Pregnancy and Spontaneous Cervical Artery Dissection:
A Propensity-matched Retrospective Cohort StudyThis section is compiled by Frank M. Painter, D.C.
Send all comments or additions to: Frankp@chiro.org
FROM: J Stroke Cerebrovasc Dis. 2023 (Nov); 32 (11): 107384 ~ FULL TEXT
OPEN ACCESS Robert J. Trager, DC • Clinton J. Daniels, DC, MS • Zachary E. Scott, DC • Jaime A. Perez, PhD
Connor Whole Health,
University Hospitals Cleveland Medical Center,
Cleveland OH, USA.
Objectives: We hypothesized that pregnant women would have an increased risk of spontaneous cervical artery dissection (sCeAD) affecting the carotid or vertebral arteries over one-year follow-up after the first trimester ultrasound compared to matched non-pregnant controls.
Materials and methods: We queried a United States research network (TriNetX, Inc.) of de-identified medical records of >111 million patients, with data spanning 2008-2023. We included women aged ?18 and excluded those with trauma and conditions potentially causative of sCeAD. Women were divided into cohorts based on a1 first trimester ultrasound and subsequent labor, delivery, or full-term pregnancy, or2 gynecological examination and no pregnancy. We used propensity matching to control for variables associated with sCeAD and calculated the risk ratio (RR) of sCeAD occurring over one-year follow-up from the index date of ultrasound or gynecological exam.
Results: After matching, the incidence rate of sCeAD in the pregnancy cohort was 8.0 (95% CI: 8.0-8.1) per 100,000 person-years, compared to 3.9 (95% CI: 3.9-3.9) per 100,000 person-years in the non-pregnancy cohort, yielding an RR (95% CI) of 2.06 (1.17-3.61; P= .0104). A cumulative incidence graph suggested that most cases of sCeAD in the pregnancy cohort occurred during pregnancy rather than the postpartum period.
Conclusions: Our findings demonstrate that women have a twofold increased risk of sCeAD during pregnancy and the postpartum period compared to non-pregnant women. Further research is needed to determine whether maternal comorbidities such as preeclampsia account for these findings, and clarify when sCeAD occurs in relation to pregnancy or the postpartum period.
Keywords: Internal carotid artery dissection; Parturition; Pregnancy; Vertebral artery dissection.
From the FULL TEXT Article:
Introduction
Cervical artery dissection (sCeAD) is a condition involving a tear in the inner lining of one of the paired carotid or vertebral arteries that supply blood to the brain, and is termed spontaneous cervical artery dissection (sCeAD) in the absence of preceding trauma. [1] This condition often first manifests as neck pain, headache, or dizziness, yet may progress to stroke within hours or days. [1] According to epidemiologic data spanning 2002 to 2020, sCeAD has an incidence rate of five per 100,000 person-years, [2] and this condition accounts for about a quarter of strokes in individuals under age 50. [3] While the etiology of sCeAD is not fully understood, it is thought to involve both genetic and environmental factors. [4, 5] Pregnancy has been reported as a risk factor for sCeAD, yet the evidence for a firm association is limited. [6]
In general, pregnancy may be a risk factor for multiple types of arterial dissection. One retrospective study, which analyzed 993 cases of arterial dissection occurring during pregnancy or within six weeks post-delivery, found that dissections most often affected a coronary artery (38%), followed by vertebral artery (23%), aortic artery (20%), or carotid artery (20%). [6]
Currently, we are aware of only a single epidemiologic study that examined the association between pregnancy and sCeAD, [7] which included 826 pregnant women with sCeAD and an equal number of matched controls with renal colic. This study demonstrated that pregnancy was associated with an increased odds of sCeAD by a factor of at least two. [7] In addition, a recent systematic review identified 77 pregnant women who developed sCeAD from published case reports and series. [8]
Several pregnancy-related physiological changes are suspected of increasing the risk of sCeAD. A progressive increase in blood volume, cardiac output, and heart rate from the first trimester to labor may stress vascular system. [6, 9] In addition, a rise in the hormones estrogen and relaxin, which increase vascular compliance, may make blood vessels more susceptible to dissection. [6, 10]
A generalized prothrombotic state, [9] positional effects on circulation (e.g., supine versus left lateral recumbent), straining during labor, [6] and postpartum return of blood from the uterus to the general circulation [10] are other potentially relevant factors. Some evidence suggests that maternal comorbidities including gestational diabetes, pre-eclampsia/eclampsia, and multiple gestation may increase risk of pregnancy-related arterial dissection, however a specific association between these comorbidities and sCeAD remains unconfirmed. [6]
Considering the growing evidence of an association between pregnancy and sCeAD, we aimed to further explore this association using a retrospective cohort design. We hypothesized that women with pregnancy confirmed via first trimester ultrasound would have an increased risk of sCeAD over a one-year follow-up compared to non-pregnant women after controlling for confounding variables.
