Teamlead.duke-nus.edu.sg

Financial Incentive–Based Approaches
for Weight Loss
A Randomized Trial
Context Identifying effective obesity treatment is both a clinical challenge and a pub-
lic 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 designan 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: monthlyweigh-ins, a lottery incentive program, or a deposit contract that allowed for partici- pant matching, with a weight loss goal of 1 lb (0.45 kg) a week for 16 weeks. Par- ticipants 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 to help reduce the rate of obesity in the 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 lotterygroup 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 con- Conclusions The use of economic incentives produced significant weight loss dur-
ing the 16 weeks of intervention that was not fully sustained. The longer-term use of Trial Registration clinicaltrials.gov Identifier: NCT00520611
efits, such as enjoying good health.4,5Drawing on prior work suggesting thatthe same decision errors that hurt Author Affiliations are listed at the end of this article.
Corresponding Author: Kevin G. Volpp, MD, PhD,
Center for Health Incentives, Leonard Davis Institute
of Health Economics, University of Pennsylvania School of Medicine and the Wharton School, VA Center for Health Equity Research and Promotion, 1232 Block-ley Hall, 423 Guardian Dr, Philadelphia, PA 19104- 2008 American Medical Association. All rights reserved.
(Reprinted) JAMA, December 10, 2008—Vol 300, No. 22 2631
Randomization Procedures
Figure 1. Flow of Study Participants
Participants were randomized evenly tothe 3 groups using a block size of 6, with 958 Potentially eligible individuals were stratification based on sex and age(30-49 years vs 50-70 years). Sealed en- velopes generated by the statisticianwere used within each of the strata so that the research coordinator who en-rolled the participants did not know the randomization assignment of the nextparticipant until it had been assigned.
dinator could be blinded to the ran-domization assignment given the na- Study Protocol
delphia Veterans Affairs Medical Cen-ter and the University of Pennsylvania.
participants were randomly assignedto participate in either a weight- monitoring program involving monthlyweigh-ins, or the same program with1 of 2 financial incentive plans (de- posit contract or lottery). At initial en- covering diet and exercise strategies for a free scale with precision to 0.2 lb (to convert to kilograms, multiply by 0.45).
given a chart at the initial visit depict- ticipation either infeasible (inability to Study Population
consent, illiteracy, participation in an- is shown in FIGURE 1. Participants
ments were recruited using mailings.
if they met or exceeded their weight loss goal. As an incentive for participants to race or ethnicity selecting from a list on 2632 JAMA, December 10, 2008—Vol 300, No. 22 (Reprinted)
2008 American Medical Association. All rights reserved.
goal at the end of the month weigh-in.
the end of each month to be weighed.
tract group could earn as little as $0 or 10/19 in the lottery group), of those, 18 but only if, prior to the lottery being re- or below their weight loss goal. The lot- who lost at least 20 lb). In addition, all Statistical Analyses
ated every day. If the first digit gener- ated was a “2” or the last digit was a“7” intervals for those failing to attain goals.
tent-to-treat analyses, adjusted for base- ber was “27” (a 1 in 100 chance), the deposit contract group, those in the lot- using F tests. All participants lost to fol- rate and still attain the final 16-lb weight higher than their weight loss goal at the start” in which the overall weight loss “binge diet” to resume receiving incen- nating, record their weights, and call in their weight to the project staff by noon.
ticipant lost 2 lb instead of 4 lb in first 2008 American Medical Association. All rights reserved.
(Reprinted) JAMA, December 10, 2008—Vol 300, No. 22 2633
tions using 2-sided t tests to compare for the stratification variables (sex and in the baseline characteristics of any of by Pearson ␹2 test for categorical vari- the groups (TABLE 1). The sample was
ables and the t test or Wilcoxon rank ticipants, a greater proportion of the in- Table 1. Characteristics of the Study Samplea
No. / Total (%)
Entire Sample
Contract
Participant Characteristics
Median total annual household income from all sources Self-rated importance of controlling weight measured on a Confidence in ability to lose weight measured on a 0-10 scale, Abbreviations: BMI, body mass index, calculated as weight in kilograms divided by height in meters squared; GED, general education development.
