Abstract
Aims: Studies showed that low-density lipoprotein cholesterol (LDL-C) to triglyceride (TG) ratio could be used as a predictive parameter of low-density lipoprotein oxidation in vivo and the level of small dense LDL-C. However, whether LDL-C/TG ratio is associated with stroke prognosis remains unclear. We investigated the associations of LDL-C/TG ratio with outcomes in patients with acute ischaemic stroke (AIS) or transient ischaemic attacks (TIA) and explored whether it produced more predictive value than LDL-C and TG.
Methods: Data were derived from the Third China National Stroke Registry (CNSR-III). Multivariable Cox regression for stroke recurrence, composite vascular events and all-cause death and logistic regression for the poor functional outcome (modified Rankin Scale score 3–6) were used.
Results: A total of 14123 patients were included. After adjusting for confounding factors, quartile 4 of LDL-C/TG ratio was associated with an increased risk of recurrent stroke (hazard ratio [HR], 1.27; 95% confidence interval [CI], 1.03–1.56), composite vascular events (HR,1.23; 95% CI, 1.00–1.52), death (HR,1.70; 95% CI, 1.13–2.54) and poor functional outcome (odds ratio, 1.34; 95% CI, 1.12–1.61) at 3 months follow-up compared with quartile 1. We also found that quartile 4 of LDL-C and TG was positively and negatively associated with poor functional outcome at 3 months, respectively. LDL-C/TG ratio performed better than LDL-C or TG in predicting clinical outcomes.
Conclusions: LDL-C/TG ratio was associated with the risk of stroke recurrence, composite vascular events, death and poor functional outcome in patients with AIS or TIA.
Qin Xu and Changjun Li contributed equally to this work.
See editorial vol. 31: 1133-1134
Introduction
Atherosclerosis is one of the most important causes of ischaemic stroke worldwide. The accumulation of low-density lipoprotein (LDL) particles in the subendothelium of arteries is a key pathophysiological process in the formation of atherosclerosis1). As the traditional biological marker of LDL, the LDL cholesterol (LDL-C) is now widely recognized as a risk factor for atherosclerotic cardiovascular disease (ASCVD)2). It is now well-established that lowering LDL-C level can reduce the risk of ischaemic stroke3-5). However, the associations of LDL-C with the prognosis of stroke remains limited and controversial. One study showed that elevated untreated baseline LDL-C level was associated with an increased risk of recurrent ischaemic stroke within 3 months among patients presenting with minor ischaemic stroke or transient ischaemic attacks (TIA)6). But others reported null association7, 8). Serum triglyceride (TG) is another important component of blood lipids, and it is involved in the pathological process of carotid intima thickening or cerebral atherosclerosis9). Elevated TG has been suggested to be an essential risk factor for ischaemic stroke9). Studies have shown that TG is closely related to LDL metabolism. TG is a powerful inverse determinant of LDL particle size, and positively correlate with the levels of small dense LDL-C (sdLDL-C)10, 11), which is a critical subclass of LDL-C and has a stronger ability to cause atherosclerosis and carotid plaques than that of other LDL subfractions12-14). Studies suggested that LDL-C/TG ratio could be used as a predictive parameter of the sdLDL-C levels and LDL oxidation (another major causative factor in the development of atherosclerosis15)) in vivo16-18). However, to our knowledge, no previous research has investigated the association between LDL-C/TG ratio and short- and long-term prognosis in patients with acute ischaemic stroke (AIS) or TIA, and whether LDL-C/TG ratio could be used as a novel biomarker to produce better predictive value for stroke prognosis remain unknown.
Aim
The present study aimed to investigate the associations of LDL-C/TG ratio, with clinical outcomes among patients with AIS or TIA based on the Third China National Stroke Registry (CNSR-III), and to compare its predictive value with LDL-C and TG to provide new evidence to further clarify their role in stroke prognosis.
Methods
Study Design and Population
The detailed study design for CNSR-III has been published elsewhere19). In brief, CNSR-III is a nationwide prospective registry in China. A total of 14146 patients with AIS and 1020 patients with TIA were enrolled from 201 hospitals between August 2015 and March 2018. The inclusion criteria were as follows: (1) aged ≥ 18 years old; (2) ischaemic stroke or TIA within 7 days from the onset of symptoms to enrollment; and (3) informed consent from the patients or legally authorized representatives. The study was approved by the ethics committee at Beijing Tiantan Hospital and all participating sites, and written informed consent was obtained from the patients or their legally authorized representatives.
Data Collection and Calculation
Baseline demographic and clinical characteristics were collected by trained research coordinators at each site through face-to-face interviews or medical records. These data included age, sex, body mass index (BMI), smoking and alcohol consumption status, medical history (hypertension, stroke or TIA, diabetes mellitus, dyslipidemia, atrial fibrillation, coronary heart disease, heart failure, peripheral vascular disease), medication history (cholesterol-lowering agents), the National Institutes of Health Stroke Scale (NIHSS) score at admission, prestroke modified Rankin Score (mRS), blood pressure at admission and treatment in hospital (antihypertensive agents, antiplatelet agents, anticoagulant agents, cholesterol lowering agents, hypoglycemic agents, rt-PA intravenous thrombolytic, mechanical thrombectomy), etc. Etiological classification was based on the criteria of the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria20).
Within 24 hours after admission, fasting venous blood (defined as at least 8 h since the last meal) was collected from the patients. Blood samples were temporarily stored in a −80℃ refrigerator and then transported to the central laboratory at Beijing Tiantan Hospital by means of an ultra-cold chain. Serum total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), LDL-C, TG and high-sensitivity C-reactive protein (hsCRP) were measured centrally. Fasting blood glucose (FBG) and glycated Hemoglobin (HbA1c) were reported from each sub-center. Testing personnel at the central laboratory were unaware of the clinical characteristics of the patients. The LDL-C/TG ratio was calculated as the ratio of LDL-C to TG.
Outcome Assessment
The trained coordinators at each site followed all enrolled patients at 3 months and 1 year to collect information on clinical outcomes, including stroke recurrence (new ischaemic or hemorrhagic stroke), composite vascular events (ischaemic stroke, hemorrhagic stroke, myocardial infarction or vascular death), poor functional outcome (defined as mRS 3–6) and all-cause death (defined as death from any cause). The definition of above-mentioned clinical outcomes was consistent with those previously described in the CNSR-III protocol19), and all reported events were reviewed and confirmed by a central adjudication committee.
Statistical Analysis
Patients were divided into four groups according to the quartiles of LDL-C/TG ratio, LDL-C, and TG. Continuous variables were presented as medians and interquartile ranges (IQR) due to skewed distribution and were compared by nonparametric Wilcoxon or Kruskal–Wallis test. Categorical variables were presented as frequencies with percentages and were compared by the chi-square test or Fisher’s exact test. Hazard ratio (HR) and 95% confidence interval (CI) were calculated for stroke recurrence, composite vascular events and all-cause death using Cox proportional hazards model. For poor functional outcome, logistic regression model was used to calculate odds ratio (OR) and 95% CI. Since patients in this study were enrolled from 201 hospitals, the hospitals were added to the models as clusters, and a robust sandwich variance estimator was used to handle the correlations. We fitted an unadjusted model and an adjusted model for each outcome. As there are many potential variables influencing the outcomes, we used the stepwise variable selection approaches to evaluate appropriate covariates for adjusted models in order to avoid overfitting. For outcomes of stroke recurrence and composite vascular events, we adjusted for age, sex, NIHSS score at admission, systolic blood pressure, FBG, hsCRP, stroke subtype, medical history (stroke or TIA, and atrial fibrillation), TOAST, medication in hospital (antiplatelet agents, and hypoglycemic agents) after stepwise process. For outcomes of death and mRS score 3-6, we adjusted for age, sex, NIHSS score at admission, systolic blood pressure, diastolic blood pressure, BMI, FBG, hsCRP, HDL-C, stroke subtype, prestroke mRS score, medical history (stroke or TIA, diabetes, and atrial fibrillation), TOAST, medication in hospital (antihypertensive agents, antiplatelet agents, and hypoglycemic agents) after stepwise process. We treated median LDL-C/TG ratio, LDL-C and TG of each quartile as continuous variables in each model and conducted linear trend tests. The Kaplan–Meier method and log-rank test were used for time-to-event data. Subgroup analysis was conducted according to treatments of cholesterol-lowering agents in hospital (yes/no) with an interaction test. Additionally, we used C statistics, integrated discrimination improvement (IDI), and net reclassification index (NRI) to evaluate the predictive performance of LDL-C/TG ratio, LDL-C, and TG beyond the basic model.
A two-sided P values ＜0.05 was considered to indicate statistical significance. All statistical analyses were performed with SAS software, version 9.4 (SAS Institute Inc.).
