Preclinical Alzheimer disease: neuropsychological test performance

up to 8 years prior to onset

J.Saxton, Ph.D., O.Lopez, M.D., G.Ratcliff, D.Phil., C.Dulberg Ph.D., M.P.H., L.P.Fried, M.D., M.Carlson Ph.D., A.Newman, M.D., L.Kuller, M.D.

Appendix (E)A-1. Exploratory Factor Analyses

Exploratory factor analysis conducted on the whole group resulted in a clear and easily interpretable three factor structure shown below. The highest loadings for each test are highlighted. The factors were identified as follows:

Factor 1:Memory/Verbal Retrieval (Visual, Verbal, General and Delayed Memory; Boston Naming Test; WMS Information/Orientation)

Factor 2:Speed/Attention/Visual (Trails A & B; Digit Symbol Substitution Test; Attention/Concentration; Block Design)

Factor 3:Verbal Productivity (P & S fluency; Fruit fluency; Vocabulary)

The factor structure makes intuitive sense. The only somewhat anomalous finding is that Block Design which is typically considered to be a constructional or visuospatial task loads on the Speed/Attention factor. This may be related to the fact that Block Design is a timed task and that the other tests loading on Factor 2 do predominantly involve both speed and the visual modality. Certainly Block Design has more in common with this factor than with those loading on factors 1 or 3. Interestingly, in another study of an elderly population in western Pennsylvania (Ratcliff et al., 2003) Trail Making, a factor 2 test, was found to load increasingly with increasing age on a factor identified as visuospatial/constructional ability.

When the analysis was restricted to the subjects who remained non-demented, the factor structure remained similar with the exception that the primary loading for WMS Information/Orientation shifted to a new fourth factor and Vocabulary shifted from factor 3 to factor 1 (see below). WMS Information/Orientation is one of the tests most influenced by education, and performance on this test approached ceiling. The fourth factor could tentatively be identified as crystallized knowledge/information. The fact that WMS Information/Orientation loaded primarily on a Memory/Verbal Retrieval factor in a group that included individuals with impending dementia may indicate that the memory component in WMS Information/Orientation, trivial or automatic for normal subjects, becomes significant for demented subjects. The shift in loading of the Vocabulary test was probably insignificant as the loadings for both factors 1 and 3 are very similar and Vocabulary as assessed in the WAIS-R could have easily loaded on either a verbal retrieval or verbal productivity factor.

References

Ratcliff G, Dodge H, Birzescu M, Ganguli M. Tracking Cognitive Functioning Over Time: Ten-Year Longitundincal Data From a Community-Based Study. Applied Neuropsychology 2003;10:2;76-88.

Factor Analysis using all 693 (621 Normal Controls and 72 preclinical AD)

Rotated Component Matrix

Component
1 / 2 / 3
General memory / .907 / .225 / .235
Delayed recall / .849 / .288 / .181
Verbal memory / .830 / .251
Visual memory / .706 / .359 / .125
Boston Naming Test / .527 / .318 / .244
Info/orient / .332 / .158
Trails A time / -.136 / -.767
Trails B time / -.305 / -.724 / -.243
Digit Symbol Substitution / .256 / .702 / .299
Attention/Concentration / .194 / .657 / .262
Block Design / .410 / .619
Fluency: S / .160 / .859
Fluency: P / .165 / .859
Fluency: fruit / .285 / .121 / .521
Vocabulary / .481 / .239 / .510

Extraction Method: Principal Component Analysis.

Rotation Method: Varimax with Kaiser Normalization.

Factor Analysis using 621 Normal Control subjects only

Rotated Component Matrix

Component
1 / 2 / 3 / 4
General memory / .927 / .235 / .179
Delayed recall / .847 / .307 / .127
Verbal memory / .844 / .187
Visual memory / .705 / .368
Vocabulary / .538 / .216 / .450 / .212
Boston Naming Test / .448 / .268 / .226 / .421
Trails A time / -.788
Trails B time / -.260 / -.717 / -.232 / -.108
Digit Symbol Substitution / .248 / .693 / .312
Attention/Concentration / .175 / .639 / .255 / .110
Block Design / .386 / .617 / .134
Fluency: S / .143 / .854
Fluency: P / .125 / .154 / .847
Fluency: fruit / .242 / .117 / .544
Info/orient / .933

Extraction Method: Principal Component Analysis.

Rotation Method: Varimax with Kaiser Normalization.

