American Journal of Hypertension Advance Access published April 4, 2014
Original Article
Prevalence and Associated Factors of Obstructive Sleep Apnea in Patients with Resistant Hypertension Elizabeth S. Muxfeldt,1 Victor S. Margallo,1 Gleison M. Guimarães,1 and Gil F. Salles1
methods A cross-sectional analysis with 422 resistant hypertensive patients (31.3% men; mean age = 62.4 ± 9.9 years) submitted to a full-night polysomnography. The presence of OSA was defined by an apnea-hypopnea index (AHI) >5 per hour and moderate/severe OSA was defined by an AHI >15. Statistical analysis included bivariable comparisons between patients with and without moderate/severe OSA and logistic regressions to assess the independent correlates of OSA severity.
circumferences, had higher prevalences of diabetes and left ventricular hypertrophy, and had higher proteinuria than patients with no/mild OSA. No difference was found in plasma aldosterone and renin activity. Nighttime systolic blood pressures and pulse pressures were higher in moderate/severe OSA, with lower nocturnal blood pressure fall. In multivariable logistic regression, male sex, older age, diabetes, obesity, increased waist and neck circumferences, and nighttime systolic blood pressure were the independent correlates of moderate/severe OSA.
conclusions Resistant hypertensive patients had a very high prevalence of OSA, and patients with moderate/severe OSA had an adverse ambulatory BP profile, with higher nighttime systolic blood pressures and pulse pressures and higher prevalence of nondipping patterns. Other correlates of OSA severity were mainly demographic-anthropometric variables.
results Three-hundred forty-seven patients (82.2%) had OSA, and 234 patients (55.5%) had moderate/severe OSA. Patients with moderate/severe OSA were more frequently elderly and obese men with larger waist and neck
Keywords: blood pressure; hypertension; obstructive sleep apnea; polysomnography; resistant hypertension.
Obstructive sleep apnea (OSA) is characterized by loud snoring, daytime sleepiness, and witnessed breathing interruptions with at least 5 obstructive respiratory events per hour of sleep.1 Affecting 2%–4% of the adult population,1 OSA is highly prevalent in patients with hypertension (range= 35%–80%) and other cardiovascular diseases,1–3 and this association seems to be proportional to the severity of the disease.2,3 Patients with resistant hypertension (RH), defined as an uncontrolled clinic blood pressure (BP) in spite of using at least 3 antihypertensive medications,4 present high cardiovascular morbidity and mortality.5 The first report on the relationships between OSA and RH was published more than a decade ago.6 Although it seems that OSA contributes to development or persistence of antihypertensive treatment resistance,2 it is not clear whether OSA should be considered a cause,7 a risk factor for RH,4,8 or just 2 associated diseases with similar phenotypes, such as obesity2–4,8–11 and nocturnal nondipping pattern.2,9,10 Moreover, RH and OSA also share common pathogenetic mechanisms, such as volume
overload,4,12–14 sympathetic activation,2,3,15,16 and aldosterone excess.3,4,14,17 Although it is generally accepted that there are strong associations between OSA and RH, this has not been systematically examined, and all previous studies evaluated rather small numbers of patients.6,7,18,19 Therefore, the objective of this study was to evaluate the prevalence of OSA in a large cohort of patients with well-characterized RH and to identify clinic-laboratorial covariables associated with its presence and severity.
Correspondence: Elizabeth Silaid Muxfedlt ([email protected]).
1Hypertension Program, University Hospital Clementino Fraga Filho, School of Medicine, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil.
Initially submitted August 25, 2013; date of first revision October 13, 2013; accepted for publication January 18, 2014.
doi:10.1093/ajh/hpu023
METHODS Patients and baseline procedures
From a total of 451 consecutive patients within a cohort of RH patients, 422 patients agreed to participate in this crosssectional study from September 2010 to January 2013 in the hypertension outpatient clinic of our tertiary-care University
© American Journal of Hypertension, Ltd 2014. All rights reserved. For Permissions, please email: [email protected]
American Journal of Hypertension 1
Downloaded from http://ajh.oxfordjournals.org/ at University of Pennsylvania Library on September 1, 2014
background Obstructive sleep apnea (OSA) syndromes are strongly associated with resistant hypertension, although this has not been systematically examined. The aim of our study was to investigate the prevalence of OSA and its associated factors in a large cohort of resistant hypertensive patients.
Muxfeldt et al.
