Planet Health

Study 1

Evaluation Methodology

Design: Evaluation of the Planet Health intervention involved a randomized, controlled trial with 5 intervention and 5 control schools. Data were collected over two years: baseline data were collected on a cohort of students at the beginning of grades 6 and 7 in fall 1995, with post-test data collected in spring 1997 (grades 7 and 8). The sample included 1,295 students from public schools in four urban Boston (MA) communities. Post-test data were collected in spring 1997 on 83% of the baseline sample. Planet Health sessions were delivered within existing curricula using classroom teachers. Lessons focused on decreasing television viewing and consumption of high-fat foods, while increasing fruit/vegetable intake and physical activity.

Schools were the unit of randomization. Recruitment of schools was based on willingness to implement the classroom and physical education curriculum. Control schools received their usual health curricula and PE classes.

Recruitment: Baseline data were collected in fall 1995 on 1,560 students in both intervention and control schools. Overall participation rate was 64.5% in control schools and 64.8% in intervention schools. A total of five schools required active parental consent. Consent rates were 58% among schools with active consent and 89% among schools with passive consent. The analysis was based on 1,295 students who completed both baseline and post-test data.

Attrition: Post-test data were collected in spring 1997 on 83% of the baseline sample (n = 1,295). Students were excluded who transferred schools, were in special education classes or were in the wrong grade. It should be noted that students who completed the Spanish-language version of the questionnaire (5% of eligible students) were excluded from the analysis.

Sample: The sample used for analysis consisted of 1,295 students who completed both baseline and post-test assessments. Of these students, 48% were female and 52% male. Average student age was approximately 12 years old. Racial/ethnic breakdown was as follows: 66% white, 13% African American, 13% Hispanic, 8% Asian/Pacific Islander, 2% American Indian and 7% other.

Measures: The primary outcome measures were obesity prevalence, incidence and obesity remission (a reduction in obesity among those already obese). Obesity was defined using a composite indicator based on a body-mass index and a triceps skinfold value greater than (or equal to) age- and sex-specific 85th percentile. Researchers measured change in obesity from baseline (fall 1995) to post-test (spring 1997).

Secondary outcomes: Measures of television viewing, physical activity and dietary intake, as well as other sociodemographic and behavioral variables were obtained from a Food and Activity Survey completed by youth. The survey included an 11-item 'Television and Video' measure (estimate of total television viewing), a 16-item 'Youth Activity Questionnaire' (an estimate of time spent in moderate and vigorous physical activity) and the 'Youth Food Frequency Questionnaire' (assessment of intake of fruits, vegetables and fat).

Analysis: Because schools were randomized, rather than students, a generalized estimated equation method was used to adjust for the individual-level covariates under cluster randomization, with schools nested within experimental conditions. Separate regression estimates were estimated for boys and girls. Analyses were conducted using an intent-to-treat protocol, with participants analyzed in their original randomized condition irrespective of the number of Planet Health sessions attended. To control for missing behavioral data, indicator variables with mean substitution were used. Researchers claim the mean substitution variables did not appear to affect the results of the analysis, but the technique is generally not appropriate.

Outcomes

The prevalence of obesity among girls in the Planet Health intervention schools was significantly reduced, when compared to girls in the control condition. There were no significant differences in obesity measures among boys.

Implementation fidelity: Implementation analysis revealed 87% of classroom teachers and 100% of PE teachers completed the training sessions. Classroom teachers reported they completed an average of 3.5 lessons per year (out of a minimum of 4 per subject). PE teachers reported they completed an average of 8.2 micro-units per year (out of 30).

Baseline equivalence: Prior to randomization, schools were matched and balanced for factors that could affect study outcomes. These included school size, ethnic composition, school food services and physical education curricula. There was a slight difference in median household income between the zip codes associated with intervention and control schools, with intervention schools averaging $36,020 and control schools averaging $34,200.

