Monocyte to high‐density lipoprotein ratio predicts clinical outcomes after acute ischemic stroke or transient ischemic attack

Abstract Aims The monocyte to high‐density lipoprotein cholesterol ratio (MHR) has emerged as a novel inflammatory biomarker of atherosclerotic cardiovascular disease. However, it has not yet been identified whether MHR can predict the long‐term prognosis of ischemic stroke. We aimed to investigate the associations of MHR levels with clinical outcomes in patients with ischemic stroke or transient ischemic attack (TIA) at 3 months and 1 year. Methods We derived data from the Third China National Stroke Registry (CNSR‐III). Enrolled patients were divided into four groups by quartiles of MHR. Multivariable Cox regression for all‐cause death and stroke recurrence and logistic regression for the poor functional outcome (modified Rankin Scale score 3–6) were used. Results Among 13,865 enrolled patients, the median MHR was 0.39 (interquartile range, 0.27–0.53). After adjustment for conventional confounding factors, the MHR level in quartile 4 was associated with an increased risk of all‐cause death (hazard ratio [HR], 1.45; 95% confidence interval [CI], 1.10–1.90), and poor functional outcome (odd ratio [OR], 1.47; 95% CI, 1.22–1.76), but not with stroke recurrence (HR, 1.02; 95% CI, 0.85–1.21) at 1 year follow‐up, compared with MHR level in quartile 1. Similar results were observed for outcomes at 3 months. The addition of MHR to a basic model including conventional factors improved predictive ability for all‐cause death and poor functional outcome validated by the C‐statistic and net reclassification index (all p < 0.05). Conclusions Elevated MHR can independently predict all‐cause death and poor functional outcome in patients with ischemic stroke or TIA.


| INTRODUC TI ON
Atherosclerosis, especially intracranial atherosclerosis, is the key etiological factor of ischemic stroke. Inflammation and lipid accumulation are recognized as fundamental to the pathophysiology and progression of atherosclerosis. 1,2 Several inflammatory factors have been associated with strokes, such as neutrophil-to-lymphocyte ratio, 3 C-reactive protein (CRP) or hypersensitive CRP (hsCRP), 4 and d-dimer. 5 Monocytes, as agents of the innate immune response, 6,7 interact primarily with platelets and endothelial cells to aggravate inflammatory, pro-thrombotic pathways which in turn cause the formation, progression, and rupture of atherosclerotic plaques. 8,9 As such, monocytes are emerging as indicators of post-ischemic stroke inflammation. 10 Lipid levels, especially the high-density lipoprotein (HDL) cholesterol, can impact monocyte development. Murine models showed that HDL can suppress stem cell proliferation and then reduce monocytosis. 11,12 HDL is inversely associated with intermediate monocyte counts 13,14 and can counteract the pro-inflammatory and pro-oxidant effects of monocytes, 15 thus reducing the risk of atherosclerotic events. Recently, the monocyte count to HDL ratio (MHR), as calculated by dividing the monocyte count by the HDL value, has been recognized as a novel inflammatory biomarker of atherosclerotic cardiovascular diseases, 16 including ischemic stroke. 17 It has been reported to be associated with poor outcomes among patients with the acute coronary syndrome, 18 the recurrence of atrial fibrillation, 19 stroke-associated pneumonia in patients with acute ischemic stroke (AIS), 20 and the risk of disability or death in patients with cerebral hemorrhage. 21 Nevertheless, to the best of our knowledge, evidence on the role of MHR in the prognosis of ischemic stroke is limited and conflicting. Bolayir et al. 22 conducted a case-control study including 466 patients with AIS and found that high MHR was associated with an increased risk of 30-day mortality after stroke. Another observational study involving 803 patients with AIS found no association between MHR and 3 months mortality. 20 In addition, it has not yet been identified whether MHR can predict recurrent stroke, poor functional outcome, and long-term mortality after ischemic stroke using prospective cohort data.
Therefore, this study aimed to investigate the associations be-  elsewhere. 23 The study was approved by the ethics committee of Beijing Tiantan Hospital and participant hospitals. Written informed consent was obtained from the patients or their legally authorized representatives.

