Video Abstract

Video Abstract

Close modal
OBJECTIVES

Our aim was to compare the effect of 2 treatment models on attendance and child weight status: a less intense guided self-help (GSH) program delivered in the primary care setting versus traditional family-based behavioral treatment (FBT) delivered in an academic center.

METHODS

We conducted a randomized clinical trial among 164 children between 5 and 13 years old with a BMI ≥85th percentile and their parents. The intervention group (GSH) received 14 individual sessions over 6 months, with 5.3 hours of treatment. The control group (FBT) received 20 group-based sessions over 6 months, with 20 hours of treatment. Main outcomes included proportion of sessions families attended and change in child BMI z-score (BMIz), percentage from the 95th BMI percentile, difference from the 95th BMI percentile at the end of treatment, and 6-month follow-up.

RESULTS

Mean age of children was 9.6 years, BMI z-score 2.1, 49% female, and >90% Latino. The odds of attending GSH compared to FBT was 2.2 (P < 0.01). Those assigned to GSH had a 67% reduced risk of attrition (hazard ratio = 0.33, 95% confidence interval 0.22–0.50, P < .001). Intent-to-treat analysis showed no between-group differences in change in BMIz and percentage from the 95th BMI percentile over time. Combined, there was a significant reduction in BMIz from baseline to posttreatment (β = −0.07 (0.01), P < .01, d: 0.60) and a slight increase from posttreatment to follow-up (β = 0.007 (0.13), P = .56).

CONCLUSIONS

This study provides support for a novel, less intense GSH model of obesity treatment, which can be implemented in the primary care setting. Future studies should examine effective approaches to dissemination and implementation of GSH in different settings to increase access to treatment.

What’s Known on the Subject:

Family-based behavioral treatment (FBT) is effective for pediatric weight loss but is intensive and difficult to access. Guided self-help (GSH) programs provide similar treatment components, are effective for many conditions, and can be delivered in the primary care office.

What This Study Adds:

GSH programs for pediatric obesity may increase access to treatment of those who are unable to attend FBT programs at tertiary care centers. Furthermore, children attending GSH made comparable decreases in weight status as those who attended FBT.

A third of children in the United States have overweight or obesity.1  Primary care providers (PCPs) have been identified as key players in the treatment and prevention of childhood obesity.24  However, the United States Preventive Services Task Force concluded that moderate intensity counseling may not be deliverable by PCPs during a 20-minute well-child visit, and that these children should be referred to intensive counseling or behavioral interventions to assist with weight control.5 

The preferred program includes family-based, multicomponent behavioral therapy using an integrated care model and multidisciplinary team with >25 hours of contact time over 6 to 12 months.6  Family-based behavioral treatment (FBT) is 1 proven approach, with programs often consisting of 20 group-based sessions (separate groups for parents and children), each lasting 60 minutes, spanning a 6-month period.79  Separate behavioral coaching sessions for each parent–child dyad can also be included to provide individualized feedback, bringing the total contact time to 26 hours. Basic nutrition and physical activity topics are discussed, as well as behavioral strategies to make healthy lifestyle changes. Unfortunately, FBT and other multicomponent programs are intensive and attrition rates can range from 27% to 73%.10  Scheduling issues, transportation difficulties, distance needed to travel, and competing work and family responsibilities (including child care for nonparticipating children) are often cited as reasons for poor attendance.11,12  Minority populations and those in underserved areas also have high rates of attrition.1315  Since many of these FBT programs are located at academic centers, availability and access may be limited.

In response to these challenges, a guided self-help (GSH) model of obesity management was developed to provide shorter treatment sessions and increased flexibility in scheduling.16  The GSH model provides a structured program for obesity management that includes standard behavioral strategies (ie, problem-solving, stimulus control, self-monitoring, and goal-setting) that are included in traditional FBT. Notably, GSH for pediatric obesity has been effective at decreasing BMI z-score (BMIz).16,17  Instead of providing a traditional group-based structure, GSH can be delivered in 14 individually scheduled visits with a health coach over 6 months. Although the first session is 1 hour long to review expectations and nutrition goals, the remaining sessions are 20 minutes each, resulting in a total of 5.3 hours of treatment. This program is based on self-regulation theory which purports that continued feedback from social and environmental interactions on goal-directed behaviors will allow patients to make adjustments over time and eventually achieve their goals.18,19  Even though the program is self-directed, meeting with a health coach promotes treatment compliance and provides accountability for the participant so they can meet their healthy eating and activity goals. One review suggests that these types of sessions may be effective at providing obesity treatment and could be an alternative to FBT.20  However, more studies are need.

The goal of the GSH obesity treatment in the doctor’s office study was to test the efficacy of delivering a GSH model of pediatric obesity treatment in the primary care setting compared with standard of care (FBT delivered in an academic tertiary care setting). The primary aim of this study was to compare the effect of both treatments on attendance and change in child weight status. We hypothesized that there would be greater attendance in GSH. We also hypothesized that there would be similar changes in weight status in both groups among those who attended. However, because more families would attend GSH, there could be greater changes in weight status in the intent-to-treat analysis.

