Mental Exercise: Ongoing Intervention Trials

George W. Rebok, Ph.D.

Symposium: Cognitive Activity from Bedside-to-Bench: Findings from the NIA R13 Conference Grant

American Geriatrics Society

Chicago, IL

May 6, 2006

Cognitive Training “in the News”

• Brain training takes aging Japan by storm

• By George Nishiyama Mon Apr 10, 10:58 AM ET

• TOKYO (Reuters) - Tamako Kondo says 10 minutes of exercise every morning keeps her fit. But the 80-year-old doesn't hit the treadmill or take aerobics classes. Instead, she sits at a desk, pencil in hand, and tackles simple arithmetic and other quizzes, part of a "brain training" program that has taken Japan by storm.

• Bookshops now have separate sections for workbooks with the exercises and video game versions are selling like hot cakes among the growing ranks of older Japanese who hope the drills will reinvigorate their gray matter.

• "I want to delay becoming senile as much as possible," said Kondo, who lives in a Tokyo home for the elderly.

“Use it or Lose it?”“It’s a fortunate person whose brainIs trained early, again and again,And who continues to use itTo be sure not to lose it,So the brain, in old age, may not wane.”

(Rosenzweig MR, Bennett EL. Behavioral Brain Research 1996;78:57-65)

“Despite the frequent assertions of the mental exercise hypothesis, its intuitive plausibility, and an understandably strong desire to believe that it is true….., there is currently little scientific evidence that differential engagement in mentally stimulating activities alters the rate of mental aging.”

(Salthouse TA. Mental exercise and mental aging: Evaluating the validity of the “Use it or lose it” hypothesis. Perspectives on Psychological Science 2006; 1:68-87.)

Growing Interest in Promoting Public Cognitive Health

• Staying Sharp project (AARP)• Maintain Your Brain (Alzheimer’s Association)

• Keep Your Brain Young (McKhann & Albert, 2002)• The New Brain (Restak, 2004)• Age-Proof Your Mind (Tan, 2005)

Outline of Talk• To present evidence on the effectiveness of ongoing

intervention trials in improving and maintaining cognitive functioning of older adults

• To explore the question of the extent to which skills acquired during cognitive training transfer to similar tasks having a more real-world component

• To discuss challenges and what steps might be taken to develop the next generation of training studies

A Taxonomy of Behavioral and Non-Behavioral Intervention Strategies (adapted from Baltes)

Goal Level Target Type Mode SettingEnrich

Prevent

Remediate

Compensate

Individual

Small group

(e.g., n =3-5)

Large group

(e.g., n > 5)

Neighborhood

Community

Cognitive ability (e.g., memory, attention, executive function)

Cognitive complaints

Efficacy beliefs

Functional skills

Cognitive training-practice

Cognitive rehabilitation

Pharmaco-therapy

Life-style modification (e.g., exercise, health habits, diet, stress reduction)

Cognitive engagement

Biomedical

Single component

Multiple component (e.g., cognitive training + pharmaco-therapy;

Cognitive training + exercise therapy)

Laboratory

Clinic

Hospital

Home

School

Workplace

Community center

Internet

Multiple settings (e.g., Clinic + Internet; clinic + home)

Training on Basic Abilities: Background• Programmatic Research on Basic Abilities: 1970’-

1980’s– Early childhood education programs - plasticity

• Does range of cognitive plasticity vary across life span?

– Adult cognitive longitudinal studies: Variability in rate of cognitive decline

• Early Basic Ability Training in Old Age: 1970-1990– Focus on abilities showing “early” decline in 60’s (abstract

reasoning, perceptual speed, working memory)– Ability-specific (single ability) training - focus on

strategies associated with ability– Significant training effect compared to no-treatment or

social contact control group (retest gain)• Training gain: 0.50-0.75 Sd

Training on Basic Abilities: Background (2)• Some evidence for temporal durability of training effects

(up to 7 yrs for reasoning; 3.5 yrs for memory; 18 months for speed)

• “New Questions” for Training Research:– Long-term clinical outcomes of interventions– “Transfer” to measures of functioning, everyday tasks

• Concerns re Generation 1 Training Research:– Representativeness of samples - regional, convenience samples;

lack of diversity in samples– Clinical Trial Design - Intent to treat design - attrition– Replicability of findings– Clinically meaningful outcomes– ACTIVE

ACTIVE - Generation 2 of Cognitive Training Studies

RFA initiated by NIA and NINR

ACTIVE - Advanced Cognitive Training for Independent and Vital Elderly Randomized Controlled Clinical Trial Common multi-site intervention protocol with “proven interventions” Include intent-to-treat analyses

