- Email: [email protected]
Incidence of cognitively defined late-onset Alzheimer's dementia subgroups from a prospective cohort study
Alzheimer’s & Dementia - (2017) 1-10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
Featured Article
Incidence of cognitively defined late-onset Alzheimer’s dementia subgroups from a prospective cohort study Q15
Paul K. Cranea,*, Emily Trittschuhb,c, Shubhabrata Mukherjeea, Andrew J. Saykind, R. Elizabeth Sandersa, Eric B. Larsone, Susan M. McCurryf, Wayne McCormicka, James D. Boweng, Thomas Grabowskih,i, Mackenzie Moorej, Julianna Baumanj, Alden L. Grossk, C. Dirk Keenel, Thomas E. Birdb,m, Laura E. Gibbonsa, Jesse Mezn, for the Executive Prominent Alzheimer’s Disease: Genetics and Risk Factors (EPAD:GRF) Investigators
a Department of Medicine, University of Washington, Seattle, WA, USA VA Puget Sound Health Care System, Geriatric Research Education and Clinical Center, Seattle, WA, USA c Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle, WA, USA d Department of Radiology and Imaging Sciences and the Indiana Alzheimer’s Disease Center, Indiana University School of Medicine, Indianapolis, IN, USA e Group Health Research Institute, Seattle, WA, USA f Department of Psychosocial and Community Health, University of Washington, Seattle, WA, USA g Department of Neurology, Swedish Medical Center, Seattle, WA, USA h Department of Radiology, University of Washington, Seattle, WA, USA i Department of Neurology, University of Washington, Seattle, WA, USA Q3 j College of Arts and Sciences, University of Washington, Seattle, WA, USA k Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA l Department of Pathology, University of Washington, Seattle, WA, USA m Department of Neurology, University of Washington, Seattle, WA, USA n Department of Neurology, Boston University School of Medicine, Boston, MA, USA
Q1 Q2
b
Abstract
Introduction: There may be biologically relevant heterogeneity within typical late-onset Alzheimer’s dementia. Methods: We analyzed cognitive data from people with incident late-onset Alzheimer’s dementia from a prospective cohort study. We determined individual averages across memory, visuospatial functioning, language, and executive functioning. We identified domains with substantial impairments relative to that average. We compared demographic, neuropathology, and genetic findings across groups defined by relative impairments. Results: During 32,286 person-years of follow-up, 869 people developed Alzheimer’s dementia. There were 393 (48%) with no domain with substantial relative impairments. Some participants had isolated relative impairments in memory (148, 18%), visuospatial functioning (117, 14%), language (71, 9%), and executive functioning (66, 8%). The group with isolated relative memory impairments had higher proportions with APOE ε4, more extensive Alzheimer’s-related neuropathology, and higher proportions with other Alzheimer’s dementia genetic risk variants. Discussion: A cognitive subgrouping strategy may identify biologically distinct subsets of people with Alzheimer’s dementia. Ó 2017 Published by Elsevier Inc. on behalf of the Alzheimer’s Association.
Keywords:
Alzheimer’s disease; Cognition; Subgroups; Endophenotypes; Heterogeneity; Genetics; Neuropathology
Q9
Q4
*Corresponding author. Tel: 11-206-744-1831; Fax: ---. E-mail address: [email protected] http://dx.doi.org/10.1016/j.jalz.2017.04.011 1552-5260/Ó 2017 Published by Elsevier Inc. on behalf of the Alzheimer’s Association. FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
2
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
110 1. Introduction 111 112 There may be considerable heterogeneity in clinical pre113 sentation among people with incident Alzheimer’s dementia 114 begging the question of whether Alzheimer’s dementia in 115 older adults should be considered a single entity or meaning116 fully subdivided into distinct disorders. Meaningfully subdi117 118 viding a condition into distinct groups is essential to the 119 strategy of personalized medicine [1–3]. Data are currently 120 lacking, demonstrating a scalable approach for 121 meaningfully subdividing Alzheimer’s dementia. 122 Recent proposed guidelines identified atypical Alz123 124 heimer’s disease subtypes usually having younger age of 125 onset, including logopenic primary progressive aphasia, dys126 executive Alzheimer’s disease, and posterior cortical atro127 phy [4]. Intriguingly, each of these subtypes is associated 128 with prominent impairment in a single nonmemory 129 130 domain—language, executive functioning, and visuospatial 131 functioning, respectively—with relatively intact memory 132 performance. These previously identified atypical Alz133 heimer’s disease subtypes may represent extremes of a spec134 trum of disease phenotypy. 135 136 We followed insights from neuropsychology, where prac137 titioners have considered patterns of relative impairments 138 across cognitive domains to facilitate diagnosis since the 139 earliest days of the field [5]. We used cognitive data to deter140 mine the distribution at Alzheimer’s dementia diagnosis for 141 142 memory, language, executive functioning, and visuospatial 143 abilities in a community-based prospective cohort study. 144 We determined individual averages across domains and 145 identified domains with substantial impairments relative to 146 that average. We defined subgroup membership based on 147 148 which domains had relative impairments. We compared de149 mographic, neuropathology, and genetic findings across sub150 groups to test the hypothesis that we could use cognitive data 151 to identify biologically distinct late-onset Alzheimer’s de152 mentia subgroups. 153 154 155 2. Methods 156 157 We followed the STROBE guidelines (Appendix A) [6]. Q5 158 All steps are summarized in Appendix B. 159 160 161 2.1. Study population 162 The source population for the Adult Changes in Thought 163 164 (ACT) study consists of community-living members of 165 Group Health, a health maintenance organization in the west166 ern Washington state. A random sample of community-living 167 Group Health members aged 65 years without established 168 dementia diagnoses was invited to an enrollment visit in 169 170 1994 to 1996. The Cognitive Abilities Screening Instrument 171 (CASI) was administered. The CASI is a 100-point scale 172 that assesses several cognitive domains. Individuals with 173 scores .85 were invited to enroll. Those with scores of 174 85 were further evaluated with a neuropsychological bat175 176 tery and comprehensive neurological evaluation. The neuro-
psychological battery included clock drawing [7], verbal fluency [8], Mattis Dementia Rating Scale [9], Boston naming [8], verbal-paired associations and recall, logical memory and recall [10], Word List Memory [8], Constructional Praxis and recall [8], Trails A and B [11], and Information and Comprehension subtest items [10]. All cognitive and clinical data were reviewed in a multidisciplinary consensus conference to determine dementia status; data from each case are discussed and forms with standardized criteria are filled out. Composite scores were not available at the time of consensus conferences and were not considered. Individuals free of dementia were invited to enroll in the longitudinal study. Identical methods were used for an expansion cohort in 2000 to 2003. In 2005, the study began continuous enrollment in which identical methods are used to enroll new participants each month. This report considers all enrollees through April 2015, the most recent data freeze. Once enrolled, participants are administered using the CASI every 2 years. The same procedures are used to identify incident dementia [12] and probable or possible Alzheimer’s disease using NINCDS-ADRDA criteria [13], Q6 referred to here as Alzheimer’s dementia. Other than being a Group Health member, being free of dementia, and volunteering for a longitudinal study, there are no additional inclusion or exclusion criteria for ACT. ACT study evaluates participants in their own homes or at a study clinic for study visits [14]. We focus here on individuals who developed incident Alzheimer’s dementia. The derivation of the analytic cohort is provided in Fig. 1. The study was reviewed by Group Health and University of Washington Institutional Review Boards. Participants gave written informed consent. 2.2. Ascertainment of subgroups An expert panel (E.T., A.J.S., and J.M.) considered each cognitive item and assigned each item to a single cognitive domain—memory, visuospatial functioning, language, executive functioning, or other. We used modern psychometric methods to obtain scores for each domain. Composite scores have been recommended to address idiosyncrasies of individual cognitive tests. Modern psychometric approaches have proven to have incrementally better validity data than scores derived from standard approaches [15–17], and they are specifically recommended for genetic analyses [18]. We recoded observed item responses to avoid sparse response categories and limit to 10 response categories (see Appendix C–F). We used Mplus 7.4 [19] to fit confirmatory factor analysis single factor or bifactor models for each domain separately. All scores were scaled to have mean 0 and standard deviation (SD) 1 in all those with incident AD who had all Q7 four scores (n 5 825). Psychometric modeling details for each domain are provided in Appendix C–F. We determined each person’s average across the four cognitive domain scores. We determined relative
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310
3
Fig. 1. Derivation of the analytic sample.
impairments by calculating differences between each domain score and the individual’s average cognition score across domains. We evaluated candidate thresholds to define “substantial” relative impairments ranging from 0.40 to 1.25 points at 0.05-point increments. For each candidate threshold, we classified people as having 0, 1, or 2 domains with substantial relative impairments; we further divided those with 1 domain according to which domain was relatively impaired. This approach is illustrated in Box 1, where we analyze domain scores for a made-up person and for two real people from ACT diagnosed with atypical Alzheimer’s disease using the same methods described previously. We further illustrate performance on a single-selected test from each domain in Box 2. As discussed in Appendix G, we empirically selected a threshold of 0.75 points. We compared characteristics of groups defined by that threshold. 2.3. Neuropathology procedures ACT neuropathology protocols are published [20,21]. We evaluated neurofibrillary tangles as measured by Braak stage [22], neuritic plaque frequency according to CERAD [23], presence of cerebral amyloid angiopathy, presence of hippocampal sclerosis, presence and location of Lewy bodies categorized as present anywhere and as present in the amygdala, presence of cystic infarcts, and presence and location of cerebral microinfarcts categorized as present anywhere, in the cortex, or in deep structures (basal ganglia or thalamus). We present findings for the group of ACT participants who died free of dementia, for everyone who died following diagnosis of Alzheimer’s dementia, and separately for those in each cognitively defined subgroup. ACT study does not identify people with mild cognitive impairment (MCI); people who died free of dementia could include some people with MCI. The neuropathology results presented here exclude people who died with non–Alzheimer’s dementia.
2.4. Genetic data APOE genotyping was available for most participants [24]. ACT participants with European ancestry were included in the Alzheimer’s Disease Genetics Consortium [24] and the International Genomics of Alzheimer’s Project (IGAP) [25] meta-analyses of genome-wide single-nucleotide variant (SNV) associations with late-onset Alzheimer’s dementia. Genotyping, quality control, and imputation procedures are detailed in those publications. We used imputed genotype dosage data for the SNV with the highest level of association at each of the 20 loci that achieved genome-wide significance in the Lambert et al. IGAP article [25] (“IGAP SNVs”). 2.5. Statistical analyses Group differences in descriptive findings were compared using linear or logistic regression, as appropriate, and a categorical group variable with “no domain” identified as the reference. We used similar models for autopsy findings, controlling for age at death, sex, and years of education. For genetic analyses, we used nondemented elderly individuals from ACT as the reference group (some of these individuals could have had MCI) and included terms for age, sex, and three principal components to account for population stratification. As in [26], we did not adjust for APOE genotype. We determined odds ratios (ORs) for each IGAP SNV for all people with Alzheimer’s dementia and separately for each cognitively defined subgroup. 3. Results 3.1. Characteristics of individuals with incident Alzheimer’s dementia As of April 2015, ACT had enrolled 5088 people, of whom 4365 had at least one follow-up. Over 32,286 person-years of follow-up (mean 8.1), 1076 people developed dementia [12],
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
4 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444
Box 1 Illustration of cognitively defined subgroup procedures.