Results
Baseline Characteristics
Of 15166 patients in the CNSR-III study, after excluding patients with missing data of LDL-C or TG (n=626) or patients who were lost to follow-up (n=417), a total of 14123 patients with AIS or TIA were included in this analysis (Supplementary Fig.1). The baseline characteristics of excluded and included patients were highly comparable as shown in Supplementary Table 1, except that the included patients tended to be younger, female, contained a higher proportion of TIA and lower levels of hsCRP. Table 1 showed the baseline clinical characteristics of the study population stratified according to the LDL-C/TG ratio quartile. Compared to patients with a lower LDL-C/TG ratio, those in the highest quartile group tended to be older, female, non-smokers, have smaller BMI, lower prevalence of medical history (including hypertension, stroke or TIA, diabetes, dyslipidemia), lower proportion of previous cholesterol-lowering agents treatment, lower TG concentration, FBG and HbA1c, and lower proportion of antihypertensive agents, antiplatelet agents, and hypoglycemic agents, but higher prevalence of atrial fibrillation, index events of TIA and cardioembolism stroke, and higher NIHSS scores and systolic blood pressure at admission, TC, HDL-C, LDL-C and hsCRP concentrations.
Supplementary Table 1.Baseline characteristics between included and excluded patients
| Characteristics |
Total |
Excluded |
Included |
P value
|
| No. of the patients |
15166 |
1043 |
14123 |
|
| Serum LDL-C/TG, median (IQR) |
1.8 (1.2-2.5) |
1.9 (1.3-2.5) |
1.8 (1.2-2.5) |
0.17 |
| Age, median (IQR), y |
63 (54-70) |
64 (54-72) |
62 (54-70) |
0.001 |
| Female, n (%)
|
4802 (31.7) |
299 (28.7) |
4503 (31.9) |
0.03 |
| BMI, median (IQR), kg/m2
|
24.5 (22.6-26.6) |
24.5 (22.6-26.6) |
24.49 (22.6-26.6) |
0.85 |
| Current smoker, n (%)
|
4752 (31.3) |
353 (33.8) |
4399 (31.2) |
0.07 |
| Current alcohol drinking, n (%)
|
6797 (44.8) |
481 (46.1) |
6316 (44.7) |
0.38 |
| Medical History, n (%)
|
|
|
|
|
| Hypertension |
9494 (62.6) |
664 (63.7) |
8830 (62.5) |
0.46 |
| Stroke or TIA |
3675 (24.2) |
266 (25.5) |
3409 (24.1) |
0.32 |
| Diabetes |
3510 (23.1) |
225 (21.6) |
3285 (23.3) |
0.21 |
| Dyslipidemia |
1191 (7.9) |
79 (7.6) |
1112 (7.9) |
0.73 |
| Atrial fibrillation |
1019 (6.7) |
82 (7.9) |
937 (6.6) |
0.13 |
| Coronary heart disease |
1608 (10.6) |
105 (10.1) |
1503 (10.6) |
0.56 |
| Heart failure |
94 (0.6) |
10 (1.0) |
84 (0.6) |
0.15 |
| Peripheral vascular disease |
118 (0.8) |
6 (0.6) |
112 (0.8) |
0.44 |
| Medication history, n (%)
|
|
|
|
|
| Cholesterol-lowering agents |
1671 (11.0) |
102 (9.8) |
1569 (11.1) |
0.19 |
| Admission stroke data
|
|
|
|
|
| NIHSS at admission, median (IQR) |
3 (1-6) |
4 (2-6) |
3 (1-6) |
＜0.001 |
| Prestroke mRS score 2-5, n (%)
|
1344 (8.9) |
102 (9.8) |
1242 (8.8) |
0.28 |
| Systolic blood pressure, median (IQR), mmHg |
148.0 (135.0-163.5) |
149.0 (134.0-163.5) |
148.0 (135.0-163.5) |
0.97 |
| Diastolic blood pressure, median (IQR), mmHg |
86.0 (79.0-95.0) |
86.0 (78.5-97.5) |
86.0 (79.0-95.0) |
0.69 |
| Stroke subtype, n (%)
|
|
|
|
0.003 |
| Ischemic stroke |
14146 (93.3) |
996 (95.5) |
13150 (93.1) |
|
| TIA |
1020 (6.7) |
47 (4.5) |
973 (6.9) |
|
| Toast, n (%)
|
|
|
|
0.17 |
| Large-artery atherosclerosis |
3856 (25.4) |
294 (28.2) |
3562 (25.2) |
|
| Cardioembolism |
917 (6.1) |
62 (5.9) |
855 (6.1) |
|
| Small-vessel occlusion |
3165 (20.9) |
203 (19.5) |
2962 (21.0) |
|
| Other determined etiology |
182 (1.2) |
8 (0.8) |
174 (1.2) |
|
| Undetermined etiology |
7046 (46.5) |
476 (45.6) |
6570 (46.5) |
|
| Laboratory data, median (IQR)
|
|
|
|
|
| TC, mmol/L |
4.1 (3.4-4.9) |
4.1 (3.3-4.8) |
4.1 (3.4-4.9) |
0.03 |
| TG, mmol/L |
1.4 (1.0-1.9) |
1.3 (1.0-1.8) |
1.4 (1.0-1.9) |
0.01 |
| HDL-C, mmol/L |
1.1 (0.9-1.3) |
1.1 (0.9-1.3) |
1.1 (0.9-1.3) |
0.04 |
| LDL-C, mmol/L |
2.4 (1.8-3.1) |
2.4 (1.8-3.1) |
2.4 (1.9-3.1) |
0.08 |
| FBG, mmol/L |
5.5 (4.9-6.9) |
5.5 (4.9-7.1) |
5.5 (4.9-6.9) |
0.36 |
| HbA1c, % |
5.9 (5.5-6.9) |
5.9 (5.5-6.8) |
5.9 (5.5-6.97) |
0.30 |
| hsCRP, mg/L |
1.8 (0.8-4.8) |
2.9 (1.1-8.0) |
1.8 (0.8-4.7) |
＜0.001 |
| Medications in hospital, n (%)
|
|
|
|
|
| Antihypertensive agents |
7000 (46.5) |
501 (48.8) |
6499 (46.3) |
0.12 |
| Antiplatelet agents |
14613 (97.1) |
984 (95.9) |
13629 (97.2) |
0.02 |
| Anticoagulant agents |
1546 (10.3) |
121 (11.8) |
1425 (10.2) |
0.10 |
| Cholesterol-lowering agents |
14506 (96.4) |
993 (96.8) |
13513 (96.3) |
0.46 |
| Hypoglycemic agents |
3792 (25.2) |
249 (24.3) |
3543 (25.3) |
0.48 |
| Acute stroke treatment, n (%)
|
|
|
|
|
| rt-PA intravenous thrombolytic |
1303 (8.6) |
108 (10.4) |
1195 (8.5) |
0.04 |
| Mechanical thrombectomy |
39 (0.3) |
0 (0.0) |
39 (0.3) |
0.09 |
| Recanalized through mechanical thrombectomy |
35 (89.7) |
0 (0.0) |
35 (89.7) |
- |
BMI indicates body mass index; FBG, fasting blood glucose; HbA1c, glycated hemoglobin; HDL-C, high-density lipoprotein cholesterol; hsCRP, high sensitivity C reactive protein; IQR, interquartile range; LDL-C, low-density lipoprotein cholesterol; mRS, modified Rankin Scale; NIHSS, the National Institutes of Health Stroke Scale; rt-PA, recombinant tissue plasminogen activator; TC, total cholesterol; TG, triglyceride; TIA, transient ischemic attack; TOAST, Trial of Org 10172 in Acute Stroke Treatment.
