Predictors of Obstructive Sleep Apnea in Consecutive Patients with Metabolic Syndrome

Abstract

Background:

Recent evidence suggests that obstructive sleep apnea (OSA) is common in patients with metabolic syndrome (MetS) and may contribute to metabolic deregulation, inflammation, and atherosclerosis in these patients. In clinical practice, however, OSA is frequently underdiagnosed. We sought to investigate the clinical predictors of OSA in patients with MetS.

Methods:

We studied consecutive patients newly diagnosed with MetS (Adult Treatment Panel-III). All participants underwent clinical evaluation, standard polysomnography, and laboratory measurements. We performed a logistic regression model, including the following variables: gender, age >50 years, neck and waist circumferences, hypertension, diabetes, body mass index (BMI) >30 kg/m², high risk for OSA by Berlin questionnaire, presence of excessive daytime somnolence (Epworth Sleepiness Scale), abnormal serum glucose, triglycerides, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol.

Results:

We studied 197 patients (60% men; age: 49 ± 10 years; BMI: 32.9 ± 5.1 kg/m²). OSA (defined by an apnea–hypopnea index ≥15 events per hour) was diagnosed in 117 patients [59%; 95% confidence interval (CI): 52–66]. In multivariate analysis, male gender [odds ratio (OR): 3.28; 95% CI: 1.68–6.41; P < 0.01], abnormal glucose levels (OR: 3.01; 95% CI: 1.50–6.03; P < 0.01), excessive daytime sleepiness (OR: 2.38; 95% CI: 1.13–5.04; P = 0.02), and high risk for OSA by Berlin questionnaire (OR: 4.33; 95% CI: 2.06–9.11; P < 0.001) were independently associated with OSA.

Conclusions:

Simple clinical and metabolic characteristics may help to improve the underdiagnosis of OSA in patients with MetS.

Introduction

Metabolic syndrome (MetS) is associated with higher cardiovascular risk than one might expect from simple addition of its individual components.^1,2 However, other factors may contribute to the high cardiovascular burden observed in patients with MetS, such as obstructive sleep apnea (OSA). This sleep disordered breathing is a chronic condition characterized by upper airway obstruction during sleep, resulting in recurrent hypoxia, hypercapnia, and arousals.³ OSA is highly prevalent in the adult population, reaching one in three adult individuals.⁴ Recent evidence suggested that OSA is very common (around 60%) and underdiagnosed in consecutive patients with the MetS.⁵ More importantly, OSA is associated with metabolic dysregulation, systemic inflammation, and markers of atherosclerosis in these patients.^5,6 Based on the potential impact of OSA in patients with MetS and the overall low awareness of OSA in the clinical management of MetS, the present study was aimed to evaluate the clinical characteristics and predictors of OSA among patients with MetS.

Materials and Methods

We studied consecutive patients with a recent diagnosis of MetS recruited from the Heart Institute (InCor), University of São Paulo Medical School for 2 years. The vast majority of these patients were included in a previous investigation addressing the frequency, metabolic, and inflammatory parameters in MetS patients.⁵ All participants were recruited from routine checkup evaluations. No sleep questionnaire was applied at the time of the recruitment. Patients with established cerebrovascular disease, coronary disease, heart failure, rheumatologic diseases, renal failure, smokers, pregnancy, and regular exercisers were excluded, as well as patients with a previous diagnosis of OSA. The local ethics committee (Institutional Review Board–Heart Institute) approved the protocol, and all participants gave written informed consent.

All participants underwent a detailed history and physical examination as previously described.⁵ Fasting blood samples were drawn for determination of glucose, total cholesterol, low-density lipoprotein (LDL), high-density lipoprotein (HDL), and triglycerides.

Definition of MetS

MetS was diagnosed according to the National Cholesterol Education Program, Adult Treatment Panel III (NCEP ATP III),² if at least three of the five following factors were present: (1) waist circumference (≥102 in men and ≥88 cm in women); (2) triglycerides ≥150 mg/dL, or patient on specific drug treatment; (3) HDL <40 in men and <50 mg/dL in women, or when on specific drug treatment; (4) arterial blood pressure ≥130 or 85 mmHg for systolic and diastolic blood pressure, respectively, or patient on antihypertensive drug treatment; and (5) fasting glucose ≥100 mg/dL or patient on specific drug treatment.

