Abstract
Background/Purpose:
To determine whether obstructive sleep apnea (OSA) is associated with intracerebral hemorrhage (ICH) risk, we assessed premorbid OSA exposure of patients with non-traumatic ICH and matched controls.
Methods:
Ethnic/Racial Variations of Intracerebral Hemorrhage is a multicenter, case-control study evaluating risk factors for ICH that recruited 3000 cases with ICH and 3000 controls. OSA status was ascertained using the Berlin Questionnaire (BQ) as a surrogate for premorbid OSA. We performed logistic regression analyses to evaluate the association between OSA and ICH.
Results:
2064 (71%) cases and 1516 (52%) controls were classified as having OSA by the BQ. Cases with OSA were significantly more likely to be male and have hypertension, heart disease, hyperlipidemia, and higher body mass index compared to those without OSA. OSA was more common among cases compared to controls, (71% versus 52%, OR 2.28, 95% CI 2.05–2.55). In a multivariable logistic regression model, OSA was associated with increased risk for ICH (odds ratio, 1.47; 95% CI 1.29-1.67).
Conclusions:
OSA is a risk factor for ICH.
Intracerebral hemorrhage (ICH) is a devastating stroke subtype, accounting for up to 20% of strokes.1 The 30-day case-fatality rate for ICH is up to 50%,2 and the majority of survivors experience severe functional disability. Only a quarter of ICH patients achieve functional independence at 90-days.3 Efforts to identify novel risk factors and treatment strategies are critical.
Obstructive sleep apnea (OSA) is a common disorder that results from intermittent upper airway collapse, causing a cycle of abnormal physiology that includes hypoxemia, sympathetic activation, hemodynamic disturbance, and distortion of sleep architecture.4, 5 The gold-standard test for diagnosis is polysomnography, but questionnaires have been created and validated to screen for those at high risk. The Berlin Questionnaire (BQ), which has a sensitivity of 86%, comprises ten questions and classifies patients as high- or low-risk for polysomnography-defined OSA.6
OSA is an independent risk factor for ischemic stroke and is present in up to 80% of these patients.7, 8 Importantly, ischemic stroke and ICH share common pathophysiology,9 providing a compelling rationale for examining the role of OSA in ICH patients. In this study, we tested the hypothesis that OSA is an independent risk factor for ICH.
Methods
Study Design
The methods for the Ethnic/Racial Variations of Intracerebral Hemorrhage study have been previously described.10 The study was approved by the institutional review board at each study site and informed consent was obtained from all subjects or their legal representatives. The data that support the findings of this study are available from the corresponding author on reasonable request.
Cases
Cases were adults with spontaneous ICH who resided within 75 miles of a recruitment site for at least six months. Spontaneous ICH was defined as a non-traumatic, abrupt onset of severe headache with an altered level of consciousness and/or focal neurologic deficit and a focal blood collection within the brain parenchyma seen on neuroimaging or autopsy.10 Exclusions included ICH due to coagulopathy, tumor, thrombosis, vascular malformation, aneurysm, and hemorrhagic conversion of recent ischemic stroke. Cases were recruited at 42 clinical sites and enrolled in the acute setting. Sources of data included patient interview, physical examination, chart review, blood work, and imaging.
Controls
Controls were selected at random, independent of exposure, to balance cases by age, sex, race/ethnicity, and metropolitan area.
Exposure Ascertainment
The BQ is a 10-item questionnaire consisting of demographics, snoring status, witnessed apneas, daytime sleepiness, and hypertension status.6 OSA risk is classified as “high” or “low” based on the score. Responses were obtained at the time of study enrollment by the subject or designated proxy.
Primary Analysis
Differences in baseline characteristics were compared using Fisher’s exact test or Student’s t-test. Only subjects with completed BQs were included in the analysis. Univariable and multivariable logistic regression analyses were performed. Multivariable model building proceeded in several steps: first, covariates with p<0.10 in univariable analyses were included; second, age, sex, and race/ethnicity were forced into the model; and third, variables from the univariate analysis with p<0.10 were retained in the model if they remained significant at the p<0.10 level.
Secondary Analyses
An analysis based on ICH location was performed using the model previously described. Sensitivity and layered analyses to elucidate the role of hypertension were also performed.
Results
Participants
A total of 6000 subjects were enrolled. Complete BQ results were available for 5808 participants (Online Figure I). Of the 3000 ICH cases, 2896 completed the BQ of whom 2064 (71.1%) had a score suggestive of OSA. BQ was obtained by self-report in 1296 cases and by proxy in 1600 cases. A sensitivity analysis was performed with no significant difference between the groups. Compared to ICH cases with low risk of OSA, those at high risk were more likely to be male (61.1% versus 53.5%, p<0.001) with hypertension (92.9% versus 64.5%, p<0.001), coronary artery disease (23.9% versus 16.0%, p>0.001), diabetes (32.0% versus 20.4%, p<0.001), hyperlipidemia (49.3% versus 37.7%, p<0.001), and higher mean body mass index (29.8 ± 8.1 versus 26.8 ± 6.5, p<0.001), and less likely to have lobar ICH (29.0% versus 35.0%, p=0.002) (Table 1). Among the 3000 controls, 2912 completed the BQ of whom 1516 (52.1%) had likely OSA.