Materials and methods
Study design
Figure 1 This study used a retrospective cohort design with an active comparator and new user features to reduce bias (Figure 1). This design was preferred over a case-control design as pregnant women could be identified at the first instance of a first trimester ultrasound, allowing for calculation of the incidence rate of sCeAD during pregnancy and the postpartum period. The data range spanned from 15 years prior to the query date of September 10, 2023 (i.e., September 10, 2008, to September 10, 2023).
Patients were included up to one year prior to the query date to allow for sufficient time to identify sCeAD during follow-up. Findings are reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline. [11] This study was deemed Not Human Subjects research by the University Hospitals Institutional Review Board (Cleveland, Ohio, USA, STUDY20230450). The methods used in this study were registered a priori on June 20, 2023. [12]
Setting and data source
We used the United States TriNetX network (TriNetX, Inc, Cambridge, MA, USA) to conduct this study, which includes routinely-collected, de-identified, aggregated data from over 111 million patients attending 78 healthcare institutions. TriNetX integrates electronic medical records and medical and pharmaceutical claims. [13] Data include both inpatient and outpatient care for insured and uninsured patients. The dataset viewing platform may be searched using standardized nomenclature such as the International Classification of Diseases, 10th Edition (ICD-10) codes, which are automatically converted to ICD-9 codes in queries of older records. [13]
TriNetX offers natural language processing which operates using machine learning technology (Averbis, Freiburg im Breisgau, DE) to search free text from clinical charts and other records (e.g., radiology reports). This feature can identify the presence of a condition and also recognizes negation (e.g., absence of a condition) and context. We implemented natural language processing considering that ICD-10 codes alone may be insufficient to capture all cases of sCeAD. [14]
Participants
We included women aged 18 and older, considering cases of sCeAD reported in relation to pregnancy have only been reported in adults. [8] Pregnant women were considered those who had a first trimester ultrasound (Supplemental Table 1) and had any subsequent code indicating a pregnancy of at least 38 weeks of gestation, labor, delivery, or postpartum care occurring within 10 months (i.e., ? 304 days) after the initial first trimester ultrasound (Supplemental Table 2).
First trimester ultrasound is often performed to confirm the presence of an intrauterine pregnancy, cardiac activity, and estimate gestational age, and is conducted between 11 and 14 weeks of gestation. [15] Women who had a positive pregnancy test but did not undergo a first trimester ultrasound were not included.
Non-pregnant women were considered to be those with a gynecological examination (Supplemental Table 3) and without a pregnancy over the year preceding and including the examination date and follow-up window. To minimize loss to follow-up and improve data completeness, we required that patients in both cohorts had at least one healthcare visit within the year following the index date of ultrasound or gynecological examination.
We excluded patients with a recorded history of cerebrovascular disease (e.g., stroke, hemorrhage, or occlusion of the cervical or cerebral arteries), sCeAD, and related conditions and procedures (Supplemental Table 4). We also excluded women with conditions that increase the risk of sCeAD (i.e., Alport syndrome, arterial tortuosity syndrome, Eagle syndrome, Ehlers-Danlos syndrome, fibromuscular dysplasia, Loeys-Dietz syndrome, Marfan syndrome, osteogenesis imperfecta), and trauma to the head or neck, and other types of trauma (e.g., motor vehicle accidents, falls). [4, 6]
These steps were intended to help ensure that observed differences in sCeAD could be attributed to pregnancy rather than other conditions. We excluded women who had a spontaneous or medically induced abortion to help standardize the duration of pregnancy in the pregnancy cohort, and further exclude pregnant women from the non-pregnancy cohort. Women in the non-pregnancy cohort were excluded who were pregnant or had additional markers suggestive of pregnancy such as oxytocin, Rho(D) immune globulin, or ultrasound for pregnancy.
