Association between non-alcoholic fatty liver disease and obstructive sleep apnea

Table 1 Summary of the 13 included studies

Ref.	Study design	Number of patients	Diagnosis and severity of OSA	Diagnosis of NAFLD	Patient characteristics	Results
Agrawal[49], 2015 (India)	Prospective	23 (3 mild OSA, 5 mode rate OSA, 15 severe OSA)	- No OSA, AHI < 5;	Abdominal ultrasound	Consecutive patients with diagnosis of OSA and abdominal obesity	- The prevalence of NAFLD in patients with OSA was 91.3%
						- AHI was an independent predictor of significant fibrosis
			- Mild OSA, 5-14.9;
						- No differences in the prevalence of NAFLD, raised transaminase levels and fibrosis according to the severity of OSA
			- Moderate OSA, 15-30;		Mean age: 46; Mean BMI: 32.2; Males: 78%
			- Severe OSA, > 30
Cakmak[51], 2015 (Turkey)	Retrospective	137 (118 OSA: -19 mild, - 39 moderate, - 60 severe, 19 no OSA)	- No OSA, AHI < 5;	Abdominal ultrasound	All consecutive patients referred to a sleep laboratory due to sleep apnea symptoms	- Severity of NAFLD increased as AHI increased and lowest SpO2, mean nocturnal SpO2 levels decreased
						- There was a strong association between NAFLD severity and a decrease in lowest SpO2 levels
			- Mild OSA, 5-14;
			- Moderate OSA, 15-29;		Mean age: 55.7; Mean BMI: 34.5 (OSA), 33.2 (no OSA); Males: 44.5%
						- Strong association between elevated liver enzymes and increase in nocturnal hypoxia severity in OSA patients
			- Severe OSA, ≥ 30
Petta[34], 2015 (Italy)	Cross-sectional	50 (25 OSA, 25 no OSA)	- No OSA, AHI < 5;	Liver biopsy	Consecutive patients with biopsy-proven NAFLD who underwent cardio-respiratory polygraphy	- Significant fibrosis was independently associated with mean nocturnal oxygen saturation < 95% in patients with NAFLD and OSA
			- OSA, AHI ≥ 5
					Mean age: 53; Mean BMI: 33.5 (OSA), 29.0 (no OSA); Males: 58%
Yu[36], 2015 (South Korea)	Cross-sectional	621 (286 OSA, 335 no OSA)	- No OSA, AHI < 5;	Abdominal CT scan	Subjects who examined the PSG and abdominal CT	- Patients with OSA were significantly older and had significantly higher BMI than those without OSA
						- The prevalence of NAFLD was 34% among patients with OSA and 21% among patients without OSA
					Mean age: 56.6; Mean BMI: 24.7; Males: 57.2%
			- OSA, AHI ≥ 5
						- Association between OSA and NAFLD independent of the visceral fat level in relatively lean individuals
						- This association was particularly strong in participants with excessive daytime sleepiness or short sleep duration regardless of visceral fat level
Arısoy[41], 2016 (Turkey)	Case-control	176 (52 mild, 34 moderate, 48 severe, 42 no OSA)	- No OSA, AHI < 5;	Abdominal ultrasound	Subjects referred to a sleep center with clinical suspicion of OSA	- Hepatosteatosis grade, ALT and AST levels, BMI differed significantly among the groups
			- Mild OSA, 5-14;			- BMI and hepatosteatosis grade increased progressively and significantly from no OSA to severe OSA
					Mean age: 45.1 (no OSA), 42.9 (mild), 47.6 (moderate), 47.0 (severe); Mean BMI: 28.3 (no OSA), 30.1 (mild), 34.1 (moderate), 32.7 (severe); Males: 73.9%
			- Moderate OSA, 15-29;			- Average desaturation and BMI were the parameters with the greatest independent effects on hepatosteatosis in the subjects with OSA
			- Severe OSA, ≥ 30
Benotti[40], 2016 (United States)	Retrospective	362 (115 mild, 80 moderate, 74 severe, 93 no OSA)	- No OSA, AHI < 5;	Liver biopsy	Bariatric surgery candidates with clinical suspicion of OSA	- OSA severity was associated with NAFLD liver histology only in patients without metabolic syndrome
			- Mild OSA, 5-14;
					Mean age: 46.2; Mean BMI: 49.9; Males: 21%
			- Moderate OSA, 15-29;
			- Severe OSA, ≥ 30
Buttacavoli[35], 2016 (Italy)	Observational	15	- Severe OSA, AHI ≥ 30	Abdominal ultrasound and elastography	Consecutive severe OSA patients at baseline and after 6-12 mo of CPAP treatment	- Most patients at diagnosis had severe liver steatosis (87%)
						- During follow-up, steatosis significantly improved in six patients without concurrent changes in the BMI range in the entire sample
						- No correlation was found between steatosis score and BMI at baseline, although a positive relationship between these variables was evident during CPAP treatment
					Mean age: 49.3; Mean BMI: 35.4; Males: 86.7%
Chen[37], 2016 (China)	Cross-sectional	319 (Group 1: 187 OSA with FLI < 60; Group 2: 132 OSA with FLI ≥ 60)	- No OSA, AHI < 5;	Fatty liver index (FLI) ≥ 60	All consecutive patients referred to a sleep center and diagnosed with OSA	- Participants with a FLI ≥ 60 tended to be significantly fatter and had higher transaminase levels and severe PSG parameters of sleep apnea
			- Mild OSA, 5-14.9;
					Mean age: 46.8 (Group 1), 42.3 (Group 2); Mean BMI: 24.5(Group 1), 28.5 (Group 2); Males: 79%