Conversion factor: To convert from pounds to kilograms, multiply by 0.45; from feet to meters, multiply by 0.3; and inches to centimeters, multiply by 2.54.
a No significant differences were found between groups in any of these conditions.
2634 JAMA, December 10, 2008—Vol 300, No. 22 (Reprinted)
2008 American Medical Association. All rights reserved.
Table 2. 16-Week Weight Loss Measures by Group
(F1,35 = 6.20; P = .02; 95% CI of the Contract
1 , 3 5 = 8.55; P = .006; 95% CI of (TABLE 2). Although only 10.5% (95%
Abbreviation: CI, confidence interval.
Conversion factor: To convert pounds to kilograms, multiply by 0.45.
a Difference between incentive and control conditions significant at PՅ.05.
contract group (␹22=8.59, P=.01). Theodds of achieving the 16-lb weight lossgoal were significantly greater in both Figure 2. Weight Loss From Enrollment
Through Intervention and 7-Month
the daily call-in rate was extremely high vealed qualitatively similar patterns re- gardless of age, income, or initial BMI.
tery group (F2.43; P=.13, 95% CI of the for blacks; t = 3.57; P = .001; 95% CI of justing for race did not affect the rela- Participants in each group regained weight following of 7 months (FIGURE 2). Although the
the conclusion of the intervention. At 4 months, those in the incentive groups lost significantly more weightthan those in the control group (P=.02 in the lottery group and P=.006 in the deposit contract group), but at the 7-month follow-up after enrollment, the weightdifference between groups was no longer statisti- for the lottery group; 6.2 lb for the de- cally significant (P=.23 for the lottery group and P=.61 in the deposit contract group). Nevertheless, those in the incentive groups experienced a net loss betweenenrollment and at the 7-month interval, whereas those were not statistically significant (t=1.21; in the control group did not. Error bars indicate 95% 2 = 5.58, P = .06). These participants P = .23; 95% CI, −3.20 to 12.66 lb for the lottery group; t = 0.52; P = .61; 95% CI, −5.17 to 8.75 lb for the deposit con- (t18= −1.97; P = .06, 95% CI, −9.19 to posit contract participants was less than at baseline to only a marginally signifi- study (t18 = −2.87, P = .01; 95% CI, cant degree (P = .10); otherwise, fol- −15.89 to −2.47 for the lottery group; t18= −2.41; P = .03; 95% CI, −11.67 to 2008 American Medical Association. All rights reserved.
(Reprinted) JAMA, December 10, 2008—Vol 300, No. 22 2635
change that is based on loss aversion, a weight loss. Lost to follow-up rates were concept of decision isolation; people re- effects identified in the behavioral eco- is the relative cost-effectiveness of dif- end of the weight loss phase and the fol- are particularly emotionally attracted to weight loss in studies of 6 to 12 month’s loss lost 4.7 pounds and participants of- frey and colleagues12,13 had earlier dem- be useful. The persistence of weight loss feedback was an intrinsic part of the in- 2636 JAMA, December 10, 2008—Vol 300, No. 22 (Reprinted)
2008 American Medical Association. All rights reserved.
Role of the Sponsor: Neither the sponsors nor the fun-
challenge of reducing cocaine abuse. In: Chaloupka ders had any role in the design and conduct of the FJ, Grossman M, Bickel WK, Saffer H, eds. The Eco- study; collection, management, analysis, and inter- nomic Analysis of Substance Use and Abuse. Cam- pretation of the data; and preparation, review, or ap- bridge, MA: National Bureau of Economic Research; Additional Contributions: We thank Daniel McDonald,
17. Giuffrida A, Torgenson DJ. Should we pay the
MS, Carnegie Mellon University, Pittsburgh, Pennsyl- patient? review of financial incentives to enhance significant weight loss. Study staff could vania, and Evan Robinson, BS, University of Waterloo, patient compliance. BMJ. 1997;315(7110):703- Waterloo, Ontario, Canada, for their assistance.
