Effect of Allowing Choice of Diet on Weight Loss: A Randomized Trial

Abstract

Background: Choosing a diet rather than being prescribed one could improve weight loss.

Objective: To examine whether offering choice of diet improves weight loss.

Design: Double-randomized preference trial of choice between 2 diets (choice) versus random assignment to a diet (comparator) over 48 weeks. (ClinicalTrials.gov: NCT01152359).

Setting: Outpatient clinic at a Veterans Affairs medical center.

Patients: Outpatients with a body mass index of at least 30 kg/m2.

Intervention: Choice participants received information about their food preferences and 2 diet options (low-carbohydrate diet [LCD] or low-fat diet [LFD]) before choosing and were allowed to switch diets at 12 weeks. Comparator participants were randomly assigned to 1 diet for 48 weeks. Both groups received group and telephone counseling for 48 weeks.

Measurements: The primary outcome was weight at 48 weeks.

Results: Of 105 choice participants, 61 (58%) chose the LCD and 44 (42%) chose the LFD; 5 (3 on the LCD and 2 on the LFD) switched diets at 12 weeks, and 87 (83%) completed measurements at 48 weeks. Of 102 comparator participants, 53 (52%) were randomly assigned to the LCD and 49 (48%) were assigned to the LFD; 88 (86%) completed measurements. At 48 weeks, estimated mean weight loss was 5.7 kg (95% CI, 4.3 to 7.0 kg) in the choice group and 6.7 kg (CI, 5.4 to 8.0 kg) in the comparator group (mean difference, -1.1 kg [CI, -2.9 to 0.8 kg]; P = 0.26). Secondary outcomes of dietary adherence, physical activity, and weight-related quality of life were similar between groups at 48 weeks.

Limitations: Only 2 diet options were provided. Results from this sample of older veterans might not be generalizable to other populations.

Conclusion: Contrary to expectations, the opportunity to choose a diet did not improve weight loss.

Figures

Appendix 3. Figure

Smoothed spline trajectories of weight over 48 weeks by diet type and arm (blue and red solid and dashed lines) and observed mean trajectories of weight over 48 weeks by diet type and arm (light blue and pink solid and dashed lines).

Figure 1

Participant flow.

Figure 2

Estimated mean weight trajectories in kg over 48 weeks for the Choice and Comparator arms from linear mixed models.