BP measurements
Clinic BP was measured twice on 2 occasions with suitably sized cuffs, with patients in the sitting position, using a digital BP monitor (HEM-907 XL; Omron Healthcare, Kyoto, Japan). The mean between the 4 readings was registered. The ABPM was recorded using Mobil O graph 2 American Journal of Hypertension
(version 12) equipment (DYNAMAPA, Cardios, São Paulo, Brazil) approved by the British Society of Hypertension.25 All patients were using their prescribed antihypertensive medications, including spironolactone, during ABPM. A reading was taken every 15 minutes throughout the day and every 30 minutes at night. The nighttime period was ascertained for each individual patient from registered diaries. Parameters evaluated were mean 24-hour, daytime, nighttime, and early morning (first 2 hours after awakening) systolic BP (SBP), diastolic BP, and pulse pressure. The nocturnal BP fall was calculated as (((daytime BP – nighttime BP) / daytime BP) × 100), and patients were classified according to the dipping pattern as: dippers (nocturnal reduction ≥10% and 41 cm in women or if >102 cm and >43 cm in men, respectively.2,21 The laboratory evaluation included fasting glycemia, serum creatinine and lipids, plasma aldosterone and renin activity, and albuminuria and proteinuria in a sterile 24-hour urine collection. The use of mineralocorticoid receptor blocker (spironolactone) was interrupted for at least 4 weeks before aldosterone and renin assays. Patients with high aldosterone-to-renin ratio (>30, calculated with renin activities 125 g/ m2 in men and >110 g/m2 in women.23 Carotid–femoral pulse wave velocity was measured in the morning just after the ABPM with the Complior equipment (Artech-Medical, Paris, France).24 Three consecutive measurements were performed, and the mean value was used. Direct carotid– femoral distance was corrected by a factor of 0.8, as recently recommended.24
Sleep Apnea in Resistant Hypertension RESULTS
A total of 422 patients (31% men; mean age = 62.4 ± 9.9 years) were enrolled. PSG identified OSA in 347 patients (82.2%; 95% CI = 74.0%–91.3%) and moderate or severe OSA in 234 patients (55.5%; 95% CI = 48.8%–63.0%). Thirteen patients (2.8%) had central sleep apnea. Tables 1 and 2 outline the baseline characteristics of all patients and those with and without moderate/severe OSA. Moderate/severe OSA was more frequent in male, elderly, obese (with increased waist and neck circumferences), and diabetic patients. Antihypertensive treatment was similar in number of drugs and drugs taken at bedtime, although patients with moderate/severe OSA were using more calcium channel blockers and less centrally acting alpha-agonists than those with no/mild OSA (Table 1). Patients with moderate/severe OSA also presented higher fasting glycemia and proteinuria and a higher prevalence of left ventricular hypertrophy. No difference was found in relation to renal function, serum lipids, or plasma aldosterone and renin activity. Aortic arterial stiffness was also similar in the two groups (Table 1). Furthermore, the nocturnal BP profile was worse in patients with moderate/severe OSA. They presented higher nighttime systolic BP and pulse pressure and lower nocturnal systolic BP fall, with a higher prevalence of nondipping patterns (both the riser and the nondipper pattern) than patients with no/mild OSA (Table 2). Figures 1 and 2 present the results of multivariable logistic regression analyses for the independent covariables associated with the presence of OSA, moderate/severe OSA, and severe OSA in all patients and in patients with uncontrolled ambulatory BPs, respectively. Older age, male sex, obesity, increased waist and neck circumferences, higher nighttime
Table 1. Baseline characteristics of patients grouped according to obstructive sleep apnea diagnosed by polysomnography
Characteristics
All patients
No/mild OSA
Moderate/severe OSA
(n = 422)
(n = 188)
(n = 234)
Clinical and demographic data Sex, % male Age, years
31.3 62.4 (9.9)
20.7 61.2 (10.4)
38.9* 63.4 (9.4)***
Age >65 years
40.0
34.1
44.7***
Snoring
70.5
64.2
75.5***
Daytime sleepiness
22.9
21.5
24.1
