant impact on cardiovascular disease outcomes in T1D, but this has not yet been definitively proven.

The endothelium, a single cell layer lining the entire vascular tree, serves as a dynamic, reactive organ engaged in endocrine, paracrine, and autocrine function which also assists the body as a transducer of circulatory components including mediators of disease risk such as glucose, free fatty acids, and pathogenic lipoproteins. As the physical barrier separating flowing blood from the vessel wall, the endothelium is uniquely positioned to control homeostatic processes including blood pressure, hemostasis, and homing of immune cells to sites of inflammation. When dysregulated, each of these processes can contribute to the development of atherosclerosis and have been especially implicated in atherosclerosis secondary to diabetes.

The control of vascular resistance by the endothelium is essential for maintaining mean arterial pressure and autoregulating flow regionally to different tissues depending on metabolic demands. Endothelial cells synthesize nitric oxide from l -arginine by the action of the Ca ^{2 +} -dependent, endothelial-specific nitric oxide synthase isoform in response to changes in blood flow. Once formed, nitric oxide activates soluble guanylate cyclase located in adjacent vascular smooth muscle cells, leading to increased cyclic guanosine monophosphate levels, smooth muscle cell relaxation, and functional vasodilation. This process is dependent on intact vascular endothelium and is a defining feature of normal endothelial function. Endothelial dysfunction, among the earliest features of atherosclerosis, particularly in atherosclerosis due to diabetes and insulin resistance, manifests as loss of flow-mediated vasodilation, which can be measured noninvasively by brachial artery ultrasound. Endothelial cells produce other important vasoactive mediators, including prostacyclin and endothelium-derived hyperpolarizing factor, that couple tissue blood flow to metabolic demands. In response to stimuli, including proinflammatory cytokines, the vascular endothelium also elaborates vasoconstrictors, including endothelin-1, angiotensin-II, thromboxane A ₂ , and isoprostanes, that increase vascular tone, permeability, hemostasis, and inflammation. The balance of these vasodilator and vasoconstrictor factors is pivotal for maintaining arteriolar resistance and establishing mean arterial blood pressure. Nitric oxide also reduces platelet aggregation and leukocyte adhesion, thereby suppressing endogenous thrombus formation, maintaining blood rheology, and suppressing leukocyte accumulation in the vessel wall. As discussed further later, extensive evidence implicates shifts in all of these components of normal endothelial function in the setting of diabetes and its associated abnormalities.

In addition to helping control vascular homeostasis, the endothelium also plays a part in host response to inflammation by regulating leukocyte trafficking to sites of injury. Proinflammatory signals including interleukin-1β, tumor necrosis factor–α, and oxidized LDL-C induce endothelial cell expression of genes involved in leukocyte homing and diapedesis, a multistep and orchestrated process collectively known as the leukocyte adhesion cascade . The induction of specific endothelial gene networks, composed of key mediators of leukocyte adhesion, such as E-selectin, P-selectin, vascular adhesion molecule-1, and intercellular adhesion molecule-1, along with chemoattractants such as interleukin-8 and monocyte chemoattractant protein-1, coordinates leukocyte rolling, firm adhesion to endothelial cells, and transmigration into the vessel wall. The ability of neutrophils, monocytes, and lymphocytes to attract to local areas of inflammation is vital to host defense during acute inflammatory responses. However, this endothelial activation becomes maladaptive in chronic states of inflammation, such as atherosclerosis, enabling monocytes or other immune cells to accumulate within the vessel wall and propagate atherosclerotic plaques as well as plaque rupture. Of note, most of the metabolic abnormalities associated with T1D, including hyperglycemia, elevated free fatty acids, hypertriglyceridemia, and hypertension, have all been linked to an activated endothelial state.

