A systematic review of available normative data on Neuropsychological Tests for Spanish speakers in Spain, the United States of America, Latin America, and the Caribbean

Abstract

Neuropsychological testing is an essential tool in clinical settings engaged in detecting, treating, or preventing neurocognitive disorders around the world. There is a need for accurate norms across cultures, including Latinx/Hispanic communities. We reviewed studies published in English or Spanish focused on acquiring normative data for Spanish-speaking individuals in Latin America, the Caribbean (LAC), the United States (U.S.), and Spain. We searched available studies from Embase, PubMed, PsycINFO APA, Science Direct, and ProQuest up to October 31^st, 2024. Studies were imported to COVIDENCE and reviewed by two Spanish-English bilingual reviewers and one proficient English reviewer. Ultimately, 75 articles were included and categorized into U.S. (n=23), LAC (n=21), Spain (n=23) and multiregional (n=8) based. Overall, most studies included a cognitively average sample to establish the normative data, adjusting or stratifying for age, education, and sex. In Spanish speakers, cognitive performance improved with advancing age in children and adolescents and declined with age in adults. Higher education was also associated with better performance on tests across regions. While this review highlights the increasing accumulation of norms for Spanish-speaking populations, there is a continued need to expand norms to other Spanish-speaking populations not included in this analysis. Future research should add variables, such as acculturation and bilingualism, to aid normative rigor. This review works as a tool to facilitate and improve the understanding of current normative data.

Introduction

Neuropsychological testing is an essential tool in clinical settings engaged in detecting, treating, or preventing neurocognitive disorders all over the world. Norms are often critical to detecting neurocognitive disorders, however, the “validity of norm-referenced interpretations depends in part on the appropriateness of the reference group to which test scores are compared” (American Educational Research Association, 2014). This motivates the need for accurate norms across cultures, including Latinx/Hispanic communities. However, the practice of neuropsychology varies widely depending on the country or region, and the presence and quality of normative data vary widely in Latinx/Hispanic communities. For example, in Latin America and the Caribbean (LAC), the practice of neuropsychology is still nascent relative to the practice in other regions. Historically, the first isolated papers about neuropsychology in Latin America were published in late 1800s and early 1900s in Mexico, Argentina, Peru and Uruguay. The first Neuropsychological Center was created in the 1980’s in Uruguay, largely parallel with the progress of neuropsychology in other regions (Ardila, 1990). A significant milestone was the creation of the “Sociedad Latinoamericana de Neuropsicología” (Latin American Neuropsychological Society, SLAN) in 1989 in Argentina. The goal of SLAN was to promote the progress and diffusion of neuropsychological knowledge in LAC (Sociedad Latinoamericana de Neuropsicología, 2011). However, the U.S. established the National Academy of Neuropsychology in 1975, and Division 40 (Society for Clinical Neuropsychology) was recognized by the American Psychological Association in 1980, showing a slight advantage for the development and progress of the discipline of Neuropsychology in the U.S.

Early neuropsychological development in LAC also benefited from the initial publications of neuropsychology textbooks from Luria in the 1970s and the increased availability of Spanish standardized neuropsychological instruments from the TEA publishing house in Spain (Ardila, 1990). Since then, significant progress has been made in developing neuropsychological test batteries for Spanish speakers in Latin America, including Esquima de Evaluación Neuropsicológico, Batería Neuropsicológica Breve en Español: NEUROPSI, NEUROPSI: ATENCIÓN Y MEMORIA, Evaluación Neuropsicológica Infantil (ENI), Batería Neuropsicológica de Funciones Ejecutivas y Lóbulos Frontales (BANFE), Batería Neuropsicológica para Preescolares (BANPE) and Batería Neuropsicológica Computarizada de Tamizaje-BNCT (Ostrosky Shejet, 2023).

The status of neuropsychology advancement in LAC was assessed in 2016 when Arango-Lasprilla and colleagues surveyed 800 participants from Spanish-speaking countries such as Colombia, Mexico, Argentina, Bolivia, Chile, Cuba, the Dominican Republic, Ecuador, Guatemala, Honduras, Panama, Paraguay, Peru, Puerto Rico, Uruguay, and Venezuela. Survey results revealed a shortage of professional neuropsychologists that were likely compounded by a dearth of neuropsychology practicums and other clinical training opportunities (e.g., relevant coursework, postdoctoral fellowships, etc.) in LAC relative to the U.S. and Canada. The stagnant development of neuropsychology as a branch of clinical psychology in LAC has resulted in a vicious cycle, limiting educational programs for neuropsychological training, restricting resources allotted to this specialty at universities and clinics, and hindering investment in neuropsychological science across many countries. In addition, the progression of neuropsychology in LAC has also been impeded by the limited availability of culturally validated and normed assessment tools appropriate to the population.

In the U.S., where neuropsychology is otherwise well-established, the needs of Spanish speakers are not well met (Dietrich & Hernandez, 2022). In response to this need, some neuropsychological tests created in the U.S. have been standardized for some Spanish-speaking populations; for example, the Wechsler Intelligence Scale has been adopted in Mexican, Puerto Rican, and Colombian dialects (Funes et al., 2016). Both the Hispanic Neuropsychological Society (HNS) and the National Academy of Neuropsychology (NAN) developed professional guidelines regarding the delivery of neuropsychological services to Spanish speakers (Judd et al., 2009). There has also been progress in training cultural neuropsychological competencies to work with Spanish speakers, such as the Foundational Curriculum and Core Guidelines for Training in Latine/a/o-Hispanic Cultural Neuropsychology Across the Lifespan proposed by MacDonald and colleagues (MacDonald et al., 2023).

Although there has been noticeable progress in the specialty to create more tests available for Spanish speakers as well as guidelines and expert training for clinicians, evaluating Spanish speakers remains challenging due to the variable Spanish proficiency level of the evaluator or translator, dialect variations, bilingualism of the participants, cultural bias, limited assessment tools for each country or region, and inadequate norms (Puente et al., 2015; Puente & Puente, 2009).

It requires substantial effort and resources to develop, translate, and adapt culturally and linguistically appropriate tests. Multiple variations of the Spanish language and unique slang contribute to the language’s complexity across regions and countries. For example, ‘banana’ can be ‘güineo’ in Panama and Colombia’s Atlantic Coast but ‘cambur’ in Venezuela. Because the Spanish language is complex and varied across regions due to cultural heterogeneity (Ardila, 2020), norms created for a specific country do not apply to all Spanish-speaking populations, which calls for more diverse and precise norms for these populations.

Additionally, the shortage of bilingual neuropsychologists often results in use translator services. However, neuropsychological tests are susceptible to score inflation when using a translator or interpreter (Casas et al., 2012). Having third party observers during the evaluation can also have effects on the performance of participants depending on the cognitive domain, visibility of the observer, and number of observers (Eastvold et al., 2012). Additionally, evaluating Spanish-speaking patients using English tests via an interpreter may result in inaccurate translation of an instrument, excluding cultural values in evaluations, a lack of culturally informed mental health assessments, underestimating educational and socioeconomic variations by participants’ countries of origin, and lack of normative data (Cervantes & Bui, 2015; Edwards & Cardemil, 2015; Puente et al., 2015). These limitations galvanize the need to compile currently available norms for Spanish speakers as well as consideration of other factors that may influence cognitive performance in these populations.

Variations between individuals who have migrated to the U.S. and non-migrants still residing in LAC including the unique psychosocial factors that pertained to migration stress and racism can further complicate the assessment of Hispanic communities in the U.S. For Spanish-speaking migrants, other circumstances related to the neuropsychological evaluation (e.g., qualifications of the professional administering the neuropsychological tests) and acculturation factors (e.g., adaptation to the country and proficiency in the language of the land of migration) can also impact their cognitive performance, urging for research on standardized norms to be able to adequately diagnose and treat a patient across nationality accounting for their migration experience.

The present systematic review sought to address the current state of neuropsychological normative data in Spanish speakers, with three primary aims: 1) review available norms of widely used neuropsychological tests for Spanish Speakers in the U.S., Spain, Latin America, and the Caribbean, 2) report measures of central tendency of commonly used neuropsychological tests and highlight discrepancies in established norms for Spanish speakers, and 3) compare available mean and standard deviation of performances on neuropsychological tests by age range.