This study was a 2-arm parallel assignment, clinical trial conducted from December 2016 to December 2019. The trial was approved by the University of California San Diego Human Research Protections Program. A participating parent provided written consent and the child provided written assent. Details regarding study protocol have been previously published.21 

The study was conducted in San Diego County, California, at 2 Children’s Primary Care Medical Group of San Diego sites and the University of California San Diego Center for Healthy Eating and Activity Research. Children and parents were recruited from the 2 Children’s Primary Care Medical Group of San Diego sites. These sites served communities with similar characteristics: largely Latino (48.9%–58.2%), median income ranging from $49 787 to $65 364, 10.4% to 18.3% living below the poverty level, and 73.3% to 81.9% of parents with a high school degree or less. The prevalence of childhood obesity in these neighborhoods ranged from 37.9% to 38.9% at the time of the study.22 

PCPs identified children aged 5 and 13 years with overweight or obesity using Centers for Disease Control and Prevention BMI growth charts during their routine well-child visits. PCPs followed their usual practice regarding weight management, and then asked if the parent and child would be interested in participating in a weight control intervention. Parents who expressed interest were referred to the study coordinator via the electronic health record. If the coordinator was in the office, the PCP would walk the family to her office. Otherwise, the study coordinator would contact the family, screen for eligibility, and go through the consent process via phone. Families were included in the study if they met the following criteria:

  1. child aged 5 to 13 years with a BMI ≥85th percentile for age and sex;

  2. parent responsible for food preparation willing to participate;

  3. parent with an English or Spanish reading level of fifth grade or higher;23 

  4. family willing to complete all assessment visits; and

  5. family not moving out of the San Diego area within the time frame of the study.

Children who were taking medications that impacted weight or had limitations in physical activity were excluded from the study. When families had more than 1 child who met eligibility criteria, a random numbers table was used to select the index child for data collection. Once written parent consent and child assent were obtained, families were scheduled to complete online or paper surveys and obtain a height and weight in the office.

Parent–child dyads were randomized at a 1-to-1 ratio within each clinic to the intervention (GSH at the primary care site) or control group (FBT at the academic center). Randomization occurred at the level of the individual to eliminate PCP effects on treatment, and children were stratified by sex to ensure equal distribution between groups. Parent–child dyads were randomly assigned to a group after they completed all baseline assessments. Only the study statistician and data coordinator, both of whom had no contact with participants, were aware of the allocation sequence and assigned families to each group.

Intervention details have been previously published.21  Treatment spanned a 6-month time frame with 6-month follow-up. Assessments occurred at baseline, posttreatment (month 6), and 6-month follow-up (month 12). All program materials were available in Spanish and English. Additional cultural adaptations were made to both programs and included: discussion of Latino foods and recipes, inclusion of other family members (mainly grandmothers) who participated in child care or feeding the child, and use of bilingual, bicultural health coaches.

Parent–child dyads assigned to GSH received 14 individual treatment sessions delivered by a health coach in Spanish or English, based on the parent’s preference. The first 4 sessions occurred weekly. The next 10 sessions occurred every other week to span a total time frame of 6 months. The first session was 1 hour long and reviewed program expectations and dietary guidelines with the participating parent and child. Subsequent sessions were 20 minutes long and continued to include both parent and child together. All participants received a separate parent and child manual that covered core nutrition concepts, physical activity behaviors, and basic behavioral strategies that are taught in FBT. Parents and children were expected to review 1 chapter per week and implement the changes at home. During the 20-minute session, the health coach answered any questions about dietary and physical activity behaviors and problem-solved with them about any issues they had implementing the program at home.

Parent–child dyads assigned to FBT attended 20 sessions over 6 months. Each session was 60 minutes in length and involved separate parent and child groups so that parenting strategies could be discussed with parents alone. Both groups discussed similar nutrition, physical activity, and behavioral strategies for developing healthy behaviors. However, material for the child group was tailored to be developmentally appropriate for age. Details regarding session topics and behavioral recommendations have been previously published.21  Spanish- and English-language parent groups were run concurrently so parents could attend group in their preferred language. Separate behavioral coaching sessions for each family was not provided because of limitations in staffing.

The primary outcomes of this study included program attendance (ie, proportion of sessions each family attended) and change in child BMIz, percentage from the 95th BMI percentile (%BMIp95), and difference from the 95th BMI percentile (ΔBMIp95) at the end of treatment (month 6) and 6-month follow-up (month 12).24,25  Attendance and anthropometric data were tracked by health coaches and study coordinators at each session. Anthropometric data were obtained using a portable Schorr height board (Schorr Inc, Olney, MD) and a Tanita Digital scale (model WB–110A). Height was recorded to the nearest 0.1 cm and the average of 2 values used for analysis. Weight was recorded to the nearest 0.1 kg and the average of 2 values used for analysis. BMI was calculated as kg/m2, and BMI percentile and BMIz were determined using Centers for Disease Control and Prevention growth charts.26,27  The %BMIp95 was calculated using the following formula: BMI ÷ BMI of the 95th percentile × 100. The ΔBMIp95 was calculated as follows: BMI – BMI of the 95th percentile. These measures were included as primary outcome measures because they have been shown to more accurately measure change in adiposity among children with high BMI percentiles (>97th percentile).24,25 

Secondary outcome included the proportion of families who dropped out, or “attrition.” This was defined as missing 3 or more consecutive sessions. Satisfaction surveys were administered at the end of treatment using 5-point Likert scales (1 = strongly disagree–5 = strongly agree). Patient demographic data (provided by the parent) included child and parent age, sex, ethnicity, parent education, income, and food security level.

