Sleep-Disordered Breathing and Stroke
Klar Yaggi, M.D., M.P.H.
Assistant Professor
Yale University School of Medicine
Clinical Epidemiology Research Center (CERC)
Medical Director, VA CT Center for Sleep Medicine
Journal of the Canadian Medical Association; 2006
No relevant conflicts of interest
Outline
• Epidemiologic evidence linking sleep-disordered
breathing to cerebrovascular outcomes
• Mechanisms of stroke in sleep-disordered breathing
• Impact of CPAP treatment on cerebrovascular risk
Outline
• Epidemiologic evidence linking sleep-disordered
breathing to cerebrovascular outcomes
Sleep Apnea Cycle
Sleep
Apnea
Hypoxia
Reoxygenation
Pleural pressure Δ
Ventilation
Arousal
Sympathetic
activation
Sleep Apnea and Diurnal Hypertension
Apnea Hyponea Index
0
>0-<5
≥ 5- <15
≥15
Adjusted* Odds Ratio
Reference
1.42
2.03
2.89
*adjusted for baseline hypertension, age, gender, BMI, waist circumference,
alcohol, and tobacco use
P for trend=0.002
Peppard; NEJM 2000
Sleep Apnea among Patients with Stroke
Source
Study Design
(N)
Confounding
adjustment
Prevalence of
Sleep Apnea
Mohsenin
1995
Cross-sectional
(20)
Age, BMI, HTN,
smoking
80%
Dyken
1997
Case-control
(43)
Age, sex
71%
Basetti
1999
Case-control
(153)
Age, BMI, diabetes,
stroke severity
63%
Parra
2001
Cross-sectional
(161)
Age, race, sex,
smoking, diabetes,
HTN, BMI, Lipids
61%
Yaggi; Lancet Neurology 2003
Stroke leading to Sleep Apnea?
Cerebrovascular
Outcomes
Askenasy; Stroke 1988
Sleep Apnea leading to Stroke?
• OSA strongly associated with Transient
Ischemic Attacks (TIA)
• Persistent OSA post-stroke with resolution of
neuromuscular weakness
• No consistent type/location of stroke in
association with sleep apnea
• Stroke does not cause most sleep apnea
Neau; Sleep Med Rev 2002
Causal Direction?
Cerebrovascular
Outcomes
Confounding?
Cerebrovascular
Outcomes
Central Obesity
-Hypertension
-Diabetes
-Lipids
Yale Study Design
(+)TIA, Stroke,
Death
(-)OSA
N=325 (32%)
Eligible Participants
(without stroke)
N=1022
(-) TIA, Stroke,
Death
(+)TIA, Stroke,
Death
(+)OSA
N=697(68%)
3-6 years of follow-up
(-) TIA, Stroke,
Death
Event-free Survival (TIA, Stroke, Death)
Kaplan-Meier Estimates of the Probability of Event-free Survival among Patients with the
Obstructive Sleep Apnea Syndrome and Controls
Yaggi, H. et al. N Engl J Med 2005;353:2034-2041
Yaggi; NEJM 2005
Dose-Response (Trend) Analysis
Severity of
Syndrome
Stroke or Death
No. of events No. of patients
Follow-up
yrs.
Hazard Ratio
(95% C.I.)
AHI ≤ 3 (ref)
13
271
3.08
1.00
AHI 4-12
21
258
3.06
1.75
(0.88-3.49)
AHI 13-36
20
243
3.09
1.74
(0.87-3.51)
AHI >36
34
250
2.78
3.30
(1.74-6.26)
P=0.005 (Chi-square test for linear trend)
Yaggi; NEJM 2005
Risk of Stroke or Death
Covariate
Unadjusted Hazard Ratio
(95% C.I.)
Adjusted Hazard Ratio
(95% C.I.)