Methods

This systematic review adhered to the PICO criteria, which describes the Patient/Problem, Intervention or Exposure, Comparison and Outcome (Richardson et al., 1995). The publication search included papers with participants who identified as Hispanic or Latinx (i.e., Latinos, Latinas, Latine, or Latinx) and spoke Spanish as their primary language, regardless of age, sex, or disease status. The search was also extended to include both monolingual and bilingual Spanish speaking participants. Papers were compared based on sample size, tests used, the mean or range of age, the mean or range of education, the percentage of women participants, and variables used to stratify or adjust norms (e.g., age, sex/gender, education, etc.).

Search Strategy

The search criteria followed the PRISMA 2020 guidelines (Page et al., 2021). Articles from Embase, PubMed, PsycINFO APA, Science Direct, and ProQuest were systematically searched using PICO criteria for all relevant publications up to October 31^st, 2024. Search terms used were cognitive assessment Latinx* OR neuropsychological assessment in Latinos* OR neuropsychological assessment in Hispanics and Latinos* OR validated neuropsychological tests for Hispanics or Latinx* OR neuropsychological tests for Spanish-speaking participants* OR cognitive tests in Hispanics and Latinos* AND normative data* OR norms.* We excluded reviews, interviews, introductory journal articles, systematic reviews, and other studies that were not human-related from the advanced research meta-analysis. Additional manuscripts were identified via literature review and included in our analysis if they met inclusion/exclusion criteria.

Inclusion and Exclusion criteria

Articles were included based on the following criteria: 1) original research article published in a peer-reviewed journal, 2) at least part of the sample included in the study was Spanish speaking and was tested in Spanish, 3) studies were conducted in the United States of America (U.S.A/U.S.), Latin America and the Caribbean (LAC), or Spain. Studies were excluded if they met the following criteria: 1) review papers or meta-analyses, 2) thesis, reports presentations, abstract of conferences or book chapters, 3) articles that did not include at least one aim oriented to building norms, 4) test or tests included in the study did not measure cognition or were not designed as neuropsychological tests, 5) Not commonly used neuropsychological tests as reflected by absence in recent test use surveys in the U.S., LAC and Spain (Arango-Lasprilla, Stevens, et al., 2017; Gasquoine et al., 2021; Olabarrieta-Landa et al., 2016; Rabin et al., 2016) as well as Spanish test translated in TEA Ediciones (https://web.teaediciones.com) or Pearson Assessments (www.pearsonassessments.com), and 6) articles that did not clearly describe whether instruments were administered in English or Spanish, or did not have a clear division of participants evaluated in Spanish vs English in the results. Some studies included more than one tests that were not a well-known or widely used neuropsychological test; however, if the study evaluated their sample with a neuropsychological battery comprised of mostly well-known tests, we included the study in this review.

Study Selection

Studies found in the databases described above were uploaded into COVIDENCE (Veritas Health Innovation, 2014) and examined by three independent reviewers, two Spanish-English bilinguals (APL and FA), and one proficient in English (GL). Titles and abstracts were initially screened followed by a full-text review.

Data Extraction

The primary author (APL) made the initial extractions, and a second Spanish-English bilingual (FA) reviewer confirmed the data added was accurate. When there was a conflict, both reviewers compared the data extracted to reach a consensus. The following information was extracted from the papers: country or region where the study was conducted, country or region where the participants originated, language in which the article was written, aim of the study, study design, year of publication, possible conflicts of interest, population description, population age ranges, inclusion criteria, exclusion criteria, specific language dominance of the participants (i.e., monolingual, bilinguals, both or not specified), method of recruitment, total number of participants that were Spanish speakers, neuropsychological tests used, neuropsychological tests results (mean, standard deviation of scores), controlled or stratified by (i.e., age, education, gender or other), comments regarding the characteristics of scoring provider, most highlighted limitations, and if there were confounding variables to take into consideration.

Risk of Bias Assessment

Risk of bias assessment was performed using QUADAS-2 (Whiting et al., 2011) an assessment tool used to evaluate for diagnostic accuracy of studies. Since there are no tools developed for cross-sectional studies, other than checklists or quality assessments producing index scores (Kelly et al., 2024), QUADAS-2 was deemed the tool that could best accomplish risk of bias assessment for this systematic review. QUADAS-2 tool is characterized by four main domains address the following questions modified to fit the goals of the study: 1) Could the selection of patients have introduced bias? (Appropriate inclusion/exclusion criteria, selection of targeted sample) 2) Could the conduct or interpretation of the index test have introduced bias? (Were the tests interpreted adequately? Was the background appropriately discussed? Aims were adequate for the methodology of the study?) 3) Could the reference standard, its conduct, or its interpretation have introduced bias? (The targeting “healthy controls” introduced bias? Is it unclear how the norms were designed?) 4) Could the patient flow have introduced bias? (Were all patients included in the analysis? Did the analysis report missing participants?).

Analyses

Descriptive analyses of studies

Descriptive analyses were divided in four sections: U.S., LAC, Spain and Multiregional if the studies were conducted in more than one region. Results were organized according to regional setting(s) of a specific study if not they were in the category ‘individual studies’, primary publication, tests used in the study, number of the healthy/control Spanish-speaking sample, mean age and standard deviation or age range, years of education or mean years of education and standard deviation and % of women in the study. To improve the flow of information, if specific research included multiple publications and an available methodology article, only the methodology paper was included in the tables.

Results

We extracted 2632 references from the databases published until October 31^st, 2024, of which 239 papers were duplicated (see Figure 1 for the PRISMA diagram). A total of 147 studies were included for abstract screening, of which 72 were excluded for full-text evaluation mostly due to a deviation of its primary goal from norm establishment (n=52) or the lack of a commonly used and validated test (n=12). Ultimately, 75 studies were included in the data extraction and risk of bias assessment. The risk of bias is presented in Figure S1 in the supplemental materials (McGuinness & Higgins, 2021). Interrater reliability was calculated with COVIDENCE using the number of agreement scores divided by the total number of scores and provided in Table 1.

Figure 1. PRISMA 2020 flow diagram for new systematic reviews which included searches of databases and registers only.

Table 1.

Interrater Reliability

	Title and Abstract Screening			Full-text Revew
	Proportionate agreement	Cohen’s Kappa	Agreement	Proportionate agreement	Cohen’s Kappa	Agreement
Reviewer 1 vs. 2	72%	0.42	fair	76%	0.53	moderate
Reviewer 1 vs. 3	74%	0.43	moderate	73%	0.47	moderate

Ten papers from the final set of studies were written in Spanish, and 65 in English. All studies included samples of “cognitively normal”/ “healthy controls”/“normal controls.” Four studies also included patients with neurological diseases (e.g., dementia).

Studies per country were summarized in Figure 2. Most studies did not specify whether participants were monolingual Spanish speakers. Studies predominantly described the need to recruit native Spanish speakers or participants whose primary language is Spanish. Some studies included broader age ranges (e.g., from 18 to 90) in their normative data, while others were narrower (e.g., only children or older adults). We divided the age ranges into specific categories: children from 0 to 11 years of age (n=9), adolescents from 12 to 17 years (n=12), younger adults from 18 to 49 years (n=47), and older adults who were 50+ years old (n=61).

Figure 2. Geographic Origin and/or Recruitment Locations of Participants.

Number of studies per country

Most normative studies used regression and correlations analyses to observe the association of demographic variables with cognitive performance. Studies were mainly used the factors of age (n=67), education (n=55), and gender (n=42), while some included another category, such as language or mean level of parental education. Most studies excluded participants due to neurological disorder (n=70), psychiatric disorder (n=70), substance abuse (n=58), head trauma or history of stroke (n=25), and analphabetism (n=23). Other studies excluded participants with physical limitations, intellectual disability, or other uncontrolled medical diseases, participants using medication that can affect cognition, and participants who did not have Spanish as their first or preferred language. From our sample we identified 43 studies that used an objective screening tool (MMSE, GDS or CDR) as inclusion criteria and 24 used clinical interview or clinical history revision, or other cognitive screening tools (i.e., Barthel Index or TONI-2 for children) as their normal cognition determinant.

