Nonalcoholic fatty liver disease (NAFLD) is the most common. Prevalence of Nonalcoholic Fatty Liver Disease in Women With Polycystic Ovary Syndrome

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2007;5: Prevalence of Nonalcoholic Fatty Liver Disease in Women With Polycystic Ovary Syndrome MAYA GAMBARIN GELWAN,* SANJIV V. KINKHABWALA, THOMAS D. SCHIANO, CAROL BODIAN, HSU CHONG YEH, and WALTER FUTTERWEIT # *Division of Gastroenterology and Hepatology, Weill Medical College of Cornell University; Division of Endocrinology, New York University School of Medicine; Division of Liver Diseases, Department of Biostatistics, Department of Radiology, and # Division of Endocrinology, Diabetes, and Bone Diseases, The Mount Sinai Medical Center, New York, NY Background & Aims: Insulin resistance has been implicated in the pathogenesis of both nonalcoholic fatty liver disease (NAFLD) and polycystic ovary syndrome (PCOS). We hypothesized that NAFLD would be common in both obese and nonobese women with PCOS. The aim of this study was to estimate the prevalence of and identify associated factors for hepatic steatosis in women with PCOS. Methods: This is a retrospective study of 88 consecutive premenopausal women with PCOS. Clinical history, height, weight, and laboratory values were obtained. Fasting measurements of serum glucose and insulin were used to calculate homeostasis model assessment of insulin resistance (HOMA-IR). Abdominal ultrasonography was used to determine the presence and severity of hepatic steatosis. Results: Of the 88 women (median age, 31.4 years), 48 (55%) had steatosis; 15 (39%) of them were lean women. The presence of steatosis was associated with a greater body mass index (BMI) (P.005) and HOMA-IR (P.033), a lower fasting high-density lipoprotein (P.003), and a greater prevalence of impaired fasting glucose, impaired glucose tolerance, and diabetes mellitus (P.013). Only 7 (15%) subjects with hepatic steatosis had abnormal liver chemistries. Conclusions: Fatty liver was identified in 55% of subjects with PCOS, nearly 40% of whom were lean women. High BMI and insulin resistance appeared to be important associated factors. Early recognition of NAFLD in this group of young patients is warranted, and further investigation including liver biopsy might be indicated. Nonalcoholic fatty liver disease (NAFLD) is the most common cause of cryptogenic cirrhosis.1 First suggested by Caldwell et al 2 in 1999, relationship between NAFLD and cryptogenic cirrhosis has been confirmed by others since then. 3 The estimated prevalence of NAFLD in the general U.S. population is 5% 34%, 4 6 whereas it approximates 75% in those with obesity or type 2 diabetes mellitus. 7,8 At the same time, moderate to severe hepatic steatosis is found in 65% and steatohepatitis in 35% of morbidly obese individuals undergoing liver biopsy at the time of gastric bypass surgery. 9 Several studies have shown that patients with NAFLD (steatosis and steatohepatitis) are indeed at risk for progressive liver disease, with its potential long-term complications of cirrhosis, liver failure, and hepatocellular carcinoma. 2,3,10 12 It has been proposed that NAFLD is associated with metabolic syndrome and varying degrees of insulin resistance Although insulin resistance and hyperinsulinemia are frequently found in obese subjects with NAFLD, both are also noted in lean subjects with fatty liver disease and normal glucose tolerance. 16 Another entity associated with metabolic syndrome and insulin resistance is polycystic ovary syndrome (PCOS). One of the most common endocrinopathies in women, PCOS is prevalent in 5% 8% of premenopausal women and was shown to be as common in black as well as white women of the southeastern United States. 17 At least 50% of patients with PCOS are obese, particularly in the United States. 21 Although population-based estimates demonstrate a 5% 6% prevalence of elevated ALT levels in moderate to severely obese women, 5 a recent review of 73 consecutive infertile women, most of whom had been nulliparous with PCOS, demonstrated that 30% had elevation of ALT and 12% in AST levels (defined as ALT 35 U/L and AST 40 U/L). 22 The authors found a significant association of elevated ALT levels and higher BMI, waist circumference, serum triglycerides, total cholesterol/high-density lipoprotein (HDL) ratio, hirsutism, and insulin resistance. The first report of biopsy-documented nonalcoholic steatohepatitis (NASH) in a young woman with PCOS and insulin resistance was described by Brown et al, 23 who noted a histopathologic improvement after lifestyle modification. This case report was followed by another study of 200 women with PCOS, 15% of whom had elevated liver enzymes. 24 Liver biopsies were performed in 6 women with increased liver enzymes who had features of metabolic syndrome, and all exhibited NASH with fibrosis. In view of the high prevalence of insulin resistance in PCOS and potential adverse consequences of NAFLD, it is important to study the prevalence of NAFLD in PCOS. The aim of this study was to estimate the prevalence of NAFLD in women with PCOS by using abdominal ultrasonography and to note the presence and severity of hepatic steatosis in both obese and nonobese women with PCOS. An additional goal was to identify factors associated with NAFLD in women with PCOS. Abbreviations used in this paper: BMI, body mass index; GGT, -glutamyltransferase; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; LDL, low-density lipoprotein; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; PCOS, polycystic ovary syndrome by the AGA Institute /07/$32.00 doi: /j.cgh