Primary Aim of ACTIVE To test the efficacy of three cognitive interventions to improve or

maintain the cognitively demanding activities of daily living. Important Shift in Major Outcome of Cognitive Training Research

Primary outcome is cognitively demanding activities, NOT Basic Cognitive Abilities. Outcome of ACTIVE trial specified by RFA

• Thus, the pre-specified ACTIVE design necessarily had to use basic intervention strategies which are known to be challenging for achieving real-world transfer

ACTIVE Steering Committee

• University of Alabama- Birmingham

Karlene Ball, Ph.D.• Hebrew Rehabilitation Center

for Aged, BostonJohn Morris, Ph.D.

• Indiana UniversityFrederick Unverzagt, Ph.D.

• Johns Hopkins UniversityGeorge Rebok, Ph.D.

• Pennsylvania State UniversitySherry Willis, Ph.D.

• University of Florida / Wayne State University

Michael Marsiske, Ph.D.

• New England Research Institutes, Coordinating Center

Sharon Tennstedt, Ph.D.

• National Institute on Aging

Jeffrey Elias, Ph.D.

• National Institute of Nursing Research

Kathy Mann-Koepke, Ph.D.

Participant Characteristics

Conceptual Model

Training CognitiveAbilities

Daily Function

ProximalOutcomes

PrimaryOutcomes

Reasoning Speed

No

Contact Memory

Refused Ineligible Screen for Eligibility

Eligible and Consenting

Baseline Measurements

Randomize to Training

Post - Test (PT)

Booster Booster

Yes No Yes

No Yes No

Booster

-

Figure 1. Study Design

Booster Booster Booster

1-Yr Test (A1)

2-Yr Test (A2)

3-Yr Test (A3)

5-Yr Test (A5)

Net Effect of ACTIVE Training on Proximal Outcome Composites

Memory Training

Reasoning Training

Speed Training

Memory Composite (+), PT 0.2566*** -0.0197 -0.0449 ", A1 0.2085*** 0.0178 -0.0499 ", A2 0.1751*** 0.0431 -0.0324 ", A3 0.2207*** -0.0103 0.0062

Reasoning Composite (+), PT -0.0019 0.4797*** 0.0014 ", A1 -0.0039 0.3998*** -0.0296 ", A2 -0.0228 0.2568*** -0.0402 ", A3 0.0132 0.3812*** -0.0370

Speed Composite (+), PT -0.0089 -0.0262 -1.4541*** ", A1 -0.0201 -0.0032 -1.2000*** ", A2 -0.0503 -0.0192 -0.8616*** ", A3 0.0456 0.0053 -0.9538***

Net effect size defined as [Training Mean - Control Mean at indicated time] - [Training mean - Control mean at baseline] divided by intra-subject standard deviation of the composite. (+) indicates direction of positive response.

*** p < 0.0001 testing for net effect significantly different from zero.

Net Effect of ACTIVE Booster Training on Cognitive Abilities

Memory Booster

Reasoning Booster

Speed Booster

Memory Composite (+), A1 0.0473 -0.0391 -0.0001 ", A2 0.0588 -0.0139 0.0413 ", A3 0.1631*** 0.0027 0.0374

Reasoning Composite (+), A1 -0.0041 0.302*** 0.1256*** ", A2 -0.04 0.151*** -0.0358 ", A3 -0.072 0.3756*** 0.0902*

Speed Composite (-), A1 -0.0352 -0.0412 -0.907*** ", A2 -0.0218 -0.066 -0.3398*** ", A3 0.0569 -0.0193 -0.9946***

Net effect size defined as [Boosted mean - Unboosted training mean at indicated time] - [Boosted mean - Unboosted training mean at baseline] divided by intra-subject standard deviation of the composite. (+) indicates direction of positive response. * p < 0.05 testing for net effect significantly different from zero. *** p < 0.0001 testing for net effect significantly different from zero.

