- Email: [email protected]
Delay Discounting Correlates with Proportional Lateral Frontal Cortex Volumes
Delay Discounting Correlates with Proportional Lateral Frontal Cortex Volumes James M. Bjork, Reza Momenan, and Daniel W. Hommer Background: Functional neuroimaging experiments in healthy control subjects have shown that choosing between small and immediate rewards versus larger but deferred rewards in delay discounting (DD) tasks recruits mesofrontal and lateral frontal cortex. Might individual differences in frontocortical gray matter morphology be related to preference for immediate reward? Methods: We related DD in a laboratory decision-making task to proportional frontocortical gray matter (GM) volumes calculated from segmented magnetic resonance images in 29 healthy adults. Results: Dorsolateral and inferolateral frontal cortex GM volumes (corrected as a proportion of whole cerebral brain volume) each correlated inversely with preference for immediate gratification during decision making, as indexed by DD constant k. Conversely, neither proportional orbitofrontal or mesofrontal cortex GM volume nor cerebral brain volume (CBV) or total intracranial volume (ICV; a measure of maximal brain growth) significantly correlated with severity of DD. Conclusions: Severity of discounting of delayed rewards correlates with proportional lateral frontocortical GM morphology but not with whole brain measures. In light of evidence of frontocortical abnormalities in substance dependence and sociopathy, future studies can assess whether reduced frontocortical volume itself is a morphological marker or risk factor for inability to delay gratification in psychiatric disorders.
Key Words: Brain, decision making, delay discounting, frontal cortex, gray matter, impulsivity, magnetic resonance imaging, MRI
T
he effective worth of a reward decreases with how long a person must wait to receive it (delay-discounting [DD]) in a hyperbolic function (1). Some individuals show an exaggerated preference for immediate rewards, even if an alternative deferred reward is much larger (1). Seminal functional magnetic resonance imaging (fMRI) experiments have indicated that mesofrontal cortex (mFC) (2-4) and inferolateral (ilFC) and dorsolateral frontal cortex (dlFC) (3,4) were recruited when healthy adults chose between smaller immediate versus larger delayed rewards in DD tasks, where ilFC and dlFC activation was greatest when larger deferred reward was chosen (4). These findings further implicate dlFC in executive behavior control and the mFC in incentive valuation. A link between frontocortical structure and valuation of deferred rewards has important psychiatric implications in that willpower to resist using drugs is thought to be frontocortically mediated (5). A deficit in either the ability to generate an internal representation of the deferred benefits of abstinence or an inability of this representation to be adequately referenced during decision making could confer vulnerability to the development of (or relapse to) substance abuse (1). Drug-dependent individuals show increased preference for immediately-presented rewards over larger but delayed rewards in DD tasks (6,7), and in longitudinal studies, both heroin-dependent pa-
From the Division of Clinical Neuroscience and Behavioral Research (JMB), National Institute on Drug Abuse; and Laboratory of Clinical and Translational Studies (RM, DWH), National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, Maryland. Address reprint requests to James M. Bjork, Ph.D., NIDA/DCNBR, 6001 Executive Boulevard, Room 3151, Bethesda, MD 20892; E-mail: jbjork@mail. nih.gov. Received September 24, 2008; revised November 3, 2008; accepted November 15, 2008.
0006-3223/09/$36.00 doi:10.1016/j.biopsych.2008.11.023
tients (8) and pregnant smokers (9) with more severe discounting of delayed rewards at baseline were more likely to relapse. Questionnaire-measured impulsivity in community-recruited persons has been linked to reduced prefrontal gray matter (GM) volume (10,11). Might reduced frontocortical GM volumes also relate directly to severity of laboratory-measured DD? We explored whether frontocortical gray matter volume relates to severity of DD among healthy subjects. We related the DD constant k, a coefficient that describes severity of discounting, to each of the following: proportional frontocortical gray matter volume, cerebral brain volume (CBV), and total intracranial volume (ICV) derived from segmented magnetic resonance imaging (MRI). Based on seminal fMRI exploration of DD, we hypothesized that reduced ilFC and dlFC volume (as a proportion of whole brain volume) would correlate with severity of DD.
