Customize your JAMA Network experience by selecting one or more topics from the list below.
Zuurbier SM, Arnold M, Middeldorp S, et al. Risk of Cerebral Venous Thrombosis in Obese Women. JAMA Neurol. 2016;73(5):579–584. doi:10.1001/jamaneurol.2016.0001
Obesity is a risk factor for deep vein thrombosis of the leg and pulmonary embolism. To date, however, whether obesity is associated with adult cerebral venous thrombosis (CVT) has not been assessed.
To assess whether obesity is a risk factor for CVT.
Design, Setting, and Participants
A case-control study was performed in consecutive adult patients with CVT admitted from July 1, 2006 (Amsterdam), and October 1, 2009 (Berne), through December 31, 2014, to the Academic Medical Center in Amsterdam, the Netherlands, or Inselspital University Hospital in Berne, Switzerland. The control group was composed of individuals from the control population of the Multiple Environmental and Genetic Assessment of Risk Factors for Venous Thrombosis study, which was a large Dutch case-control study performed from March 1, 1999, to September 31, 2004, and in which risk factors for deep vein thrombosis and pulmonary embolism were assessed. Data analysis was performed from January 2 to July 12, 2015.
Main Outcomes and Measures
Obesity was determined by body mass index (BMI). A BMI of 30 or greater was considered to indicate obesity, and a BMI of 25 to 29.99 was considered to indicate overweight. A multiple imputation procedure was used for missing data. We adjusted for sex, age, history of cancer, ethnicity, smoking status, and oral contraceptive use. Individuals with normal weight (BMI <25) were the reference category.
The study included 186 cases and 6134 controls. Cases were younger (median age, 40 vs 48 years), more often female (133 [71.5%] vs 3220 [52.5%]), more often used oral contraceptives (97 [72.9%] vs 758 [23.5%] of women), and more frequently had a history of cancer (17 [9.1%] vs 235 [3.8%]) compared with controls. Obesity (BMI ≥30) was associated with an increased risk of CVT (adjusted odds ratio [OR], 2.63; 95% CI, 1.53-4.54). Stratification by sex revealed a strong association between CVT and obesity in women (adjusted OR, 3.50; 95% CI, 2.00-6.14) but not in men (adjusted OR, 1.16; 95% CI, 0.25-5.30). Further stratification revealed that, in women who used oral contraceptives, overweight and obesity were associated with an increased risk of CVT in a dose-dependent manner (BMI 25.0-29.9: adjusted OR, 11.87; 95% CI, 5.94-23.74; BMI ≥30: adjusted OR, 29.26; 95% CI, 13.47-63.60). No association was found in women who did not use oral contraceptives.
Conclusions and Relevance
Obesity is a strong risk factor for CVT in women who use oral contraceptives.
Various studies1-6 have identified obesity as a risk factor for deep vein thrombosis of the leg and pulmonary embolism, collectively called venous thromboembolism (VTE). A body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared) of 30 or more increases the risk of VTE approximately 2-fold compared with a normal BMI (<25), and this risk increases more with higher BMIs. Individuals with a BMI greater than 40 have an approximately 3 times higher risk.
Cerebral venous thrombosis (CVT) is a rare thrombotic condition that mainly affects young adults and children.7 Risk factors for CVT partly overlap with those for VTE and include thrombophilia, cancer, and oral contraceptive use.8,9 Other risk factors, such as local infections and head trauma, are specific for CVT.10 Some diseases, such as acute lymphoblastic leukemia, are associated with CVT and VTE, but the association is stronger for CVT.11
To our knowledge, whether obesity is associated with adult CVT has not been assessed. One study12 found that obesity is more common in children with CVT (55%) compared with controls (32%), but the sample size was small (22 cases). In addition, this study12 used historical controls and did not adjust for confounding variables. Moreover, because the clinical presentation and risk factors of CVT are different for children and adults,13 this result cannot be generalized to adult patients. In the current study, we examined whether obesity is a risk factor for adult CVT.
Question: Is obesity a risk factor for cerebral venous thrombosis?
Findings: In this case-control study that included 186 cases and 6134 controls, obesity was associated with a statistically significant 3-fold increased risk of cerebral venous thrombosis. Stratification by sex revealed that in women who used oral contraceptives, the risk was increased 30-fold. In contrast, we found no association in men or women who did not use oral contraceptives.
Meaning: Obesity is a strong risk factor for cerebral venous thrombosis in women who use oral contraceptives.