Table 1.
Characteristics of ICH Patients by Likelihood of OSA
| Low Risk for OSA | High Risk for OSA | P-value | |
|---|---|---|---|
| 832 | 2064 | ||
| Mean Age (SD) | 62.50 (14.70) | 61.31 (13.72) | 0.04 |
| Female (%) | 387 (46.5) | 803 (38.9) | <0.001 |
| Race/Ethnicity (%) | 0.021 | ||
| White | 290 (34.9) | 675 (32.7) | |
| Black | 294 (35.3) | 663 (32.1) | |
| Hispanic | 248 (29.8) | 726 (35.2) | |
| Lobar (%) | 291 (35.0) | 599 (29.0) | 0.002 |
| Mean Body Mass Index (SD) | 26.79 (6.47) | 29.80 (8.09) | <0.001 |
| Hypertension (%) | 537 (64.5) | 1918 (92.9) | <0.001 |
| Atrial Fibrillation (%) | 79 (9.6) | 260 (12.6) | 0.025 |
| Coronary Artery Disease/Myocardial Infarction (%) | 132 (16.0) | 493 (23.9) | <0.001 |
| Heart Failure (%) | 69 (8.3) | 283 (13.7) | <0.001 |
| Hyperlipidemia (%) | 308 (37.7) | 1009 (49.3) | <0.001 |
| Diabetes (%) | 170 (20.4) | 661 (32.0) | <0.001 |
| Smoking Ever (%) | 420 (50.7) | 1109 (53.8) | 0.143 |
| Alcohol Use (%) | 414 (50.2) | 1103 (53.5) | 0.118 |
| Anticoagulant Use (%) | 41 (4.9) | 128 (6.2) | 0.217 |
| Antiplatelet Use (%) | 142 (17.1) | 430 (20.8) | 0.024 |
| Snoring (%) | 381 (45.8) | 1802 (87.3) | <0.001 |
Abbreviations: SD = standard deviation
Cases had significantly higher rates of OSA by BQ relative to controls, (71.3% versus 52.1%, OR 2.28, 95% CI 2.05–2.55). As expected, they also had higher rates of cardiovascular comorbidities, including hypertension (84.8% versus 53.4%, p<0.001), coronary artery disease (21.6% versus 11.8%, p<0.001), atrial fibrillation (11.7% versus 5.2%, p<0.001), and diabetes (28.7% versus 22%, p<0.001) (Table 2).
Table 2.
Characteristics of Controls versus Cases
| Control | Case | P-value | |
|---|---|---|---|
| N | 2912 | 2896 | |
| Mean Age (SD) | 61.54 (13.86) | 61.65 (14.01) | 0.749 |
| Female (%) | 1216 (41.8) | 1190 (41.1) | 0.625 |
| Race/Ethnicity (%) | 0.766 | ||
| White | 995 (34.2) | 965 (33.3) | |
| Black | 942 (32.3) | 957 (33.0) | |
| Hispanic | 975 (33.5) | 974 (33.6) | |
| Mean Body Mass Index (SD) | 29.68 (6.69) | 28.93 (7.78) | <0.001 |
| Hypertension (%) | 1556 (53.4) | 2455 (84.8) | <0.001 |
| Atrial Fibrillation (%) | 150 (5.2) | 339 (11.7) | <0.001 |
| Coronary Artery Disease/Myocardial Infarction (%) | 344 (11.8) | 625 (21.6) | <0.001 |
| Heart Failure (%) | 121 (4.2) | 352 (12.2) | <0.001 |
| Hyperlipidemia (%) | 1355 (46.7) | 1317 (46.0) | 0.609 |
| Diabetes (%) | 640 (22.0) | 831 (28.7) | <0.001 |
| Smoking Ever (%) | 1436 (49.3) | 1529 (52.9) | 0.008 |
| Alcohol Use (%) | 1689 (58.1) | 1517 (52.6) | <0.001 |
| Anticoagulant Use (%) | 51 (1.8) | 169 (5.8) | <0.001 |
| Antiplatelet Use (%) | 158 (5.4) | 572 (19.8) | <0.001 |
| Snoring (%) | 1943 (66.7) | 2183 (75.4) | <0.001 |
| Obstructive Sleep Apnea by BQ (%) | 1516 (52.1) | 2064 (71.3) | <0.001 |
Abbreviations: SD = standard deviation
Association between OSA and ICH
Table 3 shows the results of the multivariate logistic regression model demonstrating that OSA was significantly associated with increased risk of ICH (OR: 1.47; 95% CI 1.29-1.67).
