Correlation of Triglyceride-Rich Lipoprotein Cholesterol and Diabetes Mellitus in Stroke Patients

Background Previous studies have revealed that triglyceride-rich lipoprotein cholesterol (TRL-c) is closely related to diabetes mellitus (DM) in hypertensive patients. However, this relationship in stroke patients has not been reported. The aims of this study are to investigate the relationship between TRL-c and diabetes in adult Chinese stroke. Methods Patients with stroke treated in the Department of Neurology of the Second Affiliated Hospital of Nanchang University from January 2019 to January 2021 were selected. TRL-c was calculated from total cholesterol minus (high-density and low-density lipoprotein). DM was diagnosed based on previous medical history (diagnosed by secondary hospitals or above) and/or current use of hypoglycemic drugs and/or intravenous blood glucose measurement (fasting blood glucose ≥7.0 mmol/L or nonfasting blood glucose > 11.1 mmol/L). The relationship between the TRL-c and DM was determined using multivariate logistic regression, smoothing curve fitting (penalized spline method), and subgroup analysis. Results A total of 890 patients with stroke (age, 66.1 ± 11.8 years) were enrolled, including 329 females. Multivariate logistic regression analysis demonstrated that TRL-c had a positive association with DM (OR 1.88; 95% CI: 1.22 to 2.89). Strong linear associations of TRL-c with DM were confirmed by restricted cubic spline analysis. And the association between TRL-c and DM was consistent in the different subgroups. Conclusion Positive associations were found between TRL-c and DM in patients with stroke.


Introduction
Diabetes mellitus (DM) is a major public health problem in the world. In 2020, the number of adults with DM in the Chinese mainland was estimated to be 129.8 million [1]. In addition, the proportion of undiagnosed DM is also high [2]. Consequently, it is of great signifcance to further explore the risk factors of diabetes for early diagnosis, intervention, and prevention of diabetes. Diabetes is closely related to dyslipidemia, one of its risk factors. Diabetes and dyslipidemia interact and serve as a common risk factor for atherosclerotic disease. A meta-analysis has shown that although statin therapy can reduce low-density lipoprotein cholesterol, the residual risk of atherosclerotic cardiovascular disease (ASCVD) in diabetic patients is high [3]. As a result, triglyceride-rich lipoprotein cholesterol (TRL-c) is an essential factor leading to residual cardiovascular disease risk and is a newly emerging index for cardiovascular disease risk assessment [4]. TRL-c mainly includes very low-density lipoprotein, intermediate-density lipoprotein, and nonfasting chylomicrons. In addition, studies have demonstrated that TRL-c is closely related to insulin resistance [5], obesity [6], and abnormal glucose tolerance [7].
Moreover, a positive correlation between TRL-c and diabetes has been investigated in recently published studies [8,9]. However, the results obtained from another study suggested no correlation [10]. Te correlations reported in relevant studies are diferent and the populations included are only hypertensive patients, females, and public ofcials. In addition, no research has been conducted on stroke patients with ASCVD events. Terefore, this study aims to report the association between TRL-c and diabetes in patients with stroke to provide relevant evidence for secondary diabetes prevention.

Participants. Based on Te China National Stroke
Screening and Prevention Project (CNSSPP) [11]. 987 Patients with stroke were treated in the Department of Neurology of the Second Afliated Hospital of Nanchang University from January 2019 to January 2021 and were retrospectively analyzed in this study. Te inclusion criteria were as follows: (1) patients with stroke (including cerebral infarction, intracerebral hemorrhage, and subarachnoid hemorrhage) as the frst diagnosis and frst onset (the same type of diseases); (2) age ≥18 years; (3) signed informed consent. Te exclusion criteria were as follows: (1) asymptomatic lacunar infarction, multiple lacunar infarctions; (2) previous stroke history (the diferent types of diseases were not excluded, such as patients with previous intracerebral hemorrhage sufering from cerebral infarction could be included); (3) traumatic intracerebral hemorrhage, subdural/epidural hemorrhage, traumatic subarachnoid hemorrhage, unruptured aneurysm, transient ischemic attack, and others. After excluding 48 patients who previously took blood pressure-lowering drugs and 49 patients with TRL-c outliers, 890 patients with acute stroke were included in this study ( Figure 1). Tis study followed the principles of the Declaration of Helsinki and all participants received information on the study and provided written informed consent to participate.

