ADVERTISEMENT

Issue Navigator

Volume 14 No. 05
Earn CME
Accepted Papers
Classifieds





Scientific Investigations

Lack of Worsening of Sleep-Disordered Breathing After Recurrent Stroke in the BASIC Project

Devin L. Brown, MD1; Chengwei Li, MPH1,2; Brisa N. Sánchez, PhD3; Galit Levi Dunietz, PhD4; Ronald D. Chervin, MD, MS4; Erin Case, BA1,2; Nelda M. Garcia, BS1,2; Lynda D. Lisabeth, PhD1,2
1Stroke Program, University of Michigan Medical School, Ann Arbor, Michigan; 2Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, Michigan; 3Department of Biostatistics, School of Public Health, University of Michigan, Ann Arbor, Michigan; 4Sleep Disorders Center and Department of Neurology, University of Michigan Medical School, Ann Arbor, Michigan

ABSTRACT

Study Objectives:

To investigate the difference in sleep-disordered breathing (SDB) prevalence and severity after an index and recurrent stroke.

Methods:

In a sample of 40 subjects, home sleep apnea tests were performed a median of 10 days after an index ischemic stroke and 14 days after a recurrent ischemic stroke. A respiratory event index (REI) of ≥ 10 events/h (apneas plus hypopneas per hour of recording) was used to define clinically significant SDB. The relative difference in REI or relative SDB prevalence was used to compare the post-recurrent stroke measurement with that made after the index stroke, and was expressed as a rate ratio (RR) or prevalence ratio (PR). Adjusted regression models (negative binomial for REI and log binomial for SDB) included change in body mass index and time between the events.

Results:

The median time from index to recurrent stroke was 330.5 days (interquartile range [IQR]: 103.5, 766.5). The median REI was 17.5 (IQR: 9.0, 32.0) after the index stroke and 18.0 (IQR: 11.0, 25.5) after the recurrent stroke. The within-subject median difference was zero (IQR: −9, 7.5). The relative difference in REI was not significant in unadjusted or adjusted (RR: 0.97 [95% confidence interval: 0.76, 1.24]) models. The prevalence of SDB was not different after the recurrent stroke compared with the index stroke, in unadjusted or adjusted (PR: 1.10 [95% confidence interval: 0.91, 1.32]) models.

Conclusions:

In this within-subject, longitudinal study, neither severity nor prevalence of SDB worsened after recurrent stroke.

Citation:

Brown DL, Li C, Sánchez BN, Dunietz GL, Chervin RD, Case E, Garcia NM, Lisabeth LD. Lack of worsening of sleep-disordered breathing after recurrent stroke in the BASIC project. J Clin Sleep Med. 2018;14(5):835–839.


BRIEF SUMMARY

Current Knowledge/Study Rationale: The purpose of the study was to capitalize on a unique opportunity to gain insights into the pathophysiology of post-stroke sleep-disordered breathing. We performed physiological assessments of sleep-disordered breathing longitudinally, in the same subjects, after an index stroke and a recurrent stroke.

Study Impact: This within-person, repeated-measures analysis showed similar sleep-disordered breathing severity after an index and recurrent stroke. These results did not support our hypothesis that recurrent stroke has an incremental negative effect on sleep-disordered breathing severity or prevalence and generates hypotheses about the pathophysiology of the highly prevalent condition of post-stroke sleep apnea.

