An Empirically Derived Short Form of the Boston Naming Test

Archives of Clinical Neuropsychology, Vol. 14, No. 6, pp. 481–487, 1999 Copyright © 1999 National Academy of Neuropsychology Printed in the USA. All rights reserved 0887-6177/99 $–see front matter

PII S0887-6177(98)00022-5

An Empirically Derived Short Form of the Boston Naming Test Amy E. Lansing Northwestern University Medical School

Robert J. Ivnik Mayo Clinic

C. Munro Cullum University of Texas Southwestern Medical Center

Christopher Randolph Loyola University Medical Center

The Boston Naming Test (BNT) is one of the most commonly used tests of confrontation naming. The length of the test, particularly when administered to impaired patients, has prompted the derivation of several abbreviated forms. Short forms of the BNT have typically been equated in terms of difficulty, but not empirically derived for discriminating between normals and anomic patients. Furthermore, most reports to date have been limited in sample size and generalizability. The present study examined BNT data from a total of 1,044 subjects, including 719 normals and 325 patients with Alzheimer’s disease (AD). Scores were calculated for the entire 60-item version as well as for eight previously reported short forms. The scores were examined for the effects of age, education, and gender, as well as for the ability of each form to discriminate between AD patients and normals. There was a significant effect of age, education, and gender on all previously published forms, and the short forms varied in their ability to discriminate between patients and controls. A stepwise discriminant analysis was conducted to empirically derive a new, gender-neutral short form with discriminability comparable to the full 60-item test. Norms from this sample on the empirically derived short form are reported. © 1999 National Academy of Neuropsychology. Published by Elsevier Science Ltd

One of the most commonly used tests for the assessment of confrontation naming is the Boston Naming Test (BNT; Kaplan, Goodglass, & Weintraub, 1983). The BNT was derived from an earlier experimental version which consisted of 85 line drawings (Borod, Goodglass, & Kaplan, 1980; Kaplan, Goodglass, & Weintraub, 1976). The current test

Address correspondence to Christopher Randolph, 111 North Wabash, Suite 1021, Chicago, IL 60602; E-mail: [email protected]

481

482

A. E. Lansing et al.

consists of 60 line drawings selected from the 85-item test to represent a continuum of difficulty. The test was originally normed on children, and on 84 adult subjects ranging in age from 18 to 59 years (Kaplan et al., 1983). Performance has been reported to be affected by normal aging (Albert, Heller, & Milberg, 1988), and additional normative data has since been published for a broader age range (Halperin, Healey, Zeitchik, Ludman, Weinstein, 1989; Ivnik, Malec, Smith, Tangalos, & Peterson, 1996; Lichtenberg, Ross, & Christensen, 1994; Van Gorp, Satz, Kiersch, & Henry, 1986). Individual sample sizes have been limited, however; and the effects of age, education, and gender on performance have been inconsistently addressed (see Randolph, Lansing, Ivnik, Cullum, & Hermann, 1999 for a more complete review of these issues). The BNT is designed to allow full credit for uncued naming responses as well as for the correct naming of drawings in response to stimulus (semantic) cues used when patients are apparently misperceiving the target item. Phonemic cuing is also utilized, but correct responses at this level do not contribute to the overall score. Because of the number of items included in the test, together with the two-stage cuing format, the test can be time-consuming to administer to impaired patients. To address the issue of administration time as well as to provide equivalent, alternate forms of the BNT for clinical and research purposes, a variety of short forms have been proposed. There are currently three 30-item and five 15-item short form tests, derived from the 60-item test, designed to provide equivalent and/or expedient assessments (Williams, Mack, & Henderson, 1989; Mack, Freed, Williams, & Henderson, 1992; Morris, Mohs, Rogers, Fillenbaum, & Heyman, 1988). These short forms have been designed to expedite administration, and with the exception of one 30-item form (Williams et al., 1989), they have not been empirically derived to differentiate between anomics and normals. Each of the nonempirical tests was derived by retaining alternate items, staggering item assignment or selecting items within specified word frequency categories in order to achieve equivalent word frequency values within and between tests. The 30-item empirical form was developed by examining differences between group means on individual items for normal controls and patients with Alzheimer disease (Williams et al., 1989). This was based on a limited sample of 55 subjects, the number of items included was chosen arbitrarily and discriminability was not specifically tested. Mack et al. (1992) assessed the utility of the five 15-item forms and the three 30-item forms, and reported equivalency across forms and short form score extrapolation to the 60-item score for all but the Consortium to Establish a Registry for Alzheimer’s Disease (Morris et al., 1988) short form test. Franzen, Haut, Rankin, and Keefover (1995) examined internal consistency and correlations among these forms, concluding that internal consistency was adequate for all forms and correlations between forms were reasonable. These authors did caution that classification rates may vary among forms, possibly limiting their clinical utility. Although preliminary findings on a sample of 52 subjects did not indicate a significant relationship between education and performance on any version of the BNT (Mack et al., 1992), the influence of demographic variables (age, education, gender) on short form performance has not been thoroughly investigated. Although the equivalent 15-item and 30-item short forms are useful for test-retest reliability specifically in a research setting, a clinically useful short form should: 1. Be empirically derived from a sufficiently large sample of normals and anomic patients. 2. Consist of as few items as possible to expedite administration while maximally discriminating among these groups. 3. Include norms that correct for all relevant demographic influences, for the purpose of clinical interpretation.