Table 1.Baseline characteristics according to quartiles of LDL-C/TG ratio
| Characteristics |
Total |
Quartiles of LDL-C/TG ratio |
P value
|
| Q1 (＜1.21) |
Q2 (1.21-＜1.76) |
Q3 (1.76-＜2.48) |
Q4 (≥ 2.48) |
| No. of the patients |
14123 |
3567 |
3478 |
3570 |
3508 |
|
| Serum LDL-C/TG, median (IQR) |
1.8 (1.2-2.5) |
0.9 (0.6-1.1) |
1.5 (1.4-1.6) |
2.1 (1.9-2.3) |
3.1 (2.7-3.6) |
＜0.001 |
| Age, median (IQR), y |
62 (54-70) |
60 (52-67) |
62 (54-70) |
64 (56-71) |
64 (57-72) |
＜0.001 |
| Female, n (%)
|
4503 (31.9) |
1104 (31.0) |
1067 (30.7) |
1145 (32.1) |
1187 (33.8) |
0.02 |
| BMI, median (IQR), kg/m2
|
24.5 (22.6-26.6) |
25.1 (23.2-27.1) |
24.6 (22.9-26.7) |
24.3 (22.5-26.5) |
23.9 (22.0-26.0) |
＜0.001 |
| Current smoker, n (%)
|
4399 (31.2) |
1158 (32.5) |
1137 (32.7) |
1106 (31.0) |
998 (28.5) |
＜0.001 |
| Current alcohol drinking, n (%)
|
6316 (44.7) |
1619 (45.4) |
1559 (44.8) |
1627 (45.6) |
1511 (43.1) |
0.13 |
| Medical history, n (%)
|
|
|
|
|
|
|
| Hypertension |
8830 (62.5) |
2380 (66.7) |
2222 (63.9) |
2165 (60.6) |
2063 (58.8) |
＜0.001 |
| Stroke or TIA |
3409 (24.1) |
902 (25.3) |
869 (25.0) |
847 (23.7) |
791 (22.6) |
0.03 |
| Diabetes |
3285 (23.3) |
1104 (31.0) |
849 (24.4) |
739 (20.7) |
593 (16.9) |
＜0.001 |
| Dyslipidemia |
1112 (7.9) |
380 (10.7) |
287 (8.3) |
226 (6.3) |
219 (6.2) |
＜0.001 |
| Atrial fibrillation |
937 (6.6) |
142 (4.0) |
219 (6.3) |
265 (7.4) |
311 (8.9) |
＜0.001 |
| Coronary heart disease |
1503 (10.6) |
376 (10.5) |
376 (10.8) |
376 (10.5) |
375 (10.7) |
0.98 |
| Heart failure |
84 (0.6) |
18 (0.5) |
23 (0.7) |
23 (0.6) |
20 (0.6) |
0.82 |
| Peripheral vascular disease |
112 (0.8) |
25 (0.7) |
31 (0.9) |
32 (0.9) |
24 (0.7) |
0.61 |
| Medication history, n (%)
|
|
|
|
|
|
|
| Cholesterol-lowering agents |
1569 (11.1) |
492 (13.8) |
406 (11.7) |
395 (11.1) |
276 (7.9) |
＜0.001 |
| Admission stroke data |
|
|
|
|
|
|
| NIHSS at admission, median (IQR) |
3 (1-6) |
3 (1-5) |
3 (1-6) |
3 (1-6) |
3 (1-6) |
＜0.001 |
| Prestroke mRS score 2-5, n (%)
|
1242 (8.8) |
320 (9.0) |
311 (8.9) |
305 (8.5) |
306 (8.7) |
0.91 |
| Systolic blood pressure, median (IQR), mmHg |
148.0 (135.0-163.5) |
147.5 (134.5-162.5) |
147.5 (135.0-163.0) |
148.0 (135.0-163.5) |
149.5 (135.5-165.0) |
0.01 |
| Diastolic blood pressure, median (IQR), mmHg |
86.0 (79.0-95.0) |
87.0 (80.0-96.0) |
86.0 (79.0-95.0) |
85.5 (79.0-95.0) |
86.0 (79.0-95.0) |
0.03 |
| Stroke subtype, n (%)
|
|
|
|
|
|
0.04 |
| Ischemic stroke |
13150 (93.1) |
3330 (93.3) |
3246 (93.3) |
3344 (93.7) |
3230 (92.1) |
|
| TIA |
973 (6.9) |
237 (6.6) |
232 (6.7) |
226 (6.3) |
278 (7.9) |
|
| TOAST, n (%) |
|
|
|
|
|
＜0.001 |
| Large-artery atherosclerosis |
3562 (25.2) |
880 (24.7) |
898 (25.8) |
880 (24.7) |
904 (25.8) |
|
| Cardioembolism |
855 (6.1) |
145 (4.1) |
186 (5.4) |
251 (7.0) |
273 (7.8) |
|
| Small-vessel occlusion |
2962 (21.0) |
823 (23.1) |
721 (20.7) |
745 (20.9) |
673 (19.2) |
|
| Other determined etiology |
174 (1.2) |
45 (1.3) |
41 (1.2) |
48 (1.3) |
40 (1.1) |
|
| Undetermined etiology |
6570 (46.5) |
1674 (46.9) |
1632 (46.9) |
1646 (46.1) |
1618 (46.1) |
|
| Laboratory data, median (IQR) |
|
|
|
|
|
|
| TC, mmol/L |
4.1 (3.4-4.9) |
3.7 (3.1-4.6) |
3.9 (3.3-4.7) |
4.2 (3.6-4.9) |
4.6 (4.0-5.4) |
＜0.001 |
| TG, mmol/L |
1.4 (1.0-1.9) |
2.2 (1.7-3.0) |
1.5 (1.2-1.9) |
1.3 (1.0-1.5) |
1.0 (0.8-1.2) |
＜0.001 |
| HDL-C, mmol/L |
1.1 (0.9-1.3) |
1.0 (0.8-1.1) |
1.1 (0.9-1.2) |
1.1 (1.0-1.3) |
1.2 (1.0-1.4) |
＜0.001 |
| LDL-C, mmol/L |
2.4 (1.9-3.1) |
1.8 (1.3-2.3) |
2.3 (1.8-2.8) |
2.6 (2.1-3.2) |
3.1 (2.6-3.7) |
＜0.001 |
| FBG, mmol/L |
5.5 (4.9-6.9) |
5.8 (5.0-7.7) |
5.6 (4.9-6.9) |
5.5 (4.9-6.6) |
5.4 (4.8-6.3) |
＜0.001 |
| HbA1c, % |
5.9 (5.5-7.0) |
6.1 (5.5-7.7) |
6.0 (5.5-7.0) |
5.8 (5.4-6.8) |
5.8 (5.4-6.5) |
＜0.001 |
| hsCRP, mg/L |
1.8 (0.8-4.7) |
1.7 (0.8-3.8) |
1.8 (0.8-4.6) |
1.9 (0.8-5.3) |
2.0 (0.8-5.5) |
＜0.001 |
| Medications in hospital, n (%)
|
|
|
|
|
|
|
| Antihypertensive agents |
6499 (46.3) |
1776 (50.1) |
1641 (47.6) |
1579 (44.5) |
1503 (43.0) |
＜0.001 |
| Antiplatelet agents |
13629 (97.2) |
3448 (97.3) |
3376 (98.0) |
3430 (96.7) |
3375 (96.7) |
0.003 |
| Anticoagulant agents |
1425 (10.2) |
342 (9.7) |
357 (10.4) |
375 (10.6) |
351 (10.1) |
0.60 |
| Cholesterol-lowering agents |
13513 (96.3) |
3421 (96.6) |
3327 (96.6) |
3401 (95.9) |
3364 (96.3) |
0.43 |
| Hypoglycemic agents |
3543 (25.3) |
1191 (33.6) |
917 (26.6) |
810 (22.8) |
625 (17.9) |
＜0.001 |
| Acute stroke treatment, n (%)
|
|
|
|
|
|
|
| rt-PA intravenous thrombolytic |
1195 (8.5) |
288 (8.1) |
278 (8.0) |
321 (9.0) |
308 (8.8) |
0.33 |
| Mechanical thrombectomy |
39 (0.3) |
5 (0.1) |
10 (0.3) |
13 (0.4) |
11 (0.3) |
0.31 |
| Recanalized through mechanical thrombectomy |
35 (89.7) |
4 (80.0) |
9 (90.0) |
11 (84.6) |
11 (100.0) |
0.54 |
BMI indicates body mass index; FBG, fasting blood glucose; HbA1c, glycated hemoglobin; HDL-C, high-density lipoprotein cholesterol; hsCRP, high sensitivity C reactive protein; IQR, interquartile range; LDL-C, low-density lipoprotein cholesterol; mRS, modified Rankin Scale; NIHSS, the National Institutes of Health Stroke Scale; rt-PA, recombinant tissue plasminogen activator; TC, total cholesterol; TG, triglyceride; TIA, transient ischemic attack; TOAST, Trial of Org 10172 in Acute Stroke Treatment.
Totally, 857 (6.1%) and 1351 (9.6%) patients experienced recurrent stroke within 3 months and within 1 year, respectively. As shown in Fig.1, the Kaplan–Meier analysis with the log-rank test showed that patients with the highest quartile of LDL-C/TG ratio had significantly higher cumulative recurrence rate within 3 months than those with the lowest quartile (P=0.039). The risk of recurrent stroke within 3 months significantly increased in patients with the highest quartile of LDL-C/TG ratio (6.8% versus 5.3%; crude HR, 1.31; 95% CI, 1.07–1.60) compared with those with first quartile (Table 2). After adjusting for confounding factors, the association persisted (HR,1.27; 95% CI, 1.03–1.56). When treating LDL-C/TG ratio as a continuous variable, the adjusted HR per 1 standard deviation (SD) for the risk of recurrent stroke with 3 months were statistically significant (HR, 1.04; 95% CI, 1.00–1.08; P=0.04). We observed a slightly higher rate of stroke recurrence within 1 year in patients with the highest quartile of LDL-C/TG ratio compared with those in quartile 1 (10.3% versus 9.1%), however, increased but nonsignificant risk was found (adjusted HR, 1.11; 95% CI, 0.95-1.29; P=0.20). Similar results were observed for composite vascular events within 3 months and within 1 year. Multivariable-adjusted restricted spline analysis showed an approximately linear associations of LDL-C/TG ratio with the risk of recurrent stroke and composite vascular events within 3 months (Fig.2).
