This randomized clinical trial examines the effectiveness of an automated online behavioral obesity treatment program in primary care for initial weight loss, and how well 3 different maintenance schedules help reduce weight regain.
Key Points
Question
Is an automated online behavioral obesity treatment program effective for long-term weight loss and what is the most effective delivery schedule for a maintenance intervention?
Findings
This randomized clinical trial included 540 participants enrolled in a 3-month weight loss treatment program with randomization to 9 months of maintenance with 3 different strategies. Participants lost a mean of 3.6 kg and those in the 2 active maintenance groups (9 monthly sessions or two 4-week refresher sessions) had less weight regain at 12 and 24 months than those in the control group (monthly newsletters).
Meaning
Findings of this trial indicate that pragmatic implementation of an automated online behavioral obesity treatment that includes active maintenance produces clinically significant weight loss at 12 and 24 months.
Abstract
Importance
Behavioral weight loss interventions have achieved success in primary care; however, to our knowledge, pragmatic implementation of a fully automated treatment that requires little researcher oversight has not been tested. Moreover, weight loss maintenance remains a challenge.
Objective
To evaluate the long-term effectiveness of an automated, online, behavioral obesity treatment program (Rx Weight Loss [RxWL]) at 12 months (primary end point) and 24 months when delivered pragmatically in primary care and to compare the effectiveness of 3 weight loss maintenance approaches.
Design, Setting, and Participants
This randomized clinical trial of RxWL, an online weight loss program, recruited patients from a Rhode Island primary care network with approximately 60 practices and 100 physicians. Eligible participants were primary care patients aged 18 to 75 years with overweight or obesity who were referred by their nurse care manager and enrolled between 2018 and 2020. All participants were included in the intention-to-treat analysis, whereas only those who engaged with maintenance intervention were included in the per-protocol analysis. Data were analyzed from August 2022 to September 2023.
Interventions
All participants were offered the same 3-month weight loss program, with randomization to one of three 9-month maintenance programs: control intervention (monthly online newsletters), monthly intervention (9 monthly video lessons and 1 week of self-monitoring per month), or refresher intervention (an introductory session and two 4-week periods of lessons and self-monitoring at 7 and 10 months).
Main Outcomes and Measures
The primary outcome was weight change at 12 months using height and weight data collected from electronic medical records through 24 months.
Results
Among the 540 participants (mean [SD] age, 52.8 [13.4] years; 384 females [71.1%]) in the intention-to-treat analysis, mean estimated 3-month weight loss was 3.60 (95% CI, −4.32 to −2.88) kg. At the 12-month primary end point, the amount of weight regained in the monthly (0.37 [95% CI, −0.06 to 0.81] kg) and refresher (0.45 [95% CI, 0.27 to 0.87] kg) maintenance groups was significantly less than that in the newsletter control maintenance group (1.28 [95% CI, 0.85 to 1.71] kg; P = .004). The difference in weight regain between the monthly and refresher maintenance groups was not statistically significant. This pattern persisted at 24 months. In the per-protocol analysis of 253 participants, mean weight loss at the end of the initial 3-month intervention was 6.19 (95% CI, −7.25 to −5.13) kg. Similarly, at 12 months there was less weight regain in the monthly (0.61 kg) and refresher (0.96 kg) maintenance groups than in the newsletter control maintenance group (1.86 kg).
Conclusions and Relevance
Results of this randomized clinical trial indicate that pragmatic implementation of a 12-month automated, online, behavioral obesity treatment that includes 9 months of active maintenance produces clinically significant weight loss over 2 years in primary care patients with overweight or obesity. These findings underscore the importance of providing ongoing maintenance intervention to prevent weight regain.
Trial Registration
ClinicalTrials.gov Identifier: NCT03488212
Introduction
Over 50% of US adults are projected to have obesity by 2030.1 Evidence-based behavioral weight loss programs, a first-line treatment for obesity, are difficult to access.2 Efforts to expand access have centered on primary care settings, which serve over 75% of US adults,3 but many clinicians lack the resources to deliver these treatments.2,4 Online treatments overcome many of these barriers.5,6 Comprehensive online obesity treatments delivered in a primary care setting have elicited clinically significant weight losses of 3 kg to 5 kg.7,8,9,10,11
Most online obesity treatments in primary care require supplemental clinician coaching, which enhances weight loss outcomes7,8,12 but adds cost and reduces scalability.4 Thus, we developed Rx Weight Loss (RxWL), an automated online program that requires no clinician contact and produces weight loss of approximately 5 kg at 3 months.13,14 Historically, RxWL has been tested with substantial researcher involvement (eg, participation incentives, researcher interaction), which may improve patient engagement and outcomes.15,16,17 Conducting a pragmatic trial with minimal researcher involvement is an important next step for assessing the clinical effectiveness of such programs in primary care18 and moving toward disseminating this noncommercial intervention via community partners.