We determined scores for memory, visuospatial functioning, language, and executive functioning. We determined each person’s average across domains at dementia diagnosis. We determined differences from individual average scores. Our first example case is a hypothetical individual: Scores Average Difference
Memory
Visuospatial functioning
Language
Executive functioning
0.00 20.25 10.25
0.00 20.25 10.25
0.00 20.25 10.25
21.00 20.25 20.75
This person scored at the ACT average among people with Alzheimer’s dementia for memory, visuospatial functioning, and language and scored 1 SD below average for executive functioning (top row). This produces an average score of 20.25 (second row). The third row shows differences from average for each domain, with a difference of 20.75 for executive functioning, which would lead this person to be categorized in the substantial executive functioning impairment subgroup. Our second example case was a real individual with scores as shown in the following:
Scores Average Difference
Memory
Visuospatial functioning
Language
Executive functioning
11.02 20.08 11.10
20.12 20.08 20.04
21.02 20.08 20.94
20.19 20.08 20.11
This person’s average score was 20.08. Her memory score was quite a bit better than this, whereas her language score was substantially worse (difference of 20.94 points). At the ACT consensus conference, she was diagnosed with primary progressive aphasia, a previously identified atypical Alzheimer’s disease subtype characterized by intact memory and a substantial language impairment. Our third example case was another real individual with scores as shown in the following:
Scores Average Difference
Memory
Visuospatial functioning
Language
Executive functioning
12.00 10.96 11.04
10.21 10.96 20.75
11.39 10.96 10.44
10.26 10.96 20.70
This person’s average score was 10.96. His memory score and language score were higher than his executive functioning and visuospatial functioning scores. He was diagnosed with posterior cortical atrophy, another previously identified atypical Alzheimer’s disease subtype characterized by intact memory and a substantial visuospatial functioning impairment. These examples demonstrate procedures used to determine cognitively defined subgroups of typical late-onset Alzheimer’s dementia. The real examples show how these approaches work for one individual each with previously established atypical Alzheimer’s disease subtypes.
of which 869 developed probable or possible Alzheimer’s disease by NINCDS-ADRDA criteria [13], referred to here as Alzheimer’s dementia. We had sufficient data to compute all four cognitive domain scores for 825 (95%) of these; most of the remainder had only a phone-based CASI without visuospatial functioning assessment. On average, people with Alzheimer’s dementia performed much worse than people with normal cognition, especially for memory; see Box 2 for examples of test scores (one per domain) alongside published normative data [27,28]. We used means and SDs from the 825 people from ACT with incident Alzheimer’s dementia and all four domain scores to define the metric (mean 5 0, SD 5 1) for each domain. We identified a threshold of 0.75 points as a threshold indicating substantial relative impairment (See Appendix G). At
that threshold, there were 393 people (48%) who had no domains with a substantial relative impairment. There were 396 people (48%) with a substantial relative impairment in a single domain, including 148 with a substantial relative impairment in memory (18%), 117 in visuospatial functioning (14%), 71 in language (9%), and 66 in executive functioning (8%). The remaining 30 people (4%) had substantial relative impairments in 2 domains. Demographic characteristics of people with incident Alzheimer’s dementia in each cognitively defined subgroup are shown in Table 1, along with APOE genotype and cognitive data. There were no important differences across groups in demographic characteristics, including age at dementia onset. There were substantial differences in the proportion of people with incident Alzheimer’s dementia who had at least one APOE ε4 allele, ranging from 45% of those with
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578
5
Box 2 Performance of people with Alzheimer’s dementia from the ACT study compared with published normative data
Q12
Memory: Logical memory II from the WMS-R1 Raw Score Norms, previously published Normal, 25th percentile 8–9 Normal, 50th percentile 11–13 Normal, 75th percentile 16 Mean (SD) from people from ACT with incident Alzheimer’s dementia All with AD 1.7 (2.6) 0 domains with relative impairment 1.4 (2.1) 1 domain with relative impairment Memory 0.1 (0.7) Visuospatial 2.7 (3.1) Language 3.2 (3.2) Executive functioning 3.6 (3.2) 2 domains with relative impairment 3.5 (3.4)
Executive functioning: Trails B2
Visuospatial construction2
Language: Animal fluency2
8.9 10.2 10.9
12.4 15.3 18.5
8.8 (1.7) 8.9 (1.4)
10.5 (3.9) 10.4 (3.6)
231* 257*
9.7 (1.2) 7.2 (2.0) 9.3 (1.4) 9.2 (1.3) 8.0 (2.9)
10.7 (4.0) 12.0 (4.2) 8.8 (3.5) 8.8 (3.3) 10.8 (4.5)
166* 230* 225* 300* 300*
198 150 109
*Trails B is terminated at 5 minutes (300 seconds). The median and 25th percentile for everyone with Alzheimer’s dementia and for each of the subgroups was 300 seconds (it was 296 seconds for the group with single domain–memory). For example, for everyone with Alzheimer’s dementia, the 25th percentile was 300, and the median was 300, while the 75th percentile was 231. We thus show only the 75th percentile for people with Alzheimer’s dementia from ACT.
This table shows published normative data and data from people with incident Alzheimer’s dementia from the ACT study for one indicator from each domain. For the memory domain, scores for people with Alzheimer’s dementia are substantially lower on Logical Memory II than published normative data for people older than 83 years (median age 88 years)1. The group with substantial relative memory impairment had the very lowest performance, as expected. For the other three tests, we show published data from cognitively intact females aged 85 to 89 with .12 years education2, which is the modal group for the ACT study. In the visuospatial functioning domain, construction was not especially impaired for people with Alzheimer’s dementia compared with published normative data, with the exception of the people with substantial relative visuospatial functioning impairments. In the language domain, animal naming was substantially impaired compared with published normative data across all groups of people with Alzheimer’s dementia, and most profoundly among those with substantial relative language impairments, but also among those with substantial relative executive functioning impairments. In the executive functioning domain, Trails B was 300 seconds (maximum value) for more than half of people with Alzheimer’s dementia in the ACT study, and for at least three quarters of those in the isolated substantial executive functioning group and the group with more than one domain with a substantial impairment. isolated relative memory impairment to 22% of those with isolated relative executive functioning impairment; the overall c2 P value for APOE was .005. Averages for cognitive scores are also shown in Table 1. For most groups, average CASI total scores were similar to the overall average for people with incident Alzheimer’s dementia, though they were somewhat higher for people with relative visuospatial functioning impairments (P 5 .01) and perhaps somewhat lower for those with relative executive functioning impairments (P 5 .61) and especially, those with multiple domains with relative impairments (P 5 .02). The CASI total score includes 10 points for visuospatial functioning and 12 points for executive functioning, so 10% of the total score is based on executive functioning and 12% on executive functioning, while our average scores included other measures of these domains and also assigned each of them to 25% of the total average score. By design, average scores across the entire group of people with incident Alzheimer’s dementia for each cognitive domain were 0 with 1 SD. People with no domains with a substantial relative impairment had average overall scores
and average domain scores that reflected these patterns. The four groups made up of people with a single domain with relative impairments had mean scores for the indicator domain about a full standard deviation below average (about 21.0), whereas scores for the other domains were close to the average (about 0.0) or a little better than the average. On average, the group with multiple domains with relative impairments had lower scores on nonmemory domains. 3.2. Autopsy findings Autopsy findings are summarized in Table 2. Of the 825 people with incident Alzheimer’s dementia and all four cognitive domain scores, 180 had died and come to autopsy at the time of this report. We evaluated data from those individuals and people who died without a diagnosis of dementia; people with non–Alzheimer’s dementia are not included in Table 2. Most neuropathological findings were more common among people with Alzheimer’s dementia than people with no dementia (all P , .023, with the exception of Lewy
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
6 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712
Table 1 Demographic characteristics, APOE genotype, and cognitive scores across cognitively defined subgroups of 825 people with incident Alzheimer’s dementia
Characteristics
Overall (N 5 825), mean (SD) or n (%)
Demographic characteristics Female 522 (63%) Enrollment age 77 (6) Diagnosis age 86 (6) Education 14 (3) APOE genotype Any APOE ε4 243/711 (34%) alleles Cognitive scores CASI 79 (9) Average 0.0 (0.7) cognition Memory 0.0 (1.0)* Visuospatial 0.0 (1.0)* functioning Language 0.0 (1.0)* Executive 0.0 (1.0)* functioning
No domain with relative impairment (N 5 393), mean (SD) or n (%)
Single domain with relative impairment Memory (N 5 148), mean (SD) or n (%)
Visuospatial functioning (N 5 117), mean (SD) or n (%)
Language (N 5 71), mean (SD) or n (%)
Executive functioning (N 5 66), mean (SD) or n (%)
Multiple domains with relative impairments (N 5 30), mean (SD) or n (%)
248 (63%) 77 (6) 86 (6) 14 (3)
92 (62%) 76 (7) 85 (6) 14 (3)
77 (66%) 78 (6) 86 (6) 14 (3)
47 (66%) 77 (6) 86 (5) 14 (3)
37 (56%) 76 (6) 85 (6) 13 (3)
21 (70%) 78 (6) 87 (6) 13 (4)
121/340 (36%)
60/130 (45%)
28/100 (28%)
15/60 (25%)
13/58 (22%)
6/23 (26%)
78 (10) 0.0 (0.7)
79 (8) 0.2 (0.6)
81 (8) 0.0 (0.8)
80 (7) 20.1 (0.8)
77 (8) 20.1 (0.7)
74 (11) 20.4 (0.9)
0.0 (0.8) 0.0 (0.8)
20.9 (0.7) 0.7 (0.7)
0.6 (1.0) 21.2 (0.9)
0.3 (1.0) 0.4 (0.9)
0.5 (0.9) 0.2 (0.9)
0.3 (1.4) 20.4 (1.4)
0.0 (0.8) 0.0 (0.9)
0.4 (0.9) 0.6 (0.8)
0.3 (0.9) 0.1 (1.0)
21.1 (0.9) 0.0 (0.9)
0.1 (0.9) 21.1 (0.8)
20.6 (1.4) 20.8 (1.6)
Abbreviations: CASI, Cognitive Abilities Screening Instrument; SD, standard deviation. *Scores were standardized to have a mean of 0 and SD 1 among people with Alzheimer’s disease who had all four scores.