Endothelial dysfunction is a defining feature of early atherosclerosis in individuals with T1D and also occurs in the presence of other traditional cardiovascular risk factors such as hypertension and hyperlipidemia. Mechanistically, endothelial dysfunction results from a loss of nitric oxide bioavailability, which can occur through impaired production by endothelial nitric oxide synthase or increased degradation. Consequently, the athero-protective effects of nitric oxide including vasodilation, inhibition of thrombosis formation or aggregation, and suppression of leukocyte adhesion to the vessel wall are lost. There are multiple metabolic derangements common to T1D and metabolic syndrome including insulin resistance, hyperglycemia, high circulating free fatty acid concentrations, and elevated reactive oxygen species that all contribute to loss of nitric oxide and endothelial dysfunction in people with T1D.

Infusion of free fatty acids has been shown to reduce endothelium-dependent vasodilation in animal models and in humans. Free fatty acids activate protein kinase C, driving signal transduction pathways that reduce nitric oxide production by endothelial nitric oxide synthase. The accumulation of lipids in tissues and cells including free fatty acids, fatty acyl-coenzyme As, and others such as diacylglycerols is termed lipotoxicity because of the effect that these lipid mediators have on intracellular signal transduction pathways including insulin. Another major cause of reduced nitric oxide bioavailability is the formation of peroxynitrite (ONOO−) through the reaction of nitric oxide with superoxide anion. High intracellular free fatty acids result in uncoupling of fatty acid oxidation, which increases levels of free radicals such as superoxide anion (O ₂ ^−• ). Normally, superoxide is rapidly removed by scavenging enzymes such as superoxide dismutase. When superoxide anion levels rise, as occurs in people with diabetes in response to elevated free fatty acids and hyperglycemia, peroxynitrite is formed nonenzymatically at high levels. Other enzymes that increase superoxide, including nicotinamide adenine dinucleotide phosphate, reduced form nicotinamide adenine dinucleotide phosphate oxidases, and xanthine oxidases, can also indirectly promote formation of peroxynitrite in the setting of diabetes. Once generated, peroxynitrite fosters endothelial dysfunction and vascular disease in several postulated ways. Peroxynitrite can trigger apoptosis and cell death in endothelial cells and vascular smooth muscle cells, induce endothelial adhesion molecule expression, and disrupt the endothelial glycocalyx. In addition, peroxynitrite-dependent oxidation of tetrahydrobiopterin, a critical cofactor for endothelial nitric oxide synthase function, uncouples endothelial nitric oxide synthase, leading to production of superoxide instead of nitric oxide. Lastly, reactive oxygen species can enhance proinflammatory gene expression leading to endothelial activation.

Recent preclinical work provides another perspective on endothelial dysfunction, namely that this dysfunction may extend beyond altered vasomotor function to include changes in metabolism. Several recent studies have reported pathway alterations that result in changes in glucose and free fatty acid handling. For example, a loss of PPAR-γ in the endothelium changes lipid metabolism, free fatty acid levels, and adiposity with concomitant changes in insulin sensitivity. Other work has found that regulation of insulin receptor adaptor proteins in the endothelium (IRS1 and IRS2) by the Forkhead box proteins, a family of DNA binding transcription factors that regulate expression of genes involved in growth, proliferation, and metabolism, can mediate atherogenesis. In these studies, deletion of the three genes encoding Forkhead box protein isoforms conditionally in the endothelium protects against atherosclerosis while also promoting hepatic insulin sensitivity. By establishing a role for endothelial cells directly in metabolism, these and other studies force a broader definition of what endothelial dysfunction might represent. Furthermore, these observations link to clinical studies which have identified changes in the endothelium as an early and important aspect of diabetes complication development, far earlier than clinically evident atherosclerosis.

Further evaluation of the effects of T1D on endothelial function is contingent on advancing our ability to measure endothelial function, either invasively in the cardiac catheterization laboratory or noninvasively with techniques such as brachial artery ultrasound. These techniques build on the seminal observation that after removal of the endothelium in arterial preparations, the normal vasodilatory response to substances such as acetylcholine becomes paradoxical, resulting in vasoconstriction. Quantitative coronary angiography can document the change in vascular diameter in response to acetylcholine (or bradykinin, substance P, or serotonin). In people with endothelial dysfunction, the vasodilator response to acetylcholine is blunted or results in paradoxical vasoconstriction. In brachial artery ultrasound, forearm blood flow is occluded for 5 minutes with use of a sphygmomanometer and subsequently maintained at constant pressure. After release of the cuff, reactive hyperemia ensues, leading to endothelium-dependent, flow-mediated nitric oxide production and vasodilation, measured by an increase in artery diameter on ultrasound. In individuals with endothelial dysfunction, these responses are severely blunted.

