Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Body mass index vs deuterium dilution method for establishing childhood obesity prevalence, Ghana, Kenya, Mauritius, Morocco, Namibia, Senegal, Tunisia and United Republic of Tanzania

Body mass index vs deuterium dilution method for establishing childhood obesity prevalence,... Research Body mass index vs deuterium dilution method for establishing childhood obesity prevalence, Ghana, Kenya, Mauritius, Morocco, Namibia, Senegal, Tunisia and United Republic of Tanzania a b c d e Adama Diouf, Theodosia Adom, Abdel Aouidet, Asmaa El Hamdouchi, Noorjehan I Joonas, f g h f i Cornelia U Loechl, Germana H Leyna, Dorcus Mbithe, Thabisile Moleah, Andries Monyeki, j k l Hilde Liisa Nashandi, Serge MA Somda & John J Reilly Objective To compare the World Health Organization ( WHO) body mass index (BMI)-for-age definition of obesity against measured body fatness in African children. Methods In a prospective multicentre study over 2013 to 2017, we recruited 1516 participants aged 8 to 11 years old from urban areas of eight countries (Ghana, Kenya, Mauritius, Morocco, Namibia, Senegal, Tunisia and United Republic of Tanzania). We measured height and weight and calculated BMI-for-age using WHO standards. We measured body fatness using the deuterium dilution method and defined excessive body fat percentage as > 25% in boys and > 30% in girls. We calculated the sensitivity and specificity of BMI z-score > +2.00 standard deviations (SD) and used receiver operating characteristic analysis and the Youden index to determine the optimal BMI z-score cut-off for classifying excessive fatness. Findings The prevalence of excessive fatness was over three times higher than BMI-for-age-defined obesity: 29.1% (95% CI: 26.8 to 31.4; 441 children) versus 8.8% (95% CI: 7.5 to 10.4; 134 children). The sensitivity of BMI z-score > +2.00 SD was low (29.7%, 95% CI: 25.5 to 34.2) and specificity was high (99.7%, 95% CI: 99.2 to 99.9). The receiver operating characteristic analysis found that a BMI z-score +0.58 SD would optimize sensitivity, and at this cut-off the area under the curve was 0.86, sensitivity 71.9% (95% CI: 67.4 to 76.0) and specificity 91.1% (95% CI: 89.2 to 92.7). Conclusion While BMI remains a practical tool for obesity surveillance, it underestimates excessive fatness and this should be considered when planning future African responses to the childhood obesity pandemic. tifies children with the highest body fatness and the highest Introduction risk of co-morbidities. However, the indicator is conservative Childhood obesity is now a pandemic, heralding a substantial as it fails to identify children who are excessively fat, but who 1,2 7–10 burden of future noncommunicable diseases, despite the do not have high BMI-for-age. There are several problems established burden of underweight in low- and middle-income with this evidence. First, few studies tested the diagnostic countries. Changes in diet and reduced physical activity among performance of the WHO BMI-for-age definition of obesity, adolescent boys and girls have occurred across Africa, cur- focusing on definitions based on national BMI reference 9,10 rently most evident in urban areas, although rural areas are data or the International Obesity Task Force definition. also ae ff cted. The World Health Organization (WHO) Report Second, few studies assessed the diagnostic performance of of the Commission on Ending Childhood Obesity advocated BMI-for-age against a measure of body fatness with low bias more surveillance of the prevalence of obesity to plan where and acceptable individual diagnostic accuracy such as total 11,12 and when to intervene, and to measure the effectiveness of body water. Finally, the applicability of the evidence to future interventions. African children is unclear; bias in the estimation of excessive Body mass index (BMI)-for-age is a well-established in- body fatness by BMI varies across populations in adults. The dicator for surveillance of paediatric obesity. The WHO child extent to which such bias is population-specific for children growth standards define obesity in school-aged children as too is less clear, although compared with Europeans, South- BMI z-score > +2.00 standard deviations (SD). Systematic East Asian children have higher body fatness than would be 6 14 reviews have shown that, as in adults, high BMI-for-age iden- expected from their BMI. Laboratoire de Nutrition, Département de Biologie Animale, Faculté des Sciences et Techniques, Université Cheikh Anta Diop de Dakar, BP 5005 Dakar Fann, Senegal. Nutrition Research Centre, Ghana Atomic Energy Commission, Accra, Ghana. Association Tunisienne des Sciences de la Nutrition, Tunis, Tunisia. Unité Mixte de Recherche Nutrition et Alimentation CNESTEN-Université Ibn Tofail, Rabat, Morocco. Biochemistry Department; Victoria Hospital; Ministry of Health and Quality of Life, Quatre Bornes, Mauritius. International Atomic Energy Agency, Vienna International Centre, Vienna, Austria. Department of Epidemiology and Biostatistics, School of Public Health and Social Sciences, Dar el Salaam, United Republic of Tanzania. Department of Food, Nutrition and Dietetics, Kenyatta University, Nairobi, Kenya. Physical Activity, Sport and Recreation, Faculty of Health Sciences, North-West University, Potchefstroom, South Africa. Ministry of Health and Social Services, Windhoek, Namibia. Centre Muraz, Bobo-Dioulasso, Burkina Faso. School of Psychological Science and Health, University of Strathclyde, Glasgow, Scotland. Correspondence to Adama Diouf (email: [email protected]). (Submitted: 15 November 2017 – Revised version received: 25 July 2018 – Accepted: 10 August 2018 – Published online: 10 September 2018 ) 772 Bull World Health Organ 2018;96:772–781 | doi: http://dx.doi.org/10.2471/BLT.17.205948 Research Measuring childhood obesity in Africa Adama Diouf et al. The aim of this study was to com- South Africa. The height of children was are laboratory-based and impractical pare the prevalence of the WHO BMI- measured to the nearest 0.1 cm using a for large-scale epidemiological studies. for-age definition of obesity against the Seca stadiometer, and weight to 0.1 kg While not a criterion method, body prevalence of excessive body fatness in light indoor clothing using a Seca fatness measured by total body water in a relatively large sample of African scale (Seca, London, England). From is practical for large epidemiological children. the height and weight measures, we studies and provides accurate measures calculated BMI for each child as weight of fatness which are unbiased relative to 2 11,12 divided by height squared (kg/m ) and multicomponent methods. Methods then computed age- and sex-specific The total body water measures were Study design z-score relative to the WHO BMI-for- made on the same day as the height, age reference using the Stata zanthro weight and waist circumference mea- This design for this prospective, multi- package (Stata Corp., College Station, sures. Accurate measurement of total centre, data-pooling study, was agreed United States of America). We defined body water requires a normal hydration at the first meeting of the Reducing obesity as BMI z-score > +2.00 SD and status. We therefore asked participants Obesity Using Nuclear Techniques To overweight (including obesity) as BMI and their families to have normal fluid Design Interventions study in 2012. We z-score > +1.00 SD and food intake on the day before the followed the Standards for Reporting estimation of total body water and to of Diagnostic Accuracy Studies for Body water measures avoid vigorous exercise after the final the conduct and reporting of the study. We aimed to measure total body water meal of the previous day to avoid dehy- Sampling and study procedures origi- in all participants using the deuterium dration and depletion of glycogen stores. nally took place across 11 African cen- dilution method, as described previ- Deuterium oxide-labelled water (99.8% tres between 2013 and 2017. We aimed 11,14 ously. We used standard operating purity; Cambridge Isotope Laboratories to recruit around 150 participants per procedures, with training support pro- Inc., Andover, USA) accurately weighed country (a larger sample was used in the vided for all countries via a combination (0.001 g precision) was orally adminis- United Republic of Tanzania, because of residential and on-site training by tered to the children according to their some of the study aims there required experts recruited by the International body weight (0.5 g deuterium oxide per a larger sample). As the nutrition and Atomic Energy Agency. Ideally, body kg) followed by 50 mL of local tap water. physical activity transitions in Africa fatness measurement methods are mul- Children were asked not to eat or drink have disproportionately affected urban ticomponent, based on measures of total for at least 30 minutes before receiving 3,4 children, we focused the sampling in body water plus body density or total the deuterium-labelled water and to void urban areas. In each country, we used a body mineral. However, such methods their bladders before dosing. Baseline multistage random sampling method to select at least four to five urban public schools in one district or state, followed Fig. 1. Flowchart on the inclusion of participants to compare methods to measure by school sampling frames of all classes overweight in children in eight African countries, 2013–2017 corresponding to the target age group and sex. More details of the methods are available from the corresponding author. 2172 eligible participants recruited and Children meeting the inclusion BMI and body composition measured criteria were recruited to participate in the study after submission of a signed Data excluded from 656 participants informed consent form by a parent. • data not collected or not analysed Data collection was conducted during (155 participants) the school year. Ethical approval was • data rejected by quality control obtained from local research boards or (501 participants) committees in each country. Participants were eligible for inclusion if they were age 8 to 11 years and provided consent Data analysed from 1516 participants or assent for participation; they were excluded if they were outside the study age range, had ill health that would have 134 participants classified as 1382 participants classified as precluded participation or were not obese by BMI non-obese by BMI present in school after two consecutive visits. Anthropometric measures Final classification Final classification • 310 obese by body fatness • 131 obese by body fatness The study used a common protocol and • 1072 non-obese by body fatness • 3 non-obese by body fatness standard operating procedures across all countries. Before data collection BMI: body mass index. started all researchers were trained in Notes: Children were originally recruited and assessed for BMI and body composition in 11 countries. data collection methods by a team of Data were rejected for quality control reasons from Benin, Mali and Uganda. The final analysis was experienced researchers and fieldwork - therefore based on data from eight countries: Ghana, Kenya, Mauritius, Morocco, Namibia, Senegal, Tunisia and United Republic of Tanzania. ers during a 1-week residential course in Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 773 Research Adama Diouf et al. Measuring childhood obesity in Africa (pre-dose) saliva samples were collected the labelled water dose under supervi- was carried out with Fourier transform from each participant by rotating a sion and two further saliva samples were infrared spectroscopy (FTIR 8400S cotton-wool ball in the buccal cavity of collected at 3 hours and 4 hours ae ft r the spectrophotometer, Shimadzu Kyoto, the mouth until well soaked. Saliva was dose using the method described above. Japan) in accordance with International collected into a clean sterile and dry tube All saliva samples were stored at 4 °C Atomic Energy Agency protocols. using a 20 mL disposable syringe. Par- until their arrival to the laboratory for We converted measures of total ticipants were then requested to drink storage at −20 °C until analysis. Analysis body water to total body fat using es- Table 1. Age and anthropometric characteristics of participants in the study of body mass index-for-age and body fatness among children in eight African countries, 2013–2017 Sex, by country No. (%) of Median (IQR) children Age, years BMI z-score Body fat percentage Fat mass index, Fat-free mass index, 2 2 kg/m kg/m Ghana Boys 71 (37.4) 10 (9–11) −0.87 (−1.17 to −0.21) 15.92 (12.57 to 19.94) 2.37 (1.98 to 3.24) 12.98 (12.08 to 13.75) Girls 119 (62.6) 10 (9–10) −0.66 (−1.16 to 0.09) 