Introduction

Hypertriglyceridemia (HTG) is a common type of dyslipidemia, especially in China. HTG could initiate and promote the atherosclerotic process, which could in turn accelerate cardiovascular disease1. Therefore, HTG is thought to be a risk factor for ischemic stroke. A lot of clinical studies have been conducted to uncover the relationship between HTG and ischemic stroke. However, the results are controversial and no consensus has been made until now. Most of the studies using fasting triglyceride levels as a marker for ischemic stroke have been conflicting. Several studies have revealed that elevated fasting triglyceride level is a risk factor for ischemic stroke2,3. However, these results cannot be verified in the ARIC study4 and the Physicians’ Health Study5, in which fasting TG levels had no relationship with ischemic stroke. In contrast to the above inconsistence findings, nonfasting triglyceride levels were reported consistently to be a risk factor for ischemic stroke in the Copenhagen City Heart Study6,7 and Women’s Health Study8. These results indicated that nonfasting or postprandial triglyceride levels may be a more suitable biomarker for ischemic stroke and atherosclerotic disease compared to fasting triglyceride levels. However, multiple factors are associated with elevated nonfasting triglyceride levels, and no standardization or consensus has been made so far for the measurement of nonfasting triglyceride levels1. Therefore, scientists are trying to find a more stable surrogate biomarker.

Nonfasting hypertriglyceridemia is due to the accumulation of triglyceride-rich chylomicrons and their partially hydrolyzed chylomicrons, which are highly atherogenic. The absorptive cells in the small intestine produce Apolipoprotein B48 (ApoB48) to form the chylomicrons. Each chylomicron particle contains one molecule of ApoB489. In addition, previous studies have indicated that the fasting concentration of ApoB48 can be a good surrogate marker for the degree of postprandial lipidemia10,11. Therefore, the fasting level of ApoB48 may be a good biomarker for atherosclerotic diseases including cardiovascular disease and ischemic stroke. Clinical studies have proved that the fasting level of ApoB48 is associated with carotid atherosclerosis12,13,14,15,16, peripheral artery disease17 and coronary artery disease18,19. However, the association between fasting ApoB48 and ischemic stroke is still unknown. Thus, our objective in this study was to explore the relationship between fasting plasma ApoB48 levels and the large artery atherosclerotic (LAA) stroke.

Results

During the study period, 234 patients meeting inclusion criteria were included in this study. There were 147 (62.82%) males and the mean age was 59.97 ± 11.45 years. Controls were recruited with a 1:1 age-(±2), and gender-matched fashion during the same study period. The mean age of controls was 59.76 ± 11.39 years. Table 1 summarizes the demographic and clinical features of the controls and LAA stroke patients. Compared with the controls, the patients with LAA stroke had higher prevalence of vascular risk factors (hypertension, diabetes, smoking, drinking) and higher mean values of BMI, and TG levels (all p < 0.05). However, the control participants had higher plsama levels of HDL-C than the LAA stroke patients (p = 0.008). Fasting plasma ApoB48 levels ranged from 0 to 13 μg/mL in control participants, and from 0 to 18 μg/mL in LAA stroke patients (Fig. 1). The distribution of ApoB48 levels in controls and LAA stroke patients was different. Further analysis revealed that the plasma levels of ApoB48 were higher in the patients with LAA stroke than in the controls (4.76(3.46) vs 4.00(2.4), P < 0.001). When the plasma levels of ApoB48 were dichotomized at 5.29 μg/mL according to the ROC curve analysis, the patients in LAA stroke group had a significantly higher proportion of high ApoB48 levels (42.74% vs. 23.50%, p < 0.001), compared with controls.

Table 1 Clinical characteristics of the controls and patients with LAA stroke.
Full size table
Figure 1
figure 1

Distribution of fasting plasma ApoB48 Levels in controls and Patients with LAA stroke. Plasma concentrations of ApoB48 = 1 represents concentrations between 0 and 1.0 μg/mL.

Full size image

Further analysis indicated that there was a linear relationship between ApoB48 levels and TG levels in controls (p < 0.001, r = 0.584) and in LAA stroke group (p < 0.001, r = 0.464) (Fig. 2). TG level were dichotomized at 1.74 mmol/L according to the ROC curve analysis and patients were separated into four groups by combining ApoB48 and TG level: High ApoB48 Low TG, High ApoB48 High TG, Low ApoB48 Low TG, and Low ApoB48 High TG group. The high levels of both ApoB48 and TG was more common in LAA stroke patients than in controls (Figs. 3, 29.06% vs 16.67%, p < 0.001). Furthermore, the proportion of participants with High ApoB48 Low TG in LAA stroke group (13.67%) was significantly greater than in controls (6.84%).