Table 2 listed the neuropsychological tests included in the 75 reviewed studies by their domain with citations of the studies that included these tests. Adaptations to the instruments implemented by each study were included in the same bullet to have a more unified view of the neuropsychological tests.

Table 2.

Neuropsychological tests divided by cognitive domains

Cognitive Function	Test
Validity	• Test of Memory Malingering (TOMM, n=1)
Intellectual Functioning	• Ravens Progressive Matrices (n=2)
Global Cognition	• Dementia Rating Scale/Mattis Dementia Rating Scale (DRS, n=2) • Mini Mental State Examination (MMSE, n=6) • Montreal Cognitive Assessment (MoCa, n=1)
Learning and Memory	• Aprendizaje Incidental (n=1) • Brief Spanish-English Verbal Learning Test (n=1) • Brief Visuospatial Memory Test (BVMT, n=2) • CERAD: Word List (n=3) • Hopkins Verbal Learning Test (HVLT, n=4) • Rey Auditory Verbal Learning Test (RAVLT/AVLT)/WHO-UCLA AVLT (n=1) • Ruff-Light Trail Learning Test (RULIT, n=1) • Selective Reminding Test (n=2) • Visuospatial Span* (n=2) • WMS: Logical Memory (n=2)
Language	• Boston Diagnostic Aphasia Examination (n=1) • Boston Naming Test (BNT) 60 item/15 item/30 item (n=8) • Category Fluency/Semantic Fluency (n=10) • CERAD: Verbal Fluency (n=1) • FAS/Letter Fluency/ Control Word Association Test (COWAT, n=15) • Peabody Picture Vocabulary Test (PPVT, n=1) • Token Test (n=1) • Vocabulary* (n=1)
Visuospatial Abilities	• Benton Visual Retention Test (n=1) • Block Design* (n=2) • Judgement of Line Orientation (JOLO, n=2) • Rey Osterreith Complex Figure Test (ROCFT, n=5) • Rosen Drawing Test (n=1) • Test de las Campanas (Bells Test, n=1) • Visual Object and Space Perception battery (n=2)
Attention & Working Memory	• Arithmetic* (n=1) • Brief Test of Attention (n=1) • Digit Span* (n=4) • Letter Number Sequencing* (n=4) • Trail Making Test Part A (TMT Part A, n=9)
Executive Functions	• Similarities* (n=2) • Stroop Color Word Interference Test (n=4) • Tower of London Test (n=2) • Trail Making Test Part B (TMT Part B, n=8) • Wisconsin Card Sorting Test (WCST/M-WCST, n=5)
Processing Speed	• A Cancellation Test (n=1) • Digit Symbol Coding* (n=4) • Symbol Digit Modalities Test (SDMT, n=3) • Symbol Search* (n=2)
Motor Functioning	• Grooved Pegboard Test (n=1) • Finger Tapping Test (n=1) • Pin Test (n=1)
Batteries	• Uniform Data Set Version 3-Neuropsychological Battery (n=2) • NEUROPSI (n=1) • RBANS (n=2) • Barcelona Test (n=1)

Notes.

^*Subscales from the Wechsler Intelligence Scales. MMSE in this table only included MMSE used to measure cognition after screening and not during.

Normative studies in the U.S.

We extracted 23 papers from the U.S. (Table 3). Some prominent studies in this region were the Neuropsychological Screening Battery for Spanish Speakers (Pontón et al., 1996) and the Neuropsychological Norms for the US-Mexico Border Region in Spanish (NP-NUMBRS) that implemented the Spanish version of BVMT-R, HVLT-R (Díaz-Santos et al., 2021), WAIS-III Digit Symbol and Symbol Search (Rivera Mindt et al., 2021), PASAT, WAIS-III Letter Number Sequencing (Gooding et al., 2021), WCST-64 (Marquine, Yassai-Gonzalez, et al., 2021), Trail Making Test A and B (Suarez et al., 2021), verbal and animal fluency (Marquine, Morlett Paredes, et al., 2021), WAIS-R Block Design and Arithmetic (Scott et al., 2021), and Grooved Pegboard and Finger Tapping Tests (Heaton et al., 2021). We also included the Uniform Data Set (UDS) from the National Alzheimer’s Disease Coordinating Center (NACC) of the National Institute of Aging (NIA) (Benson et al., 2014; Marquine et al., 2023), the Washington Heights and Inwood Community Aging (WHICAP) study (Stricks et al., 1998), the FRONTIER study (Hall et al., 2018) , and the HCHS/SOL (Morlett Paredes et al., 2024) , as well as other studies related to creating norms for Spanish-speakers in the U.S (Acevedo et al., 2000; Arango-Lasprilla et al., 2024; Carrion et al., 2024; Cherner et al., 2007; Lyness et al., 2006; Morel Valdes et al., 2024; Olabarrieta-Landa et al., 2024; Ramos Usuga et al., 2024; Rosselli et al., 2010).

Table 3.

Normative Studies of Spanish Speakers in the U.S.

Adult (18–90+) Studies
Study	Publication	State	Country of Origin	Sample Size	Age M(SD) Range	Education M(SD) Range	Sex (%Women)	Statistical Approach	Test
FRONTIER Study	Hall et al. 2018	Texas New Mexico	Mexico	136	53.61 (8.9) 40–79	8.2 (3.97) 0–18	79	Regression Based	RBANS
Normative data for U.S. adult Spanish speaking population	Rivera et al. 2024	California, Connecticut, Florida, Indiana, New Jersey, Oregon, Virginia, Wisconsin	-	245	41.1 (14.9) 18–80	15.1 (4.2) 2–26	60.8	Bayesian approach	WHO UCLA Auditory Verbal Learning Test, TMTA & B, SDMT, BTA, Bell Test, ROCF, Digit Symbol Substitution, Verbal Fluency. M-WCST, SCWT
NP-Numbers	Cherner et al. 2021 a	Arizona California	-	254	37.3 (10.2) 19–60	10.67 (4.3) 0–20	58.7	Scaled Scores T-Scores Equations	BVMT-R, HVLT-R, Digit Symbol (WAIS-III), Symbol Search (WAIS-III), PASAT, WAIS-III: LNS, WCST-64, TMT A and B, Verbal Fluency (PMR), Animal Naming, WAIS-R: Block Design, WAIS-R: Arithmetic, Grooved Pegboard, Finger Tapping Test
SOL-INCA	Morlett Paredes et al. 2024	New York, California, Florida, Illinois	Puerto Rico, Dominican Republic, Mexico, Cuba, Central America, South America	5411	63.7 (8.3) 50–86	10.8 (4.8) 18–74	55.4	Regression Based	Digit Symbol Substitution
UDS	Benson et al. 2014	-	Mexico, Cuba, South America	276	70.6 (8.7)	10.7 (5.0)	76.5	Regression Based	UDS Neuropsychological Battery: MMSE, Logical Memory, Digit Span, TMT A & B, Category Fluency (animals and vegetables), BNT-15 item, Digit Symbol
	Marquine et al. 2023	-	Mexico, Puerto Rico, Dominican Republic, Cuba, Central America, South America	191	72.2 (8.1) 51.3–94.1	11.4 (5.1) 0–20	75.4	Regression Based	UDS 3ersion 3-NB: MoCa, Number Span Test, TMT A & B, MINT, Phonemic Fluency, Semantic Fluency (Animals and Vegetables), Benson Complex Figure, Craft Story 21.
WHICAP	Stricks et al. 1998	New York	-	416	74.7 (6.2)	6.1 (4.2)	70.4	Stratified by education and age	MMSE, Similarities (WAIS-R), MDRS, BNT-15, COWAT, Category Fluency (Animal, Foods and Clothing), Boston Diagnostic Aphasia Examination: Repetition and Auditory Comprehension, Selective Reminding Test, BVRT, Rosen Drawing Test Consonant trigram (TMX)
Individual Studies	Acevedo et al. 2000	Florida	Cuba, Colombia, Peru, Puerto Rico, Argentina	237	64.9 (7.7) 50–79	13.4 (3.2) 8–17+	69.2	Regression Based	Category Fluency (Animals, Vegetables and Fruits)
	Cherner et al. 2007	California	Mexico	127	37.5 (9.4) 20–55	9.75 (4.3) 0–20	57	Scaled Scores T-Scores Equations	BVMT-R, HVLT-R
	Lyness et al. 2006	California	Mexico, Cuba, Peru, El Salvador Other	54	70.8 (7.1) 62–87	9.4 (4.9) 8–20	77.8	Stratification by education	MDRS
Pediatric Studies (0–17 years old)
Study	Publication	State	Country of Origin	Sample Size	Age M(SD) or Range	Parental Education M(SD)	Sex (%Women)	Statistical Approach	Test
Individual Studies	Rosselli et al. 2010	Florida Colorado	Mexico Cuba Colombia Puerto Rico Peru Venezuela Ecuador Guatemala Nicaragua	108	5–13	N/A	55	Stratified by age	Block Design (WISC-IV), Vocabulary (WISC-IV), ENI
All Ages Studies
Study	Publication	State	Country of Origin	Sample Size	Age M(SD) Range	Education M(SD) Range	Sex (%Women)	Statistical Approach	Test
NeSBHIS	Pontón et al. 1996	California	Mexico, Central America	300	38.4 (13.5) 16–75	10.7 (5.1) 1–20	60	Stratification by age, education and gender	BNT (Pontón Satz), COWAT, WHO-UCLA Auditory Verbal Learning Test , ROCFT, Digit Span (EIWA), Digit Symbol (EIWA), Block Design (EIWA), Color Trials 1 and 2, Pin Test, Raven’s Standard Progressive Matrices