Methods and Materials Procedures were approved by the Institutional Review Board of the National Institute on Alcohol Abuse and Alcoholism, and all subjects provided written informed consent. Subjects All subjects were neurologically and medically healthy per physical and radiological examinations. Drug abstinence was verified in urine and breath samples. History of psychosis, seizures, or an estimated IQ ⬍ 80 was exclusionary for any applicant. Community-recruited control subjects (n ⫽ 29; age 20 –58 [mean 37.4 ⫾ 11.0]; 18 male subjects) did not meet DSM-IV criteria for any Axis I disorder in structured clinical interviews. Delay-Discounting Task Using a desktop computer, subjects were presented with 137 choices between a $10 standard reward (to be presented at delays of 0, 7, 30, 180, or 365 days) or an immediately-presented reward ranging from 25¢ to $10. The subject was told he/she would actually receive one randomly selected choice. At each delay to the standard reward, a switchpoint dollar value of the immediate alternative was calculated, below which the subject BIOL PSYCHIATRY 2009;65:710 –713 © 2009 Society of Biological Psychiatry
BIOL PSYCHIATRY 2009;65:710 –713 711
J.M. Bjork et al. would opt to wait for the full standard reward. The discounting constant k was calculated with a least-square fit with switchpoints at each delay to the standard $10 reward. The k values were then log-transformed (hereafter lnK) to achieve a normal distribution. Higher values of k (raw or log-transformed) indicate preference for immediate rewards (Supplement 1). Questionnaire Impulsivity The Barratt Impulsiveness Scale (BIS) (12) total and subscale scores were also collected as secondary measures of attentional, motor, and nonplanning impulsivity. Brain Volumetric Data Subjects were scanned with 1.5 T MRI (GE Medical Systems, Milwaukee, Wisconsin) using a fast-spoiled gradient-recalled acquisition in the steady state (GRASS) sequence and coronal acquisition. Brain volumes derived from coronal slices were rotated to a standard orientation and total intracranial volume was calculated by defining the inner skull surface with a handdriven cursor. The contents of the intracranial vault were then
A.
segmented into gray matter, white matter (WM), and cerebrospinal fluid (CSF) (13). To encompass frontocortical regions recruited during DD tasks, frontal cortex was calculated a priori as GM and WM anterior to the coronal slice that included the anterior commissure (AC). The mFC was defined as all frontal GM extending laterally from midline to the width of the lateral ventricles. Orbitolateral frontal cortex (olFC) was defined as the (remaining) nonmesial GM below the axial slice that encompassed the AC. The remaining lateral cortex was radially halved into inferolateral frontal cortex and dorsolateral frontal cortex. Frontocortical GM volumes were recalculated as an adjusted proportion of total cerebral brain volume before relation to behavioral measures. This corrected for individual and sex differences in whole brain volume, which relates to body size (14). Complete details of the brain volume calculation are in Supplement 1. Statistical Analysis The relationship between each of the four proportional frontocortical GM volumes and DD was assessed as a partial
B. -2
-2
-3
-4
-5
lnK
Beta = -.373, p < .05
-3
Beta = -.431, p < .05
-4
-5
lnK
-6
-6
-7
-7
-8
-8
-9
-9
-10
-10 2.0
2.5
3.0
3.5
3.0
Age-corrected GM volume
C.
-3
5.0
5.5
Beta = -.268, p = .17
-4 -5
-5
lnK
4.5
-2 -3
Beta = -.204, p = .32
-4
4.0
Age-corrected GM volume
D.