bodies). The frequency and mean severity of most neuropathological findings were similar across cognitively defined subgroups. Having a Braak stage of IV or higher was more common among people with substantial relative memory impairment (91%) than among everyone else with Alzheimer’s dementia (68%, P 5 .003), and the proportion of people with amyloid angiopathy was also higher, though this did not meet the traditional statistical significance threshold (50% vs. 35%, P 5 .08). 3.3. Genetic findings Genetic findings are summarized in Table 3. The fifth column of Table 3 shows ORs for the association between the 20 SNVs reported in Lambert et al. [25] and Alzheimer’s dementia case-control status in ACT. In general, findings in ACT were similar to those previously reported [25]. The remaining six columns of Table 3 show ORs for the associations between the 20 IGAP SNVs and case-control status, where “cases” were limited to individuals in a cognitively defined subgroup. While we were underpowered for genetic association analyses with data from a single study, we nevertheless marked “extreme risk ORs” (defined as OR . 1.30) in shades of red and “extreme protective ORs” (defined as OR , 1/1.30w0.77) in shades of blue. ORs for those with no domain with substantial relative impairments were broadly similar to those for the entire group of people with Alzheimer’s dementia, though there were somewhat smaller ORs for the SNV associated with CR1 and MS4A6A, and somewhat larger ORs for the SNVs associated with ABCA7 and CD33. For the subgroups of in-
dividuals with a single domain with a substantial relative impairment, findings were quite heterogeneous. There were eight extreme risk ORs for the group with an isolated relative memory impairment. There were nine extreme risk ORs (six risk, three protective) for the group with an isolated relative visuospatial functioning impairment, and nine extreme risk ORs for the group with an isolated substantial language impairment (four risk, five protective). There were six extreme risk ORs for the group with an isolated relative executive functioning impairment (four risk, two protective). The small group with multiple domains with relative impairments had 15 extreme risk ORs (nine risk, five protective). There were several SNVs with extreme protective ORs for at least one cognitively defined subgroup (blue cells) and extreme risk ORs for at least one other subgroup (red cells), including SNVs associated with SORL1, FERMT2, CASS4, PTK2B, ZCWPW1, and NME8. 4. Discussion 4.1. Key findings We developed and implemented an approach to identify subgroups of people with typical late-onset Alzheimer’s dementia on the basis of cognitive test data. We generated scores for four domains—memory, visuospatial functioning, language, and executive functioning—and characterized variation in these scores in a large group of people with incident Alzheimer’s dementia. We determined each individual’s average level of cognition at Alzheimer’s dementia
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846
7
Table 2 Autopsy findings for people with no dementia (n 5 294) and for people with clinical Alzheimer’s dementia (n 5 180)* Alzheimer’s disease Single domain with relative impairments
Autopsy finding
Q13
No dementia (N 5 294), mean (SD) or n (%)
Total, all people with Alzheimer’s disease (N 5 180), mean (SD) or n (%)
No domain with relative impairment (N 5 93), mean (SD) or n (%)
Alzheimer’s pathology, amyloid angiopathy, and hippocampal sclerosis Braak stage, 2.8 (1.6) 4.4 (1.6) 4.3 (1.7) mean (SD) Braak stage 4, 66 (22) 130 (72) 63 (68) n (%) CERAD level, 1.3 (1.1) 2.0 (1.1) 2.1 (1.0) mean (SD) 71 (24) 120 (67) 62 (67) Sufficient AD* Amyloid 61/291 (21) 67/179 (37) 33 (35) angiopathy Hippocampal 17/283 (6) 25/176 (14) 14 (15) sclerosis Lewy bodies Any Lewy bodies 41 (14) 33/179 (18) 19 (20) Amygdala 31/275 (11) 24/173 (14) 14/90 (16) Vascular brain injury Cystic infarcts 65/286 (23) 73/179 (41) 35/92 (38) Microvascular infarcts Any 100/293 (34) 99/179 (55) 50 (54) microvascular infarct Cortex 81/293 (28) 80/179 (45) 42 (45) Deep 69/292 (24) 72/178 (40) 35/92 (38)
Memory (N 5 32), mean (SD) or n (%)
Visuospatial functioning (N 5 25), mean (SD) or n (%)
Language (N 5 12), mean (SD) or n (%)
Executive functioning (N 5 12), mean (SD) or n (%)
Multiple domains with relative impairments (N 5 6), mean (SD) or n (%)
5.1 (1.4)
4.2 (1.5)
4.4 (1.7)
4.4 (1.8)
4.2 (1.2)
29 (91)
16 (64)
9 (75)
9 (75)
4 (67)
2.0 (1.2)
1.9 (1.1)
2.0 (1.2)
1.8 (1.3)
1.5 (1.0)
22 (67) 16 (50)
17 (68) 7/24 (29)
8 (67) 5 (42)
8 (67) 5 (42)
3 (50) 1 (17)
5/50 (17)
1/24 (4)
2 (17)
2 (17)
1/5 (20)
5 (16) 4/32 (13)
4/24 (17) 4/23 (17)
1 (8) 0 (0)
3 (25) 2/11 (18)
1 (17) 0 (0)
12 (38)
12 (48)
7 (58)
5 (42)
2 (33)
18 (56)
14/24 (58)
9 (75)
6 (50)
2 (33)
15 (47) 10 (31)
10/24 (42) 11/24 (46)
7 (58) 8 (67)
5 (42) 7 (58)
1 (17) 1 (17)
*Denominators are specified in column heads unless indicated. CERAD scores are 0 (none), 1 (sparse), 2 (moderate), 3 (frequent), or 4 (severe). “Sufficient Q14 AD” refers to having a Braak stage at least 3 and a CERAD level at least moderate.
diagnosis. We then determined differences from this average to identify domain(s) with substantial relative impairments. About half of people with incident Alzheimer’s dementia had no domains with a substantial relative impairment, and a similar number of people had a single domain with a substantial relative impairment. Demographic characteristics were similar across each of the cognitively defined subgroups, including age at the time of Alzheimer’s dementia diagnosis. This point emphasizes a distinction with the atypical Alzheimer’s disease subtypes such as posterior cortical atrophy and primary progressive aphasia identified by the International Working Group [4]; each of these is described as having an earlier age of onset than is typical for late-onset Alzheimer’s dementia, and certainly earlier age of onset than the mid to late 80s as we found for each subgroup and overall (see Table 1). These atypical Alzheimer’s disease subtypes may occasionally be found in older adults, as demonstrated by the fact that there was one person each with posterior cortical atrophy and with primary progressive aphasia in ACT (see Box 1). Our approach in categorizing people with typical late-onset Alzheimer’s dementia, when applied to those individuals with previously defined atypical Alzheimer’s disease subtypes, worked precisely as one would wish—the person with pos-
terior cortical atrophy had an intact memory score and a substantial relative impairment in visuospatial functioning, whereas the person with primary progressive aphasia had an intact memory score and a substantial relative impairment in language (see Box 1). Additional analyses will be necessary to determine relationships between people diagnosed with PPA, PCA, and dysexecutive Alzheimer’s disease and Q8 people we identify with isolated substantial relative impairments in language, visuospatial functioning, and executive functioning. The group with isolated substantial memory impairment stood out from the other groups. Using only cognitive testing data for our group assignments, we identified this group of 18% of all people with incident late-onset Alzheimer’s dementia who had a higher proportion with APOE ε4 alleles (45% compared with 34% for everyone with Alzheimer’s dementia), a higher proportion with Braak stage 4 (indeed, 91% of this group had Braak stage 4), and a higher proportion with amyloid angiopathy. Furthermore, this group had eight of the IGAP SNVs with ORs 1.30, and they were all in the risk direction (all shades of red in Table 3). Further analyses of this group are warranted. The other subgroups appeared to be readily distinguishable from the group with isolated substantial memory
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
8 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980
Table 3 Genetic findings
1.14 1.14 1.11 1.11
Overall in ACT 1.15 1.10 1.29† 0.85 1.17 1.11 1.25 1.07 1.08 1.01
1.11
0.86
0.89
0.93
1.07
0.65
0.55†
0.64
0.37 0.27 0.29
1.10 1.10 1.10
1.24†
1.28†
1.19 1.03
1.18 1.11
1.22 1.15 1.31
0.82 1.28 0.98
1.49 1.66† 0.57†
1.47 1.12 0.80
1.45 0.92 1.44
0.22
1.10
1.23†
1.31†
1.14
1.44
1.38
1.19
0.99
0.49 0.32 0.37 0.41 0.31
1.08 1.08 1.08 1.08 1.06
1.12 1.02 1.00 1.22† 1.18
1.05 0.97 0.90 1.23 1.33†
1.10 1.33 1.18 1.27 0.89
0.94 0.71 1.26 1.23 1.39
1.26 0.98 0.66 1.14 0.95
1.31 1.05 1.20 0.96 0.98
1.44 1.04 0.62 1.41 1.65
rs number rs11218343 rs6733839 rs6656401 rs9331896 rs4147929 rs10792832
Gene SORL1* BIN1 CR1 CLU* ABCA7 PICALM*
MAF 0.04 0.41 0.20 0.38 0.19 0.36
Reported OR 1.30 1.22 1.18 1.16 1.15 1.15
rs17125944 rs7274581 rs983392 rs11771145
FERMT2 CASS4* MS4A6A* EPHA1* HLA-DRB5HLA-DRB1 PTK2B CD2AP ZCWPW1* SLC24A4RIN3* INPP5D CELF1 NME8 MEF2C* CD33*
0.09 0.08 0.40 0.34 0.28
rs9271192 rs28834970 rs10948363 rs1476679 rs10498633 rs35349669 rs10838725 rs2718058 rs190982 rs3865444
Multiple domains with relative deicits (n = 15) 1.33 0.70 1.80 0.59 1.89 0.55 1.83 0.46 1.09 0.95
No domain with relative deicits (n = 217) 1.05 1.25 1.12 0.85 1.30† 1.14 1.16 1.06 1.02 0.95
Single domain with relative deicits Visuospatial Executive functioning functioning Memory Language (n = 65) (n = 37) (n = 85) (n = 38) 1.69 1.39 0.62 1.01 1.02 1.17 1.09 0.78 1.43† 1.65 1.49 0.98 1.14 0.71 0.71 1.12 1.12 1.13 0.87 0.94 1.36 0.96 1.03 1.20 1.26 1.39 0.67 1.30 1.52 1.67 0.96 0.59 1.39 1.12 1.09 0.84 1.32 0.76 0.82 1.45
Abbreviations: MAF, minor allele frequency; OR, odds ratio; SNV, single-nucleotide variant. NOTE. The first column shows the rs number for the 20 SNVs reported in Lambert et al. (2014) to be associated with late-onset Alzheimer’s dementia. The second column shows the closest gene to the tag SNV. The third column shows the MAF reported in Lambert et al. (2014). The fourth column shows the OR reported in Lambert et al. (2014). Some minor alleles were protective; we took the inverse of these protective ORs (marked with *) such that each OR shown here is in the risk direction. We sorted SNVs in descending order of their ORs. The fifth column shows the overall OR for the ACT study. SNVs with a single dagger (y) had a nominal P ,.05 in the ACT study. The next six columns show ORs for each subgroup in the ACT study, each compared with cognitively normal elderly controls. The color coding scheme is increasingly red for successively higher ORs (1.30–1.49 light pink, 1.50–1.74 medium pink, 1.75–1.99 dark pink, 2.00 dark red) and increasingly blue for successively lower ORs (0.77–0.67 turquoise, 0.66–0.58 light blue, 0.57–0.51 medium blue; ,0.5, dark blue). We chose to mark ORs greater than 1.30 (or less than 1/1.30w0.77) as that was the size of the largest OR reported in Lambert et al. (2014).
