When, How, and for Whom? A narrative review contextualizing bilingual effects in ADHD

1 Introduction

The goal of this review is to focus on how bilingual experience interacts with Attention-Deficit/Hyperactivity Disorder (ADHD) -related cognition and behavior, not only in terms of global advantages or disadvantages for such cognition and behavior, as has already been done, but through the lens of potential moderators such as socioeconomic status (SES), language background, typological distance of acquired languages, comorbidity with autism spectrum disorder (ASD), and developmental trajectory. More nuanced studies, we argue, can help to inform evidence-based guidelines, for doctors, therapists, teachers who still reportedly discourage bilingualism (i.e., the use of two languages) for neurodivergent children such as those with ADHD (Alfano et al., 2025). Moving forward, we point out that this narrative review does not consider the behavioral findings from only one age group with ADHD, but rather results from children, adolescents and adults are examined.

1.1 Bilingual effects on ADHD

Findings to date within the bilingual ADHD literature are mixed in regard to whether the addition of a second language or the presence of ADHD modulates executive functions (EFs) and language capabilities. Not only does the suspected origin of effect differ, i.e., bilingualism or ADHD, but studies have also demonstrated both advantageous and disadvantageous performances in EFs and language abilities in bilinguals with ADHD (Köder et al., 2022). It should be noted that the bilingual ADHD literature does not solely consist of studies which have recruited participants with ADHD diagnoses (Bialystok et al., 2017; Mor et al., 2015) but rather studies which have measured participants' attention abilities are also reviewed (Hardy et al., 2021; Sorge et al., 2016) as well as ADHD related behavior (Sharma et al., 2022).

Several studies have reviewed the performance of EFs in bilinguals with ADHD. EFs are commonly described as effortful mental processes which govern thought and goal directed behavior (Bunge, 2024; Diamond, 2013). They comprise of working memory (WM), inhibition (interference control and response inhibition) and cognitive flexibility/shifting (Miyake et al., 2000). Within Bialystok et al.'s (2017) group of young adults, Hardy et al.'s (2021) group of children and Lee and Hom's (2023) group whose ages ranged from children to mature adults, no advantage of bilingualism surfaced for inhibition in bilinguals with ADHD. ADHD and attention abilities were suggested by Bialystok et al. (2017) and Hardy et al. (2021), respectively, as underlying the observed weaker inhibition performances relative to groups either without ADHD (see Bialystok et al., 2017) or groups at lower levels of attention-problems (see Hardy et al., 2021). However, under certain conditions small bilingual advantages have surfaced, such as for visual processing at higher attention-problem levels (Hardy et al., 2021). Bilingual advantages in WM may only occur under specific modalities. Contradictory findings of a bilingual advantage in auditory WM have surfaced between a study focusing on children with ADHD (Sharma, 2019) and a study with an ADHD group ranging from childhood to adulthood (Lee and Hom, 2023). On the other hand, advantages have been found for both visual WM (Lee and Hom, 2023) and spatial WM in children (Sorge et al., 2016).

Advantages and disadvantages have been reported for bilinguals with ADHD on language tasks (e.g., vocabulary, verbal fluency) as well as several find null effects (Hardy et al., 2021; Lee and Hom, 2023; Bialystok et al., 2017; Köder et al., 2022). Structural language proficiency, which considers speech, syntax, semantics and coherence of language use, emerges as a stronger predictor of ADHD-related behavior ratings rather than bilingual status per se (Sharma et al., 2022). In adults with ADHD, pragmatic functioning may vary by language: hyperactive/impulsive behaviors can attenuate in the second (L2) or third language (L3), while challenges with inattention may increase in L3, underscoring the role of language context (Köder et al., 2024). Moderator factors such as anxiety add further heterogeneity (Gonzalez et al., 2023).

While recent syntheses have provided useful overviews, they mainly ask whether a “bilingual advantage” exists in ADHD (Köder et al., 2022; p.3). This framing risks obscuring the diversity of findings: some studies suggest compensatory effects under certain conditions, others report disadvantages on language-based tasks, and many show null results. This heterogeneity within ADHD-specific studies in not unique. As we will see in the following section, similar inconsistency has characterized the broader bilingualism literature in neurotypical populations, where reported effects have also varied according to task and study design.

1.2 Bilingual effects on neurotypical groups

Given that hypotheses and future directions presented within this review rely upon the existence of a bilingual advantage, it is necessary to even briefly bring to light the considerable heterogeneity within the bilingual literature as to the presence of any such advantage and its reliability. A meta-analysis by Grundy and Timmer (2017) revealed that children exhibited the most pronounced positive influence of bilingualism on WM, albeit no larger than medium effect sizes. However, when publication bias is adequately controlled for, the small bilingual advantage in EFs, which was found within the literature for children under the age of 18, was seen to diminish considerably. This was the case for both Lowe et al.'s (2021) and Gunnerud et al.'s (2020) meta-analyses which compiled 1,194 and 583 effect sizes, respectively. This trend also extends into the adult literature with results from Lehtonen et al.'s (2018) meta-analysis of 891 effect sizes not supporting the notion of a bilingual advantage in EFs. Furthermore, a number of primary studies have reported null results for a bilingual advantage in EFs (Dick et al., 2019; Paap and Greenberg, 2013; von Bastian et al., 2016). These meta-analyses found that positive effects surfaced most often in low precision studies (small sample sizes) (Lehtonen et al., 2018; Lowe et al., 2021), as well as null or negative results being underrepresented in such studies too (Lehtonen et al., 2018). Suggestions toward securing a more reliable observation of a bilingual effect included a shift to sample sizes greater than 20–30 participants, as well as controlling for SES, which this review shall discuss, as studies not accounting for SES had larger effect sizes (Lowe et al., 2021).

