Age Effects on Delay Discounting across the Lifespan A Meta-Analytical Approach to Theory Comparison and Model Development

Abstract

A long-running debate about the developmental trajectory of delay discounting has received growing attention since 1994. Relevant theories, ranging from developmental psychology and evolutionary biology to behavioral economics, yield contradictory predictions. Encompassing a wide age range from 6.7 to 83.1 years, we evaluated these theories based on meta-analyses of 178 effect sizes from 105 articles that examined age-dependent delay discounting, providing up-to-date the most comprehensive review of the topic. Delay discounting decreased with advancing age (Fisher’s z = −.059). However, meta-regression suggested that this linear trend masked a U-shaped function, as implied by some theoretical models. We developed a derivative-based method and recovered this nonlinear function based on 58 effects. Both the meta-regression based on all 178 effect sizes and the derivative-based method convergently demonstrated that delay discounting was the lowest for middle-aged people around 50, depending on the magnitude of the reward. The U-shaped function was steeper for people with shorter life expectancies; the turning point comes at a younger age for medium-to-large rewards, and delay discounting models explained heterogeneity across studies. We expanded the current theoretical frameworks by synthesizing the life history theory and the antagonistic pleiotropy hypothesis. Built upon the mortality–fertility trade-off model of Sozou and Seymour (2003), we postulated the role of parental investment in postponing the increase of delay discounting in late adulthood, suggesting a three-way mortality–fertility–parenting trade-off. Possible proximate mechanisms were also discussed. Overall, when allocating assets over temporal intervals, a higher delay discount rate accompanies a lower future reproductive opportunity. (PsycInfo Database Record (c) 2023 APA, all rights reserved)