Age (yrs)
1.09 (1.06-1.11)
1.08 (1.06-1.11)
Male sex
0.99 (0.62-1.60)
0.78 (0.48-1.28)
Body Mass Index
0.99 (0.97-1.02)
0.99 (0.96-1.02)
Current Smoker
1.21 (0.90-1.64)
1.46 (0.78-2.98)
Diabetes Mellitus
1.56 (1.02-2.59)
1.31 (0.76-1.26)
Hyperlipidemia
1.04 (0.64-1.68)
1.01 (0.61-1.66)
Hypertension
1.48 (0.95-2.28)
1.20 (0.75-1.90)
Atrial Fibrillation
1.56 (0.79-3.12)
0.91 (0.45-1.86)
Obstructive Sleep Apnea
2.24 (1.30-3.86)
1.97 (1.12-3.28)
Yaggi; NEJM 2005
Risk of Coronary Artery Events or CV Death
Unadjusted Hazard Ratio
(95% CI)
Adjusted Hazard Ratio
(95% CI)
Age ( per yr)
1.07 (1.04-1.09)
1.08 (1.05-1.10)
Tobacco use
1.96 (1.12-3.42)
2.52 (1.43-4.44)
Male gender
1.45 (0.87-2.41)
1.30 (0.77-2.21)
Body Mass Index
1.02 (0.99-1.04)
1.02 (0.99-1.05)
Current Smoker
1.33 (0.86-2.07)
1.92 (1.22-3.02)
Diabetes
1.94 (1.18-3.17)
1.42 (0.85-2.39)
Hyperlipidemia
1.62 (1.05-2.52)
1.46 (0.93-2.30)
Hypertension
1.83 (1.16-2.88)
1.40 (0.87-2.25)
Obstructive Sleep Apnea
2.57 (1.39-4.72)
2.05 (1.10-3.84)
Covariate
Yaggi/Shah; Sleep and Breathing 2009
Patient Prognostic Factors for
Stroke, Coronary Event, Death
Prognostic Factor
Hazard Ratio
95% C.I
Score
Age
<60
60-70
>70
1.00
2.60
5.03
-1.78-3.80
3.45-7.33
1
2
4
Previous CVD
1.75
1.25-2.45
2
Hypertension or DM
1.52
1.02-2.18
2
Smoking
1.54
1.06-2.20
2
Obstructive Sleep Apnea
1.64
1.04-2.42
2
Yaggi; J Clin Sleep Med 2009
Prognostic Groups for
Stroke, Coronary Event, or Death
Risk Group
Outcome
n/N
(%)
Low Risk (1-5)
44/881
5.0
Intermediate Risk (6-7)
90/590
15.3
High Risk (8-12)
57/168
33.9
Yaggi; J Clin Sleep Med 2009
Prognostic Groups for
Stroke, Coronary Event, or Death (cont.)
1.0
0.8
0.6
0.4
0.2
0.0
0
1 20 2 40
3 4 605 80
6
Years
(VA) Validation Cohort
Probability of Event-Free Survival
(Yale) Development Cohort
Kaplan-Meier Estimates for Event-Free Survival
(TIA, Stroke, M.I.,Death)
0 1 202
340 4 60
5 6 80
Years
Yaggi; J Clin Sleep Med 2009
Epidemiologic Studies Linking OSA to
Cardiovascular Outcomes
•
•
•
•
•
•
•
•
Cardiovascular Disease: Shahar, AJRCCM, 2001
Stroke: Arzt, AJRCCM, 2005
Stroke: Munoz, Stroke, 2006
Stroke Mortality: Sahlin, Arch Int Med, 2008
Myocardial Infraction: Peker, Eur Respir J, 2006
Fatal/Non-fatal Cardiovascular Events: Marin, Lancet, 2005
Sudden Death: Gami, NEJM, 2005
All Cause Mortality: Young/Marshall, Sleep, 2008
Outline
• Epidemiologic evidence linking sleep-disordered
breathing to cardiovascular outcomes
• Mechanisms of Stroke in sleep-disordered breathing
Mechanisms of Stroke in
Sleep-Disordered Breathing
1. Intermittent hypoxia
2. Sympathetic activation
3. Mechanical load
4. Snoring and carotid atherosclerosis
5. Impaired sleep and metabolic dysregulation
1. Intermittent Hypoxia
• Repetitive episodes of hypoxia
and reoxygenation
• Activation of vascular
inflammatory pathways
leading to atherosclerosis
• Severity of sleep apnea
correlates with carotid intimal
medial thickness
Ryan; Circulation 2005
Saletu; J Neurobiol 2007
Savransky; AJRCCM 2007
Szaboova; Resp Physiol and Neurobiol 2007
Lavie; Sleep Med Review 2004
2. Sympathetic Activation
Wolk; Circulation 2003
Circadian Blood Pressure and “Nondipping”
Apnea
No Apnea
Somers; J Clin Invest 1995
Ancoli-Israel; Chest 2002
Hla; Sleep 2008
Normal Circadian Variation in Vascular Events