Notes. BNT=Boston Naming Test; BTA=Brief Test of Attention; BVMT=Brief Visual Memory Test; BVRT=Benton Visual Retention Test; COWAT=Controlled Oral Word Association Test; ENI=Evaluación Neuropsicológica Infantil; M-WCST=Modified Wisconsin Card Sorting Task; HVLT=Hopkins Verbal Learning Test; MDRS=Mattis Dementia Rating Scale MMSE=Mini Mental State Examination; MINT=Multilingual Naming Test; MoCa=Montreal Cognitive Assessment; NeSBHIS=Normative data stratified by age and education for the Neuropsychological Screening Battery for Hispanics RBANS=Repeatable Battery for the Assessment of Neuropsychology Status; ROCF=Rey Osterrieth Complex Figure; SCWT=Stroop Color Word Test; SDMT=Symbol Digit Modalities Test; SOL-INCA= Study of Latinos-Investigation of Neurocognitive Aging; TMT=Trail Making Test; TOMM=Test of Memory Malingering; UDS=Uniform Data Set; WAIS= Wechsler Adult Intelligence Scale; WHICAP= Washington Heights and Inwood Community Aging project; WICS=Wechsler Intelligence Scale for Children.

^aPapers include Diaz-Santos et al. 2021, Gooding et al. 2021, Heaton et al. 2021, Marguine et al. 2021a, Marquine et al. 2021b, Rivera-Mindt et al. 2021, Scott et al. 2021, Suarez et al. 2021

^bArango-Lasprilla et al. 2024, Carrion et al. 2024, Morel Valdes et al. 2024, Olabarrieta-Landa et al. 2024, Ramos Usuga et al. 2024

Normative studies in LAC

Normative data studies for Spanish speakers in LAC are summarized in Table 4. Some of the studies included neuropsychological batteries such as the NEUROPSI (Ostrosky-Solís et al., 1999) and the CERAD (Sosa et al., 2009). The ‘ten neuropsychological tests in 11 countries’ project developed normative data for Spanish-speakers in LAC that included 3,9777 healthy individuals between the ages of 18 to 95 years old in the countries of Argentina, Bolivia, Chile, Cuba, El Salvador, Guatemala, Honduras, Mexico, Paraguay, Peru and Puerto Rico (Guardia-Olmos et al., 2015). This project provided a standardized norm for tests such as Rey-Osterrieth Complex Figure Test (ROCFT) (Rivera, Perrin, Morlett-Paredes, et al., 2015), Stroop Color and Word test (Rivera, Perrin, Stevens, et al., 2015), M-WCST (Arango-Lasprilla, Rivera, Longoni, et al., 2015), Trail Making Test (Arango-Lasprilla, Rivera, Aguayo, et al., 2015), Brief test of attention (Rivera, Perrin, Aliaga, et al., 2015), Phonological and semantical verbal fluency test (Olabarrieta-Landa, Rivera, Galarza-Del-Angel, et al., 2015), Boston naming test (Olabarrieta-Landa, Rivera, Morlett-Paredes, et al., 2015), Hopkins verbal learning test (Arango-Lasprilla, Rivera, Garza, et al., 2015) and Revised and Test of memory malingering (Rivera, Perrin, Weiler, et al., 2015).

Table 4.

Normative Studies of Spanish speakers in LAC

Adult (18–90+) Studies
Study	Publication	Country	Sample Size	Age M(SD) Range	Education M(SD) %Range Range	Sex (%Women)	Statistical Approach	Test
10/66 Dementia Research Group	Sosa et al. 2009	Mexico	1,823	65–80+	No education to completed tertiary education	Urban (65.6) Rural (59.9)	Adjusted for sex, education and age	CERAD-word list learning test, CERAD-verbal fluency (animal fluency), Community Screening Instrument for Dementia
		Peru	1,767			Urban (64.1) Rural (52.5)
		Venezuela	1,826			63.5
		Cuba	2,621			64.4
		Dominican Republic	1,769			65.4
10 Neuropsychological Tests in 11 Counties in LAC	Guardia-Olmos et al. 2015 c	Argentina	320	45.7 (19.5)	>12 yeard of education % 53.8	70.0	Stratified by age and education & Linear Regression	BTA, M-WCST, HVLT-R, TMT, FAS, Category Fluency, BNT-60, SCWT, ROCF & TOMM, Letter Fluency (Letter M) & Category Fluency (Fruits and Occupations)
		Bolivia	274	55.8 (22.0)	17.5	63.9
		Chile	320	55.1 (19.6)	24.7	58.1
		Cuba	306	53.0 (19.7)	23.5	53.6
		El Salvador	257	56.0 (20.7)	21.0	61.1
		Guatemala	214	53.2 (17.4)	37.9	55.6
		Honduras	184	76.1 (18.8)	23.9	63.6
		Mexico	1300	52.5 (20.5)	22.7	66.8
		Paraguay	263	53.0 (14.8)	17.9	61.6
		Peru	245	43.4 (20.6)	64.5	64.5
		Puerto Rico	294	50.9 (18.5)	45.6	57.1
Grupo de Neurociencias (GNA)	Torres et al. 2022	Colombia	2673	32.94 (12.10) 18–65+	8.89 (4.34) 1–12+	56.2	Stratified by age and education	TMT, Digit Span, WCST, CERAD-Semantic Fluency (Animals), BNT-15 item, CERAD-Word List, Constructional Praxis Copy, ROCFT, Ravens Progressive Matrices, Part A, A Cancellation Test & FAS
Individual Studies	del Cacho-Tena et al. 2023	Argentina	316	45.7 (19.5) 18–89	13.8 (4.53) 5–24	70	Regression Based	15-item Boston Naming Test
		Chile	245	57.4 (19.7) 18–90	9.55 (5.49) 2–24	58
		Cuba	292	52.7 (19.4) 18–90	11.7 (3.64) 2–22	73
		El Salvador	244	56.0 (22.1) 18–94	8.9 (5.3) 1–25	61
		Honduras	181	48.8 (18.9) 18–89	8.6 (5.6) 1–24	63
		Mexico	1129	53.5 (20.2) 18–94	9.3 (4.9) 1–26	67
		Paraguay	258	53.2 (14.8) 18–79	9.5 (4.4) 3–23	62
		Puerto Rico	163	49.0 (16.9) 21–89	13.4 (4.1) 2–25	55
	Guerrero-Berroa et al. 2016	Costa Rica	95	93.7 (3.4) 90–102	2.9 (1.8) 0–6	68.3	Stratified by age and sex	MMSE, CERAD-Word List, BNT (15 item), Fluency Test
		Puerto Rico	23	92.9 (2.6) 90–99	3.9 (1.6) 0–6	39.1
	Gaete et al. 2020	Chile	526	18–90	30.2% Superior Education	57	Regression Based	MoCa
	Labos et al. 2013	Argentina	891	64.05 (13.4) 20–97	10.6 (4.6) 2–25	62	Stratified age, education	Category Fluency (Animals and Fruits) & Letter Fluency (Letter M)
	Rodríguez-Lorenzana et al. 2020a,b	Ecuador	322	41.3 (18.2) 18–84	13.2 (4.6) 2–25	54.04	Stratified age, education and gender & Multiple Linear Regression	HVLT-R & ROCF, MMSE, Letter Fluency (A,S,M) & BNT 60 items
Pediatric Studies (0–17 years old)
Study	Publication	Country	Sample Size	Age M(SD) or Range	Parental Education M(SD)	Sex (%Women)	Statistical Approach	Test
Individual Studies	Rivera et al. 2021	Chile	381	11.5 (3.4)	12.4 (3.0)	50.1	Regression Based	Letter Fluency (M, R, P)
		Cuba	381	11.5 (3.5)	16.2 (1.8)	49.9
		Ecuador	297	11.4 (3.4)	14.4 (3.6)	58.2
		Guatemala	202	10.7 (2.5)	10.4 (4.1)	46.5
		Honduras	292	11.2 (3.2)	13.3 (4.1)	54.5
		Mexico	919	11.4 (3.5)	13.1 (3.9)	51.6
		Paraguay	297	11.6 (3.5)	14.1 (2.9)	53.9
		Peru	323	11.9 (3.3)	12.5 (2.4)	49.5
		Puerto Rico	192	12.1(3.6)	14.6 (2.5)	55.7
All Ages
Study	Publication	Country	Sample Size	Age M(SD) or Range	Education M(SD)	Sex (%Women)	Statistical Approach	Test
Individual Studies	Ostrosky-Solís et al. 1999	Mexico	800	47.7 (20.1) 16–85	6.8 (6.1) 1–24	52	Stratified by age and education	NEUROPSI
	Zegarra-Valdivia et al. 2022	Peru	2602	6–90 years old	-	-	Stratified by age, sex and education & Regression Based	Verbal Fluency (FASMRP) & Category Fluency (Animals and Fruits)

