Body Mass Index and Waist Circumference to Define Thinness ...

Body Mass Index and Waist Circumference to Define Thinness, Overweight and Obesity in Portuguese Adolescents: Comparison Between CDC, IOTF, WHO References Beatriz Minghelli1, Carla Nunes2, Raul Oliveira3 1

Beatriz Minghelli – School of Health Jean Piaget Algarve – Piaget Institute – Portugal; National School of Public Health 2

- NOVA University of Lisbon – Portugal; Carla Nunes – National School of Public Health - NOVA University of Lisbon – 3

Portugal; Raul Oliveira - Faculty of Human Kinetics - University of Lisbon – Portugal Corresponding author: Minghelli B, Escola Superior de Saúde Jean Piaget / Algarve, Enxerim 8300-025 Silves, Portugal; Email: [email protected]

Abstract

highest (K=0.82) and the lowest was between the IOTF and WC (K=0.34). In regard to evaluation of thinness, the agreement between all criteria of BMI was considered moderated to fair (K=0.33-0.51) and a poor level of agreement between the BMI and WC (K=0.14-0.16). Conclusion: The highest level of agreement for thinness, overweight and obesity classification in adolescents was obtained with the IOTF and CDC criteria.

O

bjectives: This study compared the criteria for classification of Body Mass Index (BMI) by Centers for Disease Control and Prevention (CDC), International Obesity Task Force (IOTF) and World Health Organization (WHO) references and by waist circumference (WC). Methods: The sample involved 966 students aged 10 to 16 years. The evaluation of weight status was verified according to CDC, IOTF, WHO criteria and WC curves for Portuguese adolescents. Results: For classification of overweight and obesity, the agreement between the criteria of CDC and IOTF was the

Ref: Ped Endocrinol. Rev. 2014:12(1): Keywords: adolescents, body mass index, obesity, thinness, waist circumference

Introduction The Body Mass Index (BMI) is used as a reliable indicator of body fat for most adolescents (1) but, despite its high sensitivity and specificity, an elevated number of adolescents classified as overweight or obese does not have high adiposity. (2) As an example, in individuals who have low or high stature, a highly developed musculature and/or an abnormal distribution of body fat can be classified incorrectly by BMI. (3) However, the major limitation of the calculation of BMI in

374

Pediatric Endocrinology Reviews (PER) n Volume 12

n

adolescents is the lack of consensus on which cutoff to use in order to classify the thinness, overweight and obesity (4) since in adolescents the relationship between BMI and components of fat and lean mass are not well established.(5) Several criteria classification of BMI have been proposed according to sex and age for adolescents, however there is not a criterion for classification of weight status in this type of population that is universally accepted.(6) The most widely used criteria are recommended by the Centers for Disease Control and Prevention (CDC), the International Obesity Task Force (IOTF) and the World Health Organization (WHO).(7)

No 1 n September 2014

Comparison between CDC, IOTF, WHO References Some studies conducted in different countries compared the classification of weight status in adolescents and found differences between the three classifications criteria. (8-17) The reasons for the observed differences between the classifications of weight status may be due to the methods used to construct the various references (mainly the populations in which they are based and chosen percentiles).(7,15) These differences affect the level of cutoffs and consequently the prevalence calculated using of these different cutoff points. Many countries have their own growth charts based on research in child population in previous years.(18) In Portugal it is not defined what is the best classification criteria of BMI for adolescents, however in 2012, Sardinha et al. (19) developed percentile curves for waist circumference (WC) according to sex and age for Portuguese adolescents. The aim of this study was to compare the three criteria for classification of BMI (CDC, IOTF and WHO) and relate them with the WC curves established for Portuguese adolescents.

Methods The design of this epidemiological study was observational, analytical and cross-sectional. This study was approved by the Ethics Committee of the Regional Health Administration of the Algarve, the Regional Directorate of Education Algarve, the Directorate General for Innovation and Curriculum Development, Ministry of Education, and the directions of Schools that participated in the project.

Population and Sample The population involved students enrolled in public schools (26,217 students between the fifth and ninth school years) from all counties of the Algarve region, of both sexes, aged between 10 and 16 years.(20) Considering the population, the minimum sample size was defined as 777, considering an estimate of the annual prevalence of overweight of 25%, reported in national and international studies, and with an error of 3%.(21-24) Due to administrative reasons, it was decided that 1,000 students should be invited to participate in this study, not only because the possible existence of non-response but also to invite complete classes (corresponding to a margin of error of 2.63%).(25) The inclusion criteria involved: students who were present on the days of data collection, which brought the consent of parents or guardian and who wanted to participate. It was used a stratified random sample defined by different phases: first it was considered the county level, assuming that we can have geographical heterogeneities (within each county schools were selected randomly, if there was more than one

school in the county), and then within each school classes were randomly selected, until the desired number of students per school was obtained. The dimension of samples by counties took were proportional to the number of students enrolled in each county in public school.

