- Email: [email protected]
Automated Telephone Counseling for Parents of Overweight Children
Automated Telephone Counseling for Parents of Overweight Children A Randomized Controlled Trial Paul A. Estabrooks, PhD, Jo Ann Shoup, MSW, Michelle Gattshall, MA, Padma Dandamudi, MPH, Susan Shetterly, MS, Stan Xu, PhD Background: Interactive technologies have the potential to increase the reach and frequency of practical clinical interventions that assist the parents of overweight and at-risk children to promote healthy lifestyle behaviors for their families. Design:
A practical RCT evaluated the relative effectiveness of three interventions to support parents of overweight or at-risk children to change the home environment to foster more healthful child eating and activity behaviors, thereby reducing child BMI and BMI z-scores. A secondary purpose was to determine the patterns of use and potential dose effect for the highest-intensity intervention.
Setting/ Parent-and-child (aged 8 –12 years) dyads (N⫽220) who received care from Kaiser participants: Permanente Colorado were assigned randomly to one of the three Family Connections (FC) interventions: FC-workbook, FC-group, or FC-interactive voice response (IVR) counseling. Main outcome measures:
Child BMI z-scores, as well as symptoms of eating disorders and body image, were assessed at baseline, 6 months, and 12 months.
Results:
The BMI z-scores of children assigned to the FC-IVR intervention were the only ones that decreased from baseline to 6 months (0.07 SD) and from baseline to 12 months (0.08 SD, p⬍0.05). Children whose parents completed at least six of the ten FC-IVR counseling calls had decreased BMI z-scores to a greater extent than children in the FC-workbook or FC-group interventions at both 6 months (p⬍0.05) and 12 months (p⬍0.01). No intervention increased child symptoms of eating disorders or body dissatisfaction at any time point.
Conclusions: This trial demonstrated that automated telephone counseling can support the parents of overweight children to reduce the extent to which their children are overweight. Trial registration:
NCT00433901 (Am J Prev Med 2009;36(1):35– 42) © 2009 Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine.
Introduction
O
ver the previous 3 decades, the prevalence of overweight children and the degree of overweight have increased significantly,1–3 and overweight children, while still in their youth, are at risk for hypokinetic diseases and much more likely to seek mental From the Department of Human Nutrition, Foods, and Exercise, Virginia Polytechnic Institute and State University (Estabrooks), Blacksburg, Virginia; and the Institute of Health Research, Kaiser Permanente Colorado (Estabrooks, Shoup, Gattshall, Dandamudi, Shetterly, Xu), Denver, Colorado Address correspondence and reprint requests to: Paul A. Estabrooks, PhD, Virginia Tech Riverside, 1 Riverside Circle SW, Suite #104, Roanoke VA 24016. E-mail: [email protected].
health care than healthy weight children.4 – 6 The American Academy of Pediatrics (the Academy) has released recommendations5 highlighting the need to promote healthful eating and physical activity while reducing sedentary behavior and sugared-drink consumption through behavioral strategies documented in studies of intensive family-based behavioral therapy.7–32 Although the research on familial interventions targeting childhood overweight issues is convincing, there is little evidence that these programs have been translated into regular practice or reach large segments of the population in need.33–35 This may be because it is questionable whether interventions tested in previous efficacy trials use resources and staff that are typical of
Am J Prev Med 2009;36(1) © 2009 Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine
0749-3797/09/$–see front matter doi:10.1016/j.amepre.2008.09.024
35
those in usual care.36 An uninvestigated area of research is the potential of alternative models of intervention delivery, such as interactive technologies, to achieve effects similar to previous efficacy trials while requiring fewer system resources. Interactive voice response (IVR) technology may be a viable avenue for delivering pediatric weight management–intervention strategies to parents of children who are overweight. This technology has been used in a number of healthcare contexts and has demonstrated effectiveness in enhancing physical activity and healthy eating in diabetic and older adult samples.37– 42 Further, a review of the literature indicates that IVR care was acceptable and helpful to patients,43 cost effective in maintaining diabetes selfmanagement,38,43 and supportive of caregivers in other contexts.44 The present study reflects a research–practice partnership with a goal to develop a practical and sustainable intervention based on previous efficacy trials.7–12 To achieve this goal, a practical RCT was designed to test interventions—all viable candidates for future implementation—that varied by frequency (i.e., one contact, three contacts, or 13 contacts) and channel of delivery. The specific purpose of the study was to determine the relative effectiveness of delivering an intervention through an additive process that included (1) a self-help Family Connections workbook for parents (FC-workbook); (2) the workbook plus two smallgroup sessions with a registered dietitian (FC-group); and (3) workbook, two small-group sessions, and ten automated IVR tailored-counseling sessions (FC-IVR). As there is little information on the extent to which parents would use IVR or whether parents could be engaged long enough to receive the core content of the intervention before discontinuing use, a second twofold purpose was to determine what percentage of the parents assigned to the FC-IVR intervention would complete more than half of the sessions, thus receiving the core content, and to determine if the ⬎50% dose of the intervention heightened effectiveness.
2007, and all analyses were performed from May 2007 through December 2007. A three-group practical controlled-trial design was used to allow future decision-makers with the appropriate data to determine whether a new intervention significantly surpasses the effect of the current standard care (i.e., FC-IVR versus FC-group) or matches the effect while utilizing fewer resources (i.e., FC-workbook versus FC-group).45 Through a random-numbers table, participants were assigned randomly (families/staff unblinded) to the FC-workbook, the FC-group, or the FC-IVR intervention. It was hypothesized that both the FC-group and FC-IVR interventions would demonstrate medium effects when compared to the FC-workbook intervention but would demonstrate small effects when compared to each other. Sample size calculations were completed, varying the detectable effect sizes from small to medium with a power of 0.8. The result was a need for 42 participants per intervention to detect a medium effect and 64 participants to detect a small effect. Therefore, more participants were assigned randomly to the FC-group and FC-IVR interventions to increase the statistical power for comparisons between those interventions and among participants who completed half of the intervention relative to FC-workbook and FC-group. Finally, an a priori alternate hypothesis was developed: that the FC-workbook and FC-group interventions were insufficient in either duration or frequency to achieve significant effects. In that case, collapsing the FC-workbook and FCgroup participants into a single group would enhance the power to detect a smaller effect with the FC-IVR participants.
Participants Children aged 8 –12 years with a BMI ⱖ85th percentile for their age were eligible for the study. Exclusion criteria included plans to move out of the state during the course of the study or a request by the child’s pediatrician that the family not be contacted. The initial pool of potential study participants included 1487 families (Figure 1). Of these, 656 parents were contacted during the recruitment phase of the study. Two-hundred and twenty families completed the baseline visit and subsequently were assigned randomly to FCworkbook (n⫽50); FC-group (n⫽85); or FC-IVR (n⫽85). The participation level of eligible families who were contacted by telephone was approximately 38%. Table 1 provides baseline child characteristics.
Measures
Methods Design and Procedures The Kaiser Permanente Colorado IRB approved this study in March 2004. All study participants signed parental consent or child assent forms as well as authorization for the use of medical records. Parents of overweight children were identified through electronic medical records, contacted by telephone, and invited to participate in the study. Enrollment for the study began in May 2004 and continued through January 2006. All participants completed an assessment at baseline, 6 months, and 12 months. Participants were followed for a period of 1 year. Follow-up measurement and assessment began in November 2004 and continued through February
36
Demographic information was obtained through medical records and parent reports. Children’s age and gender were verified by parents. Parent and child race/ethnicity information was provided by participating parents. BMI z-scores. Height was measured using a stadiometer, and weight was assessed using a regularly calibrated medical scale.27 BMI scores were calculated based on the value of the 50th-percentile BMI ranking and the SD attributable to appropriate age and gender samples from the CDC growth charts.46 Code available from the CDC (www.cdc.gov) was used to accurately calculate associated BMI z-scores (i.e., the number of SD units that the child’s BMI deviates from the mean reference value for age and gender) using a modified least mean square technique that approximates z-scores from
American Journal of Preventive Medicine, Volume 36, Number 1
www.ajpm-online.net
used as a screening tool for eating disorders in children.52 Individual responses were categorized into weight dissatisfaction, restricting/purging, binge eating, and body dissatisfaction components. The scale has high internal consistency.52
