+
Chapter 13
Memory and Aging
+
2
Memory as We Age

An irony – comparing our own current memory with our past
memory requires memory.

People are less able to accurately report memory lapses as
they age.


Complaints about memory in the elderly are more related to
depression than actual memory performance.


Sunderland et al. (1986)
Rabbit and Abson (1990)
Impaired memory is the earliest and best predictor of the onset
of Alzheimer’s disease.
+
3
Methods for Studying Aging
Longitudinal Studies

Cross-Sectional Studies
A representative sample of people tested
repeatedly over time

Advantages:



Different groups of people are sampled across
the age range, with each being tested once

The effect of age can be determined on
an individual basis, helping to pinpoint
precursors of disease
Disadvantages:

Expensive

Time consuming



High dropout rate, often making the
sample less representative
Participants get better at taking the
same test with repeated testings
Test #1

No re-testing

Quicker and less expensive

Lower dropout rate
Disadvantages

Performance can’t be related to
earlier/future data

Cohort Effect:

Practice Effects:

Advantages:
Test #2
Test #3
20-yearolds
People born at different times
differ due to:

Diet

Education

Number of siblings

Social factors (war,
economic depression)
Test #1
40-yearolds
30-yearolds
20-yearolds
0 yrs
+10 yrs
+20 yrs
0 yrs
+
4
Cohort Effects

Flynn Effect:

IQ has been rising across
generations (Raven’s matrices)
From Rönnlund and Nilsson (2008). Copyright © Elsevier.
Reproduced with permission.

People have become taller and
more educated, while family
size has shrunk across
generations
From Rönnlund and Nilsson (2008). Copyright © Elsevier.
Reproduced with permission.
+
5
Methods for Studying Aging
A Comparison

Cross-sectional and longitudinal studies lead to different results

Cohort and practice effects can be quite influential
Memory Decline with Age
Longitudinal
Data
Cross-Sectional
Data
Based on Rönnlund et al. (2005).
+
6
A Hybrid Method

Combine longitudinal and cross-sectional approaches

Add a new cohort at each test point

Comparison down the columns measures learning effects

e.g. 30-year-old cohort Test #1 to 20-year-old cohort Test #2


So both are then 30 years old at that point
Comparisons across a test point for each group measure cohort effects

e.g. 20-year-old cohort Test #1 to 30-year-old cohort’s Test #1
40-yearolds
30-yearolds
Test #2
Test #3
20-yearolds
0 yrs
+10 yrs
+20 yrs
+
7
Working Memory (WM) and Aging
Type of Memory Example Test
General Findings
Verbal span
Digit span
Declines by < 1 item
Visual span
Corsi block tapping
Declines by < ½ an item
Verbal working
memory
Recalling words in
alphabetical order
Modest decline
Sentence span
Small decline

WM span progressively declines with age



But it is a very small decline
Effects are larger when tasks involve speed of processing or episodic,
long-term memory
May et al. (1999)

The WM decline may be due to a build up of proactive interference that
older adults are less able to inhibit
+
8
Working Memory and Aging
Inhibiting Interference

Hasher and Zacks’ (1988) Inhibition Deficit Hypothesis of Aging:


A major cognitive effect of aging is the reduced capacity to inhibit irrelevant stimuli
Molander and Bäckman (1989):

Participants:


Task:


Make golf shots
Results:



Older and younger mini-golf players, matched in skill
Concentration (measured by heart-rate deceleration) increased in the younger
under competition conditions and performance was maintained, in contrast to a
decline in performance in the elderly. However, there were large individual
differences.
In another study, Bäckman and Molander (1986) showed that competition
decreased the accurate recall of specific shots, and increased irrelevant recall in
the older golfers, but did not influence recall in the young.
Conclusion:

Older adults are less able to shut out potential distracters.
+
9
Working Memory and Aging
Concentration and Attention

Charness (1985):

Younger chess players scan
more possible moves.

Older chess players scan
fewer moves but in greater
depth.

May reflect increased
difficulty keeping track of
multiple sources of
information.

Divided Attention in Aging:

Dual-task performance is
worse in advanced age than
on the two separate tasks.

This probably reflects
general difficulty handling
heavy cognitive loads,
however.