Although the question of a bilingual advantage in the realm of EFs is strongly contested, group differences between bilinguals and monolinguals appear to be more reliable on tasks requiring higher levels of attention (Comishen and Bialystok, 2021; Costa et al., 2009). As Bialystok (2025; p.4) noted in her review of the changes in understanding for the impact of bilingualism on cognitive processes over the years, “these results support the interpretation that it is degree of attention and not classification of task (i.e., WM, response inhibition, shifting) that determines whether there will be performance changes attributable to bilingual experience”. In line with this proposal, a meta-analysis by Adesope et al. (2010) comprising 63 studies investigating meta-linguistic awareness, meta-cognitive awareness, WM, abstract and symbolic reasoning, attentional control and problem-solving differences between monolinguals and bilinguals found that the largest between-group differences surfaced for attentional control. This is relevant as deviating attention profiles are noted for those with ADHD relative to neurotypicals. Selective attention, the capacity to filter out irrelevant stimuli in order to adequately focus on task-relevant information, has repeatedly been shown to be impaired in ADHD (Carter et al., 1995; Friedman-Hill et al., 2010; Klein et al., 2025). Attentional impairments are further reflected on the neurobiological level, as both stimulant and certain non-stimulant medications commonly prescribed for ADHD primarily target norepinephrine and dopamine neurotransmitters, with their benefit largely attributed to the enhancement of attentional focus (Del Campo et al., 2011; Parlatini et al., 2024).

1.3 Rationale

Taken together, the mixed findings reviewed above suggest that current evidence does not support the idea that bilingualism globally benefits or harms cognition or ADHD-related behavior. Rather, bilingual effects in ADHD are unlikely to be captured by simple advantage/disadvantage contrasts (Bialystok et al., 2017; Köder et al., 2022; Mor et al., 2015). This is perhaps unsurprising given the heterogeneity also observed in the literature on bilingualism in neurotypical populations. Such variability suggests that asking whether a bilingual effect exists in general may be less informative than asking under which conditions different outcomes emerge.

Unlike prior syntheses that focus primarily on the presence or absence of a bilingual effect (Köder et al., 2022), the present review therefore examines mechanisms and conditions, identifying when, how, and for whom bilingual experience may shape cognition in ADHD. As a narrative review, the aim is not to estimate a pooled effect size or conduct formal moderator analyses, but to organize evidence around key moderators often overlooked in previous work and may help clarify variability across findings. Firstly, the influential factor of SES is addressed, as it may affect language, cognitive, and attentional outcomes independently of bilingual status (Morton and Harper, 2007). We therefore review evidence suggesting that SES may influence the interpretation of bilingualism-related effects, including ADHD-related behavior (Bialystok et al., 2017; Mor et al., 2015; Sharma et al., 2022). Secondly, while bilingual research has progressed in recent years in understanding behavioral outcomes as a result of bilingual experience on a continuum, language distance should also be considered on a spectrum when analyzing such outcomes. Although research in this area remains limited, available evidence suggests that age of second language acquisition and linguistic distance between languages may shape developmental trajectories in EF- and attention-related outcomes (Gallo et al., 2024; Perovic et al., 2023; Radman et al., 2021). This further underlies the importance of measuring a speaker's language background as a multidimensional construct, rather than a simple monolingual/bilingual divide, thus strengthening precision as to the origin of the observed effect (De Cat et al., 2022a; Marian et al., 2007). Lastly, we broaden the ADHD profile by introducing the common comorbidity of autism spectrum disorder (ASD) (Hollingdale et al., 2020; Rong et al., 2021). Given that overlapping neurodevelopmental profiles may shape cognitive, communicative, and behavioral outcomes (Townes et al., 2023), comorbidity and developmental timing are considered as further dimensions that may clarify when bilingualism-related effects become observable. We hypothesize that longitudinal designs methodologically strengthen the observation of age-dependent bilingual effects for those with a broader ADHD profile, particularly amongst those with the combined presentation of ADHD relative to the inattentive or hyperactive-impulsive presentation (Grzadzinski et al., 2010). By foregrounding these dimensions, we aim to reframe the debate from one of global advantage to one of contextualized variability when EFs or ADHD-related behavior is the topic of interest in bilinguals with ADHD.

2 Methods

Literature for the core dimensions addressed in this paper was identified by the first author using Web of Science and Google Scholar by using Boolean strings such as (“attention” OR “ADHD”) AND (“attention” OR “vocabulary” OR “interference control” OR “executive function^*”), (“ADHD”) AND (“bilingual”), (“socioeconomic status”) AND (“bilingual^*” AND “attention”), (“socioeconomic status”) AND (“measure^*” OR “multidimension”), (“LEAP-Q” OR “Q-BEx”), (“linguistic distance” OR “executive function^*”) AND (“bilingual^*”), (“bilingual” AND “Igbo”), (“bilingual”) AND (“Chinese” OR “Mandarin”), (“autism” OR “autistic” OR “ASD”) AND (“bilingual^*”), (“ADHD” OR “attention-deficit/hyperactivity disorder”) AND (“autism” OR “autistic” OR “ASD”).