Another barrier to favorable outcomes of behavioral weight loss treatment is poor weight loss maintenance; weight regain of 1 to 2 kg annually is common after program completion.11,19 Over time, the natural rewards of weight loss decline and adherence may become monotonous, leading to disengagement.19,20,21 Treatments involving extended contact slow weight regain,7,19,22 but they are often delivered on a fixed interval schedule (eg, monthly) that may be associated with boredom and disengagement.23,24
Successful strategies for increasing patient engagement during weight maintenance include varying treatment content, timing, and scheduling.20,21,25 In 1 study, delivering content in occasional bursts as part of a refresher campaign led to 2.28 kg less weight regain compared with a newsletter control strategy.20 The refresher’s periodic intensive contact may have been associated with increased patient engagement, adherence, and weight loss during maintenance periods.20 However, to our knowledge, no studies have directly compared a refresher maintenance schedule with a traditional fixed-interval maintenance schedule for behavioral weight loss treatment in primary care.
This study involved a pragmatic trial26 of RxWL implemented in a statewide primary care network’s routine care. The RxWL program included a 3-month weight loss treatment phase and a 9-month maintenance phase with 3 different maintenance schedules: a control intervention (monthly newsletters only), a monthly maintenance intervention (monthly maintenance lessons with automated feedback), and a refresher maintenance intervention (two 4-week interventions at 7 and 10 months). Our primary objective was to examine the long-term effectiveness of RxWL for weight loss and to compare these 3 strategies for promoting weight loss maintenance through 12 (primary end point) and 24 months. We conducted both an intention-to-treat analysis and a per-protocol analysis of participants who met a minimal threshold for maintenance program engagement. We expected that the RxWL program would produce significant weight loss, the refresher maintenance format would reduce weight regain relative to the monthly maintenance format, and both active maintenance (monthly and refresher) interventions would be superior to the control format.
Methods
Design Overview and Procedure
This type 2 hybrid effectiveness-implementation trial assessed the long-term outcomes of RxWL among patients of the Rhode Island Primary Care Physicians Corporation, a primary care network of approximately 60 practices and 100 physicians. Practices are assigned a nurse care manager (NCM) for case management who referred patients with overweight and obesity to the RxWL program. From 2018 to 2020, interested patients visited the RxWL electronic platform for a program overview and to complete eligibility screening and provide online informed consent to participate. All eligible patients received the same initial 3-month weight loss program and a randomly assigned 9-month maintenance intervention (control, monthly, or refresher groups). Participants were randomized to a maintenance group electronically and automatically in the RxWL platform in a 1:1:1 ratio, and in blocks of 3, to maintenance interventions on study entry but were not informed of their assignment until the beginning of the maintenance phase (Supplement 1). The NCMs and research staff were not masked to the maintenance intervention. Weight data were collected from the electronic medical record (EMR) for 2 years after enrollment. Details regarding the study design, interventions, and self-reported 3-month outcomes were published previously.26,27 This study was approved by the Institutional Review Board of The Miriam Hospital. We followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.
Participants
Participants were patients referred by their NCM to the RxWL program. Eligibility criteria were broad: participants were aged 18 to 75 years (increased from age 70 years in 2018 per a request from the Rhode Island Primary Care Physicians Corporation), had a body mass index (BMI) of 25 or higher (calculated as weight in kilograms divided by height in meters squared), and had access to an internet-connected device. Exclusion criteria included current use of weight loss medications or consumption of a liquid diet; recent, current, or planned pregnancy; and inability to safely engage in unsupervised exercise.
Interventions
The Rx Weight Loss Program
The RxWL program is a noncommercial program beginning with a 3-month, fully automated online intervention that includes 12 weekly interactive 15-minute video lessons teaching evidence-based behavior change strategies for weight loss (eg, goal setting, problem solving).28,29 Participants submitted self-monitoring data including daily weight, minutes of moderate to vigorous physical activity (MVPA), and energy intake, and received weekly automated tailored feedback. Program goals included weight loss of 0.5 to 1 kg each week, a calorie prescription of 1200 to 1800 kcal/d tailored to their starting weight, and an MVPA goal that gradually increased to 150 or more minutes per week. All program components were delivered through the RxWL website platform except weekly, automated reminder emails to submit self-monitoring data. Participants could contact the study team for assistance with technical support only (eg, password reset), not for weight loss counseling. Patients received no financial compensation for their participation in the trial.
Maintenance Interventions
Participants received their assigned maintenance intervention regardless of weight change after 3 months of the RxWL treatment phase. Maintenance consisted of 3 arms: control, monthly, and refresher. Participants in the control maintenance group received 9 monthly online newsletters delivered via the online platform. Content included education and reinforced skills from the initial treatment.
