Duangkamol Singwicha, M.D.1, Chatchawan Rattanabannakit, M.D.2, Saowalak Hunnangkul2, Atthapon Raksthaput2, Sunisa Chaichanettee2, Natthamon Wongkom2, Pathitta Dujada2, Vorapun Senanarong, M.D.2,*
1Division of Neurology, Department of Medicine, Faculty of Medicine Vajira Hospital, Navamindradhiraj University, Bangkok, Thailand, 2Division
of Neurology, Department of Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand.
Corresponding author: Vorapun Senanarong E-mail: Vorapun.sen@mahidol.ac.th
Received 16 July 2024 Revised 10 September 2024 Accepted 13 September 2024 ORCID ID:http://orcid.org/0000-0002-2774-4187 https://doi.org/10.33192/smj.v76i12.270244
All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.
ABSTRACT
INTRODUCTION
Similar to many other countries, the aging society in Thailand has become an important issue. The diseases that come from cognitive degeneration, such as Alzheimer’s dementia, and Parkinson’s, will be increasingly encountered. Such neurodegenerative diseases involve physical symptoms and a decrease in patient cognition. In the course of the diseases, they will impact the cognitive memory domain. Some patients will find themselves having problems in their language abilities, such as effortful speech, word-finding difficulty, or low speech production. These effects will decrease their quality of life.1 Abnormalities in language are found in Alzheimer’s disease, frontotemporal dementia, corticobasal degeneration, and vascular pathology. These conditions can lead to primary progressive aphasia (PPA).2,3 The 3 variants of PPA are semantic variant PPA (svPPA), logopenic variant PPA (lvPPA), and nonfluent variant PPA (nfvPPA). Determining the cause of pathology, physical examination followed by neuropsychological testing is essential.
Semantic knowledge encompasses the general knowledge and features that form the concepts people acquire from their experiences.4 These experiences culminate in memories related to people, places, and the meaning of words. This semantic memory is a component of declarative memory and can be retrieved from the anterior
temporal lobe.5,6 The most common neuropsychological tests to measure semantic memory are naming tests, verbal fluency tests for category and letter fluency, and the Pyramids and Palm Trees (PPT) test.
The PPT test is a nonverbal measurement of cognition and semantic knowledge. The test is available in both word and picture modalities to assess verbal and nonverbal semantic memory, respectively. It was designed by Howard and Patterson in 1992 and standardized in England.7 Since then, it has been used across Europe. Nevertheless, applying the picture version of the test in other countries may have limitations due to cultural and education issues related to the pictures. Several studies have used the PPT test to investigate semantic memory in their population to establish normative data. The Spanish population’s formal education level correlated with the PPT test results.8 Similarly, an investigation of a Chinese population showed that schooling affected test performance.9
In Asian countries, including Thailand, the PPT test has not been used due to cultural issues related to the test pictures. This study aimed (1) to validate a Thai adaptation of the picture version of the PPT test with 2 standard tests—the Boston Naming Test (BNT) and the Verbal Fluency (VF) test—in a dementia cohort and (2) to explore its reliability in the Thai population.
MATERIALS AND METHODS
A cross-sectional study design was used. A total of 30 eligible participants were recruited from the Memory Clinic of Siriraj Hospital, Thailand. Participants were separated into 2 groups. There were 15 healthy subjects (the “non-dementia group”) and 15 subjects with mild to moderate stages of Alzheimer’s disease according to DSM-5 criteria and NIA-AA criteria of Alzheimer’s disease dementia13,14 which requires objective cognitive decline that is severe enough to interfere with activities daily living and decline of neuropsychiatric testing more than 2 standard deviation (the “dementia group”). Different inclusion criteria were applied to each group. In the case of the non-dementia group, the participants were
(1) aged between 50 and 80, (2) did not have underlying diseases related to memory or language problems, and
(3) could communicate in Thai. In the dementia group, the criteria were (1) aged between 50 and 80; (2) a mild to moderate stage of Alzheimer’s disease (defined as a Thai Mental State Examination [TMSE] score of 10–23); and
(3) capable of communicating in Thai. In both groups, the participants’ levels of education were divided into under 12 years and over 12 years of formal education.