Variables
Propensity matching was used in this study to minimize the impact of confounding variables and potential selection bias. Data available within two years preceding and including the index date of ultrasound or gynecological examination were assessed for matching. Key variables with a known positive or negative association with sCeAD were matched: adverse socioeconomic factors (i.e., problems related to education, employment, and housing/income; negative), [16, 17] body height (positive), [18] demographics (age, race, ethnicity; positive or negative) [4, 7], head/neck imaging (positive) [19], hypertensive diseases (positive) and related medications [5], and migraines (positive) [20] (Supplemental Table 5).
Considering that a recent systematic review found no significant association between sCeAD and diabetes, hyperlipidemia, and smoking, we did not match for these variables. [5] We did not match for body weight or body mass index [18] considering (1) weight fluctuates during and after pregnancy and (2) it would be unclear if weight measures were recent.
As a primary outcome, we ascertained occurrences of carotid artery dissection and vertebral artery dissection via ICD-10 diagnosis codes for these conditions (I77.71 and I77.74, respectively). Ascertainment of sCeAD was bolstered by natural language processing of free text describing these conditions. Considering these conditions are rare and have overlapping risk factors [4], we pooled them as a single outcome. We examined the occurrence of sCeAD over a follow-up window of one year after the first trimester ultrasound or gynecological examination.
Statistical methods
Statistical analysis was conducted using the TriNetX dataset viewing software. The propensity score was calculated for each subject based on their covariates using logistic regression. Using a greedy nearest-neighbor matching algorithm of 1:1 and caliper width of 0.1 pooled standard deviations, each subject from the pregnancy cohort was matched to a subject from the non-pregnancy cohort with a similar propensity score.
Baseline characteristics were compared using either a Pearson chi-squared test for categorical variables or independent samples t-test for continuous variables. Covariate balance was assessed using standardized mean difference (SMD), and absolute values of > 0.1 were considered to represent residual imbalance.
We also compared measures of data density and completeness. To calculate risk ratios of sCeAD, we divided the incidence of sCeAD in the pregnant cohort by the incidence in the non-pregnancy cohort. Statistical significance was assessed at P < 0.05. We conducted a sensitivity analysis to examine time-to-event data to provide greater insight into the timing of sCeAD throughout follow-up.
Based on peer reviewer feedback, we conducted post-hoc sensitivity analyses to examine the incidence of subtypes of sCeAD, maternal complications such as pre-eclampsia (ICD-10: O11, O13, O14), eclampsia (ICD-10:O15), diabetes mellitus arising in pregnancy, childbirth, or the puerperium (ICD-10: O24), and posterior reversible encephalopathy syndrome (ICD-10: I67.83), a potential sequela of pregnancy-related sCeAD. [21]
Also based on peer review and prior literature suggesting maternal socio-demographics and comorbidities may influence the likelihood of receiving a first trimester ultrasound [22], we tested whether requiring a first trimester ultrasound created a non-representative sample of pregnant women compared to a more available human choriogonadotropin (hCG) pregnancy test. For this analysis, we created an additional comparison cohort of women who had a positive hCG test and labor or delivery within the subsequent 12 months, with all other selection criteria remaining identical to our primary query. We examined between-group differences using SMD.
Using R (version 4.2.2, Vienna, AT [23]) we calculated 95% confidence intervals for incidence rate of sCeAD and proportions of maternal complications, and plotted created cumulative incidence, propensity score, and SMD using the ggplot2 package. [24]
We estimated a total required sample size of 129,826 using GPower (Version 3.1.9.6, Kiel University, DE) using a z-test to discern a 0.0006% between-cohort difference in incidence proportion (0.0012% vs. 0.0006%7) using an α-error of 0.05, allocation ratio of one, two tails, and power of 0.95.
Results
Participants
Our query identified pregnant women from 51 healthcare organizations and non-pregnant women from 76 healthcare organizations. Before matching, there were 462,123 patients in the pregnancy cohort and 2,339,587 patients in the non-pregnancy cohort. After matching, there were 460,187 patients in each cohort (mean age 29±6 years).
Table 1 Before propensity matching, the pregnancy cohort had a significantly lower mean age, had a lower incidence of pre-existing hypertensive disorders, and had a greater percentage of women that identified as Hispanic or Latino and White and those who had health hazards related to adverse socioeconomic and psychosocial circumstances (SMD >0.1; Table 1). After matching, there were no significant between-cohort differences with respect to any matched covariates (SMD <0.1).
