Nonalcoholic fatty liver disease (NAFLD) is now recognized as a frequent cause of chronic liver disease in obese adults and is the most common cause of elevated serum aminotransferase levels among adults in the United States (1,2). NAFLD begins with steatosis, or fat deposition in hepatocytes, and is associated with excess caloric intake and insulin resistance (3,4). In susceptible populations, hepatic steatosis progresses to inflammation, or nonalcoholic steatohepatitis (NASH), and, eventually, to liver fibrosis and cirrhosis (5). Current data based on standard liver biopsy, suggest that hepatic steatosis is present in more than two thirds of obese adults (6) and more than 90% of morbidly obese adults (those >200% of their ideal body weight) (7). Progression to NASH occurs in 19% of obese adults and nearly 50% of morbidly obese adults (6,7). Approximately 30% of adults with NASH have some degree of fibrosis or cirrhosis (8).
Historically, NAFLD was thought to occur mainly in older, obese females (9). More recently, it has been recognized in both sexes (10,11) and among obese children. The increasing rate of childhood obesity worldwide may be responsible for the increased recognition of the hepatic complications of obesity (12). NASH is associated with elevated alanine aminotransferase (ALT) in obese children (13). Recent studies report elevated aminotransferases in 6% of overweight and 10% of obese U.S. teens (14), 25% of obese children in Italy (12) and 24% of obese children in Japan (15). Disturbingly, progression to NASH may be more common in pediatric NAFLD, and pediatric patients are not spared from progression to fibrosis and even cirrhosis (16,17). A study of obese children with steatosis and elevated aminotransferases found steatohepatitis in 88% and fibrosis in 71% (16).
NAFLD in adults is associated with obesity, insulin resistance and the So-called metabolic syndrome (18,19). Similarly, elevated ALT in overweight and obese adolescents is associated with increased age, rising glycosylated hemoglobin (HbA-1c), elevated triglycerides (TG) (14) and insulin resistance (19). In comparison to adults with obesity alone, fatty liver disease may be more common in adults with type 2 diabetes (T2DM). Salmela et al. reported that 23% of adult non-insulin-dependent diabetics had elevated ALT (20). In 1989, Silverman et al. studied obese adults with T2DM, reporting steatosis in 100%, steatohepatitis in 50% and cirrhosis in 19% (21). A recent study of 83 adults with T2DM found a 63% prevalence of fatty liver by computed tomographic scan (22).
The current prevalence of NASH in adult T2DM is unclear (23), and the prevalence of any form of NAFLD in children with T2DM is unknown (24). Further increasing the importance of better a understanding NAFLD in this population is the fact that medications currently used in T2DM may possibly be hepatotoxic. Conversely, these medications could also improve fatty liver by improving insulin resistance.
Based on studies in adults with T2DM, we hypothesized that the frequency of elevated aminotransferases in pediatric T2DM, as a surrogate clinical marker of NAFLD, would be higher than reported in obese children without T2DM. In addition, we hypothesized that the presence of elevated aminotransferases would be associated with markers of insulin resistance, such as elevated TG, low high density lipoprotein cholesterol (HDL) and high body mass index (BMI). To test these hypotheses, a chart review of pediatric patients with T2DM was performed with a focus on liver tests and metabolic parameters.
MATERIALS AND METHODS
All charts at the Barbara Davis Center for Childhood Diabetes (n = 2186 charts) were reviewed retrospectively to identify patients younger than 18 years of age with T2DM diagnosed between 1996 and 2002. T2DM was defined as outlined by American Diabetes Association criteria (25), with type 1 diabetes excluded by pancreatic autoantibody determination. Patients with secondary diabetes (e.g., cystic fibrosis or medication related diabetes) or maturity onset diabetes of the young were excluded. Those with alcohol consumption by history, medications with known hepatotoxicity (other than insulin sensitizers), other liver disease (hereditary, viral, autoimmune, biliary disease or thyroid disease) or laboratory tests drawn for the specific purpose of evaluating suspected gastrointestinal pathology were also excluded.
A total of 118 pediatric patients with T2DM meeting inclusion criteria were identified. All liver-related blood tests from these patients were recorded. Three patients were excluded because of aminotransferase elevation secondary to causes other than NAFLD, (one Alstrom Syndrome with renal and cardiac failure, one hepatitis A and one sodium valproate toxicity), leaving 115 subjects eligible for analysis. Mean age of these subjects was was 15.4 ± 0.4 years (range, 11 to 17.9 years). There were 76 females and 39 males (Table 1). Ethnicity was 36% Hispanic, 29% Caucasian, 24% African-American and 11% Asian. Mean BMI was 33 ± 2 kg/m2; mean A-1c 8.8% ± 0.5%; mean total cholesterol 202 ± 6 mg/dL; mean low density lipoprotein cholesterol (LDL) 117 ± 8 mg/dL; mean HDL 36 ± 3 mg/d; and mean TG 225 ± 65 mg/dL.