Sample size was based on changes in BMIz. The effect size chosen was supported by observations in the GSH pilot study, which found a decrease in BMIz d = |1.71–1.50|/0.30 = 0.667.16  Using this information, empirical power estimates were calculated by generating multivariate random samples that were matched to the expected BMIz for each condition and variability over time. With a median between-group effect of −0.39 (s.d. = 0.13) across 1000 data sets, the planned design of recruiting 100 per group (total n = 200) would provide >0.82 power for detecting this effect, allowing for up to 20% loss to follow-up.28 

With a sample of 200 to support between-group comparisons of BMIz, we estimated we would have power >0.95 to detect a difference in the proportion of sessions attended in a generalized linear model (binomial) with 2-sided α < 0.05 if we expected >85% attendance in GSH and 60% attendance in FBT. All analyses of primary outcomes used an intention-to-treat sample including all participants who provided both baseline and posttreatment assessments.

Given the nature of the study, participants and health coaches could not be blinded to group assignment. However, participants were blinded to the specific details of the research hypothesis.

Data were analyzed using intent-to-treat models and R statistical program.29  To assess differences in the proportion of sessions parent–child dyads attended in GSH and FBT, we used a generalized linear mixed effect model (binomial family) with random intercepts to account for repeated assessments. We also explored attendance patterns using a cox proportional hazard regression model to determine patterns in “attrition,” defined as missing 3 or more consecutive sessions. To assess between-group differences in BMIz, %BMIp95, and ΔBMIp95 at posttreatment (6-month time point) and 6-month follow-up (12-month time point), we used linear mixed effects analysis adjusting for baseline values. This model makes use of all available information from each participant when estimating model parameters to address issues of missingness.30  All models included baseline weight status and fixed effects of age, ethnicity, and sex. We included site as a main effect in the model to account for any similarities of observations because of receiving care in the same clinic. All results were presented with 95% confidence intervals (CI), standard errors, or P values (2-sided with a significance of P < .05). Final estimates were taken from models across 100 multiple imputed data sets.

A total of 716 patients were referred to the program from December 2016 to August 2018; 230 (32.1%) were eligible and completed the consent process (Fig 1). Only 164 (22.9%) parent–child dyads completed baseline assessments and were randomized to participate, 82 in the intervention group (GSH in the primary care setting) and 82 in the control group (FBT at the academic center). Demographic characteristics between both groups were similar, with >90% of participants reporting Latino background and mean BMIz of 2.1 (Table 1). By the end of treatment, 3 families in the control group and 2 families in the intervention group had withdrawn from the study. Therefore, 79 (96%) and 80 (98%) families from the control and intervention groups respectively were included in the posttreatment (6-month time point) analysis.

FIGURE 1

Consolidated standards of reporting trials diagram of participant flow.

FIGURE 1

Consolidated standards of reporting trials diagram of participant flow.

Close modal
TABLE 1

Sample Characteristics

FBT (n = 79)GSH (n = 80)
Chula VistaEscondidoChula VistaEscondido
Parent sex, female, n (%) 33 (94) 41 (93) 32 (97) 43 (92) 
Parent ethnicity, n (%)     
 Latino 32 (91) 34 (77) 31 (94) 38 (81) 
 Non-Latino White 3 (9) 7 (16) 2 (6) 8 (17) 
 Other 0 (0) 3 (7) 0 (0) 1 (2) 
Parent education, college or higher, n (%) 8 (23) 6 (14%) 5 (15%) 8 (17%) 
Parent income, <185% of federal poverty level, n (%) 24 (69) 34 (77%) 27 (82%) 26 (56%) 
Food insecurity, n (%)     
 High food security 21 (60) 18 (41%) 16 (49%) 17 (36%) 
 Low food security 12 (32) 24 (55%) 15 (46%) 24 (51%) 
 Very low food security 1 (3) 1 (2%) 2 (6%) 5 (11%) 
Parent age, y, mean (SD)b 42.4 (6.6) 37.7 (11.5) 38.5 (7.8) 37.9 (7.6) 
Parent BMI, mean (SD)b 33.3 (6.4) 31.7 (6.2) 36.2 (6.9) 32.3 (7.0) 
Child sex, female, n (%) 19 (54) 22 (50) 16 (48) 21 (45) 
Child ethnicity, n (%)     
 Latino 33 (94) 33 (75) 30 (91) 38 (81) 
 Non-Latino White 1 (3) 4 (9) 2 (6) 5 (11) 
 Other 1 (3) 7 (16) 1 (3) 4 (9) 
Child age, y, mean (SD) 10.1 (2.2) 9.2 (2.2) 10.0 (1.8) 9.4 (2.2) 
Child BMI, mean (SD)b 28.3 (5.3) 25.8 (4.0) 28.2 (4.1) 25.6 (3.8) 
 BMIz, mean (SD) 2.2 (0.3) 2.1 (0.4) 2.2 (0.3) 2.1 (0.3) 
 BMI percentile, mean (SD) 98.0 (1.8) 97.3 (2.5) 98.0 (2.0) 97.6 (1.8) 
Child with overweight,cn (%) 2 (6) 8 (18) 4 (12) 5 (11) 
Child with obesity,cn (%) 33 (94) 36 (82) 29 (88) 42 (89) 
FBT (n = 79)GSH (n = 80)
Chula VistaEscondidoChula VistaEscondido
Parent sex, female, n (%) 33 (94) 41 (93) 32 (97) 43 (92) 
Parent ethnicity, n (%)     
 Latino 32 (91) 34 (77) 31 (94) 38 (81) 
 Non-Latino White 3 (9) 7 (16) 2 (6) 8 (17) 
 Other 0 (0) 3 (7) 0 (0) 1 (2) 
Parent education, college or higher, n (%) 8 (23) 6 (14%) 5 (15%) 8 (17%) 
Parent income, <185% of federal poverty level, n (%) 24 (69) 34 (77%) 27 (82%) 26 (56%) 
Food insecurity, n (%)     
 High food security 21 (60) 18 (41%) 16 (49%) 17 (36%) 
 Low food security 12 (32) 24 (55%) 15 (46%) 24 (51%) 
 Very low food security 1 (3) 1 (2%) 2 (6%) 5 (11%) 
Parent age, y, mean (SD)b 42.4 (6.6) 37.7 (11.5) 38.5 (7.8) 37.9 (7.6) 
Parent BMI, mean (SD)b 33.3 (6.4) 31.7 (6.2) 36.2 (6.9) 32.3 (7.0) 
Child sex, female, n (%) 19 (54) 22 (50) 16 (48) 21 (45) 
Child ethnicity, n (%)     
 Latino 33 (94) 33 (75) 30 (91) 38 (81) 
 Non-Latino White 1 (3) 4 (9) 2 (6) 5 (11) 
 Other 1 (3) 7 (16) 1 (3) 4 (9) 
Child age, y, mean (SD) 10.1 (2.2) 9.2 (2.2) 10.0 (1.8) 9.4 (2.2) 
Child BMI, mean (SD)b 28.3 (5.3) 25.8 (4.0) 28.2 (4.1) 25.6 (3.8) 
 BMIz, mean (SD) 2.2 (0.3) 2.1 (0.4) 2.2 (0.3) 2.1 (0.3) 
 BMI percentile, mean (SD) 98.0 (1.8) 97.3 (2.5) 98.0 (2.0) 97.6 (1.8) 
Child with overweight,cn (%) 2 (6) 8 (18) 4 (12) 5 (11) 
Child with obesity,cn (%) 33 (94) 36 (82) 29 (88) 42 (89) 
a

