Weight change among repeat participants of an Aboriginal community-based weight loss program

Erika Bohn-Goldbaum,Rose Fonua,Anne C Grunseit,Adrian Bauman,Andrew Milat,Aaron Cashmore,Leah Shepherd,Kate Reid

doi:10.1186/s12889-020-09086-6

Abstract

BackgroundCommunity-based weight loss programs may have potential to address overweight and obesity at the population level. However, participation patterns and individual outcomes from these programs are understudied. This study examined repeat participation patterns and participant weight change between contests over seven years of an Aboriginal Australian team-based program in order to identify (1) predictors of repeat participation and (2) associations with weight change between contests.MethodsData for the 12 contests from 2012 to 2018 were merged, with probabilistic record matching. A total of 7510 enrolments were registered for the 12 contests, representing 4438 unique people. Contest lengths varied from 10 to 16 weeks in duration. Non-repeat participants were those who only competed once in the program by the end of 2018, and repeaters were those who competed in at least two contests. Associations between repeat participation and participant baseline (i.e., first participation occasion) characteristics, change in diet and physical activity and percent change in weight during the first participation occasion were examined using crossed random effects (for person and team) regression adjusted for exposure to the program. Weight percentage change between contests was calculated for consecutive participation occasions occurring at least three months apart, converted to percent change per month. Weight change was regressed on number of repeat participation occasions adjusted for age, gender, baseline weight at first participation occasion, and weight percent change in the immediately preceding contest.ResultsOne-third of the 4433 participants participated more than once, with women more likely than men to repeat. A 1% reduction in weight during a competition was associated with an increase in weight of 0.05% per month between competition end and subsequent participation. Regain was smaller the heavier participants were at their first participation.ConclusionsWhile individuals benefit from weight loss through program participation, strengthening strategies for weight loss maintenance within or following the program could improve long-term weight outcomes and reduce weight cycling.

Highlights

Weight loss interventions among Aboriginal Australians have proven effective in reducing weight and noncommunicable disease risk [1,2,3,4,5]
Outcomes from community-based weight loss programs can be difficult to establish due to the variety of program designs and reporting inconsistencies [10]; never-the-less, weight reduction and reduced risk of non-communicable disease have been demonstrated in other populations [11,12,13,14]
We recently evaluated a community-based team weight loss competition, the New South Wales (NSW) Aboriginal Knockout Health Challenge (KHC) [15], a program open to Aboriginal Australians and Torres Strait Islander peoples aged 16 years or older

Summary

Methods

Data collection and treatment Participant outcomes are measured pre-post contest for prize allocation; written consent (provided by the participant, and, for those under 18 years of age, by their parent/legal guardian with the underage participant providing written assent) allows use of the data for prize calculation and research purposes. A participant weighing 100 kg at the start of contest 7 and 95 kg at the end of contest 6 (~ 7 months prior), would have a change of 5.3%; a positive number demonstrates a weight gain between contests. For likelihood of repeat participation, data comprised individuals who had formed teams of ≥20 persons at their first participation occasion (leaving 4312 people) and those whose first participation was prior to contest 12 (n = 4104 people). Associations between repeat participation and change in diet and PA (calculated by subtracting baseline from post-intervention scores for each contest) and percent change in weight during the first participation occasion were further adjusted for gender and age, as these variables have been shown previously to affect contest outcomes [15]. All analyses were conducted using Stata 15.1 (College Station, TX, USA)

Results

Introduction

Discussion

Strengths and limitations

Conclusion