Elevated blood pressure and hypertension are strong risk factors for both microvascular complications and macrovascular complications of diabetes including atherosclerotic cardiovascular disease and heart failure. Atherosclerotic cardiovascular disease, a group of cardiovascular complications including acute coronary syndrome, coronary or arterial revascularization, angina, transient ischemic attack, myocardial infarction, and peripheral arterial disease that develop secondary to atherosclerosis, is the leading cause of morbidity and mortality in people with T1D and T2D and is directly related to elevations in average blood pressure. Notably, numerous studies have demonstrated strong associations with the consistent use of antihypertensive therapy and a reduction in micro- and macrovascular complications, including atherosclerotic cardiovascular disease and heart failure.

The American Diabetes Association recommends a goal blood pressure of <140/90 mm Hg for individuals with both diabetes and high blood pressure, with a stricter goal of <130/80 mm Hg for people with additional risk factors that place them at higher risk of cardiovascular disease. Initial antihypertensive treatments include angiotensin-converting enzyme inhibitors (ACEis), angiotensin II receptor blockers (ARBs), thiazide diuretics, or Ca ²⁺ channel blockers, with an emphasis on personalized medicine unless present with other indicators of diabetic kidney disease, when ACEis and ARBs are preferred. Albuminuria and impaired glomerular filtration rate are each independent risk factors for cardiovascular disease; thus it is imperative to target treatment of both hypertension and diabetic kidney disease with complementary medications, if possible.

Stricter blood pressure goals for the prevention of future cardiovascular disease may be indicated in the setting of youth-onset diabetes due to a combination of factors including a more aggressive diabetes phenotype and longer diabetes duration. In the “Pittsburgh Epidemiology of Diabetes Complications Study,” a 10-year prospective study in adults with T1D of childhood onset (<17 years of age), the goal blood pressure range for adults with T1D was determined to be a systolic blood pressure <120 mm Hg and a diastolic blood pressure <80 mm Hg due to significantly elevated age-adjusted relative risk for mortality, coronary artery disease, lower extremity arterial disease, diabetic kidney disease, neuropathy, and retinopathy.

Metformin is an oral antihyperglycemic agent from the biguanide class that acts through both inhibition of complex 1 in the mitochondrial electron transport chain to activate 5ʹ-adenosine monophosphate-activated protein kinase dependent and independent mechanisms in a tissue-specific manner to produce a variety of effects that are beneficial for individuals with diabetes. Available since the 1950s, metformin remains a first-line therapy for the treatment of T2D due to important systemic effects including inhibition of intestinal glucose absorption, suppression of gluconeogenesis in the liver, reduction of hepatic glucose output, facilitation of glucose uptake into the tissues, and improvement in insulin sensitivity and inflammation. Metformin has been shown to consistently improve body composition and reduce the risk for cardiovascular disease in adults with T2D; yet, less is known about the effects of metformin as an ancillary agent in T1D.

