Obstructive sleep apnea in non-dialysis chronic kidney disease patients

Abstract

There is an enormous increase in the burden of chronic kidney disease both in developing and developed countries. There is a paucity of data on obstructive sleep apnea (OSA) in chronic kidney disease (CKD) patients in India. We used a cross-sectional prospective observational study to determine the prevalence of OSA in non-dialysis CKD patients. Of the 647 CKD patients 302 patients were in stage II, III and IV. The study population was screened using the Berlin questionnaire and 87 patients were positive for OSA (28%). Among the 87, 37 patients were excluded based on the exclusion criteria. Fifty patients underwent a split night sleep study. Stage II, III CKD patients were clubbed as early CKD or group one and stage IV CKD patients were clubbed as late CKD or group two. The spilt night study revealed an 88% incidence of OSA of varying severity. A sub group analysis was done to assess the severity of OSA. A statistical significance (p < 0.05) between early and late CKD group was observed with respect to AHI and ODI. An improvement in the late CKD is observed and the Z values for AHI and ODI are 4.273 and 2.307, respectively. The prevalence and incidence of OSA was found to be 28% and 88% in non-dialysis CKD patients, respectively and the risk and severity of OSA increased with the progression of CKD stages and thus necessitating the need for screening the non-dialysis CKD population.

• Berlin questionnaire
• non-dialysis CKD
• obstructive sleep apnea
• sleep apnea and renal function
• sleep study in CKD

Introduction

Chronic kidney disease (CKD) encompasses a spectrum of different pathophysiologic processes associated with abnormal kidney function and a progressive decline in glomerular filtration rate (GFR).1 Prevalence is estimated to be 8–16% worldwide.2 The Indian CKD registry and previous studies done in Indian population revealed an approximate prevalence of 17.2% with preponderance of male sex.3^,4 Obstructive sleep apnea (OSA) happens to be the most common form of sleep disordered breathing (SDB) and was recognized more than 30 years ago and presents as complete or partial blockage of the upper airway during sleep.5 Population-based studies in the developed countries have shown that up to 6% of adults are likely to have undiagnosed OSA.6^,7 Studies conducted on general population to assess the prevalence of OSA in India reveals a prevalence ranging from 7 to 20%.8^,9 Several studies in developed countries on CKD patients have shown a higher prevalence of SDB ranging from 18 to 80% with prevalence in maintenance of hemodialysis patients to be 50–80%.10–13 Studies have shown that OSA in CKD patients are present with non-specific symptoms like tiredness, headache, decreased sleep during night, increased day time somnolence, and snoring.14–16 These risks increase twofold when the two disorders CKD and OSA coexist.9–11^,15 OSA carries increased morbidity due to hypertension, coronary artery disease, cardiac arrhythmias and stroke.16

The increased prevalence of OSA in CKD patients can be attributed to reduction in upper airway dimensions in patients with renal failure.1^,14^,17–20 Beecroft et al. reported that the pharyngeal cross-sectional area measured by pharyngometry in ESRD patients was 12% less than that in the normal renal function control group matched for BMI.21 Such pharyngeal narrowing was considered to occur because of upper airway edema due to systemic fluid overload and upper airway dilator muscle dysfunction due to neuropathy or myopathy associated with chronic uremia. Ventilation control is known to be unstable in CKD patients. Chemoreflex responsiveness has been reported to be augmented in patients with CKD V, possibly because of the accumulation of uremic molecules or metabolic acidosis, explained by a high “loop gain” theory, which contributes to the pathogenesis of OSA in CKD.21 Upper airway dilator muscle dysfunction secondary to neuropathy or myopathy associated with either chronic uratemia or the underlying cause of CKD, such as diabetes mellitus can attribute to OSA.22

There is a paucity of data in the Indian population. Hence, this prospective study was undertaken to assess the prevalence of OSA in CKD patients not on dialysis.

