This article was downloaded by: [HINARI Consortium (T&F)] On: 10 February 2011 Access details: Access Details:

[subscription number 791525795] Publisher Psychology Press Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK

The Clinical Neuropsychologist

Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713721659

Diagnostic Utility of Letter Fluency, Category Fluency, and Fluency Difference Scores in Alzheimer's Disease

Jane H. Cerhan; Robert J. Ivnik; Glenn E. Smith; Eric C. Tangalos; Ronald C. Petersen; Bradley F. Boeve Online publication date: 09 August 2010

To cite this Article Cerhan, Jane H. , Ivnik, Robert J. , Smith, Glenn E. , Tangalos, Eric C. , Petersen, Ronald C. and Boeve,

Bradley F.(2002) 'Diagnostic Utility of Letter Fluency, Category Fluency, and Fluency Difference Scores in Alzheimer's Disease', The Clinical Neuropsychologist, 16: 1, 35 42 To link to this Article: DOI: 10.1076/clin.16.1.35.8326 URL: http://dx.doi.org/10.1076/clin.16.1.35.8326

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

The Clinical Neuropsychologist 2002, Vol. 16, No. 1, pp. 3542

1385-4046/02/1601-035$16.00 # Swets & Zeitlinger

Diagnostic Utility of Letter Fluency, Category Fluency, and Fluency Difference Scores in Alzheimer's Disease
1

Department of Psychiatry & Psychology, 2 Division of Community Internal Medicine, and 3 Neurology Department, Mayo Clinic, Rochester, MN, USA

Jane H. Cerhan 1, Robert J. Ivnik 1, Glenn E. Smith 1, Eric C. Tangalos 2, Ronald C. Petersen 2, and Bradley F. Boeve 3

Downloaded By: [HINARI Consortium (T&F)] At: 01:04 10 February 2011

ABSTRACT
Word generation tasks, specically letter uency and category uency, are a standard part of most test protocols used by clinical neuropsychologists to measure aspects of language and executive retrieval functions. Given the reliance on these measures as clinical tools, it is important to understand the diagnostic implications of patients' performances. In Alzheimer's disease (AD), category uency has generally been found to be disproportionately impaired, whereas letter uency is usually more mildly impaired. It has been proposed that this performance pattern occurs because access to temporal-lobe semantic stores is necessary for category uency but not letter uency. In this study, the diagnostic utility of category uency, letter uency, and difference scores (letter uency minus category uency) in AD were examined. Forty AD patients and 221 normal control subjects took category uency and letter uency tests. Traditional t-test comparisons revealed that the groups differed signicantly on the two test tasks and in terms of mean difference scores. However, using sensitivity and specicity to calculate predictive values, only category uency and letter uency (but not difference scores) were useful in predicting AD in individuals. Furthermore, category uency was superior to letter uency in this regard. Likelihood ratio tables are provided for use in calculating the odds of AD for specic category uency and letter uency scores generated by individual patients.

Word generation tasks are a standard part of most test protocols used by clinical neuropsychologists to measure aspects of language and executive retrieval functions. Most common are letter uency and category uency tasks. Letter uency tasks require generation of words starting with specic letters within a limited time. Category uency tasks require individuals to generate exemplars of specied semantic categories, such as the names of animals or fruits. Given the reliance on these measures as clinical tools, it is important to understand the diagnostic implications of patients' performances. In Alzheimer's disease (AD), category uency has frequently been found to be disproportionately impaired, whereas letter uency is usually more

mildly impaired (Crossley, D'Arcy, & Rawson, 1997; Martin & Fedio, 1983; Monsch et al., 1994; Salmon, Heindel, & Lange, 1999; Zec et al., 1990) or even intact (Butters, Granholm, Salmon, Grant, & Wolfe, 1987). This performance pattern has not been found uniformly (e.g., Hart, Smith, & Swash, 1988; Suhr & Jones, 1998), but it has been reported often. For example, Monsch et al. (1994) compared 89 patients with AD and 53 demographically similar control subjects on four uency measures. Using Receiver Operating Characteristics (ROC) curves, they found that category uency best discriminated between patients and control subjects (sensitivity 100% and specicity 92.5%) and letter uency was the

Address correspondence to: Jane H. Cerhan, Ph.D., Department of Psychiatry & Psychology (W-11-B), Mayo Clinic, Rochester, MN 55905, USA. Accepted for publication: December 3, 2001.