Both monthly and refresher maintenance groups were provided 9 maintenance sessions (1 introductory and 8 delivering content) on the online platform; content and goals were similar, but the delivery schedules differed. Lessons reinforced initial program goals and focused on successful behavioral maintenance strategies,28,29 such as coping with boredom and connecting goals with personal values. Participants were instructed to use an evidence-based23 self-regulation approach involving strategies tailored to their weight trajectory (ie, no weight regain, small regain, large regain), and to continue increasing MVPA to 200 or more minutes per week. One week of self-monitoring and automated feedback, as occurred in the initial 3-month treatment, accompanied each lesson.
Participants in the monthly maintenance group received a maintenance lesson and engaged in self-monitoring with automated feedback each month. Participants in the refresher maintenance group received an introductory lesson and two 4-week periods of intervention at 7 and 10 months. Each intervention period in the refresher maintenance group involved a 2.27-kg weight loss goal, weekly lessons, and self-monitoring with automated feedback.
Measures
Participants self-reported age, sex assigned at birth (female or male), and racial and ethnic background at enrollment. Race was categorized as American Indian or Alaska Native, Asian or Pacific Islander, Black or African American, White, or other (including multiracial or chose not to disclose). Ethnicity was categorized as Hispanic or Latinx or not Hispanic or Latinx. Data on race and ethnicity were collected and analyzed because prior research often finds effects of race or ethnicity on obesity treatment outcomes, and to be consistent with the methods used in previous studies. Height and weight were collected from the EMR. All weight measurements within 1 month (ie, 30 days) before enrolling in the RxWL program and through 24 months (ie, 720 days) after enrollment were included. Given the trial’s pragmatic nature, clinicians were not asked to weigh participants for data collection. The online platform recorded lessons viewed and self-monitoring data entered by the participant. Data were collected from June 2018 to March 2022.
Statistical Analysis
Analyses were conducted from August 2022 to September 2023 using SPSS Statistics, version 26 (IBM). A sample size of 567 was needed to detect a mean (SD) change of 2.0 (5.0) kg among weight loss maintenance interventions with a power of 0.8 and a type I error rate of 5%; we rounded the sample size to 600 participants based on expected rates of data availability.26 The threshold for statistical significance was P < .05 and the tests were 2 sided. The sample analyzed included all participants enrolled until the target sample size was achieved.
Descriptive statistics were computed for baseline characteristics. Independent samples t tests and Pearson χ2 were used to test for differences in demographics between participants with any weight data in the EMR during the 2-year study period (ie, the analysis sample) and those without data available for analysis.
The analysis examining weight change was conducted via a piecewise linear mixed-effects model using maximum likelihood estimation, following the intention-to-treat principle. The model included a knot at 3 months (ie, end of the 3-month weight loss phase and beginning of the maintenance phase) and a fixed linear effect of time on the primary outcome of weight loss. A dummy coded effect indicator for maintenance intervention was allowed to interact with the time effect to test for different trajectories in weight from 3 months to 12 and 24 months. Covariates of participant age, sex, and racial or ethnic minority status (yes or no) were included as fixed effects. The National Institutes of Health–defined race and ethnicity categories were used. The categories representative of minority status included American Indian, Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, Black or African American, Hispanic or Latino. Random intercepts accounted for nesting of observations within participants, including the variance in the number of observations each participant contributed to the analysis. Least-squares mean estimates of weight loss were calculated. We also conducted a sensitivity analysis using only participants (n = 143) who had a weight measure in the EMR both within 30 days of the start and within 30 days of the end of the 3-month treatment. Models that included the effect of practice did not converge; this parameter was therefore discarded.
Of the 540 participants included in the intention-to-treat analyses, 77 (14.3%) never engaged with any intervention during either the 3-month weight loss phase or the 9-month maintenance phase (ie, never viewed any lessons nor entered any self-monitoring information). An additional 210 participants (38.9%) engaged with the initial 3-month weight loss program but did not engage with maintenance. Thus, to compare the 3 maintenance strategies, we conducted a secondary per-protocol analysis of the 253 participants who engaged with the maintenance intervention by viewing at least 1 lesson or entering self-monitoring information at least once during the 9-month maintenance period. We compared weight regain, number of lessons viewed, and days of self-monitoring between groups. Although only participants in the monthly and refresher maintenance groups were asked to self-monitor and received feedback, the ability to enter self-monitoring information remained continuously available to all participants. Participants were categorized as having engaged with over half of the recommended maintenance activities (4 lessons or 30 days of self-monitoring) or with less than half of the activities. We then examined the association between program engagement and weight change during maintenance by adding the engagement variables to the linear mixed effects model for weight change.
Results
Patient Enrollment
As shown in Figure 1, 1765 patients were referred to the program, and 721 (41%) consented and were assessed for eligibility. Of these patients, 654 enrolled in the program and were randomized; 540 (82.5%) had any EMR weight data available, allowing them to be analyzed. We were unable to determine specific reasons for lack of data availability for the 114 participants without EMR data, but the most common reason was the patient leaving the practice network. Demographic characteristics are shown in Table 1; no differences were detected when comparing the 540 participants with EMR data vs those without such data (n = 114).