18.72 (15.65 to 22.62) 2.85 (2.33 to 3.74) 12.63 (11.77 to 13.38) Total 190 (100.0) 10 (9–11) −0.70 (−1.16 to −0.05) 18.03 (14.40 to 21.08) 2.65 (2.14 to 3.42) 12.77 (11.95 to 13.70) Kenya Boys 84 (46.9) 10 (9–11) −0.91 (−1.34 to −0.30) 22.94 (17.93 to 28.61) 3.58 (2.68 to 5.01) 12.29 (10.92 to 13.64) Girls 95 (53.1) 10 (9–11) −0.69 (−1.36 to −0.05) 24.14 (19.43 to 27.49) 3.71 (2.85 to 4.86) 12.47 (10.73 to 14.53) Total 179 (100.0) 10 (9–11) −0.82 (−1.35 to −0.15) 23.57 (19.34 to 28.11) 3.64 (2.75 to 4.86) 12.34 (10.84 to 13.90) Mauritius Boys 82 (53.6) 10 (9–11) 0.76 (−1.01 to 1.86) 25.28 (18.67 to 33.23) 4.12 (2.75 to 7.36) 13.14 (11.74 to 14.67) Girls 71 (46.4) 10 (9–11) 0.56 (−0.55 to 1.84) 32.11 (24.62 to 37.66) 5.67 (3.96 to 8.39) 12.64 (11.44 to 14.68) Total 153 (100.0) 10 (9–11) 0.68 (−0.76 to 1.84) 28.80 (21.65 to 35.48) 4.96 (3.41 to 7.71) 13.01 (11.64 to 14.68) Morocco Boys 94 (50.3) 9 (8–10) −0.24 (−1.00 to 0.51) 19.76 (16.31 to 24.57) 3.08 (2.49 to 4.17) 12.76 (12.15 to 13.66) Girls 93 (49.7) 9 (8–10) −0.33 (−0.99 to 0.42) 25.69 (21.91 to 30.11) 4.07 (3.20 to 5.07) 11.97 (11.06 to 12.62) Total 187 (100.0) 9 (8–10) −0.27 (−0.99 to 0.51) 23.23 (18.30 to 28.60) 3.70 (2.76 to 4.72) 12.36 (11.63 to 13.27) Namibia Boys 66 (43.7) 10 (9–11) −0.08 (−0.91 to 1.09) 22.84 (18.85 to 30.76) 3.60 (2.70 to 5.73) 12.92 (12.06 to 13.69) Girls 85 (56.3) 10 (9–11) 0.42 (−0.76 to 1.64) 32.69 (26.76 to 39.06) 5.38 (4.14 to 8.71) 11.94 (11.01 to 13.18) Total 151 (100.0) 10 (9–11) 0.19 (−0.84 to 1.44) 27.97 (22.32 to 37.50) 4.70 (3.38 to 7.42) 12.59 (11.44 to 13.41) Senegal Boys 70 (47.9) 10 (9–11) −1.29 (−1.84 to −0.71) 13.43 (10.64 to 19.99) 1.95 (1.49 to 3.07) 12.50 (11.72 to 13.10) Girls 76 (52.1) 10 (9–10) −1.40 (−2.15 to −0.58) 19.30 (15.80 to 24.91) 2.62 (2.13 to 3.52) 11.41 (10.71 to 12.04) Total 146 (100.0) 10 (9–10) −1.32 (−2.05 to −0.60) 16.70 (12.76 to 22.61) 2.35 (1.79 to 3.34) 11.84 (11.12 to 12.66) Tunisia Boys 80 (51.0) 9 (9–10) 0.04 (−0.65 to 0.99) 23.49 (20.30 to 26.86) 3.86 (3.19 to 5.00) 12.56 (11.89 to 13.75) Girls 77 (49.0) 10 (8–10) 0.31 (−0.65 to 1.18) 30.03 (25.57 to 33.89) 4.89 (3.99 to 6.34) 11.94 (11.13 to 12.82) Total 157 (100.0) 9 (8–10) 0.10 (−0.65 to 1.12) 26.03 (22.88 to 31.37) 4.29 (3.54 to 5.77) 12.37 (11.60 to 13.29) United Republic of Tanzania Boys 158 (44.8) 10 (9–11) 0.04 (−0.58 to 1.12) 18.50 (15.10 to 24.90) 3.00 (2.41 to 4.83) 13.42 (12.73 to 14.34) Girls 195 (55.2) 10 (9–11) 0.02 (−0.80 to 0.92) 23.30 (19.30 to 31.10) 3.73 (2.93 to 5.62) 12.70 (11.89 to 13.63) Total 353 (100.0) 10 (9–11) 0.02 (−0.67 to 0.95) 21.50 (17.00 to 29.40) 3.43 (2.60 to 5.37) 13.00 (12.27 to 14.14) Total Boys 705 (46.5) 10 (9–11) −0.37 (−1.09 to 0.69) 20.47 (15.60 to 26.09) 3.25 (2.40 to 4.57) 12.92 (12.06 to 13.90) Girls 811 (53.5) 10 (9–10) −0.33 (−1.09 to 0.72) 24.90 (19.37 to 31.46) 3.91 (2.87 to 5.64) 12.23 (11.31 to 13.31) Total 1516 10 (9–11) −0.35 (−1.09 to 0.71) 22.65 (17.43 to 29.60) 3.59 (2.60 to 5.17) 12.59 (11.64 to 13.63) (100.0) BMI: body mass index; IQR: interquartile range. Number of records removed from original samples: Ghana (4), Kenya (1), Mauritius (3), Morocco (3), Namibia (4), Senegal (10), Tunisia (2) and United Republic of Tanzania (3). 2 5 Notes: BMI was calculated as weight in kg divided by height in m and z-scores were obtained from the World Health Organization BMI-for-age child growth standards. Body fat percentage was measured using deuterium oxide dilution. Fat mass index was calculated as fat mass in kg divided by height in m , with fat mass measured from total body water. Fat free mass index was calculated as fat free mass in kg divided by height in m , with fat-free mass measured from total body water. Measures were made in 2014–2017 in Kenya and United Republic of Tanzania and in 2013–2015 in all other countries. 774 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 Research Measuring childhood obesity in Africa Adama Diouf et al. tablished age- and sex-specific constants not greatly affected by the definition of sensitivity and specificity for identify- 16,17 for the hydration of fat-free mass, as excessive fatness (data are available from ing excessive fatness. We used Spear- described elsewhere. Quality control the corresponding author). man rank-order correlation to test the procedures, with four stringent criteria association between countries’ total Data management described in detail elsewhere, were BMI-for-age z-score and total body fat applied to the measures of enrichment Training in data management, data percentage. We also made an explor- of deuterium required for the total sharing and data quality control was atory analysis of possible geographical body water measures and to the esti- provided during a 1-week data man- differences in the results by grouping mates of total body water, total body agement residential training course in the countries into three geographically fat and body fat percentage. These Benin in 2014. Throughout the study, defined sub-groups: sub-Saharan Africa quality control measures were: (i) deu- support was provided by site visits and (Ghana, Kenya, Namibia, Senegal and terium enrichment of each of the two online chat or email by a central data United Republic of Tanzania), North post-dose samples should be within management coordinator from Burkina Africa (Morocco and Tunisia) and an 2% of the mean of the two post-dose Faso recruited by the International African island (Mauritius). samples; (ii) measured enrichment Atomic Energy Agency. The weight and should lie within an expected range of height measures, BMI-for-age z-scores Results normal enrichments based on the body and total body water-derived measures weight of the child (outliers in the total of body fatness were all made prospec- Fig. 1 shows the flowchart of the study. body-water-to-height relationship were tively and independently and the results Of the 2172 children recruited to the identified and excluded); (iii) outliers in of each measure were not available at the study, eligible data were available from body fat percentage were identified and time of the other measures. 1516 (69.8%). The age and anthropo - excluded (e.g. large mismatches between metric characteristics of the eligible Analysis body fat percentage and BMI z-score participants are shown in Table 1. The or unphysiological body fat percentage We used standard diagnostic perfor- mean age was 9.6 years (95% confidence measures); and (iv) if more than 10% mance indicators to determine the interval, CI: 9.5 to 9.7) and median of total body water measures from any extent to which BMI z-score > +2.00 age was 10 years (interquartile range, centre failed to meet the quality control SD identified children with excessive IQR: 9 to 11). The median BMI-for-age criteria, then we excluded all data from fatness. We calculated sensitivity (pro- z-score was −0.35 (IQR: −1.09 to 0.71) that centre from the pooled analyses. portion of real positive values among all and median body fat percentage was Based on these criteria, we excluded data the recorded positive values), specific- 22.65% (IQR: 17.43 to 29.60). Fig. 2 from three out of 11 original participat- ity (proportion of real negative values provides more detail on the distribu- ing countries (Benin, Mali and Uganda), among all the negative values), and tion of body fatness and BMI-for-age so that the present study is based on positive and negative predictive values z-scores. The prevalence of excessive data from eight countries (Ghana, for the total sample. We used the Youden fatness was 29.1% (95% CI: 26.8 to 31.4; Kenya, Mauritius, Morocco, Namibia, index method to determine the optimal 441 children). Overall, the prevalence of Senegal, Tunisia and United Republic BMI z-score cut-off for optimizing the obesity by the WHO BMI-for-age crite- of Tanzania) and 1516 children. Among these, 2% of total body water measures Fig. 2. Relationships between body mass index-for-age z-score and body fat percentage were rejected for quality control reasons among children in eight African countries, by geographical area, 2013–2017 and were not included in the analyses reported here. We expressed total body fatness 1.00 as a percentage of body weight. Many studies have established that a high body fatness, even in childhood, has a 0.75 range of adverse health consequences, with most focusing on the cardiometa- Area under ROC curve: 0.86 bolic consequences, as summarized by 7,8 systematic reviews. One report on the 0.50 relationship between body fatness and cardio-metabolic risk in childhood used a skinfold thickness method previously validated against a multicomponent 0.25 model to measure body fatness. The researchers found a marked increase in cardiometabolic risk profile at body fat > 25% in boys and > 30% in girls, across 0.25 0.50 0.75 1.00 a wide age range. We therefore used this 1–Specificity, % definition of excessive fatness (true posi- tive in the receiver operator characteris- BMI: body mass index. tic analysis) in the present study. As in Note: Areas were defined as follows: sub-Saharan Africa (Ghana, Kenya, Namibia, Senegal and United 9,10 previous studies, the conclusions were Republic of Tanzania), North Africa (Morocco and Tunisia) and African island (Mauritius). Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 775 Sensitivity, % Research Adama Diouf et al. Measuring childhood obesity in Africa rion was 8.8% (95% CI: 7.5 to 10.4; 134 Table 2. Comparison of obesity defined by body mass index-for-age and by body fatness children) (Table 2) and of overweight among children in eight African countries, by geographical area, 2013–2017 was 19.5% (95% CI: 17.6 to 21.6; 296 children) (Table 3). Obesity defined by BMI- Obesity defined by body fatness, no. (%) of children In the whole sample, the sensitivity for-age No Yes Total of BMI z-score > +2.00 SD for identify- ing excessively fat children was 29.7% Sub-Saharan Africa (95% CI: 25.5 to 34.2), specificity was Ghana 99.7% (95% CI: 99.2 to 99.9), positive No 183 (100.0) 5 (71.4) 188 (99.0) predictive value 97.8% (95% CI: 93.6 Yes 0 (0.0) 2 (28.6) 2 (1.0) to 99.5) and negative predictive value Total 183 (100.0) 7 (100.0) 190 (100.0) 77.6% (95% CI: 75.3 to 79.7). The sen- Kenya sitivity of BMI z-score > +2.00 SD to No 125 (100.0) 50 (92.6) 175 (97.8) identify excessively fat children varied Yes 0 (0.0) 4 (7.4) 4 (2.2) little between boys and girls (66/203 Total 125 (100.0) 54 (100.0) 179 (100.0) for boys versus 65/238 for girls). In the Namibia whole sample BMI z-score > 1.00 SD No 75 (100.0) 49 (64.5) 124 (82.1) had sensitivity of 59.2% (95% CI: 54.4 Yes 0 (0.0) 27 (35.5) 27 (17.9) to 63.8) and specificity of 96.7% (95% Total 75 (100.0) 76 (100.0) 151 (100.0) CI: 95.5 to 97.7; Table 4). Analysis of the data by country and in the three popu- Senegal lation sub-groups is shown in Table 2 No 130 (100.0) 12 (75.0) 142 (97.3) and Table 3. Sensitivity was lower in the Yes 0 (0.0) 4 (25.0) 4 (2.7) North African and Island populations Total 130 (100.0) 16 (100.0) 146 (100.0) than the sub-Saharan Africans. The rank United Republic of Tanzania order correlation between country me- No 259 (98.9) 48 (52.7) 307 (87.0) dian BMI z-score and country fat mass Yes 3 (1.1) 43 (47.3) 46 (13.0) index was high (r = 0.6). Total 262 (100.0) 91 (100.0) 353 (100.0) The receiver operator characteristic All analysis is shown in Fig. 3. The optimal No 772 (99.6) 164 (67.2) 936 (91.9) cut-off point in the BMI-for-age distri- Yes 3 (0.4) 80 (32.8) 83 (8.1) bution for classifying excessive fatness Total 775 (100.0) 244 (100.0) 1019 (100.0) was a BMI z-score of +0.58 SD (Table 4). North Africa At this cut-off the area under the curve Morocco was 0.86, sensitivity was 71.9% (95% No 141 (100.0) 37 (80.4) 178 (95.2) CI: 67.4 to 76.0), specificity 91.1% (95% Yes 0 (0.0) 9 (19.6) 9 (4.8) CI: 89.2 to 92.7), positive predictive value 76.8% (95% CI: 72.4 to 80.7) and Total 141 (100.0) 46 (100.0) 187 (100.0) negative predictive value 88.8% (95% Tunisia CI: 86.7 to 90.6). No 89 (100.0) 60 (88.2) 149 (94.9) Yes 0 (0.0) 8 (11.8) 8 (5.1) Total 89 (100.0) 68 (100.0) 157 (100.0) Discussion All The present study has established the No 230 (100.0) 97 (85.1) 327 (95.1) extent to which the WHO BMI-for-age Yes 0 (0.0) 17 (14.9) 17 (4.9) definition of obesity underestimates Total 230 (100.0) 114 (100.0) 344 (100.0) the prevalence of excessive fatness in African island African children. Excessive fatness was Mauritius present in nearly a third of children, No 70 (100.0) 49 (59.0) 119 (77.8) suggesting that urban African environ- Yes 0 (0.0) 34 (41.0) 34 (22.2) ments are now highly obesogenic even Total 70 (100.0) 83 (100.0) 153 (100.0) for children. Excessive fatness was over All countries three times more common than the No 1072 (99.7) 310 (70.3) 1382 (91.2) prevalence of BMI-defined obesity. This difference is large enough to be mean- Yes 3 (0.3) 131 (29.7) 134 (8.8) ingful for public health. For example, Total 1075 (100.0) 441 (100.0) 1516 (100.0) the case for policy action to prevent and BMI: body mass index. control obesity is much weaker at an ap- We measured height and weight and calculated obesity from BMI-for-age using the World Health Organization reference z-score > +2.00 standard deviations. parent prevalence of around 8% (based We measured body fatness using the deuterium dilution method and defined excessive body fat on BMI-for-age z-score > +2.00 SD in 11,12 percentage as > 25% in boys and > 30% in girls. the present study) than at the preva- lence of around 30% (excessive fatness) 776 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 Research Measuring childhood obesity in Africa Adama Diouf et al. Table 3. Comparison of overweight defined by body mass index-for-age and obesity defined by body fatness among children in eight African countries, by geographical area, 2013–2017 Overweight defined by Obesity defined by body fatness, no. (%) of children BMI-for-age No Yes Total Sub-Saharan Africa Ghana No 182 (99.5) 2 (28.6) 184 (96.8) Yes 1 (0.5) 5 (71.4) 6 (3.2) Total 183 (100.0) 7 (100.0) 190 (100.0) Kenya No 122 (97.6) 48 (88.9) 170 (95.0) Yes 3 (2.4) 6 (11.1) 9 (5.0) Total 125 (100.0) 54 (100.0) 179 (100.0) Namibia No 75 (100.0) 26 (34.2) 101 (66.9) Yes 0 (0.0) 50 (65.8) 50 (33.1) Total 75 (100.0) 76 (100.0) 151 (100.0) Senegal No 130 (100.0) 7 (43.8) 137 (93.8) Yes 0 (0.0) 9 (56.2) 9 (9.2) Total 130 (100.0) 16 (100.0) 146 (100.0) United Republic of Tanzania No 246 (93.9) 21 (23.1) 267 (75.6) Yes 16 (6.1) 70 (76.9) 86 (24.4) Total 262 (100.0) 91 (100.0) 353 (100.0) All No 755 (97.4) 104 (42.6) 859 (84.3) Yes 20 (2.6) 140 (57.4) 160 (15.7) Total 775 (100.0) 244 (100.0) 1019 (100.0) North Africa Morocco No 137 (97.2) 21 (45.7) 158 (84.5) Yes 4 (2.8) 25 (54.3) 29 (15.5) Total 141 (100.0) 46 (100.0) 187 (100.0) Tunisia No 84 (94.4) 29 (42.6) 113 (72.0) Yes 5 (5.6) 39 (57.4) 44 (28.0) Total 89 (100.0) 68 (100.0) 157 (100.0) All No 221 (96.1) 50 (43.9) 271 (78.8) Yes 9 (3.9) 64 (56.1) 73 (21.2) Total 230 (100.0) 114 (100.0) 344 (100.0) African island Mauritius No 64 (91.4) 26 (31.3) 90 (58.8) Yes 6 (8.6) 57 (68.7) 63 (41.2) Total 70 (100.0) 83 (100.0) 153 (100.0) All countries No 1040 (96.7) 180 (40.8) 1220 (80.5) Yes 35 (3.3) 261 (59.2) 296 (19.5) Total 1075 (100.0) 441 (100.0) 1516 (100.0) BMI: body mass index. We measured height and weight and calculated overweight from BMI-for-age using the World Health Organization reference z-score > +1.00 standard deviations. We measured body fatness using the deuterium dilution method and defined excessive body fat 11,12 percentage as > 25% in boys and > 30% in girls. Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 777 Table 4. Comparison of World Health Organization body mass index-for-age cut-offs for obesity and overweight and the empirically determined optimal cut-off for identifying excessive fatness among children in eight African countries, 2013–2017 a b Diagnostic performance measure BMI z-score > +2.00 SD BMI z-score > +1.00 SD BMI z-score +0.58 SD No. of children Total no. % (95% CI) No. of children Total no. % (95% CI) No. of children Total no. % (95% CI) Sensitivity 131 441 29.7 (25.5 to 34.2) 261 441 59.2 (54.4 to 63.8) 317 441 71.9 (67.4 to 76.0) Specificity 1072 1075 99.7 (99.2 to 99.9) 1040 1075 96.7 (95.5 to 97.7) 979 1075 91.1 (89.2 to 92.7) Positive predictive value 131 134 97.8 (93.6 to 99.5) 261 296 88.2 (83.9 to 91.6) 317 413 76.8 (72.4 to 80.7) Negative predictive value 1072 1382 77.6 (75.3 to 79.7) 1040 1220 85.2 (83.1 to 87.2) 979 1103 88.8 (86.7 to 90.6) BMI: body mass index; CI: confidence interval; SD: standard deviation. We calculated values as follows: sensitivity: [true positives/(true positives + false negatives)]; specificity: [true negatives/(true negatives + false positives)]; positive predictive value: [true positives/(true positives + false positives)]; negative predictive value: [true negatives/(true negatives + false negatives)]. We calculated the optimal cut-off z-score from the receiver operating characteristic curve (area under the curve: 0.86). Research Adama Diouf et al. Measuring childhood obesity in Africa observed. To improve the estimation of Fig. 3. Receiver operator characteristic analysis of the ability of body mass index-for- prevalence, cut-off points in the BMI age z-score to identify children with excessive fatness in eight African countries, distribution lower than the z-score of 2013–2017 +2.00 SD might be considered. At BMI z-score > +1.00 SD the ability to identify Sub-Saharan Africa North Africa over-fatness was improved but not opti- 60 60 mized. The optimal BMI z-score cut-off for classifying excessive fatness (which 40 40 maximized the area under the curve) in our study was +0.58 SD. 20 20 There are no directly comparable studies in African children, or using the 0 0 WHO-recommended definition based -4 -2 0+2+4-4-20 +2 +4 on BMI, but in non-African popula- BMI-for-age z-score BMI-for-age z-score tions biases have been reported for other 9,10 African island Total BMI-based definitions of obesity The 60 60 present study adds to previous stud- ies suggesting that underestimation 40 40 of excessive fatness by BMI-for-age criteria is likely to be a global cause for 19,20 20 20 concern. Our study shows that a high proportion of African children with 0 0 apparently healthy BMI-for-age have ex- -4 -2 0+2+4 -4 -2 0+2+4 cessive body fatness. The bias observed BMI-for-age z-score BMI-for-age z-score is unlikely to be due to a high body fat Boys Girls Fitted values percentage secondary to unusually low fat-free mass (lean body mass). This is because of the consistency between the Accuracy Studies guidance in both the or delay future obesity prevention and findings of the present study and studies conduct and reporting of the study; and control efforts in Africa. Further re- 9,10 for other populations. Furthermore, the standardization and quality control search is needed to determine whether median fat mass index values, which of both the study measurement methods the sensitivity of the BMI-for-age indica- measure fatness relatively independent and data management. A key limitation tor is especially low in African children of fat-free mass, were high in the pres- of the study was that we were unable to compared with other populations. ■ ent study. Reference data for fat mass test definitively for differences in the di- index from British children of the same agnostic accuracy of BMI-for-age across Acknowledgements age (and measured in 2001, long after different populations of African chil- We thank the participating centres and the childhood obesity epidemic had af- dren. Our exploratory comparison of countries, the Ministry of Higher Educa- fected children in the United Kingdom country groups by sub-Saharan Africa, tion and Scientific Research of Senegal of Great Britain and Northern Ireland) North Africa and an island population (PAPES) and Ministry of Education were very similar to those in the present was underpowered. A further limita- (Senegal), Centre National d’Energie des study: 50th centile of 3.4 kg/m for boys tion is generalizability. The participant Sciences et des methods Nucléaires (Mo- and 4.2 kg/m for girls compared with age range of the present study limits rocco), Ministry of Health and Quality 2 2 3.25 kg/m for boys and 3.91 kg/m for our conclusions to 8 to 11 year olds, of Life and the University of Fribourg, girls in the present study. Our findings although our findings are consistent Switzerland (Mauritius), Kenyatta Uni- are consistent with the evidence that with those reported for younger and versity, Nairobi, Kenya, Ecole Supérieure body fatness of contemporary children older participants, including adults, in des Sciences et Techniques de la Santé de is higher, across the range of body fat- systematic reviews of studies from non- Tunis, Université Tunis El Manar 6,9,10 ness, than that of children in the recent African populations. 23–25 past. In conclusion, excessive fatness is Funding: e Th study was partly funded by The main strengths of the present now prevalent among urban popula- the International Atomic Energy Agency, study were the large sample size and tions of African children and is likely (RAF/6/402). narrow age range of the sample; the to have serious future public health novelty of using the WHO BMI-for-age implications. While at a group level Competing interests: None declared. definition in an African setting; the the BMI z-score and body fatness were novelty and value of having an unbiased related, BMI-for-age substantially un- definition of body fatness; the use of the derestimated the scale of the problem Standards for Reporting of Diagnostic of excessive fatness and so may hinder 778 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 Percentage fat mass Percentage fat mass Research Measuring childhood obesity in Africa Adama Diouf et al. صخلم ،برغلماو ،لاغنسلا ،ةلوفطلا ةلحرم في ةنمسلا راشتنا ديدحتل ،مويرتويدلا فيفتخ ةقيرط لباقم مسلجا ةلتك شرؤم ايبيمانو ،سويشيرومو ،اينيكو ،اناغو ،ةدحتلما اينازنت ةيروهجمو ،سنوتو نم تارم ثلاث نم رثكأ لىعأ ةطرفلما ةنمسلا راشتنا ناك جئاتنلا فيرعتو ،ةقرافلأا لافطلأا في ةسيقلما ةنمسلا ينب ةنراقم ضرغلا 29.1٪ :نسلل ةبسنلاب مسلجا ةلتك شرؤم اهدديح يتلا ةنمسلا ،نسلل ةبسنلاب ( مسلجا BMI ةلتك ) شرؤم لىع مئاقلا ةنادبلا 8.8٪ لباقملافط ( 441 ؛31.4 لىإ 26.8 :95٪ ةقثلا لصاف) .ةيلماعلا ةحصلا ةمظنلم عباتلاو ةيساسح تناك. (لافط 134 ،10.4 لىإ 7.5 :95٪ ةقثلا لصاف ) 2013 نم ةترفلا للاخ زكارلما ةددعتم ةيقابتسا ةسارد في ةقيرطلا ةلصحملل ةبسنلاب نم +2.00 بركلأا - (SD) يرايعلما فارحنلاا لىإ 8 ينب مهرماعأ حواترت ا كراشم 1516 ليجستب انمق ، لىإ2017 :95٪ ةقث لصافب ،29.7٪ ) ،ةضفخنم -مسلجا ةلتك شرؤ ةيرايعلما لم ،برغلماو ،لاغنسلا نادلب ) نيماث في ةيضرلحا قطانلما نم ا ماع 11 لصافب ،99.7٪ ) ةيلاع ةيصوصلخا تناكو (34.2 لىإ 25.5 ،سويشيرومو ،اينيكو ،اناغو ،ةدحتلما اينازنت ةيروهجمو ،سنوتو ليغشتلا صئاصخ ليلتح فشتكا.(99.9 لىإ 99.2 :95٪ ةقث مسلجا ةلتك شرؤم باسحو نزولاو لوطلا سايقب (ايبيمانو انمق. ةلصحملل -+0.58 ةبسنب يرايعلما فارحنلاا نأ بقتسملل ل .ةيلماعلا ةحصلا ةمظنم يرياعم مادختساب ،نسلل ةبسنلاب (BMI) دنع هنأو ،ةيساسلحا نم نس يح ُ فوس -مسلجا ةلتك شرؤلم ةيرايعلما ،موييرتويدلا فيفتخ ةقيرط مادختساب مسلجا ةنمس سايقب انمق ةيساسلحا تناكو ،0.86 ىنحنلما تتح ةحاسلما تناك دلحا اذه نم رثكأ نوكتل مسلجا في ةدئازلا نوهدلل ةيوئلما ةبسنلا انددحو ةيصوصخو ،(76.0 لىإ 67.4 :95٪ ةقث لصاف 71.9٪ ) ةيساسح باسحب انمق.تانبل نم ا في رثكأو 30٪ ،دلاولأا في 25٪ .(92.7 لىإ 89.2 :95٪ ةقث لصاف) 91.1٪ نم بركأ نكتل مسلجا ةلتك شرؤلم ةيرايعلما ةلصحلما ةيصوصخو ،ةنمسلا ةبقارلم ةيلمع ةادأ مسلجا ةلتك شرؤم لظي مانيب جاتنتسلاا ليلحتب انعتساو ) ةيرايعلما ،(SD تافارحنلال ةبسنلاب +2.00 دنع رابتعلاا في كلذ عضو بيجو ،ةطرفلما ةنمسلا نم للقي هنأ لاإ لثملأا دلحا ديدحتلندوي شرؤمو لبقتسملل ليغشتلا صئاصخ ةلحرم في ةنادبلا ءابول ةيلبقتسم ةيقيرفأ تاباجتسلا ةنمسلا طيطختلا فينصت فدبه ،مسلجا ةلتك شرؤلم ةيرايعلما ةلصحملل .ةلوفطلا .ةطرفلما 摘要 加纳，肯尼亚，摩洛哥，毛里求斯，纳米比亚，塞内加尔，突尼斯和坦桑尼亚联合共和国各国通过身体质量指 数与重水同位素法确定儿童肥胖患病率 目的 将世卫组织 (WHO) 年龄别身体质量指数 (BMI- 结果 过度肥胖的患病率比年龄别身体质量指数所定义 for-age) 对肥胖的定义与所测量的非洲儿童体脂率进行 的肥胖高出三倍以上 ： 29.1％（95％ 置信区间，CI ： 对比。 26.8 至 31.4 ； 441 名儿童）与 8.8％（95％ 置信区间， 方法 在 2013 年至 2017 年的一项前瞻性、多中心的研 CI ： 7.5 至 10.4 ； 134 名儿童） 。BMI z 评分 >+2.00 标 究中，我们从 8 个国家的城市地区（加纳，肯尼亚， 准差 (SD) 的敏感性低（29.7％，95％ 置信区间，CI ： 摩洛哥，毛里求斯，纳米比亚，塞内加尔，突尼斯和 25.5 至 34.2） ，特异性高（99.7 ％，95 ％ 置信区间， 坦桑尼亚联合共和国）招募了 1516 名 8 至 11 岁的参 CI ： 99.2 至 99.9） 。受试者工作特征分析发现 BMI z 评 与者。我们使用 WHO 标准测量了参与者的身高和体 分 +0.58 标准差 (SD) 将优化敏感性，并在此界限值时， 重，并计算了年龄别身体质量指数。我们使用重水同 曲线下区域为 0.86，敏感性为 71.9％（95％ 置信区间， 位素法测量了参与者的体脂，并定义过度肥胖百分位 CI ： 67.4 至 76.0） ，特异性为 91.1％（95％ 置信区间， 数为男孩 >25％，女孩 >30％。我们计算了 BMI z 评 CI ： 89.2 至 92.7）。 分 >+2.00 标准差 (SD) 的敏感性和特异性，并使用受 结论 虽然身体质量指数 (BMI) 仍然是肥胖监测的实用 试者工作特征分析和约登指数来确定过度肥胖分类的 工具，但它低估了过度肥胖的患病率，因此，在规划 最佳 BMI z 评分的界限值。 未来非洲应对儿童肥胖率快速上升的情况时应慎重考 虑。 