Characterization of the population
We analysed a population of 873 adolescents aged 13.8 years with a mean BMI of 21.5 kg/m2. Following the established definition for child overweight and obesity, 71.8% of them were normo-weight whereas 28.2% showed an excessive weight (Table 1). The mean BMI Z-score was 0.00 (-0.067-0.067) for the global population, and -0.018 (-0.10-0.07) and 0.021 (-0.08-0.12), for girls and boys, respectively (p=0.569). Also, according to the latest clinical guides of the American College of Cardiology/American Heart Association/American Diabetes Association for paediatric population [18][19], they were, on average, in the normotensive range and showed unaltered levels of the glycemic and lipid profiles (Table 1). However, further analysis in females and males pointed out significant differences between sex. Girls (52.8%) exhibited lower rates of overweight and obesity, SBP, and glycemia, but higher levels of TC, LDL-C, and HDL-C than boys (Table 1). The ratio of TG and HDL-C, as a predictor of metabolic syndrome and CV diseases [20], was significantly lower in female individuals.
Genotypic and allelic frequencies of ACE2 SNPs
We next analyzed five single nucleotide polymorphisms (SNPs) of the ACE2 gene related with CV injuries [21][9][10]. Firstly, the genotype frequencies of rs4646188, rs879922, rs233575, rs2074192, and rs2158083 were found in a Hardy's-Weinberg equilibrium. The prevalence of the minor alleles of the studied SNPs in girls and boys, respectively, were: (G) in rs4646188, 10.1% and 8.4%; (C) in rs879922, 42.4% and 39.3%; (G) in rs233575, 39.2% and 37.3%; (T) in rs2074192, 40.10% and 41.5%, and (C) in rs2158083, 38.45% and 36.3% (Table 2).
Association between ACE2 SNPs and overweight and obesity
Next, we tested whether these five ACE2 SNPs (rs4646188, rs879922, rs233575, rs2074192, and rs2158083) might associate with the anthropomorphic variables of the population. Remarkably, none of the SNPs were significantly related with the SBP and DBP, either in girls or boys (not shown). However, in females, the occurrence of overweight or obesity was significantly associated with three of the studied SNPs (Table 3). In particular, the minor allele C of the rs879922 variant was present in 26.9% of overweight or obese females [OR 1.67 (95% CI: 1.02-2.75), p=0.042]. Similarly, the G allele of rs233575 and the C allele of rs2158083 were present in 28.0% [OR 1.98 (1.21- 3.22), p=0.006] and 27.0% [OR 1.67 (1.04-2.68), p=0.032] of overweight or obese girls, respectively (Table 3). In contrast, the overweight or obesity condition was not related with any SNPs in the male population (not shown). Thus, the presence of rs879922, rs233575, and rs2158083, previously related with CV damages [21][9][10], were also associated with overweight and obesity in adolescent females.
Association between ACE2 SNPs and the glucose and lipid profiles
Further potential relationships between the SNPs and biochemical variables were also assessed. Glycemic parameters such as glycemia, plasma insulin and the HOMA-IR were not significantly associated with the presence of any SNPs in both sexes (not shown). Lipid parameters did not show any relationship with SNPs in male adolescents. However, again in females, four out of five SNPs were linked with the lipid profile. The highest levels of TG were significantly associated with the presence of heterozygous genotypes of rs879922 (p=0.020), rs233575 (p=0.017) and rs2158083 (p=0.036) (Table 4). The TG/HDL-C ratio also linked with rs879922 and rs233575. In addition, the established cut-off levels of TG (over 90 mg/dl; 75th percentile) for adolescents (10-19 years-old) [18] were related with rs879922 [OR 1.78 (95% CI: 1.06-2.96)], rs2158083 [OR 1.75 (1.08-2.82)], and rs233575 [OR 1.61 (1.00-2.62)] (Figure 2A). Also, the highest levels of TC and LDL-C were related with the heterozygous genotypes of rs2074192 (p=0.003 and p=0.03, respectively) and rs2158083 (p=0.008 and p=0.019, respectively) (Table 4). Similarly, the cut-off levels of TC (over 170 mg/dl; 75th percentile), but not that of LDL-C levels, for adolescents (10-19 years-old) [18] were associated with rs2074192 [OR 1.54 (95% CI: 1.04-2.28) and rs2158083 [OR 1.53 (1.04-2.25)] (Figure 2B). The HDL-C levels were, however, independent of the existence of these ACE2 SNPs. Therefore, the presence of ACE2 SNPs rs879922, rs233575 and rs2158083 may be useful to predict elevated levels of TG and TC in girls.
Association between ACE2-haplotypes and overweight/obesity and TG levels
Since rs879922, rs233575, rs2074192, and rs2158083 were related with overweight/obesity and lipid alterations in females, we examined whether combinations of these SNPs might also associate with higher risk of both pathologies. In particular, the haplotype composed by minor alleles of rs879922, rs233575, and rs2158083 (C-G-C) was overrepresented in girls (34%) and linked to significant higher BMI [coef. 0.01 (0.001, 0.019; 95% CI), p=0.038], in comparison with the haplotype composed by their major alleles (G-A-T) (Table 5). The C-G-C haplotype was also associated with the presence of overweight/obesity [OR 1.41 (1.01, 1.97; 95% CI), p=0.044], and elevated plasma TG [coef. 0.023 (0.002, 0.044; 95% CI), p=0.031] and TG/HDL ratio [(coef. 0.031 (0.001, 0.061; 95% CI), p= 0.045]. Interestingly, after adjusting by BMI, the C-G-C haplotype maintained its association with higher TG levels [coef. 0.023 (0.002, 0.044 95% CI), p=0.034] (Table 5). On the other hand, the haplotype composed by minor alleles of rs22074192 and rs2158083 (C-C) was overrepresented in 37% subjects but did not significantly associate with either BMI or lipid levels (not shown). Thus, C-G-C haplotype of rs879922, rs233575, and rs2158083 SNPs could be considered a risk marker for obesity and dyslipemia in females.