Figure 1. Flow of Participants Through the Trial.
BMI indicates body mass index (calculated as weight in kilograms divided by height in meters squared); EMR, electronic medical record; and ITT, intention-to-treat.
Table 1. Baseline Characteristics of Participants Included and Excluded in Analysis and by Weight Loss Maintenance Intervention.
| Characteristic | No. (%) | ||||
|---|---|---|---|---|---|
| Analysis sample (N = 540) | Excluded from analysis (n = 114) | Weight loss maintenance intervention | |||
| Monthly (n = 174) | Refresher (n = 185) | Control (n = 181) | |||
| Age, mean (SD), y | 52.8 (13.4) | 52.1 (12.8) | 53.4 (12.3) | 51.9 (14.6) | 53.2 (13.1) |
| Sex | |||||
| Female | 384 (71.1) | 76 (66.7) | 126 (72.4) | 129 (69.7) | 129 (71.3) |
| Male | 156 (28.9) | 38 (33.3) | 48 (27.6) | 56 (30.3) | 52 (28.7) |
| Race | |||||
| American Indian or Alaska Native | 1 (0.2) | 2 (1.8) | 0 | 1 (0.5) | 0 |
| Asian or Pacific Islander | 8 (1.5) | 1 (0.9) | 1 (0.6) | 5 (2.7) | 1 (0.6) |
| Black or African American | 13 (2.4) | 6 (5.3) | 2 (1.1) | 3 (1.6) | 8 (4.4) |
| White | 509 (94.3) | 102 (89.4) | 170 (97.7) | 172 (93.0) | 168 (92.8) |
| Othera | 9 (1.7) | 3 (2.6) | 1 (0.6) | 4 (2.2) | 4 (2.2) |
| Ethnicity | |||||
| Hispanic or Latinx | 15 (2.8) | 2 (1.8) | 5 (2.9) | 6 (3.2) | 4 (2.2) |
| Not Hispanic or Latinx | 525 (97.2) | 112 (98.2) | 169 (97.1) | 179 (96.8) | 177 (97.8) |
| BMI, mean (SD) | 36.0 (7.1) | 36.8 (6.0) | 36.0 (7.2) | 35.9 (6.0) | 36.2 (8.1) |
Abbreviation: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).
Other includes multiracial or chose not to disclose.
Patient Characteristics
The mean (SD) age of the 540 participants included in the intention-to-treat analysis was 52.8 (13.4) years; 384 participants were female (71.1%) and 156 were male (28.9%). One participant (0.2%) was American Indian or Alaska Native, 8 (1.5%) were Asian or Pacific Islander, 13 (2.4%) were Black or African American, 509 (94.3%) were White, and 9 (1.7%) were of other race; additionally, 15 participants (2.8%) identified as Hispanic or Latinx. The mean (SD) participant BMI was 36.0 (7.1), and participants entered their weight into the EMR a mean (SD) 4.8 (2.6) times.
Weight Change Outcomes
In the intention-to-treat analysis, the change in weight over the initial 3-month treatment period was estimated to be −3.60 (95% CI, −4.32 to −2.88] kg; P < .001) (Figure 2A). A sensitivity analysis including only the 143 participants who had entered their weight in the EMR both within 30 days of enrolling in RxWL and within 30 days of completing the 3-month treatment phase confirmed this finding (weight change, −3.86 [95% CI, −5.02 to −2.70] kg; P < .001) and yielded an estimated 3-month weight loss of 3.77% (95% CI, −4.90% to −2.65%; P < .001).
Figure 2. Weight Change Outcomes at 12 and 24 Months in Participants in the Rx Weight Loss Program.
Mean (SE) weight change over the 2-year study period for the intention-to-treat analysis (N = 540) (A) and the per-protocol analysis (B) with participants who engaged with both the initial weight loss treatment and the maintenance intervention (n = 253).
As shown in Table 2, participants assigned to the newsletter control maintenance group had regained significantly more weight at the 12-month primary end point (1.28 [95% CI, 0.85 to 1.71] kg) than those in the monthly (0.37 [95% CI, −0.06 to 0.81] kg) or refresher maintenance groups (0.45 [95% CI, 0.27 to 0.87] kg); P = .004). Weight regain in the monthly and refresher maintenance groups did not differ significantly from each other. This pattern persisted through 24 months. Collapsing across all 3 maintenance groups, age (coefficient = 0.03 [95% CI, –0.02 to 0.08]; P = .21), female sex (coefficient = –1.16 [95% CI, −2.40 to 0.80]; P = .07), and racial or ethnic minority status (coefficient = 0.65 [95% CI, −1.55 to 2.85]; P = .56) were not associated with rate of weight regain.