2 April 2007 INCREASED PREVALENCE OF NAFLD IN WOMEN WITH PCOS 497 Figure 1. Ultrasonographic criteria for severity of hepatic steatosis. (A) ABSENT: the echogenicity of the liver parenchyma is slightly greater or equal to that of the renal cortex; clear visualization of the diaphragm and intrahepatic vessels; (B) MILD: slight diffuse increase in fine echoes in the liver parenchyma with normal visualization of the diaphragm and intrahepatic vessel borders; (C) MODERATE: moderate diffuse increase in fine echoes in the liver parenchyma with slightly impaired visualization of intrahepatic vessels and diaphragm; (D) SEVERE: marked increase in fine echoes with poor or nonvisualization of the intrahepatic vessel borders, diaphragm, and posterior or deep portion of the right lobe of the liver. Methods Medical records of reproductive-age women consecutively seen between April and November 2004 in a single endocrinology practice in New York City with expertise in the evaluation and management of PCOS were retrospectively reviewed. Women were included in the study population if they were 18 years of age or older at the time of the initial evaluation and met the consensus diagnostic criteria for PCOS from the 1990 National Institutes of Health conference in Bethesda, Maryland 25 : (1) menstrual dysfunction, (2) hyperandrogenism, and (3) the exclusion of other causes. Subjects were excluded if they had a history of heavy alcohol use, defined as greater than 2 drinks or 20 g alcohol per day. Women with a history of chronic viral hepatitis, hemochromatosis, other chronic liver disease, or those taking hepatotoxic drugs were excluded. Data were collected from the medical records of the subjects office encounters, including clinical history, height, weight, and laboratory values. Fasting measurements of glucose and insulin, lipid profile, and liver chemistries before metformin treatment were recorded. All women underwent abdominal ultrasonography after the initial visit shortly after the biochemical measurements were obtained. The vast majority of women, who had an abdominal ultrasound while receiving metformin, had it performed within 1 year of initiation of therapy. The severity of hepatic steatosis was graded as absent, mild, moderate, or severe (Figure 1) on the basis of the echogenicity of the liver parenchyma and the visualization of intrahepatic vessels and the diaphragm as described previously. 10,26 Ultrasonography was performed with commercially available scanners (Advanced Technology Laboratories, Bothell, WA). Curved array transducers with wide-band frequency 4-2 or 5-2 MHz were used for routine abdominal scanning. In obese women or difficult sound beam penetration, a wide-band frequency 4-1, 4-2, or 3-2 MHz phase array transducer was used for the scanning. The scans were performed in sagittal, coronal, and oblique subcostal planes in supine and left posterior decubitus positions. Ultrasound images were graded at the time of the examination by an experienced radiologist (H C.Y.) who was blinded to the clinical characteristics of the patient, except for the diagnosis of PCOS. The subjects also underwent a pelvic sonography, and the presence or absence of polycystic ovary morphology was confirmed. Subjects were grouped into 3 categories according to their body mass index (BMI) on the basis of the World Health Organization criteria: (1) lean with a BMI 25 kg/m 2, (2) overweight with a BMI 25 but 30 kg/m 2, and (3) obese with a BMI 30 kg/m Insulin resistance was assessed by homeostasis model assessment (HOMA-IR) as calculated from fasting glucose and insulin values 28 : HOMA-IR fasting insulin (miu/l) fasting glucose (mg/dl)/405. Table 1. Demographic and Clinical Data of the Study Population Variable All subjects (N 88) Steatosis absent (N 40) Steatosis present (N 48) P value Age (y) Ethnicity.208 Ashkenazi Jewish 33 (38%) 13 (33%) 20 (42%) Other white 48 (55%) 25 (63%) 23 (48%) Hispanic 4 (5%) 0 (0%) 4 (8%) Asian 2 (2%) 1 (3%) 1 (2%) African American 1 (1%) 1 (3%) 0 (0%) BMI (kg/m 2 ) BMI 25 kg/m 2 38 (43%) 23 (61%) 15 (39%) BMI 25 but 30 kg/m 2 13 (15%) 6 (46%) 7 (54%) BMI 30 kg/m 2 37 (42%) 11 (30%) 26 (70%) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) NOTE. All data are medians or counts with proportions of the column using fasting laboratory specimens unless otherwise noted. P values reflect Mann-Whitney U tests or 2 tests, as appropriate. Percentages might not total to 100% because of rounding.