18. Volpp KG, Pauly MV, Loewenstein G, Bangsberg
REFERENCES
D. P4P4P: an agenda for research on pay for perfor- mance for patients. Health Aff (Millwood). In press.
1. Flegal KM, Carroll MD, Ogden CL, Johnson CL.
19. Schroeder SA. We can do better—improving the
Prevalence and trends in obesity among US adults, health of the American people. N Engl J Med. 2007; 1999-2000. JAMA. 2002;288(14):1723-1727.
2. Ogden CL, Carroll MD, Curtin LR, McDowell MA,
20. Ainslie G. Specious reward: a behavioral theory
Tabak CJ, Flegal KM. Prevalence of overweight and of impulsiveness and impulse control. Psychol Bull.
obesity in the United States, 1999-2004. JAMA. 2006; 21. Thaler RH. Some empirical evidence on dynamic
different incentive program designs.
3. Flegal KM, Graubard BI, Williamson DF, Gail MH.
inconsistency [letter]. Economics Stud. 1981;23: Excess deaths associated with underweight, over- weight, and obesity. JAMA. 2005;293(15):1861- 22. Loewenstein G, Prelec D. Anomalies in intertem-
poral choice: evidence and an interpretation. Q J Econ.
appear to be highly effective in inducing 4. Laibson DI. Golden eggs and hyperbolic discounting.
initial weight loss. However, this weight Q J Econ. 1997;112(2):443-477.
23. Kirby K. Bidding on the future: evidence against
5. O’Donoghue T, Rabin M. Doing it now or later.
normative discounting of delayed rewards. J Exp Psy- Am Econ Rev. 1999;89(1):103-124.
6. Loewenstein G, Brennan T, Volpp KG. Asymmet-
24. Camerer C, Ho T-H. Experience-weighted attrac-
ric paternalism to improve health behaviors. JAMA.
tion: learning in normal form games. Econometrica.
7. Ryan DH, Espeland MA, Foster GD, et al. Look
25. Loewenstein GF, Weber EU, Hsee CK, Welch N.
AHEAD (Action for Health in Diabetes): design Risk as feelings. Psychol Bull. 2001;127(2):267- Author Affiliations: Center for Health Equity Re-
and methods for a clinical trial of weight loss for search and Promotion, Philadelphia Veterans Affairs the prevention of cardiovascular disease in type 2 26. Connolly T, Butler DU. Regret in economic and
Medical Center, Philadelphia (Dr Volpp and Ms Nor- diabetes. Control Clin Trials. 2003;24(5):610- psychological theories of choice. J Behav Decis Making.
ton); Center for Health Incentives, Leonard Davis In- stitute of Health Economics (Drs Volpp, Troxel, and 8. National Institutes of Health. Clinical Guidelines on
27. Chapman GB, Coups EJ. Emotions and preven-
Loewenstein and Ms Norton), Department of Medi- the Identification, Evaluation, and Treatment of Over- tive health behavior: worry, regret, and influenza cine, University of Pennsylvania School of Medicine weight and Obesity in Adults–The Evidence Report vaccination. Health Psychol. 2006;25(1):82-90.
(Dr Volpp and Ms Norton), Department of Health Care [published correction appears in Obes Res.
28. Bandura A. Principles of Behavior Modification.
Management, the Wharton School (Dr Volpp), and 1998;6(6):464] Obes Res. 1998;6(suppl 2):51S- New York, NY: Holt, Rinehart and Winston, Inc; 1969.
the Center for Clinical Epidemiology and Biostatistics 29. Kahneman DR, Tversky A. Prospect theory: an
and Department of Biostatistics and Epidemiology, Uni- 9. Knowler WC, Barrett-Connor E, Fowler SE, et al.
