Title: Impulsivity and SelfControl
1Impulsivity and Self-Control
John Monterosso1, George Ainslie2, Edythe
London1 1 UCLA, Dept of Psychiatry2 Coatesville
VA Medical Center
"To cease smoking is the easiest thing I ever
did. I ought to know because I've done it a
thousand times." - Mark Twain
2Plan
- Summarize the intrapersonal bargaining account
of willpower - Present fMRI data from 3 studies
- Methamphetamine addicts and controls performing a
delay discounting task - Signal change in the ventral striatum of normal
participants winning money varying in amount and
delay - Craving smokers voluntarily abstaining from
available access to cigarette smoke - 12-step and recovery from addiction
3How do people say no?
4Local Vs Global framing of the contingencies
- Are you hammering a nail or building a house?
- Are you having fun, or undermining the life you
want?
5How do you move from local to global?
- Picoeconomic/ Bargaining answer -
- You come to see specific behaviors (local) as
predictive of what your own behavior will be
within the larger category of behaviors (global).
6Bargaining account of self-control
Times Now and Future Alternatives Dirty Vs
Clean
Dirty Now and Dirty Future (DD) Dirty Now and
Clean Future (DC) Clean Now and Dirty Future
(CD) Clean Now and Clean Future (CC)
Ranking
Ranking
7Tomorrow, Tomorrow!
8Future behavior is uncertain
Ranking
If my estimate of what my future behavior will be
is (sufficiently) informed by present choice then
9A thought experiment on willpower
- Would the smoker trying to quit resist todays
temptation if - She knew she was destined to relapse for good
tomorrow? - She knew she would not smoke at all starting
tomorrow?
Willpower requires there to seem to be more in
the balance than what is literally at stake.
10Internal feedback between behavior and expectation
Expectation of future choices
Current Preference
11(No Transcript)
12Delay Discounting in MA Abusers and Controls
- Administer a delay discounting task outside the
scanner. - Estimated a discount function for each
individual. - Administer delay discounting task in the scanner
with three block-types No Choice, Hard Choices,
Easy Choices.
13(K.04)
No Choice
27 Today
Hard Choice
27 Today
50 in 21 days
Easy Choice
27 Today
259 in 21 days
27 Today
29 in 21 days
14MA discounted more steeply
Geometric mean k-value for MA abusers .045
(Indifferent between 20 today and 47 in 1
month)
Geometric mean k-value for control subjects .013
(Indifferent between 20 today and 28 in 1
month)
15(No Transcript)
16Task Related Activity (Choice - No Choice)
VLPFC
AC/SMA
DLPFC
IPS
(Human Brain Mapping, 2007)
17(No Transcript)
18(Human Brain Mapping, 2007)
19Conclusions
- Meth addicts exhibit steep delay discounting
- Differences in Meth and control discount rates do
not appear to require robust signal change in
Beta regions - Meth addicts did differ from controls in
frontoparietal activity as a function of decision
difficulty
20Anomalies of human discounting data
- Inconsistent across domains
- Far more variable than other species tested
- Negative discounting for pain and some pleasure
Conceived contingencies are not limited to what
is explicitly at stake Tom
21(No Transcript)
22Non-choice based discounting
- Would it be different?
- Would it be less domain specific?
- Would we see differences in drug users relative
to controls?
23Signal change in nucleus accumbens during
anticipation of reward
Knutsen et at., 2003
24Estimating Delay Discounting from fMRI signal
change (NOT ACTUAL DATA!)
100 In 1 month
Vi
Vd
56 In 1 month
DK1
37 In 1 month
30 In 1 month
25 Today
Signal increase
27 In 1 month
26 In 1 month
25.50 In 1 month
25.25 In 1 month
K.00016
K.0004
K.001
K.0025
K.006
K.016
K.041
K.1
2510
20
30
Today
In 1 week
In 2 weeks
2610
20
30
Today
In 1 week
In 2 weeks
2710
20
30
Today
In 1 week
In 2 weeks
Red
2810
20
30
Today
In 1 week
In 2 weeks
Red
2910
20
30
Today
.29
.07
.04
In 1 week
.14
.07
In 2 weeks
30Pilot Study Design
- 10 healthy participants
- Outside the scanner subjects made choices between
all 36 pairs - During fMRI subjects completed 12 trials at each
of the 9 amount-delay combinations. - So far we just looked at signal change in the
ventral striatum during anticipation. - Treated as ordinal data
31Later larger gt 10 Now
N 10 normal healthy college students
32Later larger gt 20 Now
N 10 normal healthy college students
33Non-choice based discounting
- Would it be different?
- Would it be less domain specific?
- Would we see differences in drug users relative
to controls?
34(No Transcript)
35Smoking Self-ControlChallenge
36Would you like to smoke right now?
37Smoke Unavailable
Valve is closed 20 sec
38Smoke available, but try not to
Valve is open 20 sec
39Successful Self-Control - Unavailable
40(No Transcript)
41When people recover from addiction
- They may talk of a spiritual awakening
- If they got help, gt90 chance 12-step
- They may talk of having hit rock bottom
- They will likely experience quitting as a
struggle to gain control over their life.
?
4212-Step
- Free
- Arguably more effective then alternative
therapies.
43Crits-Christoff, 1999
44Weiss et al, 2003
45Intertemporal bargaining and 12-Steps
- Self-control requires there to be more seen in
the balance then just one. - Anything that makes a choice unique undermines
its status as precedent, making smaller sooner
more compelling. - Underconfidence in future self should produce
choice of smaller sooner. - Overconfidence in future self should produce
choice of smaller sooner. - Bright lines facilitate self-control
You cannot use in moderation
Strict
Emphasis on the process.
Every day one day closer to next relapse.
Permanent
Smoke in Future
Abstain in Future
DC
DD
Abstinence, not moderation
Smoke Now
CC
CD
Abstain Now
46THANKS
- George Ainslie, MD
- Edythe London, PhD
- Traci Mann, PhD
- Steve Engel, PhD
- Andrew Ward, PhD
- Xochitl Cordova
- Sara Cathcart
Funding from NIDA TRDRP