impairment and in terms of the pattern of findings across the SNVs they were also distinguishable from each other. Larger sample sizes will be needed to make firmer conclusions about those groups. 4.2. Possible explanations for findings One possible explanation for these findings is that the strategy we implemented to group people based on relative cognitive impairments works to identify biologically distinct Alzheimer’s dementia subtypes. This explanation seems particularly well buttressed for the group with substantial relative impairments in memory, where APOE genotype, Braak stage, and seven IGAP SNVs had extreme ORs, all in the risk direction. The genetic data also suggest support for possible biological relevance for the group with substantial relative impairments in visuospatial functioning, where there were 11 IGAP SNVs with extreme ORs. Another possibility is that our observations do not reflect biologically distinct groups and instead represent the play of chance. Replication in other samples will be very important to differentiate between these two possibilities.
4.3. Implications of findings If these findings are replicated in other samples, they suggest that a noninvasive, widely available technology, –cognitive testing, may differentiate people with Alzheimer’s disease into biologically relevant distinct subgroups. In the future, combinations of cognitive testing alongside other Q10 modalities such as structural magnetic resonance imaging, cerebrospinal fluid biofluids, or positron emission tomography scans may be used to identify subgroups of people with similar biological processes that are distinct from those of people in other subgroups. Much work remains to be done; we are hopeful this work represents initial steps toward a personalized medicine approach to Alzheimer’s dementia therapeutics. 4.4. Strengths and limitations The ACT study includes a sample that reflects the demographic characteristics of the surrounding community. We chose to categorize people based on data at their initial Alzheimer’s dementia diagnosis, so disease duration (and thus
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114
likely severity) should not vary much, though there may be differences in severity across subgroups even within this timeframe. The sample is limited in terms of ethnic heterogeneity; the genetic analyses are limited by design to people with European ancestry. The genetic findings are based on a relatively small number of study participants. Replication in additional study samples will be important. We were unable to evaluate differences in imaging parameters or in rates of cognitive change over time with the data available to us. We evaluated heterogeneity among people with clinical Alzheimer’s dementia based on NINCDS-ADRDA criteria. It has long been appreciated that there is an imperfect correlation between Alzheimer’s pathology and diagnoses of dementia and Alzheimer’s dementia, and that other forms of pathology such as vascular brain disease play important roles [29]. We found some differences in autopsy findings across cognitively defined Alzheimer’s dementia subgroups. Larger samples with autopsy data will be needed to more definitively investigate the roles disparate neurodegenerative conditions may play in the heterogeneity of clinical presentation. The categorization scheme we developed borrows from principles that have been instrumental in clinical neuropsychology for many years, since it considers patterns of impairments, compared in this case to an individual’s overall ability level (as opposed to their estimated premorbid ability). This approach successfully detected the individuals diagnosed with primary progressive aphasia and posterior cortical atrophy from the ACT study (see Box 1). These individuals can be considered “positive controls” to demonstrate the success of the categorization approach. The categorization scheme is based entirely on cognitive data, which is well tolerated and noninvasive. Most of our missing data were for visuospatial functioning; people with visual impairments severe enough that their visuospatial functioning cannot be assessed cannot be categorized using this approach. Furthermore, the visuospatial functioning domain was limited by the data we had available to form our composite scores. Future studies could investigate use of standard rather than modern psychometric approaches to determine subgroup membership. We used theory-driven approaches to categorize people with late-onset Alzheimer’s dementia rather than computer-driven approaches such as cluster analysis. Cluster analysis assigns people to categories that maximize distinctions between classes of people and result in categories that may reflect clinical experience but also may not. Subsequent research could compare our findings to those that would have been obtained with cluster analysis approaches.
9
half of the incident Alzheimer’s dementia had a single domain with a substantial relative impairment. Initial findings provide some support for the notion that these subgroups may be biologically distinct. Further work will be needed to replicate these findings. Acknowledgments Data collection was funded by U01 AG 006781 (E.B.L. and P.K.C., MPIs). Data analyses were funded by R01 AG 042437 (P.K.C., PI), P50 AG05136 (T.G., PI), and R01 AG029672 (P.K.C., PI). Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jalz.2017.04.011.
RESEARCH IN CONTEXT
1. Systematic review: We reviewed PubMed literature sources on endophenotyping in the context of Alzheimer’s dementia. We found many articles on fluid and neuroimaging biomarker strategies, but none that incorporated patterns of cognitive tests. We have contributed to the literature considering differences between memory and executive functioning among people with Alzheimer’s dementia to isolate people with prominent executive dysfunction. We are not aware of studies that have also considered language and visuospatial functioning as well. 2. Interpretation: Our findings indicate that our cognitive subtyping strategy identified groups that had heterogeneous neuropathological findings and disparate relationships with genetic variants previously identified to be associated with risk for late-onset Alzheimer’s dementia. 3. Future directions: Future research is needed to replicate these findings in other populations and determine whether a cognitive subtyping strategy may be useful to isolate biologically relevant subsets of people with Alzheimer’s dementia and thus used as an endophenotyping strategy.
References
4.5. Conclusions Here, we describe development and implementation of an approach to categorize people at the time of Alzheimer’s dementia diagnosis based on patterns of memory, visuospatial functioning, language, and executive functioning. About
[1] Gotovac K, Hajnsek S, Pasic MB, Pivac N, Borovecki F. Personalized medicine in neurodegenerative diseases: how far away? Mol Diagn Ther 2014;18:17–24. [2] Kosik KS. Personalized medicine for effective Alzheimer disease treatment. JAMA Neurol 2015;72:497–8.
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181
10 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248
Q11
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
[3] Cholerton B, Larson EB, Quinn JF, et al. Precision medicine: clarity for the complexity of dementia. Am J Pathol 2016;186:500–6. [4] Dubois B, Feldman HH, Jacova C, et al. Advancing research diagnostic criteria for Alzheimer’s disease: the IWG-2 criteria. Lancet Neurol 2014;13:614–29. [5] Lezak MD. Neuropsychological Assessment. 3 ed. NY: Oxford University Press; 1995. [6] von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007;147:573–7. [7] Spreen O, Strauss E. Compendium of Neuropsychological Tests: Administration, Norms, and Commentary. NY: Oxford UP; 1991. [8] Morris JC, Heyman A, Mohs RC, et al. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology 1989; 39:1159–65. [9] Mattis S. Dementia Rating Scale. Odessa, FL: Psychological Assessment Resources; 1988. [10] Wechsler D. WMS-R: Wechsler Memory Scale - Revised Manual. NY: Psychological Corporation/HBJ; 1987. [11] Reitan RM, Wolfson D. The Halstead-Reitan Neuropsychological Test Battery. Tucson: Neuropsychology Press; 1985. [12] American Psychiatric Association. Task Force on DSM-IV. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. 4th ed. Washington, DC: American Psychiatric Association; 1994. [13] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 1984;34:939–44. [14] Crane PK, Gibbons LE, McCurry SM, et al. Importance of home study visit capacity in dementia studies. Alzheimers Dement 2016;12:419–26. [15] Crane PK, Carle A, Gibbons LE, et al. Development and assessment of a composite score for memory in the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Brain Imaging Behav 2012;6:502–16. [16] Gibbons LE, Carle AC, Mackin RS, et al. A composite score for executive functioning, validated in Alzheimer’s Disease Neuroimaging Initiative (ADNI) participants with baseline mild cognitive impairment. Brain Imaging Behav 2012;6:517–27.
[17] Mukherjee S, Trittschuh E, Gibbons LE, Mackin RS, Saykin A, Crane PK. Dysexecutive and amnesic AD subtypes defined by single indicator and modern psychometric approaches: relationships with SNPs in ADNI. Brain Imaging Behav 2012;6:649–60. [18] van der Sluis S, Verhage M, Posthuma D, Dolan CV. Phenotypic complexity, measurement bias, and poor phenotypic resolution contribute to the missing heritability problem in genetic association studies. PLoS One 2010;5:e13929. [19] Muthen LK, Muthen BO. Mplus User’s Guide. 7th ed. LA: Muthen & Muthen; 1998-2012. [20] Sonnen JA, Larson EB, Walker RL, et al. Nonsteroidal antiinflammatory drugs are associated with increased neuritic plaques. Neurology 2010;75:1203–10. [21] Sonnen JA, Larson EB, Crane PK, et al. Pathological correlates of dementia in a longitudinal, population-based sample of aging. Ann Neurol 2007;62:406–13. [22] Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol 1991;82:239–59. [23] Mirra SS, Heyman A, McKeel D, et al. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer’s disease. Neurology 1991;41:479–86. [24] Naj AC, Jun G, Beecham GW, et al. Common variants at MS4A4/ MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer’s disease. Nat Genet 2011;43:436–41. [25] Lambert JC, Ibrahim-Verbaas CA, Harold D, et al. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet 2013;45:1452–8. [26] Lambert JC, Ibrahim-Verbaas CA, Harold D, et al. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet 2013;45:1452–8. [27] Beeri MS, Schmeidler J, Sano M, et al. Age, gender, and education norms on the CERAD neuropsychological battery in the oldest old. Neurology 2006;67:1006–10. [28] Ivnik RJ, Malec JF, Smith GE, et al. Mayo’s older americans normative studies: WMS-R norms for ages 56 to 94. Clin Neuropsychol 1992; 6:49–82. [29] Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology 2007;69:2197–204.