Among 50 participants aged 12 to 21 years with T1D in the randomized, placebo-controlled “Effects of MEtformin on cardiovasculaR function in AdoLescents with type 1 Diabetes (EMERALD)” study, metformin significantly attenuated vascular abnormalities detected with phase-contrast and 4-D flow MRI, including reductions in ascending aorta pulse wave velocity, wall shear stress, and carotid intima-media thickness-assessed atherosclerosis. These findings are clinically significant, as arterial stiffness and carotid intima-media thickness predict mortality in T1D. Additionally, reduced left ventricular contractile efficiency and discoordinated myocardial relaxation were identified in adolescents and young adults with T1D versus healthy controls along with improvements in left ventricular end-systolic and end-diastolic volume, dissynchrony, and longitudinal strain. Moreover, metformin significantly attenuated insulin resistance by hyperinsulinemic-euglycemic clamp. However, it is notable that the effects of metformin on BMI and insulin resistance were modest and glycemia did not significantly improve in T1D. In the “Reversing with Metformin Vascular Adverse Lesions (REMOVAL)” trial, a randomized, double-blind, placebo-controlled study of metformin in 428 adults with T1D with a median follow-up duration of 5 years, atherosclerosis progression, as measured by averaged maximal carotid intima-media thickness, was significantly reduced with metformin (–0.013 mm/year,–0.024 to–0.003 mm/year, P =.0093). However, it is notable that the hemoglobin A1c was minimally reduced in the metformin group versus placebo and the effect was not sustained beyond the initial 3-month treatment period. Taken together, the cardioprotective effects of metformin appear to be separate from effects on glycemia and warrant further study.

Bromocriptine, a dopamine agonist that has been widely used in the treatment of both Parkinson’s disease and prolactin-secreting adenomas, was approved by the Food and Drug Administration in 2009 for the treatment of T2D in a quick release formulation. Postulated mechanisms for the positive effects of bromocripine quick release on glycemic control in T2D include targeting of the attenuated morning dopaminergic signaling seen in T2D that helps to reset the circadian neuronal activities in the hypothalamus that increase overnight plasma glucose, free fatty acids, and triglycerides. If administered within 2 hours of wakening, bromocriptine quick release is hypothesized to result in a reduction of postprandial glycemia due to suppression of hepatic gluconeogenesis and reduction of sympathetic nervous system and renin-angiotensin system activity and may result in decreased micro- and macrovascular complications.

Bromocriptine quick release warrants further study in the T1D population as the “Cycloset Safety Trial,” a study of 1834 people with well-controlled T2D (hemoglobin A1c ≤7.0%) resulted in a 48% reduction in a composite cardiovascular disease endpoint (i.e., first myocardial infarction, stroke, coronary revascularization, or hospitalization for angina/congestive heart failure) after 12 months of treatment in an intention-to-treat analysis. Additionally, bromocriptine quick release reduced the odds of loss of glycemic control (OR: 0.63 [0.47–0.85]) and the need for therapy intensification to maintain a hemoglobin A1c ≤7.0% (OR: 0.46 [0.31–0.69]). Studies evaluating the application of bromocriptine quick release in people with T1D are ongoing. In a study of 40 adolescents and 40 adults with T1D, a 4-week treatment course of bromocriptine quick release titrated to 3.2 mg once daily was associated with a significant decrease in systolic and diastolic blood pressure in the adolescent population with T1D and a decrease in systolic blood pressure, mean arterial pressure, systemic vascular resistance, and orthostatic drop in blood pressure in the adult population with T1D, despite having a minimal overall effect on glycemia or insulin sensitivity. Analyses of the effects of bromocriptine quick release on central aortic stiffness, as evidenced by changes in aortic MRI, are forthcoming and hold promise that bromocriptine quick release may represent a treatment that could target the prevention of cardiovascular disease independent of effects on both glycemia and insulin sensitivity in T1D.