Interventions A model proposed by Golan and associates7,9,12,32— based on a social– ecologic theory that promotes changes to the home environment, parenting skills, role modeling, and positive parental health behaviors as primary factors in pediatric weight management and consistent with the Academy’s recommendations5— was used as the basis for all intervention conditions. The model stresses a health-centered approach and targets the parent exclusively as the agent of change. Interventions based on Golan’s model and Figure 1. Flow chart of participant identification and recruitment tested across a number of studies7–12,32 have demonstrated short- and long-term growth curves estimated from the 2000 National Health and 47 effectiveness in reducing child BMI and parental cardiovascular Nutrition Examination Survey (NHANES) data. This outdisease risk factors while also reducing risky eating behaviors in come was selected because it provides a comparison of children. The model also receives indirect support from studies relative decrease in the extent of overweight to the referent that have used similar family-based approaches and have dempopulation based on age and gender, and is one of the most 48 onstrated that they are effective,53 replicable,30 and produce frequently reported relative weight outcomes for children. large and reliable effects over time.54 See Appendix A, available Physical activity and sedentary behavior (Youth Behavioral online at www.ajpm-online.net, for examples of strategies develRisk Survey questions49). Children indicated the number of oped based on this model. days in the previous 7 days on which they completed moderFC-workbook. A 61-page workbook was developed to proate physical activity (MPA), vigorous physical activity (VPA), or both. Sedentary behavior was assessed based on the mote increased physical activity and the consumption of fruits numbers of hours of screen time during school days. The and vegetables in concert with decreased sugared-drink conscale reliability was acceptable. The scale does underestimate sumption and television viewing/recreational computer time. the days of MPA and overestimates the days of VPA.49 The workbook included two distinct sections. Part 1 targeted 3 days of intervention, and Part 2 targeted 2 days, each with Fruit, vegetable, and sugared-drink consumption. Each parspecific homework assignments (Appendix B, www.ajpmticipant completed the Block Kids Questionnaire, a foodonline.net). The workbook encouraged parents to complete frequency questionnaire for children and youth aged 8 –17 50,51 all 5 days of intervention across a single week. Homework years. The questionnaire was developed from the assignments were intended to encourage lasting changes in NHANES 1999 –2002 dietary recall data, and the nutrient the families. All parents randomly assigned to this intervendatabase was developed from the U.S. Department of Agrition received the workbook from study research assistants. culture nutrient database, version 1.0, for dietary studies. The food-frequency questionnaire asks individual portion sizes, FC-group. This intervention consisted of two small-group allowing for the computation of fruit, vegetable, and sugaredsessions (2 hours each, spaced 1 week apart) held at a local 50,51 drink servings. clinic and delivered by a dietitian. Each session included 10 –15 parents representing distinct children and utilized the Eating-disorder symptoms. The Kids’ Eating Disorders Survey Family Connections workbook. The first session focused on (KEDS) is a self-report questionnaire for children aged ⱖ8 years
January 2009
Am J Prev Med 2009;36(1)
37
Table 1. Participant characteristics at baseline Intervention
Age (M in years) Gender (%) Male Race/ethnicity (%) White Hispanic 6-month visit completed 12-month visit completed BMI (M kg/m2) BMI z-score (M) VPA (self-reported M days/week) MPA (self-reported M days/week) Sedentary/screen time (M hours/day) Sugared drinks (M ounces/week) Fruits (M servings/day) Vegetables (M servings/day) KEDS (total score M)
Total sample Nⴝ220
FC-workbook nⴝ50
FC-group nⴝ85
FC-group and FC-IVR nⴝ85
10.7
11.0
10.6
10.7
54
39
58
59
63 26 78 71 27.2 2.04 4.0 2.3 5.5 85.3 1.2 1.3 7.3
59 29 78 74 27.1 2.00 4.0 2.4 5.3 76.6 1.2 1.6 7.8
69 19 75 66 27.4 2.07 4.3 2.6 5.4 92.5 1.0 1.0a 7.3
60 30 80 74 27.1 2.04 3.8 2.0 5.6 83.3 1.3 1.4 7.0
p⬍0.05 for contrasts by study arm (2 for percents, ANOVA for continuous variables [Kruskal–Wallis tests significance levels comparable]). Participants in the FC-group reported consuming significantly fewer vegetables per day at baseline. FC, Family Connections; IVR, interactive voice response; KEDS, Kids’ Eating Disorders Survey; MPA, moderate physical activity; VPA, vigorous physical activity
a
parents’ behavioral health skills and knowledge of weight, nutrition, and physical activity. It also identified key parenting skills: limit setting, effective communication, and role modeling. This session concluded with role playing, problem solving, and the development of an action plan. Session 2 integrated the knowledge acquired in Session 1, the experiences associated with the action plan, and strategies for restructuring the home environment. The session again concluded with parents’ completing an action plan for parental behaviors, role modeling, and changes to the home environment that would facilitate healthy eating and physical activity. FC-IVR. Parents assigned to this intervention had completed the two-session Family Connections small-group program and were subsequently provided with ten follow-up sessions delivered via IVR. Telephone follow-up calls commenced 1 week after the second small-group session. The participants could either call the IVR system proactively or wait for the system to call at the designated time. Participant responses to IVR questions and branching logic were used to determine the content of each call. At the beginning of each call, the primary parent would hear the goal selected the previous week and rate his or her achievement of that goal. Based on the rating, the parent was given the option of hearing tips related to the topic of the prior week’s goal. Following this, the parent was able to select specific messages related to the intervention content. Each call concluded with a goal-setting procedure. See Appendix C, at www.ajpm-online.net for both descriptions of strategies that were provided via IVR and the timing of follow-up contacts. The sixth IVR-counseling call provided parents with instruction on a family goal-setting procedure based on the 5A’s model (Assess behaviors and motivation, Advise on healthy behaviors, Ask about goals and reach collaborative agreement, Assist with barrier identification and resolution, and
38
Arrange follow-up).55 During this call, parents were trained to lead their family through regular goal setting related to physical activity and eating. This sixth call was considered the final component of the core content of the intervention. Calls 7–10 reinforced the information delivered in the initial six calls.
Statistical Analyses Intent-to-treat analyses were completed across three time points (baseline, 6-month follow-up, and 12-month follow-up) to determine changes in outcomes over time. Linear mixed models56 were fitted to account for the correlation of repeated measures, assuming that any missing data were missing at random. Preplanned linear contrasts were used to determine the presence of changes across study outcomes at (1) baseline and 6 months, (2) 6 months and 12 months, and (3) baseline and 12 months. These analyses were followed by a comparison of changes among groups across time. Mixed models were used for changes in BMI z-scores and square root–transformed dietary variables. Physical activities and sedentary/screen time were measured as rates: days of physical activity per week and hours of sedentary/screen time per day. These rates were not normally distributed; therefore, nonlinear random-effect models57,58 were used to fit these rates with logit link function and binomial distribution.59 All statistical analyses were performed using SAS version 9.1. To examine the second purpose, an a priori decision was made to complete a secondary analysis comparing the participants who completed six or more of the IVR sessions to participants who (1) completed fewer than six sessions, (2) received the FC-workbook, and (3) received the FC-group interventions. Two analytic plans were developed. One plan included the potential to examine the differences among four newly formed conditions (FC-workbook; FC-group; FCIVR: 0 –5 calls; FC-IVR: 6 –10 calls) based on IVR completion.
American Journal of Preventive Medicine, Volume 36, Number 1
www.ajpm-online.net
January 2009
p⬍0.05 for BL–12 months p⬍0.05 for BL– 6 months c p⬍0.05 for 6 –12 months BL, baseline; IVR, interactive voice response; KEDS, Kids’ Eating Disorders Survey; mo, months; MPA, moderate physical activity; VPA, vigorous physical activity b
a
12 mo 6 mo BL
6 mo
12 mo
BL
FC-group (SE) FC-workbook (SE)
Table 2. Variable M (SD) according to randomized condition
Fifty-four percent of the recruited children were boys (average age⫽10.7 years, SD ⫾1.4); 63% of the recruited children were white, while 26% were Latino. Participants were primarily overweight rather than at risk of overweight (average BMI percentile⫽97.3, SD ⫾2.3). The intervention conditions did not differ on any demographic variables. Table 2 includes the descriptive and within-group intent-to-treat comparisons across study groups. Children in the FC-group significantly decreased their BMI z-scores from baseline to 6 months, but the decrease was not sustained at 12 months. Participants in the FC-IVR conditions significantly decreased their BMI z-scores from baseline to 6 months and baseline to 12 months, while those in the FC-workbook intervention significantly decreased their BMI z-scores from baseline to 12 months. There were no differences among conditions. Participants in FC-IVR reported a significant increase in the number of days they participated in MPA from baseline to 6 months and baseline to 12 months. Changes in fruit, vegetable, and sugared-drink consumption did not demonstrate a consistent pattern of change within or among groups. Regardless of the intervention condition, all children reported healthy behaviors in response to the KEDS, and no increases in unhealthy behaviors were detected over the course of the study. Table 3 includes the descriptive and within-group comparisons based on the secondary purpose to examine differences in effect according to IVR completion. Participants in the FC-IVR intervention were divided into two groups: those who completed 0 –5 calls (n⫽20) and those who completed 6 –10 calls (n⫽38). Participants in the group FC-IVR: 6 –10 calls were the only ones with significantly decreased BMI z-scores from baseline to 6 months and from 6 months to 12 months. Between-group comparisons demonstrated that children in the group FC-IVR: 6 –10 calls significantly decreased BMI z-scores when compared to those assigned to the FC-workbook or FC-group condition at both 6 months (F[3,148]⫽–1.11; p⬍0.05) and 12 months (F[3,148]⫽–2.89; p⬍0.01).
BL
Results
n 49 38 36 85 64 56 85 68 63 BMI z-score 2.04 (0.02) 1.99 (0.03) 1.98 (0.03)a 2.06 (0.04) 2.03 (0.04)b 2.04 (0.04) 2.03 (0.04) 1.96 (0.04)b 1.95 (0.04)a VPA self-report (days/week) 4.02 (2.07) 3.97 (1.94) 4.32 (2.13) 4.30 (2.08) 4.44 (2.04) 4.04 (2.05) 3.75 (2.43) 3.57 (2.24) 3.47 (2.09) MPA self-report (days/week) 2.36 (2.32) 2.47 (2.10) 2.79 (1.95) 2.56 (2.26) 2.82 (2.35) 2.36 (2.01) 2.01 (2.18) 2.64 (2.35)b 2.71 (2.21)a Sedentary/screen time 5.27 (2.15) 5.00 (2.05) 5.47 (1.96) 5.39 (2.22) 5.49 (2.55) 5.64 (2.61) 5.63 (2.54) 5.70 (2.20) 5.60 (2.04) (hours/day) Sugared drinks (ounces/week) 76.63 (83.68) 50.40 (42.35) 60.94 (55.50) 92.46 (116.05) 63.93 (56.86)b 71.81 (89.97) 83.25 (98.90) 59.10 (58.29)b 59.63 (62.74) Fruits (servings/day) 1.20 (0.90) 1.10 (1.20) 1.22 (1.32) 0.97 (0.82) 1.24 (1.28) 1.15 (1.05) 1.32 (1.14) 1.76 (2.29) 1.26 (1.88)c Vegetables (servings/day) 1.61 (1.75) 1.54 (1.73) 1.26 (1.34) 1.01 (0.99) 1.27 (1.42) 1.48 (1.80)a 1.36 (1.37) 1.50 (1.66) 1.41 (1.39) KEDS total 7.82 (3.07) 7.21 (3.00) 6.85 (3.50) 7.26 (3.57) 6.53 (3.61) 6.93 (3.75) 7.00 (2.74) 6.65 (2.87) 6.48 (3.37)
12 mo 6 mo
FC-group and FC-IVR (SE)
This analytic plan would occur if the hypothesized difference between the FC-workbook and FC-group interventions was detected in the intent-to-treat analysis as per the initial hypothesis. The second plan included the potential to examine the differences across three newly formed conditions (combined FC-workbook and FC-group; FC-IVR: 0 –5 calls; and FC-IVR: 6 –10 calls). The second plan was developed in the event that the alternative hypothesis was supported, that is, that the participants in the FC-workbook and FC-group interventions were not significantly different from each other on study outcomes, thus enhancing the statistical power for comparisons among groups.