36

JANE H. CERHAN ET AL.

least accurate. Furthermore, category uency was the only measure that lost none of its discriminative capability when applied to a subgroup of 21 mildly impaired AD patients. This apparent discrepancy in AD patients' performances on letter and category uency suggests that different cognitive operations underlie the two tasks. To explain this, it has been suggested that both category uency and letter uency require frontally mediated executiveretrieval mechanisms. Both also require access to phonological/lexical stores. But, only category uency also requires access to more widely distributed semantic stores as the subject searches for exemplars tting the target category. The poorer performance of AD patients on category uency tasks has therefore been posited to relate to the breakdown of temporal-lobe semantic stores in this disease (Butters et al., 1987). While still debated, support for this semanticdegradation hypothesis has come from a variety of sources. For example, some authors have reported that category uency and letter uency are more equally impaired in `subcortical' dementias such as Huntington's disease (Butters et al., 1987; Monsch et al., 1994), Korsakoff's syndrome (Butters et al., 1987), and Progressive Supranuclear Palsy (Rosser & Hodges, 1994), where executive retrieval impairment is assumed to be operating. Yet, other welldesigned studies have found no discrepancy in uency patterns between AD patients and patients with `subcortical' dementias including Huntington's and Parkinson's diseases (Barr & Brandt, 1996; Suhr & Jones; 1998). Salmon et al. (1999) found evidence for semantic degradation by observing that category uency is more sensitive than letter uency to decline in AD over time, and that AD patients were less likely than normal control subjects to repeat semantic but not phonemic exemplars across exams. Randolph, Braun, Goldberg, and Chase (1993) found that cueing on category uency helped patients with Huntington's and Parkinson's diseases, but not AD patients. Some intriguing experimental (Martin, Wiggs, LaLonde, & Mack, 1994) and imaging (Keilp, Gorlyn, Alexander, Stern, & Prohovnik, 1999) research has also provided support for the semantic degradation argument. See Ober (1999) for a more extended debate.

Downloaded By: [HINARI Consortium (T&F)] At: 01:04 10 February 2011

Given the frequent nding that category uency is worse than letter uency in patients with AD, and that this pattern might be distinct from other disease entities, the question arises as to whether difference scores would be useful in clinical prediction. Few studies have addressed this question. Sherman and Massman (1999) examined difference scores in 217 patients with AD. They found that, as a group, AD patients performed more poorly on category uency than letter uency, but only two thirds of the AD sample showed the expected pattern of category uency less than letter uency. The other third showed the opposite pattern. In the current study, the diagnostic utility of category uency, letter uency, and difference scores in AD is examined further. Whereas most previous studies have relied on group comparisons, we look at the diagnostic usefulness of category uency, letter uency, and difference scores in individuals by using sensitivity and specicity to generate predictive ratios (Fletcher, Fletcher, & Wagner, 1996). METHOD
Subjects were older persons who have been followed in either the Mayo Alzheimer's Disease Patient Registry (ADPR, a research project that enrolls participants from Olmsted County, MN) or the Mayo Alzheimer's Disease Research Center (ADRC, a research project that enrolls participants from the upper Midwest region). ADPR and ADRC participants are enrolled and followed either as incident cases of acquired neurocognitive impairment (i.e., dementia) or as control subjects. The ADPR and ADRC recruitment procedures have been described in detail elsewhere (Petersen, Kokmen, Tangalos, Ivnik, & Kurland, 1990). Cognitively normal participants include independently functioning, community-dwelling persons who have recently been examined by their personal physician and who have no active neurologic or psychiatric disorder with potential to affect cognition (Malec, Ivnik, & Smith, 1993). Mayo ADPR and ADRC participants receive serial neuropsychological evaluations. Diagnosis of cognitive impairment or cognitive normalcy is assigned after every evaluation (i.e., initial and all follow-ups) via consensus meetings that include a certied geriatrician, one or more

Participants

LETTER AND CATEGORY FLUENCY

37

Downloaded By: [HINARI Consortium (T&F)] At: 01:04 10 February 2011

board-certied behavioral neurologists, and one or more board-certied clinical neuropsychologists. Patients with AD included in this study carried a diagnosis of possible or probable AD. AD diagnoses were assigned in accordance with the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association criteria (McKhann et al., 1984). The uency scores used as dependent measures in this study were included in the much larger set of information available during the consensus conferences. However, these test data, obtained at the initial evaluation, did not determine diagnosis which was assigned at the most recent evaluation and included information from serial medical, laboratory, and imaging evaluations, a number of cognitive and adjustment measures, and family and patient reports.