Marler; Stroke 1989
Day-Night Pattern of Sudden Death from Cardiac Causes in 78 Persons with and 34 Persons
without Obstructive Sleep Apnea (OSA) and in the General Population
Day-night Pattern of Sudden Death in
Obstructive Sleep Apnea
Gami, A. et al. N Engl J Med 2005;352:1206-1214
Gami; NEJM 2005
Association of Nocturnal Arrhythmias
with Sleep-disordered Breathing
Arrhythmia Type
Odds ratio
(95% CI)
Nonsustained ventricular tachycardia
3.40
(1.03-11.2)
Complex ventricular ectopy
1.74
(1.11-2.74)
Atrial fibrillation
4.02
(1.03-15.74)
Mehra; AJRCCM 2006
3. Mechanical Load in Sleep Apnea
BASELINE
(Mueller maneuver)
END OF APNEA
Obstructive Apnea
• ↑ Transmural pressure (afterload)
• ↓ Cardiac output
Bradley and Floras (eds); Marcel Decker 2000
Obstructive Apneas and Cerebral Blood flow
Balfors ; AJRCCM 1994
4. Snoring and Carotid Artery Atherosclerosis
• Vibratory stimuli causes
pathologic damage to arterial wall
endothelial cells1,2
• Snores originate in the upper
airway during sleep and result in
vibrations of the pharyngeal wall
and associated structures
• The proximity of the carotid
bifurcation to the lateral
pharyngeal wall exposes it to
these vibrations and may cause
pathologic damage to the arterial
wall endothelium3
1.
2.
3.
Curry; Muscle and Nerve 2002
Puig; Sleep 2005
Hedner; Sleep and Beathing 1994
4. Snoring and Carotid Artery Atherosclerosis
• Cross-sectional study of 110
subjects who underwent PSG
with quantification of snoring,
carotid artery ultrasound
• Simultaneous quantification of
femoral artery atherosclerosis as a
distant (from the upper airway)
control artery
• There was a significant increase
in carotid artery atherosclerosis
(in a dose-response fashion) with
increased snoring, but not femoral
atherosclerosis
Lee; Sleep 2008
4. Snoring and Carotid Artery Atherosclerosis
Covariate
Odds Ratio
C.I.
Age, per decade
3.2
1.4-7.2
Male Sex
4.6
1.4-15.2
Smoking History
3.9
1.1-15.2
Hypertesnion
4.7
1.3-15.7
Snoring, % sleep time
0-25
-
25-50
1.7
0.4-6.9
>50
10.5
2.1-51.8
Lee; Sleep 2008
5. Impaired Sleep and Metabolic Dysregulation
• ↓ Glucose tolerance and ↓ insulin release1
• ↑hunger and appetite and associated changes in
leptin and ghrelin2
• Short sleep duration ↑ risk type 2 diabetes3
• Sleep apnea associated with insulin resistance and
1. Spiegel; Lancet 1999
glucose intolerance4,5
2. Spiegel; Ann Intern Med 2004
3. Yaggi; Diabetes Care 2006
4. Punjabi; AJRCCM 2001
5. IP; AJRCCM 2001
Obstructive Sleep Apnea as a
Risk Factor for Type 2 Diabetes Mellitus
Covariate
Unadjusted HR (95%CI)
Adjusted HR (95% CI)
Age
Gender
0.99 (0.98-1.02)
1.41 (0.44-4.51)
1.00 (0.98-1.02)
1.09 (0.34-3.57)
Race (non-caucasian) 1.35 (0.64-2.85)
Fasting Glucose
1.04 (1.04-1.05)
1.13 (0.56-2.30)
1.05 (1.03-1.06)
BMI
1.06 (1.03-1.09)
1.04 (1.01-1.07)
Change in BMI
Sleep Apnea*
0.73 (0.69-0.78)
1.53 (1.21-94)
0.76 (0.70-0.83)
1.43(1.10-1.86)
* Per Quartile of AHI
Botros; Am J Med 2009
Sleep Apnea and Diabetes:
The Impact of CPAP Treatment
Botros; Am J Med 2009
Metabolic Syndrome
(Syndrome X)
•High Triglycerides
•Low HDL
Sleep Apnea Syndrome
(Syndrome “Zzzz”)
•Central Obesity
•Hypertension
•Insulin Resistance
•Cyclic Hypoxia
•Nocturnal Sympathetic
Activation
•Sleep Loss
Cyclic
Hypoxia
Oxidative stress
Inflammation
Endothelial dsfxn.