			- Moderate OSA, 15-30;
						- Severity of OSA was independently associated with prevalence of NAFLD (52.1% in patients with AHI ≥ 15 vs. 20.4% in patients with AHI < 15)
			- Severe OSA, > 30
Qi[50], 2016 (China)	Cross-sectional	175 (149 OSA: - 96 NAFLD, - 53 no NAFLD, 26 no OSA: - 10 NAFLD, - 16 no NAFLD)	- No OSA, AHI < 5;	Abdominal ultrasound	All consecutive non-obese patients referred to a sleep laboratory with clinical suspicion of OSA	- Prevalence of NAFLD in OSA patients was 64%
						- BMI, lowest SpO2, and triglycerides may be risk factors for promoting NAFLD in OSA patients
			- Mild OSA, 5-14.9;
					Mean age: 52.9 (OSA and NAFLD); Mean BMI: 24.0; Males: 87.9% (OSA), 77.3% (no OSA)
			- Moderate OSA, 15-29.9;
			- Severe OSA, > 30
Chen[52], 2018 (China)	Observational	160 (42 moderate OSA, 88 severe OSA, 30 controls)	- No OSA, AHI < 5;	Abdominal ultrasound	All consecutive patients referred to a sleep laboratory with clinical suspicion of OSA	- Prevalence of liver steatosis was 64% among the groups; 59.5% and 81.8% in patients with moderate and severe OSA respectively
			- Moderate OSA, 5-30;
						- Increasing OSA severity was associated with higher BMI, waist circumference and neck circumference
					Mean age: 42.6; Mean BMI: 28.0; Males: 86.9%
			- Severe OSA, ≥ 30
						- ALT, AST and liver steatosis score increased significantly with an increase in OSA severity
						- OSA severity was independently associated with liver steatosis and elevation of serum aminotransferases, but not with liver fibrosis
						- Serum aminotransferase, as a biomarker of liver injury, decreased in OSA patients after 3 months of CPAP treatment
Kim[38], 2018 (United States)	Retrospective	351 (73 mild OSA, 102 moderate OSA, 176 severe OSA)	- No OSA, AHI < 5;	Suspected NAFLD was diagnosed if serum ALT > 30 U/L for men and > 19 U/L for women; Advanced fibrosis was identified by the AST to platelet ratio index (APRI) score	CPAP-treated OSA adult patients who had available serum ALT data before (within 3 months) and after (within 6 months) CPAP treatment	- The prevalence of suspected NAFLD was higher (90.3%) among patients with moderate to severe OSA versus among those with mild OSA (86.3%)
			- Mild OSA, 5-14.9;
						- Fibrosis was correlated with OSA severity (7.6% for mild OSA versus 12.0% moderate OSA versus 19.7% for severe OSA)
			- Moderate OSA, 15-30;
					Mean age: 57.6; Mean BMI: 32.2; Males: 59.3%
						- There was a dose-response relationship between OSA severity and improvement in ALT and AST levels and APRI score after CPAP treatment, correlating with adherence status and without differences in the obesity severity status
			- Severe OSA, > 30
Trzepizur[39], 2018 (France)	Cross-sectional	124 (34 mild, 38 moderate, 52 severe)	- No OSA, AHI < 5;	Elastography	Patients with at least one criterion for metabolic syndrome with diagnosis of OSA	- Prevalence of advanced liver fibrosis was 12%
						- Increasing OSA severity was associated with BMI, waist circumference, ODI, percentage of sleep time with SpO2 < 90% and higher proportions of male patients with metabolic syndrome
			- Mild OSA, 5-14.9;
					Mean age: 52.4; Mean BMI: 29.9; Males: 65.6%
			- Moderate OSA, 15-29.9;
						- Increasing OSA severity was also associated with higher LSM values with a marked increase between mild-to-moderate OSA and severe OSA
			- Severe OSA, ≥ 30
						- Patients with severe OSA and metabolic comorbidities are at higher risk of significant liver disease (LSM ≥ 7.3 kPa) and advanced liver fibrosis (LSM ≥ 9.6 kPa)
						- AHI and ODI were the factors with the strongest independent association with LSM
Bhatt[42], 2019 (India)	Case-control	240 (124 OSA and NAFLD, 47 OSA without NAFLD, 44 NAFLD without OSA, 25 no OSA and no NAFLD)	- No OSA, AHI < 5;	Abdominal ultrasound	Overweight/obese subjects (BMI > 23 kg/m2)	- Mean values of AST, ALT and BMI were significantly higher in OSA with NAFLD group as compared to the other groups
						- Inflammatory markers showed a significant correlation in the OSA and NAFLD group
					Mean age: 44.8 (OSA and NAFLD); Mean BMI: 33.3 (OSA and NAFLD); Males: 55.0%
			- OSA, AHI ≥ 5
						- OSA and NAFLD operate as an independent contributors to the increased systemic inflammation that occurs in overweight/obese subjects

ALT: Alanine aminotransferase; AHI: Apnea-hypopnea index; APRI: Aspartate aminotransferase to platelet ratio index; AST: Aspartate aminotransferase; BMI: Body mass index; CPAP: Continuous positive airway pressure; CT: Computed tomography; FLI: Fatty liver index; LSM: Liver stiffness measurement; NAFLD: Non-alcoholic fatty liver disease; ODI: Oxygen desaturation index; OSA: Obstructive sleep apnea; PSG: Polysomnography; SpO₂: Oxygen saturation.