Body mass index and parent-reported self-esteem in elementary school children: evidence for a causal relationship

Abstract

OBJECTIVE: To clarify relationships between body mass index (BMI) and self-esteem in young children at a population level. To assess whether low self-esteem precedes or follows development of overweight/obesity in children.

DESIGN: Prospective cohort study in elementary schools throughout Victoria, Australia. Child BMI and self-esteem were measured in 1997 and 2000.

SUBJECTS: Random sample of 1157 children who were in the first 4 y of elementary school (aged 5–10 y) at baseline.

MEASURES: BMI was calculated from measured height and weight, then transformed to z-scores. Children were classified as nonoverweight, overweight or obese based on international cut-points. Low child self-esteem was defined as a score below the 15th percentile on the self-esteem subscale of the parent-reported Child Health Questionnaire.

RESULTS: Overweight/obese children had lower median self-esteem scores than nonoverweight children at both timepoints, especially at follow-up. After accounting for baseline self-esteem, higher baseline BMI z-score predicted poorer self-esteem at follow-up (P=0.008). After accounting for baseline BMI z-score, poorer baseline self-esteem did not predict higher BMI z-score at follow-up. While nonoverweight children with low baseline self-esteem were more likely to develop overweight/obesity (OR=2.1, 95% CI=1.2, 3.6), this accounted for only a small proportion of the incidence of overweight.

CONCLUSIONS: Our data show an increasingly strong association between lower self-esteem and higher body mass across the elementary school years. Overweight/obesity precedes low self-esteem in many children, suggesting a causal relationship. This indicates that prevention and management strategies for childhood overweight/obesity need to begin early to minimise the impact on self-esteem.

Introduction

Childhood obesity, now one of the most common chronic conditions of childhood,1 is believed to be strongly associated with psychosocial morbidity.2 For individual children, the immediate psychosocial effects of social isolation, discrimination, and peer problems can accompany childhood obesity.3 By adolescence lower self-esteem,1, 4 combined with increased rates of sadness, loneliness and nervousness,5 has been reported for obese children.

In this paper, we focus on temporal relationships between self-esteem and overweight/obesity in a large population-based sample of children tracked prospectively for 3 y during their elementary schooling. A review6 of 35 studies of self-esteem and adiposity concluded that while studies of adolescents have consistently shown an inverse relationship between self-esteem and overweight or obesity, studies of younger children have been less consistent, and have generally reported weaker associations. Studies of adolescents have also suggested that obesity may be causally related to lower self-esteem with one treatment programme for adolescent obesity resulting in improved self-esteem.7 No comparable studies in younger children have been reported. A stronger relationship between adiposity and self-esteem is generally reported when body esteem or body image is the primary aspect of self-esteem being measured.6

Longitudinal research addressing causal relationships between overweight/obesity and self-esteem in young children is extremely limited. For elementary school-aged children, longitudinal studies have suggested greater decreases in self-esteem over time for obese than nonobese children,5 and an inverse relationship between change in adiposity and change in self-esteem.8, 9 We are not aware of any longitudinal studies in young children that have considered whether poor self-esteem influences changes in adiposity.

Apart from its paucity, current research literature in this area is also limited by methodological issues. Firstly, the available evidence is commonly derived from clinical rather than epidemiological samples; however, it has been indicated that lower levels of psychological concern exist in nonclinical samples of overweight and obese children.10 Many overweight/obese children present to primary care or paediatric clinics with established comorbidities, so that existing research may be heavily biased towards obese children with more severe physical and/or mental health problems.10 It is not known by what age lower self-esteem and overweight/obesity become linked. Nor is the direction of causality understood: do overweight/obese children develop lower self-esteem in response to their weight, or are children with lower self-esteem from the outset more likely to increase their body mass index (BMI), or is there an ongoing transactional relationship?

This study reports temporal relationships between self-esteem and BMI in a large community cohort of young children tracked for 3 y, from the early to late elementary school years. We aimed to assess whether heavier children consistently experience poorer self-esteem across these years. We also looked for evidence of causality in either of the postulated directions: that poorer baseline self-esteem predicts increases in BMI, vs higher baseline BMI predicting decreases in self-esteem over the elementary school years.

