Overweight and obesity in children constitute a public health problem worldwide and have shown a steady increase in recent decades.1,2 Many children are affected by these conditions very early in life, with risks of suffering from them throughout their lives.3 A high body mass index (BMI) in childhood is related to obesity in adulthood, but this does not differentiate between lean mass and fat mass, does not account for the distribution of body fat, and is a poor predictor of cardiometabolic risk.4

Abdominal or central obesity, which is determined by measuring waist circumference (WC), has been related to cardiometabolic risk, in adults, when WC is greater than 88cm in women or 102cm in men.5 In children, the WC cut-off point is not an absolute number, since it changes with age. Because of this, the at-risk WC values used for children are those in the 90th percentile of the charts published for different populations, according to gender and age.6 These high WC values have been correlated with intermediate markers of cardiometabolic risk, such as high blood pressure7, dyslipidemia, insulin resistance,8,9 metabolic syndrome,10 left ventricular dysfunction,11 increases in carotid artery thickness,12 and increases in endothelial dysfunction markers (high-sensitivity CRP, PAI-1, TNF-alpha, among others).13 Likewise, increases in WC are associated with higher mortality rates in adults, regardless of their BMI.14 Recently, a higher WC has also been associated with obstructive pulmonary function in children.15WC is a better indicator of visceral fat than BMI when the latter is measured with nuclear magnetic resonance or computed axial tomography, taken at the level of the fourth lumbar vertebra.16 WC represents the intra-abdominal fat that is thought to be more metabolically active,9 and it helps predict cardiometabolic risk even in patients with a normal BMI (“metabolically obese individuals”). WC values are influenced by diet17 and exercise,18 so measuring it is recommended to monitor obese children's responses to these behavioral inputs.

WC values that determine abdominal obesity have been established in different countries, taking into account different races, which revealed significant differences among those considered.19 As far as we know, in Chile there are no recently tabulated data of WC in children. These data would be very useful in establishing the cut-off values that represent higher cardiometabolic risk. The objective of our study is to describe the percentile distribution of WC in a representative sample of children and adolescents from Santiago, Chile.

A total of 3022 children were measured and divided into percentiles according to their WC, gender, and age. Table 1 shows sampling sizes of the studied groups stratified by age and gender, and the mean and standard deviations for weight, height, BMI, height to age ratio, and weight to age ratio. The prevalence of obesity and overweight were 20.6% and 24%, respectively.

The percentile distributions of the WC values obtained from boys and girls according to age are shown in Table 2, and in Figs. 1 and 2, respectively.

Figure 1

(0.2MB).

Figure 2

(0.2MB).

Overall, the same trend was observed for both male and female participants, i.e., the WC increased with age in both genders, as was expected. When comparing equivalent percentiles of boys and girls of the same age, it was discovered that their WC values were also similar. So, these differences were not enough to be statistically significant (p>0.05).

On the other hand, within the age groups studied, male participants increased their WC by 2.2cm per year, on average. However, females increased theirs by 2.6cm per year, which was significant (p<0.05).

The WC value that was established as a cardiometabolic risk predictor for adults, i.e., 88cm in female adults and 102cm in male adults, was reached in the 90th percentile of girls from 12 to 14 years of age. However, the 90th percentile of boys never reached the value of 102cm in any of the ages studied.

When comparing the WC of the Chilean children with the WC of children from other countries, we found that Chilean children had a larger WC than children from the USA (African-American, European-American, and Mexican-American),19 Australia,21 Great Britain,22 and Colombia.23

The reference sample of North American children was taken from the NHANES III survey, which is one of the most important surveys published to date. This survey is usually taken as reference because of the huge racial diversity in that country. Moreover, it establishes separate WC values for the different subgroups of European-American, African-American, and Mexican-American children.

The group with the most similar WC values to those of the Chilean boys and girls was the Mexican-American subgroup. However, Chilean children still had greater WC values than those of all three aforementioned American subgroups. Figs. 3 and 4 show the 50th WC percentile of Chilean boys and girls respectively, compared to Australian, British and Colombian children, as well as the three aforementioned American subpopulations. The differences between Chilean children and these other populations in all of the age groups reached statistical significance (p<0.05).

Figure 3

(0.2MB).

Figure 4

(0.19MB).

In the present study we describe WC distribution curves in percentiles of children and teenagers of both genders. The results were obtained from a representative sample of schoolchildren of the Chilean lower-middle and working classes who attended 10 different schools in Santiago, Chile. Some of the strengths of this study are its sampling design, the number of children examined, and how representative of Chilean children the sample was, because Metropolitan region has the largest number of people, its school age obesity prevalence is the same as in the rest of the country, and the evaluated socioeconomic class represents a large percentage of the country.20 Furthermore, the method used to obtain the measurements was highly effective, as it allowed the investigators to achieve reliable results, even when they were obtained by different observers.

