Diabetes is a growing worldwide pandemic: 415 million people were living with this disease in 2015, of whom 193 million were estimated to be undiagnosed.1 Type 2 diabetes (T2D) accounts for approximately 90% of all diabetes cases and is known to impose a heavy burden on healthcare services. As a multifactorial disease, T2D arises in a favorable genetic and epigenetic background in the presence of the appropriate environmental drivers, many of which are characteristic of the lifestyle of modern societies: physical inactivity, excessive sugar and fat consumption, obesity and aging.1,2

Importantly, the main cause of morbidity and mortality among people with T2D is cardiovascular disease (CVD).1,3 The relative risk for macrovascular disorders in people with diabetes is two to four times higher than that of the general population, while the relative risk for microvascular disorders is 10–20 times higher.2,4,5 Common T2D complications such as peripheral artery disease and heart failure are often underdiagnosed and undertreated,6,7 although vascular damage begins even before diagnosis in the early stages of dysglycemia.8 Moreover, mortality and the outcomes of vascular events are worse in the diabetic population than in people without diabetes.2,9,10 The reason for these trends lies mainly in the endothelial dysfunction caused by hyperglycemia, which in turn induces atherosclerosis and leads to widespread vascular lesions.4,10 Of note, a substantial proportion of the medical expenditure associated with T2D is attributable to vascular complications.4,9 The total cost of treating a T2D patient with CV complications in the USA has been estimated to be 58% higher than that of treating a T2D patient without CV complications.9 For the above reasons, detection and control of cardiometabolic risk factors among people with T2D is of utmost importance to prevent disease progression, establishment of CVD and premature death, as well as to relieve the burden on healthcare systems.

In Portugal, a study published in 2015 including 1528 T2D patients outlined the prevalence of CVD and CV risk factors among patients treated in a primary care setting: almost 15% of patients had CVD, the most prevalent of which were myocardial infarction (4.4%), angina pectoris (4.3%) and stroke (4.0%).11 Additionally, 80.3% of all patients had hypertension, while 61.6% had hypercholesterolemia.11 The present study aimed to extend and update existing knowledge by addressing the prevalence of CVD and CV risk factors among T2D patients being treated in a hospital setting. Additionally, a retrospective overview of the three years prior to inclusion was undertaken in order to assess the clinical management and resulting outcomes among patients with established CVD.

Cardiovascular disease has a major impact on T2D patients’ morbidity and mortality, and accounts for a substantial part of the healthcare costs associated with the management of these patients. Control of CV risk factors among T2D patients is thus of utmost importance. In this study we assessed the prevalence of CVD morbidity and CV risk factors among a population of T2D patients being followed in a Portuguese hospital setting. We also reviewed the clinical management of those with CVD three years prior to study inclusion, whenever data were available. Of the 715 patients enrolled in this study, 40.0% had been diagnosed with CVD, a clearly higher prevalence than the previously published figure of 15% of Portuguese T2D patients being followed in a primary care setting.11 While this result is to some extent to be expected, as patients followed in a hospital setting tend to have more comorbidities and to be at a more advanced stage of the disease, other authors have also reported lower CVD prevalence rates. For instance, macrovascular complications were present in 12.7% of the 15992 patients from 38 countries enrolled in the DISCOVER study (which included patients initiating second-line therapy and being treated in primary and specialist healthcare settings), 32% of the 248400 patients enrolled in a study based on a national registry in Scotland (Scottish Care Information-Diabetes), 25% of the 91320 patients enrolled in the CVD-REAL Nordic study (who were new users either of SGLT-2 inhibitors or of other glucose-lowering drugs), 28% of the 51286 patients from eight national cohorts enrolled in the IMPROVE study (all of whom required insulin), and 27.2% of the 66726 patients from 28 countries enrolled in the A1chieve study (also insulin users).14–18 Of note, differences in the definitions of CVD used in the different studies cannot be excluded and could partially explain the high CVD prevalence found in our sample. In fact, McGurnaghan et al. stated that the prevalence of CVD in their sample varied from 28% to 44% according to the definition used.15

