Dehydration in cerebral venous sinus thrombosis

Abstract Aims This study aimed to unravel the dehydration status of patients with cerebral venous sinus thrombosis (CVST) to facilitate the understanding of dehydration in CVST. Methods This was a multicenter retrospective study and three populations were recruited, namely, patients with CVST, CVST mimics, and healthy subjects. Blood samples were obtained 1–2 days after admission to assess dehydration status. Stata 15.1 was performed for statistical analysis. Results A total of 208 patients were diagnosed with CVST, 237 with CVST mimics, and 200 healthy individuals were enrolled. The urine specific gravity (USG, 1.020 [1.014, 1.029] vs. 1.017 [1.011, 1.021]) was higher in patients with CVST than in those with mimics (all p < 0.001). The percentage of USG >1.03 was also higher in CVST (22.6%) than in its mimics (6.3%, p < 0.001). With the development of CVST, USG (acute vs. sub‐acute vs. chronic, 1.022 [1.015, 1.033] vs. 1.021 [1.015, 1.031] vs. 1.019 [1.014, 1.025]) decreased. All dehydration‐related markers could not differentiate CVST from its mimics and healthy populations, and they were not associated with CVST severity and prognosis (p > 0.05). Conclusion High levels of USG, especially USG >1.013, were more common in patients with CVST. Dehydration‐related indices could not characterize CVST and were not associated with CVST severity and prognosis.

[9] Therefore, we hypothesized that these markers would be abnormal in patients with CVST.However, investigations on the relationship between dehydration and CVST are still lacking.Accordingly, this study aimed to unravel the dehydration status of patients with CVST to facilitate the understanding of dehydration in CVST and provide novel insights into the risk factors, diagnostic methods, and therapeutic paradigms for CVST in the future.

| Study population
This was a multicenter retrospective observational study conducted in the neurology department of Tianjin Medical University General Hospital, Xuanwu Hospital Capital Medical University, and Tianjin Huanhu Hospital.This study was conducted in accordance with the guidelines of the 1964 Declaration of Helsinki, and all procedures performed in this study involving human participants were approved by the Institutional Ethics Committee.A total of three types of populations were recruited between April 2018 to September 2023, namely, patients with CVST, CVST mimics, and healthy subjects.All the enrolled patients were assessed with magnetic resonance venography (MRV) and/or computed tomography venography (CTV) and confirmed CVST or CVST mimics using the magnetic resonance black-blood thrombus imaging (MRBTI) technique or digital subtraction angiography (DSA), which can visualize the thrombus directly. 10cording to the time of symptom onset, CVST was divided into acute (≤15 days), sub-acute (15-30 days), and chronic (>30 days) stages.1][12] In other words, CVST mimics could be also termed as nonthrombotic cerebral venous sinus stenosis.Patients were excluded if they had (1) other austere diseases; (2) a history of stroke except CVST; and (3) renal dysfunction.People who received physical examination and were without a history of renal dysfunction and severe diseases (such as stroke, cancer, and trauma) were enrolled as healthy subjects.Some chronic disorders, such as hypertension, diabetes, and hyperuricemia, were present in healthy subjects.

| Data collection
Clinical data, including patient demographics, symptoms, imaging presentations, risk factors for CVST, and causes of CVST mimics, were comprehensively recorded.The severity and prognosis of CVST were evaluated using the National Institutes of Health Stroke Scale (NIHSS) and the Modified Rankin Score (mRS) at admission and follow-up.Blood samples were obtained 1-2 days after admission to assess dehydration status, including Cr, urea (UA), USG, hematocrit, blood sodium (Na + ), blood potassium (K + ), glucose, uric acid, D-dimer, and fibrinogen.UA/Cr and plasma osmolality were also calculated to evaluate dehydration.Plasma osmolality was estimated using the following equation: [2 × (Na + + K + ) + glucose + UA].Because the enrolled patients did not have renal diseases or renal dysfunction, the dehydration-related indices assessed in this study mainly reflected blood volume and nutritional status.Peripheral blood collected from healthy subjects only included Cr, UA, hematocrit, glucose, uric acid, and fibrinogen because of incomplete examination.