Procedure Cognitive tests considered in this report include the Wechsler Adult Intelligence Scale Revised (WAISR; Wechsler, 1981), the Wechsler Memory Scale Revised (Wechsler, 1987), the Auditory Verbal Learning Test (AVLT; Rey, 1964), the Dementia Rating Scale (DRS; Mattis, 1988), the Controlled Oral Word Association Test (COWAT [CFL version]; Benton & Hamsher, 1978; Borkowski, Benton, & Spreen, 1967), and a category uency task (Lucas et al., 1998). WAISR tasks were limited to those required for computing the Mayo Cognitive Factor Scales (MCFS; Smith et al., 1994). Standard WAISR, AVLT, DRS, and COWAT administration procedures were followed. For the category uency task, participants generated as many animal names as possible in 60 s. In two subsequent trials, participants generated the names of fruits and vegetables, in that order. The number of correct responses for each of the three trials was summed to obtain the total category uency score. All tests were administered by experienced psychometrists supervised by two ABPP-certied clinical neuropsychologists (Drs. Ivnik and Smith). Statistical Analyses First, the AD and control groups were compared on demographic variables, and DRS and MCFS indices. Then, traditional null hypothesis signicance tests (NHST; i.e., t-test comparisons) were conducted, comparing the AD and control groups on category uency and letter uency scores as well as letter uency minus category uency difference scores. The same data were then analyzed with more specic reference to the diagnostic utility of these

measures in clinical prediction of AD in individual patients (Ivnik et al., 2000). To accomplish this, rst cut scores were identied for the distributions of category uency, letter uency, and difference scores. Cut scores were selected that achieved the highest `hit rate' when used to classify subjects as either normal or impaired. In other words, this is the value that rendered the highest total combined sensitivity and specicity (correctly identifying impaired persons as impaired while also classifying cognitively normal persons as normal). Using this information, odds ratios associated with each cut score, along with the corresponding 95% condence intervals, were computed that reect the degree of diagnostic utility for each measure (Knapp & Miller, 1992). Odds ratios allow for statements about the likelihood that a person with a score equal to or below a specied cutoff is impaired relative to the likelihood that a person with a score above the cutoff is impaired (i.e., how much greater is the chance that the patient is impaired if they score poorly than if they score well). Clinically, however, we have a need for statements about the likelihood that a patient is impaired given their specic score, relative to the likelihood that the same patient is not impaired (i.e., how much greater is the chance that patient with score x is impaired vs. not impaired). This type of statement is allowed by using likelihood ratios to determine predictive values (Fletcher et al., 1996). Likelihood ratios allow a more individualized estimate of the likelihood that a particular score obtained by an individual subject represents cognitive impairment. Likelihood ratios provide important diagnostic information about an individual's test score. Likelihood ratios consider sensitivity and specicity (Fletcher et al., 1996), and are relatively stable across samples when the inclusion criteria and aspects of the condition being diagnosed (including its severity) are similar. Most importantly, likelihood ratios can be transformed into posttest odds (or positive predictive values) which consider the sensitivity and specicity information contained in the likelihood ratio and the base-rate (pretest odds) of the disorder in the population from which the individual was drawn (see Sackett, Haynes, Guyatt, & Tugwell, 1991). The resulting posttest odds can address the important question facing the clinician, ``what is the likelihood that this patient with score X has the condition of interest?'' (Smith, Cerhan, & Ivnik, in press). This relationship is summarized in the formula:

posttest odds pretest odds likelihood ratio.

38 RESULTS

JANE H. CERHAN ET AL.

Downloaded By: [HINARI Consortium (T&F)] At: 01:04 10 February 2011

Demographic characteristics of the AD patients and normal control subjects are presented in Table 1. The groups did not differ on age or gender ratio. The control group had a slightly higher average education (13.7 years compared to 12.4 years for the patients). However, when the results of this study were analyzed using education-corrected T-scores for category uency and letter uency, the results were the same (in terms of p values) as when using raw scores. Therefore, raw score analyses are presented here to facilitate interpretation and practical application by the reader. DRS and MCFS scores are also presented, demonstrating, as expected, that the AD patients were signicantly globally impaired relative to the normal control subjects.