Notes. BNT=Boston Naming Test; BTA=Brief Test of Attention; BVMT=Brief Visual Memory Test; BVRT=Benton Visual Retention Test; CERAD=Consortium to Establish a Registry for Alzheimer’s Disease; COWAT=Controlled Oral Word Association Test; M-WCST=Modified Wisconsin Card Sorting Task; HVLT=Hopkins Verbal Learning Test; MDRS=Mattis Dementia Rating Scale; MMSE=Mini Mental State Examination; MoCa=Montreal Cognitive Assessment; MPE=Mean Parental Education ROCF=Rey Osterrieth Complex Figure; SCWT=Stroop Color Word Test; SDMT=Symbol Digit Modalities Test; TMT=Trail Making Test; TOMM=Test of Memory Malingering; WAIS= Wechsler Adult Intelligence Scale

^cOlabarrieta-Landa et al. 2015a, Olabarrieta-Landa et al. 2015b, Arango-Lasprilla et al. 2015a, Arango-Lasprilla et al. 2015b, Arango-Lasprilla et al. 2015c, Rivera et al. 2015a, Rivera et al. 2015b, Rivera et al. 2015c, Rivera et al. 2015d, Rivera et al. 2019

Additional studies from other LAC countries such as Argentina (Labos et al., 2013), Costa Rica (Guerrero-Berroa et al., 2016), Colombia (Torres et al., 2022; Torres et al., 2021), Cuba (delCacho-Tena et al., 2023; Sosa et al., 2009), Chile (delCacho-Tena et al., 2023; Gaete et al., 2020), Dominican Republic (Sosa et al., 2009), Ecuador (Rodriguez-Lorenzana, Benito-Sanchez, et al., 2020; Rodriguez-Lorenzana, Nunez-Fernandez, et al., 2020), El Salvador (delCacho-Tena et al., 2023), Honduras (delCacho-Tena et al., 2023), Mexico (delCacho-Tena et al., 2023; Sosa et al., 2009), Paraguay (delCacho-Tena et al., 2023), Peru (Zegarra-Valdivia et al., 2022), Puerto Rico (delCacho-Tena et al., 2023; Guerrero-Berroa et al., 2016) and Venezuela (Sosa et al., 2009) are synthesized in Table 4.

Normative studies in Spain

Results from normative studies in Spanish-speakers in Spain are reported in Table 5. One of the most renowned normative studies in Spain is the NEURONORMA project with a sample of 356 Spanish citizens of Caucasian origin from the regions of Andalusia, The Basque Country, Catalonia, Galicia, Madrid, and Murcia (Pena-Casanova, Blesa, et al., 2009). Participants were cognitively intact and functionally independent (impairments in instrumental activites of daily living= 0). Spanish citizens over the age of 49 years recruited between 2004 and 2007 from urban sources. They completed measures of health, functionality, sociodemographic characteristics, and cognition questionnaires and a Spanish neuropsychological battery that included Verbal Span, Visuospatial Span, Letter Number Sequencing, TMT and SDMT (Pena-Casanova, Quinones-Ubeda, Quintana-Aparicio, et al., 2009), Visual Object and Space Perception Battery-Abbreviated and JOLO (Pena-Casanova, Quintana-Aparicio, et al., 2009), Stroop Color-Word Test and Tower of London-Drexel (Pena-Casanova, Quinones-Ubeda, Gramunt-Fombuena, Quintana, et al., 2009), ROCF and FCSRT (Palomo et al., 2013) semantic and letter fluency (Pena-Casanova, Quinones-Ubeda, Gramunt-Fombuena, Quintana-Aparicio, et al., 2009) and BNT-15 (M. Casals-Coll et al., 2014). NEURONORMA project also consisted of studies with younger adults (Jovenes) using verbal span, visuospatial span, LNS, TMT, SDMT (Tamayo et al., 2012), Visual Object and Space Perception Battery and JOLO (Calvo et al., 2013), Stroop Color-Word Test and Tower of London-Drexel (Rognoni et al., 2013), Boston Naming Test and Token Test (Aranciva et al., 2012), and Ruff-Light Trail Learning Test (Pérez-Enríquez et al., 2021).

Table 5.

Normative Studies of Spanish speakers in Spain

Adult (18–90+) Studies
Study	Publication	Country	Sample Size	Age M(SD) or Range	Education M(SD) or %Range or Range	Sex (%Women)	Statistical Approach	Test
Envejecer en Leganés	del Ser Quijano et al., 2004	Leganés	59	78.7 (5.6) 71–86+	Primary School 24%	50.3%	Stratified by age and education	MMSE, BVRT, Bells Test, Logical Memory (WMS), Category Fluency (Animals), Aprendizaje Incidental, Clock Drawing Test, TMT-A, Recuerdo libre y facilitado de figuras & Similarities (WAIS)
NEURONORMA PROJECT	Peña Casanova et al. 2009 d	Sevilla, Bilbao, Barcelona, Terrassa, Santiago de Compostela, Madrid & Murcia	356	64.9 (9.3) 50–80+	10.4 (5.4)	59.6	Regression Based	Verbal Span, Visuospatial Span, LNS, TMT, SDMT, Visual Object and Space Perception Battery-Abbreviated, JOLO Stroop Color-Word Interference Test, Tower of London-Drexel, ROCF, Semantic Fluency, Letter Fluency, BNT-15
NEURONORMA JÓVENES	Peña Casanova et al. 2012e	Barcelona	179	34.9 (9.5)	13.4 (3,5)	63.7	Regression Based	Verbal span, visuospatial span, LNS, TMT, SDMT, Visual Object and Space Perception Battery and JOLO, Stroop Color-Word Test and Tower of London-Drexel
NEURONORMA PLUS	Perez-Enriquez et al. 2021	Barcelona	308	54.1 (19.2) 18–95	11.42 (4.2)	54.2	Correlations	M-WCST, RULIT
SABIEX	Iñesta et al., 2022	Elche	103	65.8 (6.6) 55–87	11.4 (3.46) 3–22	67	Regression Based	Digit Span, LNS, TMT, SDMT & MMSE
SCAND	López-Higes et al. 2022	-	2088	70.8 (7.2) 50–89	Higher education (31.3%)	67	Regression Based	Category Fluency (animals, fruits, kitchen tools and clothes)
Individual Studies	Benito-Cuadrado et al., 2002	Barcelona	445	57.34 (16.75) 18–92	9.07 (5.44) 1–20	55.4	Regression Based	Barcelona Test
	Muntal et al. 2020	Terrassa	609	52.42 (19.49) 20–89	11.38 (5.6)	55.2	Regression Based	RBANS Form A
All Ages Studies
Study		Publication	Country	Sample Size	Age M(SD) or Range	Education M(SD) or %Range or Range	Sex (%Women)	Statistical Approach	Test
Individual Studies	Lubrini et al. 2022	Sevilla, Córdoba, Valencia, Barcelona, Ávila, Toledo, Málaga, Bilbao, Madrid	257	46.7 (25.3) 17–100	13.1 (3.8) 3–20	65	Regression Based	Verbal Fluency (FAS) & Category Fluency (Animal and Fruits/Vegetables)