Another important factor that could have a crucial impact on CVD prevalence is disease duration, which is associated with a higher incidence of complications. A study involving 3130 T2D patients being followed in primary care centers showed that, after 20 years of diabetes duration, 46% of patients presented at least one macrovascular complication.19 Importantly, the mean disease duration of the patients included in this study was longer than that reported in the above-mentioned cohorts: 17.4 years vs. a median of 4.1 years in the DISCOVER study, a median of 7.8 years in the Scottish national registry study, an average of 6.9 years in the IMPROVE study, and an average of 8.0 years in the A1chieve study.14,15,17,18 The time elapsed since diagnosis also explains the high percentage of patients treated with insulin (73.7%), as this drug is usually prescribed in later and more severe stages of T2D. Still, the CVD rate found in our sample exceeds those of many international studies – some of which included solely insulin-treated patients – which highlights a worrisome scenario.

All patients in the FAS had at least one CV risk factor, and three of these risk factors were present in over 80% of patients: hypertension (84.4%), obesity and overweight (83.1%), and hypercholesterolemia (80.3%). Unsurprisingly, the prevalence of these risk factors was somewhat lower among Portuguese people with T2D being treated at primary care centers, typically at earlier stages of disease: the study by Cardoso et al. revealed that 80.3% patients had hypertension and 61.6% had hypercholesterolemia.11 However, the percentage of patients with CV risk factors in our sample was also higher than that seen among a population of 735 T2D patients from two Madrid districts being treated by an endocrinologist at the hospital (45% of whom were obese and 40% of whom had high levels of LDL-C and total cholesterol), as well as in 3130 patients attending primary care centers in the Catalonia region (42.9%, 42.3% and 56.2% of whom were hypertensive, obese and had high LDL-C, respectively) and also compared to patients in the Scottish Care Information-Diabetes registry (among whom the rates of hypertension, obesity and hypercholesterolemia were 57.9%, 46.5% and 22.7%, respectively).15,19,20 Nevertheless, the mean HbA1c, BMI, waist circumference, and systolic and diastolic BP found in our sample were within the range of those reported in other studies.14,15,17–22

Interestingly, patients with established CVD had slightly but significantly better mean values of key cardiometabolic risk factors (BMI [p=0.019], systolic BP [p=0.01] and diastolic BP [p<0.001]) than those without CVD. This improvement is likely the result of appropriate medical management once CVD is diagnosed. However, the scenario is different regarding dyslipidemia: in the population with established CVD there is a higher prevalence of low HDL and/or high TG, a phenotype sometimes referred to as atherogenic dyslipidemia.23 These CV risk factors, namely hypertension, hypercholesterolemia, low HDL and high TG, were also significantly associated with CVD in other studies.18,24 In light of this, there is a need to improve the lipid profile of T2D patients with CVD.

The lack of standardized reporting of CVD manifestations by investigators in this study could have resulted in underestimation of CVD, due to potential lack of clinical suspicion. Nevertheless, underdiagnosed conditions such as peripheral or carotid artery disease (29.5%) and heart failure (22.7%) often appear as the second and third most common manifestations of CVD in T2D patients in Portugal, according to our findings. The current paradigm shift with disease-modifying drugs available for both conditions highlights the need to lower the threshold to look for these diseases and optimize their medical treatment.25,26

In our study, data from the previous three years were only available for 191 of the 286 T2D patients with established CVD. This number is below the projected population size (430) that would allow precise estimates with 95% confidence intervals concerning the secondary objective, even when this number is subsequently adjusted taking into account the prevalence of CVD. As such, the trends identified in the follow-up subgroup may lack statistical precision and should be interpreted with caution.