When tasks are made
easier, dual-task
performance is not
affected by age.
+
10
Aging and Long-Term Memory
Episodic Memory


Episodic memory declines steadily through the adult years, across
the board:

Recall and recollection tests

Verbal and visual materials

Rivermead Behavioural Memory Test (everyday memory situations)

Doors and People Test (people’s names, locations)

Memory for card hands

Memorizing passages

Memory for conversations
The magnitude of the decline depends on the nature of the task and
the method of testing (recall vs recognition).
+
11
Declines in Episodic Memory
Modulating Factors (Craik, 2005)

The overall decline in episodic memory is modulated by:


Processing capacity of the learner

Elderly take longer to perceive and process materials

Elderly are less likely to develop and use complex learning
strategies
Level of environmental support provided during retrieval

Age effects are largest in tests lacking external cues (e.g. free
recall)
+
12
Declines in Episodic Memory
Limited Attention or Capacity?

Naveh-Benjamin (2000)





Task:
 A recognition test for word pairs that were either semantically related or not
Results:
 The older group had difficulty for unrelated items, but not for related
Initial Conclusion:
 Elderly are less able to form new associative links
Follow-up (Naveh-Benjamin et al., 2003):
 Gave younger group a concurrent task, which resulted in impairment for
both related and unrelated items—this didn’t match the elderly group’s
results – so it isn’t that the elderly are just slower processing information.
Final Conclusion:
 Associative Deficit Hypothesis: The differences between young and old
are attributable to basic learning capacity, rather than to attentional or
strategic differences related to processing speed.
+
13
Declines in Episodic Memory
Associative Deficit Hypothesis


An age-related difficulty in binding together
unrelated things
 Simply recognizing old faces or names is
unaffected by age
 However, a concurrent task does reduce
performance
 Recalling which name went with a face, is
diminished in the elderly, as this requires
binding
 This decline is even more pronounced
than in the divided-attention condition
Self-Performed Task Effect:
 Age effects are minimized by asking
elderly to perform an action associated
with a to-be-remembered item
 This deepens encoding, providing
auditory, visual, manual, and self-related
codes for the memory
From Naveh-Benjamin et al. (2004b). Copyright © American
Psychological Association. Reprinted with permission.
“Break the match-stick.
Shake the pen”
+
14
Declines in Episodic Memory
Level of Environmental Support at Retrieval

Age effects are clearest in recall tests, which lack external
cues, while recognition tends to be relatively preserved in the
elderly.


This difference may reflect a combination of:

Fewer retrieval cues in the recall task

A greater involvement of association in free recall
Whether recognition is impaired or not depends on the nature of
the task:

If familiarity (“knowing”) is sufficient—no deficit

If recollection (“remembering”) is necessary—some impairment
+
15
Declines in Episodic Memory
Recognition: Remembering and Knowing
Remembering
Knowing

Recollection based

Familiarity based

Involves remembering the
event in its context

Able to recognize an item as
familiar, without being able to
recall the context

Declines substantially with age

Relatively spared in the elderly

Does not represent a
difference in confidence
between young and old

Consistent with the
associative deficit hypothesis
of aging
+
16
Declines in Episodic Memory
Prospective Memory in the Laboratory

Prospective Memory:
 Remembering to carry out an
intended action in the future
without explicit reminders
 Test:
 Participants perform an
ongoing task and respond
either
 After a specified time
 After a cue occurs
 Results:
 An age-related decrement
for both time-based and
event-based tasks

Prospective memory requires:

Encoding the action to be performed

Encoding the time when it should be
performed

Maintaining the information over a
delay

More difficult in real-life situations
with divided attention

Through rehearsal and/or periodic
retrieval from LTM


An external retrieval cue helps
Actually performing the task when
appropriate
+
17
Declines in Episodic Memory
Prospective Memory in Real Life

Unlike laboratory situations, in real-life prospective memory scenarios the
elderly often perform better than younger adults.


Example Tasks:
 Ask participants to make a telephone call or send a postcard at a specified
time.
Rationale:
 Older people are more aware of their memory limitations and compensate
with various strategies, such as:
 Diaries
 Reminders
 Older people live more ordered and structured lives, making it easier to form
plans.
 Older people may have been more motivated to perform well on a memory
task; younger people can explain memory slips by “being too busy”.
+
18
Semantic Memory and Aging

Semantic memory does not
decline with age
 It actually expands with age in
some areas:
 Vocabulary
 Historical facts
 Speed of access (a more
sensitive measure) does
decline, however.

Naturalistic, longitudinal diary
study (Kemper, 1990) revealed:
 Older adults tend to use
sentences that demand less
working memory.
 But they also write “better” and
“more interesting” sentences.