Weight, kg
79.8 (16.5)
74.4 (13.8)
84.2 (17.3)*
BMI, kg/m2
31.2 (5.7)
29.5 (5.2)
32.5 (5.7)*
Obesity, BMI ≥30 kg/m2 Waist circumference, cm Increased waist circumferencea Neck circumference, cm Increased neck circumferenceb
52.7 101.9 (12.1) 76.1 37.8 (3.7)
41.8 97.5 (11.3) 70.0 36.3 (3.1)
61.6* 105.5 (11.6)* 81.0** 39.1 (3.6)*
13.1
6.2
18.5**
Diabetes
36.3
30.9
40.6***
Physical inactivity
66.8
66.0
67.5
Cardiovascular risk factors
(Continued)
American Journal of Hypertension 3
Downloaded from http://ajh.oxfordjournals.org/ at University of Pennsylvania Library on September 1, 2014
variables), Mann–Whitney test (for continuous asymmetric variables), and χ2 test (for categorical variables). To assess the independent correlates of each severity grade of OSA, multivariable logistic regressions were performed with total OSA, moderate/severe OSA, and severe OSA as the dependent variables. Candidate variables to enter the models were age, sex, obesity, increased neck and waist circumferences, presence of diabetes and cardiovascular diseases, nighttime SBP, nondipping pattern (nondippers and risers), snoring, left ventricular hypertrophy, and 24-hour proteinuria and albuminuria. A stepwise forward procedure was used to select the independent covariables (P 10 m/s
36.7 9.5 (2.1) 21.3
7.3 114 (65–196) 17 (7–52) 35.3 9.4 (2.2) 21.0
5.1 153 (87–273)** 19 (10–54) 37.8 9.7 (2.0) 21.5
2-Dimensional echocardiogram LV mass index, g/m2 LVH
139 (49) 71.5
133 (39) 66.5
144 (55)*** 75.5***
Values are presented as percentages or as means (SD), except for number of antihypertensive drugs, serum creatinine, plasma aldosterone and renin activity, proteinuria, and albuminuria, which are presented as medians (interquartile range). Abbreviations: ACE, angiotensin-converting enzyme; AR, angiotensin II receptor; BMI, body mass index; HDL, high-density lipoprotein; IQR, interquartile range; LV, left ventricular; LVH, left ventricular hypertrophy; OSA, obstructive sleep apnea. a Increased waist circumference: >88 cm in women and >102 cm in men. b Increased neck circumference: >41 cm in women and >43 cm in men. c Creatinine clearance calculated by Cockroft Gault formula. d To calculate aldosterone-to-renin ratio, plasma renin activity
Original Article
Prevalence and Associated Factors of Obstructive Sleep Apnea in Patients with Resistant Hypertension Elizabeth S. Muxfeldt,1 Victor S. Margallo,1 Gleison M. Guimarães,1 and Gil F. Salles1
methods A cross-sectional analysis with 422 resistant hypertensive patients (31.3% men; mean age = 62.4 ± 9.9 years) submitted to a full-night polysomnography. The presence of OSA was defined by an apnea-hypopnea index (AHI) >5 per hour and moderate/severe OSA was defined by an AHI >15. Statistical analysis included bivariable comparisons between patients with and without moderate/severe OSA and logistic regressions to assess the independent correlates of OSA severity.
circumferences, had higher prevalences of diabetes and left ventricular hypertrophy, and had higher proteinuria than patients with no/mild OSA. No difference was found in plasma aldosterone and renin activity. Nighttime systolic blood pressures and pulse pressures were higher in moderate/severe OSA, with lower nocturnal blood pressure fall. In multivariable logistic regression, male sex, older age, diabetes, obesity, increased waist and neck circumferences, and nighttime systolic blood pressure were the independent correlates of moderate/severe OSA.
conclusions Resistant hypertensive patients had a very high prevalence of OSA, and patients with moderate/severe OSA had an adverse ambulatory BP profile, with higher nighttime systolic blood pressures and pulse pressures and higher prevalence of nondipping patterns. Other correlates of OSA severity were mainly demographic-anthropometric variables.
results Three-hundred forty-seven patients (82.2%) had OSA, and 234 patients (55.5%) had moderate/severe OSA. Patients with moderate/severe OSA were more frequently elderly and obese men with larger waist and neck
Keywords: blood pressure; hypertension; obstructive sleep apnea; polysomnography; resistant hypertension.