-2
3.5
lnK
-6
-6
-7
-7
-8
-8
-9
-9
-10
-10 3.5
4.0
4.5
5.0
Age-corrected GM volume
1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4
Age-corrected GM volume
Figure 1. Relationship between frontocortical proportional GM volumes and preference for immediate rewards as indexed by partial correlations between (bilateral) frontocortical GM volumes and log-transformed delay-discounting constant k. After controlling for age in multiple regression, log-transformed delay-discounting constant k correlated inversely with each of (A) dorsolateral frontal GM volume (corrected as a proportion of total cerebral brain volume) and (B) inferolateral frontal GM volume, where lower proportional GM volumes were associated with increased preference for immediate presented rewards over larger but delayed rewards. The log-transformed k constant did not partially correlate with either (C) mesofrontal or (D) orbitolateral frontal GM volumes or with total cerebral brain volume or total intracranial volume. GM, gray matter.
www.sobp.org/journal
712 BIOL PSYCHIATRY 2009;65:710 –713 correlation in simultaneous multiple regression. In each analysis, lnK was the dependent variable, with age and the brain volume measure as the independent variables. Age was included as a covariate in these analyses because aging reduces GM volume (15) as well as discounts behavior (16). We then repeated these analyses with BIS scores substituted for lnK.
Results The discounting curve predicted by the hyperbolic equation significantly correlated with actual subject decision switchpoints, with a median R2 value across subjects of .80. Proportional (CBV-adjusted) ilFC (Beta ⫽ ⫺.373, p ⬍ .05) and dlFC (Beta ⫽ ⫺.423, p ⬍ .05) GM volumes showed a negative partial correlation with lnK, where lower ilFC and dlFC GM volumes were associated with greater discounting of delayed rewards (Figure 1). In contrast, lnK did not partially correlate with either mFC (Beta ⫽ ⫺.204, p ⫽ .32) or olFC (Beta ⫽ ⫺.268, p ⫽ .17) GM volume. To determine whether brain-behavior relationships were specific to the frontal lobe measures, in multiple regression, total CBV and ICV were each substituted as the brain volumetric measure, while controlling for both age and sex. Neither CBV (p ⫽ .33) nor ICV (p ⫽ .40) partially correlated with discounting. In a secondary analysis, when BIS total and subscale scores were each substituted for lnK as the dependent impulsivity variable, none correlated with any frontocortical volume. Neither BIS total nor BIS subscale scores correlated with lnK.
Discussion In healthy adults, greater discounting of delayed rewards partially correlated with smaller proportional ilFC and dlFC (while controlling for age). Critically, ilFC and dlFC encompassed the lateral frontocortical voxels recruited during DD decision making in fMRI studies (3,4). In contrast, DD behavior did not correlate with CBV or with premorbid brain size as reflected in ICV. This frontocortical specificity is in accord with findings of Yang et al. (10) and Matsuo et al. (11), who found that prefrontal GM volume, but not whole brain volumes, correlated with questionnaire impulsivity. Unlike the Yang et al. (10) and Matsuo et al. (11) studies, however, frontocortical GM volumes did not correlate with questionnaire-measured impulsivity. We note, however, that BIS scores in our sample did not correlate with actual DD behavior. Divergent findings may have also resulted from differences in GM measurement. Critically, frontocortical GM in the Matsuo et al. (11) study was calculated using voxel-based morphometry, where each brain was warped into common stereotactic space for a probability-based assessment of each voxel. We measured actual GM volume, where brains were rotated only. We note that frontocortical GM volume only accounted for a modest portion of variance in DD, and this can be attributed to several factors. First, whether DD behavior is a state or trait quality is controversial (1). For example, DD behavior can be altered with pharmacological challenges (17) and experience of receiving deferred rewards from experimenters (18) and also differs as a function of local economic conditions (19) and remission from drug dependence (20). Second, while the large frontocortical GM regions defined here likely accommodated individual differences in precise localization of recruitment (as seen in mFC [2]), their size likely compromised specificity. Third, these control subjects generally discounted less (median k ⫽ www.sobp.org/journal
J.M. Bjork et al. .0034) than control subjects in many other experiments. This limited range of discounting likely reduced the ability to detect additional relationships between frontocortical measures and DD. Finally, we caution that our correlations between proportional frontocortical GM and k would not survive correction for multiple comparisons, but are reported here as explicitly hypothesis-driven. These data offer preliminary evidence that lateral frontocortical development may be a morphological contributor to DD, since this region is recruited when subjects opt for delayed rewards over a smaller immediate alternative (3,4). These results, if extended in clinical populations, would support opponentprocess accounts of externalizing disorders (e.g., 1,5), where impulsivity is unchecked by executive control neurocircuitry. Of interest in future research would be testing severely discounting subjects (such as substance-dependent individuals), especially in studies that jointly assess both frontocortical structure and function in conjunction with laboratory impulsivity. Finally, the advent of more functionally defined (and individually customized) cortical mapping can better define GM regions. Finally, if this finding can be replicated in children and adolescents, it would implicate reduced frontolateral cortex development as a premorbid risk factor for psychiatric disorders of impulse control.
