The Childhood Obesity—Prevention of Diabetes Through Changed Eating Patterns Study (referred to as The COPE study) is a prospective nonrandomized controlled trail conducted at multicomponent lifestyle camps in Denmark. The COPE study was designed to investigate the effect of a higher protein diet on weight loss and health-related outcomes in children from 7 to 14 years of age with overweight and obesity, and the study design is presented in Fig. 1.

The figure is created with BioRender.com, and a publication license is granted

Study design of The COPE study.

The multicomponent lifestyle camps are managed by Julemærkefonden, a non-governmental organization that manage five multicomponent lifestyle camps in Denmark. Julemærkefonden and all five camps are financed by funding from investors and private companies, and children attend these camps free of charge. Children are referred to attend camp for 10 weeks by their general practitioner if they struggle with overweight or obesity, loneliness, unhappiness, or social or family-related problems. All camps are multicomponent, focusing on improving health and quality of life through social and physical activities, healthy meals, and daily exercise. At camp, children attend school for approximately three hours daily. During camp, social workers with experience in childcare support and motivate all children. Children go home to their families on weekends, and the families are invited to four family education days. After the 10-week camp, the children are supported to explore opportunities in their respective municipalities, which may include programs for obesity treatment and/or participation in sports activities. The child and their family revisit the camp for a concluding follow-up session, taking place approximately one month after their last day at camp. This provides an opportunity to engage in discussions with camp staff and peers, sharing their experiences and progress.

Julemærkefonden is responsible for allocating children to the respective camps, usually placing them in the camp nearest to their homes. The study was initiated in collaboration with two of the camp sites (Julemærkehjem Hobro and Julemærkehjem Fjordmark). All children assigned camp Hobro from October 2020, and camp Fjordmark from May 2021 until March 2022 were invited to participate, and parents/guardians received information about the present study before starting camp. Camp staff invited children and parents/guardians to an individual information meeting four to eight weeks before starting camp, and if they agreed to participate in the study, parents/guardians were asked to sign a written consent for their child to participate, and one parent/guardian were asked to sign a written consent for themselves to participate with their child. Children were excluded from the study if they were diagnosed with an eating disorder or a disease requiring a special diet. Children or parent/guardians who participated in another clinical trial or did not understand or were unwilling/unable to comply with the study protocol were also excluded.

All camps are managed according to the same rules and regulations and comply to the official Danish guidelines concerning diets and physical activity [34, 35]. According to camp dietary policy, children are served six meals daily, and all meals are provided with instructions on portion sizes. The daily energy intake ranges from 1200 to 1800 kilocalories depending on age. For example, children below 10 years of age may be served less rye bread, potatoes, or rice. At breakfast, children are allowed to drink two cups of milk and eat three half slices of rye bread with various toppings and greens/fruit. One half slice of rye bread can be substituted with a bowl of cereal. At lunch, children are served a warm prepared meal with one piece of meat, salad/vegetables/legumes, and potatoes, rice, or pasta. At dinner, children younger than 10 years of age are allowed to eat three half slices of rye bread, while children above 10 years of age are allowed to eat four half slices of rye bread with different toppings and greens/vegetables. Additionally, children are served three in-between meals: half a piece of fruit before lunch, and one crispbread/bun with high fiber content plus greens/fruit as an afternoon and evening snack. Water is served ad libitum. According to the camp dietary policy, the targeted distribution of macronutrients is 10–15 energy percent (E%) protein per day, 45–60E% carbohydrate per day, and 25–40E% fat per day in accordance with official Danish dietary guidelines [34]. Additionally, certain foods, such as candy, sweets, and soda, are restricted.

Due to ethical considerations and the daily routines at camp (Supplementary S1), it was not feasible to randomize children within camps. Both camps comply with the same dietary policy, rules, and routines concerning meals and physical activity. Since camp Hobro have the capacity to recruit 250 children per year compared to 150 children per year at camp Fjordmark, camp Fjordmark was assigned as the active control group, and camp Hobro was assigned as the intervention group. In the control group, no changes were made. In the intervention group, the aim was to replace carbohydrates-rich foods at breakfast and two in-between-meals with naturally protein-containing foods (e.g., dairy products, nuts, egg, meat-products) to increase the amount of protein from ~ 10–15E% to ~ 25E% per day with minimal changes in total caloric intake. Meal changes were planned in collaboration with the kitchen staff to increase compliance, and the kitchen staff were provided with a weekly meal schedule and instructions on how to alternate between different protein-containing foods to ensure variety and accommodate special needs. Kitchen and camp staff in the intervention group motivated all children to eat the protein-containing foods. No changes in fruit/vegetables/salads were made, and kitchen staff were instructed to serve fruit/vegetables/salads as usual.

Halfway through the camp, children and parents/guardians in both groups were invited to a family education day. Both groups participated in social and physical activities and were introduced to the official Danish dietary guidelines [36]. Furthermore, the intervention group had a 30-min educational class focusing on protein in relation to weight loss, satiety, and physical activity. They learned how to identify protein-containing foods well-known from the supermarket, received a pamphlet containing suggestions for protein-rich breakfast and in-between meals, and were encouraged to continue eating a higher protein diet after camp. The remaining family education days were carried out by camp staff as usual.

