Cambridge, MA, is a dense city of 101,355 (20) north of Boston. At baseline (preintervention) in the 2003–2004 school year, 6,444 children were enrolled in 12 kindergarten-eighth (K-8th) grade schools and one high school in the Cambridge Public Schools (CPS). Despite its reputation as a wealthy college town, 64% of the students were nonwhite (38% African-American, 15% Hispanic, 10% Asian, and 1% other) and 41% were low-income. Almost one-third (33%) of children reported speaking a language other than English at home, and 50 countries of origin were reported—Brazil, Haiti, and Central American countries among the most common. Over the course of the study, enrollment in CPS declined to 5,599 children in 2006–2007 school year due to the transience of the population and trends toward transfer to private and suburban schools. Average daily attendance remained relatively constant at ∼94% (21). According to the 2005 Middle School Health Survey (adapted from the Youth Risk Behavior Survey (22)), among 6th–8th grade children, 40.6% reported eating 5+ fruits and/or vegetables in the past 24 h; 64.7% reported watching ≤2 h of TV daily; and 40.5% reported meeting moderate and/or vigorous physical activity benchmarks (CPS, personal communication).
Community-based participatory research (CBPR) approach
The HLCK study is the result of 10 years of CBPR in Cambridge designed to develop and mobilize environmental and structural interventions within the community and school to promote healthy weight. The CBPR approach engaged community members in all aspects of the research process from research questions to design and implementation of the study and to analysis and dissemination (18). Our study involved a collaborative effort between members of The Healthy Children Task Force (Task Force) in Cambridge, including CPS, the Institute for Community Health, and the Cambridge Public Health Department. The Task Force is a multidisciplinary coalition of elected officials, educators, health care, and public health professionals, researchers, and parents that has provided a forum for collaboratively addressing children's health issues since 1990. In 2000, the Task Force prioritized healthy eating and active living and identified increasing “healthy weight” (BMI ≥5th and <85th percentile for age and gender (23)) and fitness among K-8th grade children as community goals. Task Force partners, both individually and through the institutions they represented, became involved in elements of both the intervention and the evaluation.
The HLCK intervention developed in four phases: formative, developmental/pilot, implementation, and sustainability. These elements served as building blocks in one community's effort to address childhood obesity and promote healthy eating and active living.
Formative phase (1999–2001). Several steps initiated obesity prevention work. First, in 1999, CPS collaborated with research partners at Institute for Community Health to (i) develop a computerized data system to record annual height, weight, and fitness test score data and monitor, in aggregate, BMI percentiles and fitness status of K-8 CPS children, (ii) train PE teachers and school nurses in standardized anthropometry, and (iii) purchase standardized equipment for each school.
Second, Task Force partners created 5-2-1 guidelines based on national goals and emergent research to promote healthy weight. The 5-2-1 guidelines promoted decreasing energy intake by promoting eating five or more servings of low-energy fruits and vegetables daily (24); increasing energy expenditure by limiting inactive or sedentary time to 2 h or less of TV or screen time daily (25); and increasing moderate and vigorous physical activity to at least 60 min of age-appropriate physical activity on all or most days of the week (26). The 5-2-1 slogan served as an awareness campaign and provided goals for community-level interventions.
Third, formative research including community forums and parent input clearly identified that local families were interested in improvements to school meals and PE. Subsequently, Task Force partners were mobilized to seek grants, garner resources, and pilot healthy weight interventions.
Intervention development and pilot-testing phase (2001–2004). BMI data showed that CPS children had higher rates of overweight and obesity (BMI ≥85th percentile) than national rates (27) and several years of trend data showed an ∼0.5% annual increase of high BMI occurred from 2000 (37.0%) to 2004 (39.1%) among K-8th grade children, suggesting a worsening in children's health.
In 2001, Task Force partners pilot-tested the use of individualized “BMI and fitness report cards” (BMI and fitness reports) on parents' awareness of their children's weight and fitness status and their intentions to take follow-up action (28). Based on positive results, CPS implemented BMI and fitness reports system-wide for grades K-8. Supports (follow-up phone calls, referrals for weight management) for parents of overweight and obese children were implemented by school nurses from Cambridge Public Health Department. Over time, family feedback led to adjustments in layout and language to improve the readability of the BMI and fitness reports.
Next, with additional grant dollars, a pilot-program in four elementary schools tested the feasibility and efficacy of using school-yard gardens, cafeteria taste-tests, and family education to promote fruit and vegetables. PE enhancement grants offered professional development for PE teachers and new gymnasium equipment in all schools.
Implementation phase (2005–2007). In 2005, HLCK was launched, representing the culmination of years of collaborative efforts and several successful grants. The original partnership expanded to include CitySprouts, a gardening organization, Cambridge Department of Human Service Programs, Cambridge Green Streets Initiative, and the Federation of Massachusetts Farmers' Markets. Funding came from the Department of Education Carol M. White Physical Education Program, USDA Community Food Projects, Blue Cross Blue Shield of Massachusetts, and the Massachusetts Department of Public Health.
The 3-year, multicomponent HLCK intervention continued to be guided by CBPR principles. In keeping with observations that successful interventions were more likely to use a conceptual frame and a comprehensive environmental and policy intervention approach (10) HLCK adapted the socioecological model (29) to target community, school, family, and individuals. Figure 1 summarizes key components of HLCK implementation.