3 498 GAMBARIN-GELWAN ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 5, No. 4 Figure 2. Prevalence and severity of hepatic steatosis according to the BMI group. (o) Denotes estimated population-based prevalence of hepatic steatosis by the ultrasonographic criteria. 16,31,32 Fasting insulin and glucose values measured before the onset of treatment with metformin were used where these values were noted. Data were analyzed with NCSS and SAS (release 9.1; SAS Institute Inc, Cary, NC). Significance was determined for continuous variables by using the Mann-Whitney U test and for proportions by using 2 tests and Mantel-Haenszel tests, with.05. The study was approved by the Mount Sinai Hospital institutional review board ethics committee and conforms to the ethical guidelines of the 1975 Declaration of Helsinki. Results Eighty-eight women were identified who fit our inclusion criteria. Demographic and clinical data of the study population are reported in Table 1. Forty-three percent of the study population was found to be lean, 15% overweight, and 42% obese by BMI criteria. The overall prevalence of hepatic steatosis (mild, moderate, or severe) as assessed by abdominal ultrasonography was 55% (48/88 subjects). The presence of steatosis was associated with a greater median BMI (31.1 vs 24.3 kg/m 2, P.009). The prevalence of hepatic steatosis in the lean, overweight, and obese groups was 39%, 54% (P.366 vs lean), and 70% (P.007 vs lean, P.28 vs overweight), respectively. The prevalence and severity of steatosis in relation to BMI are shown in Figure 2, which also demonstrates that patients with higher BMI have the highest frequency of severe steatosis by ultrasonographic criteria. Sixty-four percent of women were receiving some form of pharmacotherapy at the time of evaluation and abdominal ultrasonography; 27% were being treated with metformin, 11% were being treated with oral contraceptives, and 25% were taking both. Treatment with oral contraceptives was associated with a lower prevalence of steatosis (38% vs 64%, P.015). The difference remained statistically significant when controlling for levels of BMI. The prevalence of steatosis was similar in those treated with metformin compared with those not treated with metformin (54% vs 55%, P.97). The higher BMI group was associated with increasing grade of steatosis, stratifying to treatment with oral contraceptives, metformin, neither, or both (Cochran-Mantel-Haenzsel test, P.0005). Review of the laboratory (Table 2) and imaging (Table 3) characteristics of the study population revealed that the presence of steatosis was also associated with higher premetformin HOMA-IR (3.53 vs 1.50, P.03) and a lower median fasting HDL cholesterol (54 vs 64 mg/dl, P.003), with patients in the higher HOMA-IR quartiles having significantly higher prevalence and higher severity of steatosis (Figure 3). Although there were trends toward greater premetformin fasting insulin, greater triglycerides, and increased prevalence of splenomegaly, hepatomegaly, and abnormal liver chemistries (Figure 4) in those with steatosis, these differences did not reach statistical significance with a P value of.05, in large part as a result of the small number of patients. In addition, there were no significant differences in median age, pre-metformin fasting glucose, total cholesterol, or low-density lipoprotein (LDL) cholesterol. Table 2. Laboratory Characteristics of the Study Population Variable All subjects Steatosis absent Steatosis present P value Fasting glucose (mg/dl) 86 (N 79) 86 (N 36) 86 (N 43).150 Fasting insulin (miu/l) 9.7 (N 79) 7.0 (N 36) 12.6 (N 43).059 HOMA-IR 2.04 (N 79) 1.50 (N 36) 3.53 (N 43).033 OGTT, 2-hr glucose (mg/dl) 99(N 65) 93 (N 27) 101 (N 38).315 OGTT, 2-hr insulin (miu/l) 46.3 (N 62) 42.0 (N 25) 53.9 (N 37).248 Presence of IFG, IGT, or DM 16 (24%) (N 67) 2 (7.7%) (N 26) 14 (34.1%) (N 41).013 Total cholesterol (mg/dl) 193 (N 88) 195 (N 40) 187 (N 48).443 HDL cholesterol (mg/dl) 60 (N 87) 64 (N 40) 54 (N 47).003 LDL cholesterol (mg/dl) 105 (N 85) 105 (N 40) 105 (N 45).792 Triglycerides (mg/dl) 97 (N 88) 92 (N 40) 107 (N 48).155 ALT (U/L) 17 (N 88) 15 (N 40) 20 (N 48).003 GGT (U/L) 17 (N 81) 14 (N 35) 19 (N 46).023 Elevated ALT or GGT 9 (10.2%) (N 88) 2 (5.0%) (N 40) 7 (14.6%) (N 48).139 NOTE. All data are medians or counts with proportions of the column using fasting laboratory specimens unless otherwise noted. P values reflect Mann-Whitney U tests or 2 tests, as appropriate. Percentages might not total to 100% because of rounding. OGTT, oral glucose tolerance test; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; DM, diabetes mellitus.