Table (E)T-1. Neuropsychological test scores and results of non-parametric, adjustedinferential tests for all four groups

BASELINE
Neuropsychological
test / GROUPS / Kruskal-
Wallis ANOVA
p-value‡
Years to AD onset / Normal
1.5-3.4 years
(n=24) / 3.5-5.0 years
(n=20) / 5.1-8.1 years
(n=28) / (n=621)
median / mean / SD / median / mean / SD / median / mean / SD / median / mean / SD
Orientation:
WMS-R Orientation / 13.0 / 13.2 / (.83) / 13.0 / 13.1 / (.83) / 14.0 / 13.6 / (.69) / 14.0 / 13.7 / (.60) / .012
Memory:
WMS-R Verbal / 53.0 / 51.5 / (12.67) / 52.0 / 53.0 / (16.43) / 64.0 / 58.8 / (17.07) / 66.0 / 65.3 / (14.87) / <.001 / a / b
WMS-R Visual / 33.5 / 35.5 / (7.62) / 39.0 / 37.5 / (8.88) / 42.5 / 43.3 / (8.89) / 47.0 / 46.1 / (9.03) / <.001 / a / b
Immediate Memory:
WMS-R General / 87.5 / 87.0 / (16.97) / 95.5 / 90.4 / (22.83) / 106.5 / 102.1 / (22.89) / 112.0 / 111.4 / (20.39) / <.001 / a / b
Delayed Memory:
WMS-R Delayed / 40.0 / 40.0 / (13.35) / 49.5 / 46.6 / (14.11) / 47.0 / 52.8 / (18.71) / 63.0 / 62.8 / (15.96) / <.001 / a / b / c
Speed/Attention:
WMS-R: Attention / 51.0 / 52.6 / (11.06) / 56.0 / 58.0 / (10.02) / 60.0 / 59.7 / (10.55) / 62.0 / 62.0 / (10.82) / .006 / a
WAIS-R: Digit Symbol / 33.5 / 33.4 / (11.65) / 37.5 / 38.3 / (12.30) / 38.0 / 39.7 / (12.39) / 46.0 / 46.4 / (11.31) / <.001 / a
Trails Part A (time) / 54.0 / 62.0 / (32.25) / 53.5 / 55.9 / (16.84) / 44.5 / 49.9 / (19.32) / 40.0 / 43.1 / (15.63) / .001 / a / b
Trails Part B (time) / 164.5 / 194.6 / (80.15) / 142.5 / 160.2 / (76.03) / 121.5 / 134.2 / (59.01) / 94.0 / 111.6 / (56.03) / <.001 / a
Verbal Product/Vocabulary:
WAIS-R: Vocabulary / 48.0 / 46.1 / (14.29) / 48.0 / 46.5 / (14.66) / 50.5 / 49.2 / (11.11) / 52.0 / 50.0 / (11.55) / .349
Fluency: Fruit / 10.0 / 10.0 / (2.74) / 10.5 / 9.8 / (3.25) / 11.0 / 11.7 / (3.38) / 13.0 / 12.9 / (3.32) / <.001 / a / b
Fluency: P / 11.5 / 11.6 / (5.56) / 12.5 / 11.5 / (3.94) / 12.5 / 12.6 / (4.63) / 13.0 / 13.3 / (4.61) / .405
Fluency: S / 11.5 / 11.0 / (4.79) / 12.5 / 12.7 / (4.60) / 13.0 / 12.4 / (4.89) / 13.0 / 13.4 / (4.74) / .118
Naming: Boston Naming / 48.5 / 47.3 / (8.15) / 53.5 / 49.2 / (11.57) / 51.0 / 50.1 / (5.79) / 55.0 / 53.6 / (5.18) / <.001 / a / c
Visuoconstructional Ability:
WAIS-R: Block Design / 13.0 / 14.0 / (8.06) / 18.0 / 18.0 / (5.74) / 18.5 / 19.1 / (7.76) / 22.0 / 21.6 / (8.42) / .008 / a

‡ Non-parametric ANOVA conducted on residuals of each neuropsychological test score, controlling for age, gender, race and years of education

a Kruskal-Wallis planned comparison: normal vs. AD onset 1.5-3.4 years, p < 0.01

b Kruskal-Wallis planned comparison: normal vs. AD onset 3.5-5.0 years, p < 0.01

c Kruskal-Wallis planned comparison: normal vs. AD onset 5.1-8.1 years, p < 0.01

Table (E)T-2. Distribution of total number of impaired domains* by group.

Total number of impaired domains* / Normal
(n=621) / Preclinical Dementia
(n=72)
n / Col % / N / Col %
0 / 495 / 79.7% / 33 / 45.8%
1 / 84 / 13.5% / 16 / 22.2%
2 / 21 / 3.4% / 8 / 11.1%
3 / 9 / 1.4% / 4 / 5.6%
4 / 7 / 1.1% / 5 / 6.9%
5 / 5 / .8% / 4 / 5.6%
6 / 2 / 2.8%
TOTAL / 621 / 100.0% / 72 / 100.0%

* See text for definition of “impaired domain.”