Discussion
This study is the first performed in Brazil in a large population that included children of early ages from the city of Cuiabá, which is located in a central area of South America. This city presents very interesting peculiarities for study due to the ethnic mixture and an intense migratory movement of people from different regions of Brazil with different eating habits and genetic backgrounds.
Cardiovascular diseases usually manifest after the fourth decade of life, but atherosclerosis begins during an individual’s first few years [1,5,6]. This finding emphasizes the need to determine the real concentrations of lipids in children and adolescents and identify individuals with cardiovascular risk indicators to enable early intervention.
The values used to diagnose dyslipidemia vary among populations. Each country should ideally establish its own RIs [10,22]. The reference values for lipids are probably not country-specifics. However, it is believed that race, lifestyle, diet, environment and economic development among different areas could account for the differences in lipid levels in children of different countries [23].
Some studies on lipid RIs have been conducted in Brazil, but the goals were to determine the distribution of serum lipids in pediatric populations [24-27], except one study in a small town in São Paulo [28,29]. This report is the first Brazilian investigation to establish the RIs for serum lipids in children and adolescents aged 1–12 years. The currently recommended lipid RIs for Brazilian children and adolescents are based on references from American children and adolescents, who have different genetic profiles and distinct dietary habits. Therefore, this study aimed to narrow this gap. The current cohort is a specific Brazilian population, presumably with non-ideal demographic characteristics, but the results of this study are very interesting. The above-mentioned population has more characteristics in common with Brazilian children and adolescents from other regions of the country than with any North American cohort.
The NHLBI [5] combined children and adolescents 2–19 years of age in a single group to assess their TC, HDL-c, nHDL-c and LDL levels. Only two age groups, 0–10 years old and 10–19 years old, were used to assess the TG levels. In the CALIPER study [30] and the present study, however, the parameters were studied in different age groups for each analyte.
According to the NHLBI, a TC level <170 mg/dL is desirable in children and adolescents 2–19 years of age [5]. The desirable borderline and upper limit threshold values obtained in this study were lower in the age group of 1–2 years. The desirable borderline and upper concentrations for the age group of 3–8 years overlap the recommendations of the NHLBI. Higher thresholds were obtained in the group of 9–12 years than the thresholds proposed by the American guidelines (Table 3).
The CALIPER study [30] combined children and adolescents into age groups of 0–14 days, 15 days to <1 year and 1–19 years. Our results were compared only with this last age group. We obtained lower upper limit values for children up to 8 years of age and values above the CALIPER guidelines for children and adolescents aged >9 years. Genetic differences and diverse eating habits among the Brazilian, US and Canadian populations are possible factors for the observed differences. Compared with the previous Brazilian study of lipids in a single age group (2-9 years old), our values were identical (i.e., the desirable limit for this analyte was 170 mg/dL) for 3-8 year olds [28].
An HDL-c level >45 mg/dL is desirable, 40–45 mg/dL is borderline, and <40 mg/dL is low according to the NHLBI. This last value represents approximately the 10th percentile, and it is the most relevant value in medical practice [5]. The values corresponding to desirable, borderline and lower limits in this study are below the limits suggested by the NHLBI for the age groups of 1 year, 2 years and 3 years (Table 4). There are two potential reasons for this difference: the concentration of this analyte increases gradually over time, and this parameter was studied in a single age group (2–19 years) in the NHLBI study. The concentrations were higher in the age group 4–12 years, and although they were borderline and within desirable limits, they were below the threshold compared with the values proposed by the NHLBI (Table 4). The CALIPER study defined the following four age groups for HDL-c: 15 days to <1 year, 1 to <4 years, 4 to <13 years and 13 to <19 years. The lower and upper limits obtained in the present study were lower than the CALIPER concentrations in the corresponding age groups [30]. Possible explanations include the genetic differences and diverse eating habits of the two analyzed populations. The previous Brazilian study of lipids in children does not reference the 10th percentile, but the desirable limit includes greater values than the results of our study [28].
The n-HDL-c value is obtained by subtracting the HDL-c value from the TC value. Acquiring this value does not require fasting, which renders it very attractive, especially among children. Moreover, this parameter is predictive of atherosclerosis because it includes all classes of lipoproteins [5,31]. The NHLBI considers concentrations < 120 mg/dL desirable [5]. Our values are higher than those suggested by NHLBI in the three categories, namely, desirable, borderline and upper limit (Table 3). The lower HDL-c concentrations in this population (Tables 3 and 4) may explain this difference. The value for this parameter in our study was greater than that in the previous Brazilian lipid study [28]. The CALIPER study did not report this parameter in their analysis [30].
A comparison between the concentrations of LDL-c obtained in this study and the NHLBI revealed that the LDL-c 75th percentile in our study was lower than the NHLBI desirable limit. The LDL-c concentrations of the borderline upper limit for the 1-8-year-old age group overlapped with the NHLBI recommendations. The concentration obtained for this parameter in the age group 9–12 years was higher than the recommendations of the NHLBI (Table 3). One possible explanation for this discrepancy may be methodological differences because the NHLBI data were clustered into a single age group. Differences in the TC, HDL-c and TG values may contribute to these disparities. The value for this parameter in our study was higher than that in the previous Brazilian study of lipids [28]. The CALIPER study did not report this parameter in their analysis [30].
The TG thresholds in this study were systematically higher than those proposed by the NHLBI (Table 3) [5]. The fasting period may provide one explanation for this effect. The recommended fasting for the collection of this analyte is 12–14 hours, but the recommended fasting intervals (duration between feedings) in this study were 3 hours for children up to 2 years, 6 hours for children up to 5 years and 12-14 hours for children > 5 years.
The TG levels obtained in this study [30] show that the upper and lower limits were systematically lower compared with CALIPER. The fasting time may be one explanation for this effect because fasting was not required in the CALIPER study. The value for this parameter in our study was greater than that in the previous Brazilian study of lipids [28].
The weakness of this study was in not includind children from different places around the country. However, the studied children are descended from parents from various regions of Brazil and have many characteristics in common with other Brazilian children and adolescents. In addition, not including adolescents older than 13 years old and the long duration of fasting in younger children are other concerns. New studies need to fully fill this gap and include older age groups in this analysis. Further studies of the serum lipid concentrations and clinical and laboratory parameters already evaluated in this work in our population will be performed in future analyses.