Peer-reviewed journal articles published since 2000 and relevant to the fields of developmental psychology, cognitive science, linguistics, bilingualism, and neurodevelopmental disorders were considered for inclusion. Studies were included if they reported group-level behavioral or neurocognitive measures related to the constructs under investigation. Both cross-sectional and longitudinal designs were considered. Case studies and case reports were excluded because the aim was to identify patterns that could be compared across participant groups and generalized across studies.

A number of studies were excluded because their primary focus fell outside the scope of the review, including research on infancy, caregivers, aphasia, intervention outcomes, and training effects. Studies whose primary focus was on real-time processing mechanisms rather than developmental, cognitive, or linguistic outcomes were also excluded. This included a number of eye-tracking studies that examined gaze patterns as potential markers of ASD or online attentional and working-memory processes in ADHD and were therefore not retained.

Depending on the aim of the search, additional keywords such as vocabulary size, longitudinal, inhibition, working memory, and shifting were subsequently used to refine the selection process. These criteria resulted in the inclusion of 42 studies. At each stage of the selection process, the first author conducted the screening, and inclusion decisions were reviewed by the second author. Any uncertainties regarding study inclusion were resolved through discussion. See Figure 1 for an overview of the selection process.

The EF of inhibition was singled out given the proposal by Barkley (2010) that for those with ADHD, deviations in inhibition performances underly weaknesses in a broader range of executive functioning. Furthermore, inhibition in ADHD has been widely studied with weak performances consistently surfacing (Metin et al., 2012; Mullane et al., 2009; Senkowski et al., 2023).

3 Results

Table 1 provides an outline of the studies included in this narrative review. While each study is recorded once, some are categorized according to only one of their contributory reasons within this paper. This means that some studies were also cited in relation to additional reasons not provided in Table 1. For example, although Peristeri et al. (2021) was placed under “Bilingualism and ASD/ADHD” it was also cited to highlight how longitudinal studies are necessary in observing changes in performance due to increasing age.

Table 1. Included studies.

Overview of included studies. The left-hand column references the authors of each study, the central column identifies the age group investigated in each study and whether they have any diagnoses, and finally the right-hand column summarizes the main or one of the main contributions each study gave in relation to the focus of the respective section.

ADHD, Attention-Deficit/Hyperactivity Disorder; ASD, autism spectrum disorder; EFs, executive functions; NA, not applicable; RTs, reaction times; SES, socioeconomic status; TD, typically developing; ToM, Theory of Mind; WM, working memory; y. o, years old.

3.1 Methodological caveats

The studies reviewed above highlight that findings are sensitive to methodological variation, particularly in how participant groups are matched and how bilingualism itself is defined and measured. Two issues are especially relevant in this respect: SES, which may confound the interpretation of group differences if not adequately controlled, and language background, since commonly used monolingual/bilingual labels may obscure substantial variation in language experience. In the following sections, we therefore discuss these two key methodological caveats.

3.1.1 Socio-economic status: group differences, independent effects, and operationalization

In investigating potential effects of bilingualism on ADHD related behaviors, SES requires particular attention because low levels of SES have been found to negatively influence receptive vocabulary and attention abilities in TD groups. This is relevant as monolinguals with ADHD have been noted to perform poorly on such measures of language skills, receptive vocabulary and attention abilities (Bruce et al., 2006; Klein et al., 2025; Kouvava et al., 2022; Mullane et al., 2009). In direct consequence of not controlling for SES, bilingual factors such as age of second language acquisition, L2 proficiency, frequency of L2 usage and contexts in which the L2 is used, may be misinterpreted as having more or less positive impact on the dependent variable, for example attention abilities in bilinguals with ADHD, than they actually do. Variances in this control procedure were made evident in a seminal paper by Morton and Harper (2007) who challenged the notion of a bilingual advantage in children given that SES is not always balanced across groups. In calculating the cost of performing the Simon task (incongruent—congruent) in terms of errors, pooled error scores of monolingual and bilingual children revealed significantly less errors for children of higher SES in comparison to children of a lower SES. Furthermore, when comparing monolingual and bilingual group performances in search for a bilingual advantage, one was not found as a result of having balanced SES between groups.