Measures Body Mass Index (BMI) For the measurement of body weight, we used a digital scale brand SECA 780 with a capacity of 150 kg and 100 g of precision, and the measurement of height was performed using a stadiometer 200 cm. In both measurements, the students were standing erect without shoes. For measuring the height, the student was placed back to the instrument, with the head oriented in alignment with Frankfurt plane according to standard procedures.(1) Following the table of percentile BMI for age from the CDC, individuals who had values below the 5th percentile were classified as underweight, those with percentile between 85th and less than 95th classified as overweight and with percentile more than 95th with obesity.(26) According to the limits proposed by Cole et al. (IOTF) the 2 values o f the cutoff greater than or equal to 25kg/m indicated overweight and equal to or greater than 30kg/ 2 m indicated the presence of obesity.(27) The thinness was 2 defined with cutoff values less or equal to 18.5kg/m .(28) WHO uses the z-score which classifies individuals at values below -2 with thinness, while defined overweight individuals at z-score greater than 1 and less than 2, and obesity is obtained with the presence of a z-score above 2. Adolescents with a z-score between -2 and 1 are considered with an adequate weight.(29)

Waist Circumference (WC) The WC was performed in the horizontal plane, with the individual in the standing position, naked at the abdomen and after a normal expiration. Two consecutive measurements were made and the average was obtained. This evaluation was performed by a single evaluator. It was used an inelastic and flexible tape measure, with 150 cm and a resolution of 1 mm and it was positioned at the midpoint between the last rib and the upper edge of the iliac crest. The WC was related to abdominal percentile table set for Portuguese adolescents.(19) The abdominal obesity was defined with a value of abdominal percentile equal to or above percentile 90. A value between percentile 75 and percentile 90 was defined as a high value of WC.(30) Thinness was identified with WC equal to or less than percentile 5.(26,28)


n


375

Comparison between CDC, IOTF, WHO References

Data Analysis In a first approach descriptive statistics were made to all variables of this study. The Kappa statistics (29) was used to analyze the correlation between the three criteria of BMI and classification of WC. The statistical analysis was performed with Statistical Package for Social Sciences version 19.0. Statistical significance was set at 0.05. Based on the definition of two different phases of growth period, two groups were considered in analyses: group 1 students with age between 10 and 12 years; group 2 - students between 13 and 16 years. It is assumed that individuals from group 1 aren’t yet in the period of pronounced growth, primarily boys, and between 13 and 16 years the adolescents are in the period of accelerated pubertal growth.(32)

Results From 16 counties of the Algarve region, only 2 counties refused to participate in the study by logistical issues. The minimum number set as sample size with a margin of error of 3% was exceeded, with a sample of 966 students (corresponding to a margin of 2.68%), aged between 10 and 16 years (12.24±1.53 years), where 437 (45.2%) were male and 529 (54.8%) females. Students were divided into the 2 pre-defined age groups, with Boys

Criteria for classification of weight status

CDC

IOTF

WHO

WC

574 (59.4%) aged between 10 and 12 years and 392 (40.6%) with 13 to 16 years. Averages and standard deviations of weight, height, BMI and 2 WC were 48.24 Kg±11.68 kg; 1.55 m±0.11 m; 19.82 kg/m ±3.46 2 kg/m ; 73.11 cm± 9.67 cm, respectively. Table 1 show the values of the prevalence of underweight, overweight and obesity according to the BMI criteria classification of the CDC, IOTF and WHO and WC classification in accordance with sex and age group. The WHO criteria, followed by the IOTF, estimate the highest values of prevalence of overweight and obesity compared to the CDC. In regard to the values of WC, these were higher than those obtained with the WHO classification. As for thinness, the results obtained with the WHO classification showed the highest values, followed by the IOTF, CDC and WC. Table 2 shows the values of agreement between the three criteria for assessment of weight status by BMI and WC. In boys and girls, it was observed that in both age groups, the correlation between the CDC and IOTF was the highest obtained. Lower values of agreement occurred between the WC with the three criteria of BMI for the whole sample. Table 3 shows the correlation between the three criteria used to classify overweight and obesity and WC. Both in boys as in girls in both age groups showed a greater agreement (good and very good) between the CDC and IOTF criteria (K=0.79; Girls