| Statistical analysis
Stata 15.1 was performed for statistical analysis in this study.
The Kruskal-Wallis test was used to assess the data distribution.
Continuous variables following the Gaussian distribution were displayed as mean ± standard deviation (SD) and analyzed using student's t-test; otherwise, they were presented as median [interquartile range, IQR] and analyzed using the Mann-Whitney U test.
Categorical variables were depicted as numbers (percentages) and analyzed using a chi-square test.Receiver operating characteristic (ROC) analysis was used to demonstrate the sensitivity and specificity of the variables for differentiating patients with CVST from its mimic and healthy subjects.An area under the curve (AUC) >0.6 with a p-value <0.05, was considered significant.The diagnostic power was also assessed using the positive predictive value (+PV), negative predictive value (− PV), positive likelihood ratio (+LR), and negative likelihood ratio (−LR).The optimal cutoff values were confirmed with maximal sensitivity and specificity.Propensity score matching was performed to eliminate confounding factors.Spearman's rank correlation analysis was performed to identify the association between the biomarkers and CVST severity and prognosis.Missing variables were replaced with mean values (continuous variables) or modes (dichotomous variables) only if they accounted for less than 10% of the total.Statistical significance was set at p < 0.05.However, the significant level α was adjusted using the Bonferroni method to avoid type I errors when analyzing repeated measurement data.vs. 0.4%, p < 0.001) and unconsciousness (9.6% vs. 0.0%, p < 0.001) in patients with CVST were significantly higher than that in patients with CVST mimics; however, the symptoms of dizziness (20.7% vs. 39.2%, p < 0.001) and tinnitus (15.9% vs. 40.9%,p < 0.001) were more commonly seen in patients with CVST mimics.Furthermore, imaging showed that the right transverse sinus (50.0% vs. 17.3%, p < 0.001), superior sagittal sinus (58.2% vs. 2.5%, p < 0.001) and straight sinus (18.3% vs. 1.3%, p < 0.001) were more commonly seen in CVST than in its mimics, and the left transverse sinus (50.0% vs. 73.8%,p < 0.001) was more commonly seen in mimics than in CVST.

| Patients
The most common risk factor for CVST was obstetric causes, followed by anemia and oral contraceptive use.The causes of CVST mimics included sinus-filling defects, sinus atresia/hypoplasia, and giant arachnoid granules.All aforementioned are shown in Table 1.

| Dehydration-related indices in patients with CVST at different stages
We set onset-to-door time ≤15 days as an acute stage, 16-30 days as the sub-acute stage, and > 30 days as a chronic stage.There were 62 patients with CVST in the acute stage (7

| ROC analysis for differentiating CVST from CVST mimics and health subjects
Based on comparisons among patients with CVST, CVST mimics, and healthy subjects, parameters that had between-group differences with a p-value of <0.003 were included in the ROC analysis.As for differentiating CVST from mimics (Table S1), the parameters with AUC For differentiating CVST from healthy subjects (Table S2), parameters with AUC >0.6 only included fibrinogen (AUC, 95%CI: 0.605, 0.548-0.662,p < 0.001).The optimal cut-off fibrinogen level was

| Association of dehydration-related indices to CVST severity and prognosis
Spearman's correlation analysis showed that the glucose level had a mild correlation with the baseline NIHSS (r = 0.400, p < 0.001) and 7-day NIHSS (r = 0.345, p < 0.001) scores in patients with CVST.