In a study by Bialystok et al. (2017), the effect of ADHD status on vocabulary knowledge, interference control and response inhibition was investigated in monolingual adults with ADHD (N = 38; M age = 22.1; SD ± 3.8), bilingual adults with ADHD (N = 41; M age = 23; SD ± 3.9) and compared to TD monolinguals (N = 44; M age = 22.1; SD ± 3.5) and TD bilinguals (N = 45; M age = 20; SD ± 1.6). Within this study the SES of the bilinguals with and without ADHD was higher than that of the monolingual groups. It is possible that the elevated SES of the bilingual groups modulated their ADHD symptoms as results of both ADHD assessment tests, the CAARS-S:L and BAARS-IV, exhibited better scores than those of the monolingual controls. While we argue here that SES may have modulated ADHD scores and present in commencing paragraphs evidence for different levels of SES affecting receptive vocabulary and attention abilities, it is also plausible that ADHD scores of the bilingual ADHD group were lower due to hyperactivity scores. The bilingual ADHD group exhibited significantly lower hyperactivity scores relative to the monolingual ADHD group on the BAARS-IV, with results from CAARS-S:L approaching significance, p = 0.09. Similarly, an analogous SES pattern was observed in a study by Sharma et al. (2022). Unlike Bialystok et al. (2017) who investigated the influence of ADHD on language skills and EFs in adults, Sharma et al. (2022) analyzed the effects of L2 experience and structural language skills in English on ADHD-related behavior in a group of monolingual (N = 230; M age = 8.4; SD ± 1.7) and bilingual (N = 164; M age = 8.1; SD ± 1.7) primary school children. Here, bilinguals had significantly higher scores of SES relative to monolinguals and there was no significant difference in the level of ADHD-related behavior between groups. SES was balanced in a study by Mor et al. (2015) who compared bilinguals with ADHD (N = 20; M age = 25.2; SD ± 2.2) and bilinguals without ADHD (N = 20; M age = 24.1; SD ± 2.1) to monolingual with ADHD (N = 20; M age = 24.4; SD ± 2.4) and monolinguals without ADHD (N = 20; M age = 24.3; SD ± 2.5). No difference surfaced between bilinguals and monolinguals on assessments of ADHD, measured by subtests of attention performance, only between ADHD and non-ADHD groups. Furthermore, they found no evidence of a bilingual advantage, with bilinguals with ADHD underperforming monolinguals with ADHD on interference control, measured by the Numeric Stroop and Simon arrows task.

Furthermore, low SES appears to have negative implications for receptive vocabulary size in TD monolingual and bilingual children (Calvo and Bialystok, 2014). While language difficulties are not a diagnostic requirement for ADHD, children with ADHD have been noted to have poorer language skills and receptive vocabulary in comparison to TD children (Bruce et al., 2006; Kouvava et al., 2022). Weaker vocabulary performances have also been observed in TD bilinguals (Bialystok et al., 2010; Calvo and Bialystok, 2014), which is suggested to stem from the fact that they split their time between more than one language, thus experiencing diminished exposure to each language relative to monolinguals. Given the effects of low SES on receptive vocabulary and that on the other hand ADHD and bilingualism can independently account for low vocabulary scores, failure to account for SES risks wrongly attributing vocabulary weaknesses to bilingualism or ADHD.

Additionally, previous studies have shown a modulatory effect of SES on attention abilities. This is of direct relevance for individuals with ADHD as they consistently display atypical patterns of attention in comparison to TD groups, such as difficulties in directing attention away from task-irrelevant information (Klein et al., 2025; Mullane et al., 2009). In a study by Mezzacappa (2004), an Attention Network task was employed in a group of children (N = 249; M age = 5.96, SD ± 0.46) who were subsequently divided into three SES groups, low, middle and high. Analysis of general task performance revealed faster response times for more socially advantaged children in comparison to those less advantaged. Furthermore, under the task manipulation where attention was alerted by cues, results revealed longer reaction times and poorer accuracy performances for the low SES group relative to higher.

Difficulties in directing attention toward task-relevant information have also been found in linguistic tasks. D'Angiulli et al. (2024) found that participants of lower SES exhibited poorer allocation of auditory attention toward task relevant consonant-vowel pairs in comparison to task-irrelevant consonant-vowel pairs. Additionally, while SES group differences are not always visible in behavioral findings, EEG measurements have further underscored group differences in TD children with variational involvement of neural regions when relevant and irrelevant auditory stimuli must be attended to (D'Angiulli et al., 2008). These findings present behavioral overlaps in low SES groups and those with ADHD, as well as fundamental differences in neural activations for differing SES groups. Behaviorally, those of low SES and those with ADHD do not struggle with the general engagement or activation of attention, but both groups continue to show difficulty prolonging attention during tasks that require selective attention (D'Angiulli et al., 2008; Klein et al., 2025; Schibli et al., 2017). Since SES appears to influence the development and efficiency of attention networks, there is the potential for modulatory effects in ADHD populations.

