Table 3. Multivariate Models Estimating Odds Ratios of Reporting Disordered Weight-Control Behaviora at Follow-up Associated With Planet Health and Other Healthy Choices (HC) Intervention Components in 45 Massachusetts Middle Schools, 2005–2008 (N = 17,567 Students)b .
| Variable | Partially Adjusted Modelsc, OR (95% CI) | P Value | Fully Adjusted Modeld, OR (95% CI) | P Value |
|---|---|---|---|---|
| No. of Planet Health lessons taught on nutrition | 1.01 (0.99–1.02) | .42 | — | — |
| No. of student lesson-exposures to Planet Health lessons on nutrition (per 100 lesson-exposures) | 1.01 (0.97–1.04) | .67 | — | — |
| No. of Planet Health lessons taught on reducing television viewing | 1.00 (0.97–1.03) | .96 | 1.04 (1.02–1.07) | <.001 |
| No. of student lesson-exposures to Planet Health lessons on reducing television viewing (per 100 lesson-exposures) | 0.94 (0.82–1.07) | .36 | 0.80 (0.74–0.85) | <.001 |
| No. of Planet Health lessons taught on physical activity and fitness | 1.00 (0.99–1.01) | .93 | — | — |
| No. of student lesson-exposures to Planet Health lessons on physical activity and fitness (per 100 lesson-exposures) | 1.01 (1.00–1.02) | .003 | 1.01 (1.00–1.01) | .008 |
| No. of Planet Health physical education microunitse taught | 1.00 (0.99–1.01) | .82 | — | — |
| No. of teachers teaching Planet Health lessons | 0.99 (0.97–1.02) | .47 | 1.02 (1.01–1.03) | .002 |
| No. of partners from outside school involved in HC activities | 0.73 (0.62–0.86) | <.001 | 0.76 (0.68–0.84) | <.001 |
| Presence of programs addressing television viewing reduction goals with school staff | 0.54 (0.40–0.73) | <.001 | 0.38 (0.28–0.53) | <.001 |
| HC team meets definition of team | 0.77 (0.68–0.95) | .02 | 0.76 (0.66–0.86) | <.001 |
Abbreviations: OR, odds ratio; CI, confidence interval; —, not applicable.
Disordered weight-control behavior defined as report of any vomiting or use of laxatives or diet pills for weight control in past month.
Both models use individual level as unit of analysis and generalized estimating equation methods to account for school clustering.
Partially adjusted models adjust for school sociodemographic characteristics listed in Table 1 and disordered weight-control behavior at baseline.
Fully adjusted model adjusts for all variables in the table that remained significant at P < .05 after process of backward elimination, school sociodemographic characteristics, and disordered weight-control behavior at baseline.
Lessons lasting only 5 to 10 minutes.