This research was sponsored by intramural research funds of the National Institute on Alcohol Abuse and Alcoholism. We thank Michael Kerich and Khai Van for image processing and Dr. Suzanne Mitchell for her donation of the delay discounting task. The authors reported no biomedical financial interests or potential conflicts of interest. Supplementary material cited in this article is available online. 1. Bickel WK, Miller ML, Yi R, Kowal BP, Lindquist DM, Pitcock JA (2007): Behavioral and neuroeconomics of drug addiction: Competing neural systems and temporal discounting processes. Drug Alcohol Depend 90(suppl 1):S85–S91. 2. Kable JW, Glimcher PW (2007): The neural correlates of subjective value during intertemporal choice. Nat Neurosci 10:1625–1633. 3. McClure SM, Ericson KM, Laibson DI, Loewenstein G, Cohen JD (2007): Time discounting for primary rewards. J Neurosci 27:5796 –5804. 4. McClure SM, Laibson DI, Loewenstein G, Cohen JD (2004): Separate neural systems value immediate and delayed monetary rewards. Science 306:503–507. 5. Bechara A (2005): Decision making, impulse control and loss of willpower to resist drugs: A neurocognitive perspective. Nat Neurosci 8:1458 –1463. 6. Reynolds B (2006): A review of delay-discounting research with humans: Relations to drug use and gambling. Behav Pharmacol 17:651– 667. 7. Perry JL, Carroll ME (2008): The role of impulsive behavior in drug abuse. Psychopharmacology (Berl) 200:1–26. 8. Passetti F, Clark L, Mehta MA, Joyce E, King M (2008): Neuropsychological predictors of clinical outcome in opiate addiction. Drug Alcohol Depend 94:82–91. 9. Yoon JH, Higgins ST, Heil SH, Sugarbaker RJ, Thomas CS, Badger GJ (2007): Delay discounting predicts postpartum relapse to cigarette smoking among pregnant women. Exp Clin Psychopharmacol 15:176 –186. 10. Yang Y, Raine A, Lencz T, Bihrle S, LaCasse L, Colletti P (2005): Volume reduction in prefrontal gray matter in unsuccessful criminal psychopaths. Biol Psychiatry 57:1103–1108. 11. Matsuo K, Nicoletti M, Nemoto K, Hatch JP, Peluso MA, Nery FG, Soares JC (2008): A voxel-based morphometry study of frontal gray matter correlates of impulsivity [published online ahead of print May 8]. Hum Brain Mapp. 12. Patton JH, Stanford MS, Barratt ES (1995): Factor structure of the Barratt Impulsiveness Scale. J Clin Psychol 51:768 –774.
J.M. Bjork et al. 13. Momenan R, Hommer D, Rawlings R, Ruttimann U, Kerich M, Rio D (1997): Intensity-adaptive segmentation of single-echo T1-weighted magnetic resonance images. Hum Brain Mapp 5:194 –205. 14. Ho KC, Roessmann U, Straumfjord JV, Monroe G (1980): Analysis of brain weight. II. Adult brain weight in relation to body height, weight, and surface area. Arch Pathol Lab Med 104:640 – 645. 15. Bjork JM, Grant SJ, Hommer DW (2003): Cross-sectional volumetric analysis of brain atrophy in alcohol dependence: Effects of drinking history and comorbid substance use disorder. Am J Psychiatry 160: 2038 –2045. 16. Green L, Fry AF, Myerson J (1994): Discounting of delayed rewards: A lifespan comparison. Psychol Sci 5:33–36.
