Childhood obesity is a major public health issue. Its incidence in Mexico has increased dramatically over the last few decades. According to preliminary results from the National Survey on Health and Nutrition 2016, the prevalence of overweight and obesity in children aged between 6 and 12 years has exceeded 33 %( Reference Shamah-Levi, Cuevas-Nasu and Dommarco-Rivera 1 ); this places Mexico among the top countries on this topic. Excessive adipose tissue is a risk factor for developing non-communicable chronic diseases; however, a greater risk has been reported when fat is located in upper body areas( Reference Lee, Pedley and Therkelsen 2 ). Therefore, early identification of overweight and obesity in children is key for health assessment processes to achieve overall well-being in this population( Reference Kim, Lee and Laurson 3 ).
BMI, waist circumference (WC) and waist-to-hip ratio are practical measurements to assess adiposity. The former has been adopted by most health-care practitioners as well as in large-scale epidemiological studies, although it does not identify the amount and distribution of adipose tissue. On the other hand, WC estimates adiposity in the abdominal region and has been associated with dyslipidaemia and CHD( Reference Maffeis, Pietrobelli and Grezzani 4 , Reference Moschonis, Karatzi and Polychronopoulou 5 ). However, one of the limitations of this measurement is the lack of a standardized technique. Moreover, ethnic variations and practical obstacles such as clothing removal and other anthropometric variations hinder its international standardization( Reference Da Silva, Zambon and Vasques 6 ). Previously, neck circumference (NC) was proposed as an alternative for adiposity estimation( Reference Hatipoglu, Mazicioglu and Kurtoglu 7 ). The French scientist Jean Vague( Reference Vague 8 ) was the first researcher who used NC as an indicator of upper body fat. He also proposed the concepts of android and gynoid adiposity to describe local fat deposition. It has been shown that NC not only reflects subcutaneous fat depots near the respiratory tract, but also obesity grade in human populations( Reference Lee, Pedley and Therkelsen 2 ). Noteworthy, in 1995 Sjöström et al. ( Reference Sjöström, Håkangård and Lissner 9 ) concluded that upper-body fat estimated by NC measurement could indicate a greater metabolic risk compared with abdominal visceral fat.
Some authors have pointed out that NC might have a supplementary clinical value to other anthropometric measurements in the paediatric population( Reference Yang, Samarasinghe and Kane 10 ). It has been reported in Mexico that acantosis nigricans is associated with NC in the adult population( Reference Hernández-Escalante, Cabrera-Araujo and Euán-Braga 11 ). To the best of our knowledge, there are no studies of Mexican children that describe the usefulness of NC along with other anthropometric indicators of adiposity. Therefore, our aims were to assess the correlation between anthropometric adiposity indicators and NC and to establish cut-off points on NC for both sexes at different ages.
Methods
Selection and description of participants
Data presented in the current paper are part of a broader project entitled ‘Active intervention to improve feeding habits and physical activity in schoolchildren’. The participants belonged to a convenience sample of six schools in Acatlán, Jalisco, Mexico. Only children who were not attending a weight-loss programme or taking any medication, and who were not physically disabled were included. Children with history of chronic disease, neck masses, neck deformity or cervical collar were excluded from the present study.
Technical information
Prior to the data collection, researchers were trained and standardized in accordance with the Habicht method( Reference Habicht 12 ). Height was measured to the nearest 0·1 cm using a portable stadiometer (SECA, Hamburg, Germany) with the participant shoeless and the head held in the Frankfort horizontal plane. Body weight was measured to the nearest 0·05 kg using a calibrated electronic weighing scale (SECA, Hamburg, Germany) without shoes and heavy clothing. WC was measured to the nearest 0·1 cm with the participant standing, at the end of a normal expiration, using an inelastic tape at the point midway between the lowest rib and the top of the iliac crest. NC was measured at the mid-point of the neck at the level of the thyroid cartilage, with the participant’s body held erect, eyes facing forward and normal breathing, using an inelastic tape. To assess body fat composition, we used triceps and subscapular skinfold thickness according to standard techniques( Reference Stewart, Marfell-Jones and Olds 13 ) and body fat percentage was calculated using Slaughter’s equation( Reference Slaughter, Lohman and Boileau 14 ). BMI was calculated by dividing the weight in kilograms by the square of the height in metres.
All measurements were taken twice and the mean was used for the analysis. When the height, weight and circumferences differed by 1 % or more, a third measurement was obtained. This process was also carried out when skinfold thicknesses differed by 5 % or more.