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315
Featured Article
Incidence of cognitively defined late-onset Alzheimer’s dementia subgroups from a prospective cohort study Q15
Paul K. Cranea,*, Emily Trittschuhb,c, Shubhabrata Mukherjeea, Andrew J. Saykind, R. Elizabeth Sandersa, Eric B. Larsone, Susan M. McCurryf, Wayne McCormicka, James D. Boweng, Thomas Grabowskih,i, Mackenzie Moorej, Julianna Baumanj, Alden L. Grossk, C. Dirk Keenel, Thomas E. Birdb,m, Laura E. Gibbonsa, Jesse Mezn, for the Executive Prominent Alzheimer’s Disease: Genetics and Risk Factors (EPAD:GRF) Investigators
a Department of Medicine, University of Washington, Seattle, WA, USA VA Puget Sound Health Care System, Geriatric Research Education and Clinical Center, Seattle, WA, USA c Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle, WA, USA d Department of Radiology and Imaging Sciences and the Indiana Alzheimer’s Disease Center, Indiana University School of Medicine, Indianapolis, IN, USA e Group Health Research Institute, Seattle, WA, USA f Department of Psychosocial and Community Health, University of Washington, Seattle, WA, USA g Department of Neurology, Swedish Medical Center, Seattle, WA, USA h Department of Radiology, University of Washington, Seattle, WA, USA i Department of Neurology, University of Washington, Seattle, WA, USA Q3 j College of Arts and Sciences, University of Washington, Seattle, WA, USA k Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA l Department of Pathology, University of Washington, Seattle, WA, USA m Department of Neurology, University of Washington, Seattle, WA, USA n Department of Neurology, Boston University School of Medicine, Boston, MA, USA
Q1 Q2
b
Abstract
Introduction: There may be biologically relevant heterogeneity within typical late-onset Alzheimer’s dementia. Methods: We analyzed cognitive data from people with incident late-onset Alzheimer’s dementia from a prospective cohort study. We determined individual averages across memory, visuospatial functioning, language, and executive functioning. We identified domains with substantial impairments relative to that average. We compared demographic, neuropathology, and genetic findings across groups defined by relative impairments. Results: During 32,286 person-years of follow-up, 869 people developed Alzheimer’s dementia. There were 393 (48%) with no domain with substantial relative impairments. Some participants had isolated relative impairments in memory (148, 18%), visuospatial functioning (117, 14%), language (71, 9%), and executive functioning (66, 8%). The group with isolated relative memory impairments had higher proportions with APOE ε4, more extensive Alzheimer’s-related neuropathology, and higher proportions with other Alzheimer’s dementia genetic risk variants. Discussion: A cognitive subgrouping strategy may identify biologically distinct subsets of people with Alzheimer’s dementia. Ó 2017 Published by Elsevier Inc. on behalf of the Alzheimer’s Association.
Keywords:
Alzheimer’s disease; Cognition; Subgroups; Endophenotypes; Heterogeneity; Genetics; Neuropathology
Q9
Q4
*Corresponding author. Tel: 11-206-744-1831; Fax: ---. E-mail address: [email protected] http://dx.doi.org/10.1016/j.jalz.2017.04.011 1552-5260/Ó 2017 Published by Elsevier Inc. on behalf of the Alzheimer’s Association. FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
2
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
110 1. Introduction 111 112 There may be considerable heterogeneity in clinical pre113 sentation among people with incident Alzheimer’s dementia 114 begging the question of whether Alzheimer’s dementia in 115 older adults should be considered a single entity or meaning116 fully subdivided into distinct disorders. Meaningfully subdi117 118 viding a condition into distinct groups is essential to the 119 strategy of personalized medicine [1–3]. Data are currently 120 lacking, demonstrating a scalable approach for 121 meaningfully subdividing Alzheimer’s dementia. 122 Recent proposed guidelines identified atypical Alz123 124 heimer’s disease subtypes usually having younger age of 125 onset, including logopenic primary progressive aphasia, dys126 executive Alzheimer’s disease, and posterior cortical atro127 phy [4]. Intriguingly, each of these subtypes is associated 128 with prominent impairment in a single nonmemory 129 130 domain—language, executive functioning, and visuospatial 131 functioning, respectively—with relatively intact memory 132 performance. These previously identified atypical Alz133 heimer’s disease subtypes may represent extremes of a spec134 trum of disease phenotypy. 135 136 We followed insights from neuropsychology, where prac137 titioners have considered patterns of relative impairments 138 across cognitive domains to facilitate diagnosis since the 139 earliest days of the field [5]. We used cognitive data to deter140 mine the distribution at Alzheimer’s dementia diagnosis for 141 142 memory, language, executive functioning, and visuospatial 143 abilities in a community-based prospective cohort study. 144 We determined individual averages across domains and 145 identified domains with substantial impairments relative to 146 that average. We defined subgroup membership based on 147 148 which domains had relative impairments. We compared de149 mographic, neuropathology, and genetic findings across sub150 groups to test the hypothesis that we could use cognitive data 151 to identify biologically distinct late-onset Alzheimer’s de152 mentia subgroups. 153 154 155 2. Methods 156 157 We followed the STROBE guidelines (Appendix A) [6]. Q5 158 All steps are summarized in Appendix B. 159 160 161 2.1. Study population 162 The source population for the Adult Changes in Thought 163 164 (ACT) study consists of community-living members of 165 Group Health, a health maintenance organization in the west166 ern Washington state. A random sample of community-living 167 Group Health members aged 65 years without established 168 dementia diagnoses was invited to an enrollment visit in 169 170 1994 to 1996. The Cognitive Abilities Screening Instrument 171 (CASI) was administered. The CASI is a 100-point scale 172 that assesses several cognitive domains. Individuals with 173 scores .85 were invited to enroll. Those with scores of 174 85 were further evaluated with a neuropsychological bat175 176 tery and comprehensive neurological evaluation. The neuro-
psychological battery included clock drawing [7], verbal fluency [8], Mattis Dementia Rating Scale [9], Boston naming [8], verbal-paired associations and recall, logical memory and recall [10], Word List Memory [8], Constructional Praxis and recall [8], Trails A and B [11], and Information and Comprehension subtest items [10]. All cognitive and clinical data were reviewed in a multidisciplinary consensus conference to determine dementia status; data from each case are discussed and forms with standardized criteria are filled out. Composite scores were not available at the time of consensus conferences and were not considered. Individuals free of dementia were invited to enroll in the longitudinal study. Identical methods were used for an expansion cohort in 2000 to 2003. In 2005, the study began continuous enrollment in which identical methods are used to enroll new participants each month. This report considers all enrollees through April 2015, the most recent data freeze. Once enrolled, participants are administered using the CASI every 2 years. The same procedures are used to identify incident dementia [12] and probable or possible Alzheimer’s disease using NINCDS-ADRDA criteria [13], Q6 referred to here as Alzheimer’s dementia. Other than being a Group Health member, being free of dementia, and volunteering for a longitudinal study, there are no additional inclusion or exclusion criteria for ACT. ACT study evaluates participants in their own homes or at a study clinic for study visits [14]. We focus here on individuals who developed incident Alzheimer’s dementia. The derivation of the analytic cohort is provided in Fig. 1. The study was reviewed by Group Health and University of Washington Institutional Review Boards. Participants gave written informed consent. 2.2. Ascertainment of subgroups An expert panel (E.T., A.J.S., and J.M.) considered each cognitive item and assigned each item to a single cognitive domain—memory, visuospatial functioning, language, executive functioning, or other. We used modern psychometric methods to obtain scores for each domain. Composite scores have been recommended to address idiosyncrasies of individual cognitive tests. Modern psychometric approaches have proven to have incrementally better validity data than scores derived from standard approaches [15–17], and they are specifically recommended for genetic analyses [18]. We recoded observed item responses to avoid sparse response categories and limit to 10 response categories (see Appendix C–F). We used Mplus 7.4 [19] to fit confirmatory factor analysis single factor or bifactor models for each domain separately. All scores were scaled to have mean 0 and standard deviation (SD) 1 in all those with incident AD who had all Q7 four scores (n 5 825). Psychometric modeling details for each domain are provided in Appendix C–F. We determined each person’s average across the four cognitive domain scores. We determined relative
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310
3
Fig. 1. Derivation of the analytic sample.
impairments by calculating differences between each domain score and the individual’s average cognition score across domains. We evaluated candidate thresholds to define “substantial” relative impairments ranging from 0.40 to 1.25 points at 0.05-point increments. For each candidate threshold, we classified people as having 0, 1, or 2 domains with substantial relative impairments; we further divided those with 1 domain according to which domain was relatively impaired. This approach is illustrated in Box 1, where we analyze domain scores for a made-up person and for two real people from ACT diagnosed with atypical Alzheimer’s disease using the same methods described previously. We further illustrate performance on a single-selected test from each domain in Box 2. As discussed in Appendix G, we empirically selected a threshold of 0.75 points. We compared characteristics of groups defined by that threshold. 2.3. Neuropathology procedures ACT neuropathology protocols are published [20,21]. We evaluated neurofibrillary tangles as measured by Braak stage [22], neuritic plaque frequency according to CERAD [23], presence of cerebral amyloid angiopathy, presence of hippocampal sclerosis, presence and location of Lewy bodies categorized as present anywhere and as present in the amygdala, presence of cystic infarcts, and presence and location of cerebral microinfarcts categorized as present anywhere, in the cortex, or in deep structures (basal ganglia or thalamus). We present findings for the group of ACT participants who died free of dementia, for everyone who died following diagnosis of Alzheimer’s dementia, and separately for those in each cognitively defined subgroup. ACT study does not identify people with mild cognitive impairment (MCI); people who died free of dementia could include some people with MCI. The neuropathology results presented here exclude people who died with non–Alzheimer’s dementia.
2.4. Genetic data APOE genotyping was available for most participants [24]. ACT participants with European ancestry were included in the Alzheimer’s Disease Genetics Consortium [24] and the International Genomics of Alzheimer’s Project (IGAP) [25] meta-analyses of genome-wide single-nucleotide variant (SNV) associations with late-onset Alzheimer’s dementia. Genotyping, quality control, and imputation procedures are detailed in those publications. We used imputed genotype dosage data for the SNV with the highest level of association at each of the 20 loci that achieved genome-wide significance in the Lambert et al. IGAP article [25] (“IGAP SNVs”). 2.5. Statistical analyses Group differences in descriptive findings were compared using linear or logistic regression, as appropriate, and a categorical group variable with “no domain” identified as the reference. We used similar models for autopsy findings, controlling for age at death, sex, and years of education. For genetic analyses, we used nondemented elderly individuals from ACT as the reference group (some of these individuals could have had MCI) and included terms for age, sex, and three principal components to account for population stratification. As in [26], we did not adjust for APOE genotype. We determined odds ratios (ORs) for each IGAP SNV for all people with Alzheimer’s dementia and separately for each cognitively defined subgroup. 3. Results 3.1. Characteristics of individuals with incident Alzheimer’s dementia As of April 2015, ACT had enrolled 5088 people, of whom 4365 had at least one follow-up. Over 32,286 person-years of follow-up (mean 8.1), 1076 people developed dementia [12],
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
4 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444
Box 1 Illustration of cognitively defined subgroup procedures.