Baddeley et al.’s (1992) Speed and
Capacity of Language Processing
(SCOLP) Test:
 Spot-the-word vocabulary test:
 Participants pick the real word
from the pseudo-word
 Highly resistant to age and
disease
 Semantic processing test:
 Participants verify sentences as
quickly as possible
 High accuracy rates
 Reaction times are very
sensitive to age and other
factors
+
19
Implicit Learning and Memory

Results are mixed, due to the
wide range of implicit processes

Moderately impaired with
advanced age:


Priming tasks involving
response production (e.g.
stem completion)
Small/no impairment in the
elderly:

Identification tasks (e.g.
lexical decision/word
fragment)

Stronger implicit effects in the elderly
 False Fame Effect:
 Participants first see unfamiliar
names
 Then are asked to mark names
that are famous
 Previously processed, unfamiliar
names are judged as more famous
 Especially true for elderly
participants
 Due to impaired recollection,
forcing them to rely on
familiarity
 The elderly may be more susceptible
to false information and leading
questions
+
20
Implicit Learning and Memory
Motor Skills

Motor performance declines with age

Speed of perception and movement decline


Leads to slower learning rate on time-based tasks
The rate of motor learning need not decline with age, however

Young and old adults show comparable rates when learning:

A sequence of motor movements

New stimulus–response mappings

To make serial responses to a number of stimuli under self-paced
conditions

To navigate a computer maze
+
21
Implicit Learning and Memory

Generally:

When the response is
obvious and performance is
measured in terms of speed
improvements


Wilson, Cockburn, and
Baddeley (1989)

Task:

The elderly perform well
When the response is nonobvious, novel associations
must be learned


Older adults are impaired

This is often the case for
learning about new
technologies

Patients learn a simple
process of entering the
time/date into a handheld
computer
Results:

Rate of learning was
extremely sensitive to
episodic memory deficits
+
22
Resisting Age Effects

Factors tending to correlate with resistance to memory
impairment:




Good physical health
 Appropriate diet
 Regular exercise
Continued mental activity
 But rates of decline don’t differ across professors/blue-collar workers
(Christensen et al., 1997)
 Meaningful material may allow the active learner to compensate for
declining episodic memory (Shimamura et al., 1995)
 Explicit memory training (e.g. mnemonics) can help
 But young participants gain substantially more from training than
the elderly
An enriched environment (at least it works for rats …)
Better to be safe than sorry:

Follow the tips, even though evidence is still lacking in some respects
+
23
Resisting Age Effects
Memory Training Programs—Worth It?

Ball et al. (2002)
 Participants:
 2,832 older adults, divided
into four groups
 6-week Training Program:
 Group 1: Strategy training;
practice on words and
shopping lists
 Group 2: Practice on verbal
reasoning tasks
 Group 3: Speed training on
visual search and divided
attention tasks
 Group 4: Controls (no
training)
 Final testing all the task types
and on everyday tasks

Results:
 Groups 1 to 3 improved on
the skills trained
 Despite being tested on
novel materials
 No change occurred for
untrained skills, however
 Improvements did not
generalize to everyday tasks
either

Conclusion:
 Only specific skills can be
trained; no generalization
 It’s possible that training had
some protective effects
+
24
Theories of Aging
Speed Theory -- Salthouse (1996)

Many of the cognitive effects of
aging are caused by reduced
processing speed

Problems with the theory:

Based on extensive
correlational data



Digit Symbol
Substitution Test (DSST)
is a good predictor of age
deficits
The rest of the decline could
be caused by a more general
decline in cognitive
functioning (Salthouse &
Becker, 1998)
Measures that correlate with age
deficits aren’t pure speed tests


DSST also involves strategy and
working memory
Many other physical and cognitive
capacities that decline with age
could have a causal effect

Speed measures don’t always
explain the most variance

Grip strength is an even better
predictor!

Baltes and Lindenberger
(1997)
+
25
Theories of Aging
Reduced WM Capacity – Inhibition Theory

Reduced processing resources
 Dividing young participants’
attention can mimic
performance of older adults
 However, this isn’t always
true
 e.g. Naveh-Benjamin et
al. (2003; 2004)
 Episodic deficits are
more like amnesia than
an attentional limitation
 Conclusion:
 Probably one of many
factors

Reduced ability to inhibit
irrelevant information (Hasher,
Zacks, & May, 1999)

Problems:

Why would this influence
free recall?