INTRODUCTION

The complex relationship between sleep-disordered breathing (SDB) and stroke remains incompletely explored. The similarly high prevalence of post-stroke and post-transient ischemic attack (TIA) SDB,1 predominated by obstructive rather than central sleep apnea,2 and persistently high prevalence of post-stroke SDB late after stroke3,4 help support the concept that SDB often predates stroke. Furthermore, SDB is now an established risk factor for incident stroke based on laboratory- and community-based studies,58 with multiple meta-analyses to support this relationship.911 Conversely, the higher prevalence of SDB in stroke patients with dysphagia,12 and higher prevalence of SDB after brainstem versus other infarction sites,13 suggest that stroke might contribute to or exacerbate SDB. However, to address most directly and definitively whether stroke exacerbates or causes SDB would require a sleep apnea test be performed immediately before and after a stroke—an almost impossible feat to perform in a systematic investigation. Therefore, as a next best alternative, to gain insights into the effects of stroke on SDB and generate hypotheses about the pathophysiology of post-stroke SDB, we examined the within-person difference between SDB as assessed objectively by the respiratory event index (REI) after an index stroke and again after recurrent stroke. We hypothesized that recurrent stroke leaves patients with incrementally higher REIs than do index strokes.

METHODS

Detailed methods of the ongoing Brain Attack Surveillance in Corpus Christi (BASIC) Project have been published previously.14 Briefly, active and passive surveillance are used to identify acute cases of ischemic stroke and intracerebral hemorrhage from all acute care hospitals in Nueces County, Texas. Cases are validated by stroke fellowship-trained physicians after review of source documents. Residents of the county who were age 45 years or older were offered enrollment and then underwent a study interview. Those who were not pregnant and did not use oxygen, mechanical ventilation, or positive pressure ventilation before 30 days (active surveillance) or 45 days (passive surveillance) were considered eligible for an SDB substudy and were offered enrollment for SDB screening from July 2010 to May 2016 (intracerebral hemorrhage screening stopped on October 30, 2015). Subjects who had been previously enrolled in the substudy were offered repeat SDB screening in the event of a recurrent stroke if eligibility criteria were again met. In this report, the first stroke with associated SDB screening is referred to as the index stroke, but may not represent the first stroke for each individual subject (eg, if a prior stroke preceded the study time period). Written informed consent was obtained from the patient or surrogate and the study was Institutional Review Board-approved by the University of Michigan and the Corpus Christi hospital systems.

A home sleep apnea test (previously termed a portable “out of center sleep test”) was performed overnight during the acute hospitalization or in the subjects' home with the well-validated ApneaLink Plus device (ResMed, San Diego, California, United States)15 that measures oxygen saturation, nasal pressure, respiratory effort, and pulse. Detailed methods of the SDB substudy have been published.13 ApneaLink Plus software with default settings used in validation studies was used to identify apneas and hypopneas. Our hypopnea definition used the 4% threshold in keeping with the prevailing definition at the time of the study outset.13 Automated scoring of apneic events was then edited by a registered polysomnographic technologist as previously described. The software tabulated the REI, a sum of the apneas plus hypopneas per hour of recording (events/h). The SDB severity was categorized as none (REI < 5 events/h), mild (REI 5 to < 15 events/h), moderate (REI 15 to < 30 events/h), and severe (REI ≥ 30 events/h).

Age, risk factors (hypertension, diabetes mellitus, high cholesterol, atrial fibrillation, coronary artery disease, history of stroke/TIA before the study time period, body mass index [BMI]), and initial stroke severity (National Institutes of Health Stroke Scale [NIHSS]) were abstracted from the medical record for the index event, and BMI and NIHSS for both the index and recurrent events. Race/ethnicity, current smoking status, and current alcohol intake (categorical: none, < 1, 1–14, and > 14 drinks/wk), were obtained from the patient/proxy interview conducted shortly after the index event. The Berlin Questionnaire, a validated tool,16 was also administered by way of this interview, with respect to the pre-stroke state, to gain insights into the risk of SDB prior to the index event. Results were categorized in the standard fashion into high risk and low risk.