Methods

Sample

The baseline sample was drawn from the Health of Young Victorians Study (HOYVS), a large epidemiologic study of Australian children's health, well-being and anthropometry conducted between September and December 1997. HOYVS employed a two-stage stratified random sampling design to draw a sample of elementary school children representative of the state of Victoria (population 4.9 million), Australia. Methods and findings have been reported elsewhere.11, 12, 13 Briefly, 24 elementary schools were randomly selected to be representative of Victoria's government, Catholic, and independent school sectors. The final achieved sample of 3174 children (75% response) in Years Preparatory-6 mirrored Victorian census data (Australian Bureau of Statistics, 1998) for children by age distribution, gender, ethnicity (parental county of birth), and proportion of indigenous persons (Australian Aboriginal and Torres Strait Islanders). Children in the first 4 y of formal schooling (Years Preparatory-3) who provided data in the baseline survey were included in the follow-up study (when in years 3–6), conducted between October 2000 and June 2001. The time interval between baseline and follow-up assessments ranged from 3.0 to 3.7 y (mean 3.2 y, s.d. 0.2 y).

Measures

At baseline and follow-up children's height and weight were measured by trained field workers. Height was measured using the Invicta portable rigid stadiometer, which measures to the nearest 1 mm to a height of 207 cm; weight was measured using Tanita model 1597 digital scales which were regularly calibrated and shown to be accurate within 20 g. Investigation of anthropometric reliability found no evidence of systematic bias for intra-rater or inter-rater comparisons. BMI was calculated (weight (kg)/height (m)2) and children were categorised into BMI categories (nonoverweight, overweight, and obese) according to the International Obesity Task Force's gender and age-specific cutoff points.14 In addition, BMI was transformed to standardised (z) scores based on gender and exact age, using the LMS method15 and the Centres for Disease Control and Prevention 2000 Growth Chart data.16 This transformation adjusts for skew in the BMI distribution and physiological changes that occur in child BMI with age, enabling analyses across gender and age groups with BMI as a continuous variable.

Parents provided the child's date of birth from which exact age was calculated for use in the child BMI transformations described above. At both baseline and follow-up, parents completed the self-esteem scale from the Australian authorised adaptation of the Child Health Questionnaire (CHQ PF50).13 The CHQ PF50, a parent-proxy report of the functional health status of children aged 5–18 y, has been found to have good psychometric properties when used in populations of healthy children13, 17 and children with chronic diseases.18, 19 The self-esteem scale is a six-item scale enquiring about the parent's perception of the child's satisfaction with their school ability, athletic ability, friendships, appearance, family relationships, and their life overall. All items are based on 4-week recall, and response choices range from ‘very satisfied’ to ‘very dissatisfied’ on a five-point scale. A scale score is calculated and transformed to a possible range of zero to 100, with lower scores indicating poorer self-esteem, and higher scores indicating better self-esteem.

Analyses

CHQ PF50 self-esteem scale scores were calculated according to the manual.20 Unlike the normally distributed BMI z-score variable, self-esteem scores at both baseline and follow-up were left skewed with some ceiling effects (20 and 15% of the sample, respectively, attaining the highest possible score), requiring use of nonparametric statistics for analyses involving these variables. Low self-esteem scores were defined as scores which fell below the 15th percentile (a score of <66.7 for both baseline and follow-up self-esteem), which is close to the parametric concept of one standard deviation below the mean.

Characteristics of children retained and lost to follow-up were compared using independent sample t-test, Mann–Whitney two-sample test, or χ2 statistic as appropriate. Baseline and follow-up self-esteem scores were compared using the Wilcoxon matched pairs sign-rank test. Gender differences in self-esteem scores were assessed by the Mann–Whitney two-sample test, and differences by age group (5.0–5.9, 6.0–6.9, 7.0–7.9, 8.0–8.9, or 9.0–10.9 y at baseline) and by BMI category were assessed by the Kruskal–Wallis rank test. Proportions of children with low self-esteem scores were compared across BMI categories using the χ2 statistic. Linear regression was used to assess whether baseline self-esteem score predicted follow-up BMI z-score. Mean BMI z-score at follow-up was compared between children with low and nonlow self-esteem scores by the independent samples t-test. Logistic regression analyses assessed the odds of being overweight/obese at follow-up for children with low self-esteem scores at baseline.

Both studies received approval from the Royal Children's Hospital Ethics in Human Research Committee, the Victorian Department of Education, and the Catholic Education Office. Informed consent was received from a parent or guardian, and all participating children provided assent.