Why is performance on the
Peterson task not
influenced by age?

After all, forgetting on
this task is assumed to
be caused by built up
proactive interference
+
26
Theories of Aging
Frontal Lobe Deficits

The frontal lobes atrophy with advanced age

Tasks thought to be supported by the frontal lobe also tend to
decline with age


These tasks typically rely on the executive component of working
memory and/or inhibition
However, the correlation between frontal lobe atrophy and agerelated cognitive decline is weak and the theory is not well-specified
currently
+
27
The Shrinking Brain
Region
As we age, this region …
Overall brain
Shrinks
Ventricles
Expands
Frontal lobes
Shrinks most rapidly
Temporal lobes
Shrinks slowly
Shrinks slowly, then accelerates
(possibly due to disease)
Hippocampus
Loses 20–30% of its neurons by
age 80
Occipital lobes

Shrinks slowly
The latency of Evoked Response Potentials (ERP) increases with age

The P300 increases at an average of 2 ms/year

The rate of slowing becomes more dramatic in dementia
+
28
The Aging Brain
Neuroimaging Studies

Broader activations in the elderly:
 WM and visual attention yield
bilateral activation in the
elderly (Cabeza et al., 2004)
 It is strongly lateralized in
young participants
 Autobiographical memory
activation in the hippocampus
is bilateral in the elderly
(Maguire & Frith, 2003)
 It is left lateralized in young
participants
 Use of other brain structures is
thought to compensate for
overload of one brain
component

Reduced activations in the
elderly:
 Elderly don’t show occipitotemporal activation that
younger participants do in a
complex visual memory task
(Iidaka et al., 2001)
 Suggests the elderly aren’t
using visual imagery
 Elderly benefited less from
a visual mnemonic
strategy
 With complex tasks, the elderly
are no longer able to
compensate and so rely on
simpler strategies
+
29
The Aging Brain
Neurotransmitters

Dopamine



Related to numerous cognitive functions
 Agonists (e.g. bromocriptine) improve spatial working memory
 Antagonists (e.g. haloperidol) diminishes spatial working memory
 Levels correlate with episodic memory performance (Bäckman et al., 2000)
 R2 = 38% on word recognition
 R2 = 48% on face recognition
Decreases 5–10% per decade of life
 According to both post-mortem and PET studies
 Depletion is associated with cognitive deficits in Parkinson’s and
Huntington’s diseases
Covarying out dopamine level nearly eliminates the effect of age on memory
performance (Erixon-Lindroth et al., 2005)
+
Normal Changes with Aging

Normal changes with age:

Slower thinking.

Difficulty paying attention.

Need more cues - like words,
pictures, smell, etc. - to recall
information.


Common causes:

Health related:





Using fewer memorization
skills like visualization and
organization.

Associations are more
difficult.

Decline in vision and hearing.



High blood pressure
Prescription drugs
Bad nutrition
Low Blood Sugar or diabetes
Depression
Anxiety
Taking multiple prescriptions
Lifestyle – lack of sleep, lack of
activity, stress
http://www.alz.org/alzheimers_disease
_10_signs_of_alzheimers.asp
+
31
Alzheimer’s Disease (AD)


Prevalence:
 Accounts for over 50% of senile
dementia cases
 Occurs to 10% of the population
over age 65
 The rate of occurrence
increases with age
Diagnosis:
 Requires memory impairment
plus two other deficits
 Language, action control,
perception, or executive
function
 Difficult to diagnose early
 Ultimate diagnosis typically
requires post-mortem
examination finding dense
 Amyloid plaques
 Neurofibrillary tangles


Warning signs (Petersen et al., 2001):

Memory loss affecting job skills

Difficulty performing familiar tasks

Language problems

Disorientation in time and place

Poor/decreased judgment

Problems with abstract thinking

Misplacing things

Changes in mood/behavior

Changes in personality

Loss of initiative
Primary feature:

Defective episodic memory
+
32
Alzheimer’s Disease (AD)
Disease Progression
Other Brain
Regions
Temporal &
Parietal Lobes
Medial
Temporal Lobes
& Hippocampus
• Initial memory
problems
+
33
Alzheimer’s Disease (AD)
A Case Study: Novelist and Philosopher Iris Murdoch

Garrard et al. (2005) studied the progression of AD in
Murdoch’s writings.

