Anxiety and the Risk of Stroke
The Rotterdam Study
Background and Purpose—It is unclear whether anxiety is a risk factor for stroke. We assessed the association between anxiety and the risk of incident stroke.
Methods—This population-based cohort study was based on 2 rounds of the Rotterdam Study. Each round was taken separately as baseline. In 1993 to 1995, anxiety symptoms were measured using the Hospital Anxiety and Depression Scale-Anxiety (HADS-A). In 2002 to 2004, anxiety disorders were assessed using the Munich version of the Composite International Diagnostic Interview. Participants were followed up for incident stroke until January 2012.
Results—In the sample undergoing HADS-A (N=2625; mean age at baseline, 68.4 years), 332 strokes occurred during 32 720 years of follow-up. HADS-A score was not associated with the risk of stroke during complete follow-up (adjusted hazard ratio, 1.02; 95% confidence interval, 0.74–1.43; for HADS-A≥8 compared with HADS-A <8), although we did find an increased risk after a shorter follow-up of 3 years (adjusted hazard ratio, 2.68; 95% confidence interval, 1.33–5.41). In the sample undergoing the Munich version of the Composite International Diagnostic Interview (N=8662; mean age at baseline, 66.1 years), 340 strokes occurred during 48 703 years of follow-up. Participants with any anxiety disorder had no higher risk of stroke than participants without anxiety disorder (adjusted hazard ratio, 0.95; 95% confidence interval, 0.64–1.43). We also did not observe an increased risk of stroke for the different subtypes of anxiety.
Conclusions—Anxiety disorders were not associated with stroke in our general population study. Anxiety symptoms were only related to stroke in the short term, which needs further exploration.
An increasing body of evidence suggests that anxiety is associated with the risk of coronary heart disease.1,2 Literature on the association between anxiety and stroke is scarce and inconsistent. Although 1 study found no association between generalized anxiety disorder and stroke,3 another study showed that the risk of stroke increased with increasing anxiety symptom score.4 Therefore, these findings require further corroboration. We assessed the association between anxiety and the risk of stroke in the general population.
Details on Methods are available in the online-only Data Supplement.
The Rotterdam Study is a prospective population-based cohort study.5 The study consists of 14 926 participants aged ≥45 years.5 Anxiety symptoms were assessed in 1993 to 1995 in 3060 participants. Of these, 2625 participants were eligible for the analysis. Anxiety disorders were assessed from 2002 to 2004 in 9974 participants. Of these, 8662 participants were eligible for the analysis.
The Rotterdam Study has been approved by the Medical Ethics Committee of the Erasmus MC and by the Ministry of Health, Welfare and Sport of the Netherlands, implementing the Population Studies Act: Rotterdam Study. Written informed consent was obtained from all participants.
Assessment of Anxiety
We used the Hospital Anxiety and Depression Scale-Anxiety (HADS-A) to measure anxiety symptoms.6
The 1-year prevalence of anxiety disorders was assessed using a slightly adapted version of the Munich version of the Composite International Diagnostic Interview.7 This test assesses the following anxiety disorders according to the Diagnostic and Statistical Manual of Mental Disorders (4th edition, text revision) (DSM-IV-TR) criteria: generalized anxiety disorder, panic disorder, agoraphobia, social phobia, and specific phobia.7
Assessment of Stroke
Participants were continuously monitored for incident stroke, which were identified from medical records and confirmed by an experienced vascular neurologist.8
We calculated hazard ratios and 95% confidence intervals with Cox proportional hazards models. We analyzed participants based on HADS-A score (categorically using 2 cut-offs: ≥8 versus <8 and ≥11 versus <11, and continuously per SD) and on Munich version of the Composite International Diagnostic Interview (at least 1 disorder [generalized anxiety disorder, panic disorder, agoraphobia, social phobia, or specific phobia] versus no disorder, and the separate subtypes of anxiety disorder versus no disorder).
Baseline characteristics of the study samples are presented in Table 1. Average (±SD) follow-up in the sample undergoing HADS-A was 12.5 (±5.3) years during which 332 strokes occurred: 204 ischemic and 30 hemorrhagic. Average follow-up in the sample undergoing Munich version of the Composite International Diagnostic Interview was 5.6 (±2.1) years during which 340 strokes occurred: 237 ischemic and 40 hemorrhagic.