Results

In all, 83% (1943/2336) of children in Grades preparatory-three (age range 5.0–10.7 y) participated in the baseline study, of whom 81% (1569) provided data in the follow-up study. Of these, 412 children were excluded from analyses for this paper due to incomplete data providing a final sample of 1157 children (Table 1). Over the 3 y follow-up period, 102 (8.8%) of the nonoverweight children became overweight or obese, while 46 of the 220 (20.9%) overweight/obese children moved into the normal weight category.

Table 1 Baseline and follow-up characteristics of retained cohort (n=1157)

The retained cohort was similar to those lost to follow-up in terms of gender (49 vs 53% male; P=0.09), mean age (7.6 vs 7.7 y; P=0.15), median self-esteem score (87.5 vs 87.5; P=0.24) and mean BMI z-score (0.43 vs 0.50; P=0.13) at baseline. However, the crude BMI of children retained in the cohort was lower than for children lost to follow-up (16.9 vs 17.3 kg/m2; P=0.001) and a smaller proportion was categorised as overweight or obese at baseline (15.0 and 4.0%, respectively vs 16.9 and 8.2%, respectively; P<0.001).

For the retained sample, median self-esteem score decreased over the course of the study (87.5 vs 83.3; P<0.001), corresponding to a mean drop in self-esteem score of 1.4 (s.d.=15.7). The correlation between self-esteem scores at the two time points was 0.49 (Spearman r, P<0.001). There was no gender difference in median self-esteem scores at baseline (P=0.10) or follow-up (P=0.08); nor was there a difference by age group (P=0.20 and 0.58, respectively).

At baseline, the median self-esteem score for children who were overweight and obese was slightly lower than for children who were nonoverweight, and higher proportions of overweight and obese children had self-esteem scores below the 15th percentile (Table 2). After 3 y, these cross-sectional relationships between BMI category and self-esteem were much more marked: nearly half of all obese children fell below the 15th percentile for self-esteem, and median self-esteem score was 17 points lower in the obese than the nonoverweight children (Table 2). Even so, most (68%) children with low-self-esteem scores at follow-up were neither overweight nor obese.

Table 2 Self-esteem across BMI categories at baseline and follow-up

After accounting for baseline BMI z-score, poorer baseline self-esteem score did not predict higher BMI z-score at follow-up (P=0.32). However, children with low self-esteem scores at baseline were marginally more likely to be classified as overweight/obese at follow-up (OR=1.5, 95% CI=1.0, 2.6). This relationship became stronger when considering only nonoverweight children: those with low baseline self-esteem scores were more likely to develop overweight/obesity by follow-up than children with higher baseline self-esteem scores (OR=2.1, 95% CI=1.2, 3.6). However, most (78%) of the children who became overweight did not have low baseline self-esteem scores, and the vast majority (88%) of normal weight children with low baseline self-esteem scores remained nonoverweight at follow-up. Baseline self-esteem score did not predict which overweight/obese children would move into the nonoverweight category during the study (OR=1.4, 95% CI=0.7, 2.8).

The converse relationship (ie impact of baseline BMI on subsequent self-esteem) was more striking. Higher baseline BMI z-score predicted poorer self-esteem score at follow-up (P=0.008). Similarly, children classified as overweight or obese at baseline had lower self-esteem scores at follow-up (P<0.001), and were significantly more likely to fall into the low self-esteem category (OR=1.8, 95% CI=1.2, 2.6). All three analyses accounted for baseline self-esteem score.

Discussion

Self-esteem and BMI are clearly related, especially by the later elementary school years. Our longitudinal data suggest that BMI may play an important causal role in the development of lower self-esteem experienced by many overweight and obese children. While the reverse proposition—that early self-esteem would predict later BMI — was less compelling, poor self-esteem did increase the odds of nonoverweight 5–10-y-olds becoming overweight or obese later in elementary school.

Overall, self-esteem scores fell slightly as children got older. This corresponds with cross-sectional data showing that older children generally have poorer self-esteem than their younger counterparts.13 Although lower self-esteem was evident in overweight and obese children at both time points, this relationship was much more marked at follow-up, when nearly half the obese children fell in the lowest 15% of self-esteem scores. This may partly be related to the fact that children were older, and/or that they had been overweight/obese for longer (since most children did not change weight categories over the course of the study). Regardless, this finding suggests that very early interventions for childhood obesity may be needed to prevent significant impacts on self-esteem. Randomised controlled trials would be needed to demonstrate whether this is actually true, and whether interventions that effectively reduce overweight in older children could also reverse the lower self-esteem.

