Table 6. Summary of results.

PCOS = polycystic ovary syndrome; NAFLD = non-alcoholic fatty liver disease; HS = hepatic steatosis; P = p-value; BMI = basal metabolic index; WC = waist circumference; HOMA-IR = homeostasis model assessment-estimated insulin resistance; FAI = free androgen index; HDL = high-density lipoprotein; SHBG = sex hormone binding globulin; MetS = metabolic syndrome; LAP = lipid accumulation product; TG = triglycerides; LSM = liver stiffness measurement; CAP = controlled attenuation parameter; TE = transient elastography; OR = odds ratio; CI = confidence interval; FIB-4 = fibrosis-4; IR = insulin resistance; nmol/L = nanomoles per liter

Author and Year of Publication	Study Type	Purpose of Study	Number of Participants/ studies	Results and Conclusion
Vassilatou et al., 2010 [10]	Case-control study	Investigation of premenopausal PCOS patients (via abdominal ultrasonography and biochemical testing) to determine the presence of NAFLD Assessment of metabolic and hormonal factors correlating with NAFLD and PCOS.	57 premenopausal women with PCOS; 60 age- and weight-matched women without PCOS as controls	36.8% of cases had HS vs 20.0% of controls (P < 0.05) whereas 22.8% of cases had abnormal serum aminotransferases vs only 3.3% of controls (P < 0.01). All participants who had MetS also had evidence of HS. Patients with HS were determined to be 3.55 times more likely to have PCOS [i.e., OR = 3.55 with (95% CI 1.02-5.35)]. Factors that correlate HS with PCOS are: PCOS diagnosis, advanced age, elevated BMI, WC, HOMA-IR, and FAI, as well as decreased HDL and SHBG levels. It was concluded that NAFLD was common in PCOS patients, likely due to increased androgens and metabolic abnormalities.
Jones et al., 2012 [16]	Cross-sectional Case-control study	To determine if PCOS is an independent risk factor for HS and determine if HS is related to hyperandrogenemia	29 PCOS patients and 22 age and BMI-matched healthy control women	Hyperandrogenic women with PCOS were found to have greater liver fat content than non-hyperandrogenic PCOS patients and healthy controls, even after adjusting for BMI, HOMA-IR, and internal and visceral adipose tissue volume. Therefore, it was concluded that hyperandrogenism in PCOS patients, irrespective of the presence of IR and obesity, is associated with HS.
Polyzos et al., 2014 [13]	Cross-sectional study	To investigate the association of non-invasive indices of HS and fibrosis with MetS in PCOS patients versus controls from Greece.	314 PCOS women (77 with MetS and 237 without) and 78 controls	All three steatosis indices were higher in PCOS patients than controls, whereas only two of the four fibrosis indices were higher in PCOS patients. All three steatosis indices were higher in PCOS women with MetS than those without it. Still, only one fibrosis index was higher in PCOS women with MetS. Therefore, the results suggest that indices for steatosis have a greater association with MetS than indices for fibrosis, especially in the PCOS patients
Macut et al., 2016 [15]	Cross-sectional study	To determine the prevalence of NAFLD in PCOS patients from Greece and the most significant risk factors associated with progression to NAFLD in PCOS patients.	600 women with PCOS and 125 BMI-matched healthy women as controls	The prevalence of NAFLD was 50.6% in PCOS patients vs. 34.0% in controls. WC, LAP, insulin and HOMA-IR, total cholesterol, and TGs were higher in PCOS patients than in controls (P < 0.001). NAFLD-liver fat score was most significantly associated with WC, BMI, glucose levels, LAP, HOMA-IR, FAI, and TGs. HOMA-IR and LAP were deemed as independent risk factors for NAFLD in PCOS patients
Kim, JJ et al., 2017 [9]	Case-control study	To analyze the prevalence of NAFLD in non-obese women with or without PCOS and to determine the correlation between NAFLD and PCOS in non-obese Asians	275 non-obese PCOS patients from Seoul and 892 non-obese controls from Seoul	5.5% of the non-obese PCOS patients had NAFLD vs. 2.8% of controls (P = 0.027) after adjustment for age and BMI. Hyperandrogenism in the non-obese PCOS cohort was associated with NAFLD even after adjustment for lipid profile, glycemic status, and IR.