Of the 115 subjects, aminotransferase measurements were available in 42% (n = 48; 27 females, 21 males). In 65% of these (n = 31), aminotransferases were first tested at the time of T2DM diagnosis (Group I). In 35% (n = 17), aminotransferases were first tested more than 2 months after T2DM diagnosis (Group II).
HbA-1c was determined using the DCA 2000 Analyzer (Bayer Diagnostics Inc. Tarrytown, NY) with a non-diabetic range of 4.2% to 6.2% (26). Weight was measured on a balance scale and height by stadiometer.
Pancreatic autoimmunity, indicating type 1 diabetes, was excluded by assessment of glutamic acid decarboxylase and islet cell antibodies (simultaneous radioimmunoassay) and insulin autoantibodies insulin (radioimmunoassay utilizing quantification of precipitated 125I insulin) (27). All autoantibody assays were performed at the Barbara Davis Center for Childhood Diabetes. Lipid levels were obtained in the fasting state, and LDL values were calculated. Aspartate aminotransferase (AST) and ALT were measured in commercial laboratories as part of routine clinical care and were considered increased if they were above the normal range for each laboratory.
Student t-test and regression analysis using analysis of variance were used to assess differences between patients with elevated and normal aminotransferases. A P value of <0.05 was used to determine statistical significance. The t-tests were performed using Prism 4.0 (Graph Pad Software, San Diego, CA) for Windows and analysis of variance was performed using the SAS System for Windows version 8, release 8.02 TS level 02MO, Windows version 5.02.2195 (SAS, Cary, NC). Results are reported as mean ± standard error of the mean.
Aminotransferase measurements may be affected by diabetes treatment or by the metabolic derangements of uncontrolled disease. To control for these potential confounders, the group with aminotransferases measured at diagnosis (Group I) was evaluated separately from the group with initial liver enzyme measurements obtained more than 2 months after diabetes diagnosis (Group II).
Approval for the protocol was obtained from the Colorado Multiple Institutional Review Board.
The subjects in Group I had mean BMI 35 ± 2 kg/m2, mean A-1c 8.6% ± 0.4%, mean total cholesterol 177 ± 7 mg/dL, mean LDL 118 ± 12 mg/dL, mean HDL 38 ± 2 mg/dL and mean TG 205 ± 28 mg/dL (Table 2). Group I had a mean ALT of 68 ± 19 IU/L, a mean AST of 44 ± 10 IU/L (Table 2) and a 48% prevalence of elevated ALT.
In Group II, mean BMI was 30.2 ± 1 kg/m2, mean A-1c 8.4% ± 0.7%, mean total cholesterol 208 ± 9 mg/dL, mean LDL 120 ± 7 mg/dL, mean HDL 36 ± 4 mg/dL and mean TG 253 ± 118 mg/dL (Table 2). Group II had a mean ALT of 72 ± 16 IU/L, a mean AST of 45 ± 10 IU/L (Table 2) and a 47% prevalence of elevated ALT. Of the eight patients with elevated ALT, five were treated with diet therapy or insulin alone.
There were no significant differences between groups in age, BMI, A-1c, total cholesterol, LDL, HDL, TGs or aminotransferases. Thus, Groups I and II were pooled for further analysis. Of the pooled group of all patients with aminotransferase measurements, mean age was 15.0 ± 0.6 years, BMI was 33.3 ± 1.34 kg/m2, mean A-1c 8.5% ± 0.3%, mean total cholesterol 185 ± 3 mg/dL, mean LDL 119 ± 9 mg/dL, mean HDL 37 ± 2 mg/dL and mean TG 223 ± 25 mg/dL (Table 2). When the pooled group was compared to our entire population of subjects with T2DM (Table 1), there were no significant differences between the subjects who had determination of serum aminotransferases and the cohort as a whole.
In the pooled group, 48% (n = 23) had ALT elevations, with 29% (n = 14) more than two times the upper limit of normal. ALT was greater than AST in all subjects. Mean ALT was 69 ± 11 IU/L (range, 8 to 357 IU/L) and mean AST was 44 ± 6 IU/L (range, 10 to 214 IU/L) (Table 2). Among the elevated ALT measurements, 39% (n = 9) were one to two times above normal, 26% (n = 6) were two to three times normal and 35% (n = 8) were greater than three times above normal (Fig. 1).