Table 2.Associations of LDL-C/TG ratio with clinical outcomes after stroke or transient ischemic attack
| Outcomes |
Quartiles of LDL-C/TG |
P for trend
|
Per 1 SD increase |
| Q1 |
Q2 |
Q3 |
Q4 |
| At 3 months |
|
|
|
|
|
|
| Stroke recurrence |
|
|
|
|
|
|
| n (%)
|
188 (5.3) |
220 (6.3) |
209 (5.9) |
240 (6.8) |
|
|
| Unadjusted |
Reference |
1.21 (0.99-1.46) |
1.11 (0.93-1.33) |
1.31 (1.07-1.60) |
0.02 |
1.05 (1.01-1.08) |
| Adjusted＊
|
Reference |
1.18 (0.97-1.44) |
1.08 (0.90-1.29) |
1.27 (1.03-1.56) |
0.06 |
1.04 (1.00-1.08) |
| Composite vascular events |
|
|
|
|
|
|
| n (%)
|
196 (5.5) |
227 (6.5) |
217 (6.1) |
246 (7.0) |
|
|
| Unadjusted |
Reference |
1.19 (0.99-1.44) |
1.11 (0.93-1.32) |
1.29 (1.06-1.57) |
0.03 |
1.04 (1.01-1.08) |
| Adjusted＊
|
Reference |
1.17 (0.96-1.41) |
1.06 (0.89-1.27) |
1.23 (1.00-1.52) |
0.10 |
1.03 (1.00-1.07) |
| Death |
|
|
|
|
|
|
| n (%)
|
35 (1.0) |
43 (1.2) |
55 (1.5) |
77 (2.2) |
|
|
| Unadjusted |
Reference |
1.26 (0.79-2.01) |
1.58 (1.02-2.43) |
2.25 (1.52-3.33) |
＜0.001 |
1.09 (1.05-1.13) |
| Adjusted†
|
Reference |
1.19 (0.72-1.96) |
1.23 (0.79-1.92) |
1.70 (1.13-2.54) |
0.01 |
1.10 (1.05-1.14) |
| mRS score 3-6 |
|
|
|
|
|
|
| n (%)
|
386 (10.8) |
438 (12.6) |
495 (13.9) |
601 (17.1) |
|
|
| Unadjusted |
Reference |
1.19 (1.02-1.38) |
1.33 (1.13-1.56) |
1.70 (1.41-2.06) |
＜0.001 |
1.18 (1.08-1.30) |
| Adjusted†
|
Reference |
1.06 (0.90-1.25) |
1.09 (0.90-1.31) |
1.34 (1.12-1.61) |
0.002 |
1.07 (1.00-1.16) |
| At 1 year |
|
|
|
|
|
|
| Stroke recurrence |
|
|
|
|
|
|
| n (%)
|
324 (9.1) |
331 (9.5) |
336 (9.4) |
360 (10.3) |
|
|
| Unadjusted |
Reference |
1.05 (0.90-1.24) |
1.04 (0.91-1.19) |
1.14 (0.99-1.33) |
0.10 |
1.03 (1.00-1.07) |
| Adjusted＊
|
Reference |
1.03 (0.88-1.22) |
1.00 (0.87-1.16) |
1.11 (0.95-1.29) |
0.27 |
1.03 (0.99-1.07) |
| Composite vascular events |
|
|
|
|
|
|
| n (%)
|
343 (9.6) |
349 (10.0) |
361 (10.1) |
375 (10.7) |
|
|
| Unadjusted |
Reference |
1.05 (0.90-1.22) |
1.06 (0.92-1.21) |
1.13 (0.97-1.30) |
0.12 |
1.03 (1.00-1.07) |
| Adjusted＊
|
Reference |
1.03 (0.88-1.20) |
1.01 (0.87-1.16) |
1.08 (0.93-1.25) |
0.37 |
1.02 (0.98-1.07) |
| Death |
|
|
|
|
|
|
| n (%)
|
78 (2.2) |
101 (2.9) |
122 (3.4) |
152 (4.3) |
|
|
| Unadjusted |
Reference |
1.33 (0.96-1.84) |
1.57 (1.21-2.05) |
2.00 (1.53-2.63) |
＜0.001 |
1.08 (1.04-1.12) |
| Adjusted†
|
Reference |
1.23 (0.88-1.72) |
1.29 (0.98-1.71) |
1.54 (1.13-2.10) |
0.006 |
1.09 (1.05-1.13) |
| mRS score 3-6 |
|
|
|
|
|
|
| n (%)
|
364 (10.2) |
421 (12.1) |
481 (13.5) |
569 (16.2) |
|
|
| Unadjusted |
Reference |
1.21 (1.03-1.43) |
1.37 (1.17-1.60) |
1.70 (1.42-2.04) |
＜0.001 |
1.19 (1.08-1.30) |
| Adjusted†
|
Reference |
1.07 (0.90-1.28) |
1.11 (0.94-1.30) |
1.32 (1.10-1.58) |
0.003 |
1.08 (1.00-1.16) |
LDL-C, low-density lipoprotein cholesterol; mRS, modified Rankin Scale; SD, standard deviation; TG, triglyceride.
Hazard ratios (HRs) with 95% confidence intervals (CIs) were used for stroke recurrence, composite vascular events and death; odds ratios (ORs) with 95% CIs were used for mRS score 3-6.
＊For outcomes of stroke recurrence and composite vascular events, models were adjusted for age, sex, the National Institutes of Health Stroke Scale score at admission, systolic blood pressure, fasting blood glucose, high sensitivity C reactive protein, stroke subtype, medical history (stroke or transient ischemic attacks, and atrial fibrillation), Trial of Org 10172 in Acute Stroke Treatment, medication in hospital (antiplatelet agents, and hypoglycemic agents) after stepwise process.
†For outcomes of death and mRS score 3-6, models were adjusted for age, sex, the National Institutes of Health Stroke Scale score at admission, systolic blood pressure, diastolic blood pressure, body mass index, fasting blood glucose, high sensitivity C reactive protein, high-density lipoprotein cholesterol, stroke subtype, prestroke mRS score, medical history (stroke or transient ischemic attacks, diabetes, and atrial fibrillation), Trial of Org 10172 in Acute Stroke Treatment, medication in hospital (antihypertensive agents, antiplatelet agents, and hypoglycemic agents) after stepwise process.
There were 1920 (13.6%) patients with poor functional outcome, of whom 210 (1.5%) died within 3 months. At 1 year follow-up, 1835 (13.0%) patients had poor functional outcome and 453 (3.2%) died. As shown in Table 2, patients with the highest quartile of LDL-C/TG ratio had an increased risk of death (adjusted OR, 1.70; 95% CI, 1.13–2.54) and poor functional outcome (adjusted OR, 1.34; 95% CI, 1.12–1.61) at 3-month follow-up compared with quartile 1 after adjusting for potential confounders. The LDL-C/TG ratio as a continuous variable was also positively associated with death (adjusted OR per 1 SD, 1.07; 95% CI, 1.00–1.16) at 3-month follow-up. Similar results were found for death and poor functional outcome at 1-year follow-up. In addition, Fig.1 indicated the cumulative incidences of all-cause death by LDL-C/TG ratio quartiles were higher in patients with higher LDL-C/TG ratio (log-rank P＜0.001). The relationships of LDL-C/TG ratio with death and poor functional outcome were approximately linear (Fig.2).
All the associations were consistent among patients with or without treatment of cholesterol-lowering agents in hospital after adjusting for all potential confounding factors (P for interaction ＞0.05; Supplementary Table 2).