4 April 2007 INCREASED PREVALENCE OF NAFLD IN WOMEN WITH PCOS 499 Table 3. Imaging Characteristics of the Study Population Variable All subjects Steatosis absent Steatosis present P value Presence of splenomegaly 6 (6.8%) (N 88) 1 (2.5%) (N 40) 5 (10.4%) (N 48).142 Presence of hepatomegaly 4 (4.5%) (N 88) 0 (0%) (N 40) 4 (8.3%) (N 48).062 Presence of PCO morphology 64 (72.7%) (N 87) 28 (70.0%) (N 40) 36 (76.6%) (N 47).487 NOTE. All data are medians or counts with proportions of the column using fasting laboratory specimens unless otherwise noted. P values reflect Mann-Whitney U tests or 2 tests, as appropriate. Percentages might not total to 100% because of rounding. Discussion NAFLD describes a clinicopathologic condition that is characterized by significant lipid deposition in the hepatocytes of the liver parenchyma in patients with no history of excessive alcohol consumption. The pathologic features of NAFLD range from simple hepatic steatosis (fatty liver) at the most clinically benign end of the spectrum to cirrhosis at the opposite extreme. NASH is believed to be an intermediate stage of liver damage and is characterized by hepatic steatosis plus liver cell injury, death, and hepatic inflammation. Such liver damage predominates in perivenular regions of the hepatic acinus and often promotes a fibrogenic response that sometimes gradually progresses to bridging fibrosis and ultimately to cirrhosis. 1 In this cohort of premenopausal women with PCOS, we have found a high prevalence of NAFLD in both lean and obese patients. Unlike prior studies from India, China, Italy, Japan, Korea, and the United States with 3% 16% prevalence of NAFLD by ultrasound criteria in lean or general populations, 4,5,16,29 32 we report an almost 40% prevalence of steatosis in lean (BMI 25 kg/m 2 ), otherwise healthy women with PCOS. PCOS can often be a difficult diagnosis to make and might go undetected for long periods of time before timely diagnosis is made, and then subspecialty referral is made. Even though there might be a slight selection bias and the prevalence of NAFLD might be higher in our patients from a specialized endocrinology practice as compared with the general population, this is a well-described cohort of women with PCOS with stringent inclusion criteria. On the other hand, this study is consistent with previous work in the field, in which insulin resistance and Figure 3. Prevalence of steatosis by HOMA-IR quartile. Patients in the higher HOMA-IR (fasting glucose fasting insulin/450) quartiles had significantly higher prevalence and higher severity of steatosis. obesity appear to be important associated factors with NAFLD We found that the HOMA-IR score was significantly greater and HDL was significantly lower in subjects with steatosis than in those without steatosis. Both reflect greater insulin resistance in the subjects with steatosis. Most importantly, a substantial prevalence of NAFLD in lean women with PCOS might reflect the impact of insulin resistance in PCOS, whereas greater prevalence of NAFLD in the obese women with PCOS as compared with the lean women might reflect the additive effect of obesity in augmenting insulin resistance. We found a relatively low prevalence of hepatomegaly (8.3%) and liver enzyme abnormalities (15%) in subjects with demonstrated hepatic steatosis by ultrasonography. It has been shown previously, however, that advanced liver disease including cirrhosis might be present in the absence of abnormal liver chemistries 33 ; thus the presence of hepatic steatosis by ultrasound and normal liver chemistries in a large group of these young women is of concern because significant liver pathology might possibly go unrecognized in these women for extended