Descriptive data
The mean number of data points per patient was high in both cohorts (pregnancy: 1,582; non-pregnancy: 1,469). After propensity matching, there were no significant between-cohort differences in the frequency of unknown demographics with respect to unknown race (pregnancy: 20%, non-pregnancy: 19%; SMD = 0.005), unknown ethnicity (pregnancy: 21%, non-pregnancy: 20%; SMD = 0.014), and unknown age (0% per cohort; SMD = 0). A propensity score density graph showed that the propensity scores were well-matched (Supplemental Figure 1). Together, these findings suggested that there were no meaningful differences between cohorts with respect to data completeness or residual covariate imbalance.
Key results
Table 2 The likelihood of sCeAD over one year following the index date of first trimester ultrasound or wellness visit/gynecological exam was significantly greater in the pregnancy cohort compared to the non-pregnancy cohort both before and after propensity matching (Table 2). After propensity matching, the incidence rate of sCeAD in the pregnancy cohort was 8.0 (95% CI: 8.0-8.1) per 100,000 person-years, compared 3.9 (95% CI: 3.9-3.9) per 100,000 person-years in the non-pregnancy cohort, yielding an RR (95% CI) of 2.06 (1.17-3.61; P= .0104).
Sensitivity analysis
Figure 2 A cumulative incidence graph revealed that the incidence of sCeAD increased sharply in a curvilinear fashion after the index date in the pregnancy cohort, compared to the non-pregnancy cohort, which demonstrated a linear increase (Figure 2). While there was overlap between the 95% confidence intervals between the cumulative incidence curves of each cohort, there was no obvious convergence throughout follow-up, suggesting that the difference in incidence was maintained over the outcome assessment window. While 50% of sCeAD cases occurred relatively early (by 89 days; or 3 months) in the pregnancy cohort, 50% of cases occurred closer to the middle of the outcome assessment window in the non-pregnancy cohort (by 146 days; or 5 months).
Further analysis revealed that 75% of sCeAD cases in the pregnancy cohort occurred by 181 days (6 months), a period that would correspond with pregnancy and typically precede delivery and the postpartum period. For reference, we expect that delivery would be expected about 26 to 29 weeks’ follow-up (i.e., 40 weeks’ typical duration of pregnancy minus 11 or 14 weeks to account for the first trimester ultrasound), or otherwise stated, most deliveries should occur from 182 to 203 days of follow-up (Fig. 2). In the non-pregnancy cohort, 75% of cases had occurred by 208 days (7 months).
A sensitivity analysis of individual subtypes of sCeAD (vertebral versus carotid artery dissection) occurring in the pregnancy cohort, after propensity matching, revealed that both conditions had a curvilinear cumulative incidence (Supplemental Fig. 2). While vertebral artery dissection had a greater incidence rate of 5.9 (95% CI: 3.7-8.1) per 100,000 person-years compared to carotid dissection (2.8 [95% CI: 1.3-4.4] per 100,000 person-years), the 95% confidence intervals overlapped, suggesting this difference was not meaningful. In addition, these findings should be interpreted with caution due to the potential for patients with both subtypes of sCeAD to be counted twice.
After propensity matching, the incidence rate of posterior reversible encephalopathy syndrome per 100,000 person-years was greater in the pregnancy cohort compared to the non-pregnancy cohort (16.7 [95% CI: 16.7-16.7] vs. 3.3 [95% CI: 3.3-3.3]). This finding should be interpreted with caution considering our selection criteria and matched variables were not tailored to this outcome. The proportions of pregnant women affected by maternal complications after propensity matching over one-year follow-up were: pre-eclampsia 14.9% (95% CI: 14.8-15.0), eclampsia 0.5% (95% CI: 0.5-0.5), and maternal diabetes 9.7% (95% CI: 9.6-9.8).
When cross-sectionally comparing a subset of pregnant women having first trimester ultrasound to an additional group of women with pregnancy confirmed via hCG testing (n=130,513 per group), there were no meaningful between-group differences with respect to the prevalence of adverse socioeconomic factors; overweight or obesity; diseases of the circulatory system; endocrine, nutritional, and metabolic diseases; and proportion of women identifying as Asian, Hispanic or Latino, or not Hispanic or Latino (SMD<0.1 for each; Supplemental Figure 3). However, there were potentially meaningful differences in mean age (ultrasound: 29.5±5.7 years; hCG: 28.7±5.7 years; SMD=0.113), and proportion of individuals identifying as Black or African American (ultrasound: 17%; hCG=23%; SMD=0.155) and White (ultrasound: 58%; hCG: 53%; SMD=0.106).