Table 2. Estimated Weight Loss Outcomes at 3, 12, and 24 Months Among Participants in Intention-to-Treat and Per-Protocol Analyses.
| No. | Weight loss from baseline, mean (SE), kg | Weight regain from 3 mo, kg (% initial weight loss) | ||||
|---|---|---|---|---|---|---|
| 3 mo | 12 mo | 24 mo | 12 mo | 24 mo | ||
| Weight loss in all participants (ITT analysis; N = 540) | ||||||
| Monthly maintenance group | 174 | −3.60 (0.37) | −3.23 (0.22) | −2.72 (0.52) | 0.37 (10) | 0.88 (24) |
| Refresher maintenance group | 185 | −3.60 (0.37) | −3.15 (0.21) | −2.55 (0.50) | 0.45 (13) | 1.05 (29) |
| Newsletter control maintenance group | 181 | −3.60 (0.37) | −2.32 (0.22) | −0.59 (0.52) | 1.28 (36) | 3.01 (84) |
| Weight loss in participants engaged during both initial and maintenance phases (PP analysis; n = 253) | ||||||
| Monthly maintenance group | 88 | −6.19 (0.54) | −5.58 (0.32) | −4.76 (0.75) | 0.61 (10) | 1.43 (23) |
| Refresher maintenance group | 88 | −6.19 (0.54) | −5.23 (0.32) | −3.94 (0.76) | 0.96 (15) | 2.25 (36) |
| Newsletter control maintenance group | 77 | −6.19 (0.54) | −4.33 (0.33) | −1.82 (0.79) | 1.86 (30) | 4.37 (71) |
Abbreviations: ITT, intention-to-treat; PP, per-protocol.
The 253 participants in the per-protocol analysis who engaged with maintenance intervention were evenly distributed by condition (monthly group, n = 88; refresher group, n = 88; and control group, n = 77; χ2 = 2.49; P = .29) and were older than participants who did not engage with the maintenance intervention (mean [SD] age, 55.3 [12.5] years vs 50.6 [13.8] years), but otherwise did not differ on demographics or baseline BMI. The estimated weight loss in the 3-month treatment phase for the per-protocol sample was 6.19 (95% CI, −7.25 to −5.13) kg. However, although the magnitude of weight loss and regain were greater in this subgroup, the pattern of weight regain was similar to that in the primary intention-to-treat analysis, with the greatest weight regain in the control group and nonsignificant differences between the monthly and refresher maintenance groups (Figure 2B and Table 2).
Program Engagement
In the 253 participants in the per-protocol analysis, there were no differences among the maintenance groups for mean number of maintenance lessons viewed or mean number of days of self-monitoring (Table 3). However, almost half (47.7%) of participants in the monthly maintenance group viewed at least 4 of the 9 maintenance lessons compared with 28.4% and 33.8% of participants in the refresher and newsletter control maintenance groups (P = .02). Approximately 10% more participants in the monthly maintenance group self-monitored for at least 30 days during maintenance compared with the other 2 groups, but the difference was not significant (Table 3).
Table 3. Engagement in Program Maintenance Among Participants in the Per-Protocol Analysis Who Viewed at Least 1 Lesson During Maintenance or Entered Any Self-Monitoring Information on the Web Platform.
| Engagement measure | Weight loss maintenance intervention | P value | ||
|---|---|---|---|---|
| Monthly (n = 88) | Refresher (n = 88) | Control (n = 77) | ||
| No. of lessons viewed, mean (SD) | 3.7 (3.3) | 2.6 (3.1) | 2.8 (3.2) | .36 |
| No. of days of self-monitoring, mean (SD) | 88.7 (85.6) | 78.1 (88.6) | 68.9 (78.4) | .17 |
| Participants who viewed at least 4 lessons, No. (%) | 42 (47.7) | 25 (28.4) | 26 (33.8) | .02 |
| Participants who self-monitored at least 30 d, No. (%) | 58 (65.9) | 48 (54.5) | 41 (53.3) | .18 |
Collapsing data across the 3 maintenance groups, a greater number of days of self-monitoring and lessons viewed were significantly associated with less weight regain. Participants who viewed 4 or more lessons (n = 90) had less weight regain at 12 months relative to those who were less engaged (0.46 [95% CI, –0.15 to 1.06] kg vs 1.51 [95% CI, 1.01 to 2.00] kg; P = .004). Likewise, participants who monitored for 30 or more days (n = 134) had less weight regain relative to those who were less engaged (0.96 [95% CI, 0.18 to 1.20] kg vs 1.68 [95% CI, 1.11 to 2.26] kg; P = .005).
Discussion
To our knowledge, this is the first truly pragmatic trial that tested a fully automated online weight management program for primary care patients with overweight or obesity and the first trial to compare directly a refresher and a monthly maintenance schedule. Contrary to our hypothesis, these 2 active maintenance interventions produced similar results and both were more effective than the newsletter control maintenance intervention.