BNT Short Form

483

The current study was designed to examine the utility of previously developed short forms by examining data from the BNT on a large number of normal controls (NC) and AD patients and to empirically derive a set of items that optimally discriminates AD patients from NC subjects.

METHOD BNT data from 719 normal controls and 325 patients with a diagnosis of possible or probable AD made on the basis of NINCDS-ADRDA criteria (McKhann et al., 1984) were collected from three academic medical center sites. The normal controls consisted of both paid and unpaid volunteers for studies on neuropsychological functions in normal aging. Procedures for recruiting and further neuropsychological description of the majority of the normative sample is detailed in previous papers (e.g., Cullum, Thompson, & Smernoff, 1993; Ivnik et al., 1992; Malec, Ivnik & Smith, 1993; Randolph et al., 1994). Age, gender, years of education, and diagnosis were coded for all subjects, as was their score on each item of the 60-item BNT. An item was scored as correct if the correct name was given spontaneously or after the stimulus (semantic) cue. For all subjects except 76 AD patients at one site, the test was discontinued after 6 sequential failures. For the 76 AD subjects who received the full 60-item test, there was an average of 2.0 (SD 5 2.5; range 0–10) correctly named responses after failing 6 consecutive items. Because the conclusions from this study are based on individual item performance, we felt we could not be confident that 6 sequential failures was absolutely predictive of failure on all subsequent items. For this reason, we only analyzed the data from those subjects who actu-

TABLE 1 Effects of Demographic Variables on Boston Naming Test (BNT) Forms for the Normal Control Sample

BNT Version

60-item original BNT (Kaplan et al., 1983) 30-item BNT Odd Version (Williams et al., 1989) 30-item BNT Even Version (Williams et al., 1989) 30-item BNT Empirical Version (Williams et al., 1989) 15-item BNT CERAD version (Morris et al., 1988) 15-item BNT Equivalent Form 1 (Mack et al., 1992) 15-item BNT Equivalent Form 2 (Mack et al., 1992) 15-item BNT Equivalent Form 3 (Mack et al., 1992) 15-item BNT Equivalent Form 4 (Mack et al., 1992) 15-item BNT Empirical Version (present study)