We determined scores for memory, visuospatial functioning, language, and executive functioning. We determined each person’s average across domains at dementia diagnosis. We determined differences from individual average scores. Our first example case is a hypothetical individual: Scores Average Difference
Memory
Visuospatial functioning
Language
Executive functioning
0.00 20.25 10.25
0.00 20.25 10.25
0.00 20.25 10.25
21.00 20.25 20.75
This person scored at the ACT average among people with Alzheimer’s dementia for memory, visuospatial functioning, and language and scored 1 SD below average for executive functioning (top row). This produces an average score of 20.25 (second row). The third row shows differences from average for each domain, with a difference of 20.75 for executive functioning, which would lead this person to be categorized in the substantial executive functioning impairment subgroup. Our second example case was a real individual with scores as shown in the following:
Scores Average Difference
Memory
Visuospatial functioning
Language
Executive functioning
11.02 20.08 11.10
20.12 20.08 20.04
21.02 20.08 20.94
20.19 20.08 20.11
This person’s average score was 20.08. Her memory score was quite a bit better than this, whereas her language score was substantially worse (difference of 20.94 points). At the ACT consensus conference, she was diagnosed with primary progressive aphasia, a previously identified atypical Alzheimer’s disease subtype characterized by intact memory and a substantial language impairment. Our third example case was another real individual with scores as shown in the following:
Scores Average Difference
Memory
Visuospatial functioning
Language
Executive functioning
12.00 10.96 11.04
10.21 10.96 20.75
11.39 10.96 10.44
10.26 10.96 20.70
This person’s average score was 10.96. His memory score and language score were higher than his executive functioning and visuospatial functioning scores. He was diagnosed with posterior cortical atrophy, another previously identified atypical Alzheimer’s disease subtype characterized by intact memory and a substantial visuospatial functioning impairment. These examples demonstrate procedures used to determine cognitively defined subgroups of typical late-onset Alzheimer’s dementia. The real examples show how these approaches work for one individual each with previously established atypical Alzheimer’s disease subtypes.
of which 869 developed probable or possible Alzheimer’s disease by NINCDS-ADRDA criteria [13], referred to here as Alzheimer’s dementia. We had sufficient data to compute all four cognitive domain scores for 825 (95%) of these; most of the remainder had only a phone-based CASI without visuospatial functioning assessment. On average, people with Alzheimer’s dementia performed much worse than people with normal cognition, especially for memory; see Box 2 for examples of test scores (one per domain) alongside published normative data [27,28]. We used means and SDs from the 825 people from ACT with incident Alzheimer’s dementia and all four domain scores to define the metric (mean 5 0, SD 5 1) for each domain. We identified a threshold of 0.75 points as a threshold indicating substantial relative impairment (See Appendix G). At
that threshold, there were 393 people (48%) who had no domains with a substantial relative impairment. There were 396 people (48%) with a substantial relative impairment in a single domain, including 148 with a substantial relative impairment in memory (18%), 117 in visuospatial functioning (14%), 71 in language (9%), and 66 in executive functioning (8%). The remaining 30 people (4%) had substantial relative impairments in 2 domains. Demographic characteristics of people with incident Alzheimer’s dementia in each cognitively defined subgroup are shown in Table 1, along with APOE genotype and cognitive data. There were no important differences across groups in demographic characteristics, including age at dementia onset. There were substantial differences in the proportion of people with incident Alzheimer’s dementia who had at least one APOE ε4 allele, ranging from 45% of those with
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578
5
Box 2 Performance of people with Alzheimer’s dementia from the ACT study compared with published normative data
Q12
Memory: Logical memory II from the WMS-R1 Raw Score Norms, previously published Normal, 25th percentile 8–9 Normal, 50th percentile 11–13 Normal, 75th percentile 16 Mean (SD) from people from ACT with incident Alzheimer’s dementia All with AD 1.7 (2.6) 0 domains with relative impairment 1.4 (2.1) 1 domain with relative impairment Memory 0.1 (0.7) Visuospatial 2.7 (3.1) Language 3.2 (3.2) Executive functioning 3.6 (3.2) 2 domains with relative impairment 3.5 (3.4)
Executive functioning: Trails B2
Visuospatial construction2
Language: Animal fluency2
8.9 10.2 10.9
12.4 15.3 18.5
8.8 (1.7) 8.9 (1.4)
10.5 (3.9) 10.4 (3.6)
231* 257*
9.7 (1.2) 7.2 (2.0) 9.3 (1.4) 9.2 (1.3) 8.0 (2.9)
10.7 (4.0) 12.0 (4.2) 8.8 (3.5) 8.8 (3.3) 10.8 (4.5)
166* 230* 225* 300* 300*
198 150 109
*Trails B is terminated at 5 minutes (300 seconds). The median and 25th percentile for everyone with Alzheimer’s dementia and for each of the subgroups was 300 seconds (it was 296 seconds for the group with single domain–memory). For example, for everyone with Alzheimer’s dementia, the 25th percentile was 300, and the median was 300, while the 75th percentile was 231. We thus show only the 75th percentile for people with Alzheimer’s dementia from ACT.
This table shows published normative data and data from people with incident Alzheimer’s dementia from the ACT study for one indicator from each domain. For the memory domain, scores for people with Alzheimer’s dementia are substantially lower on Logical Memory II than published normative data for people older than 83 years (median age 88 years)1. The group with substantial relative memory impairment had the very lowest performance, as expected. For the other three tests, we show published data from cognitively intact females aged 85 to 89 with .12 years education2, which is the modal group for the ACT study. In the visuospatial functioning domain, construction was not especially impaired for people with Alzheimer’s dementia compared with published normative data, with the exception of the people with substantial relative visuospatial functioning impairments. In the language domain, animal naming was substantially impaired compared with published normative data across all groups of people with Alzheimer’s dementia, and most profoundly among those with substantial relative language impairments, but also among those with substantial relative executive functioning impairments. In the executive functioning domain, Trails B was 300 seconds (maximum value) for more than half of people with Alzheimer’s dementia in the ACT study, and for at least three quarters of those in the isolated substantial executive functioning group and the group with more than one domain with a substantial impairment. isolated relative memory impairment to 22% of those with isolated relative executive functioning impairment; the overall c2 P value for APOE was .005. Averages for cognitive scores are also shown in Table 1. For most groups, average CASI total scores were similar to the overall average for people with incident Alzheimer’s dementia, though they were somewhat higher for people with relative visuospatial functioning impairments (P 5 .01) and perhaps somewhat lower for those with relative executive functioning impairments (P 5 .61) and especially, those with multiple domains with relative impairments (P 5 .02). The CASI total score includes 10 points for visuospatial functioning and 12 points for executive functioning, so 10% of the total score is based on executive functioning and 12% on executive functioning, while our average scores included other measures of these domains and also assigned each of them to 25% of the total average score. By design, average scores across the entire group of people with incident Alzheimer’s dementia for each cognitive domain were 0 with 1 SD. People with no domains with a substantial relative impairment had average overall scores
and average domain scores that reflected these patterns. The four groups made up of people with a single domain with relative impairments had mean scores for the indicator domain about a full standard deviation below average (about 21.0), whereas scores for the other domains were close to the average (about 0.0) or a little better than the average. On average, the group with multiple domains with relative impairments had lower scores on nonmemory domains. 3.2. Autopsy findings Autopsy findings are summarized in Table 2. Of the 825 people with incident Alzheimer’s dementia and all four cognitive domain scores, 180 had died and come to autopsy at the time of this report. We evaluated data from those individuals and people who died without a diagnosis of dementia; people with non–Alzheimer’s dementia are not included in Table 2. Most neuropathological findings were more common among people with Alzheimer’s dementia than people with no dementia (all P , .023, with the exception of Lewy
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
6 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712
Table 1 Demographic characteristics, APOE genotype, and cognitive scores across cognitively defined subgroups of 825 people with incident Alzheimer’s dementia
Characteristics
Overall (N 5 825), mean (SD) or n (%)
Demographic characteristics Female 522 (63%) Enrollment age 77 (6) Diagnosis age 86 (6) Education 14 (3) APOE genotype Any APOE ε4 243/711 (34%) alleles Cognitive scores CASI 79 (9) Average 0.0 (0.7) cognition Memory 0.0 (1.0)* Visuospatial 0.0 (1.0)* functioning Language 0.0 (1.0)* Executive 0.0 (1.0)* functioning
No domain with relative impairment (N 5 393), mean (SD) or n (%)
Single domain with relative impairment Memory (N 5 148), mean (SD) or n (%)
Visuospatial functioning (N 5 117), mean (SD) or n (%)
Language (N 5 71), mean (SD) or n (%)
Executive functioning (N 5 66), mean (SD) or n (%)
Multiple domains with relative impairments (N 5 30), mean (SD) or n (%)
248 (63%) 77 (6) 86 (6) 14 (3)
92 (62%) 76 (7) 85 (6) 14 (3)
77 (66%) 78 (6) 86 (6) 14 (3)
47 (66%) 77 (6) 86 (5) 14 (3)
37 (56%) 76 (6) 85 (6) 13 (3)
21 (70%) 78 (6) 87 (6) 13 (4)
121/340 (36%)
60/130 (45%)
28/100 (28%)
15/60 (25%)
13/58 (22%)
6/23 (26%)
78 (10) 0.0 (0.7)
79 (8) 0.2 (0.6)
81 (8) 0.0 (0.8)
80 (7) 20.1 (0.8)
77 (8) 20.1 (0.7)
74 (11) 20.4 (0.9)
0.0 (0.8) 0.0 (0.8)
20.9 (0.7) 0.7 (0.7)
0.6 (1.0) 21.2 (0.9)
0.3 (1.0) 0.4 (0.9)
0.5 (0.9) 0.2 (0.9)
0.3 (1.4) 20.4 (1.4)
0.0 (0.8) 0.0 (0.9)
0.4 (0.9) 0.6 (0.8)
0.3 (0.9) 0.1 (1.0)
21.1 (0.9) 0.0 (0.9)
0.1 (0.9) 21.1 (0.8)
20.6 (1.4) 20.8 (1.6)
Abbreviations: CASI, Cognitive Abilities Screening Instrument; SD, standard deviation. *Scores were standardized to have a mean of 0 and SD 1 among people with Alzheimer’s disease who had all four scores.