Notes. Notes. BNT=Boston Naming Test; BTA=Brief Test of Attention; BVMT=Brief Visual Memory Test; BVRT=Benton Visual Retention Test; CERAD=Consortium to Establish a Registry for Alzheimer’s Disease; COWAT=Controlled Oral Word Association Test; HVLT=Hopkins Verbal Learning Test; LNS=Letter Number Sequencing; JOLO=Judgement of Line Orientation; M-WCST=Modified Wisconsin Card Sorting Task; MDRS=Mattis Dementia Rating Scale; MMSE=Mini Mental State Examination; MoCa=Montreal Cognitive Assessment; MPE=Mean Parental Education; ROCF=Rey Osterrieth Complex Figure; Ruff Light Trail Learning Test; SCWT=Stroop Color Word Test; SDMT=Symbol Digit Modalities Test; TMT=Trail Making Test; TOMM=Test of Memory Malingering; WAIS= Wechsler Adult Intelligence Scale; WMS=Wechsler Memory Scale.

^dPeña-Casanova et al. 2009a, Peña-Casanova et al. 2009b, Peña-Casanova et al. 2009c, Peña-Casanova et al. 2009d, Peña-Casanova et al. 2009e, Casals-Colls et al. 2014

^ePalomo et al. 2013, Rognoni et al. 2013, Tamayo et al. 2012, Calvo et al. 2013, Aranciva et al. 2012

Other studies were from Barcelona (Benito-Cuadrado et al., 2002; Lubrini et al., 2022), Leganés (Ser Quijano et al., 2004), Terrasa (Muntal et al., 2020), Granada (Carnero Pardo et al., 2022), Elche (Iñesta et al., 2022), Sevilla, Córdoba, Valencia, Ávila, Toledo, Málaga, Bilbao and Madrid (Lubrini et al., 2022).

Multiregional Studies

Rivera and colleagues (2017) designed a study to collect normative data from 6,030 clinically healthy children and adolescents from Chile, Cuba, Ecuador, Guatemala, Honduras, Mexico, Peru, Puerto, and Spain. Colombia was also included in this study; however, this sample has been published elsewhere, which reduces the sample to 4,373 children and adolescents from 6 to 17 years old (Rivera & Arango-Lasprilla, 2017). Participants recruited in this study were born in the current country where they lived, had Spanish as a primary language, and had an IQ of more than 80 on the Test of Non-Verbal Intelligence (TONI). Participants were evaluated with the following neuropsychological tests: ROCFT , Stroop Color Word Interference Test (Rivera, Morlett-Paredes, et al., 2017), Modified Wisconsin Card Sorting Test (Arango-Lasprilla, Rivera, Nicholls, et al., 2017), Trail Making Test A and B (Arango-Lasprilla, Rivera, Ramos-Usuga, et al., 2017), Symbol Digit Modalities (Arango-Lasprilla, Rivera, Trapp, et al., 2017), a Shortened version of Token Test (Olabarrieta-Landa, Rivera, Rodriguez-Lorenzana, et al., 2017), Concentration Endurance Test, Phonological and Semantic Verbal Fluency (Olabarrieta-Landa, Rivera, Lara, et al., 2017), Peabody Picture Vocabulary Test (Olabarrieta-Landa, Rivera, Ibanez-Alfonso, et al., 2017), and the Learning and Verbal Memory test which was not included in this review due to it being a relatively new cognitive test (Rivera, Olabarrieta-Landa, et al., 2017). In this study, authors operationalized the age variable in two ways: 1) average age and 2) centered age, which was included to avoid multicollinearity. Additionally, they had variables such as the sex/gender of the child and the mean level of parental education (MLPE).

Discussion

This systematic review describes the state of normative data on neuropsychological tests for Spanish speakers across the regions of the U.S., LAC and Spain. Participants recruited in the studies were mainly “healthy controls” or “normal controls” that went through diverse screening processes across studies to determine their cognitive status. Including “healthy controls” to create normative data establishes baseline estimates of how a person with similar characteristics but no apparent confounding conditions performs on a specific neuropsychological test. Studies that used a second group composed of cognitively impaired individuals(Arnold et al., 1998; Marta Casals-Coll et al., 2014; Lyness et al., 2006; Ser Quijano et al., 2004) had the purpose to confirm an adequate comparison of the “healthy controls” participants versus impaired participants in neuropsychological tests. However, defining normal cognition and identifying cognitively “normal” participants are challenging. Although most studies noted rigorous inclusion and exclusion criteria, some did not. For example, some studies required the cognitive screeners, questionnaires on depression, and health history, whereas other only used one screening tool or a clinical interview with no expansion of which variables were used during this interview to determine the “healthy” status. That said, the variability in inclusion and exclusion criteria across studies does not invalidate the results of studies with fewer exclusion criteria descriptions. However, it does raise the question of their accuracy in excluding impairment in their “cognitively healthy” or “normal controls” participants, and under what criteria this decision was made.

It should be noted that the most used screening tool to determine normal cognition across these studies and regions was the MMSE. This is expected as the MMSE is the most used global cognition screening instrument in the U.S. and Canada (Rabin et al., 2016), Latin America (Arango-Lasprilla, Stevens, et al., 2017), and Spain (Olabarrieta-Landa et al., 2016). Some studies also included mood scales or questionnaires to determine symptoms of depression, which could affect the cognitive performance of participants (Formanek et al., 2020). There was slight variability in screening symptoms of depression where some studies used questionnaires, such as the Patient Health Questionnaire (Guardia-Olmos et al., 2015) while others used the Hamilton Depression Rating Scale (HDRS; Hamilton, 1960; Peña-Casanova, Blesa, et al., 2009), likely based on the scale’s validity in a specific population. Specifically, the PHQ-9 was mostly used in the LAC populations and has been validated in countries such as Argentina, Chile, and Peru (Errazuriz et al., 2022; Urtasun et al., 2019; Villarreal-Zegarra et al., 2019) while the HDRS has been well-validated in the Spanish population (Lobo et al., 2002).

Regarding demographic factors that affect cognition, age was the most widely integrated in the analysis across normative studies. Age is known to be associated with cognitive changes, and it is a primary risk factor for the development of Alzheimer’s disease dementia (Murman, 2015). Age showed significant associations with cognitive performance across all domains in this review. Moreover, the studies included in this review with adult/older adults samples revealed a negative linear association of age and cognitive performance, suggesting a higher age was associated with a lower cognitive score. Opposite trends were observed in samples of children/adolescents (i.e., higher age meant higher cognitive scores) as is developmentally expected (Craik & Bialystok, 2006; Giedd et al., 1999). However, the aging process is accompanied by other factors that might vary depending on the region. For example, a recent study revealed that some factors that affect cognition in LAC countries include mental health symptoms, social determinants of health (e.g., health care quality/access), education, and physical activity, which are most accentuated in middle-income countries in Latin America relative to high-income countries (Santamaria-Garcia et al., 2023).

