Financial incentive-based approaches for weight loss: a randomized trial

Abstract

Context: Identifying effective obesity treatment is both a clinical challenge and a public health priority due to the health consequences of obesity.

Objective: To determine whether common decision errors identified by behavioral economists such as prospect theory, loss aversion, and regret could be used to design an effective weight loss intervention.

Design, setting, and participants: Fifty-seven healthy participants aged 30-70 years with a body mass index of 30-40 were randomized to 3 weight loss plans: monthly weigh-ins, a lottery incentive program, or a deposit contract that allowed for participant matching, with a weight loss goal of 1 lb (0.45 kg) a week for 16 weeks. Participants were recruited May-August 2007 at the Philadelphia VA Medical Center in Pennsylvania and were followed up through June 2008.

Main outcome measures: Weight loss after 16 weeks.

Results: The incentive groups lost significantly more weight than the control group (mean, 3.9 lb). Compared with the control group, the lottery group lost a mean of 13.1 lb (95% confidence interval [CI] of the difference in means, 1.95-16.40; P = .02) and the deposit contract group lost a mean of 14.0 lb (95% CI of the difference in means, 3.69-16.43; P = .006). About half of those in both incentive groups met the 16-lb target weight loss: 47.4% (95% CI, 24.5%-71.1%) in the deposit contract group and 52.6% (95% CI, 28.9%-75.6%) in the lottery group, whereas 10.5% (95% CI, 1.3%-33.1%; P = .01) in the control group met the 16-lb target. Although the net weight loss between enrollment in the study and at the end of 7 months was larger in the incentive groups (9.2 lb; t = 1.21; 95% CI, -3.20 to 12.66; P = .23, in the lottery group and 6.2 lb; t = 0.52; 95% CI, -5.17 to 8.75; P = .61 in the deposit contract group) than in the control group (4.4 lb), these differences were not statistically significant. However, incentive participants weighed significantly less at 7 months than at the study start (P = .01 for the lottery group; P = .03 for the deposit contract group) whereas controls did not.

Conclusions: The use of economic incentives produced significant weight loss during the 16 weeks of intervention that was not fully sustained. The longer-term use of incentives should be evaluated.

Trial registration: clinicaltrials.gov Identifier: NCT00520611.

Figure 1. Flow of study

Figure 2. Diagram of Sample Participant Weight Loss Goals

This figure demonstrates the sample weight loss goal for a subject who at study initiation weighed 250 pounds. Each subject was given the goal weight loss of 1 pound per week for 16 weeks. The “Fresh Start” trajectory illustrates an approach that was taken to minimize drop-out in subjects who were unable to initially stay on target. For example, if this subject lost 1 pound after 4 weeks, the slope of their weight loss goal targets was simply adjusted so that while they still had the same goal after 16 weeks, he/she would not have to suddenly lose 3 pounds to become eligible for continued incentives.

Figure 3. Weight loss by arm in control and incentive groups from enrollment to the end of 7 months

Study participants in each of the incentive arms regained weight between 4 and 7 months following enrollment, though participants in the incentive arms experienced net weight loss between enrollment and 7 months and participants in the control arm did not. Weight loss was significantly greater in the incentive conditions relative to control at 4 months (L: p = .014; DC: p = .003), but not at 7 months (L: p = .23; DC: p = .61).