Obstructive sleep apnea (OSA) is characterized by loud snoring, daytime sleepiness, and witnessed breathing interruptions with at least 5 obstructive respiratory events per hour of sleep.1 Affecting 2%–4% of the adult population,1 OSA is highly prevalent in patients with hypertension (range= 35%–80%) and other cardiovascular diseases,1–3 and this association seems to be proportional to the severity of the disease.2,3 Patients with resistant hypertension (RH), defined as an uncontrolled clinic blood pressure (BP) in spite of using at least 3 antihypertensive medications,4 present high cardiovascular morbidity and mortality.5 The first report on the relationships between OSA and RH was published more than a decade ago.6 Although it seems that OSA contributes to development or persistence of antihypertensive treatment resistance,2 it is not clear whether OSA should be considered a cause,7 a risk factor for RH,4,8 or just 2 associated diseases with similar phenotypes, such as obesity2–4,8–11 and nocturnal nondipping pattern.2,9,10 Moreover, RH and OSA also share common pathogenetic mechanisms, such as volume
overload,4,12–14 sympathetic activation,2,3,15,16 and aldosterone excess.3,4,14,17 Although it is generally accepted that there are strong associations between OSA and RH, this has not been systematically examined, and all previous studies evaluated rather small numbers of patients.6,7,18,19 Therefore, the objective of this study was to evaluate the prevalence of OSA in a large cohort of patients with well-characterized RH and to identify clinic-laboratorial covariables associated with its presence and severity.
Correspondence: Elizabeth Silaid Muxfedlt ([email protected]).
1Hypertension Program, University Hospital Clementino Fraga Filho, School of Medicine, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil.
Initially submitted August 25, 2013; date of first revision October 13, 2013; accepted for publication January 18, 2014.
doi:10.1093/ajh/hpu023
METHODS Patients and baseline procedures
From a total of 451 consecutive patients within a cohort of RH patients, 422 patients agreed to participate in this crosssectional study from September 2010 to January 2013 in the hypertension outpatient clinic of our tertiary-care University
© American Journal of Hypertension, Ltd 2014. All rights reserved. For Permissions, please email: [email protected]
American Journal of Hypertension 1
Downloaded from http://ajh.oxfordjournals.org/ at University of Pennsylvania Library on September 1, 2014
background Obstructive sleep apnea (OSA) syndromes are strongly associated with resistant hypertension, although this has not been systematically examined. The aim of our study was to investigate the prevalence of OSA and its associated factors in a large cohort of resistant hypertensive patients.
Muxfeldt et al.
BP measurements
Clinic BP was measured twice on 2 occasions with suitably sized cuffs, with patients in the sitting position, using a digital BP monitor (HEM-907 XL; Omron Healthcare, Kyoto, Japan). The mean between the 4 readings was registered. The ABPM was recorded using Mobil O graph 2 American Journal of Hypertension
(version 12) equipment (DYNAMAPA, Cardios, São Paulo, Brazil) approved by the British Society of Hypertension.25 All patients were using their prescribed antihypertensive medications, including spironolactone, during ABPM. A reading was taken every 15 minutes throughout the day and every 30 minutes at night. The nighttime period was ascertained for each individual patient from registered diaries. Parameters evaluated were mean 24-hour, daytime, nighttime, and early morning (first 2 hours after awakening) systolic BP (SBP), diastolic BP, and pulse pressure. The nocturnal BP fall was calculated as (((daytime BP – nighttime BP) / daytime BP) × 100), and patients were classified according to the dipping pattern as: dippers (nocturnal reduction ≥10% and 41 cm in women or if >102 cm and >43 cm in men, respectively.2,21 The laboratory evaluation included fasting glycemia, serum creatinine and lipids, plasma aldosterone and renin activity, and albuminuria and proteinuria in a sterile 24-hour urine collection. The use of mineralocorticoid receptor blocker (spironolactone) was interrupted for at least 4 weeks before aldosterone and renin assays. Patients with high aldosterone-to-renin ratio (>30, calculated with renin activities 125 g/ m2 in men and >110 g/m2 in women.23 Carotid–femoral pulse wave velocity was measured in the morning just after the ABPM with the Complior equipment (Artech-Medical, Paris, France).24 Three consecutive measurements were performed, and the mean value was used. Direct carotid– femoral distance was corrected by a factor of 0.8, as recently recommended.24
