ABSTRACT
Background: There is mounting evidence that the prevalence of overweight and obesity in children is reaching epidemic proportions in North America. We compared parent-report vs. measured BMI overweight and obesity prevalence estimates among 9 year olds using the 1996 NLSCY reports published by Willms et al. (2003) and anthropometric measurements from a regional population of public school children.
Methods: Body mass index (BMI) was calculated for 1,497 9-year-old children (males N=734; females N=763) from 75 public schools in the Niagara Region of Ontario, Canada. BMI from the 1996 NLSCY was based on parental reports of height and weight of 879 nine year olds. To define overweight and obese children, we used internationally accepted age- and gender-specific cut-offs as defined by Cole et al. (2000).
Results: The NLSCY overweight prevalence estimates of boys and girls may overestimate overweight boys and girls by 17% and 10%, respectively. Measured obesity prevalence estimates were similar to parent-reports.
Conclusions: Our results suggest that parental reports of height and weight may inflate prevalence estimates of overweight children, but appear reasonably accurate for estimating obesity. Since prevalence of overweight and obesity are often combined to form a global estimate, reliance on parent-reported height and weight may overstate the magnitude of the problem.
MeSH terms: Children; obesity; overweight; body weights and measures; anthropometry; body mass index; epidemiology
Recent work by Willms and colleagues (2003)1 comparing the 1981 Canadian Fitness Survey (CFS) to the 1996 National Longitudinal Survey of Children and Youth (NLSCY) demonstrates an approximately threefold increase in the national prevalence of overweight and obesity among Canadian children 7 to 13 years old during this time period. This is based on overweight and obese body mass index (BMI (kg/m2)) categories commonly used to gauge the prevalence of weight problems in youth.2-4
The 1981 CFS used direct measures of child height and weight to compute BMI5 while the 1996 NLSCY BMI calculations were based principally* on parental reports of height and weight. Previous studies have shown that adult self-reports of height and weight are unreliable6,7 as are their reports of their child's height and weight.8 For example, Spencer et al. (2002) found that obese adults were least likely to be correctly classified as obese.7 Moreover, Goodman et al. (2000) found that 34% of obese teens were not identified by selfreported or parent-reported height and weight.8
As such, whether the substantial reported rise in the prevalence of child overweight and obesity represents a true increase or is a function of the shift from anthropometric measurements in 1981 (CFS) to parental reports in 1996 (NLSCY) is open to debate. In an attempt to address this issue, we compare the 1996 BMI rates among 9 year olds calculated from parent-reported height and weight (as reported by Willms et al., 2003) to BMI rates calculated from measured height and weight in a school population of 9 year olds.
METHODS
Parent-reported height and weight sample
The calculated BMI scores based on parent-estimated height and weight in the 1996 NLSCY1 are used as the basis for comparison. The 1996 NLSCY9 contains children from newborn to age 13. Using a multi-staged, stratified random sampling procedure, participating households were chosen from Statistics Canada's Labour Force Survey. Children living in Indian reserves, military housing, institutions, and remote northern areas were excluded from the NLSCY. One person from each household was selected as the "person most knowledgeable" (PMK), usually the mother, to provide detailed information on the child, including height and weight. From this information, BMI was calculated for each child. Overweight and obese BMI categories were defined using international age- and gender-specific cut-offs.2 In order to make a comparison with the current data (see below), we restricted our analysis to the 879 nine year olds in the NLSCY (males N=432; females N=447).
Measured height and weight sample
The target population of children in our comparison sample consisted of the entire grade four cohort from the District School Board of Niagara (DSBN). Children with physical disabilities (i.e., wheelchairs and walkers) that limited their participation in the Leger shuttle run10 were excluded. Research ethics approval was granted by both Brock University and the DSBN review boards, and consent to enter schools was obtained from school principals. Of a potential cohort of 2,534 students in grade 4 from 75 schools, parents of 2,303 (91%) provided consent. In this study, we analyzed the results of 1,497 children who were 9 years old at the time of data collection in the fall of 2004 (males N=734; females N=763). We excluded 806 8- and 10-year-old children as they would not be representative of all 8 and 10 year olds, the majority of them being one grade behind (Grade 3) or ahead (Grade 5) and not included in the target population cohort.
Anthropometric measurements of height (cm) and weight (kg) were made on all children. Standing height was measured without footwear to the nearest 0.5 cm using a wall-mounted stadiometer. Weight was measured in clothing (without shoes) required for light physical exercise to the nearest 0.1 kg using a calibrated, electronic scale. Normal, overweight, and obese BMI categories, based on measured height and weight (kg/m2), were based on the same age- and gender-specific cut-offs developed by Cole and colleagues2 and used in the NLSCY analysis.1 A z-statistic was used to test for differences in prevalence estimates across samples.
