Risk of obstructive sleep apnea and hemodialysis efficacy

1Lung Diseases Research Center, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran 2 Department of Medical-Surgical Nursing, School of Nursing and Midwifery, Mashhad University of Medical Sciences, Mashhad, Iran 3Department of Nephrology, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran 4Clinical Research Unit, Mashhad University of Medical Sciences, Mashhad, Iran 5Department of Internal Medicine, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran


Background
End-stage renal disease (ESRD) is a worldwide health problem with increasing prevalence and incidence globally and also in Iranian population.Nearly 47.7% of Iranian ESRD patients undergo hemodialysis (HD ) as their renal replacement therapy with variable efficacy according to clinical and laboratory parameters.HD is accompanied by many health problems (1,2).Sleep disorder as a prevalent health problem can adversely affect the quality of life in ESRD patients (3,4).Obstructive sleep apnea (OSA) is defined as repetitive obstructive respiratory pauses (each lasting at least 10 seconds) during sleep.OSA is a common respiratory sleep disorder in HD patients affecting 30%-80% of this population compared to 3%-7% in general population (3,5).In addition to impaired sleep quality and decreased daytime function, OSA can induce important morbidities (like hypertension, coronary artery disease, and stroke) and higher mortality in HD and other patients (6)(7)(8)(9)(10).

Objectives
We conducted this study to evaluate the prevalence of OSA in our HD patients and to assess the variables correlated with risk of OSA and long term mortality.

Patients and Methods
This was a prospective cross-sectional study.ESRD patients with at least 3 months (thrice-weekly) maintenance HD referred to Imam Reza university hospital, Mashhad, Iran were included in the study between April 2013 and April 2014.Those with a history of concurrent peritoneal dialysis, serious neurological and psychiatric disorders, history of opium or alcohol abuse, concurrent use of sedatives, uncontrolled hypothyroidism, and history of upper airway surgery, severe craniofacial abnormalities and those who did not accept to participate were excluded from the study.After explaining designs and objectives of the study and receiving informed consent, demographic and anthropometric data of the subjects were gathered by a hemodialysis nurse trained in sleep apnea.History of systemic hypertension, diabetes mellitus, ischemic heart disease, cerebrovascular disease along with HD duration was recorded.Anthropometric variables included height, weight, neck, waist circumferences and body mass index (BMI) were collected.Adequacy of HD was evaluated with urea kinetic modeling and expressed as Kt/V.Hence, Kt/V ≥1.2 used as the marker of efficient dialysis according to 2006 KDOQI guidelines (11).Other laboratory data including serum urea, creatinine, calcium, phosphorus, sodium, potassium, uric acid, hemoglobin concentration (HB), iron, total iron binding capacity (TIBC), ferritin, albumin, fasting blood sugar, cholesterol, triglyceride, and parathyroid hormone were gathered from dialysis chart review.

Obstructive sleep apnea assessment
Upper airway examination was evaluated by Mallampati scale (mouth opening amount classified from class-1 with visible soft palate, uvula and pillars to class-4 with only hard palate visible) and tonsillar hypertrophy.Subjects were also categorized as low and high risk for OSA by Persian versions of STOP-BANG (12) and Berlin questionnaire (BQ).STOP-BANG is a validated popular questionnaire consisting of 8 items used for screening of OSA: Snoring, daytime tiredness, observed pauses in breath, hypertension, BMI>35 kg/m 2 , age>50 years, neck circumference>40 cm and male gender.A total STOP-BANG score ≥3 was considered as high risk for OSA (13).BQ is a validated and accurate questionnaire including 11 questions, which is divided into three sections.It addresses snoring and witnessed apnea, daytime fatigue and sleepi ness and history of hypertension, height, weight, gender and body mass index (BMI).Individuals with positive results in ≥2 out of 3 categories were considered as high risk for OSA (14).Excessive daytime sleepiness (EDS) was evaluated by Epworth Sleepiness Scale (ESS) which is an 8 items questionnaire.It asks patients to estimate their dozing propensity in different situations.Results are summed to a final score 0 to 24.ESS greater than 10 categorizes the subject as sleepy (15).

Follow up
Subjects were followed up by phone call 2 years after enrollment in the study in order to assess mortality rate.

Ethical issues
The research followed the tenets of the Declaration of Helsinki.After explaining designs and objectives of the study the informed consent was obtained.The patients' medical records were used anonymously for data gathering and this process was secret.

Statistical analysis
Normal distribution of variables was assessed by Kolmogorov-Smirnov test.Chi-square test was used to determine differences in the frequency of variables between two groups (low risk and high risk for OSA).Continuous variables according to normal distribution were tested via independent sample t test or Mann-Whitney U test.Two-sided P value ≤0.05 was considered statistically significant.Correlation of OSA risk with HD efficacy, demographic and anthropometric data was assessed.Univariable and multivariable logistic regression with backward method was used to determine the independent association of Kt/V on high-risk category OSA.SPSS (version 16, SPSS Inc., IL, USA) was used for statistical analysis.