Sodium glucose cotransporter-2 inhibitor (SGLT2i) is a glucose transporter expressed primarily in the proximal tubules of the renal cortex that is sodium-dependent and serves as the primary site for glucose reabsorption in the kidney. SGLT2is inhibit this protein and represent a therapeutic target for glycemic control in diabetes while having a low risk of hypoglycemia secondary to an intact compensatory mechanism through the use of the SGLT1 glucose transporter located in the late proximal renal tubules which are unaffected by SGLT2is. Due to the prevention of glucose reabsorption in the kidney, SGLT2is have consistently demonstrated a reduction in hemoglobin A1c, a reduction in total daily insulin dose, and weight loss when compared to placebo, which may all translate to prevention against future cardiovascular disease. Treatment with SGLT2is encourages a shift from utilization of carbohydrates for energy to lipids secondary to the reduction of reabsorbed glucose and this results in a combination of lipolysis and weight loss. However, the impact of SGLT2is on insulin resistance and BMI is modest, and importantly, SGLT2is are associated with ketogenesis in the setting of lipolysis and thus portend a higher risk of diabetic ketoacidosis in T1D, a potentially fatal complication. Consequently, SGLT2is are not frequently considered as adjunctive therapies in T1D, particularly in adolescents with T1D who are at higher risk of diabetic ketoacidosis secondary to increased insulin resistance in the setting of puberty and poor baseline glycemic control. In the recently published “Adolescent Type 1 Diabetes Treatment with SGLT2i for Hyperglycemia and Hyperfiltration (ATTEMPT)” trial, a 22-week, double-blind, randomized, placebo-controlled study of dapagliflozin 5 mg once daily in adolescents with T1D, dapagliflozin was associated with a reduction in measured GFR as well as a 0.47% decrease in hemoglobin A1c and a 9.0% increase in glycemic time in range. Furthermore, adverse events were similar between groups and only one episode of mild diabetic ketoacidosis was reported, demonstrating both safety and efficacy of SGLT2i treatment in adolescents with T1D. Further study of SGLT2is with close attention to frequent screening for ketosis is necessary in youth and adults with T1D is needed. Continuous ketone monitoring also holds promise to mitigate the risk of diabetic ketoacidosis and may facilitate safe use of SGLT2is in T1D.

The L-cells of the distal ileum release an incretin hormone called glucagon-like peptide–1 (GLP-1) in response to ingestion of glucose which binds to a 7-transmembrane G-protein-coupled receptor (GLP-1R) and induces a variety of downstream effects in the pancreas, heart, gastrointestinal tract, and kidneys before being quickly degraded over 2 to 3 minutes by dipeptidyl peptidase-4 (DPP-4). While multiple known beneficial effects of GLP-1 are pancreatic in origin and thus are less likely to be stimulated in T1D (i.e., β-cell proliferation, decreased β-cell apoptosis, postprandial secretion of insulin), multiple gastrointestinal, immune, and kidney effects remain in play in T1D. These effects include postprandial secretion of somatostatin, inhibition of glucagon secretion, decreased gastric emptying and motility, decreased gluconeogenesis and steatosis, increased natriuresis, increased lipolysis and glucose uptake in the adipose and muscle, decreased generalized inflammation, increased muscle perfusion, and increased satiety.

In animal models, GLP-1Rs are not only present in the heart and endothelial tissue, but also in the kidney and central nervous system where they may exert direct effects. Consequently, GLP-1RAs hold significant promise in the treatment of diabetes, particularly in the setting of metabolic features including overweight and obesity. Substantial weight loss (i.e., 3–9%) and reduction in insulin resistance have been reported following treatment with GLP-1RAs in adults with T2D. Similar weight loss as well as a reduction in fat mass have been demonstrated in adults with T1D in response to the GLP-1RAs exenatide and liraglutide, without increased frequency of hypoglycemia. Indeed, liraglutide has been shown to decrease hypoglycemic events in T1D, with no increased risk of diabetic ketoacidosis. In adults with T1D, an infusion of GLP-1 in the setting of both hyperglycemia and hypoglycemia has also resulted in attenuation of systemic inflammation, oxidative stress, and endothelial dysfunction, potentially lessening the risk of inflammation-induced cardiovascular disease and diabetic kidney disease progression. While a study of liraglutide in overweight adults with T1D found no significant effect on carotid intima-media thickness and peripheral pulse wave velocity, carotid intima-media thickness may take considerable time to change, and peripheral measures of arterial stiffness may be insensitive to early hemodynamic changes. Despite their minimal effects on glycemia, GLP-1RAs remain a promising avenue for adjunctive therapy for T1D due to their positive effects on weight, inflammation, and endothelial dysfunction, and further studies of cardiovascular function are needed.