Between site differences, P < .05.

b

Overweight is defined as BMI ≥85th percentile and <95th percentile. Obesity is defined as BMI ≥95th percentile.

Participants in GSH attended a mean of 7.4 (SD 4.4) out of 14 possible sessions (52.9%), whereas participants in FBT attended a mean of 4.5 (SD 2.9) out of 20 possible sessions (22.5%). The odds of attending GSH was 2.2 times greater than FBT, adjusting for covariates (P < .01) (Table 2). Overall, the probability of attending treatment sessions was significantly predicted by income (P < .01).

TABLE 2

Adjusted Odds of Attending GSH Compared With FBT

Odds Ratio95% CI
FBT 1.0 Reference 
GSH 2.22 1.21–4.06 
Treatment sessions (1 SD = 7.4) 0.45 0.38–0.54 
Male 1.0 Reference 
Female 1.56 0.95–2.58 
Age (1 SD = 2.2) 0.95 0.74–1.22 
Latino 1.0 Reference 
Non-Latino 0.68 0.32–1.45 
Income (1 SD = 9.8) 1.47 1.13–1.92 
Odds Ratio95% CI
FBT 1.0 Reference 
GSH 2.22 1.21–4.06 
Treatment sessions (1 SD = 7.4) 0.45 0.38–0.54 
Male 1.0 Reference 
Female 1.56 0.95–2.58 
Age (1 SD = 2.2) 0.95 0.74–1.22 
Latino 1.0 Reference 
Non-Latino 0.68 0.32–1.45 
Income (1 SD = 9.8) 1.47 1.13–1.92 

Model provides adjusted odds of attending GSH controlling for all variables listed in the table. SD was used as 1 unit of change for continuous variables.

When examining attrition (defined as missing 3 or more consecutive sessions), 75% of families in FBT and 48% of families in GSH met criteria. In the cox proportional hazard model, those assigned to GSH had a 67% reduced risk of attrition (P < .001) compared with those assigned to FBT (Fig 2). In the posttreatment survey, 53% of parents in GSH reported that their group assignment was “somewhat to very convenient,” whereas only 19% of parents reported that in FBT (P < .001).

FIGURE 2

Attrition plot: proportion of families who attended GSH or FBT for pediatric obesity. Families assigned to GSH had a 67% reduced risk of attrition (hazard ratio = 0.33, 95% CI 0.22–0.50, P < .001) compared to those assigned to FBT.

FIGURE 2

Attrition plot: proportion of families who attended GSH or FBT for pediatric obesity. Families assigned to GSH had a 67% reduced risk of attrition (hazard ratio = 0.33, 95% CI 0.22–0.50, P < .001) compared to those assigned to FBT.

Close modal

When examining differences in weight status at posttreatment and follow-up, there were no between-group difference in BMIz (P = .53) or %BMIp95 (P = .1) (Table 3). Intent-to-treat analysis showed a significant reduction in BMIz in both groups combined from baseline to posttreatment (β = −0.07 (0.01), P < .01, d: 0.60), and a nonsignificant increase at 6-month follow-up compared with posttreatment (β = .007 (0.01), P = .56, d = 0.09) (Table 4). Although between-group differences were noted when examining ΔBMIp95, both groups still had significant reductions in BMI percentiles. We conducted post-hoc analysis among those attending >50% of treatment sessions (21% of FBT; 43% of GSH) and there was no significant between-group difference in change in BMIz or %BMIp95.