The picture version of the PPT test consists of 55 sets of pictures. The first 3 sets are examples used to instruct participants on the test, while the remaining 52 sets are the test stimuli. Each set comprises 3 pictures that are presented on a display board. One, the “stimulus,” is positioned in the upper half of the board; the other 2 pictures (1 “target” and 1 “distractor”) are in the lower half. Participants are requested to select the lower-level picture that they consider to be more relevant to the stimulus. The maximum test score is 52, and the median score for normal cognition is 49.
An expert panel evaluated the content of the picture version of the PPT test used in English-speaking cultures to develop content for a Thai adaptation. Six experts (2 neurologists and 4 psychiatrists) agreed that 25 of the 55 picture sets needed to be replaced by ones more appropriate to Thailand (Supplementary Table A). For instance, it was decided to replace a picture of an Eskimo and igloo with one of a farmer and hut. Similarly, a picture of a shepherd’s cane and sheep was substituted with a picture of a herding stick and buffalo. Additionally, minor changes were made to some pictures, such as using a Thai bus ticket instead of a London bus ticket. Some picture images were also sharpened (eg, one of a star and the moon and another of a pillow with a wrinkle). A pilot study was performed on 6 adults of various ages and education levels. The final, Thai-oriented version was then used in this study.
The proposed Thai adaptation of the PPT test was validated with 2 tests that correlate with semantic knowledge: the VF test (both category and letter fluency) and the BNT. To establish a baseline, the TMSE was first administered to all participants. Those scoring less than 24 on the TMSE were assigned to the dementia group. At the same visit as the TMSE, the dementia group was assessed using the Thai adaptation of the PPT test, the VF test, and the BNT. All tests were administered to the participants by 2 certificated psychiatrist assessors by using papers of pictures. The administration of the 3 tests took 45 to 50 minutes.
Statistical analyses were conducted with PASW Statistics for Windows, version 18.0 (SPSS Inc, Chicago, IL, USA). Continuous data (age) are summarized using descriptive statistics (the mean ± standard deviation). Categorical variables (sex, education, and underlying diseases) are summarized using numbers (percentages). Pearson’s correlation coefficients were used to determine the validity of the Thai adaptation of the PPT test in both healthy and dementia patients, with statistical significance indicated by probability (P) values < 0.05 (P < 0.050). Spearman’s rank correlation coefficient was used to perform concurrent validation12 of the PPT test against the other instruments (TMSE, BNT, and VF). We also used Cronbach’s alpha and intraclass correlation coefficients by using 95% confidence intervals to indicate the internal consistency of the proposed PPT test. A receiver operating characteristic (ROC) curve analysis was performed, with the area under the curve (AUC) indicating the reliability of the PPT test compared with the 3 standard tests. An AUC of 0.7 to 0.8 signified (fair reliability, while 0.8 to 0.9 indicated good reliability and
0.9 to 1.0 denoted excellent reliability.
RESULTS
The 30 participants in the non-dementia and dementia groups had no significant differences in their baseline demographic characteristics (age, sex, education level, and underlying diseases; Table 1). However, the 2 groups had significant differences in their mean scores for all instruments (proposed PPT, TMSE, BNT, and VF; Table 2). The mean scores ± SD of the non-dementia and dementia groups were 46.93 ± 5.38 and 42.43 ± 6.27, respectively (P = 0.044). A Mann–Whitney U test determined that the median PPT test score was 49 for the non-dementia group and 43 for the dementia subjects (P = 0.006; Fig 1).
TABLE 1. Demographic characteristic.
Non-dementia | Dementia | P-value | |
(N=15) | (N=15) | ||
Sex | 0.456 | ||
Male, n (%) | 7(46.7) | 5(33.3) | |
Female, n (%) | 8(53.3) | 10(66.7) | |
Age (mean ± SD) | 65.47±8.58 | 72.0±10.92 | 0.790 |
Education (year), n (%) | 0.713 | ||
≤ 12 | 9(60) | 8(53.3) | |
> 12 Underlying disease, n (%) | 6(40) | 7(46.7) | |
Hypertension | 2(13.3) | 5(33.3) | 0.195 |
Dyslipidemia | 7(46.7) | 10(66.7) | 0.269 |
Diabetes Mellitus | 5(33.3) | 9(60) | 0.143 |
Atrial Fibrillation | 2(13.3) | 5(33.3) | 0.195 |
Chronic Kidney Disease | 0 | 1(6.7) | 0.309 |
Others | 0 | 1(6.7) | 0.309 |
TABLE 2. Comparison between non-dementia and dementia group in all tests.