RESULTS
Table I shows the prevalence of overweight and obesity in 9-year-old males and females based on reported and measured BMI values. Results provided from Willms et al. (2003) are weighted sample estimates from the NLSCY.
The estimate of obesity prevalence in the school sample did not differ significantly from the NLSCY. However, a significant difference in prevalence estimates was found in the overweight BMI category. The parent-reported NLSCY prevalence of overweight 9 year olds was 17% and 10% higher for males and females, respectively, compared to the measured school sample.
DISCUSSION
This analysis partially supports previous findings that parental reports of their children's height and weight are unreliable,8 specifically for the overweight category. This suggests that overweight population estimates from the 1996 NLSCY may be significantly overstated. However, our results agree with the obesity estimates reported by Willms et al. (2003).
To address this discrepancy within the overweight category, we consider one possible explanation. Accurately reporting the height and weight of one's child is quite difficult.8 Troiano and colleagues (1995)11 examined overweight childhood trends and noted that a narrow height and weight range is required to accurately classify a child as overweight. Nine-year-old children are considered overweight if their BMI falls between 19.1 and 22.77 for males and between 19.07 and 22.81 for females.2 Since the classification of obese is bounded only by a lower cut-off, slight over- or underestimates in parent-reported height and weight are not as likely to shift a child out of this category. Moreover, parents with obese children may be more aware of their weight issues and more able to provide accurate reporting.
Although our results suggest that current youth overweight prevalence rates may overestimate the true rate, we are limited by the comparison of regional data to national data. These anthropometric estimates could be different from Canadian estimates. Notwithstanding this, in a recent Statistics Canada report,12 St. Catharines was identified as having the 10th highest percentage of obese adults (17.2%) among CMAs (Census Metropolitan Area) and was higher than the Canadian average (14.9%), suggesting that the differences reported here may be conservative. While adult obesity estimates do not translate directly to child estimates, there is a link between obesity in childhood and subsequent obesity in adulthood.13 The proposed Canadian Health Measures Survey (CHMS) by Statistics Canada and the Canadian Community Health Survey (CCHS) will permit a direct national comparison of prevalence estimates derived from direct measurement and parental reports of child and adult height and weight.14 However, the CCHS 2.1 does not include children under the age of 12. Moreover, the CHMS has recently received Health Canada ethics approval and conducted a dress rehearsal in Ottawa in preparation for its full nationwide implementation (Personal Communications, CHMS Fall 2006, Issue 10, and Spring 2007, Issue 11 newsletters).
These findings are important from a public health perspective in addition to being important methodologically. When rates of obesity are reported in the media, overweight and obese categories often are combined, providing estimates that may exaggerate the extent of the problem.15 Our combined estimates indicate a third of the pediatric population is at risk, almost 20% less than that previously reported.1 The prevalence rates we find still indicate an increase in overweight and obese children from the 1981 CFS study,16 albeit smaller than previously reported. As such, it is vital from a public health perspective to note that childhood obesity is increasing and we must continue to focus on reversing this trend.
[Sidebar]
R�SUM�
Contexte : Il appara�t de plus en plus que le surpoids et l'ob�sit� chez les enfants prennent des proportions �pid�miques en Am�rique du Nord. Nous avons compar� les estimations de pr�valence du surpoids et de l'ob�sit� chez les enfants de 9 ans, selon les d�clarations parentales et selon la mesure directe de l'indice de masse corporelle (IMC) des enfants, en nous aidant des rapports tir�s de l'Enqu�te longitudinale nationale sur les enfants et les jeunes (ELNEJ) de 1996 publi�s par Willms et coll. (2003) et des mesures anthropom�triques d'une population r�gionale d'enfants fr�quentant l'�cole publique.
M�thode : Nous avons calcul� l'IMC de 1 497 enfants de 9 ans (734 gar�ons et 763 filles) fr�quentant 75 �coles publiques de la r�gion de Niagara, en Ontario (Canada). Les IMC figurant dans l'ELNEJ de 1996 �taient fond�s sur les d�clarations parentales de la taille et du poids de 879 enfants de 9 ans. Pour cat�goriser les enfants ob�ses et ceux pr�sentant un surpoids, nous avons utilis� les seuils de d�marcation selon l'�ge et le sexe d�finis par Cole et coll. (2000), qui sont internationalement accept�s.