BIOL PSYCHIATRY 2009;65:710 –713 713 17. de Wit H, Enggasser JL, Richards JB (2002): Acute administration of d-amphetamine decreases impulsivity in healthy volunteers. Neuropsychopharmacology 27:813– 825. 18. Walls RT, Smith TS (1970): Development of preference for delayed reward in disadvantaged children. J Educ Psychol 61:118 –123. 19. Tucker B (2006): A future discounting explanation for the persistence of a mixed foraging-horticulture strategy among the Mikea of Madagascar. In: Kennett DJ, Winterhalder B, editors. Behavioral Ecology and the Transition to Agriculture. Berkeley, CA: University of California Press, 22– 40. 20. Bickel WK, Odum AL, Madden GJ (1999): Impulsivity and cigarette smoking: Delay discounting in current, never, and ex-smokers. Psychopharmacology (Berl) 146:447– 454.
www.sobp.org/journal
Key Words: Brain, decision making, delay discounting, frontal cortex, gray matter, impulsivity, magnetic resonance imaging, MRI
T
he effective worth of a reward decreases with how long a person must wait to receive it (delay-discounting [DD]) in a hyperbolic function (1). Some individuals show an exaggerated preference for immediate rewards, even if an alternative deferred reward is much larger (1). Seminal functional magnetic resonance imaging (fMRI) experiments have indicated that mesofrontal cortex (mFC) (2-4) and inferolateral (ilFC) and dorsolateral frontal cortex (dlFC) (3,4) were recruited when healthy adults chose between smaller immediate versus larger delayed rewards in DD tasks, where ilFC and dlFC activation was greatest when larger deferred reward was chosen (4). These findings further implicate dlFC in executive behavior control and the mFC in incentive valuation. A link between frontocortical structure and valuation of deferred rewards has important psychiatric implications in that willpower to resist using drugs is thought to be frontocortically mediated (5). A deficit in either the ability to generate an internal representation of the deferred benefits of abstinence or an inability of this representation to be adequately referenced during decision making could confer vulnerability to the development of (or relapse to) substance abuse (1). Drug-dependent individuals show increased preference for immediately-presented rewards over larger but delayed rewards in DD tasks (6,7), and in longitudinal studies, both heroin-dependent pa-
From the Division of Clinical Neuroscience and Behavioral Research (JMB), National Institute on Drug Abuse; and Laboratory of Clinical and Translational Studies (RM, DWH), National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, Maryland. Address reprint requests to James M. Bjork, Ph.D., NIDA/DCNBR, 6001 Executive Boulevard, Room 3151, Bethesda, MD 20892; E-mail: jbjork@mail. nih.gov. Received September 24, 2008; revised November 3, 2008; accepted November 15, 2008.
0006-3223/09/$36.00 doi:10.1016/j.biopsych.2008.11.023
tients (8) and pregnant smokers (9) with more severe discounting of delayed rewards at baseline were more likely to relapse. Questionnaire-measured impulsivity in community-recruited persons has been linked to reduced prefrontal gray matter (GM) volume (10,11). Might reduced frontocortical GM volumes also relate directly to severity of laboratory-measured DD? We explored whether frontocortical gray matter volume relates to severity of DD among healthy subjects. We related the DD constant k, a coefficient that describes severity of discounting, to each of the following: proportional frontocortical gray matter volume, cerebral brain volume (CBV), and total intracranial volume (ICV) derived from segmented magnetic resonance imaging (MRI). Based on seminal fMRI exploration of DD, we hypothesized that reduced ilFC and dlFC volume (as a proportion of whole brain volume) would correlate with severity of DD.