The prevalence of elevated ALT in males (60%) was higher than in females (41%) but this difference was not statistically significant. No correlation was found between AST or ALT and age, BMI, A-1c, total cholesterol, LDL, HDL or TG.
Six patients with elevated ALT at diagnosis had measurements available before and after treatment with insulin sensitizers. Elevated ALT decreased in four patients after metformin therapy (mean therapy duration 13 months) and in one patient after pioglitazone therapy (12 months). No subjects on insulin sensitizers had new onset of elevated liver enzymes or worsening of liver enzyme elevations. Three subjects had liver biopsies; all were consistent with nonalcoholic steatohepatitis. One biopsy demonstrated prominent diffuse macrovesicular steatosis, patchy mixed inflammation and mild sinusoidal fibrosis, most evident around the central veins. Another biopsy showed prominent diffuse macrovesicular steatosis and peri-portal inflammation. A third showed marked diffuse macrovesicular steatosis, peri-portal inflammation and moderate peri-portal fibrosis. Five subjects had ultrasound, all showing a bright liver, consistent with steatosis.
A retrospective analysis of a cohort of pediatric patients with T2DM revealed a 48% prevalence of elevated serum aminotransferases, with 60% of these elevations being two or more times above the upper limit of normal. As predicted in our hypothesis, this prevalence is higher than the 10% rate reported among obese United States teens without diabetes (14). The prevalence reported here among pediatric patients is also higher than the 23% prevalence of elevated ALT reported in adults with “non-insulin dependent diabetes” in the 1980s. This discrepancy may be attributable to the inclusion of patients with type 1 diabetes in the adult study, as the diagnosis of diabetes type was unclear and the mean BMI was only 27 kg/m2 (20). In contrast, all of our patients had a definitive diagnosis of T2DM, with negative pancreatic autoantibodies and exclusion of secondary causes of diabetes. Alternatively, the prevalence in children may actually be higher than among adults, as the children in this study were more obese than adults with T2DM and therefore possibly more prone to NAFLD. A third possible explanation is that the United States population is now more obese than in the past (28), and thus the prevalence of elevated aminotransferase may have risen in all ages.
Because our study was retrospective, it is possible that the prevalence was overestimated as a result of bias in the selection of patients for determination of serum aminotransferases. In this cohort, aminotransferases were measured for a variety of reasons, most commonly as a screen before starting oral medications, as part of screening for research protocols or because of increasing awareness of NAFLD in this population. As NAFLD is a new diagnosis in pediatric T2DM clinics, aminotransferases have been measured more frequently in recent years. However, because there were no significant differences between subjects who did or did not have aminotransferases measured regarding age, BMI, A-1c, total cholesterol, LDL, HDL or TG, we feel that the results from measured patients likely reflect the entire cohort of pediatric T2DM patients at our center.
To minimize selection bias, we excluded from analysis any patients with known reasons to suspect liver pathology, abnormal liver tests other than elevated transaminases or laboratory values drawn for the purpose of evaluating suspected gastrointestinal pathology. As the majority of patients had elevated aminotransferases at baseline or while only receiving insulin therapy, the aminotransferase elevations were more likely to be caused by NAFLD than by oral insulin sensitizers.
Although our clinic follows more females than males with T2DM, male gender appears to be associated with aminotransferase elevations, as previously reported in children (29). Aminotransferase elevations did not correlate with total cholesterol, LDL, HDL, TG, age, BMI or A-1c. In contrast, elevated ALT in overweight and obese but nondiabetic adolescents is associated with increased age, increasing A-1C and elevated TG (14). Similarly, NAFLD in adults without diabetes is associated with insulin resistance and hypertriglyceridemia (19). The lack of association in this study may be a result of our small sample size or, alternatively, the relatively homogeneous population. For example, the average TG level was greater than 200, average HDL less than 40, and average BMI greater than 30 kg/m2, suggesting a more insulin-resistant group than might be seen in patients with obesity alone. Therefore, the variability in insulin sensitivity may have been insufficient to demonstrate differences in variables dependent on insulin resistance. Although we included a wide range of ages, age was not a predictor of elevated aminotransferases. This finding is worrisome, as it suggests that even very young patients with T2DM may be at risk for NASH.