Am J Prev Med 2009;36(1)
39
48.80 (39.72) 0.97 (0.80)b,c 1.30 (0.97) 6.66 (3.10) p⬍0.05 for BL– 6 months p⬍0.05 for BL–12 months c p⬍0.05 for 6 –12 months BL, baseline; IVR, interactive voice response; KEDS, Kids’ Eating Disorders Survey; mo, months; MPA, moderate physical activity; VPA, vigorous physical activity b
BMI z-score VPA self-report MPA self-report Sedentary/ screen time Sugared drinks Fruits Vegetables Total KEDS
a
40.80 (33.50)a 1.52 (1.40) 1.23 (0.89) 6.57 (2.80) 61.66 (46.56) 1.52 (1.21) 1.32 (1.11) 7.11 (2.68) 86.68 (93.42) 1.97 (3.04) 1.61 (2.02) 6.68 (4.24) 91.95 (68.89) 2.38 (3.56) 1.78 (2.49) 6.74 (3.36) 112.16 (135.75) 1.26 (1.10) 1.51 (1.46) 6.26 (2.64) 62.18 (54.98) 1.03 (0.87) 1.31 (1.49) 6.92 (3.29) 78.92 (79.17) 1.10 (0.86) 1.29 (1.52) 7.40 (3.28)
2.00 (0.04) 4.16 (2.05) 2.56 (2.02) 5.55 (2.45)
2.00 (0.04) 4.49 (1.92) 2.62 (2.33) 5.46 (2.41) 2.04 (0.04) 4.12 (2.14) 2.32 (2.21) 5.16 (2.16)
71.80 (81.11) 1.18 (1.19) 1.47 (1.67) 7.08 (3.46)
1.90 (0.06)b,c 3.70 (2.21) 2.57 (2.18)b 5.43 (2.22) 2.08 (0.07) 3.50 (2.01) 2.35 (1.87) 5.60 (2.54)
2.07 (0.08) 3.40 (2.14) 2.40 (2.14) 6.25 (2.71)
2.09 (0.08) 3.05 (1.99) 2.90 (2.49) 5.65 (1.84)
2.03 (0.05) 3.89 (2.70) 1.65 (2.11) 5.43 (2.41)
1.94 (0.06) 3.59 (2.27) 2.89 (2.54)a 5.24 (1.69)
12 mo
a
6 mo BL
6 mo
a
12 mo
b
BL
6 mo
12 mo
BL
FC-IVR: 6–10 calls nⴝ38 FC-IVR: 0–5 calls nⴝ20 FC-workbook and FC-group nⴝ82
Table 3. Variable M (SD) by IVR dose 40
Participants in the group FC-IVR: 6 –10 calls reported a significant increase in the number of days they participated in MPA from baseline to 6 months and from baseline to 12 months. However, the changes in physical activity were not significantly different among groups. Again, changes in fruit, vegetable, and sugareddrink consumption did not demonstrate a consistent pattern of changes within or among groups.
Discussion This is the first trial to attempt to achieve clinical weight management using primarily automated technology to engage the parent as a proxy in the treatment of overweight children. Exploratory analyses of withincondition effects for the FC-IVR demonstrated improvements at both 6 months and 12 months, but between-group differences were not significant. It was found that the dose of the FC-IVR intervention may be important; a higher dose of completed parental IVR calls was associated with increased MPA and decreased BMI z-scores in those children. However, dose was not systematically varied within this study, and both the increased dose and improved BMI z-scores could be indicators of the parent level of motivation to make changes to support an overweight child in living a healthier lifestyle. Still, both motivation and skill development are necessary to achieve behavior change,60 so it could be argued that motivated parents were more likely to receive a higher level of training on skill development through the increased number of IVR calls that were completed. From a scientific perspective, this provides an avenue for future effectiveness research: to determine if varying levels of dose, within an RCT, can influence changes in parental skill development and child BMI z-scores. From a practical perspective, the information from the present study suggests that if FC-IVR was offered in a managed-care setting, approximately half of the parents who received the intervention would complete at least six calls and achieve a significant reduction in BMI z-scores for their overweight children. The frequency and duration of intervention contacts may also have an influence on the magnitude of the effect achieved. For example, recent studies by Savoye and colleagues31 and Epstein et al.30 reported larger effects on child BMI z-scores than this trial. Epstein’s trials typically included anywhere from 14 to 20 face-toface sessions for children and parents. In the case of the Savoye study,31 parents and children attended two face-to-face classes per week for 6 months and an additional class every 2 weeks for the next 6 months. This totaled 60 sessions over a 12-month period. The resulting conclusion could be that very intensive interventions can achieve significant reductions in BMI z-scores, while FC-IVR with two face-to-face sessions and
American Journal of Preventive Medicine, Volume 36, Number 1
www.ajpm-online.net
at least six follow-up counseling calls may be more likely to achieve modest reductions in BMI z-scores. Conversely, the frequency and duration of intervention contacts may influence the likelihood that a given intervention is disseminated and used in typical practice.61 When this study’s FC-IVR intervention is compared to, for example, the trial by Savoye et al.,31 it is clear, simply from a human resources perspective, that the FC-IVR intervention could reach a far greater number of families. In the one instance,31 one small group of overweight children and parents engage in an intensive, 60-session intervention with healthcare personnel over the course of a year. In terms of this study’s FC-IVR intervention, 60 face-to-face sessions would allow for the engagement of 30 small groups of overweight children and parents (i.e., two small-group sessions followed by IVR for each small group)— roughly a 30:1 advantage in potential reach with identical resources. Our intent is not to suggest that a clinical weightmanagement program for children should be deemed satisfactory by simply achieving a very modest decrease in BMI z-scores. Rather, it is to point to a potentially fruitful area of future practice that may utilize a theorybased IVR intervention to reduce the systemic cost of treating families with overweight children and increase the reach of efficacious interventions into the population at large. It is noteworthy that Calls 7–10 for FC-IVR provided messages that reinforced the basic skill-building activities and information provided during earlier intervention sessions. This study’s results suggest that these calls may be important to support decreases in BMI z-scores. This study has some limitations. First, parents were not assigned randomly to a higher or lower frequency of FC-IVR. Thus, it could be that those parents who completed all of the calls were more motivated than those who did not. Second, many children declined to wear an accelerometer, so an objective assessment of physical activity was not possible, and a very brief selfreport measure was used to reduce participant burden. Third, the cost of delivery of each of the interventions was not tracked and, therefore, information cannot be provided relative to the differential effects based on the cost of delivery, tracking, and patient retention across intervention conditions. In conclusion, the completions of Calls 6 –10 of the FC-IVR demonstrated a relative advantage over standard care. While doing so, the intervention did not have any adverse consequences on child body image or eating-disorder symptoms. The study also suggests that Golan’s social– ecologic model12 can be used to tailor an automated intervention for pediatric weight management. Conversely, the clinical significance of a small reduction in BMI z-scores, like that seen in this study, is unclear. Finally, future research should focus on understanding the optimal integration of automated telephone January 2009
counseling with human intervention strategies to improve the magnitude of effect and the potential for broad reach while monitoring the cost and cost effectiveness of pediatric weight-management programs. This work was supported by the Garfield Memorial Fund, an internal funding mechanism of Kaiser Permanente Health Plans (Project #50-227). Also, the Kaiser Permanente Colorado Weight Management Program, Michele Gilson, and Helen Seagle were integral in designing the study as a practical RCT and developing the original Family Connections content. No financial disclosures were reported by the authors of this paper.
References 1. Jolliffe D. Extent of overweight among U.S. children and adolescents from 1971 to 2000. Int J Obes 2004;28:4 –9. 2. Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH. Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med 1997;337:869 –73. 3. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the U.S., 1999 –2004. JAMA 2006;295:1549 –55. 4. Bloomgarden ZT. Prevention of obesity and diabetes. Diabetes Care 2003;26:3172– 8. 5. American Academy of Pediatrics. Policy statement: prevention of pediatric overweight and obesity. Pediatrics 2003;112:424 –30. 6. Estabrooks PA, Shetterly S. The prevalence and health care use of overweight children in an integrated health care system. Arch Pediatr Adolesc Med 2007;161:222–7. 7. Golan M, Crow S. Targeting parents exclusively in the treatment of childhood obesity: long-term results. Obes Res 2004;12:357– 61. 8. Golan M, Crow S. Parents are key players in the prevention and treatment of weight related problems. Annu Rev Nutr 2004;62:39 –50. 9. Golan M, Weizman A. Familial approach to the treatment of childhood obesity: conceptual model. J Nutr Educ 2001;33:102–7. 10. Golan M, Weizman A, Fainaru M. Impact of treatment for children obesity on parental risk factors for cardiovascular disease. Prev Med 1999;29: 519 –26. 11. Golan M, Weizman A, Apter A, Fainaru M. Parents as the exclusive agents of change in the treatment of childhood obesity. Am J Clin Nutr 1998;67:1130 –5. 12. Golan M, Fainaru M, Weizman A. Role of behaviour modification in the treatment of childhood obesity with parents as the exclusive agents of change. Int J Obes 1998;22:1217–24. 13. Coleman KJ, Epstein LH. Application of generalizability theory to measurement of activity in males who are not regularly active: a preliminary report. Res Q Exerc Sport 1998;89:58 – 63. 14. Epstein LH. Exercise and obesity in children. J Appl Sport Psychol 1992; 4:120 –33. 15. Epstein LH, Coleman KJ, Myers MD. Exercise in treating obesity in children and adolescents. Med Sci Sports Exerc 1996;28:428 –35. 16. Epstein LH, Goldfield GS. Physical activity in the treatment of childhood overweight and obesity: current evidence and research issues. Med Sci Sports Exerc 1999;31(11S):S553–9. 17. Epstein LH, Myers MD, Raynor HA, Saelens BE. Treatment of pediatric obesity. Pediatrics 1998;101:554 –70. 18. Epstein LH, McCurley J, Wing RR, Valoski A. Five-year follow-up of family-based behavioral treatments for childhood obesity. J Consul Clin Psychol 1990;58:661– 4. 19. Epstein LH, Valoski A, Wing RR, McCurley J. Ten-year follow-up of behavioral family-based treatment for obese children. JAMA 1990;264: 2519 –23. 20. Epstein LH, Valoski A, Vara LS, et al. Effects of decreasing sedentary behavior and increasing activity on weight change in obese children. Health Psychol 1995;14:109 –15.