Supplementary Table 2.Subgroup analysis according to treatments of cholesterol-lowering agents in hospital
| Outcomes |
Quartiles of LDL-C/TG |
Per 1 SD increase |
| Q1 |
Q2 |
Q3 |
Q4 |
P for int
|
OR/HR (95% CI) |
P for int
|
| At 3 months |
|
|
|
|
|
|
|
| Stroke recurrence＊
|
|
|
|
|
0.69 |
|
0.23 |
| Treated with cholesterol-lowering agents |
Reference |
1.23 (0.40-3.77) |
1.43 (0.66-3.09) |
1.44 (0.56-3.74) |
|
1.16 (0.77-1.73) |
|
| Not treated with cholesterol-lowering agents |
Reference |
1.20 (0.98-1.48) |
1.10 (0.90-1.34) |
1.34 (1.08-1.67) |
|
1.05 (1.01-1.08) |
|
| Composite vascular events＊
|
|
|
|
|
0.87 |
|
0.35 |
| Treated with cholesterol-lowering agents |
Reference |
1.10 (0.37-3.29) |
1.29 (0.61-2.70) |
1.33 (0.52-3.38) |
|
1.15 (0.76-1.74) |
|
| Not treated with cholesterol-lowering agents |
Reference |
1.19 (0.97-1.46) |
1.09 (0.90-1.32) |
1.31 (1.06-1.63) |
|
1.04 (1.01-1.08) |
|
| Death†
|
|
|
|
|
0.69 |
|
0.49 |
| Treated with cholesterol-lowering agents |
Reference |
0.23 (0.02-2.49) |
0.62 (0.13-2.85) |
1.53 (0.29-8.10) |
|
1.27 (0.57-2.83) |
|
| Not treated with cholesterol-lowering agents |
Reference |
1.28 (0.76-2.17) |
1.29 (0.79-2.11) |
1.73 (1.11-2.69) |
|
1.10 (1.06-1.14) |
|
| mRS score 3-6†
|
|
|
|
|
0.53 |
|
0.72 |
| Treated with cholesterol-lowering agents |
Reference |
0.63 (0.22-1.77) |
0.61 (0.27-1.37) |
1.06 (0.33-3.41) |
|
1.12 (0.71-1.77) |
|
| Not treated with cholesterol-lowering agents |
Reference |
1.08 (0.92-1.28) |
1.10 (0.92-1.33) |
1.36 (1.13-1.63) |
|
1.07 (1.00-1.16) |
|
| At 1 year |
|
|
|
|
|
|
|
| Stroke recurrence＊
|
|
|
|
|
0.64 |
|
0.47 |
| Treated with cholesterol-lowering agents |
Reference |
1.41 (0.61-3.29) |
1.33 (0.69-2.57) |
1.30 (0.56-3.02) |
|
1.11 (0.78-1.58) |
|
| Not treated with cholesterol-lowering agents |
Reference |
1.03 (0.87-1.23) |
1.01 (0.87-1.18) |
1.15 (0.98-1.36) |
|
1.04 (1.00-1.08) |
|
| Composite vascular events＊
|
|
|
|
|
0.74 |
|
0.61 |
| Treated with cholesterol-lowering agents |
Reference |
1.30 (0.56-3.00) |
1.29 (0.70-2.39) |
1.23 (0.54-2.78) |
|
1.10 (0.78-1.55) |
|
| Not treated with cholesterol-lowering agents |
Reference |
1.03 (0.88-1.22) |
1.03 (0.88-1.20) |
1.13 (0.96-1.33) |
|
1.03 (1.00-1.07) |
|
| Death†
|
|
|
|
|
0.72 |
|
0.57 |
| Treated with cholesterol-lowering agents |
Reference |
0.51 (0.12-2.22) |
0.57 (0.18-1.82) |
1.30 (0.29-5.89) |
|
1.14 (0.57-2.29) |
|
| Not treated with cholesterol-lowering agents |
Reference |
1.26 (0.90-1.76) |
1.32 (0.99-1.77) |
1.53 (1.11-2.11) |
|
1.09 (1.05-1.13) |
|
| mRS score 3-6†
|
|
|
|
|
0.50 |
|
0.87 |
| Treated with cholesterol-lowering agents |
Reference |
0.72 (0.25-2.09) |
0.58 (0.23-1.49) |
1.00 (0.34-2.98) |
|
1.08 (0.67-1.76) |
|
| Not treated with cholesterol-lowering agents |
Reference |
1.09 (0.91-1.30) |
1.14 (0.97-1.34) |
1.34 (1.12-1.61) |
|
1.08 (1.00-1.16) |
|
LDL-C, low-density lipoprotein cholesterol; mRS, modified Rankin Scale; SD, standard deviation; TG, triglyceride.
Hazard ratios (HRs) with 95% confidence intervals (CIs) were used for stroke recurrence, composite vascular events and death; odds ratios (ORs) with 95% CIs were used for mRS score 3-6.
＊For outcomes of stroke recurrence and composite vascular events, models were adjusted for age, sex, the National Institutes of Health Stroke Scale score at admission, systolic blood pressure, fasting blood glucose, high sensitivity C reactive protein, stroke subtype, medical history (stroke or transient ischemic attacks, and atrial fibrillation), Trial of Org 10172 in Acute Stroke Treatment, medication in hospital (antiplatelet agents, and hypoglycemic agents) after stepwise process.
†For outcomes of death and mRS score 3-6, models were adjusted for age, sex, the National Institutes of Health Stroke Scale score at admission, systolic blood pressure, diastolic blood pressure, body mass index, fasting blood glucose, high sensitivity C reactive protein, high-density lipoprotein cholesterol, stroke subtype, prestroke mRS score, medical history (stroke or transient ischemic attacks, diabetes, and atrial fibrillation), Trial of Org 10172 in Acute Stroke Treatment, medication in hospital (antihypertensive agents, antiplatelet agents, and hypoglycemic agents) after stepwise process.
Fig.3 demonstrates the associations of LDL-C/TG ratio, LDL-C and TG with clinical outcomes after adjusting for potential confounders. The fourth quartile of LDL-C level was significantly associated with increased risk of recurrent stroke (adjusted HR, 1.23; 95% CI, 1.03–1.46) and composite vascular events (adjusted HR, 1.20; 95% CI, 1.01–1.43) within 3 months when the first quartile was treated as reference. The fourth quartile of TG level was negatively associated with all-cause death within 1 year (adjusted HR, 0.63; 95% CI, 0.40–1.00), poor functional outcome at 3 months and 1 year (adjusted HR, 0.84; 95% CI, 0.71–0.99; adjusted HR, 0.72; 95% CI, 0.62–0.84). Whereas, the fourth quartile of LDL-C/TG ratio was significantly associated with recurrent stroke, composite vascular events, death and poor functional outcome at 3-month follow-up, and death and poor functional outcome at 1 year follow-up.
We also compared the enhancement in prediction model for the clinical outcomes by adding LDL-C/TG ratio, LDL-C and TG into the basic model. As shown in Table 3, we observed a significant improvement of the C-statistic after addition of LDL-C/TG ratio for all-cause death within 3 months (from 0.845 to 0.850; P=0.047), but not significantly improved with the addition of LDL-C (P=0.52) and TG (P=0.33). The predictive performance validated by IDI and category-free NRI significantly improved with the addition of LDL-C/TG ratio for recurrent stroke, composite vascular events, death and poor functional outcome at 3 months follow-up, but not all significantly improved with the addition of LDL-C and TG. For instance, the LDL-C/TG ratio improved IDI (0.06%; 95%CI, 0.01-0.11; P=0.02) and category-free NRI (8.83%; 95% CI, 1.93-15.74; P=0.01) for recurrent stroke within 3 months, but not with the addition of either LDL-C (P=0.05 for IDI) or TG (P=0.17 for IDI and NRI). For outcomes at 1 year follow-up, we only observed a slightly improvement of IDI (P=0.006) and category-free NRI (P＜0.001) with addition of LDL-C/TG ratio and TG, but not with addition of LDL-C, for poor functional outcome.
Table 3.Performance of models to predict clinical outcomes after stroke or transient ischemic attack
| Model |
C-statistic |
IDI |
Category-free NRI＊
|
| Estimate (95% CI) |
P value
|
Estimate (95% CI), % |
P value
|
Estimate (95% CI), % |
P value
|
| At 3 months |
|
|
|
|
|
|
| Stroke recurrence |
|
|
|
|
|
|
| Basic model＊
|
0.641 (0.623 to 0.660) |
Reference |
Reference |
|
Reference |
|
| Basic model+TG |
0.642 (0.623 to 0.660) |
0.56 |
0.02 (-0.01 to 0.04) |
0.17 |
4.80 (-2.12 to 11.72) |
0.17 |
| Basic model+LDL-C |
0.644 (0.625 to 0.662) |
0.24 |
0.04 (0.00 to 0.08) |
0.05 |
8.82 (1.31 to 15.11) |
0.02 |
| Basic model+LDL-C/TG |
0.645 (0.626 to 0.663) |
0.09 |
0.06 (0.01 to 0.11) |
0.02 |
8.83 (1.93 to 15.74) |
0.01 |
| Composite vascular events |
|
|
|
|
|
|
| Basic model＊
|
0.642 (0.623 to 0.661) |
Reference |
Reference |
|
Reference |
|
| Basic model+TG |
0.643 (0.624 to 0.661) |
0.66 |
0.02 (-0.01 to 0.04) |
0.17 |
4.50 (-2.30 to 11.29) |
0.20 |
| Basic model+LDL-C |
0.644 (0.626 to 0.663) |
0.24 |
0.04 (-0.01 to 0.07) |
0.09 |
7.61 (0.81to 14.41) |
0.03 |
| Basic model+LDL-C/TG |
0.645 (0.627 to 0.663) |
0.11 |
0.05 (0.01 to 0.09) |
0.03 |
8.26 (1.46 to 15.06) |
0.02 |
| Death |
|
|
|
|
|
|
| Basic model†
|
0.845 (0.817 to 0.874) |
Reference |
Reference |
|
Reference |
|
| Basic model+TG |
0.847 (0.819 to 0.876) |
0.33 |
0.10 (-0.10 to 0.28) |
0.36 |
20.16 (6.48 to 33.83) |
0.004 |
| Basic model+LDL-C |
0.846 (0.818 to 0.875) |
0.52 |
0.02 (-0.12 to 0.16) |
0.81 |
9.71 (-3.92 to 23.34) |
0.16 |
| Basic model+LDL-C/TG |
0.850 (0.822 to 0.879) |
0.047 |
0.11 (-0.17 to 0.39) |
0.46 |
22.75 (9.09 to 36.41) |
0.001 |
| mRS score 3-6 |
|
|
|
|
|
|
| Basic model†
|
0.820 (0.810 to 0.831) |
Reference |
Reference |
|
Reference |
|
| Basic model+TG |
0.821 (0.811 to 0.831) |
0.31 |
0.12 (0.04 to 0.21) |
0.004 |
18.08 (13.31 to 22.84) |
＜0.001 |
| Basic model+LDL-C |
0.821 (0.811 to 0.831) |
0.43 |
0.04 (-0.01 to 0.09) |
0.10 |
0.27 (-4.30 to 4.84) |
0.91 |
| Basic model+LDL-C/TG |
0.821 (0.811 to 0.831) |
0.53 |
0.15 (0.06 to 0.24) |
0.002 |
12.86 (8.34 to 17.37) |
＜0.001 |
| At 1 year |
|
|
|
|
|
|
| Stroke recurrence |
|
|
|
|
|
|
| Basic model＊
|
0.629 (0.614 to 0.644) |
Reference |
Reference |
|
Reference |
|
| Basic model+TG |
0.630 (0.615 to 0.646) |
0.19 |
0.02 (-0.01 to 0.04) |
0.23 |
4.39 (-0.38 to 9.16) |
0.13 |