In having comparable values of SES between groups, Köder et al. (2022; p.12) point out that “the way SES was operationalized varied considerably between studies (investigating effects of bilingualism in those with attention deficits).” In Mor et al. (2015), SES was accounted for as participants' parents' average number of schooling years, whereas Bialystok et al. (2017) measured SES by mother's educational level. On the other hand, Sharma et al. (2022) and Mezzacappa (2004) both calculated a composite standardized index of SES but included caregivers differently. More specifically, Mezzacappa's (2004) SES index comprised the highest educational and occupational status and highest income of the primary caregiver, whereas Sharma et al.'s (2022) SES index considered the average educational level of both caregivers as well as scores from the Family Affluence scale (see Currie et al., 2008). While household education appears to pave the way for groupings in bilingual research, we encourage the reconceptualization of SES acknowledged by Antonoplis (2023) and others (see Eriksson et al., 2021) which moves from the concept of SES as a single measurable property to a set of socioeconomic conditions, such as income, education and occupation. Antonoplis (2023) suggests that conditions could be chosen depending on their theoretical relevance for a proposed study, however, we argue that the conditions should be included as multiple separate predictors to ensure comparability between studies, while also strengthening understanding of the effects of each predictor. For example, while parents' average years of schooling did not modulate interference control in bilinguals with ADHD relative to monolinguals with ADHD, the influence of parental income and occupation remains unexplored (Mor et al., 2015). We discourage the use of composite scores, for example, coupling income and education under a single SES term suggest that they are interchangeable, yet a person can be highly educated but of lower income, for example a teacher. Furthermore, Eriksson et al. (2021) demonstrated that composite scores of SES explain less variance in student achievement than when conditions of SES are addressed separately. The authors segmented the composite score used by the Programme for International Student Assessment (PISA) into six multidimensional predictors. These predictors included parents' highest occupational status, parents' highest educational level, wealth possessions, cultural possessions, home economic resources and books at home. The achievement of nearly 600,000 students in mathematics, science and reading from 77 countries and entities were included with the majority being born in 2002. Books at home and parents' highest occupational status had the largest effects on student achievement and so an index of these two items was made. Comparisons revealed that the PISA composite score was found to underestimate the amount of variance in student achievement relative to both the two-item index and the six multiple predictors.

Building on this point, future research should operationalize SES as a multidimensional construct rather than a single index. In line with Antonoplis (2023), income, educational attainment, and occupational status should be treated as distinct covariates and reported separately whenever possible. This enables comparability across studies and insight as to whether and how each condition interacts with bilingualism and ADHD-related cognition. If a composite SES score is nevertheless used, its components and weighting should be explicitly justified. Clear reporting of SES sources (e.g., parent report) and measurement units further enable comparability across studies, which helps disentangle the unique contributions of aspects of the socioeconomic environment from bilingual experience.

3.1.2 Need for multidimensional assessment of language backgrounds

Beyond SES, another methodological issue concerns differences in monolingual and bilingual labeling and the implications this may have for group outcomes. Mor et al. (2015) tested interference control by means of a Numeric Stroop and Simon arrows task in a group of L1 Hebrew monolinguals with ADHD (N = 20; M age = 24.4, SD ± 2.4) and without ADHD (N = 20; M age = 24.3, SD ± 2.5) to L1 Russian L2 Hebrew bilinguals with ADHD (N = 20; M age = 25.2, SD ± 2.2) and without ADHD (N = 20; M age = 28.8, SD ± 2.1). Evidence of a speed accuracy trade-off, which occurs when sufficient reaction times (RTs) within a task are achieved but at the cost of diminishing performance accuracy in order to sustain RT adequacy, surfaced on both tasks for monolinguals with ADHD relative to their TD peers. Bilinguals with ADHD, however, not only had poorer accuracy rates but also longer RTs relative to the bilingual control group. Although monolinguals with ADHD could sacrifice their performance accuracy to maintain comparable RTs to the TD control group, it is plausible that the longer RTs and weaker performance accuracy exhibited by the bilingual ADHD group were due to the addition of another language and thus, a greater burden on inhibitory control. However, the bilingual group could also have been considered as trilingual, while monolinguals were arguably bilingual. Notably, the bilingual and monolingual groups did not differ on their University Entrance Exam for English which would have been their third and second language, respectively. In a world with increasingly widespread L2 use and passive L2 exposure (Bice and Kroll, 2019; Cook, 2002), it is becoming more and more difficult to find a truly monolingual sample. Particular attention should be paid to participants' language backgrounds in order to be as concrete as possible as to whether the observed effects stem from factors related to the use of two languages or factors also affected by having a third language. Additionally, this awareness would also add to trilingual and multilingual research and its interaction with EFs and attention.

To move beyond broad categorical labels, future work should adopt multidimensional language profiling tools such as the Q-BEx¹ questionnaire or LEAP-Q² (De Cat et al., 2022a; Marian et al., 2007). Such instruments capture individual variation across key domains, including age of acquisition, proficiency in each language, frequency and context of use, literacy, and code-switching patterns. Various studies have successfully adopted these questionnaires in bilingual research (Bustos et al., 2024; Kubota et al., 2025; Verhoeven et al., 2024). As it stands a multidimensional approach to bilingualism in ADHD has not yet been conducted. Extending these questionnaires to bilingual ADHD research would allow more precise modeling of how specific aspects of bilingual experience relate to ADHD-related cognitive profiles. At a minimum, studies should report for each participant (i) the number of languages used, (ii) relative proficiency levels, (iii) dominant language in everyday contexts, (iv) age and mode of L2 acquisition, and (v) the interactional context (dual- vs. dense-code environments). Fine-grained measures of language experience not only clarify the mechanisms behind RT and accuracy differences but also permit replication across linguistically diverse samples.

3.2 Further considerations

In the following section we discuss additional factors we deem worthy of consideration for future bilingual research, namely linguistic diversity, linguistic distance, and comorbidity. In particular we focus on the importance of investigating diverse languages, possible modulatory effects of bilingualism on EFs in ADHD with comorbid ASD, as well as behavioral symptoms in those with comorbid ADHD-ASD.