Operational definitions of terms
Underweight was considered as BMI-for-age Z-score (BMIZ) <−2, normal weight as BMIZ between −2 and +1, overweight as BMIZ between +1 and +2, and obesity as BMIZ >+2, in agreement with the WHO( Reference De Onis, Onyango and Borghi 15 ). According to previous research( Reference Maffeis, Pietrobelli and Grezzani 4 , Reference Okosun, Chandra and Choi 16 , Reference Goran and Gower 17 ), elevated values of WC are related to metabolic disorders in the paediatric population. Importantly, Fernández et al. ( Reference Fernández, Redden and Pietrobelli 18 ) have suggested to pay careful attention to children and adolescents with WC values that fall on the 75th up to the 90th percentile according to their ethnic classification and sex. This becomes relevant in the identification of children at risk for various co-morbidities, including CVD, hyperinsulinaemia and type 2 diabetes; consequently, the 75th percentile was taken to propose cut-off points to identify elevated central adiposity based on WC in our examined population.
Statistics
Statistical analyses were performed using the statistical software package IBM SPSS Statistics for Windows version 22.0 and MedCalc version 9.4.2.0 (written by Frank Schoonjans, Mariakerke, Belgium). Numerical variables were reported as mean and sd. Comparisons were conducted between groups using Student’s t test for independent samples and the χ 2 test for qualitative data. Pearson’s correlation coefficient was used to measure the strength of the association between NC and anthropometric adiposity indicators such as WC, BMI, body fat percentage and skinfold thickness. P<0·05 was considered statistically significant.
Receiver-operating characteristic curve analysis( Reference Van der Schouw, Verbeek and Ruijs 19 ) was used to determine the cut-off values. Elevated central adiposity was defined as a percentile equal to or higher than the 75th percentile in WC, whereas overweight/obesity was considered as BMIZ>1 (for the latter, children with BMIZ<−2 were excluded).
Results
The study sample consisted of 1802 participants (50·3 % girls) aged between 6 and 11 years, who met the criteria described above. Comparison of mean values of age, weight, height, body fat percentage and BMI by sex did not show statistical differences. NC and WC values were significantly higher in boys than in girls. In contrast, skinfold thickness showed significantly higher values in girls (Table 1). Overweight and obesity prevalence in the studied population was 20·5 and 23·0 %, respectively. Overweight was present in 18·8 % of boys and 22·2 % of girls, whereas obesity showed a prevalence of 27·2 and 18·7 %, respectively (P<0·001). According to height-for-age Z-score, 1·4 % of participants had moderate malnutrition and 0·1 % showed severe malnutrition. On the other hand, children with elevated central adiposity (those above the 75th percentile for WC) accounted for 25·2 % in both boys and girls.
WC, waist circumference; NC, neck circumference; TSF, triceps skinfold thickness; SSF, subscapular skinfold thickness; BF%, body fat percentage; BMIZ, BMI-for-age Z-score; HAZ, height-for-age Z-score.
Pearson’s correlation coefficients between NC and anthropometric adiposity indicators by sex are presented in Table 2. NC showed a strong positive correlation with WC, BMI, body fat percentage and skinfold thickness in both sexes, which were notably higher in boys regardless of age. A noteworthy observation was that WC showed the highest significant correlation when analysed by both age and sex (Table 3).
NC, neck circumference; BF%, body fat percentage; WC, waist circumference; TSF, triceps skinfold thickness; SSF, subscapular skinfold thickness.
NC, neck circumference; BF%, body fat percentage; WC, waist circumference; TSF, triceps skinfold thickness; SSF, subscapular skinfold thickness.
As shown in Table 4, NC accurately defined elevated WC by sex and age. The area under the curve indicates that for both age and sex, the accuracy levels of NC for identifying elevated WC were over 0·90. The NC cut-off points to identify central obesity ranged from 25·7 to 30·1 cm for girls and from 27·5 to 31·7 cm for boys between the age of 6 and 11 years. The sensitivity and specificity of this screening method were 83·8–97·0 and 78·4–92·4 %, respectively, for girls and 81·1–100·0 and 83·2–92·5 %, respectively, for boys.
AUC, area under the curve; LR, likelihood ratio.
Discussion
Because of the rapid growth of obesity incidence (tenfold higher over the last 40 years)( Reference Di Cesare, Bentham and Stevens 20 ), the development of specific population-based assessment tools for estimating adiposity is becoming an important issue for diagnostic accuracy. Our outcome showed an overall combined overweight/obesity prevalence of 43·5 %; this value is higher when compared with the prevalence in Mexico( Reference Shamah-Levi, Cuevas-Nasu and Dommarco-Rivera 1 ) which is 33·2 %. Importantly, comparing our results with the population statistics in the region, more specifically those of Jalisco State( 21 ), also appears to be relevant. An example of this is that the occurrence of overweight/obesity in Jalisco is closer to the data reported herein (39·6 %). Analysis of overweight/obesity by sex showed that in boys, proportions are similar: 46·0 % in our population and 44·7 % in the state. However, girls showed a higher prevalence: 40·9 v. 34·9 %, respectively. In summary, a large proportion of boys showed obesity, whereas girls were more often overweight. This tendency is similar to the national and state statistics.