| Data collection and calculation
Baseline data on demographics and clinical characteristics, including age, sex, body mass index, smoking and alcohol consumption status, diseases history (hypertension, stroke or TIA, diabetes, dyslipidemia, atrial fibrillation, coronary heart disease, heart failure, peripheral vascular disease, and arthritis), prestroke modified Rankin Score (mRS) score, the National Institutes of Health Stroke Scale (NIHSS) score at admission, and medications provided in hospital care (antihypertensive agents, antiplatelet agents, anticoagulant agents, cholesterol-lowering agents, and hypoglycemic agents), were collected by trained research coordinators at each center via face-toface interviews or medical records. The etiological classification of the index events was performed by the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria. 24 Fasting whole blood samples from venipuncture were collected in vacutainer tubes containing EDTA within 24 h of admission.
Afterward, the monocyte count was tested by an automated hematology analyzer at each research center. The blood samples were frozen in a cryotube at −80°C refrigerator and were transported to the central laboratory in Beijing Tiantan Hospital by cold chain.
Serum total cholesterol (TC), HDL, low-density lipoprotein (LDL), Medical University, Grant/Award Number: CCMU2022ZKYXZ009 Similar results were observed for outcomes at 3 months. The addition of MHR to a basic model including conventional factors improved predictive ability for all-cause death and poor functional outcome validated by the C-statistic and net reclassification index (all p < 0.05).
Conclusions: Elevated MHR can independently predict all-cause death and poor functional outcome in patients with ischemic stroke or TIA.

K E Y W O R D S
death, functional outcome, high-density lipoprotein, monocyte count, stroke and triglyceride (TG) were analyzed centrally. All laboratory indicators were performed by laboratory personnel who were unaware of patients' clinical characteristics. MHR was calculated as the ratio of blood monocyte count to HDL concentration.

| Outcome assessment
Patients were followed up by trained research coordinators who were blinded to subjects' baseline characteristics at 3 months and 1 year after symptom onset. Data regarding all-cause death, stroke recurrence, and modified Rankin Scale (mRS) score were collected.
All-cause death was defined as death from any cause and confirmed by a death certification from the attended hospital or the local citizen registry. The stroke recurrence was defined as a new ischemic stroke or hemorrhagic stroke within 3 months and 1 year after symptom onset. Poor functional outcome was defined as mRS score ranging from 3 to 6 or from 2 to 6 at 3 months and 1 year. The definitions of the above outcomes were consistent with those previously published in the CNSR-III protocol. 23  The Kaplan-Meier method and the log-rank test were used for time-to-event data. The associations of MHR with all-cause death and stroke recurrence were explored by Cox proportion hazards models, and hazard ratios (HRs) and 95% confidence intervals (CIs) were reported. The proportionality assumption was assessed by scaled Schoenfeld residuals, and there was no violation of the assumptions. For poor functional outcomes, odds ratio (OR) with 95% CI was estimated by the logistic regression model. A robust sandwich variance estimator was used to deal with the correlations for clustering by hospital. We fitted an unadjusted model with no confounding factors and three adjusted models. Variables adjusted in model 1 included age and sex, and in model 2 further included body mass index, current smoker, current alcohol drinking, disease history (hypertension, stroke or TIAs, diabetes, dyslipidemia, atrial fibrillation, arthritis), NIHSS score at admission, stroke subtype, prestroke mRS score, stroke etiology, antihypertensive agents, cholesterollowering agents, and hypoglycemic agents. Model 3 was additionally adjusted for TC, TG, and LDL. A linear trend test was performed by treating the median MHR value of each quartile as a continuous variable in each model. Moreover, we used a restricted cubic spline with adjustment for potential covariates to evaluate the patterns of relationships between continuous MHR value and clinical outcomes.

| Statistical analysis
Additionally, C statistics, integrated discrimination improvement (IDI), and net reclassification index (NRI) were calculated to establish the predictive performance of MHR added to the basic model.
Subgroup analyses were performed according to age, sex, BMI, NIHSS score, history of stroke or TIA, and stroke etiology with an interaction test. All analyses were performed using SAS software version 9.4 (SAS Institute Inc.), and R version 4.0.2. Overall, a twosided p-value of <0.05 was assumed statistically significant.

