Is Acanthosis Nigricans a Marker of Insulin Resistance in Obese Children?

Acanthosis nigricans (AN) was proposed as an insulin resistance marker and an independent risk factor for type 2 diabetes (1). A number of studies had associated AN with insulin resistance (IR) and much higher prevalence of type 2 diabetes in childhood. (2)

Although children with AN are often obese, only few studies have considered the roles of BMI and AN as independent markers of IR. The purpose of this research was to compare several indexes of IR (i.e., homeostasis model assessment of IR [HOMA-IR], insulin-like growth factor binding protein 1 [IGFBP1] levels, and base insulinemia) in obese children with and without AN and to determine the rate of association between AN and BMI, HDL, triglycerides, and other predictors of type 2 diabetes.

A total of 1,250 Hispanic subjects (mean age 12.4 ± 1.4 years) who consulted the pediatric department between April and November 2001 were evaluated; 288 of these children were obese (BMI ≥95th percentile). Of these children, we took a randomized sample of 74 obese children (40 girls). Data for birth weight (BW), positive family history for obesity and/or type 2 diabetes, BMI, presence of AN, blood pressure, and Tanner stage were obtained. An oral glucose tolerance test (OGTT) and measurements of lipid profile, insulinemia, and IGFBP1 were performed. All of the children were of Tanner stage ≥2 (all were in puberty) and had a positive family history. There was a high rate of AN (n = 41; 55.4%). There was no statistical difference regarding age and sex between the group with AN (n = 41) and the group without AN (n = 33). In the group with AN, four were glucose intolerant; in the group without, only two were glucose intolerant. None presented type 2 diabetes. A Student’s t test was used to compare both groups. The group with AN showed a statistical difference with the other group in BMI (30.6 vs. 27.3 kg/m², P = 0.00039), basal glucose (5.3 vs. 5.0 mmol/l, P = 0.01), HDL (39.2 vs. 45.1 mg/dl, P = 0.02), and BW (3.23 vs. 3.61 kg, P = 0.0021). AN showed an univariate association with BMI (r² = 0.45, P = 0.00038), BW (r² = −0.37, P = 0.0021), basal glucose (r² = 0.30, P = 0.009), and HDL (r² = −0.25, P = 0.03). There was no difference for all the IR indexes (HOMA-IR [6.6 vs. 4.9, P = 0.19], base insulinemia [27.3 vs. 21.5, P = 0.27], IGFBP1 [8.2 vs. 8.3, P = 0.98]) between the two groups. Likewise, there was no univariate association between AN and the markers of IR (base insulinemia [r² = 016, P = 0.16], HOMA-IR [r² = 0.2, P = 0.06], and IGFBP1 [r² = 0.07; P = 0.69]). Even though there were greater fasting insulin levels and HOMA-IR in the group with AN, the difference between both groups was not significant. Consistent with these results, a previous study showed that even though fasting insulin levels and HOMA-IR in obese children with AN were twice as high as those without AN; after adjusting for fat mass, there was no difference between both groups (3). The presence of AN showed a positive correlation with BMI (odds ratio [OR]1.30, 95% CI 1.08–1.57; P = 0.018) and a negative correlation with BW (OR 0.23, 95% CI 0.07–0.71; P = 0.03) in the multivariate analysis. BMI of subjects in the group with AN was significantly greater than in the group without AN, suggesting that AN may reflect only increased obesity. We conclude that AN predicts obesity and is not an independent marker of IR in our population.