Recent small studies have supported the idea that treatment of insulin resistance with insulin sensitizers may improve NAFLD in T2DM. Three months of rosiglitazone treatment reduced hepatic triglyceride content by 39% in adults with T2DM (30), and 12 weeks of pioglitazone decreased hepatic fat content by 47% in a similar group (31). Metformin therapy for 4 months in subjects with histologic NASH decreased mean ALT concentrations by 50% and liver volume on ultrasound by 20%, compared with a 14% decrease in ALT and less than 10% decrease in liver volume in subjects receiving only weight loss and exercise therapy (32). These studies require cautious interpretation, as aminotransferase changes do not always correlate with histologic improvement (33). Posttreatment biopsy showed only mild improvement in four of seven subjects treated with troglitazone despite normalization of the ALT in 7 of 10 obese women studied with histologic NASH (34). It is critical to better characterize baseline liver tests in T2DM and correlate the effects of diabetes therapy with prebiopsy and postbiopsy findings.
Our cohort is too small to evaluate the effect of diabetes therapy on aminotransferase elevations. However, a few observations are possible. Despite comparable glycemic control, several patients receiving insulin therapy alone had no improvement in aminotransferases, whereas liver enzymes improved in four patients treated with metformin and one patient receiving pioglitazone. Historically, the presence of elevated aminotransferases has been a contraindication to oral insulin sensitizers such as metformin and TZDs because of reported liver toxicity with troglitazone (35,36). However, severe liver toxicity does not appear to be a class effect of TZDs (37,38), and no case of metformin-induced hepatotoxicity has been reported despite more than 40 years of worldwide use (38). Thus, with appropriate monitoring of aminotransferases, metformin and TZDs deserve further study as a potential treatment for adolescents with evidence of NAFLD/NASH.
Because of the high prevalence of abnormal aminotransferases in pediatric T2DM, all pediatric patients with T2DM should have aminotransferase evaluated at time of diagnosis. In those with significant elevations, other causes of liver disease should be excluded, and further confirmation of a fatty liver should be sought with ultrasound or magnetic resonance imaging. Patients with severe or persistent elevations should probably undergo liver biopsy for definitive diagnosis, but further research is needed to determine indications for biopsy. Once other causes of liver disease are excluded, aminotransferases should be monitored on diabetes therapy. Weight loss and exercise are the cornerstone for improving insulin sensitivity. However, elevation of liver enzymes does not seem to be a contraindication for the use of metformin and TZDs, which may increase insulin sensitivity and decrease fatty infiltration of the liver. A larger prospective study with liver monitoring at baseline and at regular intervals will be needed to determine the impact of different therapies for T2DM therapy on liver enzymes and histologic NAFLD.
We would like to thank Dr. George Eisenbarth for the assessment of autoantibodies, and Dr. Boris Draznin and Dr. Ronald Sokol for their constructive comments.
1. Clark JM, Brancati FL, Diehl AM. The prevalence and etiology of elevated aminotransferase levels in the United States. Am J Gastroenterol
2. Daniel S, Ben-Menachem T, Vasudevan G, et al. Prospective evaluation of unexplained chronic liver transaminase abnormalities in asymptomatic and symptomatic patients. Am J Gastroenterol
3. Fong DG, Nehra V, Lindor KD, et al. Metabolic and nutritional considerations in nonalcoholic fatty liver
4. McGarry JD, Foster DW. Regulation of hepatic fatty acid oxidation and ketone body production. Annu Rev Biochem
5. Matteoni CA, Younossi ZM, Gramlich T, et al. Nonalcoholic fatty liver
disease: a spectrum of clinical and pathologic severity. Gastroenterology
6. Wanless IR, Lentz JS. Fatty liver
hepatitis (steatohepatitis) and obesity: an autopsy study with analysis of risk factors. Hepatology
7. Silverman JF, O'Brien KF, Long S. Liver pathology in morbidly obese patients with and without diabetes. Am J Gastroenterol
8. Falck-Ytter Y, Younossi ZM, Marchesini G. Clinical features and natural history of nonalcoholic steatosis syndromes. Semin Liver Dis