Am J Prev Med 2009;36(1)
41
21. Epstein LH, Wing RR, Valoski A, DeVos D. Long-term relationship between weight and aerobic-fitness change in children. Health Psychol 1988; 7:47–53. 22. Epstein LH, Wing RR, Steranchak L, Dickson B, Michelson J. Comparison of family-based behavior modification and nutrition education for childhood obesity. J Pediatr Psychol 1980;5:25–36. 23. Epstein LH, Wing RR, Woodall K, Penner BC, Kress MJ, Koeske R. Effects of family-based behavioral treatment on obese 5-to-8 year old children. Behav Ther 1985;16:205–12. 24. Epstein LH, McKenzie SJ, Valoski A, Klein KR, Wing RR. Effects of mastery criteria and contingent reinforcement for family-based child weight control. Addict Behav 1994;19:135– 45. 25. Epstein LH, Paluch RA, Raynor HA. Sex differences in obese children and siblings in family-based obesity treatment. Obes Res 2001;9:746 –53. 26. Epstein LH, Roemmich JN, Raynor HA. Childhood and adolescent obesity: behavioral therapy in the treatment of pediatric obesity. Pediatr Clin Nor Am 2001;48:981–93. 27. Epstein LH, Paluch RA, Kilanowski CK, Raynor HA. The effect of reinforcement or stimulus control to reduce sedentary behavior in the treatment of pediatric obesity. Health Psychol 2004;23:371– 80. 28. Epstein LH, Paluch RA, Gordy CC, Saelens BE, Ernst MM. Problem solving in the treatment of childhood obesity. J Consult Clin Psychol 2000;68: 717–21. 29. Epstein MH, Valoski A, Wing RR, McCurley J. Ten-year outcomes of behavioral family-based treatment for childhood obesity. Health Psychol 1994;13:373– 83. 30. 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;26:381–91. 31. Savoye M, Shaw M, Dziura J, et al. Effect of a weight management program on body composition and metabolic parameters in overweight children: a randomized controlled trial. JAMA 2007;297:2697–704. 32. Golan M, Weizman A, Apter A, Fainaru M. Parents as the exclusive agents of change in the treatment of childhood obesity. Am J Clin Nutr 1998; 67:1130 –5. 33. Edmunds L, Waters E, Elliot EJ. Evidence based management of childhood obesity. BMJ 2001;323:916 –9. 34. Estabrooks P, Dzewaltowski DA, Glasgow RE, Klesges LM. School-based health promotion: issues related to translating research into practice. J Sch Health 2003;73:28. 35. Barlow SE, Dietz WH. Management of child and adolescent obesity: summary and recommendations based on reports from pediatricians, pediatric nurse practitioners, and registered dietitians. Pediatrics 2002;110:236 – 8. 36. Glasgow RE, Bull SS, Gillette C, Klesges LM, Dzewaltowski DA. Behavior change intervention research in health care settings: a review of recent reports, with emphasis on external validity. Am J Prev Med 2002;23:62–9. 37. Biem HJ, Turnell RW, D’Arcy C. Computer telephony: automated calls for medical care. Clin Invest Med 2003;26:259 – 68. 38. Piette JD. Patient education via automated calls: a study of English and Spanish speakers with diabetes. Am J Prev Med 2000;17:138 – 41. 39. Piette JD, McPhee SJ, Weinberger M, Mah CA, Kraemer FB. Use of automated telephone disease management calls in an ethnically diverse sample of low-income patients with diabetes. Diabetes Care 1999;22: 1302–9. 40. Piette J. Enhancing support via interactive technologies. Curr Diab Rep 2002;2:160 –5. 41. Pinto BM, Friedman R, Marcus BH, Kelley H, Tennstedt S, Gillman MW. Effects of a computer-based, telephone-counseling system on physical activity. Am J Prev Med 2002;23:113–20.
42
42. Delichatosios H, Friedman R, Glanz K, et al. Randomized trial of a “talking computer” to improve adults’ eating habits. Am J Health Promot 2001; 15:215–24. 43. Krishna S, Balas EA, Boren SA, Maglaveras N. Patient acceptance of voice messages: a review of controlled clinical studies. Methods Inf Med 2002;41:360 –9. 44. Mahoney DF, Tarlow BJ, Jones RN. Effects of an automated telephone support system on caregiver burden and anxiety: findings from the REACH for TLC intervention study. Gerontologist 2003;43:556 – 67. 45. Tunis SR, Stryer DB, Clancy CM. Practical clinical trials: increasing the value of clinical research for decision making in clinical and health policy. JAMA 2003;290:1624 –32. 46. Must A, Dallal GE, Dietz WH. Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2) and triceps skinfold thickness. Am J Clin Nutr 1991;53:839 – 46. 47. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts for the U.S.: methods and development. Vital Health Stat 2002:11;1–190. 48. Wilfley DE, Stein RI, Saelens BE, et al. Efficacy of maintenance treatment approaches for childhood overweight: a randomized controlled trial. JAMA 2007;298:1661–73. 49. Troped P, Wiecha J, Fragala M, et al. Reliability and validity of YRBS physical activity items among middle school students. Med Sci Sports Exerc 2007;39:416 –25. 50. Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardener L. A data-based approach to diet questionnaire design and testing. Am J Epidemiol 1986;124:453– 69. 51. Block G. Health habits and history questionnaire. Program evaluation handbook: nutrition education. Los Angeles: IOX Assessment Associates, 1988. 52. Childress AC, Jarrell MP, Brewerton TD. The kids’ eating disorders survey (KEDS): internal consistency, component analysis, and reliability. Eating Disorders: The Journal of Treatment and Prevention 1993;1:123–33. 53. Epstein L, Gordy C, Raynor H, Beddome M, Kilanowski C, Paluch R. Increasing fruit and vegetable intake and decreasing fat and sugar intake in families at risk for childhood obesity. Obes Res 2001;9:171– 8. 54. Young K, Northern J, Lister K, Drummond J, O’Brien W. A meta-analysis of family-behavioral weight-loss treatments for children. Clin Psychol Rev 2007;27:240 –9. 55. Estabrooks PA, Glasgow RE, Dzewaltowski DA. Physical activity promotion through primary care. JAMA 2003;289:2913– 6. 56. Laird NM, Ware JH. Random effects models for longitudinal data. Biometrics 1982;38:963–74. 57. McCullagh P, Nelter JA. Generalized linear models. Boca Raton FL: Chapman Hall/CRC, 1989. 58. Wolfinger R, O’Connell M. Generalized linear mixed models: a pseudolikelihood approach. Journal of Statistical Compututation and Simulation 1993;48:233– 43. 59. Diggle PJ, Liang KY, Zeger SL. Analysis of longitudinal data. Oxford: Oxford University Press, 1994. 60. Bandura A. Self-efficacy: the exercise of control. New York: Freeman, 1997. 61. Rogers EM. Diffusion of innovations. 5th ed. New York: Free Press, 2003.
Supplementary data Supplementary data associated with this article can be found, in the online version, at 10.1016/j.amepre.2008.09.010.
American Journal of Preventive Medicine, Volume 36, Number 1
www.ajpm-online.net
Appendix A. General intervention framework, strategies, and activities Concept
Strategy
Example of potential activity
Increased parental behavioral health skills Increased parenting skills
Enhance parental personal motivation related to weight, nutrition, and physical activity
Demonstrate personal goal setting and barrier resolution related to physical activity and healthful eating Demonstrate parental responsibilities and role in child’s weight (i.e., not the child’s fault) Demonstrate parental leadership, general parenting skills, and effective parent–child communication Model enjoyment of healthy foods and physical activity Engage with child in physical activities Practice regular meal times and schedule between-meal snacks Involve child in decision making for enjoyable physical activity
Reframe the problem Set healthy limits
Parental modeling
Parents provide an example of healthy lifestyle behaviors for child
Home environmental changes
Restructure the home environment to support healthy food and activity options while reducing options for unhealthy choices
Appendix B. Breakdown of workbook content areas by day Day
Part 1
Part 2
1
● Assessing and calculating BMI in children and adults ● Causes of overweight in children (biological, cognitive, environmental ● Five reasons children gain weight
2
● Impact of being overweight ● Helping children to avoid fad diets ● Activity to explore parental beliefs about eating and physical activity ● Healthy habits for creating a healthy family ● Defining the division of responsibility for eating and activity ● Parenting skills to support weight reduction ● Survey of the family home environment
● Ways to promote a healthy home environment ● Nutrition: reading labels, selecting healthy food options ● Physical activity: using the FITT (frequency, intensity, time, type) principles ● Sample menus ● Tips for preparing healthy snacks and meals ● Goal setting: creating a family action plan ● Process of goal setting and keeping objectives clear ● Parent’s personal action plan ● Barriers and strategies to maintaining family action plan ● Resources
3
Am J Prev Med 2009;36(1)
42.e1
Appendix C. Conceptual and content bases of interactive voice response follow-up calls Call timing (# weeks post final group session)
Target concept
Content and sample goals
1
Parenting skills
2
Behavioral skills
3
Home environment
4
Role modeling
6
Keeping motivated
8
Part 1: Goal setting using the 5A’s (assess, advise, ask, assist, arrange)
10 12
Part 2: Goal setting using the 5A’s (assess, advise, ask, assist, arrange) Community resources
16
Kids speak up!
20
Relapse prevention
Consistency and contingency: Be consistent in limiting the amount of television your child watches this week to ⬍2 hours/day. Communication: Talk to your child about the types of physical activity he or she likes to do. Healthy habits: Set a good example for your child by eating 5–9 servings of fruits and vegetables every day this week. Praise: Praise your child whenever he or she attempts to be physically active. Physical activity: Get at least 2.5 hours of moderate physical activity this week. Nutrition: Drink low-fat milk at one meal each day. Stress: Take 10-minute stress-break walks at least once per day. Physical activity: Make a list of your family’s barriers to physical activity around the house, and come up with a way to overcome them. Nutrition: Clear the kitchen cupboards of unhealthy snacks. Physical activity: Do something active with your child for 15 minutes at least three times this week. Nutrition: Prepare at least one healthy meal together with your child. Physical activity: Review the family goal sheet with your family and set measurable, specific, and objective goals this week. Nutrition: Post signs at least twice this week about your family’s positive changes with healthy eating. Sedentary behavior: Decrease your sitting time by 0.5 hours each day this week. Nutrition: Decrease your soda and sugared-drink consumption by one serving per day. Physical activity: Increase your physical activity by 1 day per week. Nutrition: Increase your servings of fruits and vegetables by three per day. Physical activity: Set a goal to take your family to a new park or trail that you haven’t visited before. Nutrition: Check your library for cookbooks, recipes, or videos that help your family to prepare nutritious meals. Parenting skills: Set a goal to make a physical activity resource kit and use it when your family is out and about or at home. Physical activity: Survey your family regarding physical activity that they would like to do as a family, and try to do that activity at least three times per week. Nutrition: Change one food item that is high-fat to a healthy snack of fruit or vegetable. Support: Contact a weight-management specialist if any problems arise. Plan: Complete the 5A’s goal-setting process every 3 months to keep your family working on healthful eating and physical activity. Commitment: Commit to your family’s new healthy lifestyle for at least the next full year.