Using EMR weight data allowed access to posttreatment weight measurements for all participants with EMR data (N = 540), including those who never started the program (14.3%) and an additional 38.9% who did not engage with the maintenance intervention. In the intention-to-treat analysis, the average 3-month weight loss was 3.6 kg. Over the next 9 months, participants randomized to 1 of the active maintenance conditions regained 10% to 13% of their initial weight loss; the control group regained 36%. At 24 months, those randomized to the newsletter control maintenance group had regained 84% of their initial weight loss, whereas those assigned to an active maintenance group regained only 24% to 29%. These findings suggest that the contact schedule was not as important as we hypothesized20,25,26 but that providing some type of ongoing intervention reduces weight regain.
In the per-protocol analysis with 273 participants who engaged with the maintenance intervention, mean weight loss was 6.19 kg at 3 months, suggesting that these engaged participants likely adhered better to initial treatment recommendations and thus achieved better weight losses than unengaged participants. These outcomes are similar to the 3-month weight loss of 5.1 kg reported in our previous analysis of self-reported weight change that included only participants who engaged with the initial 3-month RxWL intervention.27 As in the intention-to-treat analysis, the per-protocol analysis showed that participants in the refresher and monthly maintenance groups better maintained their weight loss than those in the newsletter control maintenance group. Retained weight losses at 12 months and 24 months in the 2 active maintenance groups (Table 2) suggest that a fully automated online program with an active maintenance intervention can produce clinically significant weight loss11 in primary care comparable to effects of clinician-supported online treatments.7,12,30
There have been numerous trials of online weight loss treatment in primary care, but most involved human interaction or coaching.7,8,9,12,30,31 These studies show that comprehensive online programs supplemented with human contact can produce weight loss of 3 to 5 kg at 1 year, with pragmatically implemented programs showing weight losses of 2 to 3 kg at 1 year in intention-to-treat analyses.30,31 To our knowledge, the present study is the first to show that a contactless online program can produce clinically meaningful weight loss at 12 months and 24 months (Table 2) in primary care patients with overweight or obesity who follow an active maintenance schedule (Figure 2). The weight losses associated with the RxWL program are modest but consistent with obesity treatment recommendations,11 and there are few other interventions that can be made available to nearly any primary care patient at no cost and deployed with minimal disruption to clinical workflows. Furthermore, the per-protocol analysis demonstrates that greater treatment engagement was associated with larger and more clinically important weight losses.
The high rate of intervention disengagement in the current trial is notable and common in primary care obesity treatment generally; 14.3% of the 540 participants in the intention-to-treat sample never started the RxWL program, and 38.9% of the 470 participants who started treatment never engaged during the maintenance phase. Although these findings may be due to the pragmatic design, other online programs in primary care have also been associated with low uptake and high disengagement.5,12,30,32 Research is needed to understand the reasons for low enrollment in weight loss and maintenance programs so researchers can develop strategies to address these barriers.33
In the per-protocol analysis of maintenance-engaged participants, the number of lessons viewed was low (most often, 3 of 9 lessons), replicating findings from prior studies.5,12,30,32,33 As our group reported in a previous study,27 during the treatment phase, participants in the trial viewed 6.5 of the first 12 videos, and 37% of participants watched all 12 initial lessons, suggesting a decline in engagement over time that is common in online interventions.8,12,16,30,32,33 Nevertheless, participants who watched more lessons or self-monitored on more days had better long-term weight losses. This confirms the association between engagement with weight management strategies and the outcomes achieved.16,17,33
Strengths and Limitations
This study has several strengths. To our knowledge, this is the only pragmatic trial of an automated online treatment for overweight and obesity in primary care that was delivered without any ongoing researcher or clinician contact. The control intervention in this trial was more involved than usual care; participants in the newsletter control maintenance group received monthly online newsletters reinforcing behavioral skills and could continue to self-monitor via the platform. Finally, the engagement data were collected automatically on the study website and the weight loss data were from the EMR, allowing us to examine adherence and weight changes in the full cohort of participants, regardless of whether they accessed the RxWL program.
This study also has limitations. Body weight measurements in EMR data are often not standardized (eg, taken with shoes or clothing at varied times of day) and do not occur at predetermined intervals.34 Some participants could not be included in analysis due to lack of weight information in the EMR, and it was not possible to determine whether weight loss was due to illness. The number of weight measurements also varied across individuals; however, the statistical approach used accommodates for these differences. Despite the pragmatic nature of this trial and a population in Rhode Island characterized by 17.2% racial minority membership and 17.6% Hispanic or Latino membership,35 participants were primarily White and non-Hispanic. The sample’s homogeneity underscores the need for future work to better engage patients of diverse racial and ethnic backgrounds in various clinical contexts.