Pioglitazone, an oral antihyperglycemic medication from the thiazolidinedione class, works to improve insulin sensitivity through activation of the λ isoform of the peroxisome proliferator-activated receptor to increase peripheral glucose utilization at the level of the muscle, adipose, and liver, and decrease hepatic gluconeogenesis. To date, the effects of pioglitazone on glycemia and weight in T1D are mixed. In a small cohort of 35 pubertal adolescents with T1D, a 6-month course of adjunctive pioglitazone titrated to 30 mg once daily was not associated with a significant difference in hemoglobin A1c or total daily insulin dose when compared to the placebo, but it was associated with a significant increase in BMI z-score (pioglitazone: 0.3 ± 0.3 kg/m ² [95% CI,–1.4, +1.9 kg/m ² ] vs. placebo: 0 ± 0.3 kg/m ² [95% CI,–1.4, +1.4 kg/m ² ], P =.01). In contrast, a study in 60 lean (BMI 18–24.9 kg/m ² ) youth with T1D showed a significant decrease in hemoglobin A1c in the pioglitazone treated group versus placebo after 6 months of therapy (–0.22 ± 0.29 vs.–0.06 ± 0.49, P =.03) with no change in body weight, BMI, insulin requirement, blood pressure, or lipids over time. Rosiglitazone, another drug in the thiazolidinedione class, has been associated with a significant decrease in total daily insulin dose (–5.8% vs.–9.4%, P =.02) and an increase in serum adiponectin (+0.05 mmol/L vs. 0.0 mmol/L, P =.02), but has shown no change in hemoglobin A1c or BMI in adolescents with T1D. Additional cardiovascular effects of adjunctive treatment with thiazolidinediones remain largely unstudied in individuals with T1D and warrant further evaluation.

ACEis modulate blood pressure by inhibiting the conversion of angiotensin I to angiotensin II at the level of the angiotensin-converting enzyme. Angiotensin II plays an important role in blood pressure regulation through activation of the angiotensin II type 1 receptor which promotes vasoconstriction, sodium and water retention, sympathetic activation, and cell growth. Interestingly, ACEis also prevent activation of the angiotensin II type 2 receptor and inhibit breakdown of bradykinin which both result in incremental vasodilation and antiproliferation. Higher bradykinin concentrations thereby result in increased prostaglandin synthesis and nitric oxide release which have important dilatory effects on the vasculature. ARBs, in contrast, work to displace angiotensin II from the angiotensin II type 1 receptor and in doing so may stimulate the angiotensin II type 2 receptor, thus triggering both vasodilation and natriuresis; however, this effect is considerably less significant than that seen with ACEis. In the setting of T1D and albuminuria, as evidenced by a urine albumin to creatinine ratio ≥30 mg/g, initial treatment should include either an ACEi or an ARB to reduce the risk for progression of diabetic kidney disease as this is a strong risk factor for future cardiovascular disease. In the “ACE Inhibitors in Diabetic Nephropathy Trial,” a meta-analysis of 12 trials involving 698 individuals with T1D and microalbuminuria, treatment with an ACEi reduced the odds of progression to macroalbuminuria (i.e., urine albumin to creatinine ratio ≥300 mg/g) (OR 0.38 [95% CI, 0.25–0.57] ) and increased the odds of regression to normalbuminuria (OR 3.07 [95% CI, 2.15–4.44]), with a significant decrease in urine albumin excretion by 50.5% [95% CI, 29.2–65.5%] in individuals receiving treatment with an ACEi versus placebo ( P <.001). In the absence of early diabetic kidney disease, treatment with ACEis and/or ARBs for blood pressure modification is equivalent to other antihypertensive agents.