Currently, there is no agreement regarding the anatomical location where WC should be measured. However, some studies conducted on adults show a high correlation between the WC values obtained from the highest point of the iliac crest, at the navel level, and those taken halfway between the lower costal margin and the iliac crest.24 In fact, the same study shows that both areas are similarly accurate in diagnosing cardiometabolic risk factors.

One of the present study's possible weaknesses could be the fact that the sample showed a high tendency towards obesity. However, this was an unavoidable factor in the study, given the nature of lower-middle and working-class life in Santiago. In the present sample, the tendencies towards obesity and overweight were 20.6% and a 24%, respectively. Despite the potentially confounding factor of malnutrition in our sample, these figures are very similar to data observed previously in the Mexican25 and Latin-American populations.26 Moreover, they agree with the data recorded by the Junta Nacional de Auxilio Escolar y Becas (Junaeb) (National Board of School Assistance and Scholarships),27 which represents most 1st-grade children from all of the municipal and subsidized schools in Chile.

The apparent weakness of our sample could be representative of the Chilean reality in terms of nutrition. On the other hand, it is not acceptable to exclude obese subjects of a specific group in order to obtain an “ideal” WC to compare with. Because of this, we intended to simply describe the sample in our study, which is very useful as a reference but should also be used cautiously, for the aforementioned reasons. These issues highlight the need to perform new national studies that will help define which WC percentile is most useful in diagnosing cardiometabolic risk factors. Thus, we should not only define “normality” based on the Gaussian distribution model, but we also should ideally define it in contrast to other studies that intend to identify the cut-off points that best distinguish cardiometabolic risk factors, even when these have not been universally defined and accepted yet.

The results presented here cannot be extrapolated to Chilean children from the upper-middle and upper socioeconomic classes or to the indigenous and rural populations, but it can certainly be a generalization for the rest of the Chilean population, which represents the majority of the country. We can say that our findings reflect the portions of Chilean society that are currently undergoing a post-nutritional transition process, where obesity has become the main nutritional problem for both children and adults. This is similar to what is happening in many developing countries that are also experiencing a nutritional transition process.

WC linearly increases as the subjects of both genders get older. This suggests that the waist circumference of human beings evolves in the same way as other anthropometric parameters, such as weight and height. When waist circumference reaches a value that indicates abdominal obesity, it might only be remedied with an intervention strong enough to modify the factors that have established such progression in the child. If we regard obesity in children as the result of the interaction between genetic and environmental factors, we could conclude that it is mainly caused by new behavioral factors in our country, since the genetic factors have been relatively constant in our population. Such behavioral factors would include an inappropriate diet, sedentary lifestyle, and the improvement of socioeconomic conditions.28

It was surprising to find that some girls between the ages of 12 and 14 in the Chilean population prematurely reached the WC, with its cut-off point that is regarded to be a predictor of cardiometabolic risk in female adults (88cm). This situation may even increase with age, a trend that was observed in our sample. The values for boys were very similar to those observed for girls. However, the male group never reached the cut-off point for cardiometabolic risk established for adults at any age, even beyond the 90th percentile. However, if we take into consideration that they enter puberty later than girls do, it seems logical to think that these children have not yet completed their pubertal development and therefore it is expected that their WC may increase still further in accordance with the progression of puberty.

When we compared the results of the sample of Chilean children to the data from other countries, we found that there were significant differences between Chilean children and American, Australian, British and Colombian children. These differences could be ascribed to race and the different degrees to which obesity has prevailed in these countries throughout the years. That is to say, the prevalence of obesity in each country as a whole would partly explain the differences observed in children. In this sense, differences between races do not make it possible to extrapolate WC values from one population to another. One other reason the Chilean data were significantly different from that of other populations is that the other data are 10 years old; obesity has increased a lot in the last decade, especially in children. However recent data from racially similar populations to Chile, like Aguilar de Plata et al.23 in Colombian children show that the waist (p. 50) of Chilean children is still significantly higher than the Colombian, which may be explained by the life styles, environmental factors and nutritional habits of children and Chilean population. This shows clearly the important public health problem that has turned childhood obesity in Chile, making it essential to design interventions that effectively prevent obesity.

In conclusion, the new WC percentile curves in children found herein were obtained using local data that reflect the reality of the lives of our child and adolescent populations and were of national interest or relevance. Because of this, our study is a potentially significant contribution to anthropometric measurements used in assessing cardiometabolic risk in children and teenagers, as they show visceral fat deposits, which are recognized as having the highest metabolic activity and relevance to obesity.

The authors have nothing to disclose.

The authors have no conflict of interest to declare.