Baseline data stratified by sex revealed no significant differences among the 1,560 intervention and control students in mean values of age, body-mass index, triceps skinfold or obesity. There were some differences in ethnic composition: higher percentages of African American girls (17% vs 10%) and Hispanic boys (18% vs 12%) were in control schools.

Differential attrition: Post-test data were collected in spring 1997 on 83% of the baseline sample. For girls, data were collected for 82% of control and 81% of intervention students. For boys, data were collected for 86% of control and 83% of intervention students. Among girls, there were no significant baseline differences in rates of follow-up at baseline. For boys who were obese, a lower rate of follow-up was observed in the intervention condition (87% in the intervention vs 94% in the control condition).

Examining completers only, the intervention and control group had similar baseline sociodemographic, anthropometric, diet and physical activity data. Among girls, there was a difference in the prevalence of African American students (16% control vs 10% intervention), which is about the same distribution as at baseline.

Posttest

Obesity: There were significant results in 2 of 3 obesity measures for girls. When compared to girls in the control condition, girls in the Planet Health schools had significantly reduced obesity prevalence and significantly greater obesity remission. None of the three obesity outcomes for boys were significant.

There was some evidence that the program worked better for African-American girls than white or Hispanic girls.

Secondary outcomes: Among girls, 3 of 5 secondary outcome measures achieved significance. Girls in the intervention schools reported reduced dietary intake, increased fruit and vegetable consumption and less television viewing, when compared to girls in the control condition. Among boys, 1 of 5 secondary outcome measures achieved significance. Boys in the intervention schools reported less television viewing when compared to boys in the control condition.

Study 2

Evaluation Methodology

Design: This evaluation assessed the impact of the Planet Health intervention on use of self-induced vomiting and laxative use (purging) and diet pills to control weight in adolescent girls. The study used 480 girls in the intervention and control schools who reported no use of diet pills or purging at baseline. It excluded 21 girls who reported these activities at baseline. It then examined the effects of the intervention on the risk of reporting a new case of purging or diet pill use to control weight at post-test (21 months later).

Sample: The sample included a subgroup of 480 girls, age 10-14 (mean = 11.5 years). This subgroup was taken from a larger sample described in Gortmaker et al. (1999).

Measures: Survey items on dieting, vomiting or taking laxatives (purging), and taking diet pills in the last 30 days to control weight were combined to create a single disordered weight-control variable for analysis. The composite variable was composed of the following survey questions:

• During the past 30 days, did you diet to lose weight or to keep from gaining weight?
• During the past 30 days, did you vomit or take laxatives to lose weight or to keep from gaining weight?
• During the past 30 days, did you take diet pills to lose weight or to keep from gaining weight?

Analysis: Data from 480 girls were used to examine the effects of the Planet Health intervention on the risk of reporting a new case of purging or diet pill use to control weight at post-test (21 months later). A generalized estimating equation (GEE) was used to account for design effects due to clustered sampling by school. The authors estimated the odds of reporting disordered weight-control behavior in the past 30 days at post-test among girls who did not report use of purging or diet pills at baseline. Baseline dieting, age, obesity and ethnicity were used as covariates in multivariate models. Girls who reported purging or using diet pills at baseline were excluded from the analysis.

Outcomes

Baseline equivalence: Compared with girls in the intervention schools, a greater proportion of girls in the control schools were African American (9.1% vs 15.9%) at baseline. Therefore, ethnicity was included in all multivariate models. Control and intervention participants did not differ significantly in the distribution of age or prevalence of overweight or dieting in the past month.

Differential attrition: The cases used were subject to the same differential attrition as in Gortmaker et al. (1999).

Posttest: Girls in intervention schools were less than half as likely to report purging or using diet pills at post-test compared to girls in control schools (OR = 0.41).

Results appeared particularly strong among those who were non-dieters at baseline: among baseline non-dieters, girls in intervention schools were 12 times less likely than girls in control schools to report the use of purging or diet pills to control their weight at post-test.

Long-term: No follow-up data were gathered.