A group of 253 Mexican adolescents from the Coahuila state were classified into 3 groups, according to the nutritional status using data issued by the World Health Organization (WHO; 2007) on the basis of age-specific body mass index (BMI/age): Obese (≥95th), overweight (85th-94.9th) and normal weight (5th-84.9th percentile). Written consent to participate was requested from the parents and adolescents. The study protocols were performed in accordance with the ethical standards and were approved by the University Hospital-Autonomous University of Columnbia Review Board and registered under the code 01–2011.

The study protocol included a background interview, anthropometric measurements, and venous blood sampling for the biochemical and genetic analyses.

To determine the biochemical parameters of glucose, total cholesterol, low-density lipoprotein (LDL), high-density lipoprotein, triglycerides and insulin, 5 ml of peripheral blood samples were extracted from the participants who fasted for 12 h. These biochemical values were used to calculate the homeostatic model assessment for insulin resistance (HOMA-IR) and the atherogenic index. Biochemical parameters were measured using an SLFIA immunoassay and a TOSOH AIA-600 analyzer (Tosoh Corp., Tokyo, Japan) for insulin, and a Clinical Chemistry Analyzer Diconex InCAA model (Diconex, Buenos Aire, Argentina) for the remaining analyzed parameters.

Genomic DNA was extracted from peripheral venous blood cells using the phenol-chloroform method, precipitated in ethanol and finally resuspended in Tris-EDTA (pH 7.8) at a concentration of 0.1–1.0 µg/µl.

Genotyping was performed by the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method using an MJ Mini PTC1148 thermal cycler (Bio-Rad, Hercules, CA, USA).

For the TNFα −308G/A polymorphism, a modified version of the Chen et al (19) method was used. PCR was performed in 25 µl using 0.5 µM primers (Integrated DNA Technologies, Coralville, Iowa, USA) (forward, 5′-GGGACACACAAGCATCAAGG-3′ and reverse, 5′-AATAGGTTTTGAGGGCCATG-3′), 0.2 mM dNTPs, 1.5 mM MgCl₂, 2.5 units Taq DNA polymerase (Promega, Madison, WI, USA) and a total amount of 250 ng of genomic DNA. The PCR program consisted of 35 cycles at 94°C for 30 sec, 61°C for 30 sec and 72°C for 30 sec. The 142 base-pair (bp) PCR amplicons (~1 µg) were digested overnight with NcoI (New England Biolabs, Ipswich, MA, USA) at 37°C, and the generated fragments were analyzed by electrophoresis in 3% agarose gels using ethidium bromide staining and were documented in a UVP model 2UV High Performance Transilluminator (UVP, Inc., Upland, CA, USA). The products obtained during this process were 126 bp and 16 bp fragments for the wild-type G allele, and 142 bp (non-digested PCR product) for the variant A allele.

The PTPN22 +1858C/T polymorphism analysis was performed using a modified method reported by Harrison et al (20), using 0.5 µM of each primer (Integrated DNA Technologies) (forward, 5′-ATGTTGCTTCAACGGAATTT-3′ and reverse, 5′-CATGCTGCTATTGCTCTGCT-3′), 0.2 mM dNTPs, 1.5 mM MgCl₂, 2.5 units of Taq DNA polymerase (Promega, Madison, WI, USA) and a total amount of 250 ng of genomic DNA. The PCR program consists of 33 cycles at 94°C for 1 min, 58°C for 1 min and 72°C for 1 min. The 400-bp PCR amplicons (~1 µg) were digested overnight with XcmI (New England Biolabs) at 37°C, and the generated fragments were analyzed by electrophoresis in 2% agarose gels stained with ethidium bromide and analyzed in a gel documentation system (UVP model 2UV High Performance Transilluminator). The products obtained during this process were 238 bp and 162 bp fragments for the variant T allele, and 400 bp (non-digested PCR product) for the wild-type C allele.

For the sample size calculation, it was assumed that the coefficient of variation was 22% for the BMI. Considering a significance level of 5%, a minimum power of 80%, an Ω of 0.25, and a confidence interval of 90%, a sample size of 15/group is sufficient. To detect any significant differences in the values of the biochemical parameters among the analyzed groups of the nutritional status and genotype, comparisons were made between 2 groups (normal versus overweight, normal versus obese, and overweight versus obese), and Student's t-test or the Mann-Whitney U test were used for parametric or non-parametric distributions, respectively. Differences between >2 groups were assessed by one-way analysis of variance and the Kruskal-Wallis H test for parametric or non-parametric distributions, respectively. Post hoc tests (least significant difference and Tamhane's T2) were used for pairwise comparisons. To confirm the contribution of the genetic factors to the variability of the biochemical parameters, a linear regression analysis was undertaken. Possible associations between genotypes and biological parameters were assessed using contingency tables, χ² statistics and Fisher's exact tests. Odds ratios were estimated with 95% confidence intervals. The aforementioned analyses were performed with SPSS for Windows, V.20 (IBM Corp., Armonk, NY, USA). All P-values were two-tailed. The corrected P (Pc) values were adjusted using Bonferroni's correction to exclude spurious associations. P<0.05 was considered to indicate a statistically significant difference. Hardy-Weinberg equilibrium (HWE) was obtained using a goodness-of-fit test.

The population studied consisted of 253 adolescent subjects (age range, 11–16 years; mean, 13.49 years), from Saltillo (Coahuila, Mexico), categorized according to the BMI percentile tables of WHO: Obese (161 subjects), overweight (22 subjects) and normal weight (70 subjects) (Table I).

The distribution of the average levels of glucose, insulin, lipid profile, BMI, HOMA-IR and atherogenic index in the analyzed groups are described in Table II. Statistically significant differences were identified in BMI, insulin levels, HOMA-IR and triglyceride levels among all the three groups analyzed (P≤0.05). There was a significant difference between the normal weight and obese group in all the biochemical parameters analyzed (P≤0.05) (Table II).

The molecular analysis of the TNFα −308 G/A and PTPN22 +1858C/T gene polymorphisms by PCR-RFLP showed that the predominant genotype for the polymorphism of the TNFα gene was G/G (94.9%; 65 normal weight subjects, 20 overweight and 155 obese), the heterozygous genotype G/A was present in smaller amounts (5.1%; 5 normal weight subjects, 2 overweight and 6 obese), while the homozygous genotype A/A was absent in the studied population. For the PTPN22 gene, the C/C (94.5%; 63 with normal weight subjects, 21 overweight and 155 obese) polymorphism was the predominant genotype, and the heterozygous genotype C/T was observed in minor amounts (5.5%; 7 normal weight individuals, 1 overweight and 6 with obesity), while the homozygous T/T genotype was absent in the study population. No statistically significant association was identified between polymorphisms analyzed and the presence of overweight and obese subjects in the population (P>0.05). The two genotypes were in HWE in the population studied (P>0.05).

To determine the possible effect of the genotypes identified for TNFα and PTPN22 with the biochemical values of the studied population, mean comparisons between genotype groups were performed. According to the P-values, carriers of the CT genotype of the PTPN22 gene exhibited a decrease in glucose levels, insulin levels, and HOMA-IR index (P≤0.05), and a trend towards reduced levels of LDL and cholesterol (Table III). Linear regression analysis using the aforementioned genetic models confirmed that these polymorphisms affected the variability of these biochemical parameters.