42.e2
American Journal of Preventive Medicine, Volume 36, Number 1
A practical RCT evaluated the relative effectiveness of three interventions to support parents of overweight or at-risk children to change the home environment to foster more healthful child eating and activity behaviors, thereby reducing child BMI and BMI z-scores. A secondary purpose was to determine the patterns of use and potential dose effect for the highest-intensity intervention.
Setting/ Parent-and-child (aged 8 –12 years) dyads (N⫽220) who received care from Kaiser participants: Permanente Colorado were assigned randomly to one of the three Family Connections (FC) interventions: FC-workbook, FC-group, or FC-interactive voice response (IVR) counseling. Main outcome measures:
Child BMI z-scores, as well as symptoms of eating disorders and body image, were assessed at baseline, 6 months, and 12 months.
Results:
The BMI z-scores of children assigned to the FC-IVR intervention were the only ones that decreased from baseline to 6 months (0.07 SD) and from baseline to 12 months (0.08 SD, p⬍0.05). Children whose parents completed at least six of the ten FC-IVR counseling calls had decreased BMI z-scores to a greater extent than children in the FC-workbook or FC-group interventions at both 6 months (p⬍0.05) and 12 months (p⬍0.01). No intervention increased child symptoms of eating disorders or body dissatisfaction at any time point.
Conclusions: This trial demonstrated that automated telephone counseling can support the parents of overweight children to reduce the extent to which their children are overweight. Trial registration:
NCT00433901 (Am J Prev Med 2009;36(1):35– 42) © 2009 Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine.
Introduction
O
ver the previous 3 decades, the prevalence of overweight children and the degree of overweight have increased significantly,1–3 and overweight children, while still in their youth, are at risk for hypokinetic diseases and much more likely to seek mental From the Department of Human Nutrition, Foods, and Exercise, Virginia Polytechnic Institute and State University (Estabrooks), Blacksburg, Virginia; and the Institute of Health Research, Kaiser Permanente Colorado (Estabrooks, Shoup, Gattshall, Dandamudi, Shetterly, Xu), Denver, Colorado Address correspondence and reprint requests to: Paul A. Estabrooks, PhD, Virginia Tech Riverside, 1 Riverside Circle SW, Suite #104, Roanoke VA 24016. E-mail: [email protected].
health care than healthy weight children.4 – 6 The American Academy of Pediatrics (the Academy) has released recommendations5 highlighting the need to promote healthful eating and physical activity while reducing sedentary behavior and sugared-drink consumption through behavioral strategies documented in studies of intensive family-based behavioral therapy.7–32 Although the research on familial interventions targeting childhood overweight issues is convincing, there is little evidence that these programs have been translated into regular practice or reach large segments of the population in need.33–35 This may be because it is questionable whether interventions tested in previous efficacy trials use resources and staff that are typical of
Am J Prev Med 2009;36(1) © 2009 Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine
0749-3797/09/$–see front matter doi:10.1016/j.amepre.2008.09.024
35
those in usual care.36 An uninvestigated area of research is the potential of alternative models of intervention delivery, such as interactive technologies, to achieve effects similar to previous efficacy trials while requiring fewer system resources. Interactive voice response (IVR) technology may be a viable avenue for delivering pediatric weight management–intervention strategies to parents of children who are overweight. This technology has been used in a number of healthcare contexts and has demonstrated effectiveness in enhancing physical activity and healthy eating in diabetic and older adult samples.37– 42 Further, a review of the literature indicates that IVR care was acceptable and helpful to patients,43 cost effective in maintaining diabetes selfmanagement,38,43 and supportive of caregivers in other contexts.44 The present study reflects a research–practice partnership with a goal to develop a practical and sustainable intervention based on previous efficacy trials.7–12 To achieve this goal, a practical RCT was designed to test interventions—all viable candidates for future implementation—that varied by frequency (i.e., one contact, three contacts, or 13 contacts) and channel of delivery. The specific purpose of the study was to determine the relative effectiveness of delivering an intervention through an additive process that included (1) a self-help Family Connections workbook for parents (FC-workbook); (2) the workbook plus two smallgroup sessions with a registered dietitian (FC-group); and (3) workbook, two small-group sessions, and ten automated IVR tailored-counseling sessions (FC-IVR). As there is little information on the extent to which parents would use IVR or whether parents could be engaged long enough to receive the core content of the intervention before discontinuing use, a second twofold purpose was to determine what percentage of the parents assigned to the FC-IVR intervention would complete more than half of the sessions, thus receiving the core content, and to determine if the ⬎50% dose of the intervention heightened effectiveness.
2007, and all analyses were performed from May 2007 through December 2007. A three-group practical controlled-trial design was used to allow future decision-makers with the appropriate data to determine whether a new intervention significantly surpasses the effect of the current standard care (i.e., FC-IVR versus FC-group) or matches the effect while utilizing fewer resources (i.e., FC-workbook versus FC-group).45 Through a random-numbers table, participants were assigned randomly (families/staff unblinded) to the FC-workbook, the FC-group, or the FC-IVR intervention. It was hypothesized that both the FC-group and FC-IVR interventions would demonstrate medium effects when compared to the FC-workbook intervention but would demonstrate small effects when compared to each other. Sample size calculations were completed, varying the detectable effect sizes from small to medium with a power of 0.8. The result was a need for 42 participants per intervention to detect a medium effect and 64 participants to detect a small effect. Therefore, more participants were assigned randomly to the FC-group and FC-IVR interventions to increase the statistical power for comparisons between those interventions and among participants who completed half of the intervention relative to FC-workbook and FC-group. Finally, an a priori alternate hypothesis was developed: that the FC-workbook and FC-group interventions were insufficient in either duration or frequency to achieve significant effects. In that case, collapsing the FC-workbook and FCgroup participants into a single group would enhance the power to detect a smaller effect with the FC-IVR participants.
Participants Children aged 8 –12 years with a BMI ⱖ85th percentile for their age were eligible for the study. Exclusion criteria included plans to move out of the state during the course of the study or a request by the child’s pediatrician that the family not be contacted. The initial pool of potential study participants included 1487 families (Figure 1). Of these, 656 parents were contacted during the recruitment phase of the study. Two-hundred and twenty families completed the baseline visit and subsequently were assigned randomly to FCworkbook (n⫽50); FC-group (n⫽85); or FC-IVR (n⫽85). The participation level of eligible families who were contacted by telephone was approximately 38%. Table 1 provides baseline child characteristics.
Measures
Methods Design and Procedures The Kaiser Permanente Colorado IRB approved this study in March 2004. All study participants signed parental consent or child assent forms as well as authorization for the use of medical records. Parents of overweight children were identified through electronic medical records, contacted by telephone, and invited to participate in the study. Enrollment for the study began in May 2004 and continued through January 2006. All participants completed an assessment at baseline, 6 months, and 12 months. Participants were followed for a period of 1 year. Follow-up measurement and assessment began in November 2004 and continued through February
36
Demographic information was obtained through medical records and parent reports. Children’s age and gender were verified by parents. Parent and child race/ethnicity information was provided by participating parents. BMI z-scores. Height was measured using a stadiometer, and weight was assessed using a regularly calibrated medical scale.27 BMI scores were calculated based on the value of the 50th-percentile BMI ranking and the SD attributable to appropriate age and gender samples from the CDC growth charts.46 Code available from the CDC (www.cdc.gov) was used to accurately calculate associated BMI z-scores (i.e., the number of SD units that the child’s BMI deviates from the mean reference value for age and gender) using a modified least mean square technique that approximates z-scores from
American Journal of Preventive Medicine, Volume 36, Number 1
www.ajpm-online.net
used as a screening tool for eating disorders in children.52 Individual responses were categorized into weight dissatisfaction, restricting/purging, binge eating, and body dissatisfaction components. The scale has high internal consistency.52