Sleep parameters

After a confirmation of MetS diagnosis, all participants (regardless of sleep complaints) underwent a standard overnight polysomnography as previously described.⁵ Because of a high expected prevalence of OSA in this population and the lack of evidence that mild OSA has cardiovascular consequences as confirmed by a recent international statement,⁷ the presence of OSA was restricted to the moderate-to-severe cases, that is, apnea–hypopnea index ≥15 events per hour of sleep as previously described.⁵ Daytime somnolence was evaluated by the Epworth sleepiness scale.⁸ In brief, this scale was used to assess the general level of daytime sleepiness by having patients rate the likelihood of dozing during eight different daytime situations. The scale ranges from 0 to 24, and scores >10 were considered associated with excessive daytime sleepiness. Prediction of a high and low risk of OSA using the Berlin Questionnaire⁹ was determined on the basis of the responses in three symptom categories. In category 1, high risk was defined as persistent symptoms (>3 to 4 times/week) for ≥2 questions about snoring. In category 2, high risk was defined as persistent (>3 to 4 times/week) daytime tiredness or fatigue. In category 3, high risk was defined as a history of high blood pressure or a body mass index (BMI) >30 kg/m². To be considered at high risk of OSA, a patient had to qualify for ≥2 symptom categories. Those who denied having persistent symptoms or who qualified for only one symptom category were placed in the lower risk group.

Statistical analyses

Data were analyzed with SPSS 21.0 statistical software. Normality distribution was evaluated with the Kolmogorov–Smirnov test, and the results are expressed as mean ± standard deviation, median (interquartile range), or percentage when appropriate. A two-tailed unpaired t test or Mann–Whitney U test was used for independent samples; Chi-square test was used to compare variables between patients without and with OSA when appropriate. Univariate and multivariable logistic regression models were used to evaluate the presence of OSA predictors, including age >50 years, male gender, neck circumference related to OSA risk (>41 and >43 cm for women and men, respectively), waist circumference MetS criterion (>88 and >102 cm for women and men, respectively), high risk for OSA by Berlin questionnaire, excessive daytime sleepiness by Epworth questionnaire, hypertension, BMI >30 kg/m², elevated triglycerides, low HDL, and elevated glucose levels (the last three derived from MetS criterion). A two-sided P value <0.05 was considered significant.

Results

We initially enrolled 300 consecutive patients with confirmed MetS diagnosis. The final sample comprised 197 patients, because of refusals (n = 25), sleep study failure (n = 10), previous exclusion conditions stated on methods (n = 48), and previous OSA diagnosis (n = 20). Overall, the studied sample comprised middle-aged and obese patients, with a slight majority of men (Table 1). OSA was diagnosed in 117 (59%; 95% CI: 52–66) patients. The demographic, anthropometric, laboratorial, and sleep characteristics of the total population, as well as comparisons of patients with or without OSA, are shown in Table 1. Compared to participants without OSA, those with OSA were older, predominantly male, with increased neck and waist circumferences, and with increased levels of Epworth sleepiness scale (and consequently the percentage of daytime somnolence), triglycerides, and fasting plasma glucose levels. There were no significant differences in race, office blood pressure, BMI, hypertension, and diabetes status between patients with and without OSA.

Table 1.

Patient Characteristics

	Total (n = 197)	No. of OSA (n = 80)	OSA (n = 117)	P^a
Age (years)	49 ± 10	47 ± 10	50 ± 9	0.03
Males, n (%)	118 (60)	37 (46)	81 (69)	0.001
Caucasians, n (%)	150 (76)	60 (75)	90 (77)	0.76
BMI (kg/m²)	32.9 ± 5.1	32.2 ± 4.0	33.5 ± 5.7	0.09
Hypertension, n (%)	161 (82)	68 (85)	93 (79)	0.49
Diabetes, n (%)	30 (15)	10 (13)	20 (17)	0.38
Waist circumference (cm)	107 ± 10	105 ± 10	109 ± 11	0.02
Neck circumference (cm)	40 ± 4	39 ± 4	42 ± 4	<0.001
AHI (events per hour)	20.3 (7.5–43.1)	6.1 (2.5–9.8)	36.1 (24.3–62.2)	<0.001
Lowest O₂ saturation, n (%)	83 (75–87)	87 (83–90)	79 (70–85)	<0.001
Total sleep time O₂ saturation <90%, n (%)	2.5 (0.2–12.3)	0.3 (0–2.2)	9.5 (2.0–23.4)	<0.001
Epworth sleepiness scale	10 (7–12)	9 (6–10)	10 (7–13)	0.02
Excessive daytime sleepiness,^b n (%)	68 (35)	19 (23)	49 (42)	0.01
Triglycerides (mg/dL)	195 ± 96	175 ± 81	208 ± 103	0.002
HDL cholesterol (mg/dL)	38 (34–46)	40 (34–46)	38 (34–47)	0.59
Systolic blood pressure (mmHg)	140 ± 24	139 ± 23	141 ± 24	0.66
Diastolic blood pressure (mmHg)	86 ± 14	86 ± 15	86 ± 14	0.99
Fasting glucose (mg/dL)	105 ± 22	99 ± 15	109 ± 25	0.001
Diuretics, n (%)	72 (37)	28 (35)	44 (39)	0.61
Oral antidiabetic drugs, n (%)	24 (12)	7 (9)	174 (15)	0.22
ACE inhibitors, n (%)	58 (30)	18 (23)	40 (35)	0.06
AT1 inhibitors, n (%)	16 (8.2)	6 (8)	10 (9)	0.75
B-blockers, n (%)	39 (20)	19 (24)	20 (18)	0.28
Calcium channel blockers, n (%)	32 (17)	10 (13)	22 (19)	0.21
Statins, n (%)	17 (9)	6 (8)	11 (9)	0.64
Fibrates, n (%)	3 (2)	0 (0)	3 (3)	0.15