TABLE 3

Between-Group Comparisons of Adjusted Means in Weight Status From Baseline to Posttreatment and 6-Month Follow-Up

Adjusted Means (95% CI)P
BaselinePosttreatment6-mo follow-up
BMI z-score    .53 
 FBT 2.13 (2.09–2.16) 2.05 (2.02–2.09) 2.06 (2.03–2.09)  
 GSH 2.14 (2.11–2.17) 2.06 (2.03–2.10) 2.07 (2.04–2.11)  
%BMIp95    .10 
 FBT 119 (118–120) 116 (115–118) 116 (115–118)  
 GSH 120 (119–121) 117 (116–119) 117 (116–119)  
ΔBMIp95    .03 
 FBT 4.24 (3.98–4.50) 3.82 (3.56–4.08) 3.97 (3.71–4.23)  
 GSH 4.52 (4.26–4.79) 4.11 (3.84–4.37) 4.26 (3.99–4.52)  
Adjusted Means (95% CI)P
BaselinePosttreatment6-mo follow-up
BMI z-score    .53 
 FBT 2.13 (2.09–2.16) 2.05 (2.02–2.09) 2.06 (2.03–2.09)  
 GSH 2.14 (2.11–2.17) 2.06 (2.03–2.10) 2.07 (2.04–2.11)  
%BMIp95    .10 
 FBT 119 (118–120) 116 (115–118) 116 (115–118)  
 GSH 120 (119–121) 117 (116–119) 117 (116–119)  
ΔBMIp95    .03 
 FBT 4.24 (3.98–4.50) 3.82 (3.56–4.08) 3.97 (3.71–4.23)  
 GSH 4.52 (4.26–4.79) 4.11 (3.84–4.37) 4.26 (3.99–4.52)  

Means adjusting for baseline weight status, age, ethnicity, sex, and clinic site. CI, confidence interval.

TABLE 4

Within-Group Changes in Weight Status From Baseline to Posttreatment and 6-Month Follow-Up by Treatment Group

Baseline–PosttreatmentPosttreatment–Follow-Up
Estimate (SE)PdEstimate (SE)Pd
BMI z-score       
 FBT −0.071 (0.019) <.01 0.56 −0.010 (0.02) .60 0.11 
 GSH −0.077 (0.019) <.01 0.63 0.026 (0.02) .17 0.31 
 Combined −0.074 (0.013) <.01 0.60 0.007 (0.13) .56 0.09 
%BMIp95       
 FBT −2.14 (0.66) <.01 0.51 −1.71 (0.66) .01 0.59 
 GSH −2.37 (0.66) <.01 0.56 1.58 (0.67) .02 0.54 
 Combined −2.26 (0.48) <.01 0.52 −0.063 (0.47) .89 0.021 
ΔBMIp95       
 FBT −0.42 (0.15) <.01 0.44 −0.24 (0.15) .10 0.37 
 GSH −0.40 (0.15) .01 0.41 0.54 (0.06) <.01 0.80 
Baseline–PosttreatmentPosttreatment–Follow-Up
Estimate (SE)PdEstimate (SE)Pd
BMI z-score       
 FBT −0.071 (0.019) <.01 0.56 −0.010 (0.02) .60 0.11 
 GSH −0.077 (0.019) <.01 0.63 0.026 (0.02) .17 0.31 
 Combined −0.074 (0.013) <.01 0.60 0.007 (0.13) .56 0.09 
%BMIp95       
 FBT −2.14 (0.66) <.01 0.51 −1.71 (0.66) .01 0.59 
 GSH −2.37 (0.66) <.01 0.56 1.58 (0.67) .02 0.54 
 Combined −2.26 (0.48) <.01 0.52 −0.063 (0.47) .89 0.021 
ΔBMIp95       
 FBT −0.42 (0.15) <.01 0.44 −0.24 (0.15) .10 0.37 
 GSH −0.40 (0.15) .01 0.41 0.54 (0.06) <.01 0.80 

Linear mixed effects analysis was conducted adjusting for baseline weight status, age, ethnicity, sex, and clinic site. Final estimates were taken from models across 100 multiple imputed data sets. Combined estimates were presented for BMI outcomes if there were no significant group differences.

The results of this study provide support for a novel, less intense model of obesity treatment (GSH). Attendance was greater than in traditional FBT, and both treatments attained significant changes in BMIz, %BMIp95, and ΔBMIp95. Although this program is self-directed, GSH still provides enough structure to promote treatment compliance and application of skills, much like what is suggested for stage 2 treatment by the expert committee recommendations.3  GSH also provides similar lifestyle recommendations and behavioral strategies as FBT. According to the Behavior Change Wheel,31  successful behavior change is facilitated by interventions that promote motivation to change and convey strategies that increase patient skills, efficacy, and empowerment to change. As in FBT, GSH health coaches provide regular accountability to patients/families, work with them to set goals, and problem-solve issues that arise. This type of structure has been effective in the treatment of other illnesses such as eating disorders, anxiety, and depression,3234  and may also be applicable for the treatment of childhood obesity.