ACTIVE Findings:Effects on Everyday Task & Functioning

• No Transfer from Basic Ability Training to Everyday Functioning for any of the 3 Treatment groups - Report through A3– Decline in Functioning occurs later than decline in

basic abilities– Positive selected control group - delay in onset of

functional decline

• Findings of A5 to be reported - manuscript under review

Normal Memory vs Memory Impaired:Impact on Training on Memory,

Reasoning, and Speed

Interven Time

Memory Reasoning Speed

Memory

PT .300*** -.009 -0.050

A1 .254*** .033 -0.061

A2 .214*** .052 -0.057

Reason

PT .001 .477*** 0.025

A1 .013 .416*** -0.026

A2 -.003 .262*** -0.021

Speed

PT .004 -.017 -1.488***

A1 .004 .009 -1.238***

A2 -.024 -.013 -0.886***

Inteven Time

Memory Reasoning Speed

Memory

PT -.012 -.117 0.105

A1 -.175 -.163 0.107

A2 -.100 -.015 0.400*

Reason

PT -.048 .573*** -0.277

A1 -.230 .208 -0.155

A2 -.331 .276* -0.434*

Speed

PT -.108 -.111 -1.420***

A1 -.163 -.097 -1.100***

A2 -.298 .079 -0.755***

Next-Generation Training Platforms

• Technology-based: video training, computerized training, internet-based

• Experiential/engagement: global, non-ability specific interventions

•Trainer-less Training: collaborative, interactive but little feedback provided

•Combinatorial Training: little work done on combined training (exercise and cognition, pharmacotherapy and cognition, etc.)

Experiential/Engagement

• “Engagement” hypothesis (e.g., Schooler & Mulatu, 2001; Verghese et al., 2003) – Age-related declines in cognitive functioning may to some extent be mitigated by a lifestyle marked by social and intellectual engagement

• Broad-based effects• Evidence is correlational

Experience Corps model• Volunteers 60 and older• Serve in public elementary schools: K-3• Meaningful roles; important needs• High intensity: >15 hours per wk• Reimbursement for expenses: $150/mo• Sustained dose: full school year• Critical mass, teams• Health behaviors: physical, social, and cognitive activity• Leadership and learning opportunities • Infrastructure to support program• Program evaluation• Diversity

– Freedman M, Fried LP; Experience Corps monograph, 1997

What We’ve Learned So Far• Can recruit and retain a large group of elderly

volunteers

• Volunteers accept the need for randomization

• Program perceived as widely attractive to older adults, well-accepted by participants, including principals, teachers, and children

• Results show initial positive benefit in selected areas of function among older adults:– physical: improved chair stand – cognitive: improved executive functioning

Characteristic Participants Controls

Age, mean years 68 (r: 62-78) 68 (r: 63-75)

Female, n (%) 8 (100) 9 (100)

African American, n (%) 8 (100) 9 (100)

Education, mean years 12 12

Widowed, n (%) 5 1

MMSE, mean 24.5 25.6

EC Functional Brain MRI (fMRI)Pilot Study (Drs. Carlson, Kramer, &

Colcombe): Demographics of Intervention (N=8) & Controls

(N=9)

Preliminary Conclusions

• fMRI trial is feasible

• Change in patterns of activation are evident• Behavioral RT and fMRI data correspond in showing

improved ability to selectively attend during the most demanding condition

o Increased activity in attentional control regions suggests more successful filtering/inhibiting of conflicting information

o Corresponding reduction in dACC suggests better filtering of conflicting information

• Consistent with patterns observed in a 6-month physical activity intervention (Colcombe et al., PNAS 2003)

Mean changes in voxel activity from Pre- to Post

Some Challenges• What is the acceptable transfer mechanism? - Not much

consensus on the critical domains, e.g., ACTIVE• What is the time course of expected transfer? - May not

see immediate transfer to everyday outcomes. Are we building a “reserve” of maintained cognitive ability, which may be drawn upon in the future to attenuate the rate of decline?

• Do we need training at all? - Practice may be as effective as formal training; transfer effects may be narrower

• “Bottom-up” or “Top-down” interventions – Train at the level of complex activities or basic abilities

• Single-component or multi-component interventions• Mechanisms for the delivery of interventions – computer,

internet, video, peers/couples• Learning lessons from neurorehabilitation, education, and

physical exercise research about compliance, dosing, cross-training, coaching/monitoring, etc.

To Be Determined:

• What are the best methods for specific training outcomes?

• How can current cognitive theory inform cognitive training, and vice versa?

• How should we define successful training?• Who are the best candidates for successful

training?• Does cognitive training in later adulthood develop

cognitive reserve or serve a protective function?• How do we make training appealing, accessible,

and cost-effective?

Some Caveats• Training gains are of lower magnitude than many elderly,

patients, and caregivers expect and progress may not be steady; problem of raising “false hope” and “blaming the victim” for cognitive declines

• Training effects tend to be highly task-specific and show limited generalizability; effects are reasonably durable but maintenance doesn’t automatically occur.

• Training may not prevent cognitive decline, BUT it can boost performance and may delay normative cognitive decline.

• A few sessions of cognitive training may not be sufficient to alter the life course with respect to decline, BUT it may compress the point of disability into a smaller window at the end of life.