3.2.1 Linguistic distance as a modulatory factor and recognition of linguistic diversity

There is also the matter that certain communities have been understudied. Antonoplis (2023) highlighted the fact that despite many individuals from sub-Saharan African communities living in Europe, they have been understudied within the bilingual literature. If the aim of bilingualism is to understand all potential capabilities of the bilingual mind, then inclusion of all linguistic diversities is necessary.

While bilingual research has investigated bilingual effects stemming from the use of a tonal first language (L1), such as Mandarin or Cantonese, in combination with a non-tonal second language (L2) like English or another tonal L2 (Abutalebi et al., 2015; Ma et al., 2020; Xie et al., 2024), far fewer have focused on tonal languages from sub-Saharan regions. While there is, although limited, bilingual research on the largest tonal language spoken in West Africa, Igbo, it has not contributed to understanding neurocognitive alterations or behavioral performances in bilingual speakers with L1 Igbo. Rather, sociolinguistic impacts of bilingualism have been assessed (Obasi, 2024), alongside the role of Igbo as a code-switching language (Ihemere, 2016). Given the migration pattern of people from the African continent to Europe in seek of work or education, the influence of these languages should also be investigated alongside other European or tonal languages to better understand the diverse role of linguistic distance (LD) upon neurocognitive and behavioral performances in bilinguals. With that, as bilingual research in diagnosed ADHD populations is relatively recent, studies have not yet explicitly sought out linguistically distant language pairings to test the modulatory effects of such language combinations on the performances different EFs in bilinguals with ADHD relative to bilinguals with ADHD whose languages have a smaller LD. Currently, there is reason to believe that LD has a modulatory role in the effects of bilingualism on EFs. Firstly, it is useful to clarify what LD entails. LD can be defined at several levels: (1) genetic or typological distance (family relationship, morphological complexity, phonological overlap); (2) structural or orthographic similarity (shared scripts, cognate density); and (3) usage-based or experiential distance, reflecting how often both languages are used interchangeably in daily life and begin to change over time (Bialystok, 2007; Bybee, 2023; Laketa et al., 2021). In a comparison of early balanced and age-matched bilinguals, 19 Slovak-Serbian bilinguals (typologically similar), 21 Hungarian-Serbian bilinguals (typologically dissimilar) and 71 Serbian monolinguals completed a battery of EF tasks (Perovic et al., 2023). Group comparisons revealed advantageous performances for Hungarian-Serbian bilinguals relative to the other groups. On an Attention Shifting Task they exhibited enhanced monitoring abilities as seen by decreased global reaction times and smaller global switching costs. Furthermore, inhibitory control was assessed using the Stockings of Cambridge task, which required participants to suppress inappropriate movements while replicating a pattern displayed on a screen while adhering to specified rules. Hungarian-Serbian bilinguals exhibited significantly reduced initial thinking time prior to executing solutions for simple problems, suggesting enhanced planning efficiency. As this study controlled for age of acquisition, language proficiency, language use, education and immigration status, it is suggested that the advantageous performance of Hungarian-Serbian bilinguals was due to greater cognitive effort as there was a larger lexical, phonological and grammatical distance between their languages relative to the Slovak-Serbian bilinguals (Branzi et al., 2020).

While LD has been linked to increased control demands, recent findings suggest mixed and stage-dependent patterns, with advantages sometimes emerging for closer language pairs (Radman et al., 2021; Gallo et al., 2024), pointing to linguistic distance as a dynamic rather than uniform moderator. Gallo et al.'s (2024) bilingual sample (N = 34; M age = 63.9, SD ± 3.2) consisted of speakers of various L1s but all with L2 English, who completed multiple background questionnaires and a Flanker task. L1s consisted of Spanish, Hungarian, Polish, Italian, German, Portuguese, French, Swedish, Greek, Arabic, Dutch, and Estonian. Linguistic genetic proximity software, a database comprising of 220 languages, was used to calculate the LD between languages based on their genetic proximity. The mean output was 51.9 (SD ± 16.8) with Swedish having the lowest LD (26.7) and Hungarian the highest (87.9). L2 English proficiency was tested using the online Cambridge Test for Adult Learners. Analysis of Flanker task performance revealed a positive bilingual effect with reduced RTs for speakers of more distant L1 and L2 pairings but only when L2 proficiency was low and had their L2 been acquired in recent years. Given that an individual's L2 was more distant than their L1, this bilingual advantage only appeared to surface during the early stages of their L2 journey, i.e., recently acquired L2 and thus low proficiency. However, the longer a speaker possesses a distant L2 relative to their L1 and the more fluent they become in their L2, the bilingual effect diminishes as seen by longer RTs on the Flanker task. Interestingly, inverted results surfaced for languages of a smaller LD, with reduced RTs observed when speakers had acquired their L2 a long time ago and also had higher L2 proficiency levels. Similar findings also surfaced in a study by Radman et al. (2021). Here, bilinguals (N = 29; M age = 27.1, SD ± 6) of linguistically closer languages, French-English, displayed shorter RTs on a nonlinguistic switching task relative to Persian-English bilinguals (N = 29; M age = 22.2, SD ± 3). Thus, in both studies it appears that the L2 is the influential factor affecting EFs, as observed from non-linguistic task performances. Linguistic distance from the L1 differentially moderates the degree of influence upon EFs given i) various L2 proficiency stages as well as ii) various lengths of time since an individual learned the language. In summary, advantageous effects of LD may be present when L2 representations are not yet solidified, thus validating less interference, whereas typological proximity may only warrant such benefits once L2 representations are sufficiently strengthened. However, early acquisition, i.e., before age three, of linguistically distant languages may positively influence interference control processes when a high proficiency level has also been obtained (Perovic et al., 2023).