Statistical Methods

Descriptive statistics were calculated for baseline characteristics. Given that REI is a count variable and exhibits overdispersion, negative binomial regression models were used to evaluate the relative difference in REI between the recurrent and index stroke after controlling for change in BMI and time between the events. Results are expressed as relative difference in rate ratios of the REI after the recurrent stroke compared with the REI after the index stroke. Change in BMI and time between index and recurrent stroke were pre-specified as covariates due to the possibility that they could confound the within-person change in REI, as aging and weight change can influence SDB. Stroke severity was not included in the models given its lack of association with SDB.3,17,18 To account for the correlation between measures within each subject, the models were estimated using generalized estimating equations (GEE) with a compound symmetry correlation structure. The McNemar test was used to compare the crude proportion with SDB, defined by REI ≥ 10 events/h, at each time point. A similar GEE model based on log binomial regression was used to calculate the relative difference in the prevalence of SDB (REI ≥ 10 events/h) between the index and recurrent stroke, after controlling for change in BMI and time between index and recurrent strokes. This association was expressed as a prevalence ratio, for SDB after the recurrent stroke as compared with SDB after the index stroke.

RESULTS

During the study time period, 49 subjects consented for a second SDB screening after a recurrent stroke; 40 had usable SDB data at both time points. These 40 were representative of the overall SDB screening substudy participants who had recurrent events but did not have SDB assessment after their recurrent stroke with respect to baseline characteristics including NIHSS associated with the recurrent event (data not shown). Baseline characteristics of these 40 subjects are found in Table 1. Of the 40, 1 index event and 1 recurrent event were an intracerebral hemorrhage; these events occurred within the same subject. Most subjects had prevalent SDB: 29 (73%) following the index stroke and 33 (83%) following the recurrent stroke (P = .16).

Baseline characteristics assessed at the time of the index stroke.

jcsm.14.5.835.t01.jpg

table icon
Table 1

Baseline characteristics assessed at the time of the index stroke.

(more ...)

The median time from index stroke to SDB screening was 10 days (interquartile range [IQR]: 6, 21), and from recurrent stroke to SDB screening it was 14 days (IQR: 9, 23). The median time from index to recurrent stroke was 330.5 days (IQR: 103.5, 766.5). The within-person median change in BMI from the index to recurrent stroke was −0.16 kg/m2 (IQR: −1.56, 1.23). The median initial NIHSS after the index stroke was 2.0 (IQR: 0, 4.0), and after the recurrent stroke it was 4.0 (IQR: 2.5, 7.0). The within-person median increase (from index event to recurrent event) in initial NIHSS was 1.5 (IQR: 0, 5.0).

The median REI following the index stroke was 17.5 (IQR: 9.0, 32.0), and following the recurrent stroke was 18.0 (IQR: 11.0, 25.5), with a median within-person difference of 0 (IQR: −9, 7.5). Figure 1 shows a histogram of within-person differences between REI following the recurrent stroke and index stroke. Table 2 shows details for those subjects whose SDB severity categories changed between the sleep apnea tests that followed the index and recurrent strokes. In the models, no significant relative difference in REI, between recurrent and index strokes, existed in unadjusted analysis (P = .73), or after adjustment for change in BMI and time between events (P = .79) (Table 3). Similarly, the prevalence of SDB was not different following the recurrent stroke compared with the index stroke, in both unadjusted and adjusted analyses (Table 3). Neither time between the index and recurrent strokes nor change in BMI between the events was associated with the relative difference in REI or with SDB prevalence in the adjusted models (Table 3).

Histogram of within-person differences between REI following the recurrent stroke and REI following the index stroke (recurrent minus index).

REI = respiratory event index.

jcsm.14.5.835a.jpg

jcsm.14.5.835a.jpg
Figure 1

Histogram of within-person differences between REI following the recurrent stroke and REI following the index stroke (recurrent minus index).

(more ...)

Changes in SDB severity between results following index and recurrent strokes.

jcsm.14.5.835.t02.jpg

table icon
Table 2

Changes in SDB severity between results following index and recurrent strokes.

(more ...)

Relative difference in severity and prevalence of SDB: a comparison between measurements made after recurrent and index stroke (n = 40).

jcsm.14.5.835.t03.jpg

table icon
Table 3

Relative difference in severity and prevalence of SDB: a comparison between measurements made after recurrent and index stroke (n = 40).