Correlation with Age

Correlation with Education

Correlation with Gender

r 5 2.45 (p , .0001) r 5 2.46 (p , .0001) r 5 2.41 (p , .0001) r 5 2.43 (p , .0001) r 5 2.40 (p , .0001) r 5 2.37 (p , .0001) r 5 2.39 (p , .0001) r 5 2.46 (p , .0001) r 5 2.37 (p , .0001) r 5 2.34 (p , .0001)

r 5 .24 (p , .0001) r 5 .22 (p , .0001) r 5 .25 (p , .0001) r 5 .28 (p , .0001) r 5 .09 (p , .0001) r 5 .09 (p , .0001) r 5 .31 (p , .0001) r 5 .16 (p , .0001) r 5 .28 (p , .0001) r 5 .21 (p , .0001)

r 5 .21 (p , .0001) r 5 .21 (p , .0001) r 5 .18 (p , .0001) r 5 .20 (p , .0001) r 5 .16 (p , .0001) r 5 .14 (p , .0002) r 5 .22 (p , .0001) r 5 .26 (p , .0001) r 5 .11 (p , .002) r 5 .00 NS

Note. CERAD 5 Consortium to Establish a Registry for Alzheimer’s Disease.

484

A. E. Lansing et al. TABLE 2 Means and Standard Deviations for Scores (# Correct) on Boston Naming Test (BNT) Forms for Both Groups Alzheimer’s Patients

Normal Controls BNT Version

60-item original BNT (Kaplan et al., 1983) 30-item BNT Odd version (Williams et al., 1989) 30-item BNT Even version (Williams et al., 1989) 30-item BNT Empirical Version (Williams et al., 1989) 15-item BNT CERAD version (Morris et al., 1988) 15-item BNT Equivalent Form 1 (Mack et al., 1992) 15-item BNT Equivalent Form 2 (Mack et al., 1992) 15-item BNT Equivalent Form 3 (Mack et al., 1992) 15-item BNT Equivalent Form 4 (Mack et al., 1992) 15-item BNT Empirical Version (present study)

M

SD

M

SD

52.4

6.3

39.9

14.3

25.4

3.1

19.1

5.9

26.0

3.4

20.2

6.0

24.6

4.5

16.5

7.4

14.1

1.2

12.0

2.7

13.5

1.9

10.3

3.2

13.1

1.6

10.2

3.1

12.7

1.8

9.8

3.1

13.1

2.1

9.7

3.3

13.4

1.8

9.5

3.6

Note. CERAD 5 Consortium to Establish a Registry for Alzheimer’s Disease.

ally were administered all 60 items. This left us with 717 NC subjects and 237 AD subjects. The AD patients ranged in age from 46 to 97, with a mean age of 76.0 (SD 5 9.7), mean education of 13.2 years (SD 5 3.6), and were 60% female. The NC subjects ranged in age from 50 to 98, had a mean age of 73.6 (SD 5 10.3), mean education of 13.4 years (SD 5 2.9), and were 60% female. Both groups were almost entirely (.90%) White.

RESULTS The effects of gender, age, and education were examined for the full 60-item and 8 previously published short forms, by diagnostic group (see Table 1). For the normals, there was a significant effect of gender on all forms of the test, with men performing at a higher level than women (approximate difference of .5 SD; e.g., 2.7 points difference on the full 60-item test for normals). Statistically significant gender differences persisted on all forms even after controlling for the effects of age and education through analysis of covariance. Age and education were significantly correlated with performance on all forms in the normal group. Demographic effects were similar for the AD group (see Randolph et al., 1999 for a more complete review of demographic effects by diagnostic groups). Table 2 contains means and standard deviations of scores on the various forms, by diagnostic group. A stepwise discriminant analysis on all 60 items was run between the NC and AD groups. Twenty-two items entered in with an alpha for entry set at .15. After eliminating several gender-biased items to achieve gender equity in the NC group, a new 15-item

BNT Short Form

485

TABLE 3 Results of Discriminant Analyses for the Full 60-Item Boston Naming Test (BNT) and BNT Short Form Correctly Classified Subjects

BNT Version

60-item original BNT (Kaplan et al., 1983) 15-item BNT Empirical Version (present study) 30-item BNT Odd version (Williams et al., 1989) 30-item BNT Even Version (Williams et al., 1989) 30-item BNT Empirical version (Williams et al., 1989) 15-item BNT CERAD version (Morris et al., 1988) 15-item BNT Equivalent Form 1 (Mack et al., 1992) 15-item BNT Equivalent Form 2 (Mack et al., 1992) 15-item BNT Equivalent Form 3 (Mack et al., 1992) 15-item BNT Equivalent Form 4 (Mack et al., 1992)