Table 3.
Logistic Regression Analyses of the Association Between OSA and Risk of ICH
| Covariate | OR, unadjusted (95% CI, P-value) | OR, adjusted (95% CI, P-value) |
|---|---|---|
| Age | 1.00 (1.00-1.00, p=0.749) | 0.97 (0.97-0.98, p<0.001) |
| Female | 0.97 (0.88-1.08, p=0.606) | 1.12 (1.00-1.27, p=0.057) |
| Black | 1.05 (0.92-1.19, p=0.471) | 0.81 (0.69-0.94, p=0.007) |
| Hispanic | 1.03 (0.91-1.17, p=0.644) | 0.93 (0.80-1.08, p=0.348) |
| Body Mass Index | 0.99 (0.98-0.99, p<0.001) | 0.95 (0.95-0.96, p<0.001) |
| Anticoagulant Use | 3.48 (2.55-4.82, p<0.001) | 2.39 (1.67-3.48, p<0.001) |
| Antiplatelet Use | 4.29 (3.58-5.17, p<0.001) | 3.73 (3.05-4.58, p<0.001) |
| Atrial Fibrillation | 2.44 (2.00-2.99, p<0.001) | 1.57 (1.22-2.01, p<0.001) |
| Alcohol Use | 0.80 (0.72-0.89, p<0.001) | 0.84 (0.75-0.95, p=0.006) |
| Heart Failure | 3.19 (2.59-3.97, p<0.001) | 1.87 (1.47-2.39, p<0.001) |
| Hypertension | 4.85 (4.29-5.50, p<0.001) | 4.74 (4.09-5.51, p<0.001) |
| Obstructive Sleep Apnea by BQ (%) | 2.28 (2.05-2.55, p<0.001) | 1.47 (1.29-1.67, p<0.001) |
Abbreviations: OR = odds ratio, CI = confidence interval, BQ = Berlin Questionnaire
Secondary Analysis
ICH Location
There were 2006 non-lobar ICH cases and 890 lobar cases, and 1465 (73%) and 599 (67.3%) scored high risk for OSA on the BQ, respectively. OSA remained a risk factor regardless of location in the stratified analysis (OR: lobar 1.90 (95% CI 1.62-2.22) and non-lobar 2.49 (95% CI 2.21-2.82)). This association remained significant in multivariate logistic regression analyses (OR: lobar 1.55 (95% CI 1.28-1.88) and non-lobar 1.49 (95% CI 1.29-1.72)) (Online Tables I–II).
Hypertension
A sensitivity analysis was performed by removing hypertension from the BQ in which 38% of controls and 44% of cases scored high risk for OSA (OR 1.28, 95% CI 1.16-1.43). A layered analysis demonstrated that the risk impact of a high BQ score on multivariate logistic regression increased from 1.47 (95% CI 1.29-1.67) to 2.39 (95% CI 2.13-2.70) when hypertension was not included in the multivariable model.
Discussion
Using a case-control, multicenter study, we report that OSA is a significant, independent risk factor for ICH, regardless of hypertension status or hemorrhage location. Given the shared pathophysiology between ICH and ischemic stroke, proposed mechanisms include sympathetic activation,11 oxidative stress,12 inflammation,13 and endothelial dysfunction,14 which lead to atherosclerosis and small vessel ischemic disease. Mounting evidence indicates that risk factors associated with ICH also influence the severity and outcome of this condition and as underlying small vessel disease worsens, hematoma volume increases.15
The strengths of our study include the large sample size permitting the control of multiple risk factors and robust ascertainment with nearly all participants completing the BQ. Although our analysis is cross-sectional, the BQ, was obtained during the initial intake interview, suggesting that answers are based on data predating the ICH event. Since the BQ relies on self- or proxy-report, there is a notable predisposition for diagnostic misclassification. This, however, would tend to bias the results toward the null hypothesis and does not explain our findings.
Our results invite multiple areas of further investigation, including (1) objective measurement of OSA with polysomnography to mitigate the limitations imposed by the BQ, (2) exploration of biomarkers to provide mechanistic support, (3) evaluation of the impact of OSA treatment on functional recovery, and (4) exploration of the role OSA plays in small vessel ischemic disease.
Summary
Our study provides strong evidence that OSA is an important risk factor for ICH. This finding holds true after adjusting for confounding variables, including hypertension, and when stratifying by ICH location.