(more ...)

DISCUSSION

This within-person, repeated-measures analysis showed similar SDB severity after an index and recurrent stroke. These results therefore do not support our hypothesis that recurrent stroke has an incremental negative effect on SDB severity or prevalence. These unique findings provide important new evidence about the pathophysiology of post-stroke SDB.

If stroke caused or contributed to SDB, SDB recovery patterns following stroke would be expected to parallel stroke-related neurological dysfunction. SDB has been shown to improve in the months following stroke,3,12,1922 consistent with improvement in neurological deficits after stroke. However, although neurological deficits are typically greater after a recurrent event compared with an index event,23 such as in the current study, REI was no worse after a recurrent stroke. Thus, neurological deficits accumulate with recurrent events, but REI does not appear to worsen incrementally. Possible explanations for this include: (1) stroke's effects on the airway terminate before a recurrent stroke; thus, an individual's REI is “reset” before a new stroke. The very high prevalence of SDB years after stroke, much higher than that of similarly aged people, argues against this, however.4 (2) A first stroke contributes to SDB, but additional strokes cause no further SDB exacerbation. A specific physiological mechanism by which this would occur is not obvious. Because infarction volume is not associated with apnea/hypopnea counts,17 accumulation of additional areas of infarction should not necessarily be associated with worsened SDB. (3) Or, in most cases, stroke does not contribute to SDB.

We were unable to test the immediate effects of stroke on SDB, as this would require a sleep apnea test to be performed just prior to and immediately after stroke. This of course presents substantial logistical challenges. We have no information about SDB severity between the two ApneaLink assessments, but the times from index and recurrent stroke to SDB assessments were similar, and we did in addition adjust for time between strokes. Other limitations to this work include the use of home sleep apnea testing that may underestimate the severity of SDB, given that the number of apneic events is divided by total recording time rather than sleep time, as in polysomnography. However, application to within-person changes, similarly timed post-stroke SDB assessments, and edits by a polysomnographic technologist (to edit start and stop times) mitigates this issue. Furthermore, the device used has been very well validated.15 We do not yet have longitudinal REI information on patients without recurrent events. The study population was characterized overall by low stroke severity, because the sample was derived from a community without an academic medical center, and because the study population was derived from a population-based study.24 Even population-based studies of ischemic stroke inclusive of tertiary care centers have a median NIHSS of only 3 to 4, as most strokes are mild.25 The SDB severity found here was similar to that of our much larger sample (n = 549)—mostly in the mild-to-moderate range—but tended more toward the moderate range, probably because of the higher proportion of Mexican Americans.18 Our sample shared many baseline characteristics with typical stroke patients with respect to stroke and SDB severity, and BMI category.18,24 These features improve the generalizability of the results. However, given the convenience-based sample, patients who have experienced more severe stroke may not have participated at the time of index stroke. Nonetheless, stroke severity does not appear to be associated with SDB.3,17,18 We do not have information about ischemic stroke subtype, although subtype also is not associated with post-stroke SDB in general,17 or in this particular community.26 Information about SDB treatment was not collected, although current use of positive airway pressure was exclusionary for ApneaLink assessment at the time of each stroke event.

In summary, this work extends our knowledge about post-stroke SDB. Our results demonstrated that unlike changes seen in neurological deficits following a recurrent event, SDB does not appear to worsen incrementally after a recurrent stroke. These findings raise the hypothesis that stroke may not worsen SDB and challenges the notion that stroke is a risk factor for SDB.

