A retrospective study was carried out using an unpaired case-control design.

For the cases, patients diagnosed with type 2 diabetes mellitus who were hospitalized in the internal medicine services of a hospital in Peru between 2012 and 2021 were included. Only patients who had the diagnosis of a stroke confirmed through tomography, aged 18 years or older, and with complete medical records were considered. Patients with a diagnosis of type 1 diabetes mellitus, gestational diabetes, and those with incomplete medical records were excluded.

For the controls, patients with a diagnosis of type 2 diabetes mellitus who were hospitalized in the internal medicine services in a hospital in Peru during the same period, without a clinical-radiological diagnosis of a stroke (through computed tomography scan) who were 18 years of age or older and had medical records in which the variables to be evaluated could be accurately determined were considered. Exclusion criteria were medical records of patients not being hospitalized in the specified services during the study period and those not adequately recorded.

Both the case and control groups were selected from the same study population. Controls were randomly selected in an unpaired and simple randomized manner.

The data for both clinical and laboratory variables were taken from the initial evaluations conducted upon hospital admission and recorded in the patients’ medical records at the time of the study event.

The variables considered as risk factors were sociodemographic factors (age greater than or equal to 60 years [9] and male or female gender), history of acute myocardial infarction, the presence of chronic kidney disease (CKD) (defined by a glomerular filtration rate of less than 60 ml/min/1.73 m²), elevated levels of glycated hemoglobin (HbA1c)≥8% (64 mmol/mol) [10] presence of arterial hypertension (HTN) (defined by more than 140 mmHg systolic blood pressure and/or more than 90 mmHg diastolic blood pressure according to the International Society of Hypertension [11], presence of overweight (Body Mass Index (BMI) between 25 to 29.9 kg/m²), obesity (BMI of 30 or more kg/m²), smoking, dyslipidemias (elevated levels of total cholesterol, triglycerides, LDL, and decreased HDL according to Fredrickson’s classification [12]) and elevated levels of C-reactive protein (CRP)) defined by more than 10 mg/dl [13].

The population consisted of patients hospitalized in the internal medicine services of a hospital in Peru during the 2012-2021 period.

The sample consisted of 324 patients, of which 108 corresponded to cases and 216 to controls. The sample size was calculated using a formula for unpaired case-control studies, with a frequency of exposure among controls of 50%, based on a study by Ohishi [14]. The predicted Odds Ratio for the HTN variable was 1.96 (95% CI 1.097-3.501, P=0.023) [15]. The confidence level was estimated at 95%, and the number of controls per case was set at two. The sampling method was probabilistic.

A Microsoft Excel database was created to organize the information, and the statistical analysis was performed using SPSS version 26. Initially, a descriptive analysis was conducted to establish the frequency of presentation of each study, which was variable to, at a later stage, perform data analysis using contingency tables to calculate the Odds Ratio with 95% confidence intervals and the evaluation of the risk relationship between study variables.

For the bivariate inferential analysis, the Chi-square test was applied to analyze qualitative variables and the presence of stroke, taking into consideration the value of p <0.05 as statistically significant. Logistic regression was used for multivariate analysis.

According to the data analyzed, 60 years of age or older was identified as a risk factor for developing strokes in diabetic patients. This is supported by another case-control study conducted in the general population, which found that being 60 years of age or older carried the same risk [19]. Thus, it can be stated that older age is generally associated with a higher risk of strokes in the general population without type 2 diabetes mellitus. In diabetic patients, this age difference is attenuated, suggesting that the risk of strokes in people with diabetes may extend even to individuals younger than 60 years, as found by Yen [20].

Furthermore, in a review article evaluating the epidemiology of neurological pathology in diabetic patients, it was stated that the risk of developing it is higher for men between the fifth and sixth decade of their life and women during their sixth decade, which aligns with the results of this study [21].

The pathophysiological mechanism of aging may be due to the alterations suffered by the cerebral circulation in the elderly population (such as increased arterial stiffness and endothelial dysfunction), which increases susceptibility to vascular insufficiency and ischemic injury [22, 23]. Additionally, diabetes promotes a state of cellular senescence that can lead to lipotoxicity and tissue dysfunction, as well as chronic inflammation and degradation of the cellular matrix, which is directly linked to atherosclerosis [24].