Kumarendran et al., 2018 [22]	Cohort Study	To determine the incidence of NAFLD in PCOS patients and to explore the roles of BMI and hyperandrogenism as risk factors for NAFLD	63 120 women with PCOS selected from a primary care database in the United Kingdom and 121 064 age, BMI, and location-matched controls	The hazard ratio for NAFLD in women with PCOS was 2.23 (95% CI 1.86–2.66, p < 0.001), indicating an increased rate of NAFLD in these women. Serum testosterone > 3 nmol/L and SHBG < 30 nmol/L both resulted in increased NAFLD rates. BMI, dysglycemia, and hyperandrogenism contribute to the elevated risk in these patients.
Wu et al., 2018 [28]	Meta-analysis study	To explore the effect of PCOS on NAFLD development and that if the link is direct or due to shared risk factors	17 studies published before May 2017 were included	The OR for NAFLD in PCOS patients was 2.25 (95% CI: 1.95-2.60); therefore, PCOS subjects had a significantly higher risk of developing NAFLD. Prevalence of NAFLD was more common in obese patients vs. non-obese patients. Prevalence of NAFLD in PCOS patients was highest in subjects of Europe, followed by the Asia-Pacific region, and then in America. Hyperandrogenism was deemed as the most influential risk factor, whereas obesity and geography were less influential.
Asfari et al., 2020 [4]	Cross-sectional study	To determine if PCOS is an independent risk factor of NAFLD	77 415 out of 50 785 354 female patients with PCOS, according to National Inpatient Database, from 2002 to 2014	Patients with PCOS had approximately eight times higher odds of having NAFLD  even after adjustment for various confounders. PCOS patients were younger and more obese than controls but less likely to have co-morbidities like hypertension, dyslipidemia, and type 2 diabetes mellitus.
Chakraborty et al., 2020 [14]	Cross-sectional study	To investigate the prevalence of HS in young Indian women with PCOS and to determine the efficacy of TE in the assessment of NAFLD	70 Indian women with PCOS and 60 healthy women as controls	The prevalence of HS in women with PCOS was 38.56%, whereas it was 6.67% in controls. The aminotransferase levels were also significantly higher in PCOS patients. Assessment of liver stiffness measure (LSM) and controlled attenuation parameter (CAP) on TE may predict the presence of NAFLD in PCOS patients.
Salva-Pastor et al., 2020 [6]	Cross-sectional study	To investigate the prevalence of NAFLD in Mexican women with PCOS compared to age and BMI-matched controls	49 women of reproductive age with PCOS and 49 healthy women as controls	Prevalence of NAFLD was 69.3% in PCOS patients and 34.6% in controls; (OR=4.26, 95% CI 1.83-9.93). Prevalence of NAFLD was greater in PCOS patients with phenotype A than in other phenotypes. Patients with excess androgens had higher mean CAP on TE than subjects without hyperandrogenism. It was concluded that PCOS serves as an independent risk factor for NAFLD.
Taranto et al., 2020 [12]	Cross-sectional study	To investigate the prevalence of NAFLD in PCOS patients and their associated risk factors, also investigate various indices of HS in these patients	87 Brazilian women with PCOS and 40 controls	NAFLD was discovered in 77% of the PCOS patients compared to 52.5 % of the controls, likely due to elevated serum TGs, alanine aminotransferase, and WC. FIB-4 Index did not correlate with advanced stage of fibrosis, whereas NAFLD score and TE showed some correlation (3.8% and 12% of patients, respectively).