For comparisons between MetS patients with and without OSA.

Epworth sleepiness scale score >10. Variables with normal distribution are expressed as mean ± standard deviation. Variables with skewed distribution are presented as median (interquartile range).

ACE, angiotensin converting enzyme; BMI, body mass index; HDL, high-density lipoprotein; MetS, metabolic syndrome; OSA, obstructive sleep apnea.

In the univariate analysis, we found that the presence of OSA was associated with male gender, higher neck circumference, diabetes mellitus, elevated triglycerides, and glucose. In addition, high risk for OSA in Berlin questionnaire and excessive daytime somnolence evaluated by Epworth Sleepiness Scale were also associated with OSA (Table 2). In the multivariate analysis, male gender, abnormal glucose, excessive daytime sleepiness, and high risk for OSA in Berlin questionnaire were independently associated with OSA (Table 2).

Table 2.

Logistic Regression Between Clinical Predictors and Obstructive Sleep Apnea Diagnosis

	Univariate		Multivariate
Variable	OR	P	OR	CI (95%)	P
Male	3.58	<0.001	3.28	1.68–6.41	0.001
Age >50 years	1.41	0.25
BMI >30 kg/m²	1.15	0.40
Neck circumference	2.96	<0.01
Waist circumference	1.28	0.62
Hypertension	1.16	0.66
Diabetes mellitus	2.51	0.05
Elevated triglycerides	1.94	0.03
Low HDL	1.54	0.19
Elevated glucose	2.89	<0.01	3.01	1.50–6.03	0.002
Excessive daytime somnolence	2.44	<0.01	2.38	1.13–5.04	0.02
High OSA risk (Berlin questionnaire)	4.99	<0.001	4.33	2.06–9.11	<0.001

CI, confidence interval; OR, odds ratio.

Discussion

This study evaluated predictors of OSA among patients with MetS using simple tools that may be applied in the clinical management of patients with MetS. This study confirms that OSA is extremely common among patients with MetS. The following novel findings were observed: (1) While patients with MetS and OSA presented some of the classical OSA phenotypes, including older men, increased frequency of visceral obesity, and daytime sleepiness, we did not find any differences in BMI and hypertension frequency between MetS according to OSA status; and (2) Male gender, abnormal glucose levels (derived from MetS criterion), excessive daytime sleepiness, and high risk for OSA in the Berlin questionnaire were independent predictors of OSA.

OSA is a common condition among patients with cardiovascular and endocrine diseases. For instance, we previously demonstrated a high frequency of OSA in patients with hypertension,¹⁰ resistant hypertension,¹¹ as well as with diabetes mellitus.¹² The high frequency of OSA observed in the present investigation and in the above mentioned evidence is partially justified by the coexistence of common risk factors for OSA and cardiovascular or endocrinology diseases, such as older age, male gender, and obesity. It is well established that fat accumulation in upper airways, as well as under the mandible and in the tongue, soft palate, and uvula, contributes to pharyngeal airway narrowing.^1,13,14 Obesity also contributes indirectly to upper airway narrowing during sleep due to reduced lung volumes generated by a combination of increased abdominal fat mass and recumbent posture.¹⁵ Reduction of lung volume may decrease longitudinal tracheal traction forces and pharyngeal wall tension, which predisposes to narrowing of the airway.³ Interestingly, in the MetS scenario, neither BMI nor neck/waist circumference were useful predictors for selecting OSA patients. These findings underscore the idea that the relative role of traditional risk factors for OSA is dependent of the particular characteristics of the studied population.

Recent investigations addressed the predictors of OSA in patients with MetS. In contrast to our findings, Rogers et al. showed that obesity was the strongest predictor of OSA in 1035 blacks with MetS.¹⁶ However, this study evaluated the presence of OSA using the Apnea Risk Evaluation System (ARES) questionnaire.¹⁶ In another study restricted to male subjects, OSA patients exhibited elevated fasting glucose levels in comparison to age-matched controls.¹⁷ Consistently to our findings, increased serum glucose was significantly associated with the presence of OSA. Hyperglycemia was the only component of MetS that showed an association with OSA independent of BMI.¹⁷ However, this study was primarily designed to examine the prevalence of MetS and its components among OSA subjects reporting symptoms indicating sleep-disordered breathing. Therefore, our study is the first investigation to evaluate the predictors of OSA in a consecutive population of MetS with full polysomnography, considered the gold standard method for diagnosing OSA.