9. Ludwig J, Viggiano TR, McGill DB, et al. Nonalcoholic steatohepatitis: Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc
10. Ludwig J, McGill DB, Lindor KD. Review: nonalcoholic steatohepatitis. J Gastroenterol Hepatol
11. Bacon BR, Farahvash MJ, Janney CG, et al. Nonalcoholic steatohepatitis: an expanded clinical entity. Gastroenterology
12. Franzese A, Vajro P, Argenziano A, et al. Liver involvement in obese children: Ultrasonography and liver enzyme levels at diagnosis and during follow-up in an Italian population. Dig Dis Sci
13. Schwimmer JB, Deutsch R, Rauch JB, et al. Obesity, insulin resistance
, and other clinicopathologic correlates of pediatric nonalcoholic fatty liver
disease. J Pediatr
14. Strauss RS, Barlow SE, Dietz WH. Prevalence of abnormal serum aminotransferase values in overweight and obese adolescents. J Pediatr
15. Kawasaki T, Hashimoto N, Kikuchi T, Takahashi H, Uchiyama M. The relationship between fatty liver
and hyperinsulinemia in obese Japanese children. J Pediatr Gastroenterol Nutr
16. Rashid M, Roberts EA. Nonalcoholic steatohepatitis in children. J Pediatr Gastroenterol Nutr
17. Nadeau K, Klingensmith G, Sokol RJ. Case report: nonalcoholic steatohepatitis in a teenage girl with type 2 diabetes. Curr Opin Pediatr
18. Marchesini G, Brizi M, Bianchi G, et al. Nonalcoholic fatty liver
disease: a feature of the metabolic syndrome. Diabetes
19. Marchesini G, Brizi M, Morselli-Labate AM, et al. Association of nonalcoholic fatty liver
disease with insulin resistance
. Am J Med
20. Salmela PI, Sotaniemi EA, Niemi M, et al. Liver function tests in diabetic patients. Diabetes Care
21. Silverman JF, Pories WJ, Caro JF. Liver pathology in diabetes mellitus and morbid obesity: Clinical, pathologic, and biochemical considerations. Pathol Annu
1989;24 Pt 1:275-302.
22. Kelley DE, McKolanis TM, Hegazi RA, et al. Fatty liver
in type 2 diabetes mellitus: relation to regional adiposity, fatty acids, and insulin resistance
. Am J Physiol Endocrinol Metab
23. Subramanian G, Walmsley D, Blewitt RW. Gliclazide-induced hepatitis. Pract Diab Int
24. Day CP. Non-alcoholic steatohepatitis (NASH): where are we now and where are we going? Gut
25. American Diabetes Association: clinical practice recommendations 2002. Diabetes Care
2002;25 (Suppl 1):S1-147.
26. Chase HP, Lockspeiser T, Peery B, et al. The impact of the diabetes control and complications trial and humalog insulin on glycohemoglobin levels and severe hypoglycemia in type 1 diabetes. Diabetes Care
27. Verge CF, Stenger D, Bonifacio E, et al. Combined use of autoantibodies (IA-2 autoantibody, GAD autoantibody, insulin autoantibody, cytoplasmic islet cell antibodies) in type 1 diabetes: Combinatorial Islet Autoantibody Workshop. Diabetes
28. Mokdad AH, Ford ES, Bowman BA, et al. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA
29. Baldridge AD, Perez-Atayde AR, Graeme-Cook F, et al. Idiopathic steatohepatitis in childhood: a multicenter retrospective study. J Pediatr
30. Mayerson AB, Hundal RS, Dufour S, et al. The effects of rosiglitazone on insulin sensitivity, lipolysis, and hepatic and skeletal muscle triglyceride content in patients with type 2 diabetes. Diabetes
31. Bajaj M, Suraamornkul S, Pratipanawatr T, et al. Pioglitazone reduces hepatic fat content and augments splanchnic glucose uptake in patients with type 2 diabetes. Diabetes
32. Marchesini G, Brizi M, Bianchi G, et al. Metformin in non-alcoholic steatohepatitis. Lancet
33. Angulo P, Keach JC, Batts KP. Independent predictors of liver fibrosis in patients with nonalcoholic steatohepatitis. Hepatology
34. Caldwell SH, Hespenheide EE, Redick JA, et al. A pilot study of a thiazolidinedione, troglitazone, in nonalcoholic steatohepatitis. Am J Gastroenterol
35. Menon KVN, Angulo P, Lindor KD. Severe cholestatic hepatitis from troglitazone in a patient with nonalcoholic steatohepatitis and diabetes mellitus. Am J Gastroenterol
36. Graham DJ, Drinkard CR, Shatin D. Incidence of idiopathic acute liver failure and hospitalized liver injury in patients treated with troglitazone. Am J Gastroenterol
37. Lebovitz HE, Kreider M, Freed MI. Evaluation of liver function in type 2 diabetic patients during clinical trials: evidence that rosiglitazone does not cause hepatic dysfunction. Diabetes Care
38. Albright ES, Bell DSH. The liver, liver disease and diabetes mellitus. Endocrinologist