Effects of Conjugated Linoleic Acid and Metformin on Insulin Sensitivity in Obese Children

Randomized Clinical Trial

Nayely Garibay-Nieto; Gloria Queipo-García; Flor Alvarez; Mayra Bustos; Erendira Villanueva; Fernando Ramírez; Mireya León; Estibalitz Laresgoiti-Servitje; Ravindranath Duggirala; Teresa Macías; Sergio Cuevas; Abel Jalife; Miguel Fonseca-Sánchez; Fabiola Serratos; Juan Carlos López-Alvarenga


J Clin Endocrinol Metab. 2017;102(1):132-140. 

In This Article


This study supports that CLA improves insulin sensitivity, as measured by EHCT in a group of obese children, and exceeds LIP benefits.

Because the prevalence of metabolic syndrome in our pediatric clinic averages 35% and confers an 11-fold risk of diabetes during early adult life,[19] the exploration of conventional and pharmacological strategies focusing on improving the insulin sensitivity level is imperative. Recent studies have revealed that MET has important effects on insulin sensitivity when compared with PLB, and its use in nondiabetic, obese individuals has been massively extended.[6,20,21] A systematic review conducted by Brufani et al.[6] revealed a significant but moderate benefit of MET on weight reduction and fasting insulin sensitivity compared with PLB or lifestyle interventions alone. Nonetheless, when these outcomes were evaluated by frequently sampled intravenous glucose tolerance test[22] or hyperglycemic clamp technique,[23] no significant differences were reported. Wiegand et al.[5] demonstrated in a randomized PLB-controlled trial a beneficial effect of MET on the insulin sensitivity index in obese, insulin-resistant adolescents; however, no differences in body composition, weight, or BMI were found. We found in our study a significant improvement in all anthropometric parameters, including weight, BMI, waist circumference, and body composition (fat mass and fat-free mass, data not shown); nonetheless, these results were not significantly different among the treatment groups. To our knowledge, no randomized PLB-controlled trial using EHCT had been executed for the evaluation of MET benefits on insulin sensitivity in children. In the present study, we found no differences on Rd value (mg/kg/min) when comparing MET and PLB in the postintervention period. These data could be consistent with the final results of the Diabetes Prevention Program Research Group[24] that showed that diabetes incidence was better reduced in a LIP group compared with PLB. Nonetheless, improvement in terms of BMI, waist circumference, HDL cholesterol, and triglycerides, with considerable effect sizes (72%, 65%, 37%, and 55%, respectively), favored patients treated with MET in our study.

Several studies have proposed beneficial effects of CLA isomers on body composition, inflammation, and insulin sensitivity, promoting differentiation, lipid metabolism regulation, and apoptotic mechanisms in adipocytes.[10–12] Interestingly, evidence has suggested that the trans-10,cis-12 isomer of CLA might induce insulin resistance, whereas the CLA mixture has beneficial effects on body composition and insulin sensitivity. Risérus et al.[25] demonstrated that trans-10,cis-12 isomer–treated subjects presented insulin and glucose increases and decreased HDL and insulin sensitivity measured by 2-hour EHCT compared with PLB or CLA mixture-treated groups. No differences were observed when comparing PLB and CLA mixture-treated individuals. In our study, we showed that the CLA mixture was associated with a clinically relevant effect size (37%) over the Rd value of insulin sensitivity. The confounding variables included in the ANCOVA model showed a decline on group differences. Among these adjustments, Tanner stage was the main variable that modified insulin sensitivity, and despite our small sample size, the effect size of CLA on the Rd value remained.