Non-dementia (N=15) (Mean ± SD) | Dementia (N=15) (Mean ± SD) | P-value | |
Pyramid and palm | 46.93±5.38 | 42.43±6.27 | 0.044 |
TMSE | 28.27±1.28 | 24.33±4.25 | 0.002 |
Boston Naming Test | 27.13±2.94 | 21.13±5.47 | 0.001 |
Verbal Fluency | 20.60±5.73 | 14.00±5.33 | 0.003 |
Diagnosis | Sum PPT | p-valuea | |
Median | (IQR) | ||
Dementia | 43 | (42.0-47.5) | 0.006 |
Non-dementia | 49 | (46.5-50.0) | |
aMann-Whitney U test
Fig 1. Median score of PPT test.
The 2 participant groups were combined to evaluate the correlation of the PPT test with the 3 standard tests for all patient types. Spearman’s rank correlation coefficient was used to compare the PPT test with the other instruments. The PPT test was most correlated with the VF test (r = 0.652; P < 0.001), followed by the BNT (r = 0.641; P < 0.001), and last, the TMSE (r = 0.482;
P = 0.007). Although the PPT test score also correlated with education level (r = 0.491, P = 0.006; Table 3), there was no correlation with age or sex.
Regarding the reliability of the test, an ROC analysis was performed to determine the AUC and make comparisons with the TMSE, BNT, and VF. The results showed that the proposed PPT test had an AUC of 0.85, the highest of the AUC values of the 4 instruments (Fig 2).
For the test’s internal consistency, the intraclass correlation coefficient (ICC 3, k) was 0.86 (95% CI, 0.78- 0.92), and Cronbach’s alpha coefficient was 0.87.
DISCUSSION
Several previous studies have been undertaken in various countries to determine the validity and reliability of the PPT test. Due to some cultural issues, normative data of a healthy population must be obtained before the test can be used with patients with dementia.
In Thailand, many tests are routinely used to evaluate multiple domains of cognition. In the case of the language domain, several tests are employed. The TMSE is a baseline test used to evaluate multiple domains for screening purposes, and it is administered to all patients
TABLE 3. Correlation of PPT with other tests.
Variables Correlation coefficient PPT TMSE BNT Animal Fluency Education
PPT Spearman’s Rhoa 1
Spearman’s Rhoa
p-value
0.482**
0.007
1
NA
p-value NA
BNT Spearman’s Rhoa 0.641** 0.583** 1
Spearman’s Rhoa
p-value
0.652**
<0.001
0.650**
<0.001
0.594**
0.001
1
NA
p-value <0.001 0.001 NA
Education Spearman’s Rhoa 0.491** 0.455* 0.529** 0.585** 1
p-value 0.006 0.012 0.003 0.001 NA
* Correlation is significant at the 0.05 level (2-tailed)., ** Correlation is significant at the 0.01 level (2-tailed).
a Spearman’s rank correlation coefficient
Test | AUC |
Boston Naming Test | 0.071 |
Verbal Fluency | 0.082 |
PPT | 0.085 |
TMSE | 0.080 |
Fig 2. Comparison of AUC of tests.
requiring neuropsychiatric assessment. Two tests that are more focused on language use are the BNT and the VF test (category and letter fluency). With the VF test, patients are given 1 minute to produce as many words as possible that either name animals (category fluency) or start with a given letter (letter fluency). The VF test can therefore be used to assess both language and executive functioning.
This study found that the proposed Thai adaptation of the picture version of the PPT test had a good correlation with other language tests and was highly correlated with the VF test (r = 0.652; P < 0.001). The adaptation also had a high ROC value, and the median scores for the Thai nondementia and dementia subjects were significantly different (P = 0.006). Additionally, the median PPT test score for the Thai patients without dementia was 49, which corresponds with the median score for the picture version used in English-speaking cultures. Furthermore, both the intraclass correlation and the internal consistency results were good to satisfactory (exceeding 0.8). Taken together, these results indicate that the proposed Thai adaptation is suitable for discriminating between dementia and non-dementia.
