Adult Age Differences in Neural Representations of Value in Time, Probability and Effort Discounting Tasks
Kendra Seaman
1 Yale
University,
1, Teresa
2 TU
Karrer
1,2,
3
Dresden, Vanderbilt
Background
Nickolas Brooks
University,
4 University
1,
Linh Dang
3,
Ming Hsu
4,
David H Zald
3,
Gregory R
1
Samanez-Larkin
of California Berkeley
Effort
Time
Probability
Neural Subjective Value Representation
• Real world decisions involve integrating rewards with other discounting
factors such as time delays, probability, or physical effort.
• Discounting benefits by such factors generates a subjective value
(SV) - or common currency - to maximize utility.
• A recent meta-analysis1 supports the common currency hypothesis,
showing subjective values are represented in:
• medial prefrontal cortex (mPFC)
• posterior parietal cortex (PCC)
• ventral striatum (vStr).
• However, these meta-analytic approaches
focused on healthy young adults.
• Using tasks that integrate monetary rewards with
either time delays, probability, or physical
Bartra et al.
effort, we examined adult age differences in:
• discount rates
• neural SV representations
• neural representation of reward and discounting factors.
• Age differences only appeared when SV was decomposed into reward
magnitude and each individual discounting factor.
Best fitting k was used to generate trial-by-trial parametric modulators of SV tailored to each individual. p < .005.
• Consistent with prior studies in young adults, a network of regions including
the mPFC and vStr correlated with SV2-4.
• Contrary to prior studies, there were no age differences in discount rates
(k) or the representation of SV in the brain5-6.
• Despite age differences in the representations of reward magnitude and each
discounting factor, the lack of age differences in the representation of SV and
behavior suggest that these value-related signals are integrated similarly
across adulthood.
• Controlling for age, a common network of regions including the MPFC and vStr were correlated with SV.
• No age differences in the representation of SV.
• It is possible that age differences exist, but this study design was not
sensitive enough to detect these age differences.
Neural Reward Magnitude and Discounting Factor Representation
For each task, C represents either:
• Effort: Proportion of maximum ability
• Probability: Odds against winning (1-P(win)/P(win))
• Time: Time delay in days
Data was fit with three decision rules:
• Softmax with free decision slope
• Softmax with decision slope = 1
• Hardmax + e
• Best fitting models were determined using Akaike Weights.
• Monetary rewards may not be as salient for older adults.
• Other, more relevant rewards (social or health) have been shown to elicit
age differences in behavior7.
Raw reward and discount factors were used to generate trial-by-trial parametric modulators. p < .005.
• Future studies should use these other reward domains.
• In this sample, we will also examine:
Participants
77 healthy participants completed the study at Vanderbilt University
Mean Age (SD) = 49.86 (17.94), Age Range = 22 – 83, 44 Female, 33 Male
Computational Modeling
Subjective values were modeled using a hyperbolic discount function:
• Most effects were a reduction in representation with age.
• These effects differed for each discounting factor.
Method
Procedures
• Effort: choices between a smaller reward with a lower level of
physical effort and a larger reward with a higher level of physical
effort.
• Probability: choices between a smaller reward with a higher
probability and a larger reward with a lower probability.
• Time: choices between a smaller reward with a shorter time delay
and a larger reward with a longer time delay.
Conclusions and Future Directions
• distributed representations of SV using MVPA
• the relationship between SV representations and dopamine D2 receptor
availability.
$11.41
$22.82
Level 1
Level 4
$11.41
100%
$
$22.82
25%
$11.41
$22.82
Today
6 Months
References
• Reduced representation of reward
magnitude with age in a network of
regions, including:
• bilateral inferior frontal gyrus (IFG)
• caudate
• thalamus.
• Reduced representation of probability
with age in a network of regions,
including:
• left superior parietal lobule
• caudate.
• Increased representation of time delay
with age in inferior parietal lobule.
• Reduced representation of time delay
with age in supplementary motor area.
P < .005 P < .001 P < .0005
Discount Rates and Age
1.
Bartra O, McGuire JT, Kable JW (2013) The valuation system: A coordinate-based meta-analysis of BOLD
fMRI experiments examining neural correlates of subjective value. NeuroImage 76:412–427.
2.
Peters J, Buchel C (2009) Overlapping and distinct neural systems code for subjective value during
intertemporal and risky decision making. Journal of Neuroscience 29:15727–15734.
3.
Burke CJ, Brunger C, Kahnt T, Park SQ, Tobler PN (2013) Neural integration of risk and effort costs by the
frontal pole: only upon request. Journal of Neuroscience 33:1706–1713.
4.
Massar, S. A. A., Libedinsky, C., Weiyan, C., Huettel, S. A., & Chee, M. W. L. (2015). Separate and
overlapping brain areas encode subjective value during delay and effort discounting. NeuroImage,
120:104–113.
5.
Eppinger, B., Nystrom, L. E., & Cohen, J. D. (2012). Reduced sensitivity to immediate reward during
decision-making in older than younger adults. PloS One, 7(5), e36953.
6.
Halfmann, K., Hedgcock, W., Kable, J., & Denburg, N. L. (2015). Individual differences in the neural
signature of subjective value among older adults. Social Cognitive and Affective Neuroscience, 2016:
1111-1120.
7.
Seaman, K.L., Gorlick, Vekaria, K.M., M.A.,Hsu, M., Zald, D.H., & Samanez-Larkin, G.R. (in press) Adult
age differences in decision making across domains: Increased discounting of social and health-related
rewards. Psychology and Aging.
There were no relationships between discount rates (k) and age:
Acknowledgments
r2 = 0.033; p = .113
r2 = 0.022; p = .192
r2 = 0.010; p = .378
This research was supported by National Institute on Aging Pathway to Independence Award
R00-AG042596 to GRSL and National Institute on Aging grant R01-AG044838 to DHZ and
GRSL. We thank Jaime Castrellon and Scott Perkins for assistance with data collection.
Correspondence should be addressed to: [email protected]