periods of time. Recent study by Setji et al 24 certainly adds to the concern because all 6 women with elevated liver enzymes who underwent liver biopsy had evidence of NASH and pericellular (4/6) as well as bridging fibrosis (2/6) despite their young age (range, years; median, 28.5 years). Our study has a number of limitations. It is a retrospective study, and we do not have data on serologic work-up to exclude other causes of liver enzyme abnormalities, although subjects with known risk factors for liver disease (ie, alcohol, intravenous drug use, hepatotoxic medications) were excluded. The majority of our patients were white women, and thus our results might not be applicable to the general population. There is a lack of consensus on the diagnostic criteria for PCOS. In this study, we excluded patients who might be considered to have PCOS under the 2003 revised consensus criteria of the Rotterdam workshop 34 and on the basis of several studies identifying a subgroup of women with regular menstrual cycles but frequent anovulation. 35,36 Because women with menstrual dysfunction and hyperandrogenism are more insulin resistant than those with hyperandrogenism and regular cycles, 37,38 we might have found a higher prevalence of NAFLD in PCOS because we limited this study to a likely more insulin-resistant group. There was no uniformity in the pharmacologic treatment of PCOS; some of our patients were taking metformin, others were taking oral contraceptives, whereas others were either taking both or none of the medications. Although it appears that oral contraceptive use was associated with a lower prevalence of steatosis, this possibly reflects some reluctance for oral contraceptive treatment in very obese women with PCOS.

5 500 GAMBARIN-GELWAN ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 5, No. 4 Figure 4. Prevalence of elevated ALT or -glutamyltransferase (GGT) according to the BMI group and grade of hepatic steatosis. More obese women have abnormal liver enzymes compared with lean women (P.05). Patients with steatosis are more likely to have abnormal liver enzymes (P.01), as do patients with severe compared with moderate steatosis (P.05). Finally, because we used an abdominal ultrasound rather than a liver biopsy to identify prevalence of NAFLD, we were only able to identify presence of hepatic steatosis and not other parenchymal changes (such as steatohepatitis or fibrosis). The true prevalence of NAFLD in PCOS might thus have been underestimated in our study, because ultrasound has a sensitivity of 67% 80% (a specificity of 77% for steatosis) and does not reliably detect steatosis below 33% (as compared with liver biopsy). 39,40 Saadeh et al 40 also confirmed lack of detection of features of necroinflammation (ballooning and Mallory s hyaline) and fibrosis by the state-of-the-art imaging procedures. Because imaging test criteria for fatty liver disease are less sensitive for diagnosing advanced stage NAFLD than for detecting earlier stages of the disease when hepatic steatosis predominates, 1 a normal imaging result or one that shows minimal fat might significantly underestimate this disease as well. Currently, there are no noninvasive tests to accurately diagnose and stage NAFLD. Liver biopsy remains the most sensitive diagnostic test but cannot distinguish NAFLD from other causes of fatty liver disease, such as alcohol abuse. 41 Elevated serum aminotransferase levels, when present, might suggest a diagnosis of a more severe degree of NAFLD; these tests, however, are both insensitive and nonspecific indicators of steatosisrelated liver damage. For instance, the degree of liver enzyme elevation does not correlate with the level of damage on histologic analysis; thus, these tests cannot reliably distinguish steatosis from steatohepatitis or cirrhosis. 33,42 Serum aminotransferase levels might indeed be normal or only slightly elevated, even though liver disease is advanced. 