On the other hand, our investigation found an educational dependency in that participants’ level of education (less than 12 years) was significantly associated with performance on the test. This finding is consistent with Guo and associates9, whose investigation revealed that education level correlated with PPT test results in a Chinese population. Regarding age and sex, our research found no significant differences. Although these findings are consistent with Mehri8, they conflict with the results of a study by Callahan and colleagues10 on a Quebec-French population.
A limitation of this study is the size of the sample population. The small number of participants means that the findings may not be generalizable to the whole Thai population. It is recommended that further research should recruit a much larger sample.
CONCLUSION
The concurrent validation of the Thai adaptation of the picture version of the PPT test was good compared with the VF test, TMSE, and BNT in patients with dementia. Based on our investigation results, the Thai version of the PPT test can be employed as a test of semantic memory in Thais with dementia.
DECLARATION
This project is not funded by any external sources.
The authors declare that they have no conflicts of interest.
VS and CR were responsible for the conceptualization. DS ,VS and CR handled the methodology. AR, SC, PD and NW performed data collection. SH carried out the formal analysis. VS and CR reviewed and edited the manuscript. VS supervised the project.
REFERENCES
Funnell E. Semantic memory. In: Hillis AE, Editor. The handbook of adult language disorders: integrating cognitive neuropsychology, neurology, and rehabilitation. New York, NY: Psychology Press; 2002.p.185–205.
Gorno-Tempini ML, Hillis AE, Weintraub S, Kertesz A, Mendez M, Cappa SF, et al. Classification of primary progressive aphasia and its variants. Neurology. 2011; 76(11):1006–14.
Botha H, Josephs KA. Primary Progressive Aphasias and Apraxia of Speech. Continuum (Minneap Minn). 2019;25(1):101– 27.
Kintz S, Wright HH. Semantic Knowledge Use in Discourse: Influence of Age. Discourse Process. 2017;54(8):670–81.
Dresang HC, Dickey MW, Warren TC. Semantic memory for objects, actions, and events: A novel test of event-related conceptual semantic knowledge. Cogn Neuropsychol. 2019;36 (7–8):313–35.
Willits JA, Amato MS, MacDonald MC. Language knowledge and event knowledge in language use. Cogn Psychol. 2015;78:1–27.
Howard D, Patterson K. The Pyramids and Palm Trees Test. A test of semantic access from words and pictures. Thames Valley Company, Bury St. Edmunds, 1992.
Mehri A, Mousavi SZ, Kamali M, Maroufizadeh S. Normative data for the Pyramids and Palm Trees Test in literate Persian adults. Iran J Neurol. 2018;17(1):18–23.
Guo Q, He C, Wen X, Song L, Han Z, Bi Y. Adapting the pyramids and palm trees test and the kissing and dancing test and developing other semantic tests for the Chinese population. Appl Psycholinguist 2014;35(6):1001–19.
Callahan BL, Macoir J, Hudon C, Bier N, Chouinard N, Cossette- Harvey M, et al. Normative Data for the Pyramids and Palm Trees Test in the Quebec-French Population. Arch Clin Neuropsychol. 2010;25(3):212–7.
Klein LA, Buchanan JA. Psychometric properties of the Pyramids and Palm Trees Test. J Clin Exp Neuropsychol. 2009;31(7):803–8.
Weber M, Van Ancum J, Bergquist R, Taraldsen K, Gordt K, Midolaizak AS, et al. Concurrent validity and reliability of the Community Balance and Mobility scale in young-older adults. BMC Geriatr. 2018;18:156.
Dubois B, Villain N, Frisoni GB, Rabinovici GD, Sabbagh M, Cappa S, et al. Clinical diagnosis of Alzheimer’s disease: recommendations of the International Working Group. Lancet Neurol. 2021;20(6): 484–96.
Apostolova LG. Alzheimer Disease. Continuum (Minneap Minn). 2016;22(2 Dementia):419–34.