Figure 2
figure 2

Correlation between fasting plasma ApoB48 Levels and TG levels. Pearson correlation analysis revealed that there was a linear relationship between ApoB48 levels and TG levels in controls (p < 0.001, r = 0.584) and in LAA stroke patients (p < 0.001, r = 0.464).

Full size image
Figure 3
figure 3

Plasma levels of ApoB48 and TG in LAA stroke patients and controls. Participants were divided into four groups according to their fasting levles of ApoB48 (cutoff value, 5.29 μg/ml) and TG (cutoff value, 1.74 mmol/L). A high ApoB48 level was more common in LAA stroke patients than in controls even if the TG levels were low. *p < 0.01 vs. Control group.

Full size image

The correlation of ApoB48 with LAA stroke could be confounded by various factors. Therefore, further analyses were conducted to uncover whether ApoB48 is independently associated with LAA stroke. Those variables with p < 0.10 in the univariable analyses were included into further analysis model. After adjusting for hypertension, diabetes mellitus, BMI, smoking, drinking, TG, and HDL, fasting plasma ApoB48 levels were significantly associated with LAA stroke (OR = 1.18, 95% CI = 1.04–1.35, P = 0.014) (Table 2). However, the TG levels were not associated with LAA stroke regardless of inclusion of ApoB48 in the model. When the dichotomized variable “ApoB48 > 5.29 μg/mL” was entered into the model (Supplementary Table 1), a high ApoB48 level was also significantly associated with LAA stroke (OR = 3.50, 95% CI = 1.69–7.26, P = 0.001).

Table 2 Univariate and multivariate conditional logistic regression analysis for risk Factors in patients with LAA stroke.
Full size table

Discussion

In this study, we demonstrated that, for the first time, fasting ApoB48 levels were higher in LAA stroke patients, and a higher ApoB48 level (ApoB48 > 5.29 μg/mL) was nearly two times more prevalent in LAA stroke patients (42.74%) than in controls (23.50%). In multivariable analysis, fasting ApoB48 levels (OR = 1.18, 95% CI = 1.04–1.35, P = 0.014) and a high ApoB48 level (OR = 3.50, 95% CI = 1.69–7.26, P = 0.001) were associated with LAA stroke independent of conventional risk factors. Fasting TG levels were positively correlated with ApoB48 levels; however, the fasting plasma levels of TG were not an independently risk factor for LAA stroke.

Postprandial HTG is one of the independent risk factors for atherosclerosis. Moreover, previous large cohort studies have proved that postprandial TG levels may be a more suitable marker for ischemic stroke than fasting TG level. However, there is no standardized way to measure postprandial TG levels. Postprandial TG levels are traditionally evaluated as the response to a standard rich, high-fat meal. This test needs 6 to 8 hours and requires the patients’ cooperation. Therefore, the discovery of surrogate markers of postprandial HTG, which could be easily measured in routine clinical practice, will be useful to evaluate the stroke risk. As one chylomicron particle contains one molecule of ApoB48, the plasma level of ApoB48 indicates the number of postprandial particles [9].Fasting levels of ApoB48 closely correlate with postprandial TG levels after a lipid meal [10]. This information suggests that fasting ApoB48 level is a surrogate markers of postprandial HTG. Previous studies have shown that fasting ApoB48 could be a risk factor for atherosclerosis and atherosclerotic cardiovascular diseases. The present cross-sectional study, which shows that fasting ApoB48 is associated with LAA stroke, adds new information to this connection.

Our findings that fasting ApoB48 is associated with LAA stroke indicates that ApoB48-containing lipoproteins may take part in the development of atherosclerosis20. A lot of experimental studies have indicated the possible underlying mechanisms. ApoB48 has a binding site to arterial wall proteoglycans21. As a result, ApoB48 was found in human atherosclerotic plaques from femoral and carotid endarterectomy samples22. In vitro studies have indicated that chylomicron remnants could be uptaken by mouse peritoneal macrophages and human monocyte-derived macrophages though multiple mechanisms23. Moreover, chylomicron remnants could induce monocyte chemoattractant protein-1 expression via p38 MAPK activation and regulate early growth response factor-1 in vascular smooth muscle cells24,25. In addition, chylomicron remnants could increase the production of plasminogen activator inhibitor-1 (PAI-1) and enhance apoptosis in endothelial cells26. These studies provide the physiopathological mechanisms to support our findings.