Discussion
The current study examined the association between pregnancy and sCeAD using a large propensity matched US population, including over 920,000 total women with outcome assessment window spanning more than 920,000 person-years. Our results showed that the risk of sCeAD was doubled in women during pregnancy and the postpartum period compared to non-pregnant women. Our sensitivity analysis suggested that cases of sCeAD were distributed relatively early during follow-up, during a time window corresponding with pregnancy rather than the post-partum period. In the present study, the incidence rate of sCeAD in non-pregnant women was four to six per 100,000 person-years (pre- and post-matching, respectively). These rates are comparable to previous estimates of the incidence rate of sCeAD of five per 100,000 person-years [2], and a marker of the validity of our findings.
Our results corroborate those of a previous study which identified an increase in odds of sCeAD amongst pregnant women by a factor of at least two using case-control and case-crossover models. [7] Our study identified a similar increase in likelihood (i.e., RR =2.06), despite having a different study design. However, a key methodological difference is that the previous study noted that the heightened risk was limited to the peripartum period, with sCeAD occurring a mean of 21 days after delivery [7], while our study found that the risk of sCeAD may increase earlier during pregnancy. Our identified risk period may result from starting the outcome assessment window during the first trimester, whereas the previous study ascertained pregnancy-related sCeAD beginning three months prior to and following delivery. [7]
Also contrasting our findings of the timing of sCeAD was a prior review of case reports which found that the majority (87%) of cases of pregnancy-related sCeAD occurred in the postpartum period. [8] This difference could be explained by publication bias, in that case reports may represent more readily identifiable occurrences of sCeAD, as many were documented in relation to maternal complications such as eclampsia and preeclampsia. [8] It is also unclear if clinicians would have documented a case report highlighting sCeAD occurring during an uncomplicated pregnancy, given that an epidemiologic association between pregnancy and sCeAD was only recently described in 2020. [7]
Finally, we cannot rule out a heightened period of elevated sCeAD risk in the postpartum period in our results, as a more detailed analysis was hindered by the variable timing of delivery per our methods and de-identified nature of our dataset. Regardless, given the contrast between our results and previous studies regarding the timing of pregnancy-related sCeAD, further research is needed to clarify when pregnant women are at a heightened risk of this condition.
Although sCeAD remains relatively rare during pregnancy, with an absolute risk of only eight per 100,000, we suggest that our findings may be clinically relevant. The real-world risk of sCeAD may be even higher during and pregnancy and the postpartum period, considering our study focused on a highly selected population, excluding women with connective tissue disorders who may have a greater risk of sCeAD. [6, 8, 25, 26] Healthcare providers, such as obstetricians, may benefit from heightened awareness of the increase in risk of sCeAD during pregnancy and the postpartum period.
Future studies may examine whether specific maternal comorbidities account for the association between sCeAD and pregnancy, such as maternal hypertension, gestational diabetes, and pre-eclampsia. [6, 8] Additional research is needed to confirm whether comorbidities occurring prior to pregnancy, such as connective tissue disorders, increase the risk of sCeAD during pregnancy or the post-partum period. [6, 8, 10] Considering these several possible risk factors for sCeAD, a case-control design may be ideal for such a follow-up study. A better understanding of specific risk factors for sCeAD during pregnancy could make it possible to identify pregnant women who have a heightened risk of sCeAD and thus require closer monitoring.
Limitations
As an observational study, unmeasured confounding variables may influence results. Considering our reliance upon a large, de-identified dataset which included patients from outside healthcare organizations, we were unable to perform manual chart review to confirm sCeAD diagnosis and/or implement strict diagnostic or radiological criteria for each patient. Instead, we used a multifaceted strategy of ascertaining ICD-10 diagnosis codes for sCeAD, corroborated by natural language processing of free text in patients’ charts and imaging reports, ultimately relying on the information in the patients’ medical record. Therefore, it is possible that some included patients had a false positive or negative diagnosis of sCeAD. Considering we used a 15-year data range, and the incidence of sCeAD has increased over this timespan due to improved detection [2], it is also possible that any between-cohort differences in year of inclusion of patients would affect our findings.