Atherosclerosis
Nocturnal
Sympathetic
Activation
Blood pressure surges
Cardiac arrhythmia
∆s Cerebral blood flow
Mechanical
Load
Vibration injury
↑ afterload
↓ cardiac output
Cardiovascular disease
and Stroke
Metabolic
Dysregulation
Glucose Intolerance
Insulin Resistance
Obesity
Diabetes
Outline
• Epidemiologic evidence linking sleep-disordered
breathing to cardiovascular outcomes
• Mechanisms of stroke in sleep-disordered breathing
• Impact of CPAP treatment on cerebrovascular risk
Cardiovascular outcomes in Obstructive Sleep Apnea
With and Without Treatment
Marin; Lancet 2005
Stroke Mortality in Sleep Apnea
With and Without Treatment
Martinez-Garcia; AJRCCM 2009
Impact of CPAP on Blood Pressure
Severity of Sleep Apnea
Effective CPAP Use
Haentjens; Arch Int Med 2007
Inferring Coronary Heart Disease and Stroke
Risk Reduction from Antihypertensive effect of
CPAP
Collins; Lancet 1990
CPAP Treatment and Cerebrovascular Risk
•
•
•
•
•
↓ Blood pressure1
↓ Sympathetic activity and catecholamines2,5
↓ Recurrent atrial fibrillation3
↑ Left ventricular function4
↓ Early Signs of Atherosclerosis5
1.
2.
3.
4.
5.
Haentjens,;Arch Int Med 2007
Faccenda; AJRCCM 2001
Kanagol; Circulation 2003
Kaneko;NEJM 2003
Drager; AJRCCM 2007
Study Design: CPAP Treatment Decreases
Early Signs of Atherosclerosis
No treatment
N=12
Assessed for
Eligibility
(severe OSA)
N=400
Randomized
(without comorbidity)
N=24
Exclusion N=376
Hypertension
Diabetes
Heart Failure
Coronary Artery Disease
Stroke
Smoking
CPAP treatment
N=12
Vascular Parameters
Intima-media thickness
Arterial stiffness
Carotid diameter
24 hour BP
Lipid profile
Catecholamines
C-reactive protein
4 months of follow-up
Drager; AJRCCM 2007
Primary Outcome:
Intima-Media Thickness
Drager; AJRCCM 2007
Secondary Outcome:
Carotid-Femoral Pulse-Wave Velocity
Drager; AJRCCM 2007
Serum and Blood Pressure Measures in
Control and CPAP groups
Control Group
CPAP group
Measures
Baseline
4 mo
Baseline
4 mo
p value
BMI
29.7
29.4
29.9
29.8
NS
Systolic BP
122
121
123
119
NS
Diastolic BP
66
67
73
69
NS
Total Cholesterol
226
224
242
235
NS
LDL
147
145
158
152
NS
HDL
48
47
51
49
NS
C-reactive Protein
3.1
3.3
3.7
2.0
<0.001
Catecholamines
362
357
365
205
<0.001
Drager; AJRCCM 2007
Rationale:Treatment of Sleep Apnea in TIA
• Patients with acute TIA are ideal candidates for
prevention of recurrent vascular events
– 300,000 TIAs occur annually in U.S.1
– No neurologic deficit (< 24 hours)
– High risk of poor outcomes despite prevention strategies:
• 25% will have stroke, cardiovascular event, or death 90 days
post-TIA (half of events occur within first 72 hrs)2
• New approach to reduce recurrent vascular event
rate is needed (particularly in acute post-TIA period)
1. Johnston, NEJM, 2002
2. Johnston, JAMA, 2000
Methods: Overview
Design: Randomized controlled trial, (90days)
Sample: TIA patients (focal neurologic deficit < 24 hours
confirmed by study neurologist)
Setting: Emergency departments/inpatients at 3 CT hospitals
Exclusions: Established OSA, mechanical ventilation, COPD
requiring O2, pregnancy, symptom onset >72hrs, life
Expectancy <6 months, residence outside CT