This study was limited to parent-proxy reports of children's self-esteem. Since parents of obese young children frequently neither recognise nor feel concerned about a child's established weight problems,12, 21 many parents may not perceive, and may therefore under-report a true impact on the emotional well-being of younger children. Ideally, future studies of self-esteem would include child self-report in addition to parent reports. Nonetheless, we believe that the parent-proxy reports are likely to provide an approximation sufficient to support our conclusions. Certainly for adolescents, means, medians and proportion of ceiling scores are similar when adolescents and parent-proxies concurrently report adolescent CHQ self-esteem scores.22, 23 We have also previously shown that the parent's own weight category does not seem to alter how they perceive the health and well-being of overweight/obese children, and in particular is not associated with differences on the parent-reported child self-esteem scale of the CHQ.12

The self-esteem measure used in this study was drawn from a larger health-related quality of life instrument. While the self-esteem subscale shows good internal consistency, it has not been validated against other specific measures of child self-esteem.20 However, using this instrument we observed a decrease in self-esteem with age, which corresponds to findings from cross-sectional studies employing other measures of self-esteem, giving some confidence in its validity. A further limitation of this study was the loss to follow-up of a larger proportion of overweight and obese children. While this may have reduced the power of the study to identify relationships with overweight and obesity, it should not have impacted on the internal validity of the findings.

One of the major strengths of this study is its longitudinal design, enabling possible causal relationships to be investigated and directionality to be established. The clear associations between low self-esteem and overweight/obesity in this epidemiological sample, likely to have less psychosocial morbidity than clinical samples, suggests that overweight/obesity does indeed have a strong effect on self-esteem in otherwise healthy young children.

Based on these data, approaches focusing solely on self-esteem seem unlikely to be particularly effective in the management of overweight/obesity in the elementary school years. Interventions for overweight/obese children in this age group should focus on reducing BMI, with improved self-esteem likely to be a flow-on effect. Interventions to lower the incidence of overweight in the early school years might incorporate strategies to enhance self-esteem, along with promotion of healthy nutrition and physical activity.

References

  1. 1

    Dietz WH . Health consequences of obesity in youth: childhood predictors of adult disease. Pediatrics 1998; 101: 518–525.

    CAS  Google Scholar 

  2. 2

    World Health Organisation. Obesity: preventing and managing the global epidemic. WHO: Geneva; 1998.

  3. 3

    Stunkard AJ, Wadden TA . Psychological aspects of severe obesity. Am J Clin Nutr 1992; 55 (Suppl 2): 524S–532S.

    CAS  Article  Google Scholar 

  4. 4

    Stunkard AJ, Mendelson M . Obesity and the body image. 1. Characteristics of disturbances in the body image of some obese persons. Am J Psychiatry 1967; 123: 1300.

    Article  Google Scholar 

  5. 5

    Strauss RS . Childhood obesity and self-esteem. Pediatrics 2000; 105: e15.

    CAS  Article  Google Scholar 

  6. 6

    French SA, Story M, Perry CL . Self-esteem and obesity in children and adolescents: a literature review. Obes Res 1995; 3: 479–490.

    CAS  Article  Google Scholar 

  7. 7

    Mellin LM, Slinkard LA, Irwin CE . Adolescent obesity intervention: validation of the SHAPEDOWN program. JAMA 1987; 87: 333–338.

    CAS  Google Scholar 

  8. 8

    Kolody B, Sallis JF . A prospective study of ponderosity, body image, self-concept, and psychological variables in children. J Dev Behav Pediatr 1995; 16: 1–5.

    CAS  Article  Google Scholar 

  9. 9

    Brown KM, McMahon RP, Biro FM, Crawford P, Schreiber GB, Similo SL, Waclawiw M, Striegel-Moore R . Changes in self-esteem in black and white girls between the ages of 9 and 14 years: the NHLBI Growth and Health Study. J Adolesc Health 1998; 23: 7–19.