This study has some strengths and limitations to be addressed. The strengths include, beyond the use of polysomnography, the recruitment of consecutive patients with well characterized MetS not referred to a sleep laboratory. Moreover, we explored simple tools that can be used in the clinical practice to better discriminate OSA patients for sleep studies. The following limitations deserve comments: First, our patients were middle-aged and without a history of conditions like coronary disease and stroke. Hence, our results may not be applicable to other age groups or patients with established cardiovascular diseases. Second, the study design may not allow us to define if these predictors are cost effective for improving prognosis associated with MetS and OSA treatment. Future studies are warranted.

In conclusion, we have shown that the following clinical characteristics predict significant OSA in patients with MetS: male gender, abnormal glucose, excessive daytime somnolence, and high risk of OSA by Berlin questionnaire. These simple tools may improve the underdiagnosis of OSA in these patients.

Footnotes

Acknowledgment

Fundação de Amparo à Pesquisa do Estado de São Paulo (grant no. 2012/02953-2).

Author Disclosure Statement

No conflicting financial interests exist.

References

Gami

, Witt

, Howard

, et al. Metabolic syndrome and risk of incident cardiovascular events and death: A systematic review and meta-analysis of longitudinal studies. J Am Coll Cardiol, 2007; 49:403–414.

Grundy

, Cleeman

, Daniels

, et al. Diagnosis and management of the metabolic syndrome: An American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation, 2005; 112:2735–2752. Erratum in: Circulation 2005;112:e297.

Drager

, Togeiro

, Polotsky

, et al. Obstructive sleep apnea: A cardiometabolic risk in obesity and the metabolic syndrome. J Am Coll Cardiol, 2013; 62:569–576.

Heinzer

, Vat

, Marques-Vidal

, et al. Prevalence of sleep-disordered breathing in the general population: The HypnoLaus study. Lancet Respir Med, 2015; 3:310–318.

Drager

, Lopes

, Maki-Nunes

, et al. The impact of obstructive sleep apnea on metabolic and inflammatory markers in consecutive patients with metabolic syndrome. PLoS One, 2010; 5:e12065.

Drager

, Bortolotto

, Maki-Nunes

, et al. The incremental role of obstructive sleep apnoea on markers of atherosclerosis in patients with metabolic syndrome. Atherosclerosis, 2010; 208:490–495.

Chowdhuri

, Quan

, Almeida

, et al. An Official American Thoracic Society Research Statement: Impact of mild obstructive sleep apnea in adults. Am J Respir Crit Care Med, 2016; 193:e37–e54.

Johns

. A new method for measuring daytime sleepiness: The Epworth sleepiness scale. Sleep, 1991; 14:540–545.

Netzer

, Stoohs

, Netzer

, et al. Using the BerlinQuestionnaire to identify patients at risk for the sleep apnea syndrome. Ann Intern Med, 1999; 131:485–491.

10.

Pedrosa

, Drager

, Gonzaga

, et al. Obstructive sleep apnea: The most common secondary cause of hypertension associated with resistant hypertension. Hypertension, 2011; 58:811–817.

11.

Drager

, Genta

, Pedrosa

, et al. Characteristics and predictors of obstructive sleep apnea in patients with systemic hypertension. Am J Cardiol, 2010; 105:1135–1139.

12.

Souza

, Medeiros

, Carvalho

, et al. Characteristics and predictors of obstructive sleep apnoea in patients with type 2 diabetes. Diabetes Metab, 2017; 17:30036–30068.

13.

Horner

, Shea

, McIvor

, et al. Pharyngeal size and shape during wakefulness and sleep in patients with obstructive sleep apnoea. Q J Med, 1989; 72:719–735.

14.

Shelton

, Woodson

, Gay

, et al. Pharyngeal fat in obstructive sleep apnea. Am Rev Respir Dis, 1993; 148:462–466.

15.

Isono

. Obesity and obstructive sleep apnoea: Mechanisms for increased collapsibility of the passive pharyngeal airway. Respirology, 2012; 17:32–42.

16.

Rogers

, Ravenell

, Donat

, et al. Predictors of obstructive sleep apnea risk among blacks with metabolic syndrome. J Obes Overweight, 2015; 1:1104.

17.

Papanas

, Steiropoulos

, Nena

, et al. Predictors of obstructive sleep apnea in males with metabolic syndrome. Vasc Health Risk Manag, 2010; 6:281–286.