Methods

Study design

We performed a cross-sectional, prospective observational study in a tertiary care center in South India, enrolling patients between January 2012 and July 2013. All CKD patients attending the outpatient department of general medicine and nephrology were assessed with Cockroft – Gault equation to stage the CKD. Patients in Stage II, III and IV of CKD were enrolled in the study. This population was screened for OSA with Berlin questionnaire.23 The patients who were at high risk in accordance to Berlin questionnaire were enrolled for the study. Patients were not included on the basis of exclusion criteria such as obesity (BMI > 27 kg/m²) in accordance to the Indian guidelines.24 Other exclusion were hemoglobin <12 g/dL, abnormal thyroid function tests, CNS depressant drug usage or alcohol abuse within the last one week and clinical conditions like COPD and bronchial asthma. The study population was divided into two groups; group 1 comprising of Stage II– III (early CKD) and the group 2 Stage IV (late CKD) patients. After getting the institutional ethics committee approval and a written consent from the patient about the polysomnography study, patients in both the groups underwent a diagnostic split night sleep study. The first phase of the study that is, first three hours of the study was for diagnosis of OSA and if the patient was OSA positive, a titration of continuous positive airway pressure (CPAP) was done during the second phase of the study. The diagnosis of OSA was made based on the apnea hypopnea index (AHI) and oxygen de-saturation index (ODI), which is the number of apneas and hypopneas per hour of sleep.16 An apnea is defined as the complete cessation of airflow for at least 10 s and a hypopnea is defined as a reduction in airflow (30–50%) that is followed by an arousal from sleep or a decrease in oxy-hemoglobin saturation (3–4%).22^,25–27 AHI of 5–15 is considered as mild OSA, 16–30 as moderate OSA and AHI of 30 or more as severe OSA.22 After the sleep study the OSA patients were informed about the disorder and the treatment options including CPAP therapy were emphasized.

Statistical analysis

The study was to assess the prevalence of OSA and hence descriptive statistics was used. All the results were calculated using IBM SPSS version 22.0 software (Armonk, NY). Continuous variables were presented as median values or interquartile ranges, and categorical variables were presented as frequency (number) and proportions.

The Mann–Whitney U-test statistic was applied between the two groups due to the limited population in the two groups and a Z test was applied in the aspect to the AHI and ODI. A subgroup analysis was done to assess the incidence and severity of OSA in each stage of CKD.

Results

The baseline characteristics of the study population are summarized in Table 1. A total of 647 CKD patients were screened and 302 (46.67%) patients were in stage II, III and IV of CKD. Among the 302 patients 87 (28.81%) were in stage II, 104 (34.44%) in stage III and 111 (36.75%) in stage IV of CKD. Eighty-seven (28%) among the 302 patients were at high risk for OSA in Berlin questionnaire (Figure 1). Thirty-seven (42.5%) among the 87 patients were excluded based on the exclusion criteria. Fifty patients were divided into group 1 and group 2 who underwent split night sleep study for polysomnography testing (Table 2). The prevalence of OSA was found to be 28% after screening the population with Berlin questionnaire. The incidence of OSA was found to be 88% in the Berlin questionnaire positive group after polysomnography (Figure 2). A sub-group analysis was done to assess the severity of OSA which revealed 5 (18.51%) had normal study with AHI less than 5, 14 (51.85%) had mild sleep apnea with AHI ranging between 5 and 15, 4 had moderate OSA with AHI ranging between 16 and 30 and 4 had severe OSA with AHI more than 30 in the early CKD group. In the late CKD group 1 (4.34%) had normal study with AHI less than 5, 1 (4.34%) had mild OSA with ranging between 5 and 15, 4 (17.4%) had moderate OSA with AHI ranging between 16 and 30 and 17 (73.92%) had severe OSA with AHI more than 30 (Figure 3). The prevalence of OSA in various stages of CKD was assessed. There was 6% incidence in the stage II with mild and moderate forms of OSA. The incidence of OSA in stage III was 48% with majority of the patients having mild and moderate OSA. The incidence of OSA in stage IV CKD was 46% with the majority of the patients having severe OSA than mild to moderate form of the disease (Figure 4). The Mann–Whitney U-test statistics was applied and a statistical significance (p < 0.05) between early and late CKD with respect to AHI and ODI was observed (Table 3). An improvement in the late CKD is observed and the Z values for AHI and ODI were 4.273 and 2.307, respectively (Table 4).