Data Collection.
Te basic data were collected through face-to-face interviews, physical measurements, laboratory examinations, and auxiliary examinations. (1) Basic information: demographic information, contact information, and conditions in this study. Te cause of death and relevant information were registered if the subject died. (2) Lifestyle: stroke risk factors, such as smoking, drinking, lack of exercise, and their changes were determined. (3) History: stroke history, hypertension, diabetes, atrial fbrillation, dyslipidemia, and treatment and control were identifed. (4) Physical examination: height, body weight, and blood pressure. Professionals determined whether the subjects had regular rhythms by using cardiac auscultation. Atrial fbrillation was detected by electrocardiography (ECG) for patients with irregular rhythms. (6) Laboratory examination: fasting plasma glucose, blood lipid, glycosylated hemoglobin, and homocysteine were measured.

Outcome Defnition.
Hypertension was defned as systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg, with a history of hypertension (three times on diferent days), and/or currently using antihypertensive medications [12]. Atrial fbrillation was defned as previous medical history (diagnosed by secondary hospitals or above) and/or ECG displaying the disappearance of the P wave, replaced by an irregular F wave and irregular RR intervals [13]. Smoking status was classifed into "nonsmoker" and "current smoker" (a current smoker is one who smokes at least one cigarette a day for more than six consecutive or cumulative months). Alcohol use status was categorized into "nondrinker" and "current drinker" (current drinker is one who drinks wine ≥3 times/week, ≥ 100 g/ time). TRL-c: TRL-c � total cholesterol (TC)-low-density lipoprotein (LDL)-high-density lipoprotein (HDL) [14].

Defnition of DM and
Stroke. DM was defned as previous medical history (diagnosed by secondary hospitals or above), current use of hypoglycemic drugs, or intravenous blood glucose measurement. (1) Diabetic symptoms: fasting blood glucose ≥7.0 mmol/L (126 mg/dl) or nonfasting blood glucose >11.1 mmol/L. (2) No typical diabetic symptoms: fasting blood glucose ≥7.0 mmol/L at least two times or nonfasting blood glucose >11.1 mmol/L [2]. A stroke was defned as a sudden clinical sign of persistent focal or total neurological defcit due to the blockage or rupture of the arterial system. In addition, this defect has symptoms or signs lasting for more than 24 h (unless death), or CT or Magnetic resonance imaging (MRI) confrmed defcit consistent with acute stroke after excluding other nonvascular causes (such as primary brain tumor, brain metastasis, postseizure paralysis, brain trauma, infection, or other nonischemic causes). It is divided into ischemic stroke and hemorrhagic stroke. Ischemic stroke refers to cerebral infarction, a neurological defcit caused by cerebral vascular stenosis or obstruction. Its symptoms can last at least 24 h, or a new infarct can be confrmed by imaging. It is caused by focal symptoms and signs of the nervous system are consistent with the afected vessels' blood supply. Hemorrhagic stroke, including intracerebral and subarachnoid hemorrhage, damages brain tissue caused by the rupture of the cerebral vessels. Intracerebral hemorrhage is defned as primary nontraumatic intraparenchymal hemorrhage. Subarachnoid hemorrhage refers to the blood fowing into the subarachnoid space after the rupture of blood vessels at the bottom or on the surface of the brain, causing corresponding clinical symptoms [15].

Statistical Analysis.
Te statistical analysis was performed using Empower 2.0 (R) (https://www.empowerstats. com) and the statistical package R (https://www.rproject.org, version 4.1.2). Te measurement data following the normal distribution were expressed as mean ± standard deviation (X ± S) and compared by the t-test between groups. Te measurement data inconsistent with the normal distribution were expressed as the median (P 25 -P 75 ) and compared with the Mann-Whitney test between groups. Te enumeration data were expressed as N(%) and compared using the χ 2 test between groups. Te nonconditional logistic regression models (represent results by odds ratio(OR) and 95% CI) for major covariables adjusted for the main covariates in three models were designed to assess the independent association of the TRL-c level with DM. Te covariates were selected on the basis of statistical signifcance in the univariable analyses, their clinical importance, and the estimated variables change of at least 10% of potential confounding efects. Te penalized spline regression method (a ftted smoothing curve) and the generalized additive model (GAM) conducted the dose-response relationship for the TRL-c level with DM. In addition, we conducted subgroup analyses to explore the potential factors modifying the association. P < 0.05 was considered statistically signifcant.