Résumé Indice de masse corporelle vs méthode de dilution du deutérium pour établir la prévalence de l'obésité chez l'enfant au Ghana, au Kenya, au Maroc, à Maurice, en Namibie, en République-Unie de Tanzanie, au Sénégal et en Tunisie Objectif Comparer la définition de l'obésité de l'Organisation mondiale et défini le taux de masse grasse excessive comme étant > 25% pour de la Santé basée sur l'indice de masse corporelle (IMC) selon l'âge à la les garçons et > 30% pour les filles. Nous avons calculé la sensibilité et masse grasse mesurée chez les enfants africains. la spécificité du Z-score de l'IMC > +2,00 écarts types (ET ) et utilisé une Méthodes Dans le cadre d'une étude prospective multicentrique analyse de la fonction d'efficacité du récepteur et l’indice de Youden menée entre 2013 et 2017, nous avons recruté 1516 participants âgés afin de déterminer la valeur limite optimale du Z-score de l'IMC pour de 8 à 11 ans dans des zones urbaines situées dans huit pays (Ghana, classifier la masse grasse excessive. Kenya, Maroc, Maurice, Namibie, République-Unie de Tanzanie, Sénégal Résultats La prévalence de la masse grasse excessive était plus de trois et Tunisie). Nous avons mesuré leur taille et leur poids et calculé leur fois supérieure à la prévalence de l'obésité définie en fonction de l'IMC IMC par rapport à leur âge en utilisant les normes de l'OMS. Nous avons selon l'âge: 29,1% (IC à 95%: 26,8-31,4; 441 enfants) contre 8,8% (IC à mesuré la masse grasse à l'aide de la méthode de dilution du deutérium, 95%: 7,5-10,4; 134 enfants). La sensibilité du Z-score de l'IMC > +2,00 ET Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 779 Research Adama Diouf et al. Measuring childhood obesity in Africa était faible (29,7%, IC à 95%: 25,5-34,2), tandis que la spécificité était Conclusion Si l'IMC reste est un outil pratique pour surveiller l'obésité, élevée (99,7%, IC à 95%: 99,2-99,9). L'analyse de la fonction d'efficacité il sous-évalue la masse grasse excessive. Cela doit être pris en compte du récepteur a révélé qu'un Z-score de l'IMC de +0,58 ED optimiserait la lors de la planification des futures mesures africaines de lutte contre la sensibilité, et qu'à cette valeur limite, l'aire sous la courbe était de 0,86, pandémie d'obésité chez l'enfant. la sensibilité de 71,9% (IC à 95%: 67,4-76,0) et la spécificité de 91,1% (IC à 95%: 89,2-92,7). Резюме Сравнение эффективности измерения индекса массы тела и метода дейтериевого разбавления для определения распространенности детского ожирения (Гана, Кения, Маврикий, Марокко, Намибия, Объединенная Республика Танзания, Сенегал и Тунис) Цель Сравнение определения ожирения, принятого Всемирной Результаты Распространенность избытка жировой ткани более организацией здравоохранения (ВОЗ) на основании значения чем в три раза превышала частоту ожирения, определяемую «индекс массы тела (ИМТ)-возраст», с измеренными величинами показателем «ИМТ-возраст»: 29,1% (95%-й ДИ: 26,8–31,4; упитанности африканских детей. 441 ребенок) в сравнении с 8,8% (95%-й ДИ: 7,5–10,4; 134 ребенка). Методы В перспективном многоцентровом исследовании в Чувствительность z-оценки ИМТ >+ 2,00 СО была низкой (29,7%, период с 2013 по 2017 год приняли участие 1516 участников в 95%-й ДИ: 25,5–34,2), а специфичность — высокой (99,7%, 95%- возрасте от 8 до 11 лет из городских районов восьми стран (Гана, й ДИ: 99,2–99,9). Анализ характеристических особенностей Кения, Маврикий, Марокко, Намибия, Объединенная Республика правильности обнаружения сигналов позволил обнаружить, что Танзания, Сенегал и Тунис). Детей взвешивали и измеряли их оптимизация чувствительности возможна при использовании рост, после чего вычисляли показатель «ИМТ-возраст» согласно z-оценки ИМТ + 0,58 СО и что при этом пороговом значении стандартам ВОЗ. Также измерялось содержание жировой площадь под кривой составляла 0,86, чувствительность — ткани в организме методом дейтериевого разбавления; 71,9% (95%-й ДИ: 67,4–76,0), а специфичность — 91,1% (95%-й содержание жировой ткани считалось избыточным, если ДИ: 89,2–92,7). оно превышало 25% у мальчиков и 30% у девочек. Авторы Вывод Несмотря на то что измерение ИМТ остается практическим вычислили чувствительность и специфичность z-оценки средством выявления ожирения, оно недооценивает содержание ИМТ >+ 2,00 стандартного отклонения (СО) и воспользовались избыточной жировой ткани в организме, и это следует учитывать методом анализа характеристических показателей правильности при планировании мероприятий по борьбе с пандемией детского обнаружения сигналов индексом Юдена для определения ожирения в Африке. оптимального порога z-оценки по ИМТ в вопросе классификации наличия избытка жировой ткани. Resumen Índice de masa corporal en comparación con el método de dilución de deuterio para establecer la prevalencia de la obesidad infantil, Ghana, Kenya, Marruecos, Mauricio, Namibia, República Unida de Tanzania, Senegal y Túnez Objetivo Comparar la definición de obesidad por edad del índice de Resultados La prevalencia de la obesidad excesiva fue más de tres masa corporal (IMC) de la Organización Mundial de la Salud (OMS) con veces superior a la obesidad definida por el IMC por edad: 29,1 % la grasa corporal medida en niños africanos. (IC del 95 %: 26,8 a 31,4; 441 niños) en comparación con un 8,8 % Métodos En un estudio prospectivo multicéntrico realizado entre 2013 (IC del 95 %: 7,5 a 10,4; 134 niños). La sensibilidad del IMC con DE de y 2017, se reclutaron 1516 participantes de edades comprendidas entre los valores Z > +2,00 fue baja (29,7 %, IC del 95 %: 25,5 a 34,2) y la los 8 y los 11 años de zonas urbanas de ocho países (Ghana, Kenya, especificidad fue alta (99,7 %, IC del 95 %: 99,2 a 99,9). El análisis de Marruecos, Mauricio, Namibia, República Unida de Tanzania, Senegal las características operativas del receptor encontró que un IMC z-score y Túnez). Se midieron la altura y el peso y calculamos el IMC por edad +0,58 DE optimizaría la sensibilidad, y en este corte el área bajo la curva utilizando los estándares de la OMS. Se midió la grasa corporal mediante era de 0,86, con una sensibilidad del 71,9 % (IC del 95 %: 67,4 a 76,0) y el método de dilución de deuterio y se definió el porcentaje de grasa una especificidad del 91,1 % (IC del 95 %: 89,2 a 92,7). corporal excesiva como > 25 % en los niños y > 30 % en las niñas. Se Conclusión Aunque el IMC sigue siendo una herramienta práctica para calculó la sensibilidad y especificidad del IMC con desviaciones estándar la monitorización de la obesidad, subestima el exceso de grasa y esto (DE) de los valores Z de > +2,00 y se utilizó el análisis de las características debería tenerse en cuenta a la hora de planificar las futuras respuestas operativas del receptor y el índice Youden para determinar el límite africanas a la pandemia de obesidad infantil. óptimo del IMC z-score para clasificar el exceso de grasa. References 1. Report of the Commission on Ending Childhood Obesity. Geneva: World 3. Muthuri SK, Wachira LJM, Leblanc AG, Francis CE, Sampson M, Onywera Health Organization; 2016. VO, et al. Temporal trends and correlates of physical activity, sedentary 2. Lobstein T, Jackson-Leach R. Planning for the worst: estimates of obesity behaviour, and physical fitness among school-aged children in sub-Saharan and comorbidities in school-age children in 2025. Pediatr Obes. 2016 Africa: a systematic review. Int J Environ Res Public Health. 2014 03 10;11(5):321–5. doi: http://dx.doi.org/10.1111/ijpo.12185 PMID: 27684716 20;11(3):3327–59. doi: http://dx.doi.org/10.3390/ijerph110303327 PMID: 780 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 Research Measuring childhood obesity in Africa Adama Diouf et al. 16. Introduction to body composition assessment using the deuterium dilution 4. Craig E, Reilly JJ, Bland R. Risk factors for overweight and overfatness in rural South African children and adolescents. J Public Health (Oxf ). 2016 method with analysis of saliva samples by Fourier transform infrared spectrometry. IAEA Human Health Series No. 12. Vienna: International Mar;38(1):24–33. doi: http://dx.doi.org/10.1093/pubmed/fdv016 PMID: 25742718 Atomic Energy Agency; 2010. Available from: https://www-pub.iaea.org/ books/iaeabooks/8369/Introduction-to-Body-Composition-Assessment- 5. The WHO child growth standards. Body mass index-for-age [internet]. Geneva: World Health Organization; 2007. Available from: http://www.who. Using-the-Deuterium-Dilution-Technique-with-Analysis-of-Saliva-Samples- by-Fourier-Transform-Infrared-Spectrometry [cited 2018 Aug 21]. int/childgrowth/standards/en/ [cited 2018 Aug 21]. 6. Okorodudu DO, Jumean MF, Montori VM, Romero-Corral A, Somers VK, 17. Lohman TG. Advances in body composition assessment. Champaign: Human Kinetics; 1992. Erwin PJ, et al. Diagnostic performance of body mass index to identify obesity as defined by body adiposity: a systematic review and meta- 18. Williams DP, Going SB, Lohman TG, Harsha DW, Srinivasan SR, Webber LS, et al. Body fatness and risk for elevated blood pressure, total cholesterol, and analysis. Int J Obes. 2010 May;34(5):791–9. doi: http://dx.doi.org/10.1038/ ijo.2010.5 PMID: 20125098 serum lipoprotein ratios in children and adolescents. Am J Public Health. 1992 Mar;82(3):358–63. doi: http://dx.doi.org/10.2105/AJPH.82.3.358 PMID: 7. Reilly JJ, Wilson ML, Summerbell CD, Wilson DC. Obesity: diagnosis, prevention, and treatment; evidence based answers to common questions. 1536350 19. Reilly JJ. Health effects of overweight and obesity in 195 countries. N Engl J Arch Dis Child. 2002 Jun;86(6):392–4. doi: http://dx.doi.org/10.1136/ adc.86.6.392 PMID: 12023163 Med. 2017 10 12;377(15):1496. PMID: 29020585 20. Reilly JJ, El-Hamdouchi A, Diouf A, Monyeki A, Somda SA. Determining the 8. Reilly JJ, Methven E, McDowell ZC, Hacking B, Alexander D, Stewart L, et al. Health consequences of obesity. Arch Dis Child. 2003 Sep;88(9):748–52. doi: worldwide prevalence of obesity. Lancet. 2018 May 5;391(10132):1773–4. doi: http://dx.doi.org/10.1016/S0140-6736(18)30794-3 PMID: 29739565 http://dx.doi.org/10.1136/adc.88.9.748 PMID: 12937090 9. Javed A, Jumean M, Murad MH, Okorodudu D, Kumar S, Somers VK, et al. 21. Wells JCK, Cole TJ, ALSPAC Study Team. Adjustment of fat free mass and fat mass for height in children aged 8 years. Int J Obes. 2002;26(7):947–52. doi: Diagnostic performance of body mass index to identify obesity as defined by body adiposity in children and adolescents: a systematic review and http://dx.doi.org/10.1038/sj.ijo.0802027 22. Wells JCK, Williams JE, Chomtho S, Darch T, Grijalva-Eternod C, Kennedy K, meta-analysis. Pediatr Obes. 2015 Jun;10(3):234–44. doi: http://dx.doi. org/10.1111/ijpo.242 PMID: 24961794 et al. Body-composition reference data for simple and reference methods and a 4-component model: a new UK reference child. Am J Clin Nutr. 2012 10. Reilly JJ, Kelly J, Wilson DC. Accuracy of simple clinical and epidemiological definitions of childhood obesity: systematic review and evidence appraisal. Dec;96(6):1316–26. doi: http://dx.doi.org/10.3945/ajcn.112.036970 PMID: Obes Rev. 2010 Sep;11(9):645–55. doi: http://dx.doi.org/10.1111/j.1467- 789X.2009.00709.x PMID: 20059704 23. Ruxton CH, Reilly JJ, Kirk T. Body composition of healthy 7–8 year old children and a comparison with the ‘reference child’. Int J Obes. 11. Hills A, Davidsson L. Stable isotope methods to develop and monitor nutrition interventions. Curr Nutr Food Sci. 2010;6:289–93. doi: http:// 1999;23(12):1276–81. doi: http://dx.doi.org/10.1038/sj.ijo.0801067 24. Wells JCK, Coward WA, Cole TJ, Davies PSW. The contribution of fat and fat- dx.doi.org/10.2174/157340110791233238 12. Wells JC, Fewtrell MS. Measuring body composition. Arch Dis Child. 2006 free tissue to body mass index in contemporary children and the reference child. Int J Obes Relat Metab Disord. 2002 Oct;26(10):1323–8. doi: http:// Jul;91(7):612–7. doi: http://dx.doi.org/10.1136/adc.2005.085522 PMID: 16790722 dx.doi.org/10.1038/sj.ijo.0802077 PMID: 12355328 25. Tremblay MS, Shields M, Laviolette M, Craig CL, Janssen I, Connor Gorber 13. Hruschka DJ, Hadley C. How much do universal anthropometric standards bias the global monitoring of obesity and undernutrition? Obes Rev. 2016 S. Fitness of Canadian children and youth: results from the 2007-2009 11;17(11):1030–9. doi: http://dx.doi.org/10.1111/obr.12449 PMID: 27383689 Canadian Health Measures Survey. Health Rep. 2010 Mar;21(1):7–20. PMID: 14. Liu A, Byrne NM, Kagawa M, Ma G, Poh BK, Ismail MN, et al. Ethnic 20426223 differences in the relationship between body mass index and percentage body fat among Asian children from different backgrounds. Br J Nutr. 2011 Nov;106(9):1390–7. doi: http://dx.doi.org/10.1017/S0007114511001681 PMID: 21736824 15. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig L, et al.; STARD Group. STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies. BMJ. 2015 10 28;351:h5527. doi: http://dx.doi. org/10.1136/bmj.h5527 PMID: 26511519 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 781 Research Adama Diouf et al. Measuring childhood obesity in Africa 782 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Bulletin of the World Health Organization Pubmed Central

Body mass index vs deuterium dilution method for establishing childhood obesity prevalence, Ghana, Kenya, Mauritius, Morocco, Namibia, Senegal, Tunisia and United Republic of Tanzania

Download PDF
11 pages

Loading next page...
 
/lp/pubmed-central/body-mass-index-vs-deuterium-dilution-method-for-establishing-yInxOrzSkq

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Pubmed Central
Copyright
(c) 2018 The authors; licensee World Health Organization.