Conclusions
This randomized clinical trial shows that a fully automated weight loss program can be pragmatically implemented and delivered in a large primary care system and can produce meaningful weight losses that are sustained at 2 years. These findings underscore the importance of providing ongoing maintenance intervention to prevent weight regain and suggest that a maintenance intervention can be offered successfully with different schedules of contact.
Trial Protocol and Statistical Analysis
Data Sharing Statement
References
- 1.Ward ZJ, Bleich SN, Cradock AL, et al. Projected US state-level prevalence of adult obesity and severe obesity. N Engl J Med. 2019;381(25):2440-2450. doi: 10.1056/NEJMsa1909301 [DOI] [PubMed] [Google Scholar]
- 2.Ciciurkaite G, Moloney ME, Brown RL. The incomplete medicalization of obesity: physician office visits, diagnoses, and treatments, 1996-2014. Public Health Rep. 2019;134(2):141-149. doi: 10.1177/0033354918813102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Levine DM, Linder JA, Landon BE. Characteristics of Americans with primary care and changes over time, 2002-2015. JAMA Intern Med. 2020;180(3):463-466. doi: 10.1001/jamainternmed.2019.6282 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Haire-Joshu D, Klein S. Is primary care practice equipped to deal with obesity? comment on “Preventing weight gain by lifestyle intervention in a general practice setting”. Arch Intern Med. 2011;171(4):313-315. doi: 10.1001/archinternmed.2011.3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Grady A, Yoong S, Sutherland R, Lee H, Nathan N, Wolfenden L. Improving the public health impact of eHealth and mHealth interventions. Aust N Z J Public Health. 2018;42(2):118-119. doi: 10.1111/1753-6405.12771 [DOI] [PubMed] [Google Scholar]
- 6.Hutchesson MJ, Rollo ME, Krukowski R, et al. eHealth interventions for the prevention and treatment of overweight and obesity in adults: a systematic review with meta-analysis. Obes Rev. 2015;16(5):376-392. doi: 10.1111/obr.12268 [DOI] [PubMed] [Google Scholar]
- 7.Xiao L, Yank V, Wilson SR, Lavori PW, Ma J. Two-year weight-loss maintenance in primary care-based Diabetes Prevention Program lifestyle interventions. Nutr Diabetes. 2013;3(6):e76. doi: 10.1038/nutd.2013.17 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Appel LJ, Clark JM, Yeh HC, et al. Comparative effectiveness of weight-loss interventions in clinical practice. N Engl J Med. 2011;365(21):1959-1968. doi: 10.1056/NEJMoa1108660 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.McTigue KM, Conroy MB, Hess R, et al. Using the internet to translate an evidence-based lifestyle intervention into practice. Telemed J E Health. 2009;15(9):851-858. doi: 10.1089/tmj.2009.0036 [DOI] [PubMed] [Google Scholar]
- 10.Joiner KL, Nam S, Whittemore R. Lifestyle interventions based on the Diabetes Prevention Program delivered via eHealth: a systematic review and meta-analysis. Prev Med. 2017;100:194-207. doi: 10.1016/j.ypmed.2017.04.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jensen MD, Ryan DH, Donato KA, et al. ; American College of Cardiology/American Heart Association Task Force on Practice Guidelines, Obesity Expert Panel, 2013 . Executive summary: guidelines (2013) for the management of overweight and obesity in adults. Obesity (Silver Spring). 2014;22(suppl 2):S5-S39. [DOI] [PubMed] [Google Scholar]
- 12.Baer HJ, Rozenblum R, De La Cruz BA, et al. Effect of an online weight management program integrated with population health management on weight change: a randomized clinical trial. JAMA. 2020;324(17):1737-1746. doi: 10.1001/jama.2020.18977 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Leahey TM, Thomas G, Fava JL, et al. Adding evidence-based behavioral weight loss strategies to a statewide wellness campaign: a randomized clinical trial. Am J Public Health. 2014;104(7):1300-1306. doi: 10.2105/AJPH.2014.301870 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Thomas JG, Leahey TM, Wing RR. An automated internet behavioral weight-loss program by physician referral: a randomized controlled trial. Diabetes Care. 2015;38(1):9-15. doi: 10.2337/dc14-1474 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wadden TA, Butryn ML, Hong PS, Tsai AG. Behavioral treatment of obesity in patients encountered in primary care settings: a systematic review. JAMA. 2014;312(17):1779-1791. doi: 10.1001/jama.2014.14173 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Conroy MB, McTigue KM, Bryce CL, et al. Effect of electronic health record–based coaching on weight maintenance: a randomized trial. Ann Intern Med. 2019;171(11):777-784. doi: 10.7326/M18-3337 [DOI] [PubMed] [Google Scholar]