R�sultats : Les estimations de pr�valence du surpoids chez les gar�ons et les filles selon l'ELNEJ pourraient surestimer (de 17 % chez les gar�ons et de 10 % chez les filles) la proportion d'enfants pr�sentant un surpoids. Par contre, les estimations de pr�valence de l'ob�sit� tir�es des mesures anthropom�triques et des d�clarations parentales �taient semblables.
Conclusion : Ces r�sultats donnent � penser que les d�clarations parentales de la taille et du poids pourraient gonfler les estimations de pr�valence du surpoids chez les enfants, mais que ces estimations sont raisonnablement pr�cises pour ce qui est de l'ob�sit�. Comme on combine souvent la pr�valence du surpoids et de l'ob�sit� pour produire une estimation globale, on risque de surestimer l'ampleur du probl�me si l'on se fie aux d�clarations parentales de la taille et du poids.
[Reference]
REFERENCES
1. Willms JD, Tremblay MS, Katzmarzyk PT. Geographic and demographic variation in the prevalence of overweight Canadian children. Obes Res 2003;11:668-73.
2. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: International survey. BMJ 2000;320:1-6.
3. Flouris AD, de Ruiter W, Faught BE. Establishing new cutoffs for obesity in Canadian children. CJAP 2003;28:S54.
4. Katzmarzyk PT. The Canadian obesity epidemic, 1985-1998. CMAJ 2002;166:1039-40.
5. Villeneuve PJ, Morrison HI, Craig CL, Schaubel DE. Physical activity, physical fitness, and risk of dying. Epidemiology 1998;9:626-31.
6. Rowland ML. Self-reported weight and height. Am J Clin Nutr 1990;52:1125-33.
7. Spencer EA, Appleby PN, Davey GK, Key TJ. Validity of self-reported height and weight in 4808 EPIC-Oxford participants. Public Health Nutrition 2002;5:561-65.
8. Goodman E, Hinden BR, Khandelwal S. Accuracy of teen and parental reports of obesity and body mass index. Pediatrics 2000;106:52-58.
9. National Longitudinal Survey of Children: Overview of Survey Instruments for 1994-1995, Data Collection, Cycle I. Ottawa: Statistics Canada and Human Resources Development Canada. Cat. no 95-02.
10. Leger LA, Mercier D, Gadoury C, Lambert J. The multistage 20 metre shuttle run test for aerobic fitness. J Sports Sci 1988;6:93-101.
11. Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL. Overweight prevalence and trends for children and adolescents: The National Health and Nutritional Examination Survey, 1963-1991. Arch Pediatr Adolesc Med 1995;149:1085-91.
12. Gilmore J. Trends and conditions in census metropolitan areas: Health of Canadians living in census metropolitan areas. Ottawa: Statistics Canada Health Statistics Division, 2004. Cat. no 89-613-MIE.
13. Ontario Medical Association. An ounce of prevention or a ton of trouble: Is there an epidemic of obesity in children? A position paper by the Ontario Medical Association. Ontario Medical Review 2005;32-44.
14. Shields M. Measured obesity: Overweight Canadian children and adolescents. Ottawa: Statistics Canada Health Statistics Division, 2005. Cat. no 82-620-MWE.
15. Picard A. Fattest and Fittest: St. Catharines Leads the Fat Parade. The Globe and Mail, Saturday, July 21, 2001.
16. Tremblay MS, Willms JD. Secular trends in the body mass index of Canadian children. CMAJ 2000;163:1429-33.
Received: March 14, 2006
Accepted: November 21, 2006
[Author Affiliation]
Alayna Banach, MSc1
Terrance J. Wade, PhD1,2
John Cairney, PhD1,3
John A. Hay, PhD1
Brent E. Faught, PhD1
Deborah D. O'Leary, PhD1
La traduction du r�sum� se trouve � la fin de l'article.
1. Department of Community Health Sciences, Brock University, St. Catharines, ON
2. Department of Child and Youth Studies, Brock University
3. Departments of Psychiatry and Public Health Sciences, University of Toronto and the Centre for Addiction and Mental Health, Toronto, ON
Correspondence and reprint requests: Dr. Terrance Wade, Department of Community Health Sciences, Brock University, 500 Glenridge Ave, St. Catharines, ON L2S 3A1, Tel: 905-688-5550, ext. 4146, Fax: 905-688-8954, E-mail: twade@brocku.ca
Acknowledgements: This study was supported by research grant # 66959 from the Canadian Institutes for Health Research (Principal Investigators: Dr. Hay & Dr. Cairney). Drs. Wade and Cairney are supported through the Canada Research Chair program by SSHRC and CIHR, respectively.
No comments:
Post a Comment