To the Editor,
We have read the article recently published by Escribano et al. in the Revista Española de Cardiología with great interest.1 Epidemiological studies are a starting point for performing interventions and analytical studies, and have recently recovered the scientific prestige that they had in previous decades. We would like to specify some aspects of the definition of obesity.
As many other authors have pointed out, in the clinical field obesity is indirectly defined using the body mass index (BMI) and the waist circumference (WC), with established cut-off points for both. These values are well recognized for the adult population, but underestimate the actual prevalence in children and adolescents.2 The same occurs with other cardiovascular risk factors (CVRF): arterial hypertension and hypercholesterolemia.
Pediatric societies have shown consensus on defining the prevalence of CVRF, using percentile charts validated by cross-sectional and longitudinal studies. However, there are some controversies in defining child and adolescent obesity using BMI, ie, whether age- and sex-dependent national charts with a percentile cut-off point of 97th (p97) should be used,3 or international criteria should be taken as reference.4 There is more agreement regarding abdominal obesity, fixing the cut-off point at the 90th percentile (p90).5 For that reason, Escribano et al. should have shown the actual prevalence of general and abdominal obesity in accordance with criteria specific to the 15- to 17-year-old group, showing separate data for adults.
Our group has published CVRF prevalence data for children and adolescents including a sample of 1534 individuals between 9 and 17 years of age from southern Spain.6, 7 The prevalence of obesity in the 15 to 17 age group was 9.4% according to national criteria (95% confidence interval [CI]: 7.9%-10.8%),7 6.5% being male and 11.3% being female. If we were to use BMI>30 to define obesity, we would obtain 2.6% in males and 5.5% in females. These data therefore contrast with those published for the adult population from the first age group in the Escribano et al. study. They are however in accord with the enKid studies, which serve as a national reference in Spain, and show that the prevalence of child and adolescent obesity in the center of Spain (Castile and Leon) are very similar, although slightly less than data for Andalusia, in the south.5 Prevalence of abdominal obesity for 15- to 17-year-olds was 88% in our study (95% CI: 83.8%-93.7%),7 which is very different from that found in Escribano et al.’s article.
The authors reported a relationship between obesity (mainly abdominal) and the presence of other CVRF in the adult population, which is a phenomenon that we also found in our study.7
To make it easier to define obesity in early life stages, which are subject to growth, we have validated the waist-to-height ratio, estimating 0.5 as the cut-off point for establishing the prevalence of abdominal obesity.7 As such, this method has an advantage over using absolute WC and BMI values, as complicated percentile charts for age and sex can be avoided. This new anthropometric index should be validated externally in the child and adolescent population.
We would like to conclude our letter by congratulating Escribano et al. for their study, which serves as a national reference point, and the editorial team for accepting epidemiological studies of this type in their prestigious clinical journal.
.
Corresponding author: almendrode@secardiologia.es