ISSN
0042-9686
eISSN
1564-0604
DOI
10.2471/BLT.17.205948
Publisher site
See Article on Publisher Site

Abstract

Research Body mass index vs deuterium dilution method for establishing childhood obesity prevalence, Ghana, Kenya, Mauritius, Morocco, Namibia, Senegal, Tunisia and United Republic of Tanzania a b c d e Adama Diouf, Theodosia Adom, Abdel Aouidet, Asmaa El Hamdouchi, Noorjehan I Joonas, f g h f i Cornelia U Loechl, Germana H Leyna, Dorcus Mbithe, Thabisile Moleah, Andries Monyeki, j k l Hilde Liisa Nashandi, Serge MA Somda & John J Reilly Objective To compare the World Health Organization ( WHO) body mass index (BMI)-for-age definition of obesity against measured body fatness in African children. Methods In a prospective multicentre study over 2013 to 2017, we recruited 1516 participants aged 8 to 11 years old from urban areas of eight countries (Ghana, Kenya, Mauritius, Morocco, Namibia, Senegal, Tunisia and United Republic of Tanzania). We measured height and weight and calculated BMI-for-age using WHO standards. We measured body fatness using the deuterium dilution method and defined excessive body fat percentage as > 25% in boys and > 30% in girls. We calculated the sensitivity and specificity of BMI z-score > +2.00 standard deviations (SD) and used receiver operating characteristic analysis and the Youden index to determine the optimal BMI z-score cut-off for classifying excessive fatness. Findings The prevalence of excessive fatness was over three times higher than BMI-for-age-defined obesity: 29.1% (95% CI: 26.8 to 31.4; 441 children) versus 8.8% (95% CI: 7.5 to 10.4; 134 children). The sensitivity of BMI z-score > +2.00 SD was low (29.7%, 95% CI: 25.5 to 34.2) and specificity was high (99.7%, 95% CI: 99.2 to 99.9). The receiver operating characteristic analysis found that a BMI z-score +0.58 SD would optimize sensitivity, and at this cut-off the area under the curve was 0.86, sensitivity 71.9% (95% CI: 67.4 to 76.0) and specificity 91.1% (95% CI: 89.2 to 92.7). Conclusion While BMI remains a practical tool for obesity surveillance, it underestimates excessive fatness and this should be considered when planning future African responses to the childhood obesity pandemic. tifies children with the highest body fatness and the highest Introduction risk of co-morbidities. However, the indicator is conservative Childhood obesity is now a pandemic, heralding a substantial as it fails to identify children who are excessively fat, but who 1,2 7–10 burden of future noncommunicable diseases, despite the do not have high BMI-for-age. There are several problems established burden of underweight in low- and middle-income with this evidence. First, few studies tested the diagnostic countries. Changes in diet and reduced physical activity among performance of the WHO BMI-for-age definition of obesity, adolescent boys and girls have occurred across Africa, cur- focusing on definitions based on national BMI reference 9,10 rently most evident in urban areas, although rural areas are data or the International Obesity Task Force definition. also ae ff cted. The World Health Organization (WHO) Report Second, few studies assessed the diagnostic performance of of the Commission on Ending Childhood Obesity advocated BMI-for-age against a measure of body fatness with low bias more surveillance of the prevalence of obesity to plan where and acceptable individual diagnostic accuracy such as total 11,12 and when to intervene, and to measure the effectiveness of body water. Finally, the applicability of the evidence to future interventions. African children is unclear; bias in the estimation of excessive Body mass index (BMI)-for-age is a well-established in- body fatness by BMI varies across populations in adults. The dicator for surveillance of paediatric obesity. The WHO child extent to which such bias is population-specific for children growth standards define obesity in school-aged children as too is less clear, although compared with Europeans, South- BMI z-score > +2.00 standard deviations (SD). Systematic East Asian children have higher body fatness than would be 6 14 reviews have shown that, as in adults, high BMI-for-age iden- expected from their BMI. Laboratoire de Nutrition, Département de Biologie Animale, Faculté des Sciences et Techniques, Université Cheikh Anta Diop de Dakar, BP 5005 Dakar Fann, Senegal. Nutrition Research Centre, Ghana Atomic Energy Commission, Accra, Ghana. Association Tunisienne des Sciences de la Nutrition, Tunis, Tunisia. Unité Mixte de Recherche Nutrition et Alimentation CNESTEN-Université Ibn Tofail, Rabat, Morocco. Biochemistry Department; Victoria Hospital; Ministry of Health and Quality of Life, Quatre Bornes, Mauritius. International Atomic Energy Agency, Vienna International Centre, Vienna, Austria. Department of Epidemiology and Biostatistics, School of Public Health and Social Sciences, Dar el Salaam, United Republic of Tanzania. Department of Food, Nutrition and Dietetics, Kenyatta University, Nairobi, Kenya. Physical Activity, Sport and Recreation, Faculty of Health Sciences, North-West University, Potchefstroom, South Africa. Ministry of Health and Social Services, Windhoek, Namibia. Centre Muraz, Bobo-Dioulasso, Burkina Faso. School of Psychological Science and Health, University of Strathclyde, Glasgow, Scotland. Correspondence to Adama Diouf (email: [email protected]). (Submitted: 15 November 2017 – Revised version received: 25 July 2018 – Accepted: 10 August 2018 – Published online: 10 September 2018 ) 772 Bull World Health Organ 2018;96:772–781 | doi: http://dx.doi.org/10.2471/BLT.17.205948 Research Measuring childhood obesity in Africa Adama Diouf et al. The aim of this study was to com- South Africa. The height of children was are laboratory-based and impractical pare the prevalence of the WHO BMI- measured to the nearest 0.1 cm using a for large-scale epidemiological studies. for-age definition of obesity against the Seca stadiometer, and weight to 0.1 kg While not a criterion method, body prevalence of excessive body fatness in light indoor clothing using a Seca fatness measured by total body water in a relatively large sample of African scale (Seca, London, England). From is practical for large epidemiological children. the height and weight measures, we studies and provides accurate measures calculated BMI for each child as weight of fatness which are unbiased relative to 2 11,12 divided by height squared (kg/m ) and multicomponent methods. Methods then computed age- and sex-specific The total body water measures were Study design z-score relative to the WHO BMI-for- made on the same day as the height, age reference using the Stata zanthro weight and waist circumference mea- This design for this prospective, multi- package (Stata Corp., College Station, sures. Accurate measurement of total centre, data-pooling study, was agreed United States of America). We defined body water requires a normal hydration at the first meeting of the Reducing obesity as BMI z-score > +2.00 SD and status. We therefore asked participants Obesity Using Nuclear Techniques To overweight (including obesity) as BMI and their families to have normal fluid Design Interventions study in 2012. We z-score > +1.00 SD and food intake on the day before the followed the Standards for Reporting estimation of total body water and to of Diagnostic Accuracy Studies for Body water measures avoid vigorous exercise after the final the conduct and reporting of the study. We aimed to measure total body water meal of the previous day to avoid dehy- Sampling and study procedures origi- in all participants using the deuterium dration and depletion of glycogen stores. nally took place across 11 African cen- dilution method, as described previ- Deuterium oxide-labelled water (99.8% tres between 2013 and 2017. We aimed 11,14 ously. We used standard operating purity; Cambridge Isotope Laboratories to recruit around 150 participants per procedures, with training support pro- Inc., Andover, USA) accurately weighed country (a larger sample was used in the vided for all countries via a combination (0.001 g precision) was orally adminis- United Republic of Tanzania, because of residential and on-site training by tered to the children according to their some of the study aims there required experts recruited by the International body weight (0.5 g deuterium oxide per a larger sample). As the nutrition and Atomic Energy Agency. Ideally, body kg) followed by 50 mL of local tap water. physical activity transitions in Africa fatness measurement methods are mul- Children were asked not to eat or drink have disproportionately affected urban ticomponent, based on measures of total for at least 30 minutes before receiving 3,4 children, we focused the sampling in body water plus body density or total the deuterium-labelled water and to void urban areas. In each country, we used a body mineral. However, such methods their bladders before dosing. Baseline multistage random sampling method to select at least four to five urban public schools in one district or state, followed Fig. 1. Flowchart on the inclusion of participants to compare methods to measure by school sampling frames of all classes overweight in children in eight African countries, 2013–2017 corresponding to the target age group and sex. More details of the methods are available from the corresponding author. 2172 eligible participants recruited and Children meeting the inclusion BMI and body composition measured criteria were recruited to participate in the study after submission of a signed Data excluded from 656 participants informed consent form by a parent. • data not collected or not analysed Data collection was conducted during (155 participants) the school year. Ethical approval was • data rejected by quality control obtained from local research boards or (501 participants) committees in each country. Participants were eligible for inclusion if they were age 8 to 11 years and provided consent Data analysed from 1516 participants or assent for participation; they were excluded if they were outside the study age range, had ill health that would have 134 participants classified as 1382 participants classified as precluded participation or were not obese by BMI non-obese by BMI present in school after two consecutive visits. Anthropometric measures Final classification Final classification • 310 obese by body fatness • 131 obese by body fatness The study used a common protocol and • 1072 non-obese by body fatness • 3 non-obese by body fatness standard operating procedures across all countries. Before data collection BMI: body mass index. started all researchers were trained in Notes: Children were originally recruited and assessed for BMI and body composition in 11 countries. data collection methods by a team of Data were rejected for quality control reasons from Benin, Mali and Uganda. The final analysis was experienced researchers and fieldwork - therefore based on data from eight countries: Ghana, Kenya, Mauritius, Morocco, Namibia, Senegal, Tunisia and United Republic of Tanzania. ers during a 1-week residential course in Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 773 Research Adama Diouf et al. Measuring childhood obesity in Africa (pre-dose) saliva samples were collected the labelled water dose under supervi- was carried out with Fourier transform from each participant by rotating a sion and two further saliva samples were infrared spectroscopy (FTIR 8400S cotton-wool ball in the buccal cavity of collected at 3 hours and 4 hours ae ft r the spectrophotometer, Shimadzu Kyoto, the mouth until well soaked. Saliva was dose using the method described above. Japan) in accordance with International collected into a clean sterile and dry tube All saliva samples were stored at 4 °C Atomic Energy Agency protocols. using a 20 mL disposable syringe. Par- until their arrival to the laboratory for We converted measures of total ticipants were then requested to drink storage at −20 °C until analysis. Analysis body water to total body fat using es- Table 1. Age and anthropometric characteristics of participants in the study of body mass index-for-age and body fatness among children in eight African countries, 2013–2017 Sex, by country No. (%) of Median (IQR) children Age, years BMI z-score Body fat percentage Fat mass index, Fat-free mass index, 2 2 kg/m kg/m Ghana Boys 71 (37.4) 10 (9–11) −0.87 (−1.17 to −0.21) 15.92 (12.57 to 19.94) 2.37 (1.98 to 3.24) 12.98 (12.08 to 13.75) Girls 119 (62.6) 10 (9–10) −0.66 (−1.16 to 0.09) 18.72 (15.65 to 22.62) 2.85 (2.33 to 3.74) 12.63 (11.77 to 13.38) Total 190 (100.0) 10 (9–11) −0.70 (−1.16 to −0.05) 18.03 (14.40 to 21.08) 2.65 (2.14 to 3.42) 12.77 (11.95 to 13.70) Kenya Boys 84 (46.9) 10 (9–11) −0.91 (−1.34 to −0.30) 22.94 (17.93 to 28.61) 3.58 (2.68 to 5.01) 12.29 (10.92 to 13.64) Girls 95 (53.1) 10 (9–11) −0.69 (−1.36 to −0.05) 24.14 (19.43 to 27.49) 3.71 (2.85 to 4.86) 12.47 (10.73 to 14.53) Total 179 (100.0) 10 (9–11) −0.82 (−1.35 to −0.15) 23.57 (19.34 to 28.11) 3.64 (2.75 to 4.86) 12.34 (10.84 to 13.90) Mauritius Boys 82 (53.6) 10 (9–11) 0.76 (−1.01 to 1.86) 25.28 (18.67 