Figure 1. Prevalence of OSA in Stage II, III and IV using Berlin questionnaire.

Figure 2. Incidence of OSA in the Berlin Questionnaire positive population.

Figure 3. Severity of OSA in early and late CKD.

Figure 4. Incidence of OSA in various stages of CKD.

Table 1. Baseline characteristics.

Group statistics	Group		Mean ± Std. deviation		t (Sig.) – Value
UREA	Early CKD		62.150 ± 11.595		7.204 (0.000*)
	Late CKD		96.040 ± 20.999
Creatinine	Early CKD		1.925 ± 0.363		11.714 (0.000*)
	LATECKD		3.752 ± 0.708
GFR	Early CKD		43.948 ± 9.902		10.696 (0.000*)
	Late CKD		20.595 ± 3.644
HB	Early CKD		11.966 ± 1.630		1.067 (0.291^NS)
	Late CKD		12.482 ± 1.788
TSH	Early CKD		2.133 ± 0.969		0.844 (0.403^NS)
	Late CKD		2.360 ± 0.927
BMI	Early CKD		24.970 ± 0.975		2.046 (0.046*)
	Late CKD		25.565 ± 1.079
SBP	Early CKD		151.560 ± 13.328		0.995 (0.325^NS)
	Late CKD		155.040 ± 11.101
DBP	Early CKD		90.520 ± 9.242		1.206 (0.234^NS)
	Late CKD		87.650 ± 7.228
HBA1C	Early CKD		7.174 ± 1.298		0.389 (0.699^NS)
	Late CKD		7.047 ± 0.929
Height (cm)	Early CKD		166.520 ± 7.024		0.158 (0.875^NS)
	Late CKD		166.830 ± 6.692
Weight (kg)	Early CKD		69.480 ± 6.902		0.997 (0.324^NS)
	Late CKD		71.350 ± 6.220
			Group
Sex * group Crosstabulation			Early CKD	Late CKD	Total
Sex	Female	Count	6	10	16
		%	37.5%	62.5%	100.0%
	Male	Count	21	13	34
		%	61.8%	38.2%	100.0%
	Total	Count	27	23	50
		%	54.0%	46.0%	100.0%
Albumin	Mean ± Std. deviation				t (Sig.) – Value
Early CKD	3.626 ± 0.2212				9.286(0.000*)
Late CKD	3.130 ± 0.1845

Note: *Denotes statistical significance.

Table 2. Groups 1 and 2.

Group 1 Early CKD N = 27	GFR 30–89 mL/min
Group 2 Late CKD N = 23	GFR 15–29 mL/min

Table 3. AHI between the early CKD and the late CKD groups.

	GROUP
AHI * group Crosstabulation	Early CKD	Late CKD	Total
AHI
Normal
Count	5	1	6
%	83.3%	16.7%	100.0%
Mild
Count	14	1	15
%	93.3%	6.7%	100.0%
Moderate
Count	4	4	8
%	50.0%	50.0%	100.0%
Severe
Count	4	17	21
%	19.0%	81.0%	100.0%
Total
Count	27	23	50
%	54.0%	46.0%	100.0%

Note: *Denotes statistical significance.

Table 4. AHI and ODI between the two groups.