Once a month for the first six months after camp, study staff contacted all families through text messages to ask if they would like a follow-up phone call. The aim was to motivate the intervention group to continuously consume a higher protein diet and support health-promoting behaviors in the control group. On average, study staff were in contact with the families two times within that period, and 22% of the families never received follow-up phone calls; therefore, no further analyses of this intervention strategy were performed.

Body weight (kg), body fat (%), and skeletal muscle mass (kg) were measured in light clothing without shoes using a Bioelectric impedance (InBody model 270, Hopkins Medical Products, Grand Rapids, MI, USA). Height (m) was measured using a fixed wall measuring tape. BMI-SDS was calculated using World Health Organization AnthroPlus software and considered the primary outcome. A BMI-SDS > 1SD was defined as overweight, and a BMI-SDS > 2SD was defined as obesity [37]. Furthermore, blood pressure was measured with the right arm placed at heart level using an automatic non-invasive blood pressure monitor (Omron M3, Kyoto, Japan). Camp staff were responsible for measuring anthropometry and blood pressure in all children at baseline, 10 and 52 weeks.

Blood samples were optional for all children. Educated bio-analysts at Aarhus University Hospital were responsible for collecting blood samples from a subsample of the children at baseline, 10 and 52 weeks.

Children and parents/guardians answered several questionnaires, and all questionnaires were delivered electronically to the participating parent/guardian using REDcap.org database located at Aarhus University.

Background characteristics were collected with a parent-reported questionnaire, including child sex and age, parental education, household income, diseases in the family, participation in physical activity, etc. In accordance with national physical activity guidelines [38], the authors formulated three questions to evaluate physical activity behavior among the children, specifically focusing on high intensity, moderate intensity, and the settings of physical activities. Quality of life was measured using the validated Danish version of the Pediatric Quality of Life Inventory 4.0 questionnaire (PedsQL 4.0) [39]. The Children’s Eating Habits Questionnaire-FFQ (CEHQ-FFQ) [40] was translated into Danish by the authors, and a few foods were added to investigate eating habits. Eating behavior was measured using a Danish version of the Child Eating Behavior Questionnaire (CEBQ) [41]. The prevalence of subjective binge eating and loss-of-control eating was measured using two questions from the Eating Disorder examination questionnaire (EDE-Q 6.0) [42, 43].

Child and parents/guardians answered background characteristics at baseline. Physical activity behavior among children was assessed at baseline and 52 weeks, but not at 10 weeks, as all children were engaged in physical activity in line with national recommendations during camp. All other questionnaires were answered at baseline, 10 and 52 weeks.

Separate papers will be published presenting changes in quality of life, physical activity behavior, eating habits, eating behavior and binge eating/loss-of-control eating.

As illustrated in Fig. 1, the COVID-19 pandemic forced a lockdown from December 2020 to February 2021. All children affected by the COVID-19 lockdown were sent home for five weeks with no control of dietary intake and physical activity, which is why they were excluded. Furthermore, due to the COVID-19 lockdown, meal changes in the intervention group were postponed from April to May 2021, and few children starting camp in March/April 2021 were therefore served a standard protein-diet for 4–6 weeks and a higher-protein diet for 4–6 weeks. These children were excluded whenever investigating differences in changes between groups.

The power calculation was based on a previous study investigating the effect of a higher protein intake (21E% vs. 32E%) on childhood obesity [30]. In this study, 22 and 24 children were allocated to the standard and high protein diet, respectively. Both groups had a reduction in BMI z-score during the 13-week intervention (from 2.48 ± 0.06 BMI-z to 2.10 ± 0.10 BMI-Z in the high protein group, and from 2.51 ± 0.05 BMI-Z to 2.40 ± 0.10 BMI-Z in the standard protein group), with a greater reduction in the high protein group (p = 0.03) [30]. The number of participants needed to detect a between-group difference of − 0.27 SD with a power of 80% and a significance level of 5% was calculated to be ~ 55 children in each group.

Continuous data are presented as mean ± standard deviation (SD) for parametric data and median [inter quartile range (IQR)] for non-parametric data. Categorical data are presented as absolute numbers and percentage (%). Differences between groups at baseline were tested using linear regression analysis. As this study investigate a rather homogenous group, e.g., with a lower socioeconomic position compared to children with overweight and obesity not attending camp [44], we assume that missing data are randomly distributed across variables. Therefore, unadjusted mixed effects models were performed to investigate differences in change between the groups. All mixed effect models accounted for repeated measures over time at the individual level, thus controlling for differences in children’s weight based on their own baseline measurements. Group level was not included as a random effect, but as an interaction term to evaluate potential variations in the intervention effect across groups.

QQ-plot of the residuals was used to check for normal distribution errors. Linearity and identical distribution errors were assessed by checking the residual plots versus the predicted values. If necessary, log-transformation of the data was performed to reduce the skewness of data to meet the assumptions of linearity and equal distribution errors. Sensitivity analyses including only participants with complete data were performed to investigate the robustness of the primary outcome (BMI-SDS). Additionally, intention-to-treat analyses were performed on the complete sample, potentially reflecting the real effect of the camps more accurately.

All statistical analyses were performed using Stata/MP 17.0 (StataCorp LLC, USA) with a p < 0.05 considered statistical significant.