There are some limitations in this study. First, as a cross-sectional study, this study cannot help to make causal inferences. Second, the fasting levels of ApoB48 may change after disease onset. However, we recruited patients who were admitted in our hospital within 24 hours after disease onset, and existing study show that lipid concentrations have not significant changes during the first days after stroke onset27.

In conclusion, fasting plasma ApoB48 levels were significantly correlated with the prevalence of LAA stroke. Therefore, ApoB48 may be a new marker for LAA stroke, as well as a possible therapeutic target.

Material and Methods

Patients and controls

This study was approved by the Ethics Committees of the People’s Hospital of Deyang City. Informed consent have been obtained from every participant.All methods were performed in accordance with the relevant guidelines and regulations. From February 2015 to December 2017, consecutive ischemic patients who were admitted to our hospital were prospectively screened for enrollment. The diagnosis of LAA stroke was confirmed by two neurologists. The inclusion criteria was admission within 24 hours after onset of ischemic stroke. The exclusion criteria were as follows: (1) having a previous history of stroke or ischemic heart disease; (2) having received treatment before admission including statin treatment; (3) not having fasting plasma drawn within 24 hours after admission; (4) having incomplete data for stroke etiology and/or two or more stroke etiology; (5) having systematic diseases. We recruited healthy volunteers as controls who received health examinations in our hospital during the same study period. Those volunteers who were free of history of stroke, myocardial infarction, and systematic diseases were included in this study. Ultimately, 234 LAA stroke patients and 234 healthy volunteers were recruited in this study.

Assessment of stroke risk factors

The demographic characteristics, past medical history and clinical data sof the controls and patients were recorded prospectively. The common vascular risk factors including hypertension, diabetes mellitus, drinking, smoking and heart disease were recorded. Diabetes mellitus and hypertension were defined according with the diagnosis guidelines. Smoking was defined as smoking equal to or more than one cigarette per day for one year or more. Alcohol consumption was defined as a past or current history of drinking more than once per day for more than 1 year. Heart disease was defined as if a subject had one or more heart disease, such as myocardial infarction, and atrial fibrillation. Magnetic resonance imaging (MRI) with diffusion weighted imaging, MR or CT angiography, carotid duplex ultrasonography, transthoracic echocardiography, 24-h Holter monitoring, and other routine admission laboratory tests were conducted to help to assess the stroke subtype. Transesophageal echocardiography was also performed if needed. LAA ischemic stroke was assessed by two independent neurologists according to the Trial of Org 10172 in the Acute Stroke Treatment study28.

All LAA stroke patients and controls had fasting lipid panels drawn after an overnight fast. Total cholesterol, triglycerides, HDL, LDL were measured by standard laboratory methods on fresh plasma. In addition, additional plasma of every patient and control were frozen in a −80 °C freezer for later use.

Enzyme-Linked Immunosorbent Assay

Plasma fasting ApoB48 levels were quantified by enzyme-linked immunosorbent assay (ELISA) according to the manufacturer’s instruction (Fujirebio, Tokyo, Japan)29. The concentrations of ApoB48 were measured in batches. Reference plasma samples were pooled from 20 healthy controls and added to each plate to minimize plate-to-plate variations. Board-certified laboratory technicians performed the measurement and they were blinded to group information. The reproducibility of the results was assessed via calculating the average coefficient of variation (CV) within plates and between plates. The mean intra-assay CV was <7%, and the interassay CV was <15%.

Statistical analysis

Data were given as percentages, means and standard deviations or median and interquartile range, as appropriate. The continuous variables between groups were compared by Student’s t-test. The proportions variables between groups were compared by χ2 test. Multivariate analysis was conducted by conditional logistic regression to determine if the fasting ApoB48 level was independently related with LAA stroke after adjustment for common risk factors. The differences between groups were considered significant if the p-value was less than 0.05 (two tailed). Statistical Package for Social Science (SPSS, version 22) was used to conduct the statistical analyses.

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.