Traditional NHST results obtained by comparing the groups on category uency and letter uency raw scores and difference scores are presented in Table 2. All of the differences are highly signicant. As a group, the AD patients performed more poorly on letter uency and category uency than the normal control subjects. The mean difference score (raw letter uency minus category uency) was negative in the normal control subjects indicating that, on average, control subjects generated more words on the category uency than on the letter uency task. In contrast, the AD group generated more words on letter uency and produced a positive mean difference score. This nding is consistent with the majority of previous reports in which AD patients do more poorly on category uency than letter uency.

Table 1. Subject Characteristics: Comparing Alzheimer's Patients and Normal Control Subjects. Control subjects n Age Education Gender ratio (M:F) DRS MCFSVC MCFSPO MCFSAC MCFSLRN MCFSRET 221 221 91:130 171 219 179 207 200 197 M 76.1 13.7 134.8 106.3 105.7 104.1 101.2 106.9 SD 7.0 3.0 6.0 12.5 12.4 13.7 12.8 13.3 Alzheimer's patients n 40 40 17:23 40 40 31 34 31 39 M 77.3 12.4 105.0 85.3 78.7 84.1 78.0 72.8 SD 8 2.7 17.8 14.4 16.2 17.6 12.2 12.0 t 0.95 2.43 X 2 .024 18.08 9.52 10.65 7.76 9.45 14.82 p ns ` .05 ns ` .0001 ` .0001 ` .0001 ` .0001 ` .0001 ` .0001

Note. DRS Mattis Dementia Rating Scale total scores; MCFSVC Mayo Cognitive Factor Scales (MCFS) Verbal Comprehension factor score; MCFSPO MCFS Perceptual Organization factor score; MCFSAC MCFS Attention/Concentration factor score; MCFSLRN MCFS Learning factor score; MCFSRET MCFS Retention factor score.

Table 2. Group Comparisons of Fluency Scores in Alzheimer's Patients and Control Subjects Using NHST. Control subjects Measure LF CF DIFF (LF CF) M 34.01 40.55 6.54 SD 11.31 10.71 11.52 Alzheimer's patients M 21.05 19.27 1.77 SD 15.94 8.92 12.35 t 4.9 13.43 3.94 p .00001 ` .00001 .00024

Note. LF letter uency (Controlled Oral Word Association) total score; CF category uency total score; NHST Null Hypothesis Signicance Testing.

LETTER AND CATEGORY FLUENCY

39

Table 3. Diagnostic Accuracy of Letter Fluency (LF) and Category Fluency (CF) Using Odds Ratios. Measure Cut score Sensitivity Specicity Best hit rate 75% 90.6% Odds ratio 7.9 82.6 ns Lower bound of 95% CI 2.1 24.5

LF CF DIFF (LF CF)

25 28

72.5% 92.5%

77.8% 88.7%

Downloaded By: [HINARI Consortium (T&F)] At: 01:04 10 February 2011

Analysis of the same data using odds ratios is presented in Table 3. Odds ratios are presented associated with each cut score, along with the low end of the corresponding 95% condence interval. As described above, the odds ratios reect the degree of diagnostic accuracy for each measure. For example, using the lower end of the 95% condence interval for letter uency (COWAT) would allow for the following interpretation: We can conclude with 95% condence that any person who is demographically similar to persons in this study and who earns a score at or below 25 on COWAT is 2.1 times more likely to be diagnosed with AD than a person with a score above 25. For category uency, the lower end of the 95% condence interval for the odds ratio is 24.5, demonstrating markedly greater usefulness for diagnostic classication in this sample. The odds ratio for difference scores was not significant. No cut score could be identied that would prove useful for discriminating impaired from normal subjects.