Many national groups recommend nutrition and weight-related counseling at each well-child visit.35,36  Changes in the electronic health record, such as best practice advisories and clinical decision support tools, have been successfully implemented to support these behaviors.3739  These changes have resulted in increased nutrition and exercise counseling, as well as documentation of the diagnosis of obesity.38  However, there was little change in weight status. Adopting a chronic care model40  for obesity management41,42  that includes decision support tools to link patients to community resources and other components of the health care system is critical for effective management. Nevertheless, it is also important to create well-informed patients/families who feel supported and activated to make behavior change. Without a strategy to address this aspect of the model, we will not be able to provide optimal care.

Taveras et al tested 2 programs delivered in the primary care setting that combined decision support tools with brief interventions (eg, motivational interviewing).43,44  The inclusion of behavioral coaching resulted in greater decreases in BMIz than decision support alone. These results demonstrate the importance of behavior change support during obesity treatment. Of note, children in our study had a higher mean BMIz, which can be more difficult to treat,45  than children referenced in the previous studies. Nevertheless, we were able to attain comparable changes, if not slightly greater decreases, in weight status using GSH. Our study provides evidence that lower intensity programs may be effective at attaining moderate decreases in weight status if key behavioral strategies, such as stimulus control and parental monitoring,46  and support for behavior change through feedback and accountability (either in a group or individual setting) are included. The benefit of GSH is that it may be more feasible to implement because of increased flexibility in scheduling, decreased time commitment from the parents/children, no need for large group meeting spaces, and decreased numbers of staff to implement the program. These changes can increase access to treatment of those who cannot attend intensive multicomponent programs. Without increased access, families cannot begin to learn how to implement behavioral weight loss strategies.

Although attendance was increased and changes in BMIz and %BMIp95 were similar to traditional group-based programs, there were some limitations. The study population was primarily Latino and from lower income neighborhoods, limiting generalizability to other groups. However, this is also a strength of the study since these groups are often underrepresented in clinical trials. We also had set inclusion criteria and did not test this program among children who were taking medications that might impact weight (eg, steroids or psychiatric medications) or had limitations in physical activity, which further limits the generalizability of these findings. We were unable to randomize 200 families despite receiving >700 referrals with 230 families completing the consent process. Unfortunately, this experience is common in pediatric weight control studies.4749  However, randomizing 164 families and analyzing data from 159 families still provided us with 80% power to detect an effect in BMIz and 95% power to detect an effect in attendance. Finally, anthropometric assessments were obtained by study staff who were not blinded to treatment group. However, we believe this impact to be minimal given the objective nature of these assessments.

The GSH model of treatment may be a promising option for pediatric obesity management. This format employs many of the same behavioral strategies used in FBT but allows for greater flexibility in scheduling and could be delivered in a variety of settings. In this study, we implemented GSH for pediatric obesity in a primary care clinic and were able to obtain significant decreases in BMIz, %BMIp95, and ΔBMIp95 at the 6-month time point. Future studies should investigate effective methods of disseminating and implementing GSH-type programs in the community. Factors to consider include: appropriate education/training level of the health coach delivering this treatment and whether it can be implemented in a diverse range of settings (eg, after-school programs, churches, or community centers) and patient populations (eg, teenagers).

We thank the leadership at Children’s Primary Care Medical Group and the providers who participated in development of this study. Without their support and assistance, this study would not have been feasible.

Dr Rhee was the principal investigator of this study and conceptualized the study, designed the protocol, was responsible for the recruitment of participants and acquisition of data, participated in analysis, and drafted the original manuscript; Dr Herrera participated in the design of the work, the recruitment of participants and acquisition of data, interpretation of the data, and provided critical review of the manuscript; Drs Strong and Boutelle participated in the conceptualization of the study, design of the work, analysis and interpretation of the data, and provided critical review of the manuscript; Drs Kang-Sim and Shi participated in the design of the work, analysis and interpretation of the data, and provided critical review of the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

This study is registered at ClinicalTrials.gov, NCT02976454, https://clinicaltrials.gov/ct2/show/ NCT02976454. Research data may be shared upon request 3 years after publication of the main outcomes. Deidentified data and supporting documents will be made available for secondary analyses. Data will be available for education, research, and nonprofit purposes.

FUNDING: Supported by a grant from the Health Resources and Services Administration, Maternal and Child Health Bureau (#R40MC29452) to Dr Rhee. They were not involved in any component of the study design, data collection, analysis, or writing of this manuscript.

CONFLICT OF INTEREST DISCLAIMER: The authors have indicated they have no conflicts of interest relevant to this article to disclose.