Baseline Characteristics.
A total of 890 patients with acute stroke (mean age: 66.7 ± 11.8 years; 37.0% females) were enrolled. Tere were 280 diabetic patients, with an incidence of 31.4%. Compared with the non-DM group, patients with DM had high values of SBP, NIHSS score, Fasting plasma glucose, triglyceride, TRL-c, and lower values for current smoking, HDL-c, serum homocysteine, the prevalence of hypertension, atrial fbrillation, antiplatelet drugs, and antihypertensive drugs, as shown in Table 1. Table 2 shows the univariate analysis results of age, gender, smoking, drinking, BMI, SBP, DBP, TRL-c, homocysteine, hypertension, atrial fbrillation, and antihypertensive drugs. We found that smoking, BMI, SBP, TRL-c, homocysteine, history of hypertension, atrial fbrillation, and antihypertensive drugs were signifcantly associated with DM (P < 0.05).  (Figure 2).

Discussion
Te present study found that among the patients in our hospital with a stroke, the incidence of diabetes was 31.5% and TRL-c was positively correlated with DM risk, indicating that TRL-c may be a marker for diabetes risk. Tis new fnding expands the application population compared with previous studies. Even after the occurrence of ASCVD events, TRL-c may provide a new idea for the prevention and treatment of diabetes. Te consensus statement of the European Atherosclerosis Society on triglyceride-rich lipoproteins and their residues in 2021 [16] mentioned that reducing triglyceriderich lipoproteins and their residues is a strategic method for preventing ASCVD. Tere is increasing evidence that a high concentration of TRL-c residues is a cardiovascular disease risk. Over the past 30 years, epidemiological and genetic studies have confrmed that elevated plasma TRL-c and their residues are correlated with ASCVD events [17], including myocardial infarction [18], stroke [19], coronary artery calcifcation [20], aortic stenosis [21] and all-cause death [22,23]. Moreover, some scholars [24] proposed that although it remains uncertain whether it can be used as a monitoring target for drug therapy, it should be detected in cardiovascular health screening and disease diagnosis.
In addition, increasing evidence links TRL-c to DM and other metabolic diseases. In the Woman's Genome Health Study (WGHS) [8], 15,813 participants in the baseline fasting state were analyzed, revealing that at the average 18.6-year follow-up, 1453 new diabetes cases occurred. In the Cox model with adjusted basic covariates, TRL-c signifcantly correlated with diabetes (HR � 1.39, 95% CI 1.28-1.51, P < 0.001). Zhou Wei et al. [9] in China conducted a cross-sectional analysis of the data of 13,721 hypertensive patients without lipid-lowering drugs. Tey found that TRL-c signifcantly and positively correlated with diabetes risk (OR � 1.73, 95% CI: 1.54-1.94). Tis fnding suggested that TRL-c may be a reliable diabetes marker and may provide a new strategy for the prevention and treatment International Journal of Endocrinology  of diabetes, consistent with our results. However, foreign scholars Carvalho et al. [10] included a Brazil Adult Health (ELSA-Brasil) cohort of 4463 participants. Tey showed that 366 new diabetes cases were found after an average of 3.7 years of follow-up. Also, their plasma TRL-c concentrations were not related to new diabetes risk and could not improve the prediction of diabetes. Studies have reported inconsistent conclusions on their correlation, so more studies are needed to evaluate the potential of TRL-c in preventing diabetes. Te mechanisms underlying the relationship between TRL-c and DM are not yet fully understood. DM is related to increased cardiovascular disease risk [25]. Te increased incidence of atherosclerosis in people with diabetes has multiple causes. An important cause is the high incidence of specifc dyslipidemia, the so-called "diabetic dyslipidemia" characterized by elevated TG, increased TRL-c, small and dense LDL particles, and reduced HDL. Tis pattern of dyslipidemia is prevalent in those with type 2 diabetes [17,26]. In addition to the correlation between fasting serum triacylglycerol and diabetes, there is evidence that the plasma concentration of residual TRL-c and the cholesterol content in TRL refect the pathophysiological process leading to diabetes progression [27,28]. Terefore, residual lipoproteinemia should be a risk factor for targeted type 2 diabetes treatment [29,30].
Te strengths of our study include using standardized measurements for physical and laboratory tests and relatively large stroke population research. Te present study still has limitations. First, although we carefully adjusted for potential confounders, we cannot entirely rule out the possibility of residual confounding by unmeasured factors, such as dietary variables, sleep duration, and physical activity. It is well known that the daily diet has a signifcant impact on diabetes and cerebrovascular diseases [31]. Te study population was in the "oriental diet" pattern, and they often consumed beans, which contain isofavones, which can have benefcial efects on diabetes and cardiovascular diseases. Te relationship between soy products, isofavones, diet, and CVD has become a controversial topic [32]. Tis question is undoubtedly interesting and opens up intriguing felds in the impact of diet on diabetes and cerebrovascular diseases. For example, studies have reported that one year of treatment with broom isofavin improves surrogate endpoints associated with the risk of diabetes and cardiovascular disease in postmenopausal women with MetS [33]. It was worth mentioning that, A growing number of studies concerning gut microbiota and stroke. Recent studies have revealed that changes in gut microbiota ecology are related to cerebrovascular diseases [34,35]. Gut dysbiosis may result from obesity, metabolic disorders, cardiovascular disease, and sleep disorders; Lack of physical activity is associated  with gut dysbiosis as well. Tese may coexist in various patterns in stroke patients, enhancing the risk, incidence, and progression of cerebrovascular lesions, creating a vicious circle [36]. Meanwhile, studies have shown that the prognosis of stroke patients is signifcantly improved after the use of drugs that improve gut microbiota [37]. Terefore, the next research plan is to collect the factors afecting intestinal microbial homeostasis, dietary habits, sleep duration, physical activity, et al. Moreover, drugs or nutrients that can regulate and improve gut microbial community were collected, such as Probiotics/Prebiotics and Vit B12. Second, a cohort study based on data obtained from the medical records of patients visiting for routine clinical practice has been intentionally and systematically not been collected for research and therefore has several inherent limitations. Tird, no distinction is made between types 1 and 2 diabetes in the database. Finally, the information on TRL-c subclasses was not complete in the dataset, so only TRL-c levels were selected as the exposure.