Intervention: Early/immediate Auto-CPAP (ideally within 1st
or 2nd night of TIA) for 90 days (stopped if no evidence
of sleep apnea/response to flow limitation)
Control: Usual care
Outcomes
• Primary outcomes
– Sleep apnea prevalence (baseline portable study)
– CPAP adherence rates
• No use
• Some use (< 4hr/night or < 70% of nights)
• Good use (≥ 4hr/night and ≥ 70% of nights)
• Secondary Outcome (preliminary data)
– 90-day post-TIA recurrent vascular event rate
• TIA, stroke, myocardial infarction, hospitalization for CHF, or
death
• Intention-to-treat, CPAP use category (pre-specified)
Prevalence of Obstructive Sleep Apnea
Characteristic
Baseline Sleep Study
90-day Sleep Study
OSA prevalence n/N (%)
12/21 (57)
27/45 (59)
AHI: Mean (+/-SD)
11.1 (12)
11.0 (13)
AHI: median, range
6.5, 0-63
6.7, 0-62
Yaggi; Stroke 2010 (In Press)
Time to CPAP or Portable Sleep Study
Time
From TIA symptom onset,
Hours: mean (+/-SD)
Intervention
Control
CPAP
Sleep study
(n=45)
(N=25)
39.4 (23)
45.0 (37)
8 (18)
4 (16)
≥ 24 < 48 hours: n (%)
27 (60)
15 (60)
≥ 48 hours
10 (22)
6 (24)
<24 hours: n (%)
Yaggi; Stroke 2010 (In Press)
Auto-Titrating CPAP Use
CPAP Use Category
Number hrs/night used: mean (+/-SD)
Range
CPAP Use: N (%)
None: 0 hrs/night or 0 nights
Some: <4 hrs/night or <70% nights
Good: ≥4 hrs/night and ≥ 70% nights
Intervention Patients on CPAP
N=29
5.6 (1.9)
1.5-8.5
0 (0)
14 (48)
15 (52)
Yaggi; Stroke 2010 (In Press)
Recurrent Vascular Events:
Overall
(Intention-To-Treat)
(+) Outcome (-) Outcome
Intervention
1
(2%)
44
45
Control
3
(12%)
22
25
P=0.13; 2-sided
Yaggi; Stroke 2010 (In Press)
Recurrent Vascular Events:
Evidence of Sleep Apnea
(CPAP Use Category)
(+) Outcome (-) Outcome
Good
0
(0%)
15
15
Some
1
(7%)
13
14
None
3
(16%)
16
19
P=0.09; 2-sided
Yaggi; Stroke 2010 (In Press)
Summary
•
•
•
Sleep apnea is a valid syndrome of stroke risk.
Epidemiologic studies demonstrate that sleep
apnea is independently associated with
hypertension, stroke, myocardial infarction, fatal
and non-nonfatal cardiovascular events, and
sudden death, all cause mortality
Physiological studies suggest mechanisms
(intermittent hypoxia, sympathetic activation,
changes in cerebral blood flow, mechanical load,
metabolic dyregulation) whereby sleep apnea
provides a substrate for the development of stroke.
Implications
• The high prevalence of sleep apnea confers a
high population attributable risk for
cerebrovascular outcomes
• CPAP treatment is likely result in decreased
cerebrovascular risk.
• Longer-term randomized controlled trials
examining the impact of treatment on
cerebrovascular outcome events are needed.
Acknowledgements
John Concato, M.D.
Vahid Mohsenin, M.D.
Judy Lichtman, Ph.D.
Rachel Lampert, M.D.
Dawn Bravata, M.D. (Indianapolis)
Kingman Strohl, M.D. (Cleveland)
Mark Gorman, M.D. (Vermont)
Zhu Wang, Ph.D.
Nader Botros, M.D., M.P.H.
Neomi Shah, M.D., M.P.H.
Bernardo Selim, M.D.
Frederick Struve, Ph.D
Vincent McClain, M.D.
Yale Center for Sleep Medicine
VA CT Center for Sleep Medicine
VA Clinical Epidemiology Research Center
VA HSR&D Research CDA/Merit Program
VA CSR&D Merit Program