At the community level, implementation strategies were designed to provide policy support for healthy living choices such as a city council endorsement of the “5-2-1″ guidelines and passage of a local food preference policy; to provide opportunities for community advocacy such as the 5-2-1 coalition and youth sports commission; to provide after-school providers training on implementing the policies; and to raise community awareness of the many resources available in the city to promote healthy eating and active living through a poster campaign, newsletters, 5-2-1 mini-grants, and directories of physical activities distributed to all school children.
At the school level, PE and food service policies, systems, and programs were implemented at all 12 K-8 schools similarly to improve access to appealing, appropriate physical activity opportunities, and healthy food choices for all children; school stakeholders were trained to implement new guidelines and policies; and PE programs such as Project Adventure and ballroom dancing, and innovative food service projects such as new recipe and menu development and cafeteria taste-tests were developed to promote 5-2-1. School-yard garden programs were expanded to increase student awareness of and appreciation for locally-grown produce.
A school wellness policy http:www.cpsd.uscpsdirschool_policies.cfm was developed as required by the WIC Reauthorization Act of 2004. School nutrition guidelines included restrictions on items sold in vending machines (30); limited access to a la carte foods; system-wide substitution of lower-sugar (<6 g sugar) (31) and/or higher-fiber (>2 g fiber) cereals, whole grain breads (50–100% whole grain), and low-fat yogurt without artificial colors, and products with trans fat were phased out. Principles to promote 5-2-1 were included for PE, recess, and snacks in the policy.
At the individual- and family level, strategies and policies were designed to increase the awareness of children and their families of each student's health risk due to their BMI or fitness test scores, and to provide skills and resources for addressing individual and family health risks and lifestyle choices through school-based family nights. Annual BMI and fitness reports noted results were not diagnostic and referred parents to pediatricians for follow-up. Fitness report distribution was followed by “Fit Together” family event nights, open to all families but specifically targeting families of obese children. In addition, receptive families were offered subsidized weight management counseling at a local family-oriented obesity management agency.
Weight status. Individual weight status was assessed by BMI, calculated from height and weight measurements collected annually each spring by CPS PE teachers and school nurses who were trained as professionals each year with a standard protocol (32). As noted, all schools used the same equipment. Height was measured to the nearest 0.25 inch with a wall-mounted stadiometer (Seca 216 Accu-Hite, Snoqualmie, WA). Weight was measured to the nearest 0.2 lb with an electronic scale (Seca 216 Bellisima-digital, Snoqualmie, WA) in indoor clothing without shoes. Because CPS sent BMI and fitness screening results home to families and wanted to ensure accurate information, all student data was checked for outliers during data entry, and high and low data (BMI <5th or ≥95th percentile) were reviewed by school nurses familiar with the children. BMI z-scores and percentiles based on age and gender were calculated for each student from CDC growth charts (23). As in previous studies, BMI z-scores ≤−4 and BMI z-scores ≥5 were excluded from the analysis (33). Children were classified as: obese (BMI ≥95th percentile), overweight (BMI ≥85th and <95th percentile), healthy weight (BMI ≥5th and <85th percentile), and underweight (BMI <5th percentile) (34).
Fitness. Fitness was assessed by age- and gender-adjusted scores on five fitness tests completed annually in PE each spring: endurance cardiovascular test; abdominal strength test; flexibility test; upper body strength test; and an agility test. PE teachers were trained annually with testing and scoring protocols adapted from Amateur Athletic Union (35) and Fitnessgram (Cooper Institute, Dallas, TX (36)). Children's proficiency status (Participant, Attainment, Outstanding) was assigned for each test according to Amateur Athletic Union and Cooper Institute guidelines as described elsewhere (37). For each fitness test, children were considered “passing” if they achieved “Attainment” or “Outstanding”. The mean number of fitness tests passed was calculated for each student (from 0 to 5 tests passed). Overall fitness was calculated where “passing” was defined as having “passed” all five tests and “not passing” was defined as having failed any one of the five tests. “Passing” or “not passing” the endurance cardiovascular “shuttle run” test score was also used independently because it has been correlated with obesity (37).
Personal measures. Gender, grade, age, race/ethnicity (black, Hispanic, Asian, white), and income status based on eligibility for free or reduced price school meals (free meals) under the National School Lunch Program were extracted from the school administration record system. Free meal eligibility was used as an indirect measure of family income status and was coded as a binary variable, lower income, or higher income (eligible or ineligible for free meals, respectively) (38).
Change in weight and fitness status was assessed using continuous and categorical classifications as described above. For analysis of continuous data, paired t-tests were utilized to determine changes in mean BMI z-score and the mean number of fitness tests passed between the baseline data collection point (2004) to the follow-up data collection point (2007) in the overall cohort and in samples stratified by gender, race/ethnicity, and socioeconomic status. For categorical data, McNemar tests were utilized to determine statistical significance of changes in the proportion of children in each BMI percentile category, and each dichotomous fitness category from baseline to follow-up both in the overall sample and in samples stratified by gender, race/ethnicity, and socioeconomic status. Percent change was calculated for descriptive purposes. P values <0.05 were considered statistically significant.
All analyses used SAS (version 9.1; SAS Institute, Cary, NC).