Conclusion and Future Directions
Te present study found an independent positive association between TRL-c and DM in stroke patients. Our fndings    International Journal of Endocrinology further support the possible new strategy of TRL-c for the prevention and treatment of DM. It is worth mentioning that, medical treatment has been shifted to being more prophylactic as a recent trend. Meanwhile, it has been suggested that the alarming increase in the incidence of noncommunicable diseases, including obesity, cancers, diabetes, and cerebrovascular illnesses is the epidemiologic result of a nutrition transition characterized by dietary patterns [38]. It is well known that a healthy diet and physical activity can reduce the risk of noncommunicable diseases, including diabetes [39]. On the one hand, although Mediterranean-style eating patterns approaches have proven successful as part of the treatment for obesity and cardiometabolic derangements within clinical trial scenarios, they lack efectiveness in the long term, mainly due to poor compliance [40]. On the other hand, some studies have shown that Supplementation with probiotics or prebiotics may improve poststroke outcomes, as these supplements improved lipid and glucose metabolism in overweight people and those with diabetes mellitus [41]. Indeed, these results could provide a possible further strategy to prevent complications and delay the progression of Diabetes Mellitus. Some research [42] have thus turned its attention to nutraceutics, nutrients that have the ability to modulate physiological and pathophysiological molecular mechanisms, thus resulting in favorable health outcomes. In a word, preventing or retarding the onset of diseases has become a more attractive and cost-efective strategy in the medical arena of the prophylactic efects of dietary supplements and nutrients. Te revised content is marked in red in the manuscript.

Data Availability
Te datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Disclosure
Ruiqing Luo and Huifang Jiang are the co-frst authors.

Conflicts of Interest
Te authors declare that they have no conficts of interest.