Methods and Materials Procedures were approved by the Institutional Review Board of the National Institute on Alcohol Abuse and Alcoholism, and all subjects provided written informed consent. Subjects All subjects were neurologically and medically healthy per physical and radiological examinations. Drug abstinence was verified in urine and breath samples. History of psychosis, seizures, or an estimated IQ ⬍ 80 was exclusionary for any applicant. Community-recruited control subjects (n ⫽ 29; age 20 –58 [mean 37.4 ⫾ 11.0]; 18 male subjects) did not meet DSM-IV criteria for any Axis I disorder in structured clinical interviews. Delay-Discounting Task Using a desktop computer, subjects were presented with 137 choices between a $10 standard reward (to be presented at delays of 0, 7, 30, 180, or 365 days) or an immediately-presented reward ranging from 25¢ to $10. The subject was told he/she would actually receive one randomly selected choice. At each delay to the standard reward, a switchpoint dollar value of the immediate alternative was calculated, below which the subject BIOL PSYCHIATRY 2009;65:710 –713 © 2009 Society of Biological Psychiatry
BIOL PSYCHIATRY 2009;65:710 –713 711
J.M. Bjork et al. would opt to wait for the full standard reward. The discounting constant k was calculated with a least-square fit with switchpoints at each delay to the standard $10 reward. The k values were then log-transformed (hereafter lnK) to achieve a normal distribution. Higher values of k (raw or log-transformed) indicate preference for immediate rewards (Supplement 1). Questionnaire Impulsivity The Barratt Impulsiveness Scale (BIS) (12) total and subscale scores were also collected as secondary measures of attentional, motor, and nonplanning impulsivity. Brain Volumetric Data Subjects were scanned with 1.5 T MRI (GE Medical Systems, Milwaukee, Wisconsin) using a fast-spoiled gradient-recalled acquisition in the steady state (GRASS) sequence and coronal acquisition. Brain volumes derived from coronal slices were rotated to a standard orientation and total intracranial volume was calculated by defining the inner skull surface with a handdriven cursor. The contents of the intracranial vault were then
A.
segmented into gray matter, white matter (WM), and cerebrospinal fluid (CSF) (13). To encompass frontocortical regions recruited during DD tasks, frontal cortex was calculated a priori as GM and WM anterior to the coronal slice that included the anterior commissure (AC). The mFC was defined as all frontal GM extending laterally from midline to the width of the lateral ventricles. Orbitolateral frontal cortex (olFC) was defined as the (remaining) nonmesial GM below the axial slice that encompassed the AC. The remaining lateral cortex was radially halved into inferolateral frontal cortex and dorsolateral frontal cortex. Frontocortical GM volumes were recalculated as an adjusted proportion of total cerebral brain volume before relation to behavioral measures. This corrected for individual and sex differences in whole brain volume, which relates to body size (14). Complete details of the brain volume calculation are in Supplement 1. Statistical Analysis The relationship between each of the four proportional frontocortical GM volumes and DD was assessed as a partial
B. -2
-2
-3
-4
-5
lnK
Beta = -.373, p < .05
-3
Beta = -.431, p < .05
-4
-5
lnK
-6
-6
-7
-7
-8
-8
-9
-9
-10
-10 2.0
2.5
3.0
3.5
3.0
Age-corrected GM volume
C.
-3
5.0
5.5
Beta = -.268, p = .17
-4 -5
-5
lnK
4.5
-2 -3
Beta = -.204, p = .32
-4
4.0
Age-corrected GM volume
D.
-2
3.5
lnK
-6
-6
-7
-7
-8
-8
-9
-9
-10
-10 3.5
4.0
4.5
5.0
Age-corrected GM volume
1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4
Age-corrected GM volume
Figure 1. Relationship between frontocortical proportional GM volumes and preference for immediate rewards as indexed by partial correlations between (bilateral) frontocortical GM volumes and log-transformed delay-discounting constant k. After controlling for age in multiple regression, log-transformed delay-discounting constant k correlated inversely with each of (A) dorsolateral frontal GM volume (corrected as a proportion of total cerebral brain volume) and (B) inferolateral frontal GM volume, where lower proportional GM volumes were associated with increased preference for immediate presented rewards over larger but delayed rewards. The log-transformed k constant did not partially correlate with either (C) mesofrontal or (D) orbitolateral frontal GM volumes or with total cerebral brain volume or total intracranial volume. GM, gray matter.
www.sobp.org/journal
712 BIOL PSYCHIATRY 2009;65:710 –713 correlation in simultaneous multiple regression. In each analysis, lnK was the dependent variable, with age and the brain volume measure as the independent variables. Age was included as a covariate in these analyses because aging reduces GM volume (15) as well as discounts behavior (16). We then repeated these analyses with BIS scores substituted for lnK.