Interventions A model proposed by Golan and associates7,9,12,32— based on a social– ecologic theory that promotes changes to the home environment, parenting skills, role modeling, and positive parental health behaviors as primary factors in pediatric weight management and consistent with the Academy’s recommendations5— was used as the basis for all intervention conditions. The model stresses a health-centered approach and targets the parent exclusively as the agent of change. Interventions based on Golan’s model and Figure 1. Flow chart of participant identification and recruitment tested across a number of studies7–12,32 have demonstrated short- and long-term growth curves estimated from the 2000 National Health and 47 effectiveness in reducing child BMI and parental cardiovascular Nutrition Examination Survey (NHANES) data. This outdisease risk factors while also reducing risky eating behaviors in come was selected because it provides a comparison of children. The model also receives indirect support from studies relative decrease in the extent of overweight to the referent that have used similar family-based approaches and have dempopulation based on age and gender, and is one of the most 48 onstrated that they are effective,53 replicable,30 and produce frequently reported relative weight outcomes for children. large and reliable effects over time.54 See Appendix A, available Physical activity and sedentary behavior (Youth Behavioral online at www.ajpm-online.net, for examples of strategies develRisk Survey questions49). Children indicated the number of oped based on this model. days in the previous 7 days on which they completed moderFC-workbook. A 61-page workbook was developed to proate physical activity (MPA), vigorous physical activity (VPA), or both. Sedentary behavior was assessed based on the mote increased physical activity and the consumption of fruits numbers of hours of screen time during school days. The and vegetables in concert with decreased sugared-drink conscale reliability was acceptable. The scale does underestimate sumption and television viewing/recreational computer time. the days of MPA and overestimates the days of VPA.49 The workbook included two distinct sections. Part 1 targeted 3 days of intervention, and Part 2 targeted 2 days, each with Fruit, vegetable, and sugared-drink consumption. Each parspecific homework assignments (Appendix B, www.ajpmticipant completed the Block Kids Questionnaire, a foodonline.net). The workbook encouraged parents to complete frequency questionnaire for children and youth aged 8 –17 50,51 all 5 days of intervention across a single week. Homework years. The questionnaire was developed from the assignments were intended to encourage lasting changes in NHANES 1999 –2002 dietary recall data, and the nutrient the families. All parents randomly assigned to this intervendatabase was developed from the U.S. Department of Agrition received the workbook from study research assistants. culture nutrient database, version 1.0, for dietary studies. The food-frequency questionnaire asks individual portion sizes, FC-group. This intervention consisted of two small-group allowing for the computation of fruit, vegetable, and sugaredsessions (2 hours each, spaced 1 week apart) held at a local 50,51 drink servings. clinic and delivered by a dietitian. Each session included 10 –15 parents representing distinct children and utilized the Eating-disorder symptoms. The Kids’ Eating Disorders Survey Family Connections workbook. The first session focused on (KEDS) is a self-report questionnaire for children aged ⱖ8 years
January 2009
Am J Prev Med 2009;36(1)
37
Table 1. Participant characteristics at baseline Intervention
Age (M in years) Gender (%) Male Race/ethnicity (%) White Hispanic 6-month visit completed 12-month visit completed BMI (M kg/m2) BMI z-score (M) VPA (self-reported M days/week) MPA (self-reported M days/week) Sedentary/screen time (M hours/day) Sugared drinks (M ounces/week) Fruits (M servings/day) Vegetables (M servings/day) KEDS (total score M)
Total sample Nⴝ220
FC-workbook nⴝ50
FC-group nⴝ85
FC-group and FC-IVR nⴝ85
10.7
11.0
10.6
10.7
54
39
58
59
63 26 78 71 27.2 2.04 4.0 2.3 5.5 85.3 1.2 1.3 7.3
59 29 78 74 27.1 2.00 4.0 2.4 5.3 76.6 1.2 1.6 7.8
69 19 75 66 27.4 2.07 4.3 2.6 5.4 92.5 1.0 1.0a 7.3
60 30 80 74 27.1 2.04 3.8 2.0 5.6 83.3 1.3 1.4 7.0
p⬍0.05 for contrasts by study arm (2 for percents, ANOVA for continuous variables [Kruskal–Wallis tests significance levels comparable]). Participants in the FC-group reported consuming significantly fewer vegetables per day at baseline. FC, Family Connections; IVR, interactive voice response; KEDS, Kids’ Eating Disorders Survey; MPA, moderate physical activity; VPA, vigorous physical activity
a
parents’ behavioral health skills and knowledge of weight, nutrition, and physical activity. It also identified key parenting skills: limit setting, effective communication, and role modeling. This session concluded with role playing, problem solving, and the development of an action plan. Session 2 integrated the knowledge acquired in Session 1, the experiences associated with the action plan, and strategies for restructuring the home environment. The session again concluded with parents’ completing an action plan for parental behaviors, role modeling, and changes to the home environment that would facilitate healthy eating and physical activity. FC-IVR. Parents assigned to this intervention had completed the two-session Family Connections small-group program and were subsequently provided with ten follow-up sessions delivered via IVR. Telephone follow-up calls commenced 1 week after the second small-group session. The participants could either call the IVR system proactively or wait for the system to call at the designated time. Participant responses to IVR questions and branching logic were used to determine the content of each call. At the beginning of each call, the primary parent would hear the goal selected the previous week and rate his or her achievement of that goal. Based on the rating, the parent was given the option of hearing tips related to the topic of the prior week’s goal. Following this, the parent was able to select specific messages related to the intervention content. Each call concluded with a goal-setting procedure. See Appendix C, at www.ajpm-online.net for both descriptions of strategies that were provided via IVR and the timing of follow-up contacts. The sixth IVR-counseling call provided parents with instruction on a family goal-setting procedure based on the 5A’s model (Assess behaviors and motivation, Advise on healthy behaviors, Ask about goals and reach collaborative agreement, Assist with barrier identification and resolution, and
38
Arrange follow-up).55 During this call, parents were trained to lead their family through regular goal setting related to physical activity and eating. This sixth call was considered the final component of the core content of the intervention. Calls 7–10 reinforced the information delivered in the initial six calls.
Statistical Analyses Intent-to-treat analyses were completed across three time points (baseline, 6-month follow-up, and 12-month follow-up) to determine changes in outcomes over time. Linear mixed models56 were fitted to account for the correlation of repeated measures, assuming that any missing data were missing at random. Preplanned linear contrasts were used to determine the presence of changes across study outcomes at (1) baseline and 6 months, (2) 6 months and 12 months, and (3) baseline and 12 months. These analyses were followed by a comparison of changes among groups across time. Mixed models were used for changes in BMI z-scores and square root–transformed dietary variables. Physical activities and sedentary/screen time were measured as rates: days of physical activity per week and hours of sedentary/screen time per day. These rates were not normally distributed; therefore, nonlinear random-effect models57,58 were used to fit these rates with logit link function and binomial distribution.59 All statistical analyses were performed using SAS version 9.1. To examine the second purpose, an a priori decision was made to complete a secondary analysis comparing the participants who completed six or more of the IVR sessions to participants who (1) completed fewer than six sessions, (2) received the FC-workbook, and (3) received the FC-group interventions. Two analytic plans were developed. One plan included the potential to examine the differences among four newly formed conditions (FC-workbook; FC-group; FCIVR: 0 –5 calls; FC-IVR: 6 –10 calls) based on IVR completion.
American Journal of Preventive Medicine, Volume 36, Number 1
www.ajpm-online.net
January 2009
p⬍0.05 for BL–12 months p⬍0.05 for BL– 6 months c p⬍0.05 for 6 –12 months BL, baseline; IVR, interactive voice response; KEDS, Kids’ Eating Disorders Survey; mo, months; MPA, moderate physical activity; VPA, vigorous physical activity b
a
12 mo 6 mo BL
6 mo
12 mo
BL
FC-group (SE) FC-workbook (SE)
Table 2. Variable M (SD) according to randomized condition
Fifty-four percent of the recruited children were boys (average age⫽10.7 years, SD ⫾1.4); 63% of the recruited children were white, while 26% were Latino. Participants were primarily overweight rather than at risk of overweight (average BMI percentile⫽97.3, SD ⫾2.3). The intervention conditions did not differ on any demographic variables. Table 2 includes the descriptive and within-group intent-to-treat comparisons across study groups. Children in the FC-group significantly decreased their BMI z-scores from baseline to 6 months, but the decrease was not sustained at 12 months. Participants in the FC-IVR conditions significantly decreased their BMI z-scores from baseline to 6 months and baseline to 12 months, while those in the FC-workbook intervention significantly decreased their BMI z-scores from baseline to 12 months. There were no differences among conditions. Participants in FC-IVR reported a significant increase in the number of days they participated in MPA from baseline to 6 months and baseline to 12 months. Changes in fruit, vegetable, and sugared-drink consumption did not demonstrate a consistent pattern of change within or among groups. Regardless of the intervention condition, all children reported healthy behaviors in response to the KEDS, and no increases in unhealthy behaviors were detected over the course of the study. Table 3 includes the descriptive and within-group comparisons based on the secondary purpose to examine differences in effect according to IVR completion. Participants in the FC-IVR intervention were divided into two groups: those who completed 0 –5 calls (n⫽20) and those who completed 6 –10 calls (n⫽38). Participants in the group FC-IVR: 6 –10 calls were the only ones with significantly decreased BMI z-scores from baseline to 6 months and from 6 months to 12 months. Between-group comparisons demonstrated that children in the group FC-IVR: 6 –10 calls significantly decreased BMI z-scores when compared to those assigned to the FC-workbook or FC-group condition at both 6 months (F[3,148]⫽–1.11; p⬍0.05) and 12 months (F[3,148]⫽–2.89; p⬍0.01).
BL
Results
n 49 38 36 85 64 56 85 68 63 BMI z-score 2.04 (0.02) 1.99 (0.03) 1.98 (0.03)a 2.06 (0.04) 2.03 (0.04)b 2.04 (0.04) 2.03 (0.04) 1.96 (0.04)b 1.95 (0.04)a VPA self-report (days/week) 4.02 (2.07) 3.97 (1.94) 4.32 (2.13) 4.30 (2.08) 4.44 (2.04) 4.04 (2.05) 3.75 (2.43) 3.57 (2.24) 3.47 (2.09) MPA self-report (days/week) 2.36 (2.32) 2.47 (2.10) 2.79 (1.95) 2.56 (2.26) 2.82 (2.35) 2.36 (2.01) 2.01 (2.18) 2.64 (2.35)b 2.71 (2.21)a Sedentary/screen time 5.27 (2.15) 5.00 (2.05) 5.47 (1.96) 5.39 (2.22) 5.49 (2.55) 5.64 (2.61) 5.63 (2.54) 5.70 (2.20) 5.60 (2.04) (hours/day) Sugared drinks (ounces/week) 76.63 (83.68) 50.40 (42.35) 60.94 (55.50) 92.46 (116.05) 63.93 (56.86)b 71.81 (89.97) 83.25 (98.90) 59.10 (58.29)b 59.63 (62.74) Fruits (servings/day) 1.20 (0.90) 1.10 (1.20) 1.22 (1.32) 0.97 (0.82) 1.24 (1.28) 1.15 (1.05) 1.32 (1.14) 1.76 (2.29) 1.26 (1.88)c Vegetables (servings/day) 1.61 (1.75) 1.54 (1.73) 1.26 (1.34) 1.01 (0.99) 1.27 (1.42) 1.48 (1.80)a 1.36 (1.37) 1.50 (1.66) 1.41 (1.39) KEDS total 7.82 (3.07) 7.21 (3.00) 6.85 (3.50) 7.26 (3.57) 6.53 (3.61) 6.93 (3.75) 7.00 (2.74) 6.65 (2.87) 6.48 (3.37)
12 mo 6 mo
FC-group and FC-IVR (SE)
This analytic plan would occur if the hypothesized difference between the FC-workbook and FC-group interventions was detected in the intent-to-treat analysis as per the initial hypothesis. The second plan included the potential to examine the differences across three newly formed conditions (combined FC-workbook and FC-group; FC-IVR: 0 –5 calls; and FC-IVR: 6 –10 calls). The second plan was developed in the event that the alternative hypothesis was supported, that is, that the participants in the FC-workbook and FC-group interventions were not significantly different from each other on study outcomes, thus enhancing the statistical power for comparisons among groups.