Social determinants of health (i.e., education [17]) and genetic variants [26] are poorly represented in the TriNetX dataset. Beyond specific predisposing conditions, we could not account for a family history of sCeAD. While we excluded individuals with prior and/or concurrent head and/or neck trauma, and external causes of morbidity, it remains possible that included women had minor trauma which remained unreported or undocumented in the medical record preceding the index date or during follow-up.
We were unable to examine the mean year of enrollment due to the de-identified nature of the dataset. It is beyond the scope of our study to examine potential risk factors for sCeAD within pregnancy, such as gestational hypertension, pre-eclampsia/eclampsia, a prolonged second stage of labor, or multiple gestation. [6, 8] Considering these factors cannot be propensity matched to non-pregnant controls, a follow-up study, potentially using a case-control design, is needed to examine such factors further. In such a study, evaluated risk factors would include typical risk factors for sCeAD (e.g., hypertension) as well as maternal comorbidities and complications arising during delivery or the postpartum period.
Our aim was to begin the ascertainment window as early into pregnancy as possible using a reliable time point. Considering that most positive pregnancy tests do not result in full-term pregnancy and/or delivery, we used the first trimester ultrasound as the index event. While this had the downside of potentially missing a small percentage of cases of sCeAD occurring early in the first trimester (i.e., < 3%) [8], it had the tradeoff of providing a firmer timeline and case definition of pregnancy. Due to the variable timing of first trimester ultrasound, our pregnancy cohort likely included pregnant women having preterm birth (?10%). [27]
While this adds heterogeneity to the pregnancy cohort, we feel these patients were important to include considering that cardiovascular changes throughout pregnancy may be relevant to sCeAD risk. [6, 9, 10] In addition, our initial feasibility testing suggested that limiting the pregnancy cohort to full-term deliveries would impracticably limit sample size. The results of this study may not be generalizable to other countries, which may have a different incidence of sCeAD or risk factors for this disorder. Finally, our sensitivity analysis suggested that requiring a first trimester ultrasound produced a sample under-representing individuals identifying as Black/African American and over-representing individuals identifying as White compared to hCG pregnancy testing.
Conclusions
This large retrospective cohort study found that women have a nearly two-fold increase in risk of sCeAD during pregnancy and the peripartum period compared to non-pregnant controls. Further research is needed to examine whether specific maternal comorbidities account for the heightened risk of sCeAD during pregnancy. In addition, while our sensitivity analysis suggested that most cases of sCeAD occurred during pregnancy itself rather than the postpartum period, further research is needed to clarify the timing of pregnancy-related sCeAD.
Supplementary Material
Authors’ contributions
RT, CD, ZS, and JP conceived of and designed the study protocol and methodology.
RC and JP accessed the software and dataset, collected data, and analyzed and interpreted data.
RT drafted the initial manuscript, while all authors contributed to, critically revised, and approved of the final manuscript. RT was the guarantor of the study.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This publication was made possible through the support of the Clinical Research Center of University Hospitals Cleveland Medical Center (UHCMC), the Case Western Reserve University Clinical, Translational Science Collaborative (CTSC) 4UL1TR000439, and indirectly supported with resources and use of facilities from VA Puget Sound Health Care System. Publication contents are solely the responsibility of the authors and do not necessarily represent the official views of UHCMC, National Institutes of Health, Veterans Health Administration, or U.S. Government.
References:
Arning, C Hanke-Arning, K Eckert, B.
The Clinical Features of Dissection of the Cervical Brain-Supplying Arteries
Dtsch Arztebl Int. 2022 Sep 5;
(Forthcoming):arztebl.m2022.0238Griffin, KJ Harmsen, WS Brown, RD.
Abstract WP135: Trends In The Incidence Of Spontaneous Cervical Artery Dissection
Stroke. 2023 Feb; 54(Suppl_1)
AWP135–AWP135Putaala, J Metso, AJ Metso, TM.
Analysis of 1008 Consecutive Patients Aged 15 to 49 With First-Ever Ischemic Stroke
Stroke. 2009 Apr; 40(4):1195-1203Debette, S.
Pathophysiology and risk factors of cervical artery dissection: what have we learnt from large hospital-based cohorts?
Curr Opin Neurol. 2014 Feb; 27(1):20-28Abdelnour, LH Abdalla, ME Elhassan, S.
Diabetes, hypertension, smoking, and hyperlipidemia as risk factors for spontaneous cervical artery dissection: Meta-analysis of case-control studies
Current J Neurol. 2022 Jul 1; 21(3):183-193Beyer, SE Dicks, AB Shainker, SA.