| Basic model+LDL-C |
0.630 (0.614 to 0.645) |
0.44 |
0.02 (-0.01 to 0.05) |
0.18 |
3.91 (-1.71 to 9.53) |
0.17 |
| Basic model+LDL-C/TG |
0.630 (0.614 to 0.645) |
0.33 |
0.02 (-0.01 to 0.04) |
0.12 |
5.06 (-0.43 to 10.56) |
0.08 |
| Composite vascular events |
|
|
|
|
|
|
| Basic model＊
|
0.631 (0.616 to 0.646) |
Reference |
Reference |
|
Reference |
|
| Basic model+TG |
0.632 (0.617 to 0.647) |
0.22 |
0.01 (-0.01 to 0.04) |
0.27 |
4.22 (-0.44 to 8.88) |
0.13 |
| Basic model+LDL-C |
0.631 (0.616 to 0.647) |
0.55 |
0.02 (-0.01 to 0.05) |
0.18 |
3.67 (-1.81 to 9.15) |
0.19 |
| Basic model+LDL-C/TG |
0.631 (0.616 to 0.647) |
0.35 |
0.01 (-0.01 to 0.03) |
0.24 |
3.74 (-1.54 to 9.01) |
0.18 |
| Death |
|
|
|
|
|
|
| Basic model†
|
0.808 (0.786 to 0.829) |
Reference |
Reference |
|
Reference |
|
| Basic model+TG |
0.808 (0.787 to 0.829) |
0.68 |
0.10 (-0.02 to 0.21) |
0.10 |
18.73 (9.61 to 27.85) |
＜0.001 |
| Basic model+LDL-C |
0.809 (0.787 to 0.830) |
0.33 |
0.06 (-0.04 to 0.17) |
0.25 |
4.68 (-4.70 to 14.06) |
0.33 |
| Basic model+LDL-C/TG |
0.810 (0.788 to 0.831) |
0.18 |
0.13 (-0.02 to 0.28) |
0.09 |
16.25 (6.88 to 25.63) |
＜0.001 |
| mRS score 3-6 |
|
|
|
|
|
|
| Basic model† |
0.806 (0.795 to 0.816) |
Reference |
Reference |
|
Reference |
|
| Basic model+TG |
0.807 (0.796 to 0.817) |
0.14 |
0.18 (0.07 to 0.28) |
0.001 |
18.61 (13.72 to 23.51) |
＜0.001 |
| Basic model+LDL-C |
0.806 (0.795 to 0.816) |
0.77 |
0.004 (-0.01 to 0.02) |
0.66 |
1.85 (-3.07 to 6.78) |
0.46 |
| Basic model+LDL-C/TG |
0.806 (0.795 to 0.817) |
0.41 |
0.12 (0.03 to 0.20) |
0.006 |
12.06 (7.33 to 16.79) |
＜0.001 |
CI indicates confidence interval; IDI, integrated discrimination improvement; LDL-C, low-density lipoprotein cholesterol; mRS, modified Rankin Scale; NRI, net reclassification index; TG, triglyceride.
＊Patients were divided into 3 risk categories: 0% to 5%, 5% to 20%, and 20% to 100%.
＊Basic model included adjusted for adjusted for age, sex, the National Institutes of Health Stroke Scale score at admission, systolic blood pressure, fasting blood glucose, high sensitivity C reactive protein, stroke subtype, medical history (stroke or transient ischemic attacks, and atrial fibrillation), Trial of Org 10172 in Acute Stroke Treatment, medication in hospital (antiplatelet agents, and hypoglycemic agents).
†Basic model included adjusted for age, sex, the National Institutes of Health Stroke Scale score at admission, systolic blood pressure, diastolic blood pressure, body mass index, fasting blood glucose, high sensitivity C reactive protein, high-density lipoprotein cholesterol, stroke subtype, prestroke mRS score, medical history (stroke or transient ischemic attacks, diabetes, and atrial fibrillation), Trial of Org 10172 in Acute Stroke Treatment, medication in hospital (antihypertensive agents, antiplatelet agents, and hypoglycemic agents).
Discussion
To the best of our knowledge, this is the first prospective analysis of the association between LDL-C/TG ratio and the prognosis of AIS or TIA published to date in a real-world setting. The results indicated that LDL-C/TG ratio was independently and positively associated with stroke recurrence, composite vascular events at 3-month follow-up, and with death and poor functional outcome at 3-month and 1 year follow-up. We also found that short-term poor functional outcome was positively correlated with LDL-C level and negatively correlated with TG level.
High cholesterol level is an important risk factor for the recurrence of ischaemic stroke, and lowering LDL-C levels reduces recurrent ischaemic stroke and mortality21, 22). However, whether baseline LDL-C has a definite association with stroke recurrence and functional outcome after acute ischaemic cerebrovascular events has not been well established. In the SPARCL (Stroke Prevention by Aggressive Reduction in Cholesterol Levels) trial, baseline LDL-C is not a predictor of recurrent stroke in patients with recent stroke or TIA and no coronary heart disease8). Similarly, a null association has also been reported in a subanalysis of the Perindopril Protection Against Recurrent Stroke Study (PROGRESS)7). However, our results indicated that baseline LDL-C level was positively associated with recurrent stroke and composite vascular events at 3-month follow-up. This is partly in line with the finding from a post hoc analysis of the Clopidogrel in High-Risk Patients with Acute Nondisabling Cerebrovascular Events (CHANCE) trial, which observed a higher risk of stroke recurrence in patients who do not receive lipid-lowering therapy than in those who received lipid-lowering therapy when baseline LDL-C ≥ 2.6mmol/L6). The possible reason for the inconsistency between our results and those of SPARCL and PROGRESS may due to differences in the time from stroke or TIA onset to enrollment (PROGRESS: patients with a proven TIA or stroke in the past 5 years; SPARCL: patients who had an ischaemic or hemorrhagic stroke or TIA 1-6 months prior to randomization; CNSR-III: diagnosed the index event of ischaemic stroke or TIA with a symptom onset within 7 days), the patients’ characteristics and the duration of follow-up. In our study, baseline LDL-C was only associated with short-term outcomes, but not with long-term outcomes, possibly due to the fact that the administration of lipid-lowering therapy during the follow-up periods could improve the adverse effects of LDL-C on stroke outcomes, as suggested by previous findings6). These results suggest that, for patients with high baseline LDL-C, more aggressive lipid-lowering treatment strategies in the hyperacute phase of stroke to reduce their LDL-C level sufficiently in the shortest possible time may improve the prognosis of more patients. The above perspective still needs to be confirmed by further well-designed clinical trials.
Similar to the results of previous studies23-28), we also found patients with lower TG levels had higher all-cause death and worse functional outcome, which was known as the lipid paradox25). The paradox between TG levels and stroke outcomes may have plausible explanations. First, TG level is an indicator of nutritional status29), and low level of TG may indicate the patient’s previous nutritional deficiency, which causes the poor prognosis of stroke23, 30). Besides, TG level is not simply a reflection of nutritional status27). The accumulation of TG in non-adipose cells can counteract inflammatory response31) and fatty acid-induced lipotoxicity32). Therefore, the decrease of TG level may block the accumulation of TG, thereby aggravating the inflammatory response and lipotoxicity, leading to worse clinical outcomes. Previous study found that increased fat intake was associated with reduced risk of ischaemic stroke in men33). For stroke patients with low TG, whether increasing fat intake can improve the prognosis and how to determine the amount of fat intake are issues worthy of further exploration.
The application of LDL-C/TG ratio in the fields of stroke prognosis remains limited and inconclusive. Previously, a few small sample studies have reported a negative association of LDL-C/TG ratio with LDL oxidation degree16), or sdLDL (substituted by LDL-migration index)17, 18), both of which have been suggested to be major promoters of atherogenesis15) and positively correlated with poor functional outcome at 3 months in patients with AIS34). Our study demonstrated that LDL-C/TG ratio was positively associated with recurrent stroke and poor functional outcome in terms of both short-term and long-term effects, which might be inconsistent with the clues provided by the above studies. However, previous finding about the relationship of LDL-C/TG ratio with LDL oxidation degree and sdLDL were not from stroke patients and they did not adjust any confounders16-18), so the correlations need to be further explored in stroke patients. In addition, it has been shown that oxidized LDL (oxLDL), an oxidative modification of LDL under the oxidative stress, exerts a bidirectional regulation of apoA-I gene expression, which acts as a cofactor of lecithin-cholesterol acyltransferase, displays antioxidant properties, and suppresses inflammatory response35). Thus, the role of oxLDL in the early stage of stroke deserves further investigation. Furthermore, this result was in line with our results of single biomarker, that is, the increased risk of clinical outcomes after stroke was associated with increase in LDL-C and decrease in TG, which means an increase in LDL-C/TG ratio. Our results also suggested LDL-C/TG ratio was better than TG or LDL-C in predicting short-term prognosis of stroke, indicating LDL-C/TG ratio may contain more information, which deserves further verification.