43 As demonstrated by Sorrentino et al, 44 more than 70% of patients with metabolic syndrome and normal liver enzymes might have varying degrees of NASH, with 33% having evidence of fibrosis and 10% of cirrhosis. Liver biopsy in patients with hepatic steatosis on ultrasound and normal liver chemistries remains generally controversial and associated with clinical risks. Our study has shown that NAFLD is common in PCOS, regardless of BMI; however, as expected, obesity and insulin resistance are associated with even higher prevalence of NAFLD, similar to the studies in patients with metabolic syndrome and insulin resistance. 14,15 Despite varying degrees of hepatic steatosis, elevated liver chemistries are uncommon in PCOS. Because NASH is a risk factor for the development of cirrhosis and hepatocellular carcinoma, the high prevalence of hepatic steatosis in young lean and obese women with PCOS mostly in the absence of abnormal liver chemistries is of concern. It appears prudent to perform liver ultrasonography in patients with PCOS, regardless of their BMI and liver chemistry test results. Because insulin sensitizers such as peroxisome proliferatoractivated receptor agonists (thiazolidinediones) and metformin are part of the armamentarium to treat PCOS and possibly NASH, their use might be a viable option in PCOS patients with hepatic steatosis noted on ultrasonography. Because there might be appreciable histologic injury from fatty liver in patients with normal liver chemistry test results, recognition of hepatic steatosis and steatohepatitis in this often young age group of patients is warranted. Further controlled trials prospectively evaluating liver histology and presence of NASH and fibrosis are needed. References 1. Clark JM, Diehl AM. Nonalcoholic fatty liver disease: an underrecognized cause of cryptogenic cirrhosis. JAMA 2003;289: Caldwell SH, Oelsner DH, Iezzoni JC, et al. Cryptogenic cirrhosis: clinical characterization and risk factors for underlying disease. Hepatology 1999;29: Poonawala A, Nair SP, Thuluvath PJ. Prevalence of obesity and diabetes in patients with cryptogenic cirrhosis: a case-control study. Hepatology 2000;32: Clark JM, Brancati FL, Diehl AM. The prevalence and etiology of elevated aminotransferase levels in the United States. Am J Gastroenterology 2003;98: Ruhl C, Everhart J. Determinations of the association of overweight with elevated serum alanine aminotransferase activity in the United States. Gastroenterology 2003;124: Browning JD, Szczepaniak LS, Dobbins R, et al. Prevalence of

6 April 2007 INCREASED PREVALENCE OF NAFLD IN WOMEN WITH PCOS 501 hepatic steatosis in an urban population in the United States: impact of ethnicity. Hepatology 2004;40: Angulo P. Nonalcoholic fatty liver disease. N Engl J Med 2002; 16: Farell GC, Larter CZ. Nonalcoholic fatty liver disease: from steatosis to cirrhosis. Hepatology 2006;43(Suppl 1):S99 S Beymer C, Kowdley KV, Larson A, et al. Prevalence and predictors of asymptomatic liver disease in patients undergoing gastric bypass surgery. Arch Surg 2003;138: Matteoni CA, Younossi ZM, Gramlich T, et al. Nonalcoholic fatty liver disease: a spectrum of clinical and pathological severity. Gastroenterology 1999;116: Bugianesi E, Leone N, Vanni E, et al. Expanding the natural history of nonalcoholic steatohepatitis: from cryptogenic cirrhosis to hepatocellular carcinoma. Gastroenterology 2002;123: Ong JP, Younossi ZM. Is hepatocellular carcinoma part of the natural history of nonalcoholic steatohepatitis? Gastroenterology 2002;123: Friis-Liby I, Aldenborg F, Jerlstad P, et al. High prevalence of metabolic complications in patients with non-alcoholic fatty liver disease. Scand J Gastroenterol 2004;39: Marchesini G, Brizi M, Bianchi G, et al. Non-alcoholic fatty liver disease: a feature of the metabolic syndrome. Diabetes 2001; 50: Angelico F, Del Ben M, Conti R, et al. Insulin resistance, the metabolic syndrome, and non-alcoholic fatty liver disease. J Clin Endocrinol Metab 2005;90: Kim HJ, Kim HJ, Kwang EL, et al. Metabolic significance of non-alcoholic fatty liver disease in nonobese, nondiabetic adults. Arch Intern Med 2004;164: Knochenhauer ES, Key TJ, Kahsar-Miller M, et al. Prevalence of the polycystic ovary syndrome in unselected black and white women of the southeastern United States: a prospective study. J Clin Endocrinol Metab 1998;83: Diamanti-Kandarakis E, Kouli CR, Bergiele AT, et al. A survey of the