It is essential to highlight that in this study, the presence of arterial hypertension was the factor that increased the risk of developing a stroke in diabetic patients by more than 5 times. Likewise, a cross-sectional study with more than 11,000 diabetic individuals found a similar result (OR 2.82; 95%CI 0.29-17.3) [25]. Furthermore, Berenguer et al. [19] found that HTN is associated with the development of such neurological pathology in the general population (OR 6.6; 95% CI 3.05- 14.41).

Lifestyle is strongly related to blood pressure values due to several factors: high salt intake, smoking, reduced physical activity, and an inadequate diet that can lead to overweight, obesity, and of course, type 2 diabetes mellitus [26, 27].

Additionally, it has been suggested that HTN plays a leading role in the promotion of oxidative stress via the production of reactive oxygen species, the dysfunction of baroreceptors, making them less sensitive to their excitatory stimuli (the increase in blood pressure), thus promoting inflammation, via molecules such as CRP, interleukins 1 and 6, leukocyte esterase and intracellular adhesion molecule-1, and thus leading to functional changes of the cerebral vasculature, which generate hypoperfusion and eventually, ischemia [28].

Among the vascular changes associated with hypertension is the deterioration of cerebral blood flow autoregulation, a mechanism that normally allows cerebral blood flow to maintain a stable perfusion pressure despite variations in systemic pressure. In addition, one of the most important consequences of HTN is its effect on cerebral arteries and arterioles, leading to vascular remodeling, increasing their tone and decreasing their lumen, which increases the severity and extent of the penumbra zone during a stroke [29].

On the other hand, HTN provides potent vasoconstrictor stimuli, such as angiotensin II, which acts via its AT1 receptors and causes deleterious effects at different levels; endothelin-1, which, added to the nitric oxide deficit, also induced by HTN, promotes an intense increase in vascular tone; and the activation of endothelial calcium channels and in vascular smooth muscle [29].

The presence of elevated C-reactive protein levels in diabetic patients slightly increases the risk of developing a stroke. This is supported by the findings of a case-control study, where elevated CRP levels were found to increase the risk of developing a stroke in diabetic patients (OR 2.61; 95%CI 2.31-6.42) and in patients without such pathology (OR 2.41; 95%CI 1.93-4.21) [30]. This is in addition to the findings of The Emerging Risks Factors Collaboration consortium, which analyzed over 130 prospective studies [31] and concluded that elevated CRP levels had a relative risk (RR) of 1.46 for the development of ischemic stroke.

Precisely, the role of inflammation lies in its behavior as a precipitating factor of increased blood viscosity, an atherogenic factor in all its phases, both by CRP and by tumor necrosis factor-alpha and interleukin 6 [32]. In response to these stressors, the vascular endothelium produces cell adhesion proteins that facilitate leukocyte adhesion, while tissue macrophages increase the release of cytokines. These molecules will then stimulate the migration of muscle cells from the adventitia to the other arterial layers to form the fibrous cap of the atheromatous plaque, the elementary lesion of atherosclerosis [33].

Elevated HbA1c levels were found to be a risk factor in diabetic patients for the development of strokes in diabetic patients [34]. In the same sense, a cross-sectional analytical study in hospitalized diabetic patients found that individuals with elevated HbA1c levels were six times more likely to develop such a neurological event in diabetic patients (OR 6.29; 95%CI 1.71-23) [35]. In addition, a meta-analysis and systematic review concluded that increased levels of this molecule were associated with the development of the event in patients with the same endocrine pathology (HR 1.19; 95%CI 0.87-1.62) [10]. Likewise, it was found in a cohort study with more than 270,000 diabetic patients. It was found that HbA1c above its target values was the strongest predictor for the development of strokes in diabetic patients [36].

HbA1c is used as an assessment of glycemic control from 8 to 12 weeks before measurement [37]. Therefore, its elevation increases the risk of developing both microvascular and macrovascular complications related to diabetes mellitus [35]. This is relatively frequent since, for instance, in the United States [38] and some Latin American countries [39], these increased HbA1c levels are present in more than 20% of the diabetic population.

On the other hand, the variables studied that did not reach significance after multivariate analysis were dyslipidemia, history of acute myocardial infarction, presence of CKD, overweight, obesity, and smoking. The lack of significance in the case of smoking may be attributed to the fact that stroke had the lowest percentage of events attributable to smoking in the population of the region studied [40]. Regarding the lack of significant association of obesity, it may be because diabetic patients with normal weight could be metabolically obese (i.e., the result of the interaction between hyperinsulinemia, insulin resistance and dyslipidemia) or have insufficient insulin secretion as a result of the progression of the disease, which can lead to catabolism [41].