Am J Prev Med 2009;36(1)
39
48.80 (39.72) 0.97 (0.80)b,c 1.30 (0.97) 6.66 (3.10) p⬍0.05 for BL– 6 months p⬍0.05 for BL–12 months c p⬍0.05 for 6 –12 months BL, baseline; IVR, interactive voice response; KEDS, Kids’ Eating Disorders Survey; mo, months; MPA, moderate physical activity; VPA, vigorous physical activity b
BMI z-score VPA self-report MPA self-report Sedentary/ screen time Sugared drinks Fruits Vegetables Total KEDS
a
40.80 (33.50)a 1.52 (1.40) 1.23 (0.89) 6.57 (2.80) 61.66 (46.56) 1.52 (1.21) 1.32 (1.11) 7.11 (2.68) 86.68 (93.42) 1.97 (3.04) 1.61 (2.02) 6.68 (4.24) 91.95 (68.89) 2.38 (3.56) 1.78 (2.49) 6.74 (3.36) 112.16 (135.75) 1.26 (1.10) 1.51 (1.46) 6.26 (2.64) 62.18 (54.98) 1.03 (0.87) 1.31 (1.49) 6.92 (3.29) 78.92 (79.17) 1.10 (0.86) 1.29 (1.52) 7.40 (3.28)
2.00 (0.04) 4.16 (2.05) 2.56 (2.02) 5.55 (2.45)
2.00 (0.04) 4.49 (1.92) 2.62 (2.33) 5.46 (2.41) 2.04 (0.04) 4.12 (2.14) 2.32 (2.21) 5.16 (2.16)
71.80 (81.11) 1.18 (1.19) 1.47 (1.67) 7.08 (3.46)
1.90 (0.06)b,c 3.70 (2.21) 2.57 (2.18)b 5.43 (2.22) 2.08 (0.07) 3.50 (2.01) 2.35 (1.87) 5.60 (2.54)
2.07 (0.08) 3.40 (2.14) 2.40 (2.14) 6.25 (2.71)
2.09 (0.08) 3.05 (1.99) 2.90 (2.49) 5.65 (1.84)
2.03 (0.05) 3.89 (2.70) 1.65 (2.11) 5.43 (2.41)
1.94 (0.06) 3.59 (2.27) 2.89 (2.54)a 5.24 (1.69)
12 mo
a
6 mo BL
6 mo
a
12 mo
b
BL
6 mo
12 mo
BL
FC-IVR: 6–10 calls nⴝ38 FC-IVR: 0–5 calls nⴝ20 FC-workbook and FC-group nⴝ82
Table 3. Variable M (SD) by IVR dose 40
Participants in the group FC-IVR: 6 –10 calls reported a significant increase in the number of days they participated in MPA from baseline to 6 months and from baseline to 12 months. However, the changes in physical activity were not significantly different among groups. Again, changes in fruit, vegetable, and sugareddrink consumption did not demonstrate a consistent pattern of changes within or among groups.
Discussion This is the first trial to attempt to achieve clinical weight management using primarily automated technology to engage the parent as a proxy in the treatment of overweight children. Exploratory analyses of withincondition effects for the FC-IVR demonstrated improvements at both 6 months and 12 months, but between-group differences were not significant. It was found that the dose of the FC-IVR intervention may be important; a higher dose of completed parental IVR calls was associated with increased MPA and decreased BMI z-scores in those children. However, dose was not systematically varied within this study, and both the increased dose and improved BMI z-scores could be indicators of the parent level of motivation to make changes to support an overweight child in living a healthier lifestyle. Still, both motivation and skill development are necessary to achieve behavior change,60 so it could be argued that motivated parents were more likely to receive a higher level of training on skill development through the increased number of IVR calls that were completed. From a scientific perspective, this provides an avenue for future effectiveness research: to determine if varying levels of dose, within an RCT, can influence changes in parental skill development and child BMI z-scores. From a practical perspective, the information from the present study suggests that if FC-IVR was offered in a managed-care setting, approximately half of the parents who received the intervention would complete at least six calls and achieve a significant reduction in BMI z-scores for their overweight children. The frequency and duration of intervention contacts may also have an influence on the magnitude of the effect achieved. For example, recent studies by Savoye and colleagues31 and Epstein et al.30 reported larger effects on child BMI z-scores than this trial. Epstein’s trials typically included anywhere from 14 to 20 face-toface sessions for children and parents. In the case of the Savoye study,31 parents and children attended two face-to-face classes per week for 6 months and an additional class every 2 weeks for the next 6 months. This totaled 60 sessions over a 12-month period. The resulting conclusion could be that very intensive interventions can achieve significant reductions in BMI z-scores, while FC-IVR with two face-to-face sessions and
American Journal of Preventive Medicine, Volume 36, Number 1
www.ajpm-online.net
at least six follow-up counseling calls may be more likely to achieve modest reductions in BMI z-scores. Conversely, the frequency and duration of intervention contacts may influence the likelihood that a given intervention is disseminated and used in typical practice.61 When this study’s FC-IVR intervention is compared to, for example, the trial by Savoye et al.,31 it is clear, simply from a human resources perspective, that the FC-IVR intervention could reach a far greater number of families. In the one instance,31 one small group of overweight children and parents engage in an intensive, 60-session intervention with healthcare personnel over the course of a year. In terms of this study’s FC-IVR intervention, 60 face-to-face sessions would allow for the engagement of 30 small groups of overweight children and parents (i.e., two small-group sessions followed by IVR for each small group)— roughly a 30:1 advantage in potential reach with identical resources. Our intent is not to suggest that a clinical weightmanagement program for children should be deemed satisfactory by simply achieving a very modest decrease in BMI z-scores. Rather, it is to point to a potentially fruitful area of future practice that may utilize a theorybased IVR intervention to reduce the systemic cost of treating families with overweight children and increase the reach of efficacious interventions into the population at large. It is noteworthy that Calls 7–10 for FC-IVR provided messages that reinforced the basic skill-building activities and information provided during earlier intervention sessions. This study’s results suggest that these calls may be important to support decreases in BMI z-scores. This study has some limitations. First, parents were not assigned randomly to a higher or lower frequency of FC-IVR. Thus, it could be that those parents who completed all of the calls were more motivated than those who did not. Second, many children declined to wear an accelerometer, so an objective assessment of physical activity was not possible, and a very brief selfreport measure was used to reduce participant burden. Third, the cost of delivery of each of the interventions was not tracked and, therefore, information cannot be provided relative to the differential effects based on the cost of delivery, tracking, and patient retention across intervention conditions. In conclusion, the completions of Calls 6 –10 of the FC-IVR demonstrated a relative advantage over standard care. While doing so, the intervention did not have any adverse consequences on child body image or eating-disorder symptoms. The study also suggests that Golan’s social– ecologic model12 can be used to tailor an automated intervention for pediatric weight management. Conversely, the clinical significance of a small reduction in BMI z-scores, like that seen in this study, is unclear. Finally, future research should focus on understanding the optimal integration of automated telephone January 2009
counseling with human intervention strategies to improve the magnitude of effect and the potential for broad reach while monitoring the cost and cost effectiveness of pediatric weight-management programs. This work was supported by the Garfield Memorial Fund, an internal funding mechanism of Kaiser Permanente Health Plans (Project #50-227). Also, the Kaiser Permanente Colorado Weight Management Program, Michele Gilson, and Helen Seagle were integral in designing the study as a practical RCT and developing the original Family Connections content. No financial disclosures were reported by the authors of this paper.
References 1. Jolliffe D. Extent of overweight among U.S. children and adolescents from 1971 to 2000. Int J Obes 2004;28:4 –9. 2. Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH. Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med 1997;337:869 –73. 3. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the U.S., 1999 –2004. JAMA 2006;295:1549 –55. 4. Bloomgarden ZT. Prevention of obesity and diabetes. Diabetes Care 2003;26:3172– 8. 5. American Academy of Pediatrics. Policy statement: prevention of pediatric overweight and obesity. Pediatrics 2003;112:424 –30. 6. Estabrooks PA, Shetterly S. The prevalence and health care use of overweight children in an integrated health care system. Arch Pediatr Adolesc Med 2007;161:222–7. 7. Golan M, Crow S. Targeting parents exclusively in the treatment of childhood obesity: long-term results. Obes Res 2004;12:357– 61. 8. Golan M, Crow S. Parents are key players in the prevention and treatment of weight related problems. Annu Rev Nutr 2004;62:39 –50. 9. Golan M, Weizman A. Familial approach to the treatment of childhood obesity: conceptual model. J Nutr Educ 2001;33:102–7. 10. Golan M, Weizman A, Fainaru M. Impact of treatment for children obesity on parental risk factors for cardiovascular disease. Prev Med 1999;29: 519 –26. 11. Golan M, Weizman A, Apter A, Fainaru M. Parents as the exclusive agents of change in the treatment of childhood obesity. Am J Clin Nutr 1998;67:1130 –5. 12. Golan M, Fainaru M, Weizman A. Role of behaviour modification in the treatment of childhood obesity with parents as the exclusive agents of change. Int J Obes 1998;22:1217–24. 13. Coleman KJ, Epstein LH. Application of generalizability theory to measurement of activity in males who are not regularly active: a preliminary report. Res Q Exerc Sport 1998;89:58 – 63. 14. Epstein LH. Exercise and obesity in children. J Appl Sport Psychol 1992; 4:120 –33. 15. Epstein LH, Coleman KJ, Myers MD. Exercise in treating obesity in children and adolescents. Med Sci Sports Exerc 1996;28:428 –35. 16. Epstein LH, Goldfield GS. Physical activity in the treatment of childhood overweight and obesity: current evidence and research issues. Med Sci Sports Exerc 1999;31(11S):S553–9. 17. Epstein LH, Myers MD, Raynor HA, Saelens BE. Treatment of pediatric obesity. Pediatrics 1998;101:554 –70. 18. Epstein LH, McCurley J, Wing RR, Valoski A. Five-year follow-up of family-based behavioral treatments for childhood obesity. J Consul Clin Psychol 1990;58:661– 4. 19. Epstein LH, Valoski A, Wing RR, McCurley J. Ten-year follow-up of behavioral family-based treatment for obese children. JAMA 1990;264: 2519 –23. 20. Epstein LH, Valoski A, Vara LS, et al. Effects of decreasing sedentary behavior and increasing activity on weight change in obese children. Health Psychol 1995;14:109 –15.