Across studies, LD has shown dynamic effects: greater distance can heighten control demands during early L2 acquisition, whereas closer languages may support faster switching once both systems are automatized. These stage-dependent patterns suggest that LD is not a fixed trait but a variable interacting with proficiency, recency of L2 use, and cognitive control. For individuals with ADHD, who already display weaknesses in interference control (Mullane et al., 2009) and attentional shifting (Salehinejad et al., 2020), future research should test whether moderate or high LD pairings could provide compensatory training effects on these executive domains. One suggestion would be to compare early ADHD bilingual speakers of two distant languages to aged-matched late ADHD bilinguals with the same language pairings. Specifically, to see whether high exposure and frequency of use from birth attenuates inhibition and switching and if the same effect can be found for ADHD bilinguals with a more recently acquired L2. Furthermore, given that many tonal languages of the sub-Saharan region are also typologically dissimilar to Indo-European languages and that many from such communities have emigrated to Europe, it seems fitting to include such language pairings as well. This would encourage close work alongside community members while simultaneously addressing the under-representation of African linguistic communities in bilingual research.

3.2.2 Bilingual effects in ADHD with co-occurring ASD

Recognition of comorbid ADHD and ASD has changed over time, as it used to be the case that their dual diagnosis was prohibited; however, the revised Diagnostic and Statistical Manual of Mental Disorders, DSM-5, now enables such diagnoses (American Psychiatric Association, 2013). ASD is also a neurodevelopmental disorder, however, its clinical phenotype differs in that individuals struggle to interact and communicate while also exhibiting repetitive or restricted behaviors and/or activities (American Psychiatric Association, 2013). ADHD is itself clinically heterogeneous and can present as inattentive, hyperactive/impulsive, or combined (inattentive and hyperactive/impulsive). Inattentive symptoms typically include distractibility, forgetfulness and difficulties organizing and planning, whereas hyperactive/impulsive symptoms tend to emerge as physical restlessness, fidgeting, and talkativeness (Schiros et al., 2023). This distinction is relevant for interpreting the studies reviewed below on how ASD symptoms relate to different ADHD presentations.

Research on this comorbidity has grown in recent years not only due to increasing prevalence rates but also the individual, societal and family challenges it poses. Furthermore, for monolingual children and adolescents of both disorders, they appear to share common deviations in EFs in comparison to their TD peers. In comparison to a group of TD controls, results of a meta-analysis highlighted significantly worse performance by a pure ASD group and a pure ADHD group on tasks testing response inhibition, cognitive flexibility, attention and visuospatial abilities. This meta-analysis was comprised of 58 studies where groups had a primary ASD or ADHD diagnosis according to the DSM or the International Classification of Diseases. Internally, all studies included participants under the age of 19 and were age and gender balanced (Townes et al., 2023). The question of interest here is not only whether bilingualism could help mediate some of the cognitive challenges imposed by ADHD but also for individuals with comorbid ADHD and ASD.

Building on this, the present review advances the hypothesis that bilingual experience may buffer certain executive and communicative vulnerabilities associated with ADHD–ASD comorbidity. Specifically, managing two linguistic systems could engage attentional shifting, updating, and inhibition mechanisms that are often compromised in both conditions, potentially yielding compensatory benefits over time. These effects are likely to depend on an individual's developmental stage and bilingual profile, emerging most clearly in the inattentive and combined presentations of ADHD, where deviations in executive control are most pronounced in comparison to TD individuals. Testing this hypothesis will require longitudinal designs capable of capturing how bilingual exposure interacts with maturation and comorbidity severity.

