Delay Discount Rate Moderates a Physical Activity Intervention Testing Immediate Rewards

Abstract

Financial incentives can increase physical activity (PA), but differences in the immediacy of reward delivery and individual differences in delay discount rates (i.e., higher discount values associated with less tolerance for delayed rewards) may explain differential responding. The current study tested whether delay discount rate moderated the relative effectiveness of immediate financial rewards on increasing daily PA. Inactive, overweight adults (ages 18-60, N = 96) were randomized to receive either smaller, immediate goal-contingent rewards or larger, delayed rewards for participation. Delay discount rates were derived for those who completed the Monetary Choice Questionnaire (N = 85). Linear mixed models tested interactions between discount rate and intervention arm on changes in mean daily Fitbit-measured steps from baseline to intervention phases, and rates of change during the intervention phase. Across all groups, participants increased by 2258 steps/day on average from baseline to intervention and declined by 9 steps/day across the 4-month intervention phase. The mean increase in daily steps was greater for immediate reward-arm participants across all discount rates. Descriptive exploration of reward effects by delay discount rate suggested that the magnitude of reward effects decreased at higher discount rates. During the 4-month intervention phase, rates of decline in daily steps were similar in both reward arms, but declines became more pronounced at higher discount rates. Overall, intervention efficacy decreased with less tolerance for delays. The importance of financial reward immediacy for increasing PA appears to increase with greater delay discount rates.

Keywords: delay discounting; financial incentives; impulsive behavior; intervention; physical activity; reward schedule.

Figure 1.

Model-estimated mean daily steps by reward arm and delay discount rate

Figure 2.

Model-estimated daily steps across the intervention phase by reward arm at delay discount rates = −2.0, −1.0, 0, 1.0, and 2.0. DDz = delay discount rate (ln geomean k) Z-scores. Higher delay discount rate values indicate lower tolerance for delayed rewards.

Figure 3.

Percent decrease in model-estimated mean daily steps from peak attainment to end of intervention phase by reward arm at all estimated delay discount rates