% Alzheimer Patients

% Normal Older Adults

67.5

84.4

70.9

84.8

68.8

84.1

68.8

80.1

67.9

83.5

68.4

76.9

58.2

87.0

62.5

84.9

68.8

81.6

67.1

80.9

Note. CERAD 5 Consortium to Establish a Registry for Alzheimer’s Disease.

empirical version was compared to existing forms for discriminability. Discriminability varied somewhat among the short forms (See Table 3). The empirically derived 15-item short form was comparable to the full 60-item form in terms of overall discriminability. Table 4 includes the normative data for the new, empirically derived 15-item version of the BNT, on which there is no statistically significant difference in performance between the males and females in this sample. The data were broken down by age groups using the technique of overlapping midpoints, to conform to the Mayo Older American Normative Studies tables (e.g., Ivnik et al., 1992), for comparative purposes. After ensuring that the raw score distributions for each midpoint age group were normal, we elected to present the data as means and standard deviations. Education effects were controlled for by adding 1 point to the score of all individuals with less than 12 years of education; after this correction, there was no statistically significant difference among the groups of normals with less than 12 (adjusted M 5 13.5), exactly 12 (M 5 13.3), or more than 12 years of education (M 5 13.6): F(2, 714) 5 2.4, ns. Table 5 contains a list of the 15 items retained on the empirically derived short form, presented in order of entry into the discriminant function.

DISCUSSION To our knowledge, this is the largest normative sample as well as the largest AD sample of data ever reported on the BNT. The utility of previously formulated short forms

486

A. E. Lansing et al. TABLE 4 Normative Data for the New Empircally Derived 15-Item Version of the Boston Naming Test Normal Controls Midpoint Age (years)

55 58 61 64 67 70 73 76 79 82 85 88

Age Range (years)

N

M

SD

50–60 53–63 56–66 59–69 62–72 65–75 68–78 71–81 74–84 77–87 80–90 831

98 147 183 202 201 211 226 251 273 264 222 156

14.1 14.1 14.1 14.1 14.0 13.9 13.7 13.5 13.2 13.0 12.9 12.7

1.2 1.1 1.1 1.0 1.2 1.4 1.5 1.8 1.9 2.1 2.2 2.3

Note. Scores adjusted for education by adding 1 point to the total score of all individuals with ,12 years of education completed.

in distinguishing these groups was examined. A new, empirically derived short form is presented, along with associated normative data for a large sample aged 50 to 98, broken down by age, with a correction for education. All previous forms of the test were biased with respect to gender, with males outperforming females as a result of gender differences in naming of specific items (see Randolph et al., 1999 for further discussion). The newly derived 15-item short form adequately discriminates between normal controls and AD patients, without gender bias. For clinicians and researchers who are interested in using a single short form of the BNT, the present data should prove useful, allowing relatively rapid detection of clinically significant anomia without requiring separate normative data by gender group.

TABLE 5 Items in the Empirical Derived 15-Item Boston Naming Test Short Form, Listed in Order of Entry into Discriminant Function Item Number

37 38 35 53 39 47 27 22 32 24 49 14 21 58 57

Item Description

Escalator Harp Dominoes Scroll Hammock Accordion Globe Snail Acorn Seahorse Asparagus Mushroom Racquet Palette Trellis

BNT Short Form

487

It is important to note that the data presented in this study were derived from the administration of the entire original 60-item BNT, and not by separate administration of each short form. The scores for the various short forms were computed from the 60-item test, and it is always possible that this technique for deriving a short form may differ somewhat from actual administration out of the context of the 60-item administration. Given the nature of the task, however, this seems unlikely. The current study is also limited by the fact that the normative sample was almost entirely White sample, fairly well educated, and relatively homogeneous. Caution should be exercised in the interpretation of BNT test performance of patients who are demographically dissimilar from this sample.