Am J Prev Med 2009;36(1)
41
21. Epstein LH, Wing RR, Valoski A, DeVos D. Long-term relationship between weight and aerobic-fitness change in children. Health Psychol 1988; 7:47–53. 22. Epstein LH, Wing RR, Steranchak L, Dickson B, Michelson J. Comparison of family-based behavior modification and nutrition education for childhood obesity. J Pediatr Psychol 1980;5:25–36. 23. Epstein LH, Wing RR, Woodall K, Penner BC, Kress MJ, Koeske R. Effects of family-based behavioral treatment on obese 5-to-8 year old children. Behav Ther 1985;16:205–12. 24. Epstein LH, McKenzie SJ, Valoski A, Klein KR, Wing RR. Effects of mastery criteria and contingent reinforcement for family-based child weight control. Addict Behav 1994;19:135– 45. 25. Epstein LH, Paluch RA, Raynor HA. Sex differences in obese children and siblings in family-based obesity treatment. Obes Res 2001;9:746 –53. 26. Epstein LH, Roemmich JN, Raynor HA. Childhood and adolescent obesity: behavioral therapy in the treatment of pediatric obesity. Pediatr Clin Nor Am 2001;48:981–93. 27. Epstein LH, Paluch RA, Kilanowski CK, Raynor HA. The effect of reinforcement or stimulus control to reduce sedentary behavior in the treatment of pediatric obesity. Health Psychol 2004;23:371– 80. 28. Epstein LH, Paluch RA, Gordy CC, Saelens BE, Ernst MM. Problem solving in the treatment of childhood obesity. J Consult Clin Psychol 2000;68: 717–21. 29. Epstein MH, Valoski A, Wing RR, McCurley J. Ten-year outcomes of behavioral family-based treatment for childhood obesity. Health Psychol 1994;13:373– 83. 30. 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;26:381–91. 31. Savoye M, Shaw M, Dziura J, et al. Effect of a weight management program on body composition and metabolic parameters in overweight children: a randomized controlled trial. JAMA 2007;297:2697–704. 32. Golan M, Weizman A, Apter A, Fainaru M. Parents as the exclusive agents of change in the treatment of childhood obesity. Am J Clin Nutr 1998; 67:1130 –5. 33. Edmunds L, Waters E, Elliot EJ. Evidence based management of childhood obesity. BMJ 2001;323:916 –9. 34. Estabrooks P, Dzewaltowski DA, Glasgow RE, Klesges LM. School-based health promotion: issues related to translating research into practice. J Sch Health 2003;73:28. 35. Barlow SE, Dietz WH. Management of child and adolescent obesity: summary and recommendations based on reports from pediatricians, pediatric nurse practitioners, and registered dietitians. Pediatrics 2002;110:236 – 8. 36. Glasgow RE, Bull SS, Gillette C, Klesges LM, Dzewaltowski DA. Behavior change intervention research in health care settings: a review of recent reports, with emphasis on external validity. Am J Prev Med 2002;23:62–9. 37. Biem HJ, Turnell RW, D’Arcy C. Computer telephony: automated calls for medical care. Clin Invest Med 2003;26:259 – 68. 38. Piette JD. Patient education via automated calls: a study of English and Spanish speakers with diabetes. Am J Prev Med 2000;17:138 – 41. 39. Piette JD, McPhee SJ, Weinberger M, Mah CA, Kraemer FB. Use of automated telephone disease management calls in an ethnically diverse sample of low-income patients with diabetes. Diabetes Care 1999;22: 1302–9. 40. Piette J. Enhancing support via interactive technologies. Curr Diab Rep 2002;2:160 –5. 41. Pinto BM, Friedman R, Marcus BH, Kelley H, Tennstedt S, Gillman MW. Effects of a computer-based, telephone-counseling system on physical activity. Am J Prev Med 2002;23:113–20.
42
42. Delichatosios H, Friedman R, Glanz K, et al. Randomized trial of a “talking computer” to improve adults’ eating habits. Am J Health Promot 2001; 15:215–24. 43. Krishna S, Balas EA, Boren SA, Maglaveras N. Patient acceptance of voice messages: a review of controlled clinical studies. Methods Inf Med 2002;41:360 –9. 44. Mahoney DF, Tarlow BJ, Jones RN. Effects of an automated telephone support system on caregiver burden and anxiety: findings from the REACH for TLC intervention study. Gerontologist 2003;43:556 – 67. 45. Tunis SR, Stryer DB, Clancy CM. Practical clinical trials: increasing the value of clinical research for decision making in clinical and health policy. JAMA 2003;290:1624 –32. 46. Must A, Dallal GE, Dietz WH. Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2) and triceps skinfold thickness. Am J Clin Nutr 1991;53:839 – 46. 47. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts for the U.S.: methods and development. Vital Health Stat 2002:11;1–190. 48. Wilfley DE, Stein RI, Saelens BE, et al. Efficacy of maintenance treatment approaches for childhood overweight: a randomized controlled trial. JAMA 2007;298:1661–73. 49. Troped P, Wiecha J, Fragala M, et al. Reliability and validity of YRBS physical activity items among middle school students. Med Sci Sports Exerc 2007;39:416 –25. 50. Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardener L. A data-based approach to diet questionnaire design and testing. Am J Epidemiol 1986;124:453– 69. 51. Block G. Health habits and history questionnaire. Program evaluation handbook: nutrition education. Los Angeles: IOX Assessment Associates, 1988. 52. Childress AC, Jarrell MP, Brewerton TD. The kids’ eating disorders survey (KEDS): internal consistency, component analysis, and reliability. Eating Disorders: The Journal of Treatment and Prevention 1993;1:123–33. 53. Epstein L, Gordy C, Raynor H, Beddome M, Kilanowski C, Paluch R. Increasing fruit and vegetable intake and decreasing fat and sugar intake in families at risk for childhood obesity. Obes Res 2001;9:171– 8. 54. Young K, Northern J, Lister K, Drummond J, O’Brien W. A meta-analysis of family-behavioral weight-loss treatments for children. Clin Psychol Rev 2007;27:240 –9. 55. Estabrooks PA, Glasgow RE, Dzewaltowski DA. Physical activity promotion through primary care. JAMA 2003;289:2913– 6. 56. Laird NM, Ware JH. Random effects models for longitudinal data. Biometrics 1982;38:963–74. 57. McCullagh P, Nelter JA. Generalized linear models. Boca Raton FL: Chapman Hall/CRC, 1989. 58. Wolfinger R, O’Connell M. Generalized linear mixed models: a pseudolikelihood approach. Journal of Statistical Compututation and Simulation 1993;48:233– 43. 59. Diggle PJ, Liang KY, Zeger SL. Analysis of longitudinal data. Oxford: Oxford University Press, 1994. 60. Bandura A. Self-efficacy: the exercise of control. New York: Freeman, 1997. 61. Rogers EM. Diffusion of innovations. 5th ed. New York: Free Press, 2003.
Supplementary data Supplementary data associated with this article can be found, in the online version, at 10.1016/j.amepre.2008.09.010.
American Journal of Preventive Medicine, Volume 36, Number 1
www.ajpm-online.net
Appendix A. General intervention framework, strategies, and activities Concept
Strategy
Example of potential activity
Increased parental behavioral health skills Increased parenting skills
Enhance parental personal motivation related to weight, nutrition, and physical activity
Demonstrate personal goal setting and barrier resolution related to physical activity and healthful eating Demonstrate parental responsibilities and role in child’s weight (i.e., not the child’s fault) Demonstrate parental leadership, general parenting skills, and effective parent–child communication Model enjoyment of healthy foods and physical activity Engage with child in physical activities Practice regular meal times and schedule between-meal snacks Involve child in decision making for enjoyable physical activity
Reframe the problem Set healthy limits
Parental modeling
Parents provide an example of healthy lifestyle behaviors for child
Home environmental changes
Restructure the home environment to support healthy food and activity options while reducing options for unhealthy choices
Appendix B. Breakdown of workbook content areas by day Day
Part 1
Part 2
1
● Assessing and calculating BMI in children and adults ● Causes of overweight in children (biological, cognitive, environmental ● Five reasons children gain weight
2
● Impact of being overweight ● Helping children to avoid fad diets ● Activity to explore parental beliefs about eating and physical activity ● Healthy habits for creating a healthy family ● Defining the division of responsibility for eating and activity ● Parenting skills to support weight reduction ● Survey of the family home environment
● Ways to promote a healthy home environment ● Nutrition: reading labels, selecting healthy food options ● Physical activity: using the FITT (frequency, intensity, time, type) principles ● Sample menus ● Tips for preparing healthy snacks and meals ● Goal setting: creating a family action plan ● Process of goal setting and keeping objectives clear ● Parent’s personal action plan ● Barriers and strategies to maintaining family action plan ● Resources
3
Am J Prev Med 2009;36(1)
42.e1
Appendix C. Conceptual and content bases of interactive voice response follow-up calls Call timing (# weeks post final group session)
Target concept
Content and sample goals
1
Parenting skills
2
Behavioral skills
3
Home environment
4
Role modeling
6
Keeping motivated
8
Part 1: Goal setting using the 5A’s (assess, advise, ask, assist, arrange)
10 12
Part 2: Goal setting using the 5A’s (assess, advise, ask, assist, arrange) Community resources
16
Kids speak up!
20
Relapse prevention
Consistency and contingency: Be consistent in limiting the amount of television your child watches this week to ⬍2 hours/day. Communication: Talk to your child about the types of physical activity he or she likes to do. Healthy habits: Set a good example for your child by eating 5–9 servings of fruits and vegetables every day this week. Praise: Praise your child whenever he or she attempts to be physically active. Physical activity: Get at least 2.5 hours of moderate physical activity this week. Nutrition: Drink low-fat milk at one meal each day. Stress: Take 10-minute stress-break walks at least once per day. Physical activity: Make a list of your family’s barriers to physical activity around the house, and come up with a way to overcome them. Nutrition: Clear the kitchen cupboards of unhealthy snacks. Physical activity: Do something active with your child for 15 minutes at least three times this week. Nutrition: Prepare at least one healthy meal together with your child. Physical activity: Review the family goal sheet with your family and set measurable, specific, and objective goals this week. Nutrition: Post signs at least twice this week about your family’s positive changes with healthy eating. Sedentary behavior: Decrease your sitting time by 0.5 hours each day this week. Nutrition: Decrease your soda and sugared-drink consumption by one serving per day. Physical activity: Increase your physical activity by 1 day per week. Nutrition: Increase your servings of fruits and vegetables by three per day. Physical activity: Set a goal to take your family to a new park or trail that you haven’t visited before. Nutrition: Check your library for cookbooks, recipes, or videos that help your family to prepare nutritious meals. Parenting skills: Set a goal to make a physical activity resource kit and use it when your family is out and about or at home. Physical activity: Survey your family regarding physical activity that they would like to do as a family, and try to do that activity at least three times per week. Nutrition: Change one food item that is high-fat to a healthy snack of fruit or vegetable. Support: Contact a weight-management specialist if any problems arise. Plan: Complete the 5A’s goal-setting process every 3 months to keep your family working on healthful eating and physical activity. Commitment: Commit to your family’s new healthy lifestyle for at least the next full year.
42.e2
American Journal of Preventive Medicine, Volume 36, Number 1