Thiazide diuretics promote natriuresis and diuresis through inhibition of the sodium/chloride channel in the proximal segment of the distal convoluted tubule in the nephron, resulting in prevention of up to 3% to 5% reabsorption of luminal sodium. In the “Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack (ALLHAT)” trial, there was no evidence of treatment superiority with thiazide diuretics versus ACEis versus Ca ²⁺ channel blockers for treatment of hypertension in individuals with diabetes, impaired fasting glucose, or normoglycemia. Similar rates of cardiovascular events were also seen in individuals with diabetes receiving treatment with thiazide diuretics when compared to either ACEis or Ca ²⁺ channel blockers. However, potential side effects of treatment should be considered when selecting a thiazide diuretic as a primary or secondary agent for blood pressure management. Thiazide diuretics exhibit a well-characterized relationship with hypokalemia due to increased renal potassium excretion as a function of effects on the sodium/chloride channel in the distal convoluted tubule of the nephron. It is estimated that up to 50% of individuals receiving treatment with thiazide diuretics will develop some degree of hypokalemia while on treatment. Additionally, at the level of the β-cells in the pancreas, hypokalemia also causes cellular hyperpolarization and this subsequently results in decreased insulin secretion and potential hyperglycemia. This effect is most important to consider in the setting of T2D and early T1D when residual β-cell function and associated insulin secretion is retained. Furthermore, historical observational studies have proposed an association between diuretic-induced hypokalemia and cardiac arrhythmias, although this has not been conclusively proven and is likely a dose-dependent effect.

The term “β-blockers” refers to a group of drugs which act to block the action of endogenous catecholamines on multiple different types of β-adrenergic receptors which function as a part of the autonomic nervous system. First-generation β-blockers are nonselective and exhibit affinity for both the β1 and β2 receptors (i.e., propranolol) while second-generation β-blockers more selectively block the β1 receptor over the β2 receptor (i.e., atenolol). Third-generation β-blockers are further advanced and demonstrate more intrinsic vasodilation (i.e., nebivolol). β-blockers have been associated with long-term improvements in both mortality and cardiovascular disease in people with either heart failure or acute myocardial infarction but a reduction in mortality has not been shown in the absence of these conditions. In the “Action to Control Cardiovascular Risk in Diabetes (ACCORD)” trial, cardiovascular event rates in people with T2D on β-blockers were significantly lower in people receiving intensive therapy (i.e., metformin, short- and long-acting insulin, sulfonylureas, acarbose, meglitinides, thiazolidinediones) to target a hemoglobin A1c <6.0% versus people on standard therapy to target a hemoglobin A1c 7.0% to 7.9%. In that same study, targeting a systolic blood pressure of <120 mm Hg did not reduce the rate of a composite cardiovascular outcome after 4.7 years of follow-up as compared to targeting a systolic blood pressure of <140 mm Hg after a mean follow-up of 4.7 years in people with T2D at high risk for cardiovascular disease.

However, β-blockers have been reported to increase the risk for severe hypoglycemia by blunting the adrenergic symptoms that are earliest warning sign of impending hypoglycemia. Nonselective β-blockers act on the β2 receptor, thereby blocking catecholamine-induced arterial vasodilation and allowing α-stimulation that is unopposed in the setting of hypoglycemia. However, β1-selective blockade in people with diabetes has been shown to prevent the hypoglycemia-induced impairment in autonomic response that follows a hypoglycemic event and β-blockers decrease the risk of hypoglycemia-associated cardiac arrhythmias and death. Further studies evaluating the use of β-blockers in individuals with T1D are needed.

Calcium channel blockers are a group of heterogeneous organic compounds that block the calcium channel, thereby preventing the transit of Ca ²⁺ into the cell and thus reducing subsequent cellular excitability. Ca ²⁺ channel blockers can be stratified into the following three categories: (1) dihydropyridinic agents which act as vasodilators in the periphery, (2) phenilalchilaminic agents which are cardiac negative inotropes and chronotropes, and (3) benzothiazepinic agents which are intermediate in their action profile. An abundance of randomized controlled trials have demonstrated the safety and efficacy of dihydropyridinic Ca ²⁺ channel blockers in the management of hypertension and the prevention of cardiovascular events, further supporting the use of Ca ²⁺ channel blockers as a first-line therapy in individuals with T1D and an ideal agent in dual and triple combination therapy regimens. However, Ca ²⁺ channel blockers have also been shown to have blood pressure–independent effects on glycemic control in people with T1D, particularly in the earliest phase of T1D after diagnosis. In a phase 2, randomized, placebo-controlled trial of individuals with recently diagnosed T1D, administration of a 12-month course of the Ca ²⁺ channel blocker verapamil resulted in improvements in area under the curve for a 2-hour mixed-meal tolerance test c-peptide, a marker of endogenous β-cell function, as well as lower insulin requirements and decreased hypoglycemia when compared to placebo. Of note, in this trial, verapamil was not associated with any adverse events or episodes of hypotension. To further evaluate the molecular basis for these glycemic changes in T1D, a study of human islet cells and murine models for T1D showed that verapamil inhibited the expression of proapoptotic β-cell thioredoxin-interacting protein (TXNIP) in human islets and INS-1 cells to improve the survival and function of residual β-cells. Additionally, a global serum proteomics analysis in T1D identified the T1D autoantigen chromogranin A as the primary protein that is altered by verapamil to reduce elevations in circulating proinflammatory T-follicular-helper cell markers, thereby promoting β-cell health and function and delaying T1D progression for the duration of treatment. Consequently, chromogranin A remains a potential therapeutic target to further prevent T1D progression and verapamil may represent an important avenue for both T1D prevention and delay of progression.