Table 4 shows the likelihood ratios for scores on category uency generated in these samples. Using likelihood ratios allows a more precise estimate of the chances that a person's score represents cognitive impairment. In this case, likelihood ratios reect the odds of each test score in persons with AD as opposed to the odds of that score occurring in a person without AD (Sackett et al., 1991). For example, in a group demographically similar to persons in this sample, a score of 23 is 34 times more likely to occur in the presence of AD than in the absence of AD. The same information, for letter uency, is shown in Table 5. The likelihood ratio values in Tables 4 and 5 are conservative. They are whole integers that were determined by truncating, not by rounding, decimal places from the computed likelihood ratio values. When used with estimates of pre-test odds (prevalence), these likelihood ratios can be used to determine the odds of AD in individual patients (post-test odds pretest odds likelihood ratio).

Table 4. Likelihood Ratios for the Distribution of Category Fluency (CF) Scores in Alzheimer's Patients and Control Subjects. Number of subjects in each score range Total category uency score Below 20 2021 22 23 24 2526 2728 2930 3135 Above 35 Control subjects n (%) 0 (0.0) 1 (0.5) 2 (0.9) 2 (0.9) 4 (1.8) 9 (4.0) 7 (3.2) 9 (4.0) 41 (18.6) 146 (66.0) Alzheimer's patients n (%) 20 8 2 1 0 2 4 0 0 3 (50.0) (20.0) (5.0) (2.5) (0.0) (5.0) (10.0) (0.0) (0.0) (7.5) Likelihood ratio b 127 127 55 34 19 10 8 6 2 `1

40

JANE H. CERHAN ET AL.

Table 5. Likelihood Ratios for the Distribution of Letter Fluency (LF) Scores in Alzheimer's Patients and Control Subjects. No. of subjects in each score range Total letter uency score 04 510 1115 16 17 18 19 2022 2325 2628 Above 28 Control subjects n (%) 0 (0.0) 2 (0.9) 6 (2.7) 1 (0.5) 3 (1.4) 5 (2.3) 3 (1.4) 14 (6.3) 20 (9.1) 23 (10.4) 144 (65.2) Alzheimer's patients n (%) 5 (12.5) 7 (17.5) 5 (12.5) 0 (0.0) 1 (2.5) 1 (2.5) 0 (0.0) 6 (15.0) 4 (10.0) 1 (2.5) 10 (25.0) Likelihood ratio b 33 33 11 9 8 6 5 4 3 2 `1

Downloaded By: [HINARI Consortium (T&F)] At: 01:04 10 February 2011

DISCUSSION This study examined the diagnostic utility of category uency, letter uency, and letter uency minus category uency difference scores in AD. In our sample, AD patients differed very significantly from normal control subjects on all three measures. As a group, the patients scored much lower on letter uency and category uency than the normal control subjects. And, using NHST comparison, the mean difference scores (raw letter uency minus category uency) also differed signicantly. In fact, the average letter uency minus category uency difference score was positive for the AD patients and negative for the control subjects, consistent with previous authors' reports of a relatively greater decit in category uency in AD (Crossley et al., 1997; Monsch et al., 1994; Salmon et al., 1999). These NHST results establish that each measure is sensitive to the adverse effects of AD on uency, but they do not address the diagnostic usefulness of these measures for detecting AD in individuals. A complimentary way to analyze the same data is to consider each measure's diagnostic sensitivity and specicity. In this study, odds ratios and likelihood ratios, both ways of reecting combined sensitivity and specicity, were used. This method may be more relevant to clinical practice than NHST because the analyses occur at the level of

individuals (not groups) and more directly assess the utility of category uency and letter uency in diagnosing AD in patients. The odds ratios demonstrated that category uency's ability to classify AD patients and normal control subjects was dramatically better than that of letter uency in this sample. Specically, it was demonstrated that a person scoring at or below the cut score on letter uency was 2.1 times more likely to be diagnosed with AD than was a person scoring above the cutoff. On the other hand, for category uency, a person at or below the cutoff was at least 24.5 times more likely to have AD. This marked difference between category uency and letter uency is difcult to glean from the traditional NHST analysis, but is clearly demonstrated using odds ratios. This nding of category uency's greater sensitivity to AD is consistent with previous research (Crossley et al., 1997; Monsch et al., 1994) and could arguably lend further support to the temporal-lobe semantic-degradation hypothesis (Butters et al., 1987; Salmon et al., 1999). Interestingly, no diagnostically useful cut score was identied for letter uency minus category uency difference scores, so odds ratios for the difference scores were not calculated. Although NHST results showed that the mean letter uency minus category uency difference scores were signicantly greater in the AD