In our study, we described for the first time in Mexican children the usefulness of NC for assessing elevated central adiposity, evidenced by its high correlation with WC. Many studies conducted in South America, the USA, Europe and a few Middle Eastern countries have evaluated the validity of NC( Reference Hatipoglu, Mazicioglu and Kurtoglu 7 , Reference Nafiu, Burke and Lee 22 – Reference Kelishadi, Djalalinia and Motlagh 28 ). These reports have also proposed it as a rapid, easily performed method to identify overweight and obesity. Importantly, one of the most remarkable features of NC is the almost no need for additional measurements as LaBerge et al. have reported( Reference LaBerge, Vaccani and Gow 29 ).
Remarkably, the correlation coefficients presented herein are consistent with the published reports conducted by Hatipoglu et al. ( Reference Hatipoglu, Mazicioglu and Kurtoglu 7 ) and Nafiu et al.( Reference Nafiu, Burke and Lee 22 ). However, our outcome showed higher correlations between NC and both WC (r=0·91) and BMI (r=0·86) than those of these studies. This might be explained by the selection criteria and the nature of the population itself. On the one hand, our evaluated population was obtained from six schools in the south-western state of Jalisco. Unlike the methodology applied by those studies, we did not use case–control design, nor did we recruit children from clinical facilities. Indeed, our data are widely supported by the size and heterogeneity of the population studied.
We found high correlations in anthropometric adiposity indicators by sex and age. In girls, the correlations ranged from 0·72 to 0·90, while in boys from 0·79 to 0·93. In this regard, WC and BMI showed the highest correlations. Although other authors have reported similar tendencies, correlations in their studies have been lower. For instance, Coutinho et al. ( Reference Coutinho, Longui and Monte 24 ) reported correlations ranging from 0·64 to 0·86 in girls and from 0·44 to 0·86 in boys. Of note, correlations in boys were markedly lower compared with girls. On the other hand, Lou et al.( Reference Lou, Yin and Wang 30 ) reported lower correlations for WC in girls ranging from 0·56 to 0·74. This finding might be due to the ethnic differences in the studied populations and the different overweight and obesity prevalence rates (40·9 % in our study v. 27·4 % in Lou et al.’s study). This suggests that overweight/obesity prevalence in a given population might play an important role in the analysis of correlations.
A plausible explanation for the high correlation we observed between NC and other anthropometric adiposity indicators might be the fact that this body region exerts a high lipolytic activity. According to several studies, serum NEFA are mostly released from the upper body region, and even in a greater manner than from visceral fat( Reference Guo, Hensrud and Johnson 31 , Reference Torriani, Gill and Daley 32 ). Other reports have shown that NC is a useful tool for identifying upper-body adiposity( Reference Stabe, Vasques and Lima 33 – Reference Lee, Pedley and Therkelsen 35 ).
Previous research indicates that up to a certain point, neck adipose tissue should be regarded as an ectopic fat depot. This concept has its origin in the following hypothesis: when an excess of TAG cannot further expand into metabolically favourable depots, they infiltrate into tissues that are able to store only a small amount of fat. In fact, ectopic fat is considered to have properties of dysfunctional adipose tissue and it is closely related to metabolic alterations( Reference Mittendorfer 36 ).
Several authors have proposed NC as a metabolic syndrome indicator for pre-school and school children, as well as adolescents( Reference Da Silva, Zambon and Vasques 6 , Reference Androutsos, Grammatikaki and Moschonis 23 , Reference Kelishadi, Heidari-Beni and Qorbani 37 – Reference Kurtoglu, Hatipoglu and Mazicioglu 43 ). NC has been related to non-alcoholic fatty liver and cut-off points for its identification have been proposed( Reference Hatipoğlu, Doğan and Mazıcıoğlu 44 ). Figueroa-Sosa et al.( Reference Figueroa-Sosa, García-Rojas and Oropeza-Priego 45 ) recently highlighted the importance of having an easy-to-apply tool such as NC, mainly for use in marginalized communities where often there is no access to anthropometric equipment such as a scale or stadiometer.
One of the aims of the present study was to propose NC cut-off points based on WC to identify elevated central adiposity. In this regard, the International Diabetes Federation states that WC≥90th percentile is an indicator for metabolic risk in children( Reference Alberti, Zimmet and Kaufman 46 ). However, as stated in the ‘Methods’ section, we decided to use the 75th percentile as a cut-off point to identify children with elevated central adiposity. The reason for this is that Fernández et al.( Reference Fernández, Redden and Pietrobelli 18 ) found on the 75th and 90th percentiles that Mexican-American girls evidenced the fastest overall WC increase of all girls. Besides, at any of the percentiles considered, Mexican-American people showed the highest overall WC and the fastest overall rate of WC increase with age. Therefore, due the high prevalence of overweight and obesity in Mexico, we considered that early identification of population at risk is highly important.