Linguistic problems increased as the disease progressed

Difficultly coming up with specific words

Spelling deteriorated

Sentences became shorter and used fewer
low-frequency words


This was probably a compensatory strategy
Semantic difficulties

Major problems in defining words appropriately

Naming pictures became difficult

Difficulty generating items from a semantic category
+
34
Alzheimer’s Disease (AD)
Episodic Impairments

Deficits in/for:
 Recall
 Recognition
 Verbal materials
 Visual materials
 Everyday memory

Recency is relatively preserved
(like amnesic syndrome)
 At late stages, recency also
tends to decline

Performance on earlier items is
grossly impaired
Data from Greene et al. (1996).
+
35
Alzheimer’s Disease (AD)
Forgetting

Although AD patients have great difficulty acquiring new
information, forgetting rates match those of normal elderly

Kopelman (1985):


Task:

Matched picture recognition performance to normals’ (at a 5minute delay) by increasing exposure time

Re-tested picture recognition after 24 hours
Results:

Equivalent performance for AD patients and control elderly
individuals
+
36
Alzheimer’s Disease (AD)
Semantic Difficulties

Hodges, Patterson, and Tyler (1994) found that the degree of
temporal lobe atrophy predicts deficits in semantic memory, as
measured by a battery testing:





Picture (objects or animals) naming
Picture–name matching
Describing characteristics of named/pictured objects
Verifying sentences about objects
The semantic deficits in AD are not as severe as they are for
semantic dementia, in which:


Episodic memory is relatively spared
Atrophy primarily occurs in the left temporal lobe
 Damage in AD tends to be more medial
+
37
Alzheimer’s Disease (AD)
Implicit Memory
Study
Task
Finding
Heindel et al. (1989)
Motor skill:
Pursuit rotor
•AD patients were worse to
begin with
•Yet, they improved at the
usual rate
Moscovitch (1982)
Skill learning:
Mirror reading
Negligible impairment in
improvement rate for AD
patients
Fleischman et al. (1997)
Priming:
Lexical decision
Normal priming
Fleischman et al. (1997)
Priming:
Stem completion
Impaired priming, unlike in
classic amnesia syndrome
Beauregard et al. (2001)
Priming:
Stem completion
• Normal priming with shallow
processing
• Reduced priming with deep
processing
Salmon et al. (1988)
Semantic priming:
Cued associate
Reduced priming
+
38
Alzheimer’s Disease (AD)
Working Memory


Working memory is relatively less impaired than episodic memory;
however, there are modest deficits in:

Digit span

Corsi block tapping (visuo-spatial memory)
AD patients can maintain small amounts of material over unfilled
delays, so long as they can use the phonological loop to rehearse

Articulatory suppression causes rapid forgetting


Normal elderly individuals are only affected by intellectually demanding
filler tasks (e.g. counting backward by threes)
The capacity for sustained attention is not significantly compromised
+
39
Alzheimer’s Disease (AD)
Working Memory

Baddeley et al. (1986)

Task:

Matched digit load to
normal performance

Also matched secondary,
nonverbal rotary pursuit
task

Had to perform digit span
task either:

Alone

Simultaneously with the
tracking task

Results:
 Young and normal elderly:
 A slight performance
decrease in dual-task
condition
 AD patients:
 Showed a drastic decrease
in the dual-task condition
 Deficits increase as
disease progresses
(Baddeley et al., 2001)
 All three groups behave
similarly:
 In the single-task condition,
regardless of difficulty level
+
40
Alzheimer’s Disease (AD)
Treatment
Pharmacological

Donepezil, Rivastigmine, and
Galantamine

Inhibit cholinesterase

Prevents the breakdown of
acetylcholine

A neurotransmitter that normally
is depleted in AD
Behavioral

In the disease’s early stages:
 Rely on preserved procedural and
implicit memory to teach skills
 Errorless learning
 Fading cues
 Remember: “JOHNNY”
 Fill in “JOH_ _ Y”
 Fill in “J_H_ _ Y”
 Fill in “J_ _ _ _Y”
 Use simple memory aids like
 Message boards
 Calendars
 Modify the environment to make it
clear where things should go
+
41
Alzheimer’s Disease (AD)
Training Programs

Spector et al. (2000)



Reality Orientation Training (ROT)


Improved performance on specific areas trained
 Does not generalize to everyday memory performance
Reduced depression
Helps the patient orient in time and place
Reminiscence Therapy




Helps patients maintain a sense of personal identity
Incorporates photographs and other reminders of their past life
 Can have them construct a personal life-story book
Can be done in groups
Facilitates patient–therapist interactions