Results The discounting curve predicted by the hyperbolic equation significantly correlated with actual subject decision switchpoints, with a median R2 value across subjects of .80. Proportional (CBV-adjusted) ilFC (Beta ⫽ ⫺.373, p ⬍ .05) and dlFC (Beta ⫽ ⫺.423, p ⬍ .05) GM volumes showed a negative partial correlation with lnK, where lower ilFC and dlFC GM volumes were associated with greater discounting of delayed rewards (Figure 1). In contrast, lnK did not partially correlate with either mFC (Beta ⫽ ⫺.204, p ⫽ .32) or olFC (Beta ⫽ ⫺.268, p ⫽ .17) GM volume. To determine whether brain-behavior relationships were specific to the frontal lobe measures, in multiple regression, total CBV and ICV were each substituted as the brain volumetric measure, while controlling for both age and sex. Neither CBV (p ⫽ .33) nor ICV (p ⫽ .40) partially correlated with discounting. In a secondary analysis, when BIS total and subscale scores were each substituted for lnK as the dependent impulsivity variable, none correlated with any frontocortical volume. Neither BIS total nor BIS subscale scores correlated with lnK.
Discussion In healthy adults, greater discounting of delayed rewards partially correlated with smaller proportional ilFC and dlFC (while controlling for age). Critically, ilFC and dlFC encompassed the lateral frontocortical voxels recruited during DD decision making in fMRI studies (3,4). In contrast, DD behavior did not correlate with CBV or with premorbid brain size as reflected in ICV. This frontocortical specificity is in accord with findings of Yang et al. (10) and Matsuo et al. (11), who found that prefrontal GM volume, but not whole brain volumes, correlated with questionnaire impulsivity. Unlike the Yang et al. (10) and Matsuo et al. (11) studies, however, frontocortical GM volumes did not correlate with questionnaire-measured impulsivity. We note, however, that BIS scores in our sample did not correlate with actual DD behavior. Divergent findings may have also resulted from differences in GM measurement. Critically, frontocortical GM in the Matsuo et al. (11) study was calculated using voxel-based morphometry, where each brain was warped into common stereotactic space for a probability-based assessment of each voxel. We measured actual GM volume, where brains were rotated only. We note that frontocortical GM volume only accounted for a modest portion of variance in DD, and this can be attributed to several factors. First, whether DD behavior is a state or trait quality is controversial (1). For example, DD behavior can be altered with pharmacological challenges (17) and experience of receiving deferred rewards from experimenters (18) and also differs as a function of local economic conditions (19) and remission from drug dependence (20). Second, while the large frontocortical GM regions defined here likely accommodated individual differences in precise localization of recruitment (as seen in mFC [2]), their size likely compromised specificity. Third, these control subjects generally discounted less (median k ⫽ www.sobp.org/journal
J.M. Bjork et al. .0034) than control subjects in many other experiments. This limited range of discounting likely reduced the ability to detect additional relationships between frontocortical measures and DD. Finally, we caution that our correlations between proportional frontocortical GM and k would not survive correction for multiple comparisons, but are reported here as explicitly hypothesis-driven. These data offer preliminary evidence that lateral frontocortical development may be a morphological contributor to DD, since this region is recruited when subjects opt for delayed rewards over a smaller immediate alternative (3,4). These results, if extended in clinical populations, would support opponentprocess accounts of externalizing disorders (e.g., 1,5), where impulsivity is unchecked by executive control neurocircuitry. Of interest in future research would be testing severely discounting subjects (such as substance-dependent individuals), especially in studies that jointly assess both frontocortical structure and function in conjunction with laboratory impulsivity. Finally, the advent of more functionally defined (and individually customized) cortical mapping can better define GM regions. Finally, if this finding can be replicated in children and adolescents, it would implicate reduced frontolateral cortex development as a premorbid risk factor for psychiatric disorders of impulse control.