We performed an unmatched case-control study. The controls were recruited from March 1, 1999, through September 31, 2004, and the cases were recruited from July 1, 2006 (Amsterdam), and October 1, 2009 (Berne), through December 31, 2014. Data analysis was performed from January 2 to July 12, 2015.
Cases were patients with CVT included in 2 prospective cohorts from the Academic Medical Center in Amsterdam, the Netherlands, and Inselspital University Hospital in Berne, Switzerland. In these hospitals, data on consecutive adult patients with CVT have been recorded since July 2006 (Amsterdam) and October 2009 (Berne). We included patients who were admitted until December 2014. The diagnosis of CVT was confirmed with computed tomography venography, magnetic resonance imaging and magnetic resonance venography, angiography, or autopsy in all patients in accordance with international guidelines.14
Controls were healthy individuals who participated in the Dutch Multiple Environmental and Genetic Assessment of Risk Factors for Venous Thrombosis (MEGA) study.15 The MEGA study is a case-control study performed in the Netherlands that included 4956 consecutive patients aged 18 to 70 years with a first deep vein thrombosis of the leg or pulmonary embolism from March 1, 1999, through September 31, 2004. Details of this study have been reported.15 Partners of patients and individuals identified by random digit dialing were asked to participate as controls. In total, 6297 controls (3297 partners and 3000 identified by random digit dialing) were included. These participants were between the ages of 18 and 70 years and had no history of venous thrombosis. We excluded women (cases and controls) who were pregnant or had given birth less than 12 weeks earlier.
The Medical Ethics Committee Academic Medical Center, Committee Medical Ethics Leiden University Medical Centre, and Cantonal Ethics Commission Bern approved the study. Written informed consent was obtained from all participants included in the MEGA study and the Berne cohort. For the patients with CVT included in the Amsterdam cohort, written informed consent was not required under Dutch law because only observational data were collected.
Baseline characteristics, risk factors for thrombosis, imaging findings, and clinical outcome were recorded using a standardized case report form for cases. Each control completed a detailed questionnaire on acquired risk factors for thrombosis, which included self-reported current height and weight. In accordance with the definitions of the World Health Organization, BMI was categorized as follows: normal weight, BMI less than 25; overweight, BMI of 25 to 29.99; and obesity, BMI of 30 or greater.
A multiple imputation procedure was used for missing data on height, weight, oral contraceptive use, smoking status, history of cancer, and ethnicity.16 In total, 5 data sets were imputed, and results were pooled according to Rubin’s rules. We applied multivariate logistic regression analysis to study the association between obesity and CVT. Individuals with a normal BMI (<25) were the reference category. In a separate analysis, we included BMI as a continuous variable in the model. Unadjusted and adjusted odds ratios (ORs) with 95% CIs were calculated. We adjusted for the following prespecified variables: sex, age (as a continuous variable), history of cancer, ethnicity, smoking status, and oral contraceptive use. On the basis of a previous report,4 we performed predefined subgroup analyses in which we studied the influence of sex and oral contraceptive use on the association between obesity and CVT. P < .05 was considered statistically significant. All data were analyzed with SPSS statistical software, version 20 (SPSS Inc).
There were 192 cases and 6297 controls within the specified period. We excluded 6 cases and 163 controls because of pregnancy or recent delivery; therefore, the study population consisted of 186 cases and 6134 controls. The numbers of patients for whom data were missing were as follows: height, 547 (64 cases and 483 controls); weight, 539 (57 cases and 482 controls); oral contraceptive use, 32 (6 cases and 26 controls); history of cancer, 20 (1 case and 19 controls); smoking status, 487 (57 cases and 430 controls); and ethnicity, 423 (38 cases and 385 controls). In total, BMI could not be calculated without imputation in 69 cases because height or weight was not available. Eleven of these cases (15.9%) had obesity listed in their medical history.
Cases were younger (median age, 40 vs 48 years), more often female (133 [71.5%] vs 3220 [52.5%]), more often used oral contraceptives (97 [72.9%] vs 758 [23.5%] women), and more frequently had a history of cancer (17 [9.1%] vs 235 [3.8%]) compared with controls (Table 1). The most common clinical manifestations of patients with CVT are provided in Table 2.
Mean BMI was higher in cases than controls (26.7 vs 25.6, P = .01). After adjustment for confounding variables, the risk of CVT was increased in patients with obesity (BMI ≥30; adjusted OR, 2.63; 95% CI, 1.53-4.54) compared with patients with a normal BMI (Table 3). Overweight (BMI 25.0-29.99) was not associated with CVT (adjusted OR, 1.37; 95% CI, 0.80-2.36). When included as a continuous variable, BMI was also associated with an increased risk of CVT (adjusted OR per 1-unit increase in BMI, 1.09; 95% CI, 1.05-1.13; P < .001).
