The research design is a 3
4 factorial with one between subjects factor (treatment condition) with three levels and one within subjects factor (time) with four levels (baseline, 3-, 6-, and 12-months). Participants will be randomized to one of three groups: a) Standard behavioral weight loss program (STND); b) technology-supported behavioral weight loss program (TECH); or c) self-guided behavioral weight loss program (SELF).
A total of 96 participants who meet the inclusion and exclusion criteria listed in Table
will be recruited in two cohorts. Participants will be between 18–60 years old with a body mass index (BMI) between 30–40 kg/m2, weight stable for the past 6 months, not currently enrolled in a formal weight loss program or taking weight loss medications, and be willing to self-monitor and wear an accelerometer daily for 6 months. Participants will be excluded if they have any unstable medical conditions, are currently pregnant, or taking any medications known to influence weight. See Table
for a complete list of exclusion criteria.
ENGAGED inclusion and exclusion criteria
Recruitment and screening process
All study procedures are approved by the Institutional Review Board at Northwestern University. Participants will be recruited via flyers and advertisements in local newspapers and on public transportation. Interested candidates will be directed to a study website which includes study-related information and an online screening questionnaire to assess initial eligibility. Figure
illustrates the screening process.
ENGAGED Study Screening Flowchart.
Participants will complete an initial online consent process prior to completing the online screener, which assesses height and weight, physical activity readiness, group session availability, and exclusionary weight loss practices (i.e. participation in a weight loss program or use of weight loss medications). Study staff will contact eligible participants within 1–4 days of completing an online screener to conduct a telephone screen.
The telephone screen will review details of the study and review additional eligibility criteria including health status and medication use. Eligible and interested participants will be invited to attend a study orientation session and will be instructed to begin a one-week run-in period of dietary self-monitoring. If participants indicate a health condition such as type II diabetes or hypertension, medical clearance from a physician will be required prior to the pretreatment session in order to continue participation.
In-person orientation session
During the orientation session, complete details of the study will be given and an equipoise induction will be performed to ensure that participants are aware of the pros and cons of the treatment conditions and to minimize the risk of differential attrition
]. Participant questions will be answered and informed consent will be obtained from interested participants. Consented participants will complete the final in-person screening and baseline assessment, which includes assessing blood pressure and anthropometric measurements. In addition, participants will complete a smartphone competency task to evaluate their ability to use basic functions of a smartphone after instruction and will complete a brief interview to assess: binge eating disorder, bulimia nervosa, substance abuse and dependence, attention deficit hyperactivity disorder, and major depressive disorder. Eligible participants will also complete baseline questionnaires and wear an accelerometer for 10 days to measure current activity levels.
At the pretreatment session, participants will return the accelerometer, meet with a dietician to complete a dietary recall, and engage in a brief treadmill activity. During the treadmill activity, participants will walk on a treadmill at a moderate intensity while reviewing ratings of perceived exertion to become familiar with the physiological signs of physical activities at this intensity.
Eligible participants will be scheduled for a group time and assigned to a team with seven other participants, all of whom will be randomized to the same condition. Group randomization will be completed by a statistician, and neither participants, nor coaches, will find out group assignment until the first group session. The intervention timeline is presented in Figure
Standard weight loss program (STND)
STND participants will be given a goal of 7% weight loss to be achieved through changes in diet and activity, similar to the DPP lifestyle intervention
]. Based on baseline weight, participants will be given daily calorie goal between 1200–2000 kcal/day and a 25% fat gram goal between 32–55 grams/day. These goals are designed to produce a one to two pound weight loss per week. Participants will also be given a weekly physical activity goal, beginning with 45 minutes of moderate intensity physical activity per week and progressing weekly to 175 minutes per week. Participants will be provided with “The Complete Book of Food Counts”
] and paper Keeping Track diaries during months 1–6 to facilitate self-monitoring of dietary intake, physical activity, and body weight.
The original sixteen session DPP protocol will be condensed to eight group sessions (see Table
) and, similar to the Look AHEAD trial
], group sessions will be combined with individual telephone coaching to increase social support and reduce costs
]. Group sessions will run 90 minutes in duration and be held weekly for the first eight weeks of the study. Each group session will begin with a private weigh-in and participants will submit their Keeping Track diary. All group sessions will be led by a trained lifestyle interventionist and will include setting an agenda, reviewing homework and self-monitoring, presentation of new material, group discussions, and assignment of homework. See Table
for a list of session topics. All group sessions will end with an optional supervised group walking session.
ENGAGED Study Group Session Topics
In addition to in-person group sessions, STND participants will receive individual coaching calls. Coaching calls will be scheduled on a weekly basis during the initial eight weeks of the study and monthly during months 3–6. Coaches will review submitted Keeping Track diaries and determine the participant’s level of achievement for each goal using standardized adherence criteria for the study. Coaching calls will last approximately 10–15 minutes and follow a structured script aiming to reinforce success and problem solve around barriers to change. Four content areas will be covered during each call: (1) weight loss, (2) adherence to dietary and physical activity self-monitoring, (3) attainment of diet and activity goals, and (4) goal setting.