Descriptive analysis
A total number of 65 HD patients (63% men) with mean ± SD age of 53.0 ± 16.5 years were enrolled in the study.Among them 49 patients (75.4%) were hypertensive and 21 (32.3%) had diabetes type 2. Mean ± SD for BMI and waist circumference was 23.1 ± 4.2 kg/m² and 89.4 ± 13.2 cm, respectively.Eleven patients (19%) were anemic (Hb <10 g/dL) with mean ± SD hemoglobin concentration of 11.2 ± 1.9 g/dL in all patients.HD duration was ranged from 3 to 255 months.HD was efficient in 41 (67.2%) of patients with mean ± SD Kt/V of 1.33 ± 0.29.Grade 2 and 3 tonsillar enlargement were seen in 21.5% and 1.5%, respectively.Mallampati grade 3-4 was observed in 29 patients (44.6%).The range of STOP-BANG and BQ scores were 1-6 and 0-32, respectively.High risk OSA was determined in 55 subjects (84.6%) and 43 (68.3%) according to STOP-BANG and BQ, respectively.ESS score ranged from 0 to 16.Most of the subjects were in non-EDS group (74.2%) and 25.8% had EDS.Patients' demographic data are summarized in Table 1.

Correlation analysis
Subjects in high risk group (based on STOP-BANG) was significantly older (mean age 58.6 vs. 41 years, P < 0.001).Chi-square analysis showed male predominance (70.9% versus 20%, P = 0.004) and higher diabetes prevalence (38.2% versus 0.0%, P = 0.024) in high-risk versus lowrisk group (based on STOP-BANG).STOP-BANG questionnaire screening showed high risk OSA is associated with age, neck, waist circumference, Kt/V, male gender and diabetes.BQ screening showed high risk OSA is correlated with waist circumference, Kt/V and erythropoietin administration (Table 1).Efficient HD was more prevalent in low-risk group (based on BQ) (90% versus 56.1%, P = 0.009) and also in no EDS group (73.8% versus 43.8%, P = 0.031).Although anemia was not more prevalent in high risk group (according to BQ), using erythropoietin was more frequent in them (88.6% versus 66.7% in lowrisk, P = 0.04).Lower hemoglobin concentrations were observed in EDS group (10.2 ± 2.01 versus 11.6±1.85,P = 0.032).Other clinical and laboratory variables were not significantly different in OSA (Table 1) and EDS groups.Univariable analysis for prediction of OSA based on STOP-BANG showed significant correlation with age, male gender, waist circumference and Kt/V (Table 2).However, after multivariable logistic regression we found only Kt/V and age were associated with OSA based on STOP-BANG (Table 2).Based on BQ, we found that waist circumference significantly associated with OSA risk after controlling for EDS, dialysis duration and Kt/V (OR: 1.96; 95% CI: 1.01-1.12,P = 0.02) (Table 3).

Discussion
Nocturnal HD has been shown to improve OSA severity (16).Volume overload and night time rostral fluid shift have been suggested as the pathogenic mechanisms for the high prevalence of OSA in HD patients (17).Our study reproduced the results of other studies in terms of high prevalence of OSA risk in HD patients (3,(18)(19)(20).Using STOP-BANG most of our patients (84.6%) were assessed as high risk of OSA which is much higher than general Iranian population (21).This high prevalence has been observed in other studies with a range of 64% to 80.6% (20,19).Wali et al highlighted that BQ is a less sensitive tool for screening OSA in 44.2% of HD patients (21) while 68.3% of our patients were screened as high risk of OSA.Univariable analysis showed a significant correlation between waist circumference and Kt/V.In multivariable analysis OSA was significantly associated with age and waist circumference (OR: 1.09, 95% CI: 1.02-1.16,P=0.01 and OR: 1.06, 95% CI: 1.00-1.12,P = 0.025, respectively).Patients who were screened as high risk of OSA had lower Kt/V in comparison to the low risk of OSA group (Table 1).
Multivariate analysis revealed a significant negative independent correlation of OSA with Kt/V (OR=0.007,CI 95%: 0.0-0.36)which can be interpreted as a protective role of adequate dialysis for OSA.Similar results were observed in a Chinese cohort with reported lower dialysis dose (single-pool Kt/V) increases the risk of sleep apnea and likelihood of daytime sleepiness (EDS) (22).Although EDS was not common in our patients (25.8%), a wide range of EDS prevalence has been reported in individuals undergoing HD (22.2 and 77%) (23,18).This wide range of EDS can be explained by the subjective nature of ESS questionnaire.EDS can also be explained by day-night sleep reversal due to uremic encephalopathy and abnormal metabolism (24).This study showed anemia is significantly more prevalent in the EDS group.In other studies EDS was also correlated to older age, upper airways obstruction and Erythropoietin use (25).OSA is considered as a well-known independent risk factor for all-cause mortality (8-9,26).Our results showed no significant increase in mortality of high risk OSA subjects.Survey of a larger cohort with the use of polysomnography as the gold standard of OSA diagnosis can be more informative in terms of mortality.

Conclusions
Nephrologists should be more vigilant about early diagnosis of OSA in ESRD subjects in order to improve HD efficacy and prevent cardiovascular complications of OSA.STOP-BANG and BQ are simple tools for

Table 1 .
Comparison of variables between low risk and high risk OSA groups based on BQ and STOP-BANG

Characteristics a Questionnaire BQ STOP BANG Low risk OSA High risk OSA P value b Low risk OSA High risk OSA P value b
a Values are expressed as mean ± SD. b P values were tested by t test and Mann-Whitney U test.c P value ≤0.05 was considered as statistically significant.d Values were analyzed by chi-square test and expressed as number (percent).

Table 3 .
Univariable and multivariable logistic regression of variables associated to OSA (based on Berlin Questionnaire) P value ≤0.05 was considered as statistically significant. a

Table 2 .
Univariable and multivariable logistic regression of variables associated to OSA (based on STOP-BANG)