Stratification by sex revealed no statistically significant association between obesity and the risk of CVT in men (adjusted OR, 1.16; 95% CI, 0.25-5.30; Table 4). In women, overweight and obesity were associated with CVT (BMI 25-29.9: adjusted OR, 1.71; 95% CI, 1.01-2.91; BMI ≥30: adjusted OR, 3.50; 95% CI, 2.00-6.14).
We also stratified for oral contraceptive use (Table 5). Among women who did not use oral contraceptives, we found no association between obesity and CVT (BMI ≥30: adjusted OR, 1.29; 95% CI, 0.46-3.66). In contrast, in women who used oral contraceptives, obesity was strongly associated with the risk of CVT. Compared with women with a normal weight who did not use oral contraceptives, obese women taking oral contraceptives had a 29-fold increased risk of CVT (adjusted OR, 29.26; 95% CI, 13.47-63.60). The risk of CVT was also increased in overweight women who used oral contraceptives (BMI 25-29.99: adjusted OR, 11.87; 95% CI, 5.94-23.74).
Our study indicates that obesity (BMI ≥30) is associated with an increased risk of CVT. This association appears to be fully attributable to a strongly increased risk in women who use oral contraceptives. Among these women, obesity was associated with an almost 30-fold increased risk of CVT compared with women of normal weight who did not use oral contraceptives. In men and women who do not use oral contraceptives, we found no association between CVT and obesity.
In VTE, the association with obesity is also stronger in women than men. In a large cohort study,5 a BMI in the highest quartile increased the risk of VTE 2.8-fold for women and 1.7-fold for men. Another study3 confirmed similar effect sizes. Only a few studies4,17,18 have examined the interaction between oral contraceptive use and obesity on the risk of VTE. Pomp et al4 found an OR of 24 among women with a BMI of 30 or greater who used oral contraceptives compared with nonusers of oral contraceptives who were of normal weight. In another study,17 the risk of VTE in oral contraceptive users with a BMI of 30 or greater was increased approximately 10-fold. A possible interaction between oral contraceptives and obesity has also been observed in ischemic stroke. In a case-control study by Kemmeren et al,18 the risk of ischemic stroke in women who used oral contraceptives was approximately doubled in those with obesity.
Among oral contraceptive users, the risk of CVT was higher in those with obesity (BMI ≥30) than in women who were overweight (BMI 25-29.99). This dose-response effect, in combination with the magnitude of the effect size and the evidence that obesity is also associated with VTE,2,4,19 favors a causal association among obesity, oral contraceptive use, and CVT.20 One mechanism by which obesity could increase the risk of thrombosis are changes in coagulation factor levels. Compared with women with a normal weight, obese women have higher plasma concentrations of prothrombotic factors, such as plasminogen activator inhibitor 1 and von Willebrand factor.21 Obesity is also associated with increased activated protein C resistance and higher concentrations of factor VIII, which are risk factors for thrombosis.22 Use of oral contraceptives also leads to increased activated protein C resistance,23 which might explain the synergistic effect of both risk factors that we observed.
It is important to determine whether the control group we used is representative of the healthy population. In our study, the prevalence of a BMI of 30 or greater was 12.6% and 13.8% among male and female controls, respectively. These percentages are similar to the findings of a study24 among healthy adults from the Netherlands that was performed in the same period as the MEGA study. The prevalence of a BMI above 25 among men and women is also comparable to the results of that study.
Of interest, obesity is also a risk factor for idiopathic intracranial hypertension (IIH). Patients with IIH most often present with headache, papilledema, and decreased visual acuity. Identical symptoms occur in intracranial hypertension owing to impaired venous return in patients with chronic CVT.25 Before the introduction of computed tomography and magnetic resonance venography, it was not uncommon for patients with CVT to be misdiagnosed as having IIH.26,27 Like CVT, IIH is far more common in women than in men, and the association between IIH and obesity is also stronger in women than men.28-30 Given these resemblances, it is intriguing to speculate on a possible common pathogenesis by which obesity increases the risk of both conditions in women. It has been hypothesized that IIH is caused by decreased outflow from the cerebral venous system, possibly owing to stenosis of the transverse sinuses or insufficiency of the valves in the jugular veins.31,32 Studies31,33 have suggested that obesity enhances this mechanism by transmittance of increased intra-abdominal pressure to the cerebral venous system. On the other hand, there are also data indicating that the sinus stenosis seen in IIH is secondary to the increased intracranial pressure and is reversible after removal of cerebrospinal fluid.34 The effect of stenting of the cerebral sinuses has been examined in patients with IIH but only in small and uncontrolled series, and the data are inconclusive.35,36 Prospective studies37,38 evaluating the efficacy of cerebral venous sinus stenting in IIH are currently ongoing.