| Baseline characteristics
Among the enrolled patients in the CNSR-III, we excluded patients without baseline monocyte (n = 293), HDL (n = 601), or who were lost to follow-up (n = 407), and accordingly, a total of 13,865 patients were included in our analyses. Table S1 showed that baseline characteristics of included and excluded patients were largely comparable, except that the included patients were more likely to be younger, females, non-smokers, had lower NIHSS scores at admission, and received a higher proportion of TIA and antiplatelet agents but a lower proportion of anticoagulants and rt-PA intravenous thrombolytic.
The baseline characteristics of our study population stratified according to the MHR quartile are shown in Table 1. Compared to patients with a lower MHR, those in the highest quartile group were more likely to be younger, males, current smokers, alcohol drinkers, having higher BMI and prestroke mRS score, higher prevalence of medical history (including hypertension, diabetes, dyslipidemia, and atrial fibrillation) and large-artery atherosclerosis ischemic stroke, higher TG concentration, and higher proportion of antihypertensive agents and hypoglycemic agents, but lower arthritis, TC concentration, and LDL concentration.

| Incremental predictive value of MHR
We evaluated the performance of models with MHR to predict all-cause death, stroke recurrence, and poor functional outcome (   Abbreviations: MHR indicates monocyte-to-HDL ratio; mRS, modified Rankin Scale; SD, standard deviation. the general population. 25 Another prospective observational study suggested that MHR was an independent predictor of major adverse cardiovascular events' occurrence in patients undergoing coronary angiography. 16 Recently, a meta-analysis that involved 9 studies indicated that for patients with coronary heart disease, increased MHR value might be associated with higher long-term mortality and major adverse cardiovascular events. 26 Nevertheless, some studies have explored the relationship between MHR and short-term stroke prognosis based on small sample size data from one single center but reported inconsistent results for hemorrhagic transformation 27,28 and for all-cause mortality. 20,22 The discrepancy might be due to the differences in the patients' characteristics, and the small sample size with limited statistical power. Furthermore, evidence surrounding the relationship between MHR and the long-term prognosis of stroke was limited. Our study confirmed the correlations between MHR and short-and long-term poor functional outcomes and F I G U R E 2 Association between monocyte to HDL ratio and clinical outcomes. Adjusted for age, sex, body mass index, current smoker, current alcohol drinking, disease history (hypertension, stroke or transient ischemic attacks, diabetes, dyslipidemia, atrial fibrillation, and arthritis), the National Institutes of Health Stroke Scale score at admission, stroke subtype, prestroke modified Rankin Scale score, stroke etiology, antihypertensive agents, cholesterol-lowering agents, hypoglycemic agents, total cholesterol, triglyceride, and low-density lipoprotein.
all-cause death in ischemic stroke or TIA patients, especially among younger or LAA and cardiogenic embolism patients, in accordance with the previously reported relationship between higher levels of MHR and poor functional outcome at 3 months. 29,30 In addition, we further proved the predictive value of MHR for poor functional outcome and all-cause death after stroke using comprehensive evaluation indicators including C statistics, IDI, and NRI. To the best of our knowledge, no previous study investigated the relationship between MHR and post-stroke recurrence. Our study showed a null association between MHR and recurrent stroke in patients with a minor ischemic stroke or TIA; however, further research is warranted to validate this result.
Atherosclerosis is a highly prevalent cause of ischemic stroke, 31 due to associated inflammation and lipid accumulation. 1,2 Monocytes are an important source of pro-inflammatory substances during atherosclerosis, 8,9,32 and have found to be obviously associated with elevated risks of death and poor functional outcomes at 3-months and 1-year follow-up in patients with acute ischemic cerebrovascular events. 33 In the onset of AIS, monocytes bind to adhesion molecules on injured vascular endothelium, migrate to the sub-endothelial space, and mature into macrophages, then phagocytose oxidized LDL and differentiate into foam cells. 34 Foam cells can release a variety of pro-inflammatory and prooxidative cytokines at the inflammatory site, attracting T lymphocytes and more monocytes to gather. 35 Monocytes directly or indirectly participate in all stages of atherosclerosis from fat streak formation to plaque rupture. 22 HDL has anti-inflammatory, antioxidant, and antithrombotic effects, and is a classic antiatherosclerotic factor. 36,37 A prospective cohort study found that changes in HDL proteins during the early acute phase of stroke was associated with recovery. 