icon star paper   Articles  
Back grey_arrow_rt.gif
 
 
Alanine Aminotransferase Levels Predict Insulin Resistance in HIV Lipodystrophy
 
 
  JAIDS Journal of Acquired Immune Deficiency Syndromes 2003; 34(5):534-536
 
Raymond T. Chung, MD; Deborah R. Casson; Gisela Murray; Sunbin Song; Steven Grinspoon, MD; Colleen Hadigan, MD, MPH. Gastrointestinal Unit, Department of Medicine Massachusetts General Hospital, Harvard Medical School, Boston, MA (Dr Chung, Ms Casson) and The Program in Nutritional Metabolism, Massachusetts General Hospital, Harvard Medical School, Boston, MA (Ms Murray, Ms Song, Dr Grinspoon, Dr Hadigan)
 
"...Here, ALT was a strong positive predictor of insulin resistance, independent of hepatitis status and with measures of impaired glucose tolerance independent of viral hepatitis….. HCV co-infection was a predictor of insulin and glucose response to OGTT…HCV & HBV may exacerbate glucose intolerance-- interpretation is limited in part by the relatively small number of subjects with HBV and HCV co-infection in our cohort…. Our data demonstrate that even small increases in ALT are directly predictive of impaired insulin sensitivity in patients with HIV-related fat redistribution and that hepatitis co-infection may significantly increase the risk of insulin resistance. ALT levels were strongly associated with indices of glucose and insulin metabolism independent of viral hepatitis and may be a useful clinical indicator for insulin resistance in patients with HIV and lipodystrophy."
 
To the Editor:
 
HIV-associated lipodystrophy is characterized by fat redistribution, hyperlipidemia, and insulin resistance. Hepatic injury related to HIV infection, drug toxicity, hepatitis, or fatty infiltration of the liver may contribute to insulin resistance observed in HIV lipodystrophy. For example, increasing evidence in patients with chronic hepatitis shows that hepatocellular damage is directly related to the development of insulin resistance and diabetes mellitus. A strong relationship between alanine aminotransferase (ALT) elevations, fatty infiltration of the liver, and insulin resistance was identified in the setting of obesity. In the US National Health and Nutrition Examination Survey, abnormal ALT was positively correlated with insulin levels in obese subjects. Here, we evaluate the possible links between insulin resistance in HIV lipodystrophy, and markers of liver disease, including presence or absence of viral hepatitis.
 
METHODS
 
Subjects
 
Subjects were physician- or self-referred for evaluation of fat redistribution. Subjects were asked if they had experienced loss of facial fat, increased fat in the neck, increased abdominal girth, increased chest/breast fat, or loss of fat in the arms or legs. Those who identified a change in >1 body areas were invited to participate; fat redistribution was confirmed by a physical examination for all subjects by a single investigator (CH). Informed consent was obtained from each subject. Metabolic and nutritional data from a subset of this group were reported previously.
 
Preexisting diabetes was exclusionary. Each subject had a complete history and physical examination. After a 12-hour overnight fast, a standard 75-g oral glucose tolerance test (OGTT) was performed. Fasting total, low-density, and high-density cholesterol and triglyceride levels were obtained. Diabetes was defined as a fasting glucose > 126 mg/dL or a 2-hour postchallenge glucose > 200 mg/dL. Subjects were not excluded based on current or past alcohol consumption.Thirteen reported a history of alcohol abuse but current abstinence; no subject reported current alcohol abuse. Liver transaminases, alkaline phosphatase, and total bilirubin were determined, as well as hepatitis B surface antigen (HBsAg) and anti-hepatitis C virus (HCV) antibody. If a subject was anti-HCV antibody positive, HCV RNA was measured by polymerase chain reaction. Insulin resistance was assessed with the homeostatic model for assessment of insulin resistance (HOMA-IR), a mathematical estimate of insulin resistance from values of fasting glucose and insulin.
 
Statistical Analysis
 
Multivariate linear regression analyses were used to determine predictors of insulin resistance, glucose, and insulin response to OGTT. Age, sex, body mass index (BMI), waist-to-hip ratio, current protease inhibitor (PI) use, ALT, and viral hepatitis status were included in the model. Potential interactions between ALT and hepatitis status were also assessed.
 
RESULTS
 
Ninety-one subjects between the ages of 26-60 years with evidence of fat redistribution were evaluated. The majority of subjects presented with evidence of both lipoatrophy and central fat accumulation (67%); 16% had lipoatrophy only and 16% had fat accumulation only. Of the 91 subjects, 5 were positive for HBsAg, 9 were positive for HCV RNA by polymerase chain reaction, and no one had both. Overall, the mean bilirubin, ALT, and alkaline phosphatase levels were not elevated. However, the mean aspartate aminotransferase (AST) was increased slightly above the upper limit of the normal range (10-37 U/L).
 
