Brief report
Efficacy of a Weight-Loss Website Based on Positive Deviance: A Randomized Trial

https://doi.org/10.1016/j.amepre.2011.08.012Get rights and content

Background

Web-based interventions for weight control could promote more-widespread dissemination of weight-loss opportunities; however, they remain limited in effectiveness. Incorporating weight-control practices used by individuals with long-term weight-loss success (“positive deviants”) may improve the efficacy of web-based weight control interventions.

Purpose

To evaluate the efficacy of AchieveTogether, a web-based weight-loss intervention for adults based on user-generated weight-loss strategies from successful weight losers.

Design

In 2009–2010, participants were randomized to either a 12-week web-based intervention, AchieveTogether, or a wait-list control condition.

Setting/participants

100 overweight or obese adults participated in the study.

Intervention

AchieveTogether was designed to help individuals implement weight control practices used by others who successfully lost and maintained weight.

Main outcome measures

The primary outcome was change in weight. Secondary outcomes included blood pressure, daily caloric intake, quality of life, and use of weight control practices. ANCOVA, with adjustment for baseline values on outcome measures, was used to assess differences between groups in primary and secondary outcomes. Statistical analyses were conducted in 2010–2011.

Results

Most participants were women (69.7%) and white (90.8%), with a mean age of 50.3 years and baseline BMI of 33.2; 88% completed post-program assessments. Mean weight loss among intervention participants was −1.4 kg (95% CI= −2.2, −0.5), compared with a mean weight gain of 0.6 kg (95% CI= −0.3, 1.4) in control participants (p<0.01).

Conclusions

User-generated weight-loss practices from “positive deviants” could promote weight control in web-based interventions; however, further research is needed to improve program efficacy.

Trial registration

This study is registered at ClinicalTrials.gov NCT00709501.

Introduction

The obesity epidemic has widespread adverse health and economic effects.1 Internet-based weight-loss programs could help address the obesity epidemic, as they can be widely disseminated with low costs. Existing Internet-based weight-loss programs have largely promoted weight-loss strategies designed by health professionals, including goal-setting and feedback, and features to promote social interaction.2, 3, 4, 5, 6 These programs have had modest short-term effects, suggesting that exploring alternative approaches may be beneficial.

The present study evaluated the efficacy of using a positive deviance framework to guide the design of an Internet-based weight-loss intervention.7 The central premise of positive deviance is that solutions to population-based problems often exist within a population, and that population members' solutions can be generalized to improve other members' performance.8 Positive deviance uses qualitative research to identify the behaviors of individuals who have obtained rare success (typically top 10% of the population) at achieving specific outcomes. These behaviors are then disseminated through community engagement and behavior modification strategies. Positive deviance has been used in diverse health-related interventions, but has not previously been used in weight-loss interventions.9, 10, 11, 12, 13, 14

The present Internet-based intervention evaluated the efficacy of promoting weight-loss behaviors that were generated from a prior qualitative study of individuals who maintained long-term weight loss of at least 30 lbs (positive deviants).7 It was hypothesized that an Internet-based intervention that disseminated weight-loss behaviors of these positive deviants could promote weight loss among overweight and obese adults.

Section snippets

Design

The study was conducted at a single institution (Penn State Hershey Medical Center) in 2009–2010 with 100 participants randomized to the 12-week AchieveTogether intervention or wait-list control condition with in-person assessments at baseline and post-program (12 weeks). This sample size enabled 90% power to detect a 2.0-kg difference between conditions in the primary outcome of weight loss,3, 6 assuming 20% attrition. Randomization took place at the baseline visit using concealed envelopes.

Participants

Baseline characteristics of participants are shown in Table 1; no significant group differences were observed. At post-program, treatment completers (n=88) were older than those lost to follow-up (n=12; mean age = 50.3 vs 43.4 years, p<0.05); no other differences were noted.

Weight Change and Health Outcomes

Table 2 shows weight-loss and health outcomes among study completers. There was a 2.0-kg difference in weight change between groups (p<0.01). The difference in weight change between groups remained significant with

Discussion

The present study evaluated the efficacy of a website based on positive deviance for disseminating weight-control behaviors used by individuals successful with weight loss.7 In contrast to the expert-driven approach that has guided the design of many other weight-loss websites,3, 6 the AchieveTogether website design was driven by successful solutions that individuals have used to manage their weight.

The use of aggregate qualitative data on weight-loss behaviors from those who successfully lost

References (29)

  • D.F. Tate et al.

    Using Internet technology to deliver a behavioral weight loss program

    JAMA

    (2001)
  • H.L. Stuckey et al.

    Using positive deviance for determining successful weight-control practices

    Qual Health Res

    (2010)
  • D.R. Marsh et al.

    The power of positive deviance

    BMJ

    (2004)
  • D.R. Marsh et al.

    The positive deviance approach to improve health outcomes: experience and evidence from the fieldIntroduction

    Food Nutr Bull

    (2002)
  • Cited by (0)

    View full text