to 33.23) 4.12 (2.75 to 7.36) 13.14 (11.74 to 14.67) Girls 71 (46.4) 10 (9–11) 0.56 (−0.55 to 1.84) 32.11 (24.62 to 37.66) 5.67 (3.96 to 8.39) 12.64 (11.44 to 14.68) Total 153 (100.0) 10 (9–11) 0.68 (−0.76 to 1.84) 28.80 (21.65 to 35.48) 4.96 (3.41 to 7.71) 13.01 (11.64 to 14.68) Morocco Boys 94 (50.3) 9 (8–10) −0.24 (−1.00 to 0.51) 19.76 (16.31 to 24.57) 3.08 (2.49 to 4.17) 12.76 (12.15 to 13.66) Girls 93 (49.7) 9 (8–10) −0.33 (−0.99 to 0.42) 25.69 (21.91 to 30.11) 4.07 (3.20 to 5.07) 11.97 (11.06 to 12.62) Total 187 (100.0) 9 (8–10) −0.27 (−0.99 to 0.51) 23.23 (18.30 to 28.60) 3.70 (2.76 to 4.72) 12.36 (11.63 to 13.27) Namibia Boys 66 (43.7) 10 (9–11) −0.08 (−0.91 to 1.09) 22.84 (18.85 to 30.76) 3.60 (2.70 to 5.73) 12.92 (12.06 to 13.69) Girls 85 (56.3) 10 (9–11) 0.42 (−0.76 to 1.64) 32.69 (26.76 to 39.06) 5.38 (4.14 to 8.71) 11.94 (11.01 to 13.18) Total 151 (100.0) 10 (9–11) 0.19 (−0.84 to 1.44) 27.97 (22.32 to 37.50) 4.70 (3.38 to 7.42) 12.59 (11.44 to 13.41) Senegal Boys 70 (47.9) 10 (9–11) −1.29 (−1.84 to −0.71) 13.43 (10.64 to 19.99) 1.95 (1.49 to 3.07) 12.50 (11.72 to 13.10) Girls 76 (52.1) 10 (9–10) −1.40 (−2.15 to −0.58) 19.30 (15.80 to 24.91) 2.62 (2.13 to 3.52) 11.41 (10.71 to 12.04) Total 146 (100.0) 10 (9–10) −1.32 (−2.05 to −0.60) 16.70 (12.76 to 22.61) 2.35 (1.79 to 3.34) 11.84 (11.12 to 12.66) Tunisia Boys 80 (51.0) 9 (9–10) 0.04 (−0.65 to 0.99) 23.49 (20.30 to 26.86) 3.86 (3.19 to 5.00) 12.56 (11.89 to 13.75) Girls 77 (49.0) 10 (8–10) 0.31 (−0.65 to 1.18) 30.03 (25.57 to 33.89) 4.89 (3.99 to 6.34) 11.94 (11.13 to 12.82) Total 157 (100.0) 9 (8–10) 0.10 (−0.65 to 1.12) 26.03 (22.88 to 31.37) 4.29 (3.54 to 5.77) 12.37 (11.60 to 13.29) United Republic of Tanzania Boys 158 (44.8) 10 (9–11) 0.04 (−0.58 to 1.12) 18.50 (15.10 to 24.90) 3.00 (2.41 to 4.83) 13.42 (12.73 to 14.34) Girls 195 (55.2) 10 (9–11) 0.02 (−0.80 to 0.92) 23.30 (19.30 to 31.10) 3.73 (2.93 to 5.62) 12.70 (11.89 to 13.63) Total 353 (100.0) 10 (9–11) 0.02 (−0.67 to 0.95) 21.50 (17.00 to 29.40) 3.43 (2.60 to 5.37) 13.00 (12.27 to 14.14) Total Boys 705 (46.5) 10 (9–11) −0.37 (−1.09 to 0.69) 20.47 (15.60 to 26.09) 3.25 (2.40 to 4.57) 12.92 (12.06 to 13.90) Girls 811 (53.5) 10 (9–10) −0.33 (−1.09 to 0.72) 24.90 (19.37 to 31.46) 3.91 (2.87 to 5.64) 12.23 (11.31 to 13.31) Total 1516 10 (9–11) −0.35 (−1.09 to 0.71) 22.65 (17.43 to 29.60) 3.59 (2.60 to 5.17) 12.59 (11.64 to 13.63) (100.0) BMI: body mass index; IQR: interquartile range. Number of records removed from original samples: Ghana (4), Kenya (1), Mauritius (3), Morocco (3), Namibia (4), Senegal (10), Tunisia (2) and United Republic of Tanzania (3). 2 5 Notes: BMI was calculated as weight in kg divided by height in m and z-scores were obtained from the World Health Organization BMI-for-age child growth standards. Body fat percentage was measured using deuterium oxide dilution. Fat mass index was calculated as fat mass in kg divided by height in m , with fat mass measured from total body water. Fat free mass index was calculated as fat free mass in kg divided by height in m , with fat-free mass measured from total body water. Measures were made in 2014–2017 in Kenya and United Republic of Tanzania and in 2013–2015 in all other countries. 774 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 Research Measuring childhood obesity in Africa Adama Diouf et al. tablished age- and sex-specific constants not greatly affected by the definition of sensitivity and specificity for identify- 16,17 for the hydration of fat-free mass, as excessive fatness (data are available from ing excessive fatness. We used Spear- described elsewhere. Quality control the corresponding author). man rank-order correlation to test the procedures, with four stringent criteria association between countries’ total Data management described in detail elsewhere, were BMI-for-age z-score and total body fat applied to the measures of enrichment Training in data management, data percentage. We also made an explor- of deuterium required for the total sharing and data quality control was atory analysis of possible geographical body water measures and to the esti- provided during a 1-week data man- differences in the results by grouping mates of total body water, total body agement residential training course in the countries into three geographically fat and body fat percentage. These Benin in 2014. Throughout the study, defined sub-groups: sub-Saharan Africa quality control measures were: (i) deu- support was provided by site visits and (Ghana, Kenya, Namibia, Senegal and terium enrichment of each of the two online chat or email by a central data United Republic of Tanzania), North post-dose samples should be within management coordinator from Burkina Africa (Morocco and Tunisia) and an 2% of the mean of the two post-dose Faso recruited by the International African island (Mauritius). samples; (ii) measured enrichment Atomic Energy Agency. The weight and should lie within an expected range of height measures, BMI-for-age z-scores Results normal enrichments based on the body and total body water-derived measures weight of the child (outliers in the total of body fatness were all made prospec- Fig. 1 shows the flowchart of the study. body-water-to-height relationship were tively and independently and the results Of the 2172 children recruited to the identified and excluded); (iii) outliers in of each measure were not available at the study, eligible data were available from body fat percentage were identified and time of the other measures. 1516 (69.8%). The age and anthropo - excluded (e.g. large mismatches between metric characteristics of the eligible Analysis body fat percentage and BMI z-score participants are shown in Table 1. The or unphysiological body fat percentage We used standard diagnostic perfor- mean age was 9.6 years (95% confidence measures); and (iv) if more than 10% mance indicators to determine the interval, CI: 9.5 to 9.7) and median of total body water measures from any extent to which BMI z-score > +2.00 age was 10 years (interquartile range, centre failed to meet the quality control SD identified children with excessive IQR: 9 to 11). The median BMI-for-age criteria, then we excluded all data from fatness. We calculated sensitivity (pro- z-score was −0.35 (IQR: −1.09 to 0.71) that centre from the pooled analyses. portion of real positive values among all and median body fat percentage was Based on these criteria, we excluded data the recorded positive values), specific- 22.65% (IQR: 17.43 to 29.60). Fig. 2 from three out of 11 original participat- ity (proportion of real negative values provides more detail on the distribu- ing countries (Benin, Mali and Uganda), among all the negative values), and tion of body fatness and BMI-for-age so that the present study is based on positive and negative predictive values z-scores. The prevalence of excessive data from eight countries (Ghana, for the total sample. We used the Youden fatness was 29.1% (95% CI: 26.8 to 31.4; Kenya, Mauritius, Morocco, Namibia, index method to determine the optimal 441 children). Overall, the prevalence of Senegal, Tunisia and United Republic BMI z-score cut-off for optimizing the obesity by the WHO BMI-for-age crite- of Tanzania) and 1516 children. Among these, 2% of total body water measures Fig. 2. Relationships between body mass index-for-age z-score and body fat percentage were rejected for quality control reasons among children in eight African countries, by geographical area, 2013–2017 and were not included in the analyses reported here. We expressed total body fatness 1.00 as a percentage of body weight. Many studies have established that a high body fatness, even in childhood, has a 0.75 range of adverse health consequences, with most focusing on the cardiometa- Area under ROC curve: 0.86 bolic consequences, as summarized by 7,8 systematic reviews. One report on the 0.50 relationship between body fatness and cardio-metabolic risk in childhood used a skinfold thickness method previously validated against a multicomponent 0.25 model to measure body fatness. The researchers found a marked increase in cardiometabolic risk profile at body fat > 25% in boys and > 30% in girls, across 0.25 0.50 0.75 1.00 a wide age range. We therefore used this 1–Specificity, % definition of excessive fatness (true posi- tive in the receiver operator characteris- BMI: body mass index. tic analysis) in the present study. As in Note: Areas were defined as follows: sub-Saharan Africa (Ghana, Kenya, Namibia, Senegal and United 9,10 previous studies, the conclusions were Republic of Tanzania), North Africa (Morocco and Tunisia) and African island (Mauritius). Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 775 Sensitivity, % Research Adama Diouf et al. Measuring childhood obesity in Africa rion was 8.8% (95% CI: 7.5 to 10.4; 134 Table 2. Comparison of obesity defined by body mass index-for-age and by body fatness children) (Table 2) and of overweight among children in eight African countries, by geographical area, 2013–2017 was 19.5% (95% CI: 17.6 to 21.6; 296 children) (Table 3). Obesity defined by BMI- Obesity defined by body fatness, no. (%) of children In the whole sample, the sensitivity for-age No Yes Total of BMI z-score > +2.00 SD for identify- ing excessively fat children was 29.7% Sub-Saharan Africa (95% CI: 25.5 to 34.2), specificity was Ghana 99.7% (95% CI: 99.2 to 99.9), positive No 183 (100.0) 5 (71.4) 188 (99.0) predictive value 97.8% (95% CI: 93.6 Yes 0 (0.0) 2 (28.6) 2 (1.0) to 99.5) and negative predictive value Total 183 (100.0) 7 (100.0) 190 (100.0) 77.6% (95% CI: 75.3 to 79.7). The sen- Kenya sitivity of BMI z-score > +2.00 SD to No 125 (100.0) 50 (92.6) 175 (97.8) identify excessively fat children varied Yes 0 (0.0) 4 (7.4) 4 (2.2) little between boys and girls (66/203 Total 125 (100.0) 54 (100.0) 179 (100.0) for boys versus 65/238 for girls). In the Namibia whole sample BMI z-score > 1.00 SD No 75 (100.0) 49 (64.5) 124 (82.1) had sensitivity of 59.2% (95% CI: 54.4 Yes 0 (0.0) 27 (35.5) 27 (17.9) to 63.8) and specificity of 96.7% (95% Total 75 (100.0) 76 (100.0) 151 (100.0) CI: 95.5 to 97.7; Table 4). Analysis of the data by country and in the three popu- Senegal lation sub-groups is shown in Table 2 No 130 (100.0) 12 (75.0) 142 (97.3) and Table 3. Sensitivity was lower in the Yes 0 (0.0) 4 (25.0) 4 (2.7) North African and Island populations Total 130 (100.0) 16 (100.0) 146 (100.0) than the sub-Saharan Africans. The rank United Republic of Tanzania order correlation between country me- No 259 (98.9) 48 (52.7) 307 (87.0) dian BMI z-score and country fat mass Yes 3 (1.1) 43 (47.3) 46 (13.0) index was high (r = 0.6). Total 262 (100.0) 91 (100.0) 353 (100.0) The receiver operator characteristic All analysis is shown in Fig. 3. The optimal No 772 (99.6) 164 (67.2) 936 (91.9) cut-off point in the BMI-for-age distri- Yes 3 (0.4) 80 (32.8) 83 (8.1) bution for classifying excessive fatness Total 775 (100.0) 244 (100.0) 1019 (100.0) was a BMI z-score of +0.58 SD (Table 4). North Africa At this cut-off the area under the curve Morocco was 0.86, sensitivity was 71.9% (95% No 141 (100.0) 37 (80.4) 178 (95.2) CI: 67.4 to 76.0), specificity 91.1% (95% Yes 0 (0.0) 9 (19.6) 9 (4.8) CI: 89.2 to 92.7), positive predictive value 76.8% (95% CI: 72.4 to 80.7) and Total 141 (100.0) 46 (100.0) 187 (100.0) negative predictive value 88.8% (95% Tunisia CI: 86.7 to 90.6). No 89 (100.0) 60 (88.2) 149 (94.9) Yes 0 (0.0) 8 (11.8) 8 (5.1) Total 89 (100.0) 68 (100.0) 157 (100.0) Discussion All The present study has established the No 230 (100.0) 97 (85.1) 327 (95.1) extent to which the WHO BMI-for-age Yes 0 (0.0) 17 (14.9) 17 (4.9) definition of obesity underestimates Total 230 (100.0) 114 (100.0) 344 (100.0) the prevalence of excessive fatness in African island African children. Excessive fatness was Mauritius present in nearly a third of children, No 70 (100.0) 49 (59.0) 119 (77.8) suggesting that urban African environ- Yes 0 (0.0) 34 (41.0) 34 (22.2) ments are now highly obesogenic even Total 70 (100.0) 83 (100.0) 153 (100.0) for children. Excessive fatness was over All countries three times more common than the No 1072 (99.7) 310 (70.3) 1382 (91.2) prevalence of BMI-defined obesity. This difference is large enough to be mean- Yes 3 (0.3) 131 (29.7) 134 (8.8) ingful for public health. For example, Total 1075 (100.0) 441 (100.0) 1516 (100.0) the case for policy action to prevent and BMI: body mass index. control obesity is much weaker at an ap- We measured height and weight and calculated obesity from BMI-for-age using the World Health Organization reference z-score > +2.00 standard deviations. parent prevalence of around 8% (based We measured body fatness using the deuterium dilution method and defined excessive body fat on BMI-for-age z-score > +2.00 SD in 11,12 percentage as > 25% in boys and > 30% in girls. the present study) than at the preva- lence of around 30% (excessive fatness) 776 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 Research Measuring childhood obesity in Africa Adama Diouf et al. Table 3. Comparison of overweight defined by body mass index-for-age and obesity defined by body fatness among children in eight African countries, by geographical area, 2013–2017 Overweight defined by Obesity defined by body fatness, no. (%) of children BMI-for-age No Yes Total Sub-Saharan Africa Ghana No 182 (99.5) 2 (28.6) 184 (96.8) Yes 1 (0.5) 5 (71.4) 6 (3.2) Total 183 (100.0) 7 (100.0) 190 (100.0) Kenya No 122 (97.6) 48 (88.9) 170 (95.0) Yes 3 (2.4) 6 (11.1) 9 (5.0) Total 125 (100.0) 54 (100.0) 179 (100.0) Namibia No 75 (100.0) 26 (34.2) 101 (66.9) Yes 0 (0.0) 50 (65.8) 50 (33.1) Total 75 (100.0) 76 (100.0) 151 (100.0) Senegal No 130 (100.0) 7 (43.8) 137 (93.8) Yes 0 (0.0) 9 (56.2) 9 (9.2) Total 130 (100.0) 16 (100.0) 146 (100.0) United Republic of Tanzania No 246 (93.9) 21 (23.1) 267 (75.6) Yes 16 (6.1) 70 (76.9) 86 (24.4) Total 262 (100.0) 91 (100.0) 353 (100.0) All No 755 (97.4) 104 (42.6) 859 (84.3) Yes 20 (2.6) 140 (57.4) 160 (15.7) Total 775 (100.0) 244 (100.0) 1019 (100.0) North Africa Morocco No 137 (97.2) 21 (45.7) 158 (84.5) Yes 4 (2.8) 25 (54.3) 29 (15.5) Total 141 (100.0) 46 (100.0) 187 (100.0) Tunisia No 84 (94.4) 29 (42.6) 113 (72.0) Yes 5 (5.6) 39 (57.4) 44 (28.0) Total 89 (100.0) 68 (100.0) 157 (100.0) All No 221 (96.1) 50 (43.9) 271 (78.8) Yes 9 (3.9) 64 (56.1) 73 (21.2) Total 230 (100.0) 114 (100.0) 344 (100.0) African island Mauritius No 64 (91.4) 26 (31.3) 90 (58.8) Yes 6 (8.6) 57 (68.7) 63 (41.2) Total 70 (100.0) 83 (100.0) 153 (100.0) All countries No 1040 (96.7) 180 (40.8) 1220 (80.5) Yes 35 (3.3) 261 (59.2) 296 (19.5) Total 1075 (100.0) 441 (100.0) 1516 (100.0) BMI: body mass index. We measured height and weight and calculated overweight from BMI-for-age using the World Health Organization reference z-score > +1.00 standard deviations. We measured body fatness using the deuterium dilution method and defined excessive body fat 11,12 percentage as > 25% in boys and > 30% in girls. Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 777 Table 4. Comparison of World Health Organization body mass index-for-age cut-offs for obesity and overweight and the empirically determined optimal cut-off for identifying excessive fatness among children in eight African countries, 2013–2017 a b Diagnostic performance measure BMI z-score > +2.00 SD BMI z-score > +1.00 SD BMI z-score +0.58 SD No. of children Total no. % (95% CI) No. of children Total no. % (95% CI) No. of children Total no. % (95% CI) Sensitivity 131 441 29.7 (25.5 to 34.2) 261 441 59.2 (54.4 to 63.8) 317 441 71.9 (67.4 to 76.0) Specificity 1072 1075 99.7 (99.2 to 99.9) 1040 1075 96.7 (95.5 to 97.7) 979 1075 91.1 (89.2 to 92.7) Positive predictive value 131 134 97.8 (93.6 to 99.5) 261 296 88.2 (83.9 to 91.6) 317 413 76.8 (72.4 to 80.7) Negative predictive value 1072 1382 77.6 (75.3 to 79.7) 1040 1220 85.2 (83.1 to 87.2) 979 1103 88.8 (86.7 to 90.6) BMI: body mass index; CI: confidence interval; SD: standard deviation. We calculated values as follows: sensitivity: [true positives/(true positives + false negatives)]; specificity: [true negatives/(true negatives + false positives)]; positive predictive value: [true positives/(true positives + false positives)]; negative predictive value: [true negatives/(true negatives + false negatives)]. We calculated the optimal cut-off z-score from the receiver operating characteristic curve (area under the curve: 0.86). Research Adama Diouf et al. Measuring childhood obesity in Africa observed. To improve the estimation of Fig. 3. Receiver operator characteristic analysis of the ability of body mass index-for- prevalence, cut-off points in the BMI age z-score to identify children with excessive fatness in eight African countries, distribution lower than the z-score of 2013–2017 +2.00 SD might be considered. At BMI z-score > +1.00 SD the ability to identify Sub-Saharan Africa North Africa over-fatness was improved but not opti- 60 60 mized. The optimal BMI z-score cut-off for classifying excessive fatness (which 40 40 maximized the area under the curve) in our study was +0.58 SD. 20 20 There are no directly comparable studies in African children, or using the 0 0 WHO-recommended definition based -4 -2 0+2+4-4-20 +2 +4 on BMI, but in non-African popula- BMI-for-age z-score BMI-for-age z-score tions biases have been reported for other 9,10 African island Total BMI-based definitions of obesity The 60 60 present study adds to previous stud- ies suggesting that underestimation 40 40 of excessive fatness by BMI-for-age criteria is likely to be a global cause for 19,20 20 20 concern. Our study shows that a high proportion of African children with 0 0 apparently healthy BMI-for-age have ex- -4 -2 0+2+4 -4 -2 0+2+4 cessive body fatness. The bias observed BMI-for-age z-score BMI-for-age z-score is unlikely to be due to a high body fat Boys Girls Fitted values percentage secondary to unusually low fat-free mass (lean body mass). This is because of the consistency between the Accuracy Studies guidance in both the or delay future obesity prevention and findings of the present study and studies conduct and reporting of the study; and control efforts in Africa. Further re- 9,10 for other populations. Furthermore, the standardization and quality control search is needed to determine whether median fat mass index values, which of both the study measurement methods the sensitivity of the BMI-for-age indica- measure fatness relatively independent and data management. A key limitation tor is especially low in African children of fat-free mass, were high in the pres- of the study was that we were unable to compared with other populations. ■ ent study. Reference data for fat mass test definitively for differences in the di- index from British children of the same agnostic accuracy of BMI-for-age across Acknowledgements age (and measured in 2001, long after different populations of African chil- We thank the participating centres and the childhood obesity epidemic had af- dren. Our exploratory comparison of countries, the Ministry of Higher Educa- fected children in the United Kingdom country groups by sub-Saharan Africa, tion and Scientific Research of Senegal of Great Britain and Northern Ireland) North Africa and an island population (PAPES) and Ministry of Education were very similar to those in the present was underpowered. A further limita- (Senegal), Centre National d’Energie des study: 50th centile of 3.4 kg/m for boys tion is generalizability. The participant Sciences et des methods Nucléaires (Mo- and 4.2 kg/m for girls compared with age range of the present study limits rocco), Ministry of Health and Quality 2 2 3.25 kg/m for boys and 3.91 kg/m for our conclusions to 8 to 11 year olds, of Life and the University of Fribourg, girls in the present study. Our findings although our findings are consistent Switzerland (Mauritius), Kenyatta Uni- are consistent with the evidence that with those reported for younger and versity, Nairobi, Kenya, Ecole Supérieure body fatness of contemporary children older participants, including adults, in des Sciences et Techniques de la Santé de is higher, across the range of body fat- systematic reviews of studies from non- Tunis, Université Tunis El Manar 6,9,10 ness, than that of children in the recent African populations. 23–25 past. In conclusion, excessive fatness is Funding: e Th study was partly funded by The main strengths of the present now prevalent among urban popula- the International Atomic Energy Agency, study were the large sample size and tions of African children and is likely (RAF/6/402). narrow age range of the sample; the to have serious future public health novelty of using the WHO BMI-for-age implications. While at a group level Competing interests: None declared. definition in an African setting; the the BMI z-score and body fatness were novelty and value of having an unbiased related, BMI-for-age substantially un- definition of body fatness; the use of the derestimated the scale of the problem Standards for Reporting of Diagnostic of excessive fatness and so may hinder 778 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 Percentage fat mass Percentage fat mass Research Measuring childhood obesity in Africa Adama Diouf et al. صخلم ،برغلماو ،لاغنسلا ،ةلوفطلا ةلحرم في ةنمسلا راشتنا ديدحتل ،مويرتويدلا فيفتخ ةقيرط لباقم مسلجا ةلتك شرؤم ايبيمانو ،سويشيرومو ،اينيكو ،اناغو ،ةدحتلما اينازنت ةيروهجمو ،سنوتو نم تارم ثلاث نم رثكأ لىعأ ةطرفلما ةنمسلا راشتنا ناك جئاتنلا فيرعتو ،ةقرافلأا لافطلأا في ةسيقلما ةنمسلا ينب ةنراقم ضرغلا 29.1٪ :نسلل ةبسنلاب مسلجا ةلتك شرؤم اهدديح يتلا ةنمسلا ،نسلل ةبسنلاب ( مسلجا BMI ةلتك ) شرؤم لىع مئاقلا ةنادبلا 8.8٪ لباقملافط ( 441 ؛31.4 لىإ 26.8 :95٪ ةقثلا لصاف) .ةيلماعلا ةحصلا ةمظنلم عباتلاو ةيساسح تناك. (لافط 134 ،10.4 لىإ 7.5 :95٪ ةقثلا لصاف ) 2013 نم ةترفلا للاخ زكارلما ةددعتم ةيقابتسا ةسارد في ةقيرطلا ةلصحملل ةبسنلاب نم +2.00 بركلأا - (SD) يرايعلما فارحنلاا لىإ 8 ينب مهرماعأ حواترت ا كراشم 1516 ليجستب انمق ، لىإ2017 :95٪ ةقث لصافب ،29.7٪ ) ،ةضفخنم -مسلجا ةلتك شرؤ ةيرايعلما لم ،برغلماو ،لاغنسلا نادلب ) نيماث في ةيضرلحا قطانلما نم ا ماع 11 لصافب ،99.7٪ ) ةيلاع ةيصوصلخا تناكو (34.2 لىإ 25.5 ،سويشيرومو ،اينيكو ،اناغو ،ةدحتلما اينازنت ةيروهجمو ،سنوتو ليغشتلا صئاصخ ليلتح فشتكا.(99.9 لىإ 99.2 :95٪ ةقث مسلجا ةلتك شرؤم باسحو نزولاو لوطلا سايقب (ايبيمانو انمق. ةلصحملل -+0.58 ةبسنب يرايعلما فارحنلاا نأ بقتسملل ل .ةيلماعلا ةحصلا ةمظنم يرياعم مادختساب ،نسلل ةبسنلاب (BMI) دنع هنأو ،ةيساسلحا نم نس يح ُ فوس -مسلجا ةلتك شرؤلم ةيرايعلما ،موييرتويدلا فيفتخ ةقيرط مادختساب مسلجا ةنمس سايقب انمق ةيساسلحا تناكو ،0.86 ىنحنلما تتح ةحاسلما تناك دلحا اذه نم رثكأ نوكتل مسلجا في ةدئازلا نوهدلل ةيوئلما ةبسنلا انددحو ةيصوصخو ،(76.0 لىإ 67.4 :95٪ ةقث لصاف 71.9٪ ) ةيساسح باسحب انمق.تانبل نم ا في رثكأو 30٪ ،دلاولأا في 25٪ .(92.7 لىإ 89.2 :95٪ ةقث لصاف) 91.1٪ نم بركأ نكتل مسلجا ةلتك شرؤلم ةيرايعلما ةلصحلما ةيصوصخو ،ةنمسلا ةبقارلم ةيلمع ةادأ مسلجا ةلتك شرؤم لظي مانيب جاتنتسلاا ليلحتب انعتساو ) ةيرايعلما ،(SD تافارحنلال ةبسنلاب +2.00 دنع رابتعلاا في كلذ عضو بيجو ،ةطرفلما ةنمسلا نم للقي هنأ لاإ لثملأا دلحا ديدحتلندوي شرؤمو لبقتسملل ليغشتلا صئاصخ ةلحرم في ةنادبلا ءابول ةيلبقتسم ةيقيرفأ تاباجتسلا ةنمسلا طيطختلا فينصت فدبه ،مسلجا ةلتك شرؤلم ةيرايعلما ةلصحملل .ةلوفطلا .ةطرفلما 摘要 加纳，肯尼亚，摩洛哥，毛里求斯，纳米比亚，塞内加尔，突尼斯和坦桑尼亚联合共和国各国通过身体质量指 数与重水同位素法确定儿童肥胖患病率 目的 将世卫组织 (WHO) 年龄别身体质量指数 (BMI- 结果 过度肥胖的患病率比年龄别身体质量指数所定义 for-age) 对肥胖的定义与所测量的非洲儿童体脂率进行 的肥胖高出三倍以上 ： 29.1％（95％ 置信区间，CI ： 对比。 26.8 至 31.4 ； 441 名儿童）与 8.8％（95％ 置信区间， 方法 在 2013 年至 2017 年的一项前瞻性、多中心的研 CI ： 7.5 至 10.4 ； 134 名儿童） 。BMI z 评分 >+2.00 标 究中，我们从 8 个国家的城市地区（加纳，肯尼亚， 准差 (SD) 的敏感性低（29.7％，95％ 置信区间，CI ： 摩洛哥，毛里求斯，纳米比亚，塞内加尔，突尼斯和 25.5 至 34.2） ，特异性高（99.7 ％，95 ％ 置信区间， 坦桑尼亚联合共和国）招募了 1516 名 8 至 11 岁的参 CI ： 99.2 至 99.9） 。受试者工作特征分析发现 BMI z 评 与者。我们使用 WHO 标准测量了参与者的身高和体 分 +0.58 标准差 (SD) 将优化敏感性，并在此界限值时， 重，并计算了年龄别身体质量指数。我们使用重水同 曲线下区域为 0.86，敏感性为 71.9％（95％ 置信区间， 位素法测量了参与者的体脂，并定义过度肥胖百分位 CI ： 67.4 至 76.0） ，特异性为 91.1％（95％ 置信区间， 数为男孩 >25％，女孩 >30％。我们计算了 BMI z 评 CI ： 89.2 至 92.7）。 分 >+2.00 标准差 (SD) 的敏感性和特异性，并使用受 结论 虽然身体质量指数 (BMI) 仍然是肥胖监测的实用 试者工作特征分析和约登指数来确定过度肥胖分类的 工具，但它低估了过度肥胖的患病率，因此，在规划 最佳 BMI z 评分的界限值。 未来非洲应对儿童肥胖率快速上升的情况时应慎重考 虑。 Résumé Indice de masse corporelle vs méthode de dilution du deutérium pour établir la prévalence de l'obésité chez l'enfant au Ghana, au Kenya, au Maroc, à Maurice, en Namibie, en République-Unie de Tanzanie, au Sénégal et en Tunisie Objectif Comparer la définition de l'obésité de l'Organisation mondiale et défini le taux de masse grasse excessive comme étant > 25% pour de la Santé basée sur l'indice de masse corporelle (IMC) selon l'âge à la les garçons et > 30% pour les filles. Nous avons calculé la sensibilité et masse grasse mesurée chez les enfants africains. la spécificité du Z-score de l'IMC > +2,00 écarts types (ET ) et utilisé une Méthodes Dans le cadre d'une étude prospective multicentrique analyse de la fonction d'efficacité du récepteur et l’indice de Youden menée entre 2013 et 2017, nous avons recruté 1516 participants âgés afin de déterminer la valeur limite optimale du Z-score de l'IMC pour de 8 à 11 ans dans des zones urbaines situées dans huit pays (Ghana, classifier la masse grasse excessive. Kenya, Maroc, Maurice, Namibie, République-Unie de Tanzanie, Sénégal Résultats La prévalence de la masse grasse excessive était plus de trois et Tunisie). Nous avons mesuré leur taille et leur poids et calculé leur fois supérieure à la prévalence de l'obésité définie en fonction de l'IMC IMC par rapport à leur âge en utilisant les normes de l'OMS. Nous avons selon l'âge: 29,1% (IC à 95%: 26,8-31,4; 441 enfants) contre 8,8% (IC à mesuré la masse grasse à l'aide de la méthode de dilution du deutérium, 95%: 7,5-10,4; 134 enfants). La sensibilité du Z-score de l'IMC > +2,00 ET Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 779 Research Adama Diouf et al. Measuring childhood obesity in Africa était faible (29,7%, IC à 95%: 25,5-34,2), tandis que la spécificité était Conclusion Si l'IMC reste est un outil pratique pour surveiller l'obésité, élevée (99,7%, IC à 95%: 99,2-99,9). L'analyse de la fonction d'efficacité il sous-évalue la masse grasse excessive. Cela doit être pris en compte du récepteur a révélé qu'un Z-score de l'IMC de +0,58 ED optimiserait la lors de la planification des futures mesures africaines de lutte contre la sensibilité, et qu'à cette valeur limite, l'aire sous la courbe était de 0,86, pandémie d'obésité chez l'enfant. la sensibilité de 71,9% (IC à 95%: 67,4-76,0) et la spécificité de 91,1% (IC à 95%: 89,2-92,7). Резюме Сравнение эффективности измерения индекса массы тела и метода дейтериевого разбавления для определения распространенности детского ожирения (Гана, Кения, Маврикий, Марокко, Намибия, Объединенная Республика Танзания, Сенегал и Тунис) Цель Сравнение определения ожирения, принятого Всемирной Результаты Распространенность избытка жировой ткани более организацией здравоохранения (ВОЗ) на основании значения чем в три раза превышала частоту ожирения, определяемую «индекс массы тела (ИМТ)-возраст», с измеренными величинами показателем «ИМТ-возраст»: 29,1% (95%-й ДИ: 26,8–31,4; упитанности африканских детей. 