Technology-supported weight loss program (TECH)
Similar to STND, TECH will attend eight in-person group sessions, receive regular coaching calls, and follow the same weight, physical activity, and dietary goals. However rather than paper Keeping Track diaries, TECH will be provided with a study smartphone, ENGAGED application, and an accelerometer to use during the first 6 months of the intervention. In addition to covering the topics discussed during the first group meeting (Table
), TECH will be instructed on how to self-monitor dietary intake, use the accelerometer, and communicate virtually with their teammates and coach using the ENGAGED application. Time-stamped dietary and physical activity data will be continuously downloaded to a secure server, and participant data will be summarized automatically on the secure coaching application and website, allowing coaches to have up-to-date access to participant data, enabling corrective or supportive feedback to be given via text message or during scheduled coaching calls.
The ENGAGED technology includes a study smartphone (Motorola Droid™), pre-loaded with the ENGAGED application, and an accelerometer (SHIMMER™). TECH participants will use the ENGAGED application, which contains the CalorieKing® food database, a comprehensive nutritional source containing over 50,000 food entries, to self-monitor daily dietary intake. Goal thermometers will display participant’s goal and actual amount of calories and fat grams consumed (Figure
). The ENGAGED application also includes a Team tab that allows participants to view team members’ adherence to self-monitoring and accelerometer usage. In addition, the Team tab on the application includes a message board and peer-to-peer messaging to facilitate communication and support among teammates. Participants will also wear an accelerometer that connects via Bluetooth to the smartphone to provide real-time objective data on the accumulation of moderate-to-vigorous intensity physical activity. In the event participants engage in an activity not captured by the accelerometer (i.e. cycling, swimming), users may manually input physical activity.
Screen shot of ENGAGED smartphone application.
Self-guided behavioral weight loss program (SELF)
SELF will attend one 60-minute in-person group session during but will not attend any additional group sessions nor receive coaching calls throughout the intervention. Similar to STND, SELF will be given “The Complete Book of Food Counts”
] and paper Keeping Track diaries. Rather than attend in-person group sessions, SELF will receive three Group Lifestyle Balance™ (GLB) Program DVDs
] that contain 12 group sessions adapted from the original DPP curriculum as well as the corresponding GLB participant handouts. SELF will be given their individual 7% weight loss goals and instructed to watch the DVDs to get their calorie and fat gram goals.
Team weight loss competition
In order to encourage members of each weight loss team to support each other’s behavioral adherence and weight loss, a team weight loss competition will be held at 3- and 6-months during each cohort. The team with the greatest weight loss percentage at each time point will win $400, to be divided evenly among the group members ($50 per team member).
Fidelity checklists will be prepared for group and telephone coaching sessions. Checklists for each session will specify a) good counseling practices; b) prescribed session content; and c) proscribed session content. Coaches will undergo training to become familiarized with the modified DPP protocol and to role-play potential conflicts and situations that may arise during group and telephone sessions. Study staff will also attend weekly clinical meetings to discuss any participant conflicts that may arise. All group and telephone coaching sessions will be audio recorded, and each month a random sample of 10% of the recordings will be coded by study staff to evaluate treatment fidelity. If fidelity falls below 90% on any of the fidelity checks, staff will be retrained.
Assessments will be completed at baseline, 3-, 6-, and 12-months. Participants will receive $40 for completing each of the 3-, 6, and 12-month assessments. The procedures and measures are described below:
Height, weight, BMI, and waist circumference
Height will be measured using a wall-mounted stadiometer to the nearest 0.01 cm. Body weight will be measured without shoes, wearing light clothing on a calibrated beam balance scale to the nearest 0.01 kg. BMI (kg/m2) will be calculated using height and weight measurements. Waist circumference will be assessed by positioning an anthropometric tape midway between the palpated iliac crest and the palpated lowest rib margin in the mid-axillary lines. Two measurements will be taken and the mean value of each site will be used.
Dietary intake adherence
Dietary intake will be assessed at baseline, 3-, and 6-months via interview by a non-intervention registered dietician using the computer-assisted telephone interview (CATI) 24 hour recalls
] using the multiple pass method to assess calorie and fat intakes. At each time point, three days (one weekend and two weekdays) will be randomly selected over a three week period to conduct the recalls. The data derived from the recalls will be analyzed using the annually updated University of Minnesota Nutrition Coordinating Center’s (NCC) Nutrition Data System for Research (NDS-R-2005). Only STND and TECH will complete dietary recalls at 3- and 6-months.
Physical activity adherence
Physical activity will be objectively measured by an accelerometer (Actigraph #7164, Pensacola, FL) over 7 days at baseline, 3-, and 6-months. The accelerometer will be attached to an elastic belt and worn on the right hip. Minutes of moderate-to-vigorous intensity physical activity will be measured in a modified bout lasting a minimum of eight consecutive minutes. Adherence will be ascertained on the basis of how well the participant’s bout-corrected accelerometer count corresponds to the weekly activity goal.