REFERENCES Albert, M. S., Heller, H. S., & Milberg, W. (1988). Changes in naming ability with age. Psychology and Aging, 3, 173–178. Borod, J. C., Goodglass, H., & Kaplan, E. (1980). Normative data on the Boston Aphasia Examination, Parietal Lobe Battery, and the Boston Naming Test. Journal of Clinical Neuropsychology, 2, 209–215. Cullum, C. M., Thompson, L. T., & Smernoff, E. N. (1993). Three-word recall as a measure of memory. Journal of Clinical and Experimental Neuropsychology, 15, 321–329. Franzen, M. D., Haut, M. W., Rankin, E., & Keefover, R. (1995). Empirical comparison of alternate forms of the Boston Naming Test. The Clinical Neuropsychologist, 9, 225–229. Halperin, J. M., Healey, J. M., Zeitchik, E., Ludman, W. L., & Weinstein, L. (1989). Developmental aspects of linguistic and mnestic abilities in normal children. Journal of Clinical and Experimental Neuropsychology, 11, 518–528. Ivnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. G., & Peterson, R. C. (1996). Neuropsychological test norms above age 55: COWAT, BNT, MAE Token, WRAT-R reading, AMNART, Stroop, TMT, & JLO. The Clinical Neuropsychologist, 10, 262–278. Ivnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. G., Peterson, R. C., Kokmen, E., & Kurland, L. T. (1992). Mayo’s older American normative studies: WAIS-R norms for ages 56 to 97. The Clinical Neuropsychologist, 6(Suppl. 1), 1–30. Kaplan, E., Goodglass, H., & Weintraub, S. (1976). Boston Naming Test, Experimental Edition. Boston: Veterans Administration Hospital. Kaplan, E., Goodglass, H., & Weintraub, S. (1983). Boston Naming Test. Philadelphia: Lea & Febiger. Lichtenberg, P. A., Ross, T., & Christensen, B. (1994). Preliminary normative data on the Boston Naming Test for an older urban population. The Clinical Neuropsychologist, 8, 109–111. Mack, W. J., Freed, D. M., Williams, B. W., & Henderson, V. W. (1992). Boston Naming Test: Shortened versions for use with Alzheimer’s disease. Journal of Gerontology, 47, 154–158. Malec, J. F., Ivnik, R. J., & Smith, G. E. (1993). Neuropsychology and normal aging. In R. W. Parks, R. F. Zec, & R. S. Wilson (Eds.), Neuropsychology of Alzheimer’s disease and other dementias (pp. 81–111). New York: Oxford University Press. McKhann, G., Drachmann, P., Folstein, M., Katzman, R., Price, D., & Stadler, E. (1984). Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA work group under the auspices of the Department of Health and Human Services task force on Alzheimer’s disease. Neurology, 34, 939–944. Morris, J. C., Mohs, R. C., Rogers, H., Fillenbaum, G., & Heyman, A. (1988). Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) clinical and neuropsychological assessment of Alzheimer’s disease. Psychopharmacology Bulletin, 24(4), 641–652. Randolph, C., Gold, J. M., Kozora, K., Cullum, C. M., Hermann, B. P., & Wyler, A. R. (1994). Estimating memory function: Disparity of Wechsler Memory Scale-Revised and California Verbal Learning Test indices in clinical and normal samples. The Clinical Neuropsychologist, 8, 99–108. Randolph, C., Lansing, A., Ivnik, R. J., Cullum, C. M., Hermann, B. P. (1999). Determinants of confrontation naming performance. Archives of Clinical Neuropsychology, 14, 489–496. Van Gorp, W. G., Satz, P., Kiersch, M. E., & Henry, R. (1986). Normative data on the Boston Naming Test for a group of normal older adults. Journal of Clinical and Experimental Neuropsychology, 8, 702–705. Williams, B. W., Mack, W., & Henderson, V. W. (1989). Boston Naming Test in Alzheimer’s disease. Neuropsychologia, 27(8), 1073–1079.