MRI Volumetric Results

1
Cognitive and Brain Aging in the Baltimore
Longitudinal Study of Aging
Susan M. Resnick, Ph.D. Laboratory of Personality
and Cognition National Institute on Aging
2

Cognitive and Brain Aging in Older Adults

• What is the background upon which drug abuse is
  superimposed?
• Which aspects of cognition show age-related
  decline in individuals without dementia?
• How does the brain change with age?
• Structural changes
• Functional changes
• fMR probes of specific regions

3
Baltimore Longitudinal Study of Aging BLSA

• Study initiated in 1958
• Women studied since 1978
• Highly educated community-dwelling sample
• GRC visits every 2 years for 2 1/2 days
• Behavioral and physical assessments
• Prospective diagnoses of dementia
• Information on alcohol and smoking but no
  systematic information on other substance abuse

4
Age Effects Vary Across Specific Cognitive
Functions

• Some abilities are preserved throughout the
  lifespan, e.g. over-learned skills such as
  Vocabulary

Vocabulary
5
Age Effects Vary Across Specific Cognitive
Functions

• Other specific functions show declines
• Different abilities may begin declining at
  different ages
• Different abilities may decline at different rates

Benton Visual Retention Test BVRT
6
Age Differences and Longitudinal Changes on the
BVRT
Cross-sectional
Longitudinal
7
California Verbal Learning Test
8
Longitudinal Change in Delayed Verbal Memory
N 266 N 345
Neurology 2003
9
CVLT Long Delay Free Recall Rates of Change are
Variable Across Individuals

Age gt 60 Mean 72.3
10
LPC Neuroimaging Study Early Markers of
Alzheimers Disease and Cognitive Decline
Prior Cognitive Testing
Annual Evaluation
Men and Women (Age 55-85)
Without Prior Neurologic or Severe
Cardiovascular Disease


MRI
PET-CBF
Neuropsychological
Brain Structure
Rest
Testing
Ischemic Change
Verbal Memory
Figural Memory
11
LPC Neuroimaging Assessments 2/10/94 through
6/03/04
ASSESSMENT MEN WOMEN TOTAL 1 94 69
163 2 88 63 151 3 82
59 141 4 77 59
136 5 73 57 130 6 71
55 126 7 68 53 121 8
61 45 106 9 48 38
86 10 16 13 29
TOTAL 678 511 1189

12

Goals

• To determine the rates of structural and
  functional brain changes as a prerequisite for
  identification of disease.
• To determine whether some regions are more
  vulnerable to tissue loss and functional changes.
• To identify brain changes that predict cognitive
  impairment and dementia.
• To identify factors that modify brain-behavior
  associations in aging.

13
Variability in BLSA Brain Morphology (N 18)

















14
MR Image Processing Using RAVENS
Original
Automated Skull-Stripping
Manual Editing
1
2
3
Segmentation
158 Average
Model
4
15
Cross-sectional Both Gray and White Matter
Volumes Are Negatively Correlated with Age
White Volume
Gray Volume
700




700
r -.29
r -.23
600
600
cm3
cm3
500
500
400
400
300
300
50
60
70
80
90
50
60
70
80
90
AGE
AGE
WOMEN
MEN
N 116 Mean Age 70.4 (7.5)
Resnick et al. Cerebral Cortex 200010464
16


Longitudinal Brain and CSF Volumes are Measured
with High Reliability over Four Years
Brain (GW)
Ventricles
(n 92, Mean Age 70.4)
17
Annual Changes in Brain Volumes over 4 Years
Resnick et al. J Neuroscience 2003
18
Longitudinal Brain Changes are Evident in
Younger and Older Individuals



Annual Rate of Change (cm3)
p lt .001
19
4-year Gray Matter Loss in Specific Regions
20

Longitudinal Decreases and Increases in GM
21

Qualitative Changes in Tissue Composition
Measured by Signal Intensity
22
Age Effects on Tissue Composition Decreased
Gray-White Signal Contrast


Cross-sectional
Longitudinal
Tissue Contrast
p lt .001
23

Age Differences and Age Changes in Regional
Cerebral Blood Flow (rCBF)
24
PET Sample Characteristics

• Men Women
• N 46 37
• Age (yrs) 70.9 7.3 70.6 7.9
• ApoE e4 (No. -/) 34/12 24/13
• Mild memory loss
• (No. -/) 41/5 31/6
• Clinical Dementia Rating (CDR) gt 0.5

25
PET Results Cross-sectional Effects ofAge on
Resting rCBF
Older compared with younger individuals show
selective decreases in rCBF in insular (INS),
cingulate (CG), and inferior temporal (IT)
regions.

26
Longitudinal Age Changes in Resting rCBF
Longitudinal declines in rCBF over 4 years are
observed in bilateral superior temporal, right
middle temporal, inferior parietal and midline
occipital regions.

27
Age Influences the RATE of rCBF Decline in the
Mesial Temporal Lobe
Older individuals show faster rates of decline in
mesial temporal rCBF.