- 17.Bennett GG, Herring SJ, Puleo E, Stein EK, Emmons KM, Gillman MW. Web-based weight loss in primary care: a randomized controlled trial. Obesity (Silver Spring). 2010;18(2):308-313. doi: 10.1038/oby.2009.242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Thorpe KE, Zwarenstein M, Oxman AD, et al. A pragmatic-explanatory continuum indicator summary (PRECIS): a tool to help trial designers. J Clin Epidemiol. 2009;62(5):464-475. doi: 10.1016/j.jclinepi.2008.12.011 [DOI] [PubMed] [Google Scholar]
- 19.MacLean PS, Wing RR, Davidson T, et al. NIH working group report: innovative research to improve maintenance of weight loss. Obesity (Silver Spring). 2015;23(1):7-15. doi: 10.1002/oby.20967 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Dutton GR, Gowey MA, Tan F, et al. Comparison of an alternative schedule of extended care contacts to a self-directed control: a randomized trial of weight loss maintenance. Int J Behav Nutr Phys Act. 2017;14(1):107. doi: 10.1186/s12966-017-0564-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Levy RL, Jeffery RW, Langer SL, et al. Maintenance-tailored therapy vs. standard behavior therapy for 30-month maintenance of weight loss. Prev Med. 2010;51(6):457-459. doi: 10.1016/j.ypmed.2010.09.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tsai AG, Felton S, Wadden TA, Hosokawa PW, Hill JO. A randomized clinical trial of a weight loss maintenance intervention in a primary care population. Obesity (Silver Spring). 2015;23(10):2015-2021. doi: 10.1002/oby.21224 [DOI] [PubMed] [Google Scholar]
- 23.Wing RR, Tate DF, Gorin AA, Raynor HA, Fava JL. A self-regulation program for maintenance of weight loss. N Engl J Med. 2006;355(15):1563-1571. doi: 10.1056/NEJMoa061883 [DOI] [PubMed] [Google Scholar]
- 24.Dombrowski SU, Knittle K, Avenell A, Araújo-Soares V, Sniehotta FF. Long term maintenance of weight loss with non-surgical interventions in obese adults: systematic review and meta-analyses of randomised controlled trials. BMJ. 2014;348:g2646. doi: 10.1136/bmj.g2646 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Jeffery RW, Levy RL, Langer SL, et al. A comparison of maintenance-tailored therapy (MTT) and standard behavior therapy (SBT) for the treatment of obesity. Prev Med. 2009;49(5):384-389. doi: 10.1016/j.ypmed.2009.08.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Espel-Huynh HM, Wing RR, Goldstein CM, Thomas JG. Rationale and design for a pragmatic effectiveness-implementation trial of online behavioral obesity treatment in primary care. Contemp Clin Trials. 2019;82:9-16. doi: 10.1016/j.cct.2019.05.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Thomas JG, Panza E, Espel-Huynh HM, et al. Pragmatic implementation of a fully automated online obesity treatment in primary care. Obesity (Silver Spring). 2022;30(8):1621-1628. doi: 10.1002/oby.23502 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Diabetes Prevention Program (DPP) Research Group . The Diabetes Prevention Program (DPP): description of lifestyle intervention. Diabetes Care. 2002;25(12):2165-2171. doi: 10.2337/diacare.25.12.2165 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wadden TA, West DS, Delahanty L, et al. ; Look AHEAD Research Group . The Look AHEAD study: a description of the lifestyle intervention and the evidence supporting it. Obesity (Silver Spring). 2006;14(5):737-752. doi: 10.1038/oby.2006.84 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hoerster KD, Hunter-Merrill R, Nguyen T, et al. Effect of a remotely delivered self-directed behavioral intervention on body weight and physical health status among adults with obesity: the D-ELITE randomized clinical trial. JAMA. 2022;328(22):2230-2241. doi: 10.1001/jama.2022.21177 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Little P, Stuart B, Hobbs FDR, et al. An internet-based intervention with brief nurse support to manage obesity in primary care (POWeR+): a pragmatic, parallel-group, randomised controlled trial. Lancet Diabetes Endocrinol. 2016;4(10):821-828. doi: 10.1016/S2213-8587(16)30099-7 [DOI] [PubMed] [Google Scholar]
- 32.Neve MJ, Collins CE, Morgan PJ. Dropout, nonusage attrition, and pretreatment predictors of nonusage attrition in a commercial Web-based weight loss program. J Med Internet Res. 2010;12(4):e69. doi: 10.2196/jmir.1640 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kozak AT, Buscemi J, Hawkins MA, et al. Technology-based interventions for weight management: current randomized controlled trial evidence and future directions. J Behav Med. 2017;40(1):99-111. doi: 10.1007/s10865-016-9805-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Gianfrancesco MA, Goldstein ND. A narrative review on the validity of electronic health record–based research in epidemiology. BMC Med Res Methodol. 2021;21(1):234. doi: 10.1186/s12874-021-01416-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.US Census Bureau . QuickFacts Rhode Island. Accessed November 22, 2023. https://www.census.gov/quickfacts/fact/table/RI/PST045222
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Trial Protocol and Statistical Analysis
Data Sharing Statement