GSH

guided self-help

FBT

family-based behavioral treatment

PCP

primary care provider

BMIz

BMI z-score

%BMIp95

percentage from the 95th BMI percentile

ΔBMIp95

difference from the 95th BMI percentile

1
Ogden
CL
,
Carroll
MD
,
Kit
BK
,
Flegal
KM
.
Prevalence of childhood and adult obesity in the United States, 2011-2012
.
JAMA
.
2014
;
311
(
8
):
806
814
2
Krebs
NF
,
Jacobson
MS
.
American Academy of Pediatrics Committee on Nutrition
.
Prevention of pediatric overweight and obesity
.
Pediatrics
.
2003
;
112
(
2
):
424
430
3
Barlow
SE
;
Expert Committee
.
Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: summary report
.
Pediatrics
.
2007
;
120
(
suppl 4
):
S164
S192
4
National Initiative for Children’s Healthcare Quality CPRC, and Child and Adolescent Health Measurement Initiative
.
Childhood Obesity Action Network
.
Available at: www.nichq.org/obesityactionnetwork. Accessed August 22, 2008
5
Barton
M
;
US Preventive Services Task Force
.
Screening for obesity in children and adolescents: US Preventive Services Task Force recommendation statement
.
Pediatrics
.
2010
;
125
(
2
):
361
367
6
Wilfley
DE
,
Staiano
AE
,
Altman
M
, et al.
Improving Access and Systems of Care for Evidence-Based Childhood Obesity Treatment Conference Workgroup
.
Improving access and systems of care for evidence-based childhood obesity treatment: Conference key findings and next steps
.
Obesity (Silver Spring)
.
2017
;
25
(
1
):
16
29
7
Epstein
LH
,
Paluch
RA
,
Roemmich
JN
,
Beecher
MD
.
Family-based obesity treatment, then and now: twenty-five years of pediatric obesity treatment
.
Health Psychol
.
2007
;
26
(
4
):
381
391
8
Whitlock
EP
,
O’Connor
EA
,
Williams
SB
,
Beil
TL
,
Lutz
KW
.
Effectiveness of weight management interventions in children: a targeted systematic review for the USPSTF
.
Pediatrics
.
2010
;
125
(
2
):
e396
e418
9
Mead
E
,
Brown
T
,
Rees
K
, et al
.
Diet, physical activity and behavioural interventions for the treatment of overweight or obese children from the age of 6 to 11 years
.
Cochrane Database Syst Rev
.
2017
;
6
(
6
):
CD012651
10
Skelton
JA
,
Beech
BM
.
Attrition in paediatric weight management: a review of the literature and new directions
.
Obes Rev
.
2011
;
12
(
5
):
e273
e281
11
Hampl
S
,
Paves
H
,
Laubscher
K
,
Eneli
I
.
Patient engagement and attrition in pediatric obesity clinics and programs: results and recommendations
.
Pediatrics
.
2011
;
128
(
suppl 2
):
S59
S64
12
Kwitowski
M
,
Bean
MK
,
Mazzeo
SE
.
An exploration of factors influencing attrition from a pediatric weight management intervention
.
Obes Res Clin Pract
.
2017
;
11
(
2
):
233
240
13
Davis
AM
,
Daldalian
MC
,
Mayfield
CA
, et al
.
Outcomes from an urban pediatric obesity program targeting minority youth: the Healthy Hawks program
.
Child Obes
.
2013
;
9
(
6
):
492
500
14
Cui
Z
,
Seburg
EM
,
Sherwood
NE
,
Faith
MS
,
Ward
DS
.
Recruitment and retention in obesity prevention and treatment trials targeting minority or low-income children: a review of the clinical trials registration database
.
Trials
.
2015
;
16
:
564
15
Yancey
AK
,
Ortega
AN
,
Kumanyika
SK
.
Effective recruitment and retention of minority research participants
.
Annu Rev Public Health
.
2006
;
27
:
1
28
16
Boutelle
KN
,
Norman
GJ
,
Rock
CL
,
Rhee
KE
,
Crow
SJ
.
Guided self-help for the treatment of pediatric obesity
.
Pediatrics
.
2013
;
131
(
5
):
e1435
e1442
17
Bergmann
K
,
Mestre
Z
,
Strong
D
, et al
.
Comparison of two models of family-based treatment for childhood obesity: a pilot study
.
Child Obes
.
2019
;
15
(
2
):
116
122
18
Kanfer
F
,
Karoly
P
.
Self-control: A behavioristic excursion into the lion’s den
.
Behav Ther
.
1972
;
3
:
398
416
19
Karoly
P
,
Kanfer
F
.
Self-Management and Behavior Change: From Theory to Practice
.
New York
:
Pergamon
;
1982
20
Lenders
CM
,
Manders
AJ
,
Perdomo
JE
,
Ireland
KA
,
Barlow
SE
.
Addressing pediatric obesity in ambulatory care: where are we and where are we going?
Curr Obes Rep
.
2016
;
5
(
2
):
214
240
21
Rhee
KE
,
Herrera
L
,
Strong
D
,
DeBenedetto
AM
,
Shi
Y
,
Boutelle
KN
.
Design of the GOT Doc study: a randomized controlled trial comparing a guided self-help obesity treatment program for childhood obesity in the primary care setting to traditional family-based behavioral weight loss
.
Contemp Clin Trials Commun
.
2021
;
22
:
100771
22
California Center for Public Health Advocacy
.
Overweight and obesity among children by California City – 2010
.
23
Wilkinson
GS
,
Robertson
GJ
.
Wide Range Achievement Test 4 Professional Manual
.
Lutz, FL
:
Psychological Assessment Resources
;
2006
24
Freedman
DS
,
Davies
AJG
,
Kompaniyets
L
, et al
.
A longitudinal comparison of alternatives to body mass index z-scores for children with very high body mass indexes
.
J Pediatr
.
2021
;
235
:
156
162
25