38 Another study enrolled AIS patients with diabetes reported that HDL levels were negatively correlated with the risk of recurrent stroke and major adverse cardiovascular events within 1 year. 39 HDL has also been found to TA B L E 3 Performance of models with MHR to predict all-cause death, stroke recurrence and poor functional outcome. be an independent risk factor for 3-month poor outcomes in patients with ischemic stroke. 40 This may be related to the fact that HDL can prevent monocytes from recruiting to the arterial wall, thereby inhibiting the adhesion of monocytes to endothelial cells, reducing the activation of monocytes and the proliferation of monocyte progenitor cells, 41 thereby protecting endothelial cells from inflammation and oxidative stress. 19 Therefore, we speculate that when monocyte levels increase and/or HDL levels decrease, patients with ischemic stroke would be more adversely affected by inflammation and lipid accumulation, and the degree of atherosclerosis could be more severe. That is, high MHR levels are more likely to lead to irreversible damage to brain tissue, which in turn affects all-cause death and poor functional outcomes in ischemic stroke. This is consistent with our analyses of subgroups of stroke subtypes. Furthermore, younger, LAA, and cardioembolism patients were more likely to have higher MHR levels in our study.
The associations between MHR and clinical outcomes were more pronounced among younger patients after multivariable adjustment for important potential confounders. This may be related to the following mechanism. Hearps et al. found that the expression of CD11b, L-selectin, and TLR4 in the elderly may affect the migration of monocytes and lead to more dysfunction of monocytes than in young people. 42 In addition, it was found that the level of P38 in older mice was significantly lower than that in young mice in an animal study. 43 HDL further restrains monocytes by inhibiting the expression of P38 and CD11b. 41,44 Therefore, eld may be a protective factor. In addition to LAA, we also obtained similar results among the cardioembolism patients, which are different from previous studies. 29,30 We speculate that the degree of correlation between MHR levels and 1-year poor functional prognosis in cardioembolism is related to the inflammatory potential of monocytes or HDL. Monocytes adhered to the surface of endothelial cells can induce the production of various cytokines in damaged tissues and stimulate the proliferation of fibroblasts. 45 Extensive proliferation of fibroblasts can increase atrial fibrosis, alter atrial conduction characteristics, 46 induce atrial fibrillation, and cause shedding of cardiac embolus. However, HDL can have a protective effect on atrial muscle cells by increasing the directional transport of cholesterol. 19 The interaction between inflammation and lipids can not only promote the occurrence and development of atherosclerosis but also drive the healing response of vascular injury, which is not easy to cause stroke recurrence and may play a protective role, so we have not found a correlation between high levels of MHR and stroke recurrence. 1,2,47 Recently, with the continuous progress of gene-level research, the differences in the transcriptional process of monocytes after ischemic stroke may be another potential reason for the different tendencies of different subtypes of stroke. 48 In addition, recent studies reported some novel prognostic biomarkers for predicting stroke outcomes, such as circulating immunoregulatory lymphocytes, 49 neurofilament light chain protein, 50 and circulating microRNA. 51 However, we could not adjust them due to the lack of data. Therefore, we cannot conclude the casual relationship. We cannot establish a causal relationship between MHR with clinical outcomes after stroke. Finally, HDL-specific genotypes or phenotypes (e.g., familial hypercholesterolemia or hyperalphalipoproteinemia) were not screened in our study, our results may not be generalizable to a more heterogeneous global population.

| CON CLUS IONS
In conclusion, our study suggested that elevated MHR levels at admission substantially increased risks of all-cause death and poor functional outcomes in patients with ischemic stroke or TIA at 3month and 1-year follow-ups. The finding underscored the utility and predictive potential of MHR, which can be swiftly applied clinically to determine the prognosis of stroke.

AUTH O R CO NTR I B UTI O N S
Q.X. and Q.W. conducted the literature review, interpreted the data, and drafted the initial manuscript; Q.X. performed the statistical analyses; L.C., interpreted data, reviewed, and revised the manuscript. X.M. and A.W. designed the study and were involved in data interpretation and manuscript preparation; H.L., X.T., X.X., Y.Z., X.Z, and Y.J.W. contributed to the acquisition of data; Y.L., Y.P.W., and Y.J.W. revised the manuscript; and all authors have read and approved the submitted manuscript.

ACK N OWLED G M ENTS
We thank all participating hospitals, their physicians, and nurses, CNSR III Steering Committee members, and all the participants of the present study.