bodies). The frequency and mean severity of most neuropathological findings were similar across cognitively defined subgroups. Having a Braak stage of IV or higher was more common among people with substantial relative memory impairment (91%) than among everyone else with Alzheimer’s dementia (68%, P 5 .003), and the proportion of people with amyloid angiopathy was also higher, though this did not meet the traditional statistical significance threshold (50% vs. 35%, P 5 .08). 3.3. Genetic findings Genetic findings are summarized in Table 3. The fifth column of Table 3 shows ORs for the association between the 20 SNVs reported in Lambert et al. [25] and Alzheimer’s dementia case-control status in ACT. In general, findings in ACT were similar to those previously reported [25]. The remaining six columns of Table 3 show ORs for the associations between the 20 IGAP SNVs and case-control status, where “cases” were limited to individuals in a cognitively defined subgroup. While we were underpowered for genetic association analyses with data from a single study, we nevertheless marked “extreme risk ORs” (defined as OR . 1.30) in shades of red and “extreme protective ORs” (defined as OR , 1/1.30w0.77) in shades of blue. ORs for those with no domain with substantial relative impairments were broadly similar to those for the entire group of people with Alzheimer’s dementia, though there were somewhat smaller ORs for the SNV associated with CR1 and MS4A6A, and somewhat larger ORs for the SNVs associated with ABCA7 and CD33. For the subgroups of in-
dividuals with a single domain with a substantial relative impairment, findings were quite heterogeneous. There were eight extreme risk ORs for the group with an isolated relative memory impairment. There were nine extreme risk ORs (six risk, three protective) for the group with an isolated relative visuospatial functioning impairment, and nine extreme risk ORs for the group with an isolated substantial language impairment (four risk, five protective). There were six extreme risk ORs for the group with an isolated relative executive functioning impairment (four risk, two protective). The small group with multiple domains with relative impairments had 15 extreme risk ORs (nine risk, five protective). There were several SNVs with extreme protective ORs for at least one cognitively defined subgroup (blue cells) and extreme risk ORs for at least one other subgroup (red cells), including SNVs associated with SORL1, FERMT2, CASS4, PTK2B, ZCWPW1, and NME8. 4. Discussion 4.1. Key findings We developed and implemented an approach to identify subgroups of people with typical late-onset Alzheimer’s dementia on the basis of cognitive test data. We generated scores for four domains—memory, visuospatial functioning, language, and executive functioning—and characterized variation in these scores in a large group of people with incident Alzheimer’s dementia. We determined each individual’s average level of cognition at Alzheimer’s dementia
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846
7
Table 2 Autopsy findings for people with no dementia (n 5 294) and for people with clinical Alzheimer’s dementia (n 5 180)* Alzheimer’s disease Single domain with relative impairments
Autopsy finding
Q13
No dementia (N 5 294), mean (SD) or n (%)
Total, all people with Alzheimer’s disease (N 5 180), mean (SD) or n (%)
No domain with relative impairment (N 5 93), mean (SD) or n (%)
Alzheimer’s pathology, amyloid angiopathy, and hippocampal sclerosis Braak stage, 2.8 (1.6) 4.4 (1.6) 4.3 (1.7) mean (SD) Braak stage 4, 66 (22) 130 (72) 63 (68) n (%) CERAD level, 1.3 (1.1) 2.0 (1.1) 2.1 (1.0) mean (SD) 71 (24) 120 (67) 62 (67) Sufficient AD* Amyloid 61/291 (21) 67/179 (37) 33 (35) angiopathy Hippocampal 17/283 (6) 25/176 (14) 14 (15) sclerosis Lewy bodies Any Lewy bodies 41 (14) 33/179 (18) 19 (20) Amygdala 31/275 (11) 24/173 (14) 14/90 (16) Vascular brain injury Cystic infarcts 65/286 (23) 73/179 (41) 35/92 (38) Microvascular infarcts Any 100/293 (34) 99/179 (55) 50 (54) microvascular infarct Cortex 81/293 (28) 80/179 (45) 42 (45) Deep 69/292 (24) 72/178 (40) 35/92 (38)
Memory (N 5 32), mean (SD) or n (%)
Visuospatial functioning (N 5 25), mean (SD) or n (%)
Language (N 5 12), mean (SD) or n (%)
Executive functioning (N 5 12), mean (SD) or n (%)
Multiple domains with relative impairments (N 5 6), mean (SD) or n (%)
5.1 (1.4)
4.2 (1.5)
4.4 (1.7)
4.4 (1.8)
4.2 (1.2)
29 (91)
16 (64)
9 (75)
9 (75)
4 (67)
2.0 (1.2)
1.9 (1.1)
2.0 (1.2)
1.8 (1.3)
1.5 (1.0)
22 (67) 16 (50)
17 (68) 7/24 (29)
8 (67) 5 (42)
8 (67) 5 (42)
3 (50) 1 (17)
5/50 (17)
1/24 (4)
2 (17)
2 (17)
1/5 (20)
5 (16) 4/32 (13)
4/24 (17) 4/23 (17)
1 (8) 0 (0)
3 (25) 2/11 (18)
1 (17) 0 (0)
12 (38)
12 (48)
7 (58)
5 (42)
2 (33)
18 (56)
14/24 (58)
9 (75)
6 (50)
2 (33)
15 (47) 10 (31)
10/24 (42) 11/24 (46)
7 (58) 8 (67)
5 (42) 7 (58)
1 (17) 1 (17)
*Denominators are specified in column heads unless indicated. CERAD scores are 0 (none), 1 (sparse), 2 (moderate), 3 (frequent), or 4 (severe). “Sufficient Q14 AD” refers to having a Braak stage at least 3 and a CERAD level at least moderate.
diagnosis. We then determined differences from this average to identify domain(s) with substantial relative impairments. About half of people with incident Alzheimer’s dementia had no domains with a substantial relative impairment, and a similar number of people had a single domain with a substantial relative impairment. Demographic characteristics were similar across each of the cognitively defined subgroups, including age at the time of Alzheimer’s dementia diagnosis. This point emphasizes a distinction with the atypical Alzheimer’s disease subtypes such as posterior cortical atrophy and primary progressive aphasia identified by the International Working Group [4]; each of these is described as having an earlier age of onset than is typical for late-onset Alzheimer’s dementia, and certainly earlier age of onset than the mid to late 80s as we found for each subgroup and overall (see Table 1). These atypical Alzheimer’s disease subtypes may occasionally be found in older adults, as demonstrated by the fact that there was one person each with posterior cortical atrophy and with primary progressive aphasia in ACT (see Box 1). Our approach in categorizing people with typical late-onset Alzheimer’s dementia, when applied to those individuals with previously defined atypical Alzheimer’s disease subtypes, worked precisely as one would wish—the person with pos-
terior cortical atrophy had an intact memory score and a substantial relative impairment in visuospatial functioning, whereas the person with primary progressive aphasia had an intact memory score and a substantial relative impairment in language (see Box 1). Additional analyses will be necessary to determine relationships between people diagnosed with PPA, PCA, and dysexecutive Alzheimer’s disease and Q8 people we identify with isolated substantial relative impairments in language, visuospatial functioning, and executive functioning. The group with isolated substantial memory impairment stood out from the other groups. Using only cognitive testing data for our group assignments, we identified this group of 18% of all people with incident late-onset Alzheimer’s dementia who had a higher proportion with APOE ε4 alleles (45% compared with 34% for everyone with Alzheimer’s dementia), a higher proportion with Braak stage 4 (indeed, 91% of this group had Braak stage 4), and a higher proportion with amyloid angiopathy. Furthermore, this group had eight of the IGAP SNVs with ORs 1.30, and they were all in the risk direction (all shades of red in Table 3). Further analyses of this group are warranted. The other subgroups appeared to be readily distinguishable from the group with isolated substantial memory
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
8 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980
Table 3 Genetic findings
1.14 1.14 1.11 1.11
Overall in ACT 1.15 1.10 1.29† 0.85 1.17 1.11 1.25 1.07 1.08 1.01
1.11
0.86
0.89
0.93
1.07
0.65
0.55†
0.64
0.37 0.27 0.29
1.10 1.10 1.10
1.24†
1.28†
1.19 1.03
1.18 1.11
1.22 1.15 1.31
0.82 1.28 0.98
1.49 1.66† 0.57†
1.47 1.12 0.80
1.45 0.92 1.44
0.22
1.10
1.23†
1.31†
1.14
1.44
1.38
1.19
0.99
0.49 0.32 0.37 0.41 0.31
1.08 1.08 1.08 1.08 1.06
1.12 1.02 1.00 1.22† 1.18
1.05 0.97 0.90 1.23 1.33†
1.10 1.33 1.18 1.27 0.89
0.94 0.71 1.26 1.23 1.39
1.26 0.98 0.66 1.14 0.95
1.31 1.05 1.20 0.96 0.98
1.44 1.04 0.62 1.41 1.65
rs number rs11218343 rs6733839 rs6656401 rs9331896 rs4147929 rs10792832
Gene SORL1* BIN1 CR1 CLU* ABCA7 PICALM*
MAF 0.04 0.41 0.20 0.38 0.19 0.36
Reported OR 1.30 1.22 1.18 1.16 1.15 1.15
rs17125944 rs7274581 rs983392 rs11771145
FERMT2 CASS4* MS4A6A* EPHA1* HLA-DRB5HLA-DRB1 PTK2B CD2AP ZCWPW1* SLC24A4RIN3* INPP5D CELF1 NME8 MEF2C* CD33*
0.09 0.08 0.40 0.34 0.28
rs9271192 rs28834970 rs10948363 rs1476679 rs10498633 rs35349669 rs10838725 rs2718058 rs190982 rs3865444
Multiple domains with relative deicits (n = 15) 1.33 0.70 1.80 0.59 1.89 0.55 1.83 0.46 1.09 0.95
No domain with relative deicits (n = 217) 1.05 1.25 1.12 0.85 1.30† 1.14 1.16 1.06 1.02 0.95
Single domain with relative deicits Visuospatial Executive functioning functioning Memory Language (n = 65) (n = 37) (n = 85) (n = 38) 1.69 1.39 0.62 1.01 1.02 1.17 1.09 0.78 1.43† 1.65 1.49 0.98 1.14 0.71 0.71 1.12 1.12 1.13 0.87 0.94 1.36 0.96 1.03 1.20 1.26 1.39 0.67 1.30 1.52 1.67 0.96 0.59 1.39 1.12 1.09 0.84 1.32 0.76 0.82 1.45
Abbreviations: MAF, minor allele frequency; OR, odds ratio; SNV, single-nucleotide variant. NOTE. The first column shows the rs number for the 20 SNVs reported in Lambert et al. (2014) to be associated with late-onset Alzheimer’s dementia. The second column shows the closest gene to the tag SNV. The third column shows the MAF reported in Lambert et al. (2014). The fourth column shows the OR reported in Lambert et al. (2014). Some minor alleles were protective; we took the inverse of these protective ORs (marked with *) such that each OR shown here is in the risk direction. We sorted SNVs in descending order of their ORs. The fifth column shows the overall OR for the ACT study. SNVs with a single dagger (y) had a nominal P ,.05 in the ACT study. The next six columns show ORs for each subgroup in the ACT study, each compared with cognitively normal elderly controls. The color coding scheme is increasingly red for successively higher ORs (1.30–1.49 light pink, 1.50–1.74 medium pink, 1.75–1.99 dark pink, 2.00 dark red) and increasingly blue for successively lower ORs (0.77–0.67 turquoise, 0.66–0.58 light blue, 0.57–0.51 medium blue; ,0.5, dark blue). We chose to mark ORs greater than 1.30 (or less than 1/1.30w0.77) as that was the size of the largest OR reported in Lambert et al. (2014).