The second most common demographic factor used to adjust norms was education, with higher years of education generally predicting better cognitive performance. This association held true across studies analyzed in this review, regardless of the region or country, and even when level of education of parents was used as a demographic adjustment. However, education systems differ in each country regarding years of education typically attained, which may be especially relevant in lower education vs higher education countries. This systematic review revealed significant discrepancies in mean years of education across LAC countries/regions, which may introduce bias when pooling data across different regions and assessing neuropsychological performance. For example, a one-year change in low education countries may be more profound than a one-year change in high education countries. Additionally, years of education may not be a good indicator of quality of education, which was not systematically assessed in the studies, and may introduce further bias when generalizing across regions and countries.

The third most used factor to adjust normative data was sex/gender as an adjustment factor. For example, for the Judgment of Line Orientation test, two studies shared the conclusion studies that men performed better than women on this test (Calvo et al., 2013). Notably, both studies were from the same region and project but were performed in different age ranges of adults (i.e., 18–49 vs. 50+). However, some studies found no relationship between sex/gender and performance on many tests (Arango-Lasprilla, Rivera, Aguayo, et al., 2015; Benito-Cuadrado et al., 2002; Labos et al., 2013; Ser Quijano et al., 2004). This might be due to the overall changes in the role of women in society, where women now get more education and more diverse job opportunities (Agree, 2017). Nevertheless, we observed discrepancies across studies on sex/gender impact the same or similar tests. For example, on category fluency, one study noted a difference between the performance of men and women in a certain category (i.e., fruits) but another study (Benito-Cuadrado et al., 2002; Rodriguez-Lorenzana, Benito-Sanchez, et al., 2020) did not. This discrepancy might be explained by the cultural differences where both studies were conducted (i.e., the first was conducted in Argentina while the latter was in Ecuador), spurious findings, or underpowering. Additionally, normative studies should consider whether the magnitude of a demographic difference is sufficient to warrant adjustment even if statistically significant (Smith et al., 2008).

Normative studies in the U.S.

A previous review of normative data on Spanish speakers in the U.S. (Morlett Paredes et al., 2021) provided an updated status regarding neuropsychological norms for Hispanics in the U.S. As Morlett-Paredes and colleagues (2021) pointed out in their review, most of the normative studies have been performed in populations of Spanish speakers in California, Florida, Texas, and New York. Our results are consistent with the above stated findings, and most normative studies on Spanish speakers in the U.S. were conducted in Arizona, California, Texas, Florida, and New York, states with the most robust representation of Spanish speakers in the U.S. (U.S. Census Data). Future work in this area may explore any differential performances on neuropsychological tests by state of residence (within the U.S.) and contributions of cognitive or social factors.

Regarding participants’ country of origin, most came from countries such as Mexico, Cuba, and other parts of the Caribbean. Some studies listed Mexico as the country of emigration and grouped non-Mexican Spanish speakers into broader categories such as “Central America” or “South America” (Acevedo et al., 2000; Benson et al., 2014; Breton et al., 2021; Pontón et al., 1996). The use of these broad categories facilitates recruitment and reduces the number of categories, but it does not acknowledge the heterogeneity of the recruited population. As previously noted, the Spanish language can vary greatly across countries. Also, there are limited comparisons between participants born in the U.S. vs. migrated from LAC or elsewhere. Moreover, there was insufficient information about the years of migration and where the education was obtained. These variables might seem irrelevant when the participant has a good English proficiency level or seems well-acculturated, but studies have shown that non-native English speakers tend to score lower on neuropsychological performance compared to their native English speakers’ counterparts (Kisser et al., 2012; Stalhammar et al., 2022). Moreover, a higher level of acculturation shows better performance of Hispanic participants in cognitive tests that measure learning and memory, attention, working memory, and executive functions (Mendoza et al., 2022).

We highlight that there is evident progress observed in the specialty of building more diverse and sensitive norms to measure neuropsychological performance in Spanish-speaking people in the U.S. Most recent studies, like the NP-NUMBRS, have included detailed demographic information about participants such as education in the country of origin. The need to build more diverse norms in the specialty of neuropsychology is an ongoing conversation (Manly & Echemendia, 2007; Pedraza & Mungas, 2008) that has become more pressing over the years after the latest criticism related to the race-based normative data (Possin et al., 2021). Based on the most recent studies we analyzed in this review, there is an improvement in cross-cultural neuropsychology in the U.S.

Normative studies in LAC

Neuropsychology is an emerging specialty in LAC. Our review identified 26 normative studies from LAC, including multiregional studies, with sample size ranging from 189 to 3977 participants. This reflects progress in LAC neuropsychology that became more prominent in the 2000s. Application of country-based norms in clinical settings requires a critical analysis of several factors, including the characteristics of the normative sample. This is particularly true when applying existing norms to interpret performance for patients from other countries without available normative data (e.g., using norms from Guatemala to interpret performance for a person from El Salvador). As we saw in the articles of the 11 Latin American countries (Guardia-Olmos et al., 2015), notable variabilities across countries were present (e.g., differences in mean scores of some tests), and some variables might be more relevant to, or reflective of, cognitive health in one country compared to others. While Spanish is the official language in most Latin America countries, regional differences in the use of Spanish exist across regions and countries that must be considered during clinical evaluations.

Consideration of differences between participants who live in urban versus rural areas was approached in at least one study (Sosa et al., 2009). Participants living in a metropolitan area might be affected by variables (e.g., pollution) not commonly seen in rural areas and vice versa. Exposures in urban areas versus rural areas is very different depending on the country of Latin America, which can be an additional variable to explore in future studies.

Despite these limitations, the specialty of neuropsychology is showing significant improvement. Now, there are tests developed by native Spanish speakers, such as the NEUROPSI, used by 38% of Latin American neuropsychologists (Arango-Lasprilla, Stevens, et al., 2017). Continued development of original tests for Spanish speakers in Latin America that encompasses the cultural diversity in LAC remains a future direction of study.

Normative studies in Spain

Many neuropsychological test adaptations used in Latin America originated in Spain. We found 24 normative data studies including Spanish participants. This number may represent an underestimation due to missing studies published in other sources that might not be peer-reviewed or excluding other datasets used to find information in Spanish, such as Redalyc. However, it is also important to note that the three papers that did not accomplish the quality index were from Spain, and this happened mainly because the NEURONORMA project papers often did not have a description of demographic data, missing variables, and in some cases having no clearly stated limitations of the study. This also happened with some studies in Latin America, in which demographic information was omitted from several articles and only referenced from the protocol papers of the studies. Understandably, protocol papers are a good source of demographic information, but in some cases, studies did not report changes related to samples in the following articles. It is not possible to discern whether there were no changes in sample size from publication to publication or if their sample size changes were simply not reported in the consequent papers. Additionally, the lack of individual limitations from the studies restricted us from understanding the flaws and area of improvement for normative data studies Spanish speakers.

Nevertheless, the NEURONORMA Project was a seminal study conducted in Spain to obtain normative data. The NEURONORMA has extended from its original goal of getting data for older adults to obtaining data for younger adults on the NEURONORMA Plus battery. Aside from NEURONORMA, we found other studies that included sample sizes ranging from 103–1055 participants. There was a variability in the cognitive tests analyzed, but not much variability in the age ranges (mean ages ranging from 57.34–86.2) of the samples studied or the percentage of women participants included (rate of women participants from 50.3%–67%). Studies included cities such as Barcelona, Seville, Bilbao, Terrassa, Santiago de Compostela, Madrid, Murcia, Leganés, Granada, and Elche, which shows that studies are spread all over the country and not mainly in regions such as Sevilla and Madrid which have the largest populations. A limitation of the NEURONORMA studies that is important related to their population is that they noted that participants were mainly Spanish Whites, which reduced the diversity of the sample with different cultural backgrounds.

Overall, studies in Spain showed significant variability of tests like those seen in LAC and the U.S. There were also tests that are not typical for the populations in other regions, such as the Recuerdo libre y facilitado de figures, which is a modified version of the FCSRT, or Figura Compleja Modificada de Taylor (FCMT; Modified Complex Figure of Taylor), or the Ruff-Light Trail Learning Test (RULIT). In this case, we can see that there are more adaptations of tests in Spain compared to LAC, which is expected given Spain’s progress in the specialty, an essential process in adapting other well-known neuropsychological tests in the future.