Sleep Apnea in Resistant Hypertension RESULTS
A total of 422 patients (31% men; mean age = 62.4 ± 9.9 years) were enrolled. PSG identified OSA in 347 patients (82.2%; 95% CI = 74.0%–91.3%) and moderate or severe OSA in 234 patients (55.5%; 95% CI = 48.8%–63.0%). Thirteen patients (2.8%) had central sleep apnea. Tables 1 and 2 outline the baseline characteristics of all patients and those with and without moderate/severe OSA. Moderate/severe OSA was more frequent in male, elderly, obese (with increased waist and neck circumferences), and diabetic patients. Antihypertensive treatment was similar in number of drugs and drugs taken at bedtime, although patients with moderate/severe OSA were using more calcium channel blockers and less centrally acting alpha-agonists than those with no/mild OSA (Table 1). Patients with moderate/severe OSA also presented higher fasting glycemia and proteinuria and a higher prevalence of left ventricular hypertrophy. No difference was found in relation to renal function, serum lipids, or plasma aldosterone and renin activity. Aortic arterial stiffness was also similar in the two groups (Table 1). Furthermore, the nocturnal BP profile was worse in patients with moderate/severe OSA. They presented higher nighttime systolic BP and pulse pressure and lower nocturnal systolic BP fall, with a higher prevalence of nondipping patterns (both the riser and the nondipper pattern) than patients with no/mild OSA (Table 2). Figures 1 and 2 present the results of multivariable logistic regression analyses for the independent covariables associated with the presence of OSA, moderate/severe OSA, and severe OSA in all patients and in patients with uncontrolled ambulatory BPs, respectively. Older age, male sex, obesity, increased waist and neck circumferences, higher nighttime
Table 1. Baseline characteristics of patients grouped according to obstructive sleep apnea diagnosed by polysomnography
Characteristics
All patients
No/mild OSA
Moderate/severe OSA
(n = 422)
(n = 188)
(n = 234)
Clinical and demographic data Sex, % male Age, years
31.3 62.4 (9.9)
20.7 61.2 (10.4)
38.9* 63.4 (9.4)***
Age >65 years
40.0
34.1
44.7***
Snoring
70.5
64.2
75.5***
Daytime sleepiness
22.9
21.5
24.1
Weight, kg
79.8 (16.5)
74.4 (13.8)
84.2 (17.3)*
BMI, kg/m2
31.2 (5.7)
29.5 (5.2)
32.5 (5.7)*
Obesity, BMI ≥30 kg/m2 Waist circumference, cm Increased waist circumferencea Neck circumference, cm Increased neck circumferenceb
52.7 101.9 (12.1) 76.1 37.8 (3.7)
41.8 97.5 (11.3) 70.0 36.3 (3.1)
61.6* 105.5 (11.6)* 81.0** 39.1 (3.6)*
13.1
6.2
18.5**
Diabetes
36.3
30.9
40.6***
Physical inactivity
66.8
66.0
67.5
Cardiovascular risk factors
(Continued)
American Journal of Hypertension 3
Downloaded from http://ajh.oxfordjournals.org/ at University of Pennsylvania Library on September 1, 2014
variables), Mann–Whitney test (for continuous asymmetric variables), and χ2 test (for categorical variables). To assess the independent correlates of each severity grade of OSA, multivariable logistic regressions were performed with total OSA, moderate/severe OSA, and severe OSA as the dependent variables. Candidate variables to enter the models were age, sex, obesity, increased neck and waist circumferences, presence of diabetes and cardiovascular diseases, nighttime SBP, nondipping pattern (nondippers and risers), snoring, left ventricular hypertrophy, and 24-hour proteinuria and albuminuria. A stepwise forward procedure was used to select the independent covariables (P 10 m/s
36.7 9.5 (2.1) 21.3
7.3 114 (65–196) 17 (7–52) 35.3 9.4 (2.2) 21.0
5.1 153 (87–273)** 19 (10–54) 37.8 9.7 (2.0) 21.5
2-Dimensional echocardiogram LV mass index, g/m2 LVH
139 (49) 71.5
133 (39) 66.5
144 (55)*** 75.5***
Values are presented as percentages or as means (SD), except for number of antihypertensive drugs, serum creatinine, plasma aldosterone and renin activity, proteinuria, and albuminuria, which are presented as medians (interquartile range). Abbreviations: ACE, angiotensin-converting enzyme; AR, angiotensin II receptor; BMI, body mass index; HDL, high-density lipoprotein; IQR, interquartile range; LV, left ventricular; LVH, left ventricular hypertrophy; OSA, obstructive sleep apnea. a Increased waist circumference: >88 cm in women and >102 cm in men. b Increased neck circumference: >41 cm in women and >43 cm in men. c Creatinine clearance calculated by Cockroft Gault formula. d To calculate aldosterone-to-renin ratio, plasma renin activity