DISCLOSURE STATEMENT

Work for this study was performed in the Corpus Christi Medical Center and CHRISTUS Spohn hospitals, CHRISTUS Health System, in Corpus Christi, Texas and University of Michigan. All authors have seen and approved the manuscript. Funding was provided by grants: R01NS070941 and R01HL098065. Brown: received or receives funding from R01HL126700, R01HL123379, R01 NS070941, R01 HL098065 related to this work and has received travel support from the American Academy of Sleep Medicine. Li: has received funding from U10 NS086526, R01 NS070941, and R01NS091112. Sánchez: received funding from R01HL126700, R01HL123379, R01 NS070941, R01 HL098065. Dunietz: supported by a T32 Grant from the National Institute of Neurological Disorders and Stroke (NIH/NINDS T32 NS007222). Chervin: received funding from R01HL126700, R01HL123379, R01 NS070941, R01 HL098065, R01 HL105999, T32HL110952, and R43 HL117421.

He was involved with unrestricted educational gifts to the University of Michigan from Philips Respironics and Fisher Paykel; has served on the Board of Governors (unpaid) for the non-profit Sweet Dreamzzz; has consulted for Proctor & Gamble and Zansors; serves as an editor for UpToDate; and has produced copyrighted material, patents, and patents pending, owned by the University of Michigan, focused on assessment or treatment of sleep disorders. He serves on the Boards of Directors for the International Pediatric Sleep Association and the American Academy of Sleep Medicine for which he is also the immediate Past President. Case: has received funding from R01NS091112, R01 NS070941, R01NS038916, R01HL126700. Garcia: has received funding from R01NS038916, R01NS070941, R01HL126700 Lisabeth: has received funding from R01NS038916, R01HL126700, R01HL123379, R01 NS070941, R01 HL098065.

ABBREVIATIONS

BASIC

Brain Attack Surveillance in Corpus Christi

BMI

body mass index

GEE

generalized estimating equations

IQR

interquartile range

NIHSS

National Institutes of Health Stroke Scale

REI

respiratory event index

SDB

sleep-disordered breathing

TIA

transient ischemic attack

REFERENCES

1 

Bassetti C, Aldrich M. Sleep apnea in acute cerebrovascular diseases: final report on 128 patients. Sleep. 1999;22(2):217–223. [PubMed]

2 

Johnson KG, Johnson DC. Frequency of sleep apnea in stroke and TIA patients: a meta-analysis. J Clin Sleep Med. 2010;6(2):131–137. [PubMed Central][PubMed]

3 

Bassetti CL, Milanova M, Gugger M. Sleep-disordered breathing and acute ischemic stroke: diagnosis, risk factors, treatment, evolution, and long-term clinical outcome. Stroke. 2006;37:967–972. [PubMed]

4 

Cadilhac DA, Thorpe RD, Pearce DC, et al. Sleep disordered breathing in chronic stroke survivors. A study of the long term follow-up of the SCOPES cohort using home based polysomnography. J Clin Neurosci. 2005;12(6):632–637. [PubMed]

5 

Campos-Rodriguez F, Martinez-Garcia MA, Reyes-Nuñez N, Caballero-Martinez I, Catalan-Serra P, Almeida-Gonzalez CV. Role of sleep apnea and continuous positive airway pressure therapy in the incidence of stroke or coronary heart disease in women. Am J Respir Crit Care Med. 2014;189(12):1544–1550. [PubMed]

6 

Munoz R, Duran-Cantolla J, Martinez-Vila E, et al. Severe sleep apnea and risk of ischemic stroke in the elderly. Stroke. 2006;37(9):2317–2321. [PubMed]

7 

Redline S, Yenokyan G, Gottlieb DJ, et al. Obstructive sleep apnea hypopnea and incident stroke: the sleep heart health study. Am J Respir Crit Care Med. 2010;182(2):269–277. [PubMed Central][PubMed]

8 

Yaggi HK, Concato J, Kernan WN, Lichtman JH, Brass LM, Mohsenin V. Obstructive sleep apnea as a risk factor for stroke and death. N Engl J Med. 2005;353(19):2034–2041. [PubMed]

9 

Dong JY, Zhang YH, Qin LQ. Obstructive sleep apnea and cardiovascular risk: meta-analysis of prospective cohort studies. Atherosclerosis. 2013;229(2):489–495. [PubMed]