Total (n=966)

Total (n= 437)

1 0 - 1 2 1 3 - 1 6 T o t a l 1 0 - 1 2 1 3 - 1 6 (n=234) (n=203) (n=529) (n=340) (n=189)

Thinness

18 (1.9%)

5 (1.1%)

1 (0.4%)

4 (2%)

13 (2.5%)

9 (2.6%)

4 (2.1%)

Normal Weight

704 (72.9%)

337 (77.1%)

164 (70.1%)

173 (85.2%)

367 (69.4%)

222 (65.3%)

145 (76.7%)

Overweight

153 (15.8%)

57 (13%)

40 (17.1%)

17 (8.4%)

96 (18.1%)

65 (19.1%)

31 (16.4%)

Obesity

91 (9.4%)

38 (8.7%)

29 (12.4%)

9 (4.4%)

53 (10%)

44 (12.9%)

9 (4.8%)

Thinness

28 (2.9%)

7 (1.6%)

2 (0.9%)

5 (2.5%)

21 (4%)

18 (5.3%)

3 (1.6%)

Normal Weight

708 (73.3%)

340 (77.8%)

169 (72.2%)

171 (84.2%)

368 (69.6%)

222 (65.3%)

146 (77.2%)

Overweight

178 (18.4%)

73 (16.7%)

50 (21.4%)

23 (11.3%)

105 (19.8%)

72 (21.2%)

33 (17.5%)

Obesity

52 (5.4%)

17 (3.9%)

13 (5.6%)

4 (2%)

35 (6.6%)

28 (8.2%)

7 (3.7%)

Thinness

81 (8.4%)

31 (7.1%)

8 (3.4%)

23 (11.3%)

50 (9.5%)

33 (9.7%)

17 (9%)

Normal Weight

580 (60%)

274 (62.7%)

136 (58.1%)

138 (68%)

306 (57.8%)

181 (53.2%)

125 (66.1%)

Overweight

200 (20.7%)

84 (19.2%)

53 (22.6%)

31 (15.3%)

116 (21.9%)

79 (23.2%)

37 (19.6%)

Obesity

105 (10.9%)

48 (11%)

37 (15.8%)

11 (5.4%)

57 (10.8%)

47 (13.8%)

10 (5.3%)

Thinness

24 (2.5%)

15 (3.4%)

9 (3.8%)

6 (3%)

9 (1.7%)

7 (2.1%)

2 (1.1%)

Normal Weight

545 (56.4%)

269 (61.6%)

136 (58.1%)

133 (65.5%)

276 (52.2%)

178 (52.4%)

98 (51.9%)

Overweight

182 (18.8%)

86 (19.7%)

41 (17.5%)

45 (22.2%)

96 (18.1%)

52 (15.3%)

44 (23.3%)

Obesity

215 (22.3%)

67 (15.3%)

48 (20.5%)

19 (9.4%)

148 (28%)

103 (30.3%)

45 (23.8%)

Table 1: Values of relative and absolute frequencies of weight status according to the CDC, IOTF and WHO criteria for classification of BMI and to the WC classification

376


n


Comparison between CDC, IOTF, WHO References K=0.84). Lower values of agreement occurred between the WC with the three criteria of BMI in the whole sample. Table 4 presents the agreement for the classification of thinness for BMI according to three criteria and the WC. Age group 10-12 years

13-16 years

Total

Gender

CDC vs IOTF

CDC vs WHO

The agreement between all criteria of BMI was considered moderate to fair for both sexes and age groups (K=0.33-0.51). The comparisons between the criteria of BMI and WC had poor level agreement (K=0.14-0.16). IOTF vs WHO

CDC vs WC

IOTF vs WC

WHO vs WC

Boys

0.763

0.680

0.534

0.499

0.388

0.506

Girls

0.763

0.700

0.684

0.391

0.294

0.391

Total

0.763

0.693

0.628

0.433

0.331

0.435

Boys

0.764

0.524

0.515

0.317

0.297

0.288

Girls

0.847

0.710

0.695

0.301

0.259

0.257

Total

0.827

0.620

0.608

0.313

0.281

0.273

Boys

0.768

0.624

0.532

0.430

0.354

0.417

Girls

0.797

0.706

0.690

0.361

0.282

0.345

Total

0.786

0.672

0.625

0.391

0.314

0.376

Table 2: Kappa coefficient for the classification of weight status (thinness, normal weight, overweight, obesity) according to the three criteria for the classification of BMI and WC. Note: All Kappa coefficients were statistical significant (p