On the other hand, the previously mentioned reports have addressed the validity of NC by creating cut-off points based on the Centers for Disease Control and Prevention growth charts for BMI. Nevertheless, Hassan et al. ( Reference Hassan, Atef and El-Masry 47 ) pointed out the importance of cautiousness when using NC as a screening tool. In the first place, a gold standard would be optimal for this determination, and in the second place, BMI by itself is not a good proxy for regional adiposity. In this regard, we propose, for the first time, using NC cut-off values based on WC. We found that NC cut-off points for elevated central adiposity could be proposed as follows: for girls, 25·7–30·1 cm in the age range of 6–11 years old, while for boys, 27·5–31·7 cm on the age range 6–11 years old. In fact, values of area under the curve based on WC were slightly higher than those based on BMI, as reported in other studies where similar cut-off points were proposed( Reference Taheri, Kajbaf and Taheri 26 , Reference Lou, Yin and Wang 30 ). Importantly, due to the high prevalence of overweight and obesity in the studied population, the cut-off points could only be useful in geographical regions with similar prevalence. As mentioned before, Jalisco State and our study showed similar statistics.
This evidence supports the accuracy of NC for the prediction of elevated central adiposity. Although WC is not a gold standard, it has been proved to be an accurate indicator of central adiposity, which in turn is highly associated with metabolic risk regardless of age( Reference Maffeis, Pietrobelli and Grezzani 4 , Reference Moschonis, Karatzi and Polychronopoulou 5 ). The authors of the present paper want to emphasize that despite the difficulties in measuring WC, when applied optimally, it provides valuable information on central adiposity. On the other hand, NC presents certain technical advantages compared with WC. First, it does not require clothing removal, which makes it more convenient in cold weather and feasible for some cultures. Second, it is not altered by respiratory movements or postprandial abdominal distension, diminishing the need for additional confirmatory measurements.
One report found that it is feasible to effectively obtain accurate self-measurements using a flexible, inelastic, paper tape measure self-assembled from a PDF file downloaded from the Internet for large-scale consumer surveys where participants could not easily visit anthropometric labs for measurement by trained technicians. The authors of that report demonstrated a simple, inexpensive method for teaching novice mothers of young children to obtain their own body circumferences in which NC was measured, resulting in accurate and reliable data. They concluded that collecting self-measured and self-reported circumference data in future studies might be a useful approach for body composition research( Reference Barrios, Martin-Biggers and Quick 48 ). These results might suggest that NC measurement could be performed by parents or teachers given that there is no need for previous specialized training for this anthropometric variable, which in turn facilitates the identification of elevated central adiposity in children both at home and in school.
We acknowledge the worldwide importance and acceptance of both BMI and WC as practical tools. However, our report attempts to highlight the supplementary validity of NC as a time-saving tool in large epidemiological studies as well as in daily clinical practice for the identification of increased adiposity. We also acknowledge that to achieve practical acceptance, additional studies should be conducted in other strategic regions given the admixture background of its population; consequently, our proposed cut-off points provide high accuracy but are limited to the population in the studied region. Finally, further studies on NC including metabolic risk parameters could strengthen the evidence of its usefulness.
Conclusion
NC could be used as a simple, inexpensive and non-invasive indicator for central obesity assessment in Mexican schoolchildren.
Acknowledgements
Acknowledgements: The authors acknowledge the financial funding from the Tresmontes Lucchetti Company and the support of the National Institute of Public Health. The authors thank Roberto Rodriguez-Echevarria, PhD, for manuscript review. Financial support: The broader project where this work belongs was funded by the Tresmontes Lucchetti Company, Universidad de Guadalajara and the Instituto Tecnológico de Estudios Superiores de Occidente. These funding agencies had no role in the design, analysis or writing of this article. Conflict of interest: None. Authorship: E.V.-S. and C.C.-P. conceptualized and designed the study and drafted the initial manuscript. A.L.-H., E.R.-V., E.M.V.-G. and E.V.-S. carried out the initial analyses and reviewed and revised the manuscript. C.O.R.-G., E.V.-S. and C.C.-P. designed the data collection instruments, coordinated and supervised the data collection, and critically reviewed the manuscript. All authors approved the final manuscript as submitted and agreed to be accountable for all aspects of the work. Ethics of human subject participation: This study was conducted according to the guidelines of the Declaration of Helsinki. All procedures involving human subjects were approved by the Comité de Ética en Investigación del Centro Universitario de Tonalá (003-2016). Verbal informed consent was obtained and formally recorded from all schoolchildren and their tutors. Written informed consent was obtained from all school directors.