polycystic ovary syndrome in the Greek island of Lesbos: hormonal and metabolic profile. J Clin Endocrinol Metab 1999; 84: Asunción M, Calvo RM, San Milan JL, et al. A prospective study of the prevalence of polycystic ovary syndrome from Spain. J Clin Endocrinol Metab 2000;85: Azziz R, Yildiz B, Woods KS, et al. The prevalence of polycystic ovary syndrome among unselected consecutive premenopausal women. J Clin Endocrinol Metab 2004;89: Carmina E, Legro RS, Stamets K, et al. Difference in body weight between American and Italian women with polycystic ovary syndrome. Human Reprod 2003;18: Schwimmer JB, Khorram O, Chiu V, et al. Abnormal aminotransferase activity in women with polycystic ovary syndrome. Fert Steril 2005;83: Brown AJ, Tendler DA, McMurray RG, et al. Polycystic ovary syndrome with nonalcoholic steatohepatitis: beneficial effects of modest weight loss and exercise on liver biopsy findings. Endocr Pract 2005;11: Setji TL, Holland ND, Sanders LL, et al. Nonalcoholic steatohepatitis and nonalcoholic fatty liver disease in young women with polycystic ovary syndrome. J Clin Endocrinol Metab 2006;91: Zawadsky JK, Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach. In: Dunaif A, Givens JR, Haseltine FP, Merriam GR, eds. Polycystic ovary syndrome. Boston: Blackwell Scientific Publications, 1992: Scatarige JC, Scott WW, Donovan PJ, et al. Fatty infiltration of the liver: ultrasonographic and computed tomographic correlation. J Ultrasound Med 1984;3: World Health Organization. Obesity: preventing and managing the global epidemic: WHO/NUT/98.1. Geneva, Switzerland: World Health Organization, Matthews DR, Hosker JP, Rudenski AS, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting to plasma glucose and insulin concentrations in man. Diabetologia 1985;28: Shen L, Fan J-G, Shao Y, et al. Prevalence of nonalcoholic fatty liver among administrative officers in Shanghai: an epidemiologic survey. World J Gastroenterol 2003;9: Singh SP, Nayak S, Swain M, et al. Prevalence of nonalcoholic fatty liver disease in coastal eastern India: a preliminary ultrasonographic survey. Trop Gastroenterol 2004;25: Omagari K, Kadokawa Y, Masuda J-I, et al. Fatty liver in nonalcoholic non-overweight Japanese adults: incidence and clinical characteristics. J Gastroenterol Hepatol 2002;17: Bellentani S, Saccoccio G, Masutti F, et al. Prevalence of and risk factors for hepatic steatosis in northern Italy. Ann Intern Med 2000;132: Mofrad P, Contos MJ, Haque M, et al. Clinical and histologic spectrum of nonalcoholic fatty liver disease associated with normal ALT values. Hepatology 2003;37: Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004;81: Carmina E, Lobo RA. Do hyperandrogenic women with normal menses have polycystic ovary syndrome? Fertil Steril 1999;71: Carmina E, Lobo RA. Polycystic ovaries in hirsute women with normal menses. Am J Med 2001;111: Eriksson S, Eriksson KF, Bondesson L. Nonalcoholic steatohepatitis in obesity: a reversible condition. Acta Med Scand 1986; 220: Robinson S, Kiddy D, Gelding SV, et al. The relationship of insulin sensitivity to menstrual pattern in women with hyperandrogenism and polycystic ovaries. Clin Endocrinol 1993;39: Graif M, Yanuka M, Baraz M, et al. Quantitative estimation of attenuation in ultrasound video images: correlation with histology in diffuse liver disease. Invest Radiol 2000;35: Saadeh S, Younossi ZM, Remer EM, et al. The utility of radiological imaging in nonalcoholic fatty liver disease. Gastroenterology 2002;123: Bravo AA, Sheth SG, Chopra S. Liver biopsy. N Engl J Med 2001;344: Ratziu V, Giral P, Charlotte F, et al. Liver fibrosis in overweight patients. Gastroenterology 2000;118: Brunt EM. Nonalcoholic steatohepatitis: definition and pathology. Semin Liver Dis 2001;21: Sorrentino P, Tarantino G, Conca P, et al. Silent non-alcoholic fatty liver disease: a clinical-histological study. J Hepatol 2004; 41: Address requests for reprints to: Maya Gambarin-Gelwan, MD, Weill Medical College of Cornell University, 450 E 69 St, New York, NY mag2046@med.cornell.edu; fax: (212) Presented in part at the Annual Meeting of the Endocrine Society ENDO 2005, San Diego, CA, June 5, 2005 (abstract P2-365).