impairment and in terms of the pattern of findings across the SNVs they were also distinguishable from each other. Larger sample sizes will be needed to make firmer conclusions about those groups. 4.2. Possible explanations for findings One possible explanation for these findings is that the strategy we implemented to group people based on relative cognitive impairments works to identify biologically distinct Alzheimer’s dementia subtypes. This explanation seems particularly well buttressed for the group with substantial relative impairments in memory, where APOE genotype, Braak stage, and seven IGAP SNVs had extreme ORs, all in the risk direction. The genetic data also suggest support for possible biological relevance for the group with substantial relative impairments in visuospatial functioning, where there were 11 IGAP SNVs with extreme ORs. Another possibility is that our observations do not reflect biologically distinct groups and instead represent the play of chance. Replication in other samples will be very important to differentiate between these two possibilities.
4.3. Implications of findings If these findings are replicated in other samples, they suggest that a noninvasive, widely available technology, –cognitive testing, may differentiate people with Alzheimer’s disease into biologically relevant distinct subgroups. In the future, combinations of cognitive testing alongside other Q10 modalities such as structural magnetic resonance imaging, cerebrospinal fluid biofluids, or positron emission tomography scans may be used to identify subgroups of people with similar biological processes that are distinct from those of people in other subgroups. Much work remains to be done; we are hopeful this work represents initial steps toward a personalized medicine approach to Alzheimer’s dementia therapeutics. 4.4. Strengths and limitations The ACT study includes a sample that reflects the demographic characteristics of the surrounding community. We chose to categorize people based on data at their initial Alzheimer’s dementia diagnosis, so disease duration (and thus
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114
likely severity) should not vary much, though there may be differences in severity across subgroups even within this timeframe. The sample is limited in terms of ethnic heterogeneity; the genetic analyses are limited by design to people with European ancestry. The genetic findings are based on a relatively small number of study participants. Replication in additional study samples will be important. We were unable to evaluate differences in imaging parameters or in rates of cognitive change over time with the data available to us. We evaluated heterogeneity among people with clinical Alzheimer’s dementia based on NINCDS-ADRDA criteria. It has long been appreciated that there is an imperfect correlation between Alzheimer’s pathology and diagnoses of dementia and Alzheimer’s dementia, and that other forms of pathology such as vascular brain disease play important roles [29]. We found some differences in autopsy findings across cognitively defined Alzheimer’s dementia subgroups. Larger samples with autopsy data will be needed to more definitively investigate the roles disparate neurodegenerative conditions may play in the heterogeneity of clinical presentation. The categorization scheme we developed borrows from principles that have been instrumental in clinical neuropsychology for many years, since it considers patterns of impairments, compared in this case to an individual’s overall ability level (as opposed to their estimated premorbid ability). This approach successfully detected the individuals diagnosed with primary progressive aphasia and posterior cortical atrophy from the ACT study (see Box 1). These individuals can be considered “positive controls” to demonstrate the success of the categorization approach. The categorization scheme is based entirely on cognitive data, which is well tolerated and noninvasive. Most of our missing data were for visuospatial functioning; people with visual impairments severe enough that their visuospatial functioning cannot be assessed cannot be categorized using this approach. Furthermore, the visuospatial functioning domain was limited by the data we had available to form our composite scores. Future studies could investigate use of standard rather than modern psychometric approaches to determine subgroup membership. We used theory-driven approaches to categorize people with late-onset Alzheimer’s dementia rather than computer-driven approaches such as cluster analysis. Cluster analysis assigns people to categories that maximize distinctions between classes of people and result in categories that may reflect clinical experience but also may not. Subsequent research could compare our findings to those that would have been obtained with cluster analysis approaches.
9
half of the incident Alzheimer’s dementia had a single domain with a substantial relative impairment. Initial findings provide some support for the notion that these subgroups may be biologically distinct. Further work will be needed to replicate these findings. Acknowledgments Data collection was funded by U01 AG 006781 (E.B.L. and P.K.C., MPIs). Data analyses were funded by R01 AG 042437 (P.K.C., PI), P50 AG05136 (T.G., PI), and R01 AG029672 (P.K.C., PI). Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jalz.2017.04.011.
RESEARCH IN CONTEXT
1. Systematic review: We reviewed PubMed literature sources on endophenotyping in the context of Alzheimer’s dementia. We found many articles on fluid and neuroimaging biomarker strategies, but none that incorporated patterns of cognitive tests. We have contributed to the literature considering differences between memory and executive functioning among people with Alzheimer’s dementia to isolate people with prominent executive dysfunction. We are not aware of studies that have also considered language and visuospatial functioning as well. 2. Interpretation: Our findings indicate that our cognitive subtyping strategy identified groups that had heterogeneous neuropathological findings and disparate relationships with genetic variants previously identified to be associated with risk for late-onset Alzheimer’s dementia. 3. Future directions: Future research is needed to replicate these findings in other populations and determine whether a cognitive subtyping strategy may be useful to isolate biologically relevant subsets of people with Alzheimer’s dementia and thus used as an endophenotyping strategy.
References
4.5. Conclusions Here, we describe development and implementation of an approach to categorize people at the time of Alzheimer’s dementia diagnosis based on patterns of memory, visuospatial functioning, language, and executive functioning. About
[1] Gotovac K, Hajnsek S, Pasic MB, Pivac N, Borovecki F. Personalized medicine in neurodegenerative diseases: how far away? Mol Diagn Ther 2014;18:17–24. [2] Kosik KS. Personalized medicine for effective Alzheimer disease treatment. JAMA Neurol 2015;72:497–8.
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181
10 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248
Q11
P.K. Crane et al. / Alzheimer’s & Dementia - (2017) 1-10
[3] Cholerton B, Larson EB, Quinn JF, et al. Precision medicine: clarity for the complexity of dementia. Am J Pathol 2016;186:500–6. [4] Dubois B, Feldman HH, Jacova C, et al. Advancing research diagnostic criteria for Alzheimer’s disease: the IWG-2 criteria. Lancet Neurol 2014;13:614–29. [5] Lezak MD. Neuropsychological Assessment. 3 ed. NY: Oxford University Press; 1995. [6] von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007;147:573–7. [7] Spreen O, Strauss E. Compendium of Neuropsychological Tests: Administration, Norms, and Commentary. NY: Oxford UP; 1991. [8] Morris JC, Heyman A, Mohs RC, et al. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology 1989; 39:1159–65. [9] Mattis S. Dementia Rating Scale. Odessa, FL: Psychological Assessment Resources; 1988. [10] Wechsler D. WMS-R: Wechsler Memory Scale - Revised Manual. NY: Psychological Corporation/HBJ; 1987. [11] Reitan RM, Wolfson D. The Halstead-Reitan Neuropsychological Test Battery. Tucson: Neuropsychology Press; 1985. [12] American Psychiatric Association. Task Force on DSM-IV. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. 4th ed. Washington, DC: American Psychiatric Association; 1994. [13] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 1984;34:939–44. [14] Crane PK, Gibbons LE, McCurry SM, et al. Importance of home study visit capacity in dementia studies. Alzheimers Dement 2016;12:419–26. [15] Crane PK, Carle A, Gibbons LE, et al. Development and assessment of a composite score for memory in the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Brain Imaging Behav 2012;6:502–16. [16] Gibbons LE, Carle AC, Mackin RS, et al. A composite score for executive functioning, validated in Alzheimer’s Disease Neuroimaging Initiative (ADNI) participants with baseline mild cognitive impairment. Brain Imaging Behav 2012;6:517–27.
[17] Mukherjee S, Trittschuh E, Gibbons LE, Mackin RS, Saykin A, Crane PK. Dysexecutive and amnesic AD subtypes defined by single indicator and modern psychometric approaches: relationships with SNPs in ADNI. Brain Imaging Behav 2012;6:649–60. [18] van der Sluis S, Verhage M, Posthuma D, Dolan CV. Phenotypic complexity, measurement bias, and poor phenotypic resolution contribute to the missing heritability problem in genetic association studies. PLoS One 2010;5:e13929. [19] Muthen LK, Muthen BO. Mplus User’s Guide. 7th ed. LA: Muthen & Muthen; 1998-2012. [20] Sonnen JA, Larson EB, Walker RL, et al. Nonsteroidal antiinflammatory drugs are associated with increased neuritic plaques. Neurology 2010;75:1203–10. [21] Sonnen JA, Larson EB, Crane PK, et al. Pathological correlates of dementia in a longitudinal, population-based sample of aging. Ann Neurol 2007;62:406–13. [22] Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol 1991;82:239–59. [23] Mirra SS, Heyman A, McKeel D, et al. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer’s disease. Neurology 1991;41:479–86. [24] Naj AC, Jun G, Beecham GW, et al. Common variants at MS4A4/ MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer’s disease. Nat Genet 2011;43:436–41. [25] Lambert JC, Ibrahim-Verbaas CA, Harold D, et al. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet 2013;45:1452–8. [26] Lambert JC, Ibrahim-Verbaas CA, Harold D, et al. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet 2013;45:1452–8. [27] Beeri MS, Schmeidler J, Sano M, et al. Age, gender, and education norms on the CERAD neuropsychological battery in the oldest old. Neurology 2006;67:1006–10. [28] Ivnik RJ, Malec JF, Smith GE, et al. Mayo’s older americans normative studies: WMS-R norms for ages 56 to 94. Clin Neuropsychol 1992; 6:49–82. [29] Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology 2007;69:2197–204.
FLA 5.4.0 DTD JALZ2397_proof 14 June 2017 5:01 pm ce
1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315