In a multivariate analysis adjusted for age, sex, fat distribution (i.e., waist-to-hip ratio), BMI, and PI use, ALT was a strong positive predictor of HOMA-IR (P < 0.0001), as were age (P = 0.0003) and BMI (P = 0.005). ALT was a significant positive predictor of fasting glucose (P = 0.0008), 2-hour glucose (P = 0.003), fasting insulin (P < 0.0001), and insulin area under the curve (AUC) following OGTT (P = 0.001) using the same model parameters to adjust for hepatitis and other potential confounders.
 
HCV was associated with increased insulin AUC (P = 0.008) and 2-hour glucose (P = 0.03). The interaction between hepatitis B virus (HBV) and ALT was a strong positive indicator of fasting glucose (P = 0.01) and fasting insulin (P < 0.0001). PI use was positively associated with insulin AUC (P = 0.02) and 2 hour glucose (P = 0.02).
 
Clinical Characteristics of 91 Subjects with Lipodystrophy
 
Men/women- 65/26 Age: 43 Duration of HIV: 7.2 yrs
Duration of ARV: 4.7 yrs
HIV RNA: 65% <400 c/ml
BMI (kg/m2): 26.2
Waist-to-hip ratio: 0.97
Fasting glucose (mg/dL): 92
Fasting insulin: (uIU/L): 19
HOMA IR: 4.6
Cholesterol: 226 mg/dL
Triglycerides: 332 mg/dL
LDL: 131 mg/dL
HDL: 41 mg/dL
AST: 51 U/L
ALT: 32 U/L
Alkaline phosphatase: 81 U/L
Bilirubin: 0.7 mg/dL
HCV+ antibody: 13 (14%)
HCV RNA+: 9 (10%)
HBV antibody+: 5 (5%)
 
Three of 5 HBV-positive subjects and 1 HCV-positive subject had previously unrecognized diabetes. As such, HBV-positive subjects had significantly increased fasting glucose levels and increased 2-hour OGTT glucose compared with HIV/HCV subjects and HIV-positive controls (Table 2). Fasting insulin and HOMA-IR were also significantly different between HIV/HBV and HIV-positive controls. Liver transaminases did not differ between groups, but bilirubin was significantly increased in HBV.
 
Table 2. Hepatic and Metabolic Oarameters According to Viral Hepatitis Status
 
 
 
  HIV+ HIV/HBV HIV/HCV
Subjects 77 5 9
Age 43 46 47
BMI (kg/m2) 25.9 25.3 29.3
WHR 0.97 0.96 0.99
ALT 30 31 39
AST 47 67 72
Alk Phos 81 83 76
Bilirubin 0.6 1.4 0.7
Fasting glucose 90 114 93
2-hr glucose 132 195 133
Fasting insulin 17 39 23
HOMAIR 4.0 12.8 5.4
Diabetes 6.5% 60% 11%
 
 
  AUTHOR DISCUSSION
 
DISCUSSION
 
Previously, Duong et al reported a relationship between insulin resistance and HCV among HIV/HCV-positive patients with ALT elevations. Here, ALT was a strong positive predictor of insulin resistance, independent of hepatitis status and controlling for age, BMI, and PI use. ALT was also positively associated with measures of impaired glucose tolerance independent of viral hepatitis. These data coincide with reports that increased ALT is predictive of the development of diabetes and decreased hepatic insulin sensitivity in Pima Indians. One possible mechanism for the relationship between ALT, chronic viral hepatitis, and insulin resistance is hepatic steatosis. Steatosis is a common feature in congenital lipodystrophy, and insulin resistance is present in up to 98% of cases of biopsy-proven steatohepatitis. While investigators have speculated that metabolic symptoms in HIV lipodystrophy include steatosis, there are limited imaging or biopsy data available on the true incidence of hepatic steatosis in this population. Sutinen et a identified increased liver fat in 25 HIV-positive men with lipodystrophy and showed a positive correlation between liver fat content and fasting insulin levels. In mice, exposure to ritonavir led to increased lipids, as well as significant hepatomegaly and fatty liver. HIV-infected patients treated with PI-containing highly active antiretroviral therapy regimens and those experiencing fat redistribution may be at particular risk for steatosis and insulin resistance.
 
HCV co-infection was a predictor of insulin and glucose response to OGTT. Hepatocellular damage and cirrhosis in chronic viral hepatitis have been implicated in the development of insulin resistance. Increased rates of diabetes are noted in HCV in the absence of HIV infection. Our data are in agreement with observations in HIV and non-HIV-infected individuals and suggest potential effects of HCV and HBV to exacerbate glucose intolerance. However, interpretation is limited in part by the relatively small number of subjects with HBV and HCV co-infection in our cohort.
 
Fat redistribution may confer additional hepatic injury and thereby place HIV-positive patients with lipodystrophy or hepatitis co-infection at particularly high risk for insulin resistance.
 
 
 
 
  icon paper stack View Older Articles   Back to Top   www.natap.org