| DISCUSS ION
This study investigated the association between dehydration and CVST for the first time and found that (1) a high level of USG, especially for USG >1.013, was more common in patients with CVST than in patients with CVST mimics; (2) with CVST phase development, the level of USG, as well as D-dimer and fibrinogen, gradually decreased close to the levels seen in patients with CVST mimics; and (3) all dehydration-related markers were underpowered to differentiate CVST from its mimics and healthy populations, and they were not associated with CVST severity and prognosis.
Previously, more attention has been paid to dehydration in patients with AIS. 7 It is likely that dehydration can cause contraction of the total plasma volume, increase blood viscosity, reduce cardiac output, and retard collateral circulation formation in the brain. 13,14erefore, dehydration is regarded as a risk factor for AIS and is associated with AIS severity and prognosis. 15In theory, dehydration promotes thrombus formation not only in the arteries but also in the veins. 5Kelly et al. found that dehydration after AIS was strongly and independently correlated to venous thromboembolism. 5However, it remains unclear whether dehydration plays an important role in CVST.6][7] Therefore, we hypothesized that dehydration is correlated with the occurrence, development, and prognosis of CVST.
A total of 645 people were recruited for this study in order to identify a link between dehydration and CVST.In general, USG, UA/Cr, plasma osmolality, and hematocrit levels can reflect dehydration status. 7Therefore, we investigated the differences in these indices among the cohorts.We found that the level of USG and the percentage of patients with USG >1.03 were significantly higher in patients with CVST than in those with CVST mimics.Even after propensity score matching, the percentage of patients with USG >1.03 was still higher in the CVST group than that in the CVST mimic group.
intracranial pressure, which causes vomiting and nausea, leading to malnutrition.This seems to explain why UA levels were lower in patients with CVST.UA/Cr and BUN/Cr are acknowledged indicators for assessing dehydration status. 7However, the UA/Cr ratio was significantly higher in patients with CVST mimics than in patients with CVST or the healthy population.This result seemed to contrast with the USG results.However, the actual differences in UA/Cr among the three cohorts were small despite statistical significance, and the differences diminished after propensity score matching.Hence, the UA/Cr ratio could not accurately reflect the dehydration status of the CVST patients.Our results showed that the differences in plasma osmolality and Na + concentration between patients with CVST and those with its mimics were statistically significant, even after propensity score matching.However, the differences were too small to guide a clinical diagnosis.Previous studies have found that plasma osmolality is associated with the occurrence, development, and prognosis of AIS. 8 However, our results did not show a correlation in patients with CVST.Plasma osmolality reflects dehydration status, but it is also affected by abnormal ions, glucose, and proteins.
Moreover, unlike direct detection, the plasma osmolality estimated using the equation used in this study may have influenced the final results.Interestingly, the concentrations of Cr and hematocrit levels and the number of patients with hematocrit levels >45% were higher in the healthy population.Patients with CVST and CVST mimics frequently experience nausea, vomiting, and headache, which lead to loss of appetite and even malnutrition, resulting in lower Cr and hematocrit levels >45%.
We also found that as the stage progressed, dehydration-related markers in patients with CVST gradually approached the levels in patients with CVST mimics.The thrombus in the CVST in the chronic stage is organized and develops into stenosis. 10Although patients with CVSS This observational study has several limitations.First, dehydration-related indices detected in this study were scarce.
Some dehydration-related indices such as urine osmolality and glomerular filtration rate were not assessed in the enrolled participants.Furthermore, some dehydration-related markers, including USG, Na + , K + , D-dimer, and plasma osmolality were not detected in healthy individuals.As in any retrospective observational study, the lack of assessment is a major limitation, especially in rare diseases such as CVST.Second, despite the large sample size in this study, the results still seem to be in doubt because the elevation of USG reflected dehydration status in patients with CVST, but reduction of plasma osmolality and hematocrit seemed to indicate a nondehydration status in these patients.We considered that the differences in plasma osmolality among the three cohorts were too small to reach significance, and the lower hematocrit in CVST was likely to be attributed to malnutrition, but the hypothesis was not accurate enough to form convincing conclusions.However, owing to inconsistent evidence, we cannot confirm that dehydration was observed in patients with CVST.Future multicenter, well-designed prospective studies with large sample sizes are needed to further verify these conclusions.

ACK N OWLED G M ENTS
We would like to thank all patients and doctors who participated in this study for their cooperation.This study was sponsored by the National Natural Science Foundation of China (82201432), the Natural Science Foundation of Tianjin City (19JCYBJC27500), and Tianjin Key Medical Discipline (Specialty) Construction Project.

CO N FLI C T O F I NTE R E S T S TATE M E NT
JD, DL, XZ, XL, YD, XT, ZL, XY, MZ, and RM report no conflicts of interest.

TA B L E 5
The association of dehydration related indices with the severity and the prognosis of CVST.
High levels of USG, especially USG >1.013, were more common in patients with CVST, which might indicate dehydration in CVST, than in patients with CVST mimics; however, the conclusion was not convincing because other dehydration-related indices were inconsistent.With the development of the CVST phase, USG levels gradually decreased to levels close to those in patients with CVST mimics.No evidence revealed that dehydration-related indices could differentiate CVST from its mimics and healthy populations or were associated with CVST severity and prognosis.
Lin et al. demonstrated that USG-based hydration could prevent END in patients with AIS. 18Therefore, we consider alleviat-Differences of dehydration-related indices among patients with CVST, patients with CVST mimics, and health subjects after propensity score matching.Note: p < 0.002 indicates significance. 1CVST vs CVSS; 2 CVST vs Health people; 3 CVSS vs Health people.p-value <0.003 indicates statistical significance.
ing dehydration based on USG levels as a novel therapeutic strategy for CVST, which should be verified in the future.Furthermore, the concentration of UA was lower in patients with CVST than in those with CVST mimics and healthy populations.Urea levels reflect renal function and nutritional status.Patients with CVST suffer from high TA B L E 3 Abbreviations: Cr, creatinine; UA, uric acid; USG, Urine specific gravity.