LETTER AND CATEGORY FLUENCY

41

patients than the control subjects, no specic cut score could be reliably used for classication. This non-intuitive nding is due to the signicant overlap that exists between the AD and control groups' difference-score distributions. Sherman and Massman (1999) suggested that a similar nding in their study related to the heterogeneity of pathology location and progression in AD. Using Tables 4 and 5, clinicians can determine likelihood ratios for specic scores obtained from individual patients under similar demographic conditions. The likelihood ratios reect the odds of each test score occurring in persons with AD over the odds of that score occurring in a person without AD. Using likelihood ratios allows us to take into account the greater predictive power of test scores at the extreme ends of a distribution. Likelihood ratios reect the sensitivity and specicity of a test in a given sample. In actual practice, the diagnostic value of a test will also depend on the prevalence of disease in the population being tested. So, for a given test score, the actual odds of disease are equal to the likelihood ratio corresponding to that score times pretest odds (prevalence) in the population from which the subject was drawn (Fletcher et al., 1996). In this study, the prevalence of disease was 15%, which might be considerably lower than a clinical neuropsychological testing practice. The expectation, therefore, would be that the predictive utility of the category uency and letter uency measures would be somewhat higher in the typical clinical neuropsychological setting than in this study, and somewhat lower in the general population than in this study. In actual neuropsychology practice, many clinicians use both category uency and letter uency, along with other measures, as part of comprehensive assessment of their patients. This approach will naturally generate more powerful prediction than any single test score. Further research into the cognitive processes underlying these tasks, and into their performance in and ability to differentiate between different disease processes, will help in the continual effort to rene and streamline assessment and make diagnostic interpretation more accurate.

ACKNOWLEDGMENTS This study was supported by the people of Rochester, Minnesota, as well as by the National Institute on Aging funded Mayo Alzheimer's Disease Center (AG08031) and Mayo Alzheimer's Disease Patient Registry (AG06786). We gratefully acknowledge the diligence of the Mayo Clinic's Alzheimer's Disease Center staff and their professionalism in test administration, data collection and data management. REFERENCES
Barr, A., & Brandt, J. (1996). Word-list generation decits in dementia. Journal of Clinical and Experimental Neuropsychology, 18, 810822. Benton, A.L., & Hamsher, K.deS. (1978). Multilingual aphasia examination manual. Iowa City, IA: University of Iowa. Borkowski, J.G., Benton, A.L., & Spreen, O. (1967). Word uency and brain damage. Neuropsychologia, 5, 135140. Butters, N., Granholm, E.L., Salmon, D.P., Grant, I., & Wolfe, J. (1987). Episodic and semantic memory: A comparison of amnesic and demented patients. Journal of Clinical and Experimental Neuropsychology, 9, 479497. Crossley, M., D'Arcy, C., & Rawson, N.S.B. (1997). Letter and category uency in community-dwelling Canadian seniors: A comparison of normal participants to those with dementia of the Alzheimer or vascular type. Journal of Clinical and Experimental Neuropsychology, 19, 5262. Fletcher, R.H., Fletcher, S.W., & Wagner, E.H. (1996). Clinical epidemiology: The essentials. Baltimore: Williams & Wilkins. Hart, S., Smith, C.M., & Swash, M. (1988). Word uency in patients with early dementia of the Alzheimer type. British Journal of Clinical Psychology, 27, 115124. Ivnik, R.J., Smith, G.E., Petersen, R.C., Boeve, B.F., Kokmen, E., & Tangalos, E.G. (2000). Diagnostic accuracy of four approaches to interpreting neuropsychological test data. Neuropsychology, 14, 163177. Keilp, J.G., Gorlyn, M., Alexander, G.E., Stern, Y., & Prohovnik, I. (1999). Cerebral blood ow patterns underlying the differential impairment in category vs letter uency in Alzheimer's disease. Neuropsychologia, 37, 12511261. Knapp, R.G., & Miller, M.C. (1992). Clinical epidemiology and biostatistics. Baltimore, MD: Williams & Wilkins.

Downloaded By: [HINARI Consortium (T&F)] At: 01:04 10 February 2011

42

JANE H. CERHAN ET AL.

