In this population of obese, insulin resistant adolescents, we indeed found that the presence of the MS was associated with an increased prevalence of IGT and markers for NAFLD. Presence of the MS using the modified Cook definition was extremely common in subjects with morbid obesity (BMI > 40 kg/m2) but was also present in more than half of these obese adolescents with a BMI < 40 kg/m2. Since all of these obese subjects met waist circumference criteria for MS, adolescents with the presence of 2 or more additional MS criteria were significantly more likely to also have abnormal glucose regulation and evidence of more severe NAFLD. Of concern, even of subjects without evidence of MS, nearly half had evidence of NAFLD, either ALT > 40 U/l or fatty liver by US.
Since all of our subjects were obese, waist circumference was not helpful in differentiating subjects with increased risk of co-morbidities. However, elevated waist circumference may be helpful in the overweight category, where athletes with increased muscle mass may be incorrectly identified as overweight by BMI criteria alone. A study of waist circumference in 6–13 year old children showed that only 28% of children with BMI 85–95% have a waist circumference > 90%, compared to 87% of those with a BMI > 95%. [20
] A large study of almost 2000 children in Greece showed waist circumference to be the most significant predictor for cardiovascular disease risk factors (total cholesterol, triglycerides, high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C) and blood pressure). [21
] Another study of 2597 American youth showed that within a given category of weight, subjects with high waist circumference were more likely to have elevated triglycerides, elevated insulin levels, and the MS than those in the low waist circumference group in the same weight category. [22
] A recent Australian study did not find waist circumference at age 8 years to be helpful in predicting metabolic risk at age 15. [23
] Thus, although universal measurement of waist circumference is unnecessary, measuring waist circumference may be a useful part of the assessment in the subgroup of muscular adolescent athletes with BMI in the 85–95% range. Reference values are available for 90% cut-offs for waist circumference in European-American, African-American and Mexican-American adolescents. [24
Although obesity has been shown to be a risk factor for NAFLD in adults, [25
] insulin resistance appears to be even more tightly correlated with NAFLD. An Italian study of 46 adults with NAFLD showed insulin resistance to be the strongest predictor of NAFLD, with fasting insulin levels nearly twice as high in patients with NAFLD as in controls. [26
] NAFLD is also associated with an elevated serum triglycerides [27
], low HDL-C [27
], abnormal glucose regulation [29
], and central adiposity [27
], all features of the metabolic syndrome. A study of 30 non-obese adults with NAFLD and normal glucose tolerance showed these subjects to have significant insulin resistance and features of the MS, again implicating insulin resistance, even without obesity. [30
] In our study population, all subjects had evidence of insulin resistance, which may explain the high rates of NAFLD even in those subjects without the MS.
As IFG or IGT are combined as one of the 5 MS criteria, IGT in this study is a component of the diagnosis of MS and also is an outcome measure. If we focused only on results available from the baseline fasting sample to diagnose the MS, 6 of the 7 subjects with IGT alone still met 3 or more criteria for the metabolic syndrome. When the data are analyzed with that one subject changing categories (from 3 or more to < 3 criteria), MS as defined by fasting laboratory values alone is still predictive of IGT (p = 0.019).
Elevated ALT and fatty liver on liver US, while not the gold standard for diagnosis of NAFLD, can provide non-invasive evidence of NAFLD, limiting liver biopsies in pediatric patients. [31
] US has been shown to identify fatty liver with a sensitivity of 89–95% and a specificity of 84–93%. [32
] Current NAFLD literature uses a cut-off for ALT of 40 U/l to define elevated ALT. [14
]. However, clinical laboratories have different cut-offs for the upper limit of normal for ALT as well as other laboratory markers of NAFLD. Using the cut-off of 40 U/l for ALT rather than the upper limit of normal for our laboratory improved the predictive value of 3 or more MS criteria for identifying elevated ALT, however this did not reach statistical significance. ROC curves were used to determine which cut-off for ALT was associated with metabolic syndrome in our population; presence of the metabolic syndrome was associated with an ALT > 35 U/l. Of 5 black subjects with ALT > 40 U/l, only 2 had fatty liver on US. No black subjects had ALT above the upper limit of normal for our laboratory. In contrast, of 17 Hispanic subjects with ALT > 40 U/l, 13 (76%) had fatty liver on ultrasound. Previous research has shown a decreased prevalence of NAFLD in black obese children. [33
Presence of the MS was associated with an increase in fatty liver on US, as well as in increase in ALT and GGT. The majority of the NAFLD literature reports elevated ALT as more highly associated with NAFLD than GGT. In our subjects however, the presence of MS predicted elevated GGT as well as ALT. A large Italian study of obese adults also showed elevated GGT to be associated with elevated triglycerides and hyperglycemia. [34
]. In addition, a study of severely obese adults undergoing bariatric surgery showed a decrease in GGT to be the best predictor of improvement in inflammation, fibrosis, and non-alcoholic steatohepatitis (NASH). [35
This group of adolescents was selected for insulin resistance, thus these results may be different in unselected obese adolescents. In addition, surrogate markers were used to measure insulin sensitivity, rather than a gold standard such as an insulin clamp. Many of the MS criteria vary depending on race and ethnicity. The subjects in this study were largely Hispanic and black, therefore the results may be different in other racial groups. Subjects with ALT > 3 times the upper limit of normal were excluded, therefore our results may underestimate the prevalence of NAFLD in youth with the MS, although the one subject initially excluded for ALT was confirmed to have NAFLD by biopsy and later enrolled in the study after his ALT improved. In addition, undocumented alcohol use could have contributed to GGT elevations. Finally, not all of our subjects had ultrasounds available for analysis and US itself is not the most sensitive tool for assessing fatty liver, especially in severely obese subjects. However, due to the young age of the subjects, US is preferable in avoiding the risk of biopsy or radiation exposure associated with CT scans, and allows shorter imaging times when compared with magnetic resonance imaging.
In the pediatric population, identification of MS based on fasting results should lead the provider to monitor patients more closely for abnormal glucose regulation and NAFLD. Patients with extreme abnormalities in individual MS criteria (for example, triglycerides > 400 mg/dl or severe hypertension) should be evaluated and treated as necessary. Separate from the importance of individual components, the diagnosis of MS may have several beneficial effects. The diagnosis of MS may facilitate identification by the primary care provider of abnormal glucose regulation or NAFLD. These diagnoses may lead the provider to consider additional treatment modalities beyond what would likely be used for obesity alone. Early identification and intervention for these disorders may minimize long term morbidity and mortality. In addition, patients and/or parents may be more likely to make significant lifestyle changes if there is a specific diagnosis rather than just being told that they are overweight.
In contrast, in this era of cost containment, obese patients without evidence of the MS may require less frequent monitoring for abnormal glucose regulation, especially if they do not continue to gain weight over time. Future research should assess longitudinal risk of development of insulin resistance related co-morbidities in obese youth with and with out the MS.