Our study has several limitations. First, we could include only a relatively small number of patients with CVT, especially men and women without oral contraceptive use. We cannot exclude that the absent association between obesity and CVT in these groups is the result of this limited sample size. Second, we were unable to examine whether genetic thrombophilia influences the association between obesity and CVT because screening for thrombophilia was not performed in most cases. Thrombophilia should be the focus of a future study given that there is a synergistic effect between oral contraceptives and thrombophilia on the risk of CVT9,39 and between obesity and thrombophilia on the risk of VTE.4 Third, data on the key variables of height and weight were missing in a substantial proportion of cases because these measurements were not recorded in all patients. This lack of data on height and weight could have underpowered the study and biased the results if the reason the data were missing was not random. To minimize this potential bias, we used a multiple imputation procedure.16,40 The fact that 15.9% of cases in which no BMI could be calculated had obesity listed in their medical history also suggests that obesity was not infrequent among cases with missing BMI. Fourth, instead of measuring height and weight, we used self-reported data to calculate BMI, which could be less accurate, although a previous study6 found an excellent correlation between self-reported and measured BMI. Moreover, because measurements were self-reported for cases and controls, any inaccuracy would be expected to affect both groups, which would decrease the risk of bias. Fifth, there was a time difference in recruitment between cases and controls. The controls were recruited from March 1, 1999, through September 31, 2004, whereas the cases were recruited from July 1, 2006 (Amsterdam), and October 1, 2009 (Berne), through December 31, 2014. A change in prevalence of obesity in that period could theoretically have biased the results. However, the time difference seems to be too small to explain the results and would also not explain why we only observed an association in women who used oral contraceptives. Finally, we cannot exclude the possibility that residual confounding influenced the results. Some variables, such as obstructive sleep apnea syndrome, were not available for controls. Obstructive sleep apnea syndrome is a risk factor for obesity and VTE, although, to our knowledge, an association with CVT has never been reported.41,42
The increased risk of VTE and CVT associated with oral contraceptives in the presence of obesity might make physicians reluctant to prescribe oral contraceptives to obese women. However, although the relative risks are increased substantially, the absolute risks of CVT are still small.43 Moreover, withholding oral contraceptives may lead to an increase in unintended pregnancies and thus the number of pregnancy-related thrombosis cases.44 Nevertheless, obese women should be informed about the increased risk of thrombosis if they use oral contraceptives, especially if other risk factors are present. Alternative methods of contraception that are not associated with thrombosis, such as an intrauterine device, might be offered to these women.
To our knowledge, this is the first case-control study that examined the association between obesity and CVT. Our results suggest that obesity is associated with a substantially increased risk of CVT in women who use oral contraceptives. This increased risk should be taken into consideration when prescribing oral contraceptives to obese women.
Accepted for Publication: December 21, 2015.
Corresponding Author: Jonathan M. Coutinho, MD, PhD, Department of Neurology, Academic Medical Centre, Meibergdreef 9, 1105 AZ, PO Box 22660, Amsterdam, the Netherlands (firstname.lastname@example.org).
Published Online: March 14, 2016. doi:10.1001/jamaneurol.2016.0001.
Author Contributions: Drs Arnold and Zuurbier contributed equally to this work. Dr Coutinho had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Zuurbier, Arnold, Middeldorp, Stam, Cannegieter, Coutinho.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Zuurbier, Coutinho.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Zuurbier, Silvis, Nemeth, Cannegieter, Coutinho.
Obtained funding: Arnold.
Administrative, technical, or material support: Broeg-Morvay, Silvis, Heldner, Meisterernst, Meulendijks, Cannegieter.
Study supervision: Arnold, Middeldorp, Stam, Coutinho.
Conflict of Interest Disclosures: Dr Arnold reported received honoraria for lectures and advisory boards from Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Pfizer, and Covidien. No other disclosures were reported.
Funding/Support: This study is supported by grant 021.001.045 from The Netherlands Organisation for Scientific Research, grant 2012-2 from the Dutch Thrombosis Society, and the Remmert Adriaan Laan Foundation (Dr Coutinho) and the Swiss Heart Foundation (Dr Arnold).
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.