Freedman
DS
,
Woo
JG
,
Ogden
CL
,
Xu
JH
,
Cole
TJ
.
Distance and percentage distance from median BMI as alternatives to BMI z score
.
Br J Nutr
.
2020
;
124
(
5
):
493
500
26
Kuczmarski
RJ
,
Ogden
CL
,
Grummer-Strawn
LM
, et al
.
CDC growth charts: United States
.
Adv Data
.
2000
;
314
:
1
27
27
Ogden
CL
,
Kuczmarski
RJ
,
Flegal
KM
, et al
.
Centers for Disease Control and Prevention 2000 growth charts for the United States: improvements to the 1977 National Center for Health Statistics version
.
Pediatrics
.
2002
;
109
(
1
):
45
60
28
Campbell
MK
,
Mollison
J
,
Steen
N
,
Grimshaw
JM
,
Eccles
M
.
Analysis of cluster randomized trials in primary care: a practical approach
.
Fam Pract
.
2000
;
17
(
2
):
192
196
29
R: A language and environment for statistical computing
.
[computer program]
.
Vienna, Austria
:
R Foundation for Statistical Computing
;
2020
30
Hedeker
D
,
Gibbons
R
.
Longitudinal Data Analysis
.
New York
:
Wiley
;
2006
31
Michie
S
,
van Stralen
MM
,
West
R
.
The behaviour change wheel: a new method for characterising and designing behaviour change interventions
.
Implement Sci
.
2011
;
6
:
42
32
Traviss-Turner
GD
,
West
RM
,
Hill
AJ
.
Guided self-help for eating disorders: a systematic review and metaregression
.
Eur Eat Disord Rev
.
2017
;
25
(
3
):
148
164
33
Cuijpers
P
,
Donker
T
,
van Straten
A
,
Li
J
,
Andersson
G
.
Is guided self-help as effective as face-to-face psychotherapy for depression and anxiety disorders? A systematic review and meta-analysis of comparative outcome studies
.
Psychol Med
.
2010
;
40
(
12
):
1943
1957
34
Cuijpers
P
,
Noma
H
,
Karyotaki
E
,
Cipriani
A
,
Furukawa
TA
.
Effectiveness and acceptability of cognitive behavior therapy delivery formats in adults with depression: a network meta-analysis
.
JAMA Psychiatry
.
2019
;
76
(
7
):
700
707
35
Health Resources and Services Administration
.
EPSDT overview
.
Available at: http://mchb.hrsa.gov/epsdt/overview.html. Accessed July 20, 2012
36
U.S. Department of Health and Human Services
.
Healthy people 2020
.
Available at: www.healthypeople.gov/2020/topicsobjectives2020/default. Accessed November 18, 2014
37
Bordowitz
R
,
Morland
K
,
Reich
D
.
The use of an electronic medical record to improve documentation and treatment of obesity
.
Fam Med
.
2007
;
39
(
4
):
274
279
38
Coleman
KJ
,
Hsii
AC
,
Koebnick
C
, et al
.
Implementation of clinical practice guidelines for pediatric weight management
.
J Pediatr
.
2012
;
160
(
6
):
918
22.e1
39
Rattay
KT
,
Ramakrishnan
M
,
Atkinson
A
,
Gilson
M
,
Drayton
V
.
Use of an electronic medical record system to support primary care recommendations to prevent, identify, and manage childhood obesity
.
Pediatrics
.
2009
;
123
(
suppl 2
):
S100
S107
40
Wagner
EH
,
Austin
BT
,
Von Korff
M
.
Organizing care for patients with chronic illness
.
Milbank Q
.
1996
;
74
(
4
):
511
544
41
Jacobson
D
,
Gance-Cleveland
B
.
A systematic review of primary healthcare provider education and training using the Chronic Care Model for childhood obesity
.
Obes Rev
.
2011
;
12
(
5
):
e244
e256
42
Ely
AC
,
Banitt
A
,
Befort
C
, et al
.
Kansas primary care weighs in: a pilot randomized trial of a chronic care model program for obesity in 3 rural Kansas primary care practices
.
J Rural Health
.
2008
;
24
(
2
):
125
132
43
Taveras
EM
,
Marshall
R
,
Sharifi
M
, et al
.
Comparative effectiveness of clinical-community childhood obesity interventions: a randomized clinical trial
.
JAMA Pediatr
.
2017
;
171
(
8
):
e171325
44
Taveras
EM
,
Marshall
R
,
Kleinman
KP
, et al
.
Comparative effectiveness of childhood obesity interventions in pediatric primary care: a cluster-randomized clinical trial
.
JAMA Pediatr
.
2015
;
169
(
6
):
535
542
45
Ryder
JR
,
Fox
CK
,
Kelly
AS
.
Treatment options for severe obesity in the pediatric population: current limitations and future opportunities
.
Obesity (Silver Spring)
.
2018
;
26
(
6
):
951
960
46
Boutelle
KN
,
Kang Sim
DE
,
Rhee
KE
,
Manzano
M
,
Strong
DR
.
Family-based treatment program contributors to child weight loss
.
Int J Obes
.
2020
;
45
(
1
):
77
83
47
Riggs
KR
,
Lozano
P
,
Mohelnitzky
A
,
Rudnick
S
,
Richards
J
.
An adaptation of family-based behavioral pediatric obesity treatment for a primary care setting: group health family wellness program pilot
.
Perm J
.
2014
;
18
(
3
):
4
10
48
Raynor
HA
,
Osterholt
KM
,
Hart
CN
,
Jelalian
E
,
Vivier
P
,
Wing
RR
.
Evaluation of active and passive recruitment methods used in randomized controlled trials targeting pediatric obesity
.
Int J Pediatr Obes
.
2009
;
4
(
4
):
224
232
49
Raynor
HA
,
Barroso
C
,
Propst
S
,
Berlin
K
,
Robson
S
,
Khatri
P
.
Reach of a low-intensity, multicomponent childhood overweight and obesity intervention delivered in an integrated primary care setting
.
Transl Behav Med
.
2020
;
10
(
3
):
760
769