TABLE 4

Exit From Obesity Among Obese Fifth-Gradersa and Entry Into Obesity Among Overweight Fifth-Gradersb

Fifth-Grade CharacteristicExit From ObesityEntry Into Obesity
BivariateMultivariatec c = 0.759;
P < .001BivariateMultivariatec c = 0.744;
P < .001
Male 0.85 (0.62–1.16)0.86 (0.61–1.21)1.31 (0.82–2.09)1.51 (0.94–2.44)
Race/ethnicity
 Latino0.58 (0.38–0.89)*1.03 (0.63–1.70)0.93 (0.44–2.00)0.50 (0.22–1.14)
 Non-Latino black0.77 (0.49–1.20)1.31 (0.77–2.24)2.11 (1.03–4.32)*0.93 (0.48–1.78)
 Non-Latino white
 Other0.91 (0.51–1.63)1.01 (0.56–1.82)0.90 (0.32–2.53)0.52 (0.19–1.42)
No 4-year college graduate in household0.47 (0.36–0.63)***0.60 (0.42–0.86)**2.13 (1.16–3.91)*1.83 (0.88–3.79)
Responding parent BMI
 Obese0.54 (0.39–0.74)***0.75 (0.52–1.07)3.59 (2.07–6.25)***2.79 (1.58–4.91)***
 Missing0.80 (0.42–1.52)1.18 (0.62–2.25)2.85 (1.30–6.21)**2.38 (1.05–5.37)*
 Not obese
Child body image discrepancy much heavier than ideal0.30 (0.18–0.48)***0.54 (0.33–0.87)*2.99 (0.84–10.60)1.78 (0.66–4.79)
Child high fast-food and soda consumption0.84 (0.64–1.08)0.92 (0.68–1.26)1.65 (1.01–2.68)*1.13 (0.68–1.89)
Child exercise (days past week)1.03 (0.97–1.09)0.99 (0.92–1.06)0.90 (0.82–0.99)*0.93 (0.83–1.03)
Child TV (per 10 h/wk)0.88 (0.77–1.00)0.95 (0.81–1.11)1.47 (1.22–1.77)***1.26 (1.03–1.53)*
  • All data are presented as OR (95% CI). * P < .05, ** P < .01, *** P < .001 for association, from logistic regression of exit from/entry into obesity on fifth-grade characteristics. Categorical characteristics (ie, race/ethnicity, parent BMI) entered together for bivariate analysis.

  • a Models for exit from obesity include 1014 obese fifth-graders and compare characteristics between the 369 students who were no longer obese (35%) and those who remained obese in 10th grade.

  • b Models for entry into obesity include 739 overweight but non-obese fifth-graders and compare characteristics between the 105 students who became obese (14%) and those who did not become obese in 10th grade.

  • c In addition, multivariate models controlled for continuous child BMI percentile at fifth grade to ensure that results do not merely reflect a factor’s association with exact baseline BMI within a BMI category. Obese fifth-grade children who had higher BMI percentile were less likely to exit obesity by 10th grade (OR, 0.52; 95% CI, 0.45– 0.59) and overweight fifth-grade children who had higher BMI percentile were more likely to enter obesity (OR, 1.23; 95% CI, 1.12–1.36); P < .001 for both results.