Care2BWell: Worksite Wellness for Child Care

The proposed study will evaluate the efficacy of a 6-month, child care-based intervention to increase workers' MVPA and other health-related behaviors (Healthy Lifestyles) compared to an attention control condition (Healthy Finances). This study will use a two-arm, cluster-randomized controlled trial (RCT) and a sample of 104 child care centers and 416 workers (4 workers/center). The primary outcome will be workers' MVPA; and secondary outcomes include workers' weight, dietary intake, smoking status, sleep habits, and emotional health as well as child care centers' programs, policies and environment around worksite wellness and teachers' interactions with children around physical activity. All primary and secondary outcomes will be assessed at baseline, post-intervention (6 months), and maintenance (18 months).

For this study, 104 child care centers will be recruited from a mix of urban/suburban and rural areas. Potential child care centers in these counties will be identified using an online database of licensed child care facilities maintained by the North Carolina (NC) Division of Child Development and Early Education. State and local community partners will be engaged to help inform centers in targeted counties about the research study. Direct recruitment of centers will employ a variety of strategies (e.g., mail, email, telephone, website, video, in-person contacts). Once a center expresses interest, the center director will be engaged to recruit their staff (child care workers) to participate as a team.

Outcome Measures:

A series of measurements will be collected on participating child care center directors and staff at three time points - baseline, follow-up (6 months), and maintenance (18 months). Primary outcome measures will assess moderate to vigorous physical activity (MVPA) of center staff. Secondary outcomes measures will assess staff members' behavioral health risk factors as well as centers' programs, policies and environment around worksite wellness and teachers' interactions with children around physical activity. These measures will be collected during a one-day, on-site visit to the child care center, and supplemented with physical health assessments completed at the kick-off event. Center visits will be conducted by research assistants who have undergone extensive training and certification on all measurement procedures and are blinded to study-arm assignment. Similar data collection protocols will be used at all three time points - baseline, follow-up and maintenance.

Randomization:

Centers will be randomly assigned (1:1) into the intervention or control arm. Randomization will occur during local kick-off events (each center must attend one event). Each kick-off event will include a short wellness fair in the morning (including stations to collect baseline biomedical health assessments). Lunch will be provided, followed by random assignment into either the intervention or control arm. Timing allows for all of baseline measures to be complete prior to randomization. Centers in both the intervention and control arms will immediately start implementation of their respective programs.

Program Details/Intervention:

Healthy Lifestyles Intervention will kick off with an educational workshop and tailored feedback based on responses to the CHART assessment tool. The educational workshop and personalized health assessment summary will be delivered during the afternoon of the kick-off event in a group format with other participating centers. Following this launch, the intervention will provide 3 campaigns, each of which will include the following components: center support materials, director webinar, staff magazine, and behavioral goal setting/monitoring and weekly personalized feedback. The center support materials, director webinars, and staff magazines will be delivered directly to the center to support the center-based intervention. The behavioral goal setting/monitoring will be delivered directly to participating staff. While campaign structure is constant, topics of each campaign will vary. Campaign 1 focuses on increasing daily physical activity, cutting back on unhealthy snacking, and switching from sugary and diet drinks to water. Campaign 2 focuses on eating a diet rich in fruit and vegetables, whole grains, and lean protein, varying daily physical activity, and increasing strength training. Campaign 3 focuses on maintaining an active lifestyle, managing stress in a healthy way, and improving sleep habits.

Program Details/Control:

Healthy Finances Intervention will kick-off with an educational workshop and personalized health assessment snapshot at the kickoff community event. The educational workshop will be delivered during the afternoon of the kick-off event in a group format with other participating centers randomized to the Healthy Finances arm. The personalized health assessment snapshot will be provided during that workshop, but not reviewed in detail. It is intended to provide only a basic comparison of current behaviors to recommendations. Following this launch, the intervention will provide 3 campaigns, each of which will include the following components: director webinar, staff magazine, and a contest focused on quizzing knowledge gained from the magazines. The director webinars and staff magazines will be delivered directly to the center to support the center-based program. The contest and knowledge quizzes will be delivered directly to participating staff. While campaign structure is constant, topics of each campaign will vary. Campaign 1 focuses on the link between financial health and personal well-being (physical, mental), record-keeping: logging income and expenses, and thoughtful spending: creating and sticking to a spending plan. Campaign 2 focuses on creating and making the most of a personal savings account, and benefits of long-term investing: rules of thumb for picking investments. Campaign 3 focuses on understanding your credit report and credit score, understanding what good credit can do for you, and tips and strategies for improving your credit.

Analysis:

Investigators will compare the difference in mean change in MVPA between baseline and post intervention between Healthy Lifestyle and Healthy Finance arms, controlling for baseline physical activity. The primary analysis will test the hypothesis under the intent-to-treat principle using Generalized Linear Mixed Models (GLMM) that will account for the correlation induced by the clustering of staff within centers. Each GLMM will include a random intercept for centers (β 0) and fixed effects for the baseline value of the primary outcome (β1) and the intervention (β2) to test if the difference in mean change in the primary outcome is zero where β0 is the fixed intercept and e is error.

[Change in Primary Outcome6mo = β0 + β1PrimaryOutcomebaseline + β2Intervention + b0 + e] Including baseline score as a covariate in an analysis of covariance (ANCOVA), in our case a GLLM ANCOVA analysis, is a more powerful test than a group comparison of baseline to post-intervention change. Additionally, ANCOVA is not distorted by regression towards the mean bias, whereas a change analysis is. To further explore intervention effect, investigators will fit GLMMs that: 1) adjust for baseline covariates, considered a priori, to be relevant to change in MVPA; 2) adjust for baseline variables distributed differently between Healthy Lifestyle and Healthy Finance arms; 3) test interaction terms between treatment group and other covariates; and 4) examine completers only. While investigators will make every effort to minimize attrition at the center and participant level; some drop out is expected. Investigators will explore the potential bias from non-ignorable, non-response using two methods: 1) compare respondents and non- respondents to see if they differ systematically on values of non-missing variables; and 2) assess whether data are missing completely at random (MCAR), or missing at random (MAR) or missing not at random (MNAR) - non-ignorable. If missing data are ignorable (MAR and/or MCAR), investigators will consider imputing missing endpoint data using multiple imputation techniques. However, it is often hard to determine if missing data are ignorable; therefore, investigators will consider selection models (MNAR models) to assess if data are missing not at random and estimate the effect of the interventions after removing selection bias. Investigators will also consider using pattern mixture models to assess the sensitivity of our results to various assumptions of missing data patterns. Investigators' previous work in child care centers and worksites demonstrate good retention.