10 

Li M, Hou WS, Zhang XW, Tang ZY. Obstructive sleep apnea and risk of stroke: a meta-analysis of prospective studies. Int J Cardiol. 2014;172(2):466–469. [PubMed]

11 

Loke YK, Brown JW, Kwok CS, Niruban A, Myint PK. Association of obstructive sleep apnea with risk of serious cardiovascular events: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes. 2012;5(5):720–728. [PubMed]

12 

Martinez-Garcia MA, Galiano-Blancart R, Soler-Cataluna JJ, Cabero-Salt L, Roman-Sanchez P. Improvement in nocturnal disordered breathing after first-ever ischemic stroke: role of dysphagia. Chest. 2006;129(2):238–245. [PubMed]

13 

Brown DL, McDermott M, Mowla A, et al. Brainstem infarction and sleep-disordered breathing in the BASIC Sleep Apnea Study. Sleep Med. 2014;15(8):887–891. [PubMed Central][PubMed]

14 

Morgenstern LB, Smith MA, Lisabeth LD, et al. Excess stroke in Mexican Americans compared with non-Hispanic Whites: the Brain Attack Surveillance in Corpus Christi Project. Am J Epidemiol. 2004;160(4):376–383. [PubMed Central][PubMed]

15 

Ng SS, Chan TO, To KW, et al. Validation of a portable recording device (Apnealink) for identifying patients with suspected obstructive sleep apnea syndrome. Intern Med J. 2009;39(11):757–762. [PubMed]

16 

Netzer NC, Hoegel JJ, Loube D, et al. Prevalence of symptoms and risk of sleep apnea in primary care. Chest. 2003;124(4):1406–1414. [PubMed]

17 

Iranzo A, Santamaria J, Berenguer J, Sánchez M, Chamorro A. Prevalence and clinical importance of sleep apnea in the first night after cerebral infarction. Neurology. 2002;58(6):911–916. [PubMed]

18 

Lisabeth LD, Sánchez BN, Chervin RD, et al. High prevalence of post-stroke sleep disordered breathing in Mexican Americans. Sleep Med. 2017;33:97–102. [PubMed]

19 

Parra O, Abroix A, Bechich S, et al. Time course of sleep-related breathing disorders in first-ever stroke or transient ischemic attack. Am J Respir Crit Care Med. 2000;161(2 Pt 1):375–380. [PubMed]

20 

Dziewas R, Hopmann B, Humpert M, et al. Positional sleep apnea in patients with ischemic stroke. Neurol Res. 2008;30(6):645–648. [PubMed]

21 

Harbison J, Ford GA, James OF, Gibson GJ. Sleep-disordered breathing following acute stroke. QJM. 2002;95(11):741–747. [PubMed]

22 

Hui D, Choy D, Wong L, et al. Prevalence of sleep-disordered breathing and continuous positive airway pressure compliance: results in Chinese patients with first-ever ischemic stroke. Chest. 2002;122(3):852–860. [PubMed]

23 

Moerch-Rasmussen A, Nacu A, Waje-Andreassen U, Thomassen L, Naess H. Recurrent ischemic stroke is associated with the burden of risk factors. Acta Neurol Scand. 2016;133(4):289–294. [PubMed]

24 

Lisabeth LD, S+ínchez BN, Baek J, et al. Neurological, functional, and cognitive stroke outcomes in Mexican Americans. Stroke. 2014;45(4):1096–1101. [PubMed Central][PubMed]

25 

Madsen TE, Khoury JC, Alwell KA, et al. Analysis of tissue plasminogen activator eligibility by sex in the Greater Cincinnati/Northern Kentucky Stroke Study. Stroke. 2015;46(3):717–721. [PubMed Central][PubMed]

26 

Brown DL, Mowla A, McDermott M, et al. Ischemic stroke subtype and presence of sleep-disordered breathing: the BASIC Sleep Apnea Study. J Stroke Cerebrovasc Dis. 2015;24(2):388–393. [PubMed]