441 ребенок) в сравнении с 8,8% (95%-й ДИ: 7,5–10,4; 134 ребенка). Методы В перспективном многоцентровом исследовании в Чувствительность z-оценки ИМТ >+ 2,00 СО была низкой (29,7%, период с 2013 по 2017 год приняли участие 1516 участников в 95%-й ДИ: 25,5–34,2), а специфичность — высокой (99,7%, 95%- возрасте от 8 до 11 лет из городских районов восьми стран (Гана, й ДИ: 99,2–99,9). Анализ характеристических особенностей Кения, Маврикий, Марокко, Намибия, Объединенная Республика правильности обнаружения сигналов позволил обнаружить, что Танзания, Сенегал и Тунис). Детей взвешивали и измеряли их оптимизация чувствительности возможна при использовании рост, после чего вычисляли показатель «ИМТ-возраст» согласно z-оценки ИМТ + 0,58 СО и что при этом пороговом значении стандартам ВОЗ. Также измерялось содержание жировой площадь под кривой составляла 0,86, чувствительность — ткани в организме методом дейтериевого разбавления; 71,9% (95%-й ДИ: 67,4–76,0), а специфичность — 91,1% (95%-й содержание жировой ткани считалось избыточным, если ДИ: 89,2–92,7). оно превышало 25% у мальчиков и 30% у девочек. Авторы Вывод Несмотря на то что измерение ИМТ остается практическим вычислили чувствительность и специфичность z-оценки средством выявления ожирения, оно недооценивает содержание ИМТ >+ 2,00 стандартного отклонения (СО) и воспользовались избыточной жировой ткани в организме, и это следует учитывать методом анализа характеристических показателей правильности при планировании мероприятий по борьбе с пандемией детского обнаружения сигналов индексом Юдена для определения ожирения в Африке. оптимального порога z-оценки по ИМТ в вопросе классификации наличия избытка жировой ткани. Resumen Índice de masa corporal en comparación con el método de dilución de deuterio para establecer la prevalencia de la obesidad infantil, Ghana, Kenya, Marruecos, Mauricio, Namibia, República Unida de Tanzania, Senegal y Túnez Objetivo Comparar la definición de obesidad por edad del índice de Resultados La prevalencia de la obesidad excesiva fue más de tres masa corporal (IMC) de la Organización Mundial de la Salud (OMS) con veces superior a la obesidad definida por el IMC por edad: 29,1 % la grasa corporal medida en niños africanos. (IC del 95 %: 26,8 a 31,4; 441 niños) en comparación con un 8,8 % Métodos En un estudio prospectivo multicéntrico realizado entre 2013 (IC del 95 %: 7,5 a 10,4; 134 niños). La sensibilidad del IMC con DE de y 2017, se reclutaron 1516 participantes de edades comprendidas entre los valores Z > +2,00 fue baja (29,7 %, IC del 95 %: 25,5 a 34,2) y la los 8 y los 11 años de zonas urbanas de ocho países (Ghana, Kenya, especificidad fue alta (99,7 %, IC del 95 %: 99,2 a 99,9). El análisis de Marruecos, Mauricio, Namibia, República Unida de Tanzania, Senegal las características operativas del receptor encontró que un IMC z-score y Túnez). Se midieron la altura y el peso y calculamos el IMC por edad +0,58 DE optimizaría la sensibilidad, y en este corte el área bajo la curva utilizando los estándares de la OMS. Se midió la grasa corporal mediante era de 0,86, con una sensibilidad del 71,9 % (IC del 95 %: 67,4 a 76,0) y el método de dilución de deuterio y se definió el porcentaje de grasa una especificidad del 91,1 % (IC del 95 %: 89,2 a 92,7). corporal excesiva como > 25 % en los niños y > 30 % en las niñas. Se Conclusión Aunque el IMC sigue siendo una herramienta práctica para calculó la sensibilidad y especificidad del IMC con desviaciones estándar la monitorización de la obesidad, subestima el exceso de grasa y esto (DE) de los valores Z de > +2,00 y se utilizó el análisis de las características debería tenerse en cuenta a la hora de planificar las futuras respuestas operativas del receptor y el índice Youden para determinar el límite africanas a la pandemia de obesidad infantil. óptimo del IMC z-score para clasificar el exceso de grasa. References 1. Report of the Commission on Ending Childhood Obesity. Geneva: World 3. Muthuri SK, Wachira LJM, Leblanc AG, Francis CE, Sampson M, Onywera Health Organization; 2016. VO, et al. Temporal trends and correlates of physical activity, sedentary 2. Lobstein T, Jackson-Leach R. Planning for the worst: estimates of obesity behaviour, and physical fitness among school-aged children in sub-Saharan and comorbidities in school-age children in 2025. Pediatr Obes. 2016 Africa: a systematic review. Int J Environ Res Public Health. 2014 03 10;11(5):321–5. doi: http://dx.doi.org/10.1111/ijpo.12185 PMID: 27684716 20;11(3):3327–59. doi: http://dx.doi.org/10.3390/ijerph110303327 PMID: 780 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 Research Measuring childhood obesity in Africa Adama Diouf et al. 16. Introduction to body composition assessment using the deuterium dilution 4. Craig E, Reilly JJ, Bland R. Risk factors for overweight and overfatness in rural South African children and adolescents. J Public Health (Oxf ). 2016 method with analysis of saliva samples by Fourier transform infrared spectrometry. IAEA Human Health Series No. 12. Vienna: International Mar;38(1):24–33. doi: http://dx.doi.org/10.1093/pubmed/fdv016 PMID: 25742718 Atomic Energy Agency; 2010. Available from: https://www-pub.iaea.org/ books/iaeabooks/8369/Introduction-to-Body-Composition-Assessment- 5. The WHO child growth standards. Body mass index-for-age [internet]. Geneva: World Health Organization; 2007. Available from: http://www.who. Using-the-Deuterium-Dilution-Technique-with-Analysis-of-Saliva-Samples- by-Fourier-Transform-Infrared-Spectrometry [cited 2018 Aug 21]. int/childgrowth/standards/en/ [cited 2018 Aug 21]. 6. Okorodudu DO, Jumean MF, Montori VM, Romero-Corral A, Somers VK, 17. Lohman TG. Advances in body composition assessment. Champaign: Human Kinetics; 1992. Erwin PJ, et al. Diagnostic performance of body mass index to identify obesity as defined by body adiposity: a systematic review and meta- 18. Williams DP, Going SB, Lohman TG, Harsha DW, Srinivasan SR, Webber LS, et al. Body fatness and risk for elevated blood pressure, total cholesterol, and analysis. Int J Obes. 2010 May;34(5):791–9. doi: http://dx.doi.org/10.1038/ ijo.2010.5 PMID: 20125098 serum lipoprotein ratios in children and adolescents. Am J Public Health. 1992 Mar;82(3):358–63. doi: http://dx.doi.org/10.2105/AJPH.82.3.358 PMID: 7. Reilly JJ, Wilson ML, Summerbell CD, Wilson DC. Obesity: diagnosis, prevention, and treatment; evidence based answers to common questions. 1536350 19. Reilly JJ. Health effects of overweight and obesity in 195 countries. N Engl J Arch Dis Child. 2002 Jun;86(6):392–4. doi: http://dx.doi.org/10.1136/ adc.86.6.392 PMID: 12023163 Med. 2017 10 12;377(15):1496. PMID: 29020585 20. Reilly JJ, El-Hamdouchi A, Diouf A, Monyeki A, Somda SA. Determining the 8. Reilly JJ, Methven E, McDowell ZC, Hacking B, Alexander D, Stewart L, et al. Health consequences of obesity. Arch Dis Child. 2003 Sep;88(9):748–52. doi: worldwide prevalence of obesity. Lancet. 2018 May 5;391(10132):1773–4. doi: http://dx.doi.org/10.1016/S0140-6736(18)30794-3 PMID: 29739565 http://dx.doi.org/10.1136/adc.88.9.748 PMID: 12937090 9. Javed A, Jumean M, Murad MH, Okorodudu D, Kumar S, Somers VK, et al. 21. Wells JCK, Cole TJ, ALSPAC Study Team. Adjustment of fat free mass and fat mass for height in children aged 8 years. Int J Obes. 2002;26(7):947–52. doi: Diagnostic performance of body mass index to identify obesity as defined by body adiposity in children and adolescents: a systematic review and http://dx.doi.org/10.1038/sj.ijo.0802027 22. Wells JCK, Williams JE, Chomtho S, Darch T, Grijalva-Eternod C, Kennedy K, meta-analysis. Pediatr Obes. 2015 Jun;10(3):234–44. doi: http://dx.doi. org/10.1111/ijpo.242 PMID: 24961794 et al. Body-composition reference data for simple and reference methods and a 4-component model: a new UK reference child. Am J Clin Nutr. 2012 10. Reilly JJ, Kelly J, Wilson DC. Accuracy of simple clinical and epidemiological definitions of childhood obesity: systematic review and evidence appraisal. Dec;96(6):1316–26. doi: http://dx.doi.org/10.3945/ajcn.112.036970 PMID: Obes Rev. 2010 Sep;11(9):645–55. doi: http://dx.doi.org/10.1111/j.1467- 789X.2009.00709.x PMID: 20059704 23. Ruxton CH, Reilly JJ, Kirk T. Body composition of healthy 7–8 year old children and a comparison with the ‘reference child’. Int J Obes. 11. Hills A, Davidsson L. Stable isotope methods to develop and monitor nutrition interventions. Curr Nutr Food Sci. 2010;6:289–93. doi: http:// 1999;23(12):1276–81. doi: http://dx.doi.org/10.1038/sj.ijo.0801067 24. Wells JCK, Coward WA, Cole TJ, Davies PSW. The contribution of fat and fat- dx.doi.org/10.2174/157340110791233238 12. Wells JC, Fewtrell MS. Measuring body composition. Arch Dis Child. 2006 free tissue to body mass index in contemporary children and the reference child. Int J Obes Relat Metab Disord. 2002 Oct;26(10):1323–8. doi: http:// Jul;91(7):612–7. doi: http://dx.doi.org/10.1136/adc.2005.085522 PMID: 16790722 dx.doi.org/10.1038/sj.ijo.0802077 PMID: 12355328 25. Tremblay MS, Shields M, Laviolette M, Craig CL, Janssen I, Connor Gorber 13. Hruschka DJ, Hadley C. How much do universal anthropometric standards bias the global monitoring of obesity and undernutrition? Obes Rev. 2016 S. Fitness of Canadian children and youth: results from the 2007-2009 11;17(11):1030–9. doi: http://dx.doi.org/10.1111/obr.12449 PMID: 27383689 Canadian Health Measures Survey. Health Rep. 2010 Mar;21(1):7–20. PMID: 14. Liu A, Byrne NM, Kagawa M, Ma G, Poh BK, Ismail MN, et al. Ethnic 20426223 differences in the relationship between body mass index and percentage body fat among Asian children from different backgrounds. Br J Nutr. 2011 Nov;106(9):1390–7. doi: http://dx.doi.org/10.1017/S0007114511001681 PMID: 21736824 15. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig L, et al.; STARD Group. STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies. BMJ. 2015 10 28;351:h5527. doi: http://dx.doi. org/10.1136/bmj.h5527 PMID: 26511519 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948 781 Research Adama Diouf et al. Measuring childhood obesity in Africa 782 Bull World Health Organ 2018;96:772–781| doi: http://dx.doi.org/10.2471/BLT.17.205948

Journal

Bulletin of the World Health Organization – Pubmed Central

Published: Sep 10, 2018

References

  • Planning for the worst: estimates of obesity and comorbidities in school-age children in 2025.
  • Temporal trends and correlates of physical activity, sedentary behaviour, and physical fitness among school-aged children in sub-Saharan Africa: a systematic review.
  • Risk factors for overweight and overfatness in rural South African children and adolescents.
  • Diagnostic performance of body mass index to identify obesity as defined by body adiposity: a systematic review and meta-analysis.
  • Obesity: diagnosis, prevention, and treatment; evidence based answers to common questions.
  • Health consequences of obesity.
  • Diagnostic performance of body mass index to identify obesity as defined by body adiposity in children and adolescents: a systematic review and meta-analysis.
  • Accuracy of simple clinical and epidemiological definitions of childhood obesity: systematic review and evidence appraisal.
  • Stable isotope methods to develop and monitor nutrition interventions.
  • Measuring body composition.
  • How much do universal anthropometric standards bias the global monitoring of obesity and undernutrition?
  • Ethnic differences in the relationship between body mass index and percentage body fat among Asian children from different backgrounds.
  • STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies.
  • Body fatness and risk for elevated blood pressure, total cholesterol, and serum lipoprotein ratios in children and adolescents.
  • Health effects of overweight and obesity in 195 countries.
  • Determining the worldwide prevalence of obesity.
  • Adjustment of fat free mass and fat mass for height in children aged 8 years.
  • Body-composition reference data for simple and reference methods and a 4-component model: a new UK reference child.
  • Body composition of healthy 7–8 year old children and a comparison with the ‘reference child’.
  • The contribution of fat and fat-free tissue to body mass index in contemporary children and the reference child.
  • Fitness of Canadian children and youth: results from the 2007-2009 Canadian Health Measures Survey.