Pregnancy-associated arterial dissections: a nationwide cohort study
Eur Heart J. 2020 Nov 21; 41(44):4234-4242Salehi Omran, S Parikh, NS Poisson, S.
Association Between Pregnancy and Cervical Artery Dissection
Ann Neurol. 2020 Sep; 88(3):596-602Abdelnour, LH Kurdy, M Idris, A.
Systematic review of postpartum and pregnancy-related cervical artery dissection
J Matern Fetal Neonatal Med. 2022 Dec; 35(26):10287-10295Tan, EK Tan, EL.
Alterations in physiology and anatomy during pregnancy
Best Pract Res Clin Obstetr Gynaecol. 2013 Dec 1; 27(6):791-802Sanghavi, M Rutherford, JD.
Cardiovascular Physiology of Pregnancy
Circulation. 2014 Sep 16; 130(12):1003-1008Von Elm, E Altman, DG Egger, M.
The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies
Ann Intern Med. 2007; 147(8):573-577Trager RJ, Daniels C, Scott ZE, Perez JA.
Pregnancy and spontaneous cervical artery dissection in the United States: a propensity-matched retrospective cohort study.
2023 Jun 20 [cited 2023 Jun 23]; Available from: https://osf.io/wm4zc.Topaloglu, U Palchuk, MB.
Using a federated network of real-world data to optimize clinical trials operations
JCO Clin Cancer Inform. 2018; 2:1-10Paulus, JK Thaler, DE.
Does case misclassification threaten the validity of studies investigating the relationship between neck manipulation and vertebral artery dissection stroke? Yes
Chiropr Man Therap. 2016 Nov 5; 24(1):42Mei, JY Afshar, Y Platt, LD.
First-Trimester Ultrasound
Obstetr Gynecol Clinics. 2019 Dec 1; 46(4):829-852Grau, AJ Brandt, T Buggle, F.
Association of Cervical Artery Dissection With Recent Infection
Arch Neurol. 1999 Jul 1; 56(7):851-856Kellert, L Grau, A Pezzini, A.
University education and cervical artery dissection
J Neurol. 2018 May; 265(5):1065-1070Arnold, M Pannier, B Chabriat, H.
Vascular risk factors and morphometric data in cervical artery dissection: a case-control study
J Neurol Neurosurg Psychiatry. 2009 Feb; 80(2):232-234Huang, YC Chen, YF Wang, YH.
Cervicocranial arterial dissection: experience of 73 patients in a single center
Surg Neurol. 2009 Dec 1; 72:S20-S27De Giuli, V Grassi, M Lodigiani, C.
Association Between Migraine and Cervical Artery Dissection: The Italian Project on Stroke in Young Adults
JAMA Neurol. 2017 May 1; 74(5):512-518Nishimura, M Hiraoka, E Kanazawa, K.
Postpartum vertebral artery dissection with posterior reversible encephalopathy syndrome
BMJ Case Rep. 2015 Jul 6; 2015, bcr2014207332Abdullah, P Landy, CK McCague, H.
Factors associated with the timing of the first prenatal ultrasound in Canada
BMC Pregnancy Childbirth. 2019 May 10; 19:164R Core Team
R: A Language and Environment for Statistical Computing [Internet]
R Foundation for Statistical Computing, Vienna, Austria, 2022
Available from https://www.R-project.org/Wickham, H.
ggplot2: Elegant Graphics for Data Analysis [Internet]
Springer-Verlag, New York, 2016
Available from https://ggplot2.tidyverse.orgDebette, S Goeggel Simonetti, B Schilling, S.
Familial occurrence and heritable connective tissue disorders in cervical artery dissection
Neurology. 2014 Nov 25; 83(22):2023-2031Traenka, C Kloss, M Strom, T.
Rare genetic variants in patients with cervical artery dissection
Eur Stroke J. 2019 Dec; 4(4):355-362Purisch, SE Gyamfi-Bannerman, C.
Epidemiology of preterm birth
Semin Perinatol. 2017 Nov 1; 41(7):387-391Schneeweiss, S A Rassen, J S Brown, J.
Graphical Depiction of Longitudinal Study Designs in Health Care Databases
Ann Intern Med. 2019 Mar 12; 170(6):398-406[an error occurred while processing this directive] 
Return to STROKE
Since 9-18-2026