This research was sponsored by intramural research funds of the National Institute on Alcohol Abuse and Alcoholism. We thank Michael Kerich and Khai Van for image processing and Dr. Suzanne Mitchell for her donation of the delay discounting task. The authors reported no biomedical financial interests or potential conflicts of interest. Supplementary material cited in this article is available online. 1. Bickel WK, Miller ML, Yi R, Kowal BP, Lindquist DM, Pitcock JA (2007): Behavioral and neuroeconomics of drug addiction: Competing neural systems and temporal discounting processes. Drug Alcohol Depend 90(suppl 1):S85–S91. 2. Kable JW, Glimcher PW (2007): The neural correlates of subjective value during intertemporal choice. Nat Neurosci 10:1625–1633. 3. McClure SM, Ericson KM, Laibson DI, Loewenstein G, Cohen JD (2007): Time discounting for primary rewards. J Neurosci 27:5796 –5804. 4. McClure SM, Laibson DI, Loewenstein G, Cohen JD (2004): Separate neural systems value immediate and delayed monetary rewards. Science 306:503–507. 5. Bechara A (2005): Decision making, impulse control and loss of willpower to resist drugs: A neurocognitive perspective. Nat Neurosci 8:1458 –1463. 6. Reynolds B (2006): A review of delay-discounting research with humans: Relations to drug use and gambling. Behav Pharmacol 17:651– 667. 7. Perry JL, Carroll ME (2008): The role of impulsive behavior in drug abuse. Psychopharmacology (Berl) 200:1–26. 8. Passetti F, Clark L, Mehta MA, Joyce E, King M (2008): Neuropsychological predictors of clinical outcome in opiate addiction. Drug Alcohol Depend 94:82–91. 9. Yoon JH, Higgins ST, Heil SH, Sugarbaker RJ, Thomas CS, Badger GJ (2007): Delay discounting predicts postpartum relapse to cigarette smoking among pregnant women. Exp Clin Psychopharmacol 15:176 –186. 10. Yang Y, Raine A, Lencz T, Bihrle S, LaCasse L, Colletti P (2005): Volume reduction in prefrontal gray matter in unsuccessful criminal psychopaths. Biol Psychiatry 57:1103–1108. 11. Matsuo K, Nicoletti M, Nemoto K, Hatch JP, Peluso MA, Nery FG, Soares JC (2008): A voxel-based morphometry study of frontal gray matter correlates of impulsivity [published online ahead of print May 8]. Hum Brain Mapp. 12. Patton JH, Stanford MS, Barratt ES (1995): Factor structure of the Barratt Impulsiveness Scale. J Clin Psychol 51:768 –774.
J.M. Bjork et al. 13. Momenan R, Hommer D, Rawlings R, Ruttimann U, Kerich M, Rio D (1997): Intensity-adaptive segmentation of single-echo T1-weighted magnetic resonance images. Hum Brain Mapp 5:194 –205. 14. Ho KC, Roessmann U, Straumfjord JV, Monroe G (1980): Analysis of brain weight. II. Adult brain weight in relation to body height, weight, and surface area. Arch Pathol Lab Med 104:640 – 645. 15. Bjork JM, Grant SJ, Hommer DW (2003): Cross-sectional volumetric analysis of brain atrophy in alcohol dependence: Effects of drinking history and comorbid substance use disorder. Am J Psychiatry 160: 2038 –2045. 16. Green L, Fry AF, Myerson J (1994): Discounting of delayed rewards: A lifespan comparison. Psychol Sci 5:33–36.
BIOL PSYCHIATRY 2009;65:710 –713 713 17. de Wit H, Enggasser JL, Richards JB (2002): Acute administration of d-amphetamine decreases impulsivity in healthy volunteers. Neuropsychopharmacology 27:813– 825. 18. Walls RT, Smith TS (1970): Development of preference for delayed reward in disadvantaged children. J Educ Psychol 61:118 –123. 19. Tucker B (2006): A future discounting explanation for the persistence of a mixed foraging-horticulture strategy among the Mikea of Madagascar. In: Kennett DJ, Winterhalder B, editors. Behavioral Ecology and the Transition to Agriculture. Berkeley, CA: University of California Press, 22– 40. 20. Bickel WK, Odum AL, Madden GJ (1999): Impulsivity and cigarette smoking: Delay discounting in current, never, and ex-smokers. Psychopharmacology (Berl) 146:447– 454.
www.sobp.org/journal