Neither anxiety symptoms nor anxiety disorders were associated with the risk of stroke after complete follow-up (Table 2; Table I in the online-only Data Supplement): the multivariate adjusted hazard ratio was 1.02 (95% confidence interval, 0.74–1.43) for anxiety symptoms (HADS-A≥8 compared with HADS-A<8) and 0.95 (0.64–1.43) for any versus no anxiety disorder. The stricter dichotomization of HADS-A (≥11 versus <11) resulted in stronger effects although still not significant. We also did not observe any association for the separate subtypes of anxiety disorder or when we examined anxiety symptoms on a continuous scale (Table II in the online-only Data Supplement; Table 2).
Inspection of the survival curve (Figure I in the online-only Data Supplement) pointed toward a possible short-term effect for anxiety symptoms. Indeed, an increased HADS was associated with an increased risk of stroke in the first 3 years of follow-up (hazard ratio, 2.55 [1.45–4.46] for HADS-A≥8 compared with HADS-A<8; Table III in the online-only Data Supplement). After additional adjustment for depressive symptoms, the association remained similar.
In this prospective population-based cohort study, we found no long-term association between anxiety and the risk of stroke.
Two previous cohort studies assessed the association between anxiety and stroke: in one study, the risk of stroke increased with increasing anxiety symptom score,4 whereas in the other study, psychological distress was associated with stroke but not with generalized anxiety disorder.3 A reason for the differences may be the different scores to assess anxiety. Anxiety symptoms coincide with depressive symptoms and stress symptoms. Vice versa, depressive symptoms and stress symptoms have been related to stroke before.9,10 In our study, we detected only a short-term increased risk of stroke in people with anxiety symptoms, independent of depressive symptoms. This may be a true effect with possible mechanisms being activation of the hypothalamic–pituitary–adrenal axis and sympathetic nervous system or adverse health behaviors.1 This is further supported by the finding that a stricter dichotomization of the anxiety symptom score resulted in stronger effects. Moreover, there may have been some nondifferential misclassification in our assessment of anxiety, which could have resulted in an underestimation of the true effect. In contrast, the lack of a long-term effect argues against a true effect. Moreover, we found no associations with anxiety disorders, for which any mechanism should have a similar effect. An alternative explanation may, thus, be reversed causality as suggested previously for mortality.11 Given that patients with stroke have a high cardiovascular risk, they may experience a worsening health in the years preceding stroke and get anxious in response.
Strengths of our study are the population-based setting, the long follow-up period, the stringent stroke monitoring, and the limited loss of follow-up. A limitation is that due to lacking neuroimaging after stroke, 20% of all strokes remained of unspecified subtype. In addition, we did not perform a formal evaluation of the reliability of our anxiety assessments. Furthermore, we may have missed weak associations: assuming α=0.05, our study had a power of 0.77 to detect the hazard ratio of 1.43 that was found previously for high versus low anxiety (in our study HADS-A score, ≥8 versus <8).4,12
In conclusion, anxiety disorders were not associated with stroke in this general population study. Anxiety symptoms were only associated to stroke in the short term, which needs further exploration.
Sources of Funding
The Rotterdam Study was supported by the Erasmus Medical Center Rotterdam, the Erasmus University Rotterdam, the Netherlands Organisation for Scientific Research, the Netherlands Organisation for Health Research and Development (ZonMw), the Research Institute for Diseases in the Elderly, the Ministry of Education, Culture and Science, the Ministry of Health, Welfare and Sports, the European Commission, and the Municipality of Rotterdam.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.115.012361/-/DC1.
- Received December 8, 2015.
- Accepted January 27, 2016.
- © 2016 American Heart Association, Inc.
- Lambiase MJ,
- Kubzansky LD,
- Thurston RC.
- Fransson EI,
- Nyberg ST,
- Heikkilä K,
- Alfredsson L,
- Bjorner JB,
- Borritz M,
- et al
- Rothman K.