Our results indicated LDL/TG ratio was more associated with functional outcomes than recurrent event risk after ischemic stroke or TIA. There are several plausible explanations for the association between LDL-C/TG ratio and functional outcomes. Studies indicate that deficiency in LDL receptor, resulting in reduced degradation of LDL-C in the bloodstream and elevated levels of LDL-C, may contribute to an abnormal increase in blood-brain barrier permeability36). This can lead to leakage of plasma macromolecules into the intercellular space of brain tissue, infiltration of peripheral inflammatory cells into the brain parenchyma, and activation of microglia37, 38), further triggering signal coupling changes between neurons, astrocytes and neuronal apoptosis. In addition, LDL receptor deficiency can promote the release of inflammatory mediators such as interleukin-1β and interleukin-18 39). As mentioned above, the reduction in TG levels contributes to an elevation in the LDL-C/TG ratio, leading to a diminished capacity for combating inflammation31). The imbalance in inflammation regulation may further exacerbate neurological dysfunction. Furthermore, potential nutritional deficiencies among patients with reduced TG levels may impede functional recovery after stroke to some extent23, 30). These findings indicate a potential association between increased blood-brain barrier permeability, dysregulated inflammation, compromised nutritional status and impaired functional recovery among AIS or TIA patients presenting a high LDL-C/TG ratio at admission. Further studies are required to investigate whether interventions targeting inflammation regulation, blood-brain barrier permeability and nutritional status can enhance functional recovery for AIS or TIA patients with a high LDL-C/TG ratio at admission.
Although this study is a prospective multi-center registry study with a relatively large sample size, it still has certain limitations. First, we excluded 1043 (6.9%) patients because of missing LDL-C data on admission or lost to follow-up, which may cause selection bias, although Table S1 in the Supplement shows the baseline characteristics between included and excluded patients were almost comparable and we adjusted multiple confounding factors in the analysis. Secondly, although a previous study enrolling 110 type 2 diabetes patients with hypertriglyceridemia showed that TG/LDL-C ratio was a better predictive marker of LDL-migration index (an indicator of sdLDL) in patients not treated with statins than those treated with statins17), our subgroup analysis showed the associations of LDL-C/TG ratio with clinical outcomes after stroke were not modified by treatment of cholesterol-lowering agents in hospital, indicating the associations were independent of cholesterol-lowering therapy. However, the high usage rates of cholesterol-lowering agents (statins [96.1%], fibrates [0.3%], bile acid chelating resins [0.11%], nicotinic acid and its derivatives [0.11%], and so on) during hospitalization might confuse the true prognostic effects of LDL-C/TG ratio. We only collected baseline LDL-C and TG levels, and we can’t evaluate whether the change of LDL-C/TG ratio would affect the clinical prognosis due to the lack of information. The exploration of the dynamic change in LDL-C/TG ratio might provide more valuable information in future studies. Thirdly, we did not measure the levels of oxLDL and sdLDL, and we cannot determine whether the association between LDL-C/TG and oxLDL or sdLDL in the present population is consistent with the results of previous studies16-18). Furthermore, recent studies have suggested that genetic factors40), cognitive reserve41) and psychological factors42) are related to the clinical prognosis of stroke. This study did not obtain relevant data, and cannot adjust for the above factors, so the existence of residual interference factors cannot be excluded. Finally, the patients included in this registry were mainly mild to moderate stroke, and whether similar results could be obtained in more severe stroke patients needs further investigation.
Conclusion
In conclusion, this explorative study found that LDL-C/TG ratio was associated with the risk of stroke recurrence, composite vascular events, all-cause death and poor functional outcome at 3-month and 1 year follow-up in patients with AIS or TIA, indicating that LDL-C/TG ratio can be served as a simplified, effective, and routine risk stratification tool for the risk of stroke prognosis. Further studies are required to reveal deeper mechanisms to explain the relationship between LDL-C/TG ratio and outcomes following AIS or TIA.
Acknowledgements
We thank all participating hospitals, their physicians, and nurses, CNSR III Steering Committee members and all the participants of the present study.
Conflict of Interest
None.
Financial Support
This work was supported by National Key Research and Development Program of China (2022YFC2502400), National Natural Science Foundation of China (81870905, U20A20358, 82111530203), Training Fund for Open Projects at Clinical Institutes and Departments of Capital Medical University (CCMU2022ZKYXZ009), Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (2019-I2M-5-029), the Capital’s Funds for Health Improvement and Research (2020-1-2041).
Authors’ Contributions
QX and CL conducted the literature review, interpreted the data, and drafted the initial manuscript; QX performed the statistical analyses; XM and AW designed the study and were involved in data interpretation and manuscript preparation; HL, XT, XX, YZ, and XZ contributed to the acquisition of data; PJ, YW interpreted data, reviewed, and revised the manuscript; all authors have read and approved the submitted manuscript.
References
- 1) Khatana C, Saini NK, Chakrabarti S, Saini V, Sharma A, Saini RV and Saini AK: Mechanistic insights into the oxidized low-density lipoprotein-induced atherosclerosis. Oxid Med Cell Longev, 2020; 2020: 5245308
- 2) Michos ED, McEvoy JW and Blumenthal RS: Lipid management for the prevention of atherosclerotic cardiovascular disease. N Engl J Med, 2019; 381: 1557-1567
- 3) Cholesterol Treatment Trialists C, Baigent C, Blackwell L, Emberson J, Holland LE, Reith C, Bhala N, Peto R, Barnes EH, Keech A, Simes J and Collins R: Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet, 2010; 376: 1670-1681
- 4) Silverman MG, Ference BA, Im K, Wiviott SD, Giugliano RP, Grundy SM, Braunwald E and Sabatine MS: Association between lowering LDL-C and cardiovascular risk reduction among different therapeutic interventions: a systematic review and meta-analysis. JAMA, 2016; 316: 1289-1297
- 5) Schwartz GG, Steg PG, Szarek M, Bhatt DL, Bittner VA, Diaz R, Edelberg JM, Goodman SG, Hanotin C, Harrington RA, Jukema JW, Lecorps G, Mahaffey KW, Moryusef A, Pordy R, Quintero K, Roe MT, Sasiela WJ, Tamby JF, Tricoci P, White HD, Zeiher AM, Committees OO and Investigators: Alirocumab and cardiovascular outcomes after acute coronary syndrome. N Engl J Med, 2018; 379: 2097-2107
- 6) Pan Y, Wangqin R, Li H, Jin A, Li J, Lin J, Meng X, Xian Y, Laskowitz DT and Wang Y: LDL-C levels, lipid-lowering treatment and recurrent stroke in minor ischaemic stroke or TIA. Stroke Vasc Neurol, 2022; 7: 276-284
- 7) Patel A, Woodward M, Campbell DJ, Sullivan DR, Colman S, Chalmers J, Neal B and MacMahon S: Plasma lipids predict myocardial infarction, but not stroke, in patients with established cerebrovascular disease. Eur Heart J, 2005; 26: 1910-1915
- 8) Amarenco P, Goldstein LB, Callahan A, 3rd, Sillesen H, Hennerici MG, O’Neill BJ, Rudolph AE, Simunovic L, Zivin JA, Welch KM and Investigators S: Baseline blood pressure, low- and high-density lipoproteins, and triglycerides and the risk of vascular events in the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial. Atherosclerosis, 2009; 204: 515-520
- 9) Mach F, Baigent C, Catapano AL, Koskinas KC, Casula M, Badimon L, Chapman MJ, De Backer GG, Delgado V, Ference BA, Graham IM, Halliday A, Landmesser U, Mihaylova B, Pedersen TR, Riccardi G, Richter DJ, Sabatine MS, Taskinen MR, Tokgozoglu L, Wiklund O and Group ESCSD: 2019 ESC/EAS Guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk. Eur Heart J, 2020; 41: 111-188