Summary of observed limitations across studies

Multiple test versions. Attention should be given to the differences and modifications of tests across regions and even in the same regions. As we have observed, some tests like the Boston Naming Test have different versions in English and is a test that typically could have different adaptations depending on the region of use. This has been observed in the literature before for other non-Western cultures and languages (Ballard et al., 2019; Miotto et al., 2010). Therefore, caution should be made when trying to generalize norms from one region to the other due to the differences in adaptations that are not always disclosed in publications.

Limited pediatric studies. Notably, studies included in this review were mostly oriented to building normative data in older adults (50+) and younger adults (18–49). This reflects the state of the neuropsychology specialty in the U.S. and LAC, in which most neuropsychologists (78–88%) currently work with those between 19 to 65+ years old (Arango-Lasprilla, Stevens, et al., 2017; Rabin et al., 2016). The lesser availability of normative data for the pediatric population implies a great need for research in Spanish-speaking pediatric populations to strengthen normative data across the lifespan.

Unmeasured socioeconomic status. Socioeconomic status has proven to be a key factor for brain development and affects domains such as executive function (Dennis et al., 2022). Higher socioeconomic status includes resources for people to access health enhancing materials such as housing, high quality health care and food and safe working conditions as well as the possibility of obtaining higher education (Glymour & Manly, 2008). Recent literature has also incorporated Social Determinants of Health (SDOH), which include socioeconomic status and other determinants such as a quality of education. Education quality was found to be a more relevant factor compared to others when measuring SDOH in normative models (Schneider et al., 2024).

Language of test administration. Many studies involving Hispanic/Latinx participants were not included in this review because they did not report whether these participants were evaluated in Spanish or English. This information is fundamental for understanding the differences between Latinx/Hispanics bilinguals evaluated in English or Spanish. Some of these studies also did not take acculturation as an important factor for the evaluation of Spanish speakers nor report frequency of language preference in bilinguals. Although it is of great importance to do what the participant feels comfortable with, there should also be an objective component that reflects the language proficiency of the participants.

Bilingualism effects. Research has shown that bilingualism may be an important factor in predicting cognitive performance. A broad issue, affecting studies in Latin America, Spain, and the U.S., is the absence of recording bilingualism as a variable in many research designs. In some studies, researchers assume Spanish is the primary language simply because participants reside in a Spanish-speaking country. This raises a critical question: How does cognitive performance differ between individuals whose primary language is not Spanish but who live in Spanish-speaking countries, compared to those whose primary language is Spanish? However, across regions, bilingualism was often not accounted for. For example, in Spain, bilingualism is not commonly measured as a variable in neuropsychological studies, despite the fact that Catalan, a co-official language in Cataluña, is widely spoken in certain regions (Oficina para las Lenguas Oficiales en la Administración General del Estado, 2024). Although most of Spain’s population speaks Spanish, the multilingual context should be accounted for in future studies on normative data.

In the U.S., Spanish-English bilinguals tend to outperform monolingual Spanish speakers in areas such as processing speed, attention, working memory, and executive functions (Rivera Mindt et al., 2008). While bilingualism is often studied in Spanish speakers in the U.S., it is also relevant for those in other regions (Guardia-Olmos et al., 2015). The exclusion of bilingualism as a variable in many LAC and Spanish studies involving Spanish speakers may be due to the reduced necessity for a second language in daily life (e.g., work, education, social interactions) in Spanish-speaking countries. In contrast, Spanish speakers in the U.S. must learn English because it is the official language, and about 80% of the population in the U.S. is monolingual (U.S. Census, 2019).

Beyond considering bilingualism and a simple yes-no variable, language proficiency is also an important issue. Effects of bilingualism on cognitive performance may vary with language proficiencies in primary and secondary languages (Rosselli et al., 2023). In addition to language proficiency, cultural differences and education levels in the participant’s country of origin should also be considered. For example, the NP-NUMBRS study (Cherner et al., 2021) incorporates education levels as a variable, as differences in educational systems may explain variations in cognitive performance among Spanish speakers. Thus, some culture-specific factors such as familism and family cohesion can affect cognition and predict cognitive decline in Hispanics/Latinos (Estrella et al., 2024).

Summary. Future research for the development of normative data in neuropsychological tests should focus on 1) developing normative data that can be more generalized or trying to include more regions that do not currently have available normative data for specific populations; 2) considering bilingualism including language proficiencies as a relevant variable that might affect cognitive performance in Spanish-speakers across regions; 3) increase sample sizes and representativeness including multiethno-multicultural samples to improve regional generalization, 4) provide more fulsome descriptions of “healthy cognitive” group; 6) report social determinants of health such as socioeconomic status, acculturation or social determinants of health; 7) explore cultural differences like rurality that can act as confounding variables.

Limitations of the present review

This review has several limitations. The criteria used in this review excluded studies with helpful information about neuropsychological profiles of the Spanish-speaking population but did not achieve the primary goal of building normative data. The criteria were conservative and included only peer-reviewed articles but not book chapters, grey literature, systematic reviews, meta-analysis, or presentations. The exclusion of other sources was mainly due to focus on finding peer-reviewed literature and/or the availability of information needed to assess bias, confirming normal status, and reporting central tendency/variability statistics. However, this approach may have unnecessarily excluded reports that would meaningfully add to the corpus of norms research for Spanish speakers. Future research should focus on the inclusion of different information formats to see if this additional available data might be helpful for other researchers or clinicians. Additionally, most of the tests included in this review were commonly used in clinical settings that either had an available form in Spanish by a publishing company such as TEA Ediciones (https://web.teaediciones.com) or Pearson Assessments (www.pearsonassessments.com) or had been mentioned in one of the survey studies of the regions examined. However, a few novel and less common tests developed for the Spanish-speaking population may be a valuable clinical tool (i.e., CLOX, Brief Spanish-English Verbal Learning Test, Texas Spanish Naming Test) and should be validated pending available data in the future.

As mentioned previously, the databases used in this normative data review might have restricted the sample of studies. Other journal databases (e.g., Redalyc) with more information on studies written in Spanish might thus increase the sample of studies. Additionally, using only peer-reviewed studies might impact the inclusion of studies from low to middle-income LAC countries due to limited funding resources to build normative data or to publish in international journals. A more ‘lenient’ search strategy in Spanish would likely increase the number of available studies for review.

Another limitation of the current review is the Cohen’s Kappa obtained for the interrater reliability for the title, abstract, and full-text reviews, which ranged from fair to moderate agreement, This could be explained by the Cohen’s Kappa paradox, which challenges the assumption that the value of kappa statistic increases agreement in data (Zec et al., 2017). In the future, we will incorporate other agreement indexes, such as the Bennet, Alpert, and Goldstein S, to confirm an adequate level of agreement (Bennett et al., 1954).

A major limitation of this review is the lack of appropriate risk of bias assessments for normative data tests in neuropsychology, specifically of cross-sectional studies. Checklists like the STROBE (von Elm et al., 2008) help to appraise the quality of the review, but not the bias inherent in the studies. A QUADAS-2 specifically for normative studies in neuropsychology might be valuable. Future directions could be oriented to building such a risk of bias tool that, if possible, adds the input of researchers oriented to building normative data and can objectively specify the most relevant variables to determine the “quality” of a normative data study in Spanish speaking samples.

Conclusion

Neuropsychological tests give key information for diagnosing and treating many neurological diseases. Moreover, these tests can give us further information on affected areas of cognition and how we can help improve patients’ lives. However, to make an accurate measure of a patient’s cognitive state, neuropsychological tests must have norms that can compare a patient to other healthy peers with similar characteristics. For years, neuropsychological tests have predominantly been used in Western, White, and predominantly English-speaking cultures. There is now greater effort to translate, adapt, and create normative data for Spanish-speaking individuals. This review detailed the latest normative research for Spanish-speaking populations in the U.S., LAC, and Spain. Notably, studies included in this review were primarily oriented to building normative data in older adults (50+) and younger adults (18–49). Taken together, although there is a noticeable improvement in normative data for Spanish-speaking populations, we acknowledge there are many avenues for further improvement.

Supplementary Material

Figure S1

Fig S1 Risk of bias graph

Availability of data and materials

Data was deposited on COVIDENCE and Excel and is available by contacting the first author.