The incidence of hypoglycemic episodes remained approximately stable throughout the three-year period analyzed; however, while mean HbA1c levels saw no significant changes, the incidence of hyperglycemic episodes rose from 12.8% to 22.5%. The percentage of patients experiencing diabetes complications presented a similar trend, increasing from 64.4% to 72.0%. While this could be related to the natural progression of T2D, following the expected progressive loss of beta-cell function, the possibility cannot be excluded that lack of therapeutic optimization or low adherence to medication may have played a role in this dynamic, at least for some patients. Moreover, and unlike the increase in the frequency of complications, the number of hospitalization episodes and hospital appointments remained relatively stable throughout the 36 months before the beginning of the study, suggesting suboptimal utilization of healthcare resources. By contrast, with regard to cardiometabolic risk factors, small but significant improvements were seen in total cholesterol levels, waist circumference, BMI and diastolic BP. Together with the absence of deterioration in other parameters – especially eGFR – this suggests appropriate clinical management of CVD patients, particularly for preservation of renal function and promotion of a healthy lifestyle.

It is worth mentioning that the mean HbA1c level in our cohort (7.7±1.3%) indicates inadequate glycemic control in at least some patients. Although the latest clinical guidelines recommend that the glycemic goal should be individualized taking into account various patient and disease features, a target of <7% is considered the gold standard.27 Despite the large proportion of patients in our sample already treated with insulin, the high HbA1c values suggest that there is room for improvement regarding therapeutic optimization and intensification.

According to the most recent American Diabetes Association (ADA) and European Association for the Study of Diabetes (EASD) consensus, treatment with GLP-1 receptor agonists or SGLT2 inhibitors is recommended in T2D patients with established CVD, regardless of HbA1c levels, due to the CV benefits some drugs of these classes have demonstrated.22 Importantly, 84 (53.8%) of the follow-up group treated with antidiabetic drugs in our study were using GLP-1 receptor agonists and/or SGLT2 inhibitors. Although this demonstrates rapid adoption of these recommendations within the hospital setting, almost half of these CVD patients could still benefit from an antidiabetic therapy with proven CV benefits.

Our study has several potential limitations that should be considered when interpreting its findings. One is the lack of information concerning patients’ lifestyle habits (especially alcohol consumption and regular exercise) and socioeconomic data (such as income and occupational status), which may amount to residual confounding factors. Moreover, we did not define thresholds for CV risk factors, and so their reporting was based on physicians’ clinical judgment, and some degree of heterogeneity is likely to be present.

Overall, our study detected a high prevalence of CVD and CV risk factors among a cohort of Portuguese T2D patients being followed in a hospital setting. While the retrospective analysis of a small subset of patients with established CVD showed small but significant improvements in key cardiometabolic indicators, the overall scenario suggests the need to detect and manage CV risk factors in T2D patients who are not yet diagnosed with CVD, in order to prevent these complications. Optimized use of antidiabetic drugs with proven CV benefits could also help to improve outcomes in patients with established CVD and to prevent a primary event in patients with CV risk factors. Better diabetes control (including individualized therapeutic optimization and routine reassessment), as well as more rational use of healthcare resources, should also lead to better HbA1c control and thereby prevent CVD and improve the clinical outcomes of T2D patients.

H.C. reports consultancy work for Lilly. C.T.B. reports speaker fees from Merck Sharp & Dohme, Bial and consultancy work for Medinfar. L.A. reports membership of the advisory board and consultancy/speaker fees from AstraZeneca, Lilly, Boehringer-Ingelheim, Novo Nordisk and Bial.

F.S.R. reports consultancy work for Medinfar. J.L. reports speaker fees from Merck Sharp & Dohme, AstraZeneca, Novo Nordisk, Lilly, Medinfar and Boehringer-Ingelheim. C.N. reports speaker fees from Ascensia and Novartis. E.R. reports speaker fees from Merck Sharp & Dohme. N.B.V. reports speaker fees from AstraZeneca, Lilly, Bial, Servier, Merck and Merck Sharp & Dohme. T.C. is an employee of Novo Nordisk.

All the authors participated in the drafting, critical revision, and approval of the final version of the manuscript. T.C. was involved in study design. H.C., C.T.B., L.A., F.S.R., J.L., C.N., E.R. and N.B.V. were investigators in the study. All the authors were involved in the interpretation of study results.