Statins primarily work by inhibiting the hydroxymethylglutaryl coenzyme A (HMG-CoA), the first and rate-limiting step of the cholesterol biosynthesis pathway, thereby preventing HMG-CoA from converting into mevalonic acid. Not only does inhibition of this step in the cholesterol biosynthesis pathway result in 20% to 55% reductions in serum LDL-C concentrations, it also upregulates cell surface LDL-C receptor expression to promote LDL-C clearance and may also improve endothelial cell function, decrease generalized inflammation, stabilize atherosclerotic plaques, exhibit antithrombotic effects, and reduce the risk for dementia secondary to inhibition of additional isoprenoid intermediates as a part of the mevalonate pathway.

The American Diabetes Association recommends consideration of moderate-dose statin therapy, in addition to lifestyle modification, for all individuals with T1D aged 40 years and older, and in individuals less than 40 years of age in the setting of cardiovascular risk factors. High-dose statin therapy is recommended for any individual, regardless of age, in the setting of known cardiovascular disease. The American Heart Association mirrors these recommendations and recommends moderate-intensity statin therapy in all individuals with diabetes who are age 40 to 75 years of age and have an LDL-C of ≥70 mg/dL, with a titration to a high-intensity statin to reduce the LDL-C by ≥50% in people with diabetes who are at high risk. T1D is considered a “risk enhancer” for cardiovascular disease for all individuals who have had diabetes for ≥20 years with at least one additional risk factor for cardiovascular disease and initiation of a statin for cardiovascular protection should be considered at an earlier age in this population. Of note, analyses suggest a 22% relative risk reduction in major vascular and coronary artery events with every 38.7 mg/dL (1 mmol/L) reduction in LDL-C in all adults, making initiation and maintenance of statin therapy an excellent tool in the prevention of cardiovascular disease in T1D, despite the risk for potential side effects, including arthralgias, myalgias, myopathy, myositis, rhabdomyolysis, etc.

Fibrates, lipid-lowering agents used to treat hypertriglyceridemia, are thought to work as PPAR-α agonists. Of note, other endogenous lipolytic pathways including adipose tissue triglyceride lipase and hepatic lipase as well as fatty acid synthase can generate PPAR ligands in different physiologic contexts as well. These lines of evidence suggest that in diabetes, loss of endogenous lipoprotein lipase action decreases activation of the PPAR-α–regulated gene cassette, which would be predicted to result in decreased expression of apolipoprotein (apo) A-I, which is involved in HDL function, and increased endothelial inflammation. It is important to note that fibrates, as synthetic PPAR-α agonists, may not faithfully replicate cellular responses to natural PPAR-α ligands. Of interest, the potential role of lipoprotein lipase has expanded to include other proteins involved in lipoprotein lipase action. For example, apolipoprotein C-III is an endogenous inhibitor of lipoprotein lipase activity. Recent studies implicate apo C-III in promoting proatherogenic, proinflammatory responses, which may occur through various mechanisms, including potential modulation of endogenous PPAR responses, as outlined previously, as well as other means.