analysis of decision under risk. Econometrica. 1979; versity of Pennsylvania, Philadelphia (Drs Troxel and Reduction in the incidence of type 2 diabetes with life- Fassbender); and the Department of Social and De- style intervention or metformin. N Engl J Med. 2002; 30. Thaler RH, Tversky A, Kahneman DR, The AS. The
cision Sciences, Carnegie Mellon University, Pitts- effect of myopia and loss aversion on risk taking: an 10. Committee to Develop Criteria for Evaluating the
experimental test. Q J Econ. 1997;112(2):647- Author Contributions: Dr Volpp had full access to all
Outcomes of Approaches to Prevent and Treat Obe- of the data in the study and takes responsibility for sity, Institute of Medicine. Weighing the Options: Cri- 31. Rizzo JA, Zeckhauser RJ. Reference incomes, loss
the integrity of the data and the accuracy of the data teria for Evaluating Weight-Management Programs. aversion, and physician behavior. Rev Econ Stat. 2003; Thomas PR, ed; Washington, DC: National Acad- Study concept and design: Volpp¸ Loewenstein.
32. Camerer C. Three cheers—psychological, theo-
Acquisition of data: Volpp, John, Norton, Fassbender 11. Finkelstein EA, Linnan LA, Tate DF, Birken BE. A
retical, empirical—for loss aversion. J Mark Res. 2005; Analysis and interpretation of data: Volpp, John, pilot study testing the effect of different levels of fi- nancial incentives on weight loss among overweight 33. Read D, Loewenstein G, Rabin M. Choice bracketing.
Drafting of the manuscript: Volpp, John, Loewenstein.
employees. J Occup Environ Med. 2007;49(9): J Risk Uncertain. 1999;19(1-3):171-197.
Critical revision of the manuscript for important in- 34. McTigue KM, Harris R, Hemphill B, et al. Screen-
tellectual content: Volpp, John, Troxel, Norton, 12. Jeffery RW, Gerber WM, Rosenthal BS, Lindquist
ing and interventions for obesity in adults: summary RA. Monetary contracts in weight control: effective- of the evidence for the US Preventive Services Task Statistical analysis: Volpp, John, Troxel, Loewenstein.
ness of group and individual contracts of varying size.
Force. Ann Intern Med. 2003;139(11):933-949.
Obtained funding: Volpp, Loewenstein.
J Consult Clin Psychol. 1983;51(2):242-248.
35. Strategic plan for NIH Obesity Research: A re-
Administrative, technical, or material support: Volpp, 13. Jeffery RW, Thompson PD, Wing RR. Effects on
port of the NIH Obesity Research Task Force. August weight reduction of strong monetary contracts for calo- 2006; http://www.obesityresearch.nih.gov.Ac- Study supervision: Volpp, Loewenstein.
rie restriction or weight loss. Behav Res Ther. 1978; Financial Disclosures: None reported.
36. Obesity: Preventing and Managing the Global
Funding/Support: This work was supported by grant P30
14. Higgins ST, Wong CJ, Badger GJ, Ogden DE,
Epidemic. Geneva, Switzerland: World Health Orga- AG12836 from the National Institute on Aging, the Dantona RL. Contingent reinforcement increases co- Boettner Center for Pensions and Retirement Security caine abstinence during outpatient treatment and 1 37. Mokdad AH, Bowman BA, Ford ES, Vinicor F,
at the University of Pennsylvania, grant R24 HD-044964 year of follow-up. J Consult Clin Psychol. 2000; Marks JS, Koplan JP. The continuing epidemics of obe- from National Institute of Child Health and Develop- sity and diabetes in the United States. JAMA. 2001; ment Population Research Infrastructure Program, all 15. Sindelar JL. Paying for performance: the power
at the University of Pennsylvania, and grant 58-4000- of incentives over habits. Health Econ. 2008;17 38. Encinosa WE, Bernard DM, Chen CC, Steiner CA.
70058 from the US Department of Agriculture, Eco- Healthcare utilization and outcomes after bariatric nomic Research Service, and the Hewlett Foundation.
16. Higgins ST. Applying behavioral economics to the
surgery. Med Care. 2006;44(8):706-712.
2008 American Medical Association. All rights reserved.
(Reprinted) JAMA, December 10, 2008—Vol 300, No. 22 2637

Source: https://teamlead.duke-nus.edu.sg/vapfiles_ocs/2011/MSIII/Deciphering%20Medical%20Literature/Session4/Financial%20Incentive%20Based%20Approaches.pdf