Lucas, J.A., Ivnik, R.J., Smith, G.E., Bohac, D.L., Tangalos, E.G., Graff-Radford, N.R., & Petersen, R.C. (1998). Mayo's Older Americans Normative Studies: Category uency norms. Journal of Clinical and Experimental Neuropsychology, 20, 194200. Malec, J.F., Ivnik, R.J., & Smith, G.E. (1993). Neuropsychology and normal aging. In R.F.Z.R.W. Parks & R.S. Wilson (Eds.), Neuropsychology of Alzheimer's disease and other dementias (pp. 81 111). New York: Oxford University Press. Martin, A., & Fedio, P. (1983). Word production and comprehension in Alzheimer's disease: The breakdown of semantic knowledge. Brain and Language, 19, 124141. Martin, A., Wiggs, C.L., LaLonde, F., & Mack, C. (1994). Word retrieval to letter and semantic cues: A double dissociation in normal subjects using interference tasks. Neuropsychologia, 32, 14871494. Mattis, S. (1973). Mattis Dementia Rating Scale (MDRS). Odessa, FL: Psychological Assessment Resources. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., & Stadlan, E.M. (1984). 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, 34, 939 944. Monsch, A.U., Bondi, M.W., Butters, N., Paulsen, J.S., Salmon, D.P., Brugger, P., & Swenson, M.R. (1994). A comparison of category and letter uency in Alzheimer's disease and Huntington's disease. Neuropsychology, 8, 2530. Ober, B.A. (1999). Symposium: Semantic memory in Alzheimer's disease: Loss of knowledge or decits in retrieval? Journal of the International Neuropsychological Society, 5, 623625. Petersen, R.C., Kokmen, E., Tangalos, E.G., Ivnik, R.J., & Kurland, L.T. (1990). Mayo Clinic's Alzheimer's Disease Patient Registry. Aging, 2, 408415. Randolph, C., Braun, A.R., Goldberg, T.E., & Chase, T.N. (1993). Semantic uency in Alzheimer's, Parkinson's, and Huntington's disease: Dissociation of storage and retrieval failures. Neuropsychology, 7, 8288.

Rey, A. (1964). L'examen clinique en psychologie. Paris: Presses universitaires de France. Rosser, A., & Hodges, J.R. (1994). Initial letter and semantic category uency in Alzheimer's disease, Huntington's disease, and Progressive Supranuclear Palsy. Journal of Neurology, Neurosurgery, and Psychiatry, 57, 13891394. Sackett, D.L., Haynes, R.B., Guyatt, G.H., & Tugwell, P. (1991). Clinical epidemiology: A basic science for clinical medicine (2nd ed.). Boston: Little, Brown & Co. Salmon, D.P., Heindel, W.C., & Lange, K.L. (1999). Differential decline in word generation from phonemic and semantic categories during the course of Alzheimer's disease: Implications for the integrity of semantic memory. Journal of the International Neuropsychological Society, 5, 692703. Sherman, A.M., & Massman, P.J. (1999). Prevalence and correlates of category versus letter uency discrepancies in Alzheimer's disease. Archives of Clinical Neuropsychology, 14, 411418. Smith, G.E., Cerhan, J.H., & Ivnik, R.J. (in press). Diagnostic Validity. In D.S. Tulsky & D. Saaklofske (Eds.), Clinical interpretation of the WAIS-III and WMS-III. Academic Press. Smith, G.E., Ivnik, R.J., Malec, J.F., Petersen, R.C., Kokmen, E., & Tangalos, E.G. (1994). The Mayo Cognitive Factor Scales (MCFS): Derivation of a short battery and norms for factor scores. Neuropsychology, 8, 194202. Suhr, J.A., & Jones, R.D. (1998). Letter and semantic uency in Alzheimer's, Huntington's, and Parkinson's dementias. Archives of Clinical Neuropsychology, 13, 447454. Wechsler, D. (1981). Wechsler Adult Intelligence Scale Revised manual. San Antonio, TX: The Psychological Corporation. Wechsler, D. (1987). Wechsler Memory Scale Revised manual. San Antonio, TX: The Psychological Corporation. Zec, R.F., Andrise, A., Vicari, S., Feldman, E., Belman, J., Landreth, E., & Markwell, S. (1990). A comparison of phonemic and semantic word uency in Alzheimer patients and elderly controls. Journal of Clinical and Experimental Neuropsychology, 12, 18.

Downloaded By: [HINARI Consortium (T&F)] At: 01:04 10 February 2011