Supplementary Material
Acknowledgements
Geer: study design, manuscript draft/revision
Vanent, Leasure: data analysis
Woo, Molano, Sansing, Pisani: manuscript revision
Langefeld: data analysis, manuscript revision
Falcone, Yaggi, Sheth: study design, manuscript revision
Funding
Geer: NIH
Falcone: NIH, American Heart Association, Neurocritical Care Society Research Fellowship
Woo: NIH and National Institute of Neurological Disorders and Stroke
Langefeld: National Institute of Neurological Disorders and Stroke
Yaggi: NIH, Veteran’s Administration Merit
Sheth: NIH
Disclosures
Sheth: Research grants from NIH, AHA, Bard, Novartis, Biogen, Hyperfine, equity from Alva, DSMB Chair for Zoll study, and consulting for NControl and Ceribell.
Non-standard Abbreviations and Acronyms
- ICH
Intracerebral hemorrhage
- OSA
Obstructive sleep apnea
- BQ
Berlin Questionnaire
Footnotes
References
- 1.Sacco S, Marini C, Toni D, Olivieri L, Carolei A. Incidence and 10-year survival of intracerebral hemorrhage in a population-based registry. Stroke. 2009;40:394–399 [DOI] [PubMed] [Google Scholar]
- 2.Flaherty ML, Haverbusch M, Sekar P, Kissela B, Kleindorfer D, Moomaw CJ, Sauerbeck L, Schneider A, Broderick JP, Woo D. Long-term mortality after intracerebral hemorrhage. Neurology. 2006;66:1182–1186 [DOI] [PubMed] [Google Scholar]
- 3.Rost NS, Smith EE, Chang Y, Snider RW, Chanderraj R, Schwab K, FitzMaurice E, Wendell L, Goldstein JN, Greenberg SM, et al. Prediction of functional outcome in patients with primary intracerebral hemorrhage the func score. Stroke. 2008;39:2304–2309 [DOI] [PubMed] [Google Scholar]
- 4.Epstein LJ, Kristo D, Strollo PJ, Friedman N, Malhotra A, Patil SP, Ramar K, Rogers R, Schwab RJ, Weaver EM, et al. Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. Journal of Clinical Sleep Medicine. 2009;5:263–276 [PMC free article] [PubMed] [Google Scholar]
- 5.Hermann DM, Bassetti CL. Role of sleep-disordered breathing and sleep-wake disturbances for stroke and stroke recovery. Neurology. 2016;87:1407–1416 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Netzer NC, Stoohs RA, Netzer CM, Clark K, Strohl KP. Using the berlin questionnaire to identify patients at risk for the sleep apnea syndrome. Annals of Internal Medicine. 1999;131:485-+ [DOI] [PubMed] [Google Scholar]
- 7.Yaggi HK, Concato J, Kernan WN, Lichtman JH, Brass LM, Mohsenin V. Obstructive sleep apnea as a risk factor for stroke and death. New England Journal of Medicine. 2005;353:2034–2041 [DOI] [PubMed] [Google Scholar]
- 8.Pasic Z, Smajlovic D, Dostovic Z, Kojic B, Selmanovic S. Incidence and types of sleep disorders in patients with stroke. Med Arh. 2011;65:225–227 [DOI] [PubMed] [Google Scholar]
- 9.Ariesen MJ, Claus SP, Rinkel GJE, Algra A. Risk factors for intracerebral hemorrhage in the general population - a systematic review. Stroke. 2003;34:2060–2065 [DOI] [PubMed] [Google Scholar]
- 10.Woo D, Rosand J, Kidwell C, McCauley JL, Osborne J, Brown MW, West SE, Rademacher EW, Waddy S, Roberts JN, et al. The ethnic/racial variations of intracerebral hemorrhage (erich) study protocol. Stroke. 2013;44:E120–E125 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Somers VK, Dyken ME, Clary MP, Abboud FM. Sympathetic neural mechanisms in obstructive sleep-apnea. Journal of Clinical Investigation. 1995;96:1897–1904 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lavie L Obstructive sleep apnoea syndrome - an oxidative stress disorder. Sleep Medicine Reviews. 2003;7:35–51 [DOI] [PubMed] [Google Scholar]
- 13.Kohler M, Stradling JR. Mechanisms of vascular damage in obstructive sleep apnea. Nature Reviews Cardiology. 2010;7:677–685 [DOI] [PubMed] [Google Scholar]
- 14.Dyugovskaya L, Lavie P, Lavie L. Increased adhesion molecules expression and production of reactive oxygen species in leukocytes of sleep apnea patients. American Journal of Respiratory and Critical Care Medicine. 2002;165:934–939 [DOI] [PubMed] [Google Scholar]
- 15.Falcone GJ, Biffi A, Devan WJ, Brouwers HB, Anderson CD, Valant V, Ayres AM, Schwab Km Rost NS, Goldstein JN, et al. Burden of blood pressure-related alleles is associated with larger hematoma volume and worse outcome in intracerebral hemorrhage. Stroke. 2013;44:321–+ [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.