28
Functional Brain Changes with PET

• Regional decreases in resting rCBF are observed
  in older individuals, with the greatest
  differences apparent in insular, cingulate and
  temporal regions, including hippocampus.
• These changes reflect a combination of structural
  and functional brain changes.
• With increasing longitudinal interval, we are
  investigating associations between specific brain
  and cognitive changes.

29

fMR Probes for Regions Vulnerable to Aging
30
fMR OFC Sample

• Younger Adults Older Adults
• n 20 (10M/10F) 20 (10M/10F)
• Age (yrs) 28.7 (6.4) 69.3 (5.2)
• (range) 20-40 60-80
• Education (yrs) 15.2 (2.4) 15.0 (3.6)
• MMSE 29.4 (1.0) 28.9 (1.6)
• CESD 7.5 (4.1) 4.7 (5.2)

31
Younger Adults Activate Predicted OFC
RegionsLamar, Yousem, Resnick NeuroImage 2004
Match - NonMatch
NonMatch - Match
Medial OFC (plt.01)
Lateral OFC (p.01)
32
Older Adults Activate Posterior RegionsLamar,
Yousem, Resnick NeuroImage 2004
Match - NonMatch
NonMatch - Match
Association Cortices (ns)
DLPFC (p.006)
33

Conclusions

• Cognitive and brain changes associated with drug
  abuse in the elderly will be superimposed upon a
  changing brain.
• Some but not all cognitive functions show age
  changes.
• Many but not all individuals show age changes in
  cognition and brain structure and function.

34

BLSA Cognition and Neuroimaging Study
Possibilities for Studies of Drug Abuse

• Assessments of older adults continue and
  neuroimaging studies will be expanded with the
  NIA IRP MRI facility.
• Potential to include additional assessments.
• Abuse of prescription medications, including
  pain-killers will be most informative in this
  sample.

35
Collaborators Neuroimaging Project
NIA JHU Alberto Goldszal, PhD Christos
Davatzikos, PhD Dzung Pham, PhD Michael Kraut,
MD, PhD Melissa Lamar, PhD R. Nick Bryan, MD,
PhD Scott Moffat, PhD Jerry Prince,
PhD Stephanie Golski, PhD JHU PET
Facility Robert Dannals, PhD Hayden
Ravert, PhD
36

Neuroimaging Participant Selection

• Inclusion
• Age 55-85
• Prior cognitive and memory assessment
• Exclusion
• Existing neurologic disease, including dementia
• (mild cognitive decline is not exclusionary)
• Severe cardiovascular disease
• (hypertension alone is not exclusionary)
• Metastatic cancer
• Weight greater than 300 lbs or other factors
  precluding neuroimaging assessment

37
Risk Factor APOE ?4 genotype is associated with
accelerated hippocampal volume loss
N 13
N 13
1.0
?4-
?4
0.5
0.0
-0.5
-1.0
Annual Rate of Change ()
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
Neurology 200055134-136.
38
Increases in Temporal Horn Volumes Predict Mild
Cognitive Impairment by CDR Score
Temporal Horn
Brain
N 82
N 12
N 82
N 12

39
Annual percentage changes from baseline,
respectively for brain, gray, white, and
ventricular volumes entire sample -0.55 (0.5),
-0.42 (0.9), -0.67 (1.5), 4.82 (2.5) subgroup
with some medical problems -0.62 (0.5), -0.50
(0.9), -0.73 (1.6), 4.97 (2.7) very healthy
-0.36 (0.4), -0.19 (0.8), -0.48 (1.4), 4.39
(1.5).
40

Automated analysis of specific regions with
HAMMER and brain atlas

 

Template brain (regional outlines provided by
Noor Kabani)
BLSA brain
41

Gray
White
42
Age Differences and Age Changes in Spatial
Rotation Card Rotations Test
43

Age Effects on Regional White Matter Signal
Intensities
Age differences
4-year decline
Davatzikos and Resnick. Cerebral Cortex
200212767-771
44

Analysis of 4-Year Change in MRI Volumes Sample
Characteristics
MEN WOMEN
TOTALN 50 42 92 Age (yrs) 70.5
6.4 70.4 7.7 70.4 7.0 Education (yrs) 16.0
3.2 16.2 2.4 16.1 2.8 Handedness (RL)
473 402 875 Race (White Nonwhite)
482 366 848
45
PET Analysis

• IMAGE PROCESSING AND STATISTICAL ANALYSIS
• Preprocessing using SPM99 and the STAR algorithm
  for elastic stereotaxic normalization
• Voxel-based statistical analysis using customized
  SPM99 software
• Cross-sectional analysis mean CBF across Years
  1, 3, and 5
• Longitudinal analysis rates of change over time
• Significance threshold p lt .01 and cluster size
  gt 35

46

Summary Age-Related Structural Changes

• Regional brain structure can be measured reliably
  over time.
• Both gray and white matter volumes show
  longitudinal declines even in the healthy
  elderly.
• Increases in ventricular volumes are greater in
  older than younger individuals.
• There are regional patterns to both tissue loss
  and qualitative changes in tissue composition.