- 10) Hayashi T, Koba S, Ito Y and Hirano T: Method for estimating high sdLDL-C by measuring triglyceride and apolipoprotein B levels. Lipids Health Dis, 2017; 16: 21
- 11) Hirano T, Ito Y, Koba S, Toyoda M, Ikejiri A, Saegusa H, Yamazaki J and Yoshino G: Clinical significance of small dense low-density lipoprotein cholesterol levels determined by the simple precipitation method. Arterioscler Thromb Vasc Biol, 2004; 24: 558-563
- 12) Ivanova EA, Myasoedova VA, Melnichenko AA, Grechko AV and Orekhov AN: Small dense low-density lipoprotein as biomarker for atherosclerotic diseases. Oxid Med Cell Longev, 2017; 2017: 1273042
- 13) Higashioka M, Sakata S, Honda T, Hata J, Yoshida D, Hirakawa Y, Shibata M, Goto K, Kitazono T, Osawa H and Ninomiya T: Small dense low-density lipoprotein cholesterol and the risk of coronary heart disease in a Japanese community. J Atheroscler Thromb, 2020; 27: 669-682
- 14) Ikezaki H, Lim E, Cupples LA, Liu CT, Asztalos BF and Schaefer EJ: Small dense low-density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study. J Am Heart Assoc, 2021; 10: e019140
- 15) Yoshida H and Kisugi R: Mechanisms of LDL oxidation. Clin Chim Acta, 2010; 411: 1875-1882
- 16) Brizzi P, Tonolo G, Carusillo F, Malaguarnera M, Maioli M and Musumeci S: Plasma lipid composition and LDL oxidation. Clin Chem Lab Med, 2003; 41: 56-60
- 17) Ouchi G, Komiya I, Taira S, Wakugami T and Ohya Y: Triglyceride/low-density-lipoprotein cholesterol ratio is the most valuable predictor for increased small, dense LDL in type 2 diabetes patients. Lipids Health Dis, 2022; 21: 4
- 18) Yoshida A, Kouwaki M, Matsutani Y, Fukuchi Y and Naito M: Usefulness of serum total cholesterol/triglyceride ratio for predicting the presence of small, dense LDL. J Atheroscler Thromb, 2004; 11: 215-219
- 19) Wang Y, Jing J, Meng X, Pan Y, Wang Y, Zhao X, Lin J, Li W, Jiang Y, Li Z, Zhang X, Yang X, Ji R, Wang C, Wang Z, Han X, Wu S, Jia Z, Chen Y and Li H: The Third China National Stroke Registry (CNSR-III) for patients with acute ischaemic stroke or transient ischaemic attack: design, rationale and baseline patient characteristics. Stroke Vasc Neurol, 2019; 4: 158-164
- 20) Adams HP, Jr., Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL and Marsh EE, 3rd: Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke, 1993; 24: 35-41
- 21) Amarenco P, Kim JS, Labreuche J, Charles H, Abtan J, Bejot Y, Cabrejo L, Cha JK, Ducrocq G, Giroud M, Guidoux C, Hobeanu C, Kim YJ, Lapergue B, Lavallee PC, Lee BC, Lee KB, Leys D, Mahagne MH, Meseguer E, Nighoghossian N, Pico F, Samson Y, Sibon I, Steg PG, Sung SM, Touboul PJ, Touze E, Varenne O, Vicaut E, Yelles N, Bruckert E and Treat Stroke to Target I: A comparison of two LDL cholesterol targets after ischemic stroke. N Engl J Med, 2020; 382: 9
- 22) Bohula EA, Wiviott SD, Giugliano RP, Blazing MA, Park JG, Murphy SA, White JA, Mach F, Van de Werf F, Dalby AJ, White HD, Tershakovec AM, Cannon CP and Braunwald E: Prevention of stroke with the addition of ezetimibe to statin therapy in patients with acute coronary syndrome in IMPROVE-IT (Improved Reduction of Outcomes: Vytorin Efficacy International Trial). Circulation, 2017; 136: 2440-2450
- 23) Dziedzic T, Slowik A, Gryz EA and Szczudlik A: Lower serum triglyceride level is associated with increased stroke severity. Stroke, 2004; 35: e151-152
- 24) Weir CJ, Sattar N, Walters MR and Lees KR: Low triglyceride, not low cholesterol concentration, independently predicts poor outcome following acute stroke. Cerebrovasc Dis, 2003; 16: 76-82
- 25) Cheng KH, Lin JR, Anderson CS, Lai WT, Lee TH and Group S: Lipid paradox in statin-naive acute ischemic stroke but not hemorrhagic stroke. Front Neurol, 2018; 9: 541
- 26) Jain M, Jain A, Yerragondu N, Brown RD, Rabinstein A, Jahromi BS, Vaidyanathan L, Blyth B and Stead LG: The triglyceride paradox in stroke survivors: a prospective study. Neurosci J, 2013; 2013: 870608
- 27) Ryu WS, Lee SH, Kim CK, Kim BJ and Yoon BW: Effects of low serum triglyceride on stroke mortality: a prospective follow-up study. Atherosclerosis, 2010; 212: 299-304
- 28) Li W, Liu M, Wu B, Liu H, Wang LC and Tan S: Serum lipid levels and 3-month prognosis in Chinese patients with acute stroke. Adv Ther, 2008; 25: 329-341
- 29) Adlouni A, Ghalim N, Benslimane A, Lecerf JM and Saile R: Fasting during Ramadan induces a marked increase in high-density lipoprotein cholesterol and decrease in low-density lipoprotein cholesterol. Ann Nutr Metab, 1997; 41: 242-249
- 30) Collaboration FT: Poor nutritional status on admission predicts poor outcomes after stroke: observational data from the FOOD trial. Stroke, 2003; 34: 1450-1456
- 31) van Dierendonck X, Vrieling F, Smeehuijzen L, Deng L, Boogaard JP, Croes CA, Temmerman L, Wetzels S, Biessen E, Kersten S and Stienstra R: Triglyceride breakdown from lipid droplets regulates the inflammatory response in macrophages. Proc Natl Acad Sci U S A, 2022; 119: e2114739119
- 32) Listenberger LL, Han X, Lewis SE, Cases S, Farese RV, Jr., Ory DS and Schaffer JE: Triglyceride accumulation protects against fatty acid-induced lipotoxicity. Proc Natl Acad Sci U S A, 2003; 100: 3077-3082
- 33) Gillman MW, Cupples LA, Millen BE, Ellison RC and Wolf PA: Inverse association of dietary fat with development of ischemic stroke in men. JAMA, 1997; 278: 2145-2150
- 34) Zhou P, Liu J, Wang L, Feng W, Cao Z, Wang P, Liu G, Sun C, Shen Y, Wang L, Xu J, Meng P, Li Z, Xu WY and Lan X: Association of small dense low-density lipoprotein cholesterol with stroke risk, severity and prognosis. J Atheroscler Thromb, 2020; 27: 1310-1324
- 35) Nekrasova EV, Larionova EE, Danko K, Kuzmina DO, Shavva VS, Kudriavtsev IV and Orlov SV: Regulation of apolipoprotein A-I gene expression in human macrophages by oxidized low-density lipoprotein. Biochemistry (Mosc), 2021; 86: 1201-1213
- 36) de Oliveira J, Engel DF, de Paula GC, Dos Santos DB, Lopes JB, Farina M, Moreira ELG and de Bem AF: High cholesterol diet exacerbates blood-brain barrier disruption in LDLr-/- Mice: impact on cognitive function. J Alzheimers Dis, 2020; 78: 97-115
- 37) Yu LE, Lai CL, Lee CT and Wang JY: Highly electronegative low-density lipoprotein L5 evokes microglial activation and creates a neuroinflammatory stress via Toll-like receptor 4 signaling. J Neurochem, 2017; 142: 231-245
- 38) Damisah EC, Rai A and Grutzendler J: TREM2: Modulator of lipid metabolism in microglia. Neuron, 2020; 105: 759-761
- 39) Sun R, Peng M, Xu P, Huang F, Xie Y, Li J, Hong Y, Guo H, Liu Q and Zhu W: Low-density lipoprotein receptor (LDLR) regulates NLRP3-mediated neuronal pyroptosis following cerebral ischemia/reperfusion injury. J Neuroinflammation, 2020; 17: 330
- 40) Ibanez L, Heitsch L, Carrera C, Farias FHG, Del Aguila JL, Dhar R, Budde J, Bergmann K, Bradley J, Harari O, Phuah CL, Lemmens R, Viana Oliveira Souza AA, Moniche F, Cabezas-Juan A, Arenillas JF, Krupinksi J, Cullell N, Torres-Aguila N, Muino E, Carcel-Marquez J, Marti-Fabregas J, Delgado-Mederos R, Marin-Bueno R, Hornick A, Vives-Bauza C, Navarro RD, Tur S, Jimenez C, Obach V, Segura T, Serrano-Heras G, Chung JW, Roquer J, Soriano-Tarraga C, Giralt-Steinhauer E, Mola-Caminal M, Pera J, Lapicka-Bodzioch K, Derbisz J, Davalos A, Lopez-Cancio E, Munoz L, Tatlisumak T, Molina C, Ribo M, Bustamante A, Sobrino T, Castillo-Sanchez J, Campos F, Rodriguez-Castro E, Arias-Rivas S, Rodriguez-Yanez M, Herbosa C, Ford AL, Gutierrez-Romero A, Uribe-Pacheco R, Arauz A, Lopes-Cendes I, Lowenkopf T, Barboza MA, Amini H, Stamova B, Ander BP, Sharp FR, Kim GM, Bang OY, Jimenez-Conde J, Slowik A, Stribian D, Tsai EA, Burkly LC, Montaner J, Fernandez-Cadenas I, Lee JM and Cruchaga C: Multi-ancestry GWAS reveals excitotoxicity associated with outcome after ischaemic stroke. Brain, 2022; 145: 2394-2406
- 41) Gallucci L and Umarova RM: Association of cognitive reserve with stroke outcome: a protocol for a systematic review. BMJ Open, 2022; 12: e059378
- 42) Gurkova E, Sturekova L, Mandysova P and Sanak D: Factors affecting the quality of life after ischemic stroke in young adults: a scoping review. Health Qual Life Outcomes, 2023; 21: 4