Dietary self-monitoring adherence
Dietary self-monitoring adherence will be assessed as the number of days reporting dietary intake on the paper Keeping Track diaries (STND and SELF) or smartphone application (TECH). In order to receive full adherence credit for a day of self-monitoring, participants must have ≥ 1000 calories recorded and ≥ 25 grams of fat.
Physical activity self-monitoring adherence
For STND and SELF, physical activity self-monitoring adherence will be assessed as the number of days reporting physical activity in the paper Keeping Track diaries, divided by the number of possible days. For TECH participants, self-monitoring adherence will be operationalized as the number of days the accelerometer was worn ≥ 8 hours, divided by the number of possible days.
Coaches will monitor the number of minutes preparing for and conducting individual coaching calls throughout the first 6 months of the intervention.
Data analytic plan
Linear mixed modeling
Data will be analyzed on an intent-to-treat basis. The basic modeling approach will entail linear mixed models for longitudinal data. The primary aim is to test for treatment group differences on weight loss at 3-, 6-, and 12-months as well as differences in behavioral adherence and goal attainment at 3- and 6-months. Linear mixed models do not require that participants are measured at all time points, and therefore can include participants with missing data across time. In these analyses, we will examine changes across time by treating time as a nominal variable in the analysis to allow for non-linear trends. In addition to the effect of time, the main independent variable is treatment groups, a between-subjects factor with 3 levels: STND, TECH and SELF. A priori Helmert contrasts
] will be used to test the following comparisons at 3- and 6-months: 1) STND and TECH vs. SELF and 2) TECH vs. STND. We will also examine group by time interactions to test whether change in outcome across time differs by group.
We hypothesize that the weight loss advantage for those assigned to TECH would be explained (entirely or partially) by TECH’s higher rates of adherence to self-monitoring and attainment of behavioral goals. To test this mediation hypothesis, we will fit a linear mixed model predicting weight loss similar to the model described above, but now including in the regression equation the potential mediating variables corresponding to adherence to self-monitoring. To the extent that adherence mediate the effects of the experimental condition, in the model with adherence included, we expect the coefficient on the treatment group by time interaction variable to be reduced compared to the treatment group by time interaction coefficient from the model without adherence. This difference provides an estimate of the extent to which adherence may mediate the effects of the experimental condition on weight loss. Bootstrapping will be used to estimate the sampling variability of the mediation effect.
Sample size and power
We will randomize n=32 participants to each treatment group and assume 6% of participants will drop out by the end of the study so that there will be n=30 participants in each group at the final endpoint. The power calculations below are based on the second Helmert contrast: TECH vs. STND because this contrast contains the fewest number of subjects (30 in each group) and we expect this contrast to have the smallest effect sizes. The first Helmert contrast: STND and TECH vs. SELF has a larger sample size (60 vs. 30) and we anticipate larger effects sizes in this contrast due to the limited intensity of the SELF intervention. Therefore, by powering the study based on the second contrast, we will have more than adequate power for the first contrast.
We measure weight loss at 3, 6, and 12 months. Using data from a previous weight loss trial
], we saw standard deviations of 8.4 and 10.7 pounds at months 3 and 6, respectively, with a correlation of 0.86 between the two time points. With 80% power and an n=30 in both groups, we will be able to detect a difference of 6.6 pounds between the two groups.
Diet self-monitoring adherence
We measure dietary self-monitoring adherence at 3 and 6 months as the ratio of actual days of diet self-monitoring relative to the possible number of days. Using data from a previous trial
], we observed a standard deviation of 0.31 and 0.34 at 3 and 6 months, respectively, as well as a correlation of 0.93 between the two time points. Thus, with 80% power and an n=30 in both groups, we will be able to detect a difference in adherence ratios of 0.23 between the two groups.
Physical activity self-monitoring adherence
Physical activity self-monitoring will be assessed as the number of days reporting physical activity on paper diaries (STND and SELF) or wearing the accelerometer (TECH), divided by the number of possible days. Data from a previous trial
] shows a standard deviation of 0.30 at 3 months and 0.33 at 6 months and a correlation of 0.91 between the two time points. With 80% power and an n=30 in both groups, we will be able to detect a difference in percent adherence of 22% between the two groups.
Dietary intake adherence
We measure adherence to calorie goals at 3 and 6 months using 3 day 24 hour recalls. Using pilot data from a previous trial, we observed a standard deviation of change in calories of 498 at month 3 and 431 at month 6, with a correlation of 0.9 between the two time points. With 80% power and an n=30 in both groups, we will be able to detect a calorie reduction difference of 375 calories.
Physical activity adherence
Physical activity adherence is measured by accelerometer at months 3 and 6 as the number of minutes of moderate to vigorous intensity physical activity per week. Using data from a previous trial
], we assume a standard deviation of 142 minutes/week at months 3 and 6, with a correlation of 0.9 between the two time points. With 80% power and an n=30 in both groups, we will be able to detect a 100 minute/week difference between the two groups.