Despite the fact that several CLA isomers might have deleterious effects on insulin sensitivity and resistance, certain mixtures may neutralize negative effects and even induce a synergistic positive response on these parameters, as well as on metabolic and anthropometric values. Some effects of the trans-10,cis-12 isomer promote a blunted glucose uptake that depends on decreased expression of glucose transporter 4 (GLUT-4).[26] Moreover, decreased incorporation of free fatty acids into the cells may be induced by CLA, a mechanism that could be related to diminished expression of peroxisome proliferator-activated receptor-γ in adipocytes.[27] CLA has been proposed as an apoptotic accelerator of adipocytes inmammals that liberates and increases fatty acid oxidation elsewhere in the body.[28] Evidence of deleterious effects has mainly been reported in animal models, in which administered doses of CLA are superior to those used in humans (0.2 to 3 g/kg vs 0.015 to 0.1 g/kg, respectively).[29] These effects, if present in humans, could be hyperglycemia and hyperlipidemia, which may predispose an individual to diabetes and nonalcoholic fatty liver disease.[30,31] However, few studies have been published regarding the molecular mechanisms of CLA in skeletal muscle that could explain increased glucose uptake in our treated patients. On this matter, Vaughan et al. (32), using a rabdomyosarcoma cell line, have reported that omega-3 fatty acids and CLA activate mitochondrial proliferation and glycolytic activation pathways probably by apoptosis induction and subsequent upregulation of GLUT-4. Furthermore, animal models have shown the beneficial effects of CLA on insulin sensitivity and overexpression of peroxisome proliferator-activated receptor-γ and GLUT-4 in the muscle of supplemented rats.[27] In the present study we were able to demonstrate that postintervention IRS2 expression in the skeletal muscle was significantly upregulated in CLA-treated patients. To our knowledge, no studies have been published regarding the effects of CLA or CLA–isomer mixtures on insulin receptor substrate molecules. Xu et al.[33] reported that MET upregulates insulin receptor β expression and the downstream IRS2/phosphatidylinositol 3-kinase/Akt signaling transduction in an insulin-resistant rat model of nonalcoholic steatohepatitis and cirrhosis. Our results evidenced a nonsignificant but marginal (P = 0.055) IRS2 upregulation in MET-treated children. The insulin-sensitizing effects of MET have been mainly described in liver tissue. Although CLA effects have been mainly focused on adipose tissue modeling, the present study demonstrates that molecular mechanisms, particularly IRS2 upregulation, might mediate insulin-sensitizing effects on skeletal muscle. This phenomenon could explain the increased glucose infusion rate tolerability in our patients treated with CLA throughout the EHCT. Moreover, significant HOMA-IR improvement observed only in CLA-treated patients denotes a significant performance in skeletal muscle that promotes a lower pancreatic insulin secretion.

A recently published meta-analysis demonstrated that the deleterious effects of CLA consumption might be negligible, whereas its benefits, although subtle, seem to be clinically relevant regarding weight and fat mass loss.[34] In our study, BMI improvement was significant in all groups, although not significantly different among them. Nonetheless, MET displayed the highest effects over BMI (72%, compared with 43%in the PLB group and 41% in CLA-treated patients) and waist circumference (70%, compared with 60% in PLB and 30% in the CLA group). Total body fat did not improve in any group, but leptin levels significantly decreased in all patients (P<0.014, data not shown).

Racine et al. (10) reported a clinical trial in a pediatric population randomly assigned to CLA (3 g/d, c9,t11-t10- c12, 50:50) or PLB for 6 months that showed a decrease in total body fat in the CLA group and a significant decrease in HDL cholesterol levels in CLA-treated patients. Our trial demonstrated a significant improvement in HDL cholesterol levels in PLB-treated patients (baseline vs postintervention, P = 0.045). In the CLA group, we noticed a decline in the HDL cholesterol concentration that was not statistically significant when compared with PLB.

One of the limitations of this study was the high rate of participants' withdrawal, as well as the difficulties related to the EHCT, both of which contributed to the final small number of participants, as we did not have enough power for seizing small size effects associated with the treatment.

The strength of this study is supported by its own design. For example, inclusion and, particularly, elimination criteria were strictly applied. Baseline characteristics of participants were similar regarding anthropometric and metabolic condition, particularly those related to subrogated indexes of insulin resistance. Additionally, the main outcome was evaluated by the gold standard EHCT, and the benefits of the overall LIP were evident and similar, regardless of treatment allocation. Although the withdrawal of participants was high in our study, the elimination was random and homogeneous in all groups.