    CAS  Article  Google Scholar 

  10. 10

    Braet C, Mervielde I, Vandereycken W . Psychological aspects of childhood obesity: a controlled study in a clinical and nonclinical sample. J Pediatr Psychol 1997; 22: 59–71.

    CAS  Article  Google Scholar 

  11. 11

    Lazarus R, Wake M, Hesketh K, Waters E . Change in body mass index in Australian primary school children, 1985–1997. Int J Obes Relat Metab Disord 2000; 24: 679–684.

    CAS  Article  Google Scholar 

  12. 12

    Wake M, Salmon L, Waters E, Wright M, Hesketh K . Parent-reported health status of overweight and obese Australian primary school children: a cross-sectional population survey. Int J Obes Relat Metab Disord 2002; 26: 717–724.

    CAS  Article  Google Scholar 

  13. 13

    Waters E, Salmon L, Wake M, Hesketh K, Wright M . The Child Health Questionnaire in Australia: reliability, validity and population means. Aust NZ J Public Health 2000; 24: 207–210.

    CAS  Article  Google Scholar 

  14. 14

    Cole TJ, Bellizzi MC, Flegal KM, Dietz WH . Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320: 1240–1243.

    CAS  Article  Google Scholar 

  15. 15

    Cole TJ, Green PJ . Smoothing reference centile curves: the LMS method and penalized likelihood. Stat Med 1992; 11: 1305–1319.

    CAS  Article  Google Scholar 

  16. 16

    Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS, Wei R, Mei Z, Curtin LR, Roche AF, Johnson CL . CDC growth charts: United States. Adv Data 2000; 314: 1–27.

    Google Scholar 

  17. 17

    Waters E, Salmon L, Wake M . The parent-form Child Health Questionnaire in Australia: comparison of reliability, validity, structure and norms. J Pediatr Psychol. 2000; 25: 381–391.

    CAS  Article  Google Scholar 

  18. 18

    Wake M, Hesketh K, Cameron F . The Child Health Questionnaire in children with diabetes: cross- sectional survey of parent and adolescent-reported functional health status. Diabet Med 2000; 17: 700–707.

    CAS  Article  Google Scholar 

  19. 19

    Waters EB, Wake MA, Hesketh KD, Ashley DM, Smibert E . Health-related quality of life of children with acute lymphoblastic leukaemia: comparisons and correlations between parent and clinician reports. Int J Cancer 2003; 103: 514–518.

    CAS  Article  Google Scholar 

  20. 20

    Landgraf JM, Abetz L, Ware JE . The Child Health Questionnaire (CHQ): A user's manual. The Health Institute: Boston; 1996.

    Google Scholar 

  21. 21

    Baughcum AE, Chamberlin LA, Deeks CM, Powers SW, Whitaker RC . Maternal perceptions of overweight preschool children. Pediatrics 2000; 106: 1380–1386.

    CAS  Article  Google Scholar 

  22. 22

    Waters E, Salmon L, Wake M, Wright M, Hesketh K . The interpretation guide for the Australian authorised Adaptation of the Child Health Questionnaire: Adolescent/self report form: CHQ CF-80 & CF-50, Centre for Community Child Health. Melbourne, Victoria, May 2001.

  23. 23

    Waters E, Salmon L, Wake M, Wright M, Hesketh K . The interpretation guide for the Australian Authorised Adaptation of the Child Health Questionnaire: Parent/proxy form: CHQ PF-50 and PF-28, Centre for Community Child Health. Melbourne, Victoria July 1999.

Download references

Acknowledgements

This study was supported by grants from the National Heart Foundation, Financial Markets for Children, and Murdoch Childrens Research Institute. KH is supported by a National Health and Medical Research Council Public Health Postgraduate Scholarship. We acknowledge the contribution to data collection of Bibi Gerner, Louisa Salmon, Naomi Paine, Susan Gallagher, Jennifer van Gemert, Sarah Barker and Winnie Lau.

Author information

Affiliations

Authors

Corresponding author

Correspondence to K Hesketh.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hesketh, K., Wake, M. & Waters, E. Body mass index and parent-reported self-esteem in elementary school children: evidence for a causal relationship. Int J Obes 28, 1233–1237 (2004). https://doi.org/10.1038/sj.ijo.0802624

Download citation

Keywords

  • self-esteem
  • body mass index
  • child
  • longitudinal studies

Further reading

Search