	Group 1 Early CKD	Group 2 Late CKD
AHI mean SD	37.49 ± 24.80	58.37 ± 33.98
ODI Mean SD	39.09 ± 37.13	51.43 ± 47.31

The pattern of cardiac rhythms that were observed in the ECG during the study was sinus rhythm, premature ventricular complexes, and sinus tachycardia. Of the 27 patients in early CKD group 2 (7%) had sinus tachycardia, 3 (12%) had premature ventricular complexes and 22 (81%) had sinus rhythm. In the late CKD group 2 (9%) had sinus tachycardia, 1 (4%) had premature ventricular complexes and 20 (87%) had sinus rhythm. The number of arousals was determined by arousal index which was determined by EEG. The transition from delta or theta wave to alpha or beta wave was considered an arousal. In the early CKD group 13 (48%) had an arousal and 14 (52%) had no arousal. In the late CKD group 14 (61%) had arousal and 9 (39%) patients had no arousal. One patient in the late CKD group had periodic leg movement syndrome which was determined by periodic leg movement index.

Discussion

Our study revealed the prevalence and incidence of OSA in non-dialysis CKD patients to be 28% and 88%, respectively. This shows a high incidence of OSA in non-dialysis CKD which had not been looked at in Indian CKD population. This study has limitations because it is a cross-sectional analysis. We found that there is an increased risk of OSA associated with a decline of kidney functions in CKD patients.

To our knowledge this is the first study in South Indian population to assess the prevalence of OSA in non-dialysis CKD patients. Previous studies conducted in India on maintenance hemodialysis showed a prevalence of 24.6%, however, the true incidence was not assessed.9^,17 A study by Sakaguchi et al. conducted in non-dialysis CKD population in Japan revealed a high prevalence of 65%, though an initial screening test was not employed as we have done and confounding factors like anemia, thyroid disorders and COPD were not excluded.19 Another study by Markou et al., showed a prevalence of OSA at 54.3% in 35 non-dialysis CKD patients.20

A study by Roumelioti et al., with a median age of 58.1 years in stage IV to V CKD patients demonstrated a high incidence of OSA.18 The increase in age has been attributed as an independent risk factor for OSA in various studies.28^,29 The median age of our study population was 56.5 years which was low when compared to other studies done in non-dialysis patient.19^,20^,28 The study population was screened using Berlin questionnaire to assess the prevalence of OSA and was found to be 28%. The incidence of OSA after sleep study in screened population was 88%. Ramachandran et al. and Sharma et al. concluded that Berlin questionnaire has better sensitivity and specificity than Epworth sleepiness scale and hence Berlin questionnaire was used as a screening tool to identify the prevalence of OSA which is being concurred by this study.23^,30

We stratified our study population into two groups with male preponderance of 68%. Bixler et al. observed that OSA is more prevalent in male population as in our study.31^,32 The mean BMI in our study group was 25.244 kg/m² which was over weight as per Indian standards and was not suggestive of obesity as in other studies, thus excluding obesity as an individual risk factor for OSA.24^,33^,34 Anemia was not a confounder as per CKD stage.35 The mean SBP and DBP in the group 1 was 151.560 ± 13.328 mmHg and 90.520 ± 9.242 mmHg. Hypoalbuminemia was not a confounding factor as the mean value was 3.626 ± 0.2212 g/dL and 3.130 ± 0.1845 g/dL in the two groups. The mean SBP and DBP in group 2 was 155.040 ± 11.101 mmHg and 87.650 ± 7.228 mm Hg, respectively, and had not influenced the results of the study.36

As reported in other studies we observed an increased risk of OSA with progression of CKD stage which was statistically significant (p < 0.05) This is consistent with the previous studies conducted by Sakaguchi et al., Morkou et al., Roumelioti et al., which shows similar results.18–20

Conclusion

This study of OSA in non-dialysis CKD patients is unique, which revealed the high incidence and prevalence of OSA. OSA increases with progressive stage of CKD. However, further prospective longitudinal study with large sample sizes are needed with intervention in controlling progression of CKD, fluid electrolyte balance anemia, cardio vascular disease and targeted therapy with CPAP.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.