Findings in the bilingual literature display conflicting outcomes for those with ADHD, with some highlighting the addition of bilingualism to be a burden in comparison to TD individuals (Mor et al., 2015) while others maintain that bilingualism neither improves nor worsens attention abilities in children and adults with ADHD (Hardy et al., 2021; Köder et al., 2022). However, there is evidence to suggest that bilingualism may have a positive impact on second-order Theory of Mind (see Bianco et al., 2025 for information on second-order Theory of Mind) for those with ASD (Andreou et al., 2020; Montgomery et al., 2021; Peristeri et al., 2021). In a longitudinal study of 6 years (Peristeri et al., 2024), first- and second-order Theory of Mind abilities were tested in monolingual Greek (N = 21) and bilingual Albanian-Greek (N = 21) children with ASD. Groups were matched for age, verbal IQ and SES and were tested 3 times over the 6-year period, at age six, age nine and age 12. All testing occasions consisted of assessments of expressive vocabulary, WM measured by a n-back task, first-order Theory of Mind measured by the Sally-Ann task and the Smarties test and second-order Theory of Mind measured by children's responses to two stories in which a character x had false beliefs about the beliefs of character y. Additionally, verbal IQ and performance IQ were continuously tested as previous research has highlighted modulatory effects of general intelligence, thus verbal and non-verbal, on Theory of Mind performances. Consistently, at each testing occasion bilinguals with ASD significantly outperformed monolinguals with ASD on performances of WM. Furthermore, in comparing group performances in second-order Theory of Mind, effects of age were observed. Only on the second and third testing occasion bilinguals with ASD were found to significantly outperform monolingual controls. This highlights the sensitivity of longitudinal studies in demonstrating how increases in age modulate bilingual effects within a group, suggesting the emergence of bilingual effects at certain developmental time-points. Goh et al. (2020) also highlighted the observation of bilingual effects longitudinally with differing degrees of bilingual language exposure yielding contrastive results for the likelihood of an ADHD diagnosis at age 54 months. The study included 408 toddlers of Chinese, Malay or Indian ethnicity and bilingual exposure was measured as the percentage of time (out of 100%) each child was exposed to languages in their environment before the age of 24 months. The bilingual exposure score reflected the percentage of time exposed to the least frequent language for each child. Furthermore, language delay was assessed based on expressive and receptive language scores from the Bayley Scales of Infant and Toddler Development test. The test was administered in the child's dominant language, however, test items were administered in a child's non-dominant language if examiners felt items were misunderstood. It was found that children with a language delay at 24 months and who were exposed to more than one language at a high percentage, i.e., primarily exposed to dual-language environments, were less likely to receive an ADHD diagnosis at 54 months relative to children with a language delay at 24 months and primarily exposed to one language. With a possible bilingual advantage for those with ASD, the following section highlights the prevalence rates of comorbid ASD to a primary diagnosis of ADHD, prevalence rates of comorbid ADHD to a primary diagnosis of ASD, as well as which ADHD presentations may be more susceptible to comorbid ASD.

Thus far, there appears to be a higher likelihood of ADHD symptoms in those with ASD rather than ASD in ADHD (Antshel et al., 2016; Grzadzinski et al., 2016). On the one hand, 53 from a total of 63 studies in a meta-analysis estimated that 38.5% of individuals with ASD are also experiencing ADHD, and that 13 of the total 63 studies estimated 40.2% of individuals with ASD were expected to experience ADHD symptoms within their lifetime (Rong et al., 2021). On the other hand, results of a meta-analysis consisting of 13 studies estimate that 21% of children and adolescents with ADHD meet the criteria for ASD (Hollingdale et al., 2020). Findings also suggest that comorbid ADHD in those with ASD can exacerbate their social disabilities, measured by the Social Responsiveness Scale (SRS), with reports of greater autistic symptoms in this group relative to those with only an ASD diagnosis (Sprenger et al., 2013). Regarding the prevalence of ASD in those with ADHD, it appears that those with the inattentive or combined presentation of ADHD are more likely to exhibit ASD symptoms. In controlling for age, IQ and comorbidity rate, Grzadzinski et al. (2010) compared an ADHD group with traits specific to ASD to an ADHD group who did not exhibit such traits (total N = 75, 7–17 years of age). The authors found that 67% of participants who displayed autistic traits were diagnosed with the combined presentation of ADHD, whereas 21% of participants from the same group had the inattentive presentation of ADHD. For comparison, 43% of the ADHD group without autistic traits had the combined presentation of ADHD and 43% the inattentive presentation. In line with these findings, in a sample of 352 participants aged 0–17 years old who held an ADHD and ASD diagnosis, 42.7% had a combined ADHD presentation while 27.4% had an inattentive presentation. It should be noted that parents of 24.2% of participants reported no ADHD presentation, however, after controlling for age and gender, analyses confirmed that children with ADHD and ASD were more likely to have a combined ADHD presentation (Zablotsky et al., 2020). An explanation as to why the hyperactive/impulsive presentation alone is less prone to a comorbid ASD diagnosis may stem from the fact that hyperactivity in ADHD strongly appears to decrease with age (Döpfner et al., 2015).

Similar to comorbid ADHD exacerbating social disabilities in those with ASD (Sprenger et al., 2013), children with ADHD (N = 821) with the highest traits of ASD were found to be more likely to struggle with reading, language and/or motor disorders (Mulligan et al., 2009). It appears clear that this bidirectional comorbidity creates a burden whereby each disorder exacerbates the other. Given that bilingualism may positively influence naturally deviated EFs in ASD relative to TD individuals (Peristeri et al., 2024), similar benefits could potentially emerge in comorbid ASD with primary inattentive or combined ADHD, potentially alleviating the substantial language, social, and personal developmental challenges faced by this population. Taken together, these findings underline the need to interpret bilingual effects in ADHD not in isolation, but in light of co-occurring neurodevelopmental profiles such as ASD.

4 Discussion

The relationship between bilingualism and ADHD remains contested, with studies showing conflicting results ranging from additional burden (Mor et al., 2015) to no significant impact on EFs (Hardy et al., 2021; Köder et al., 2022). As we have seen, emerging studies suggest more nuanced outcomes than either a benefit of bilingualism or the absence of such a benefit. Rather, they point toward the possibility that bilingual effects may be shaped by context, language use, and developmental profiles. These potential modifiers would therefore require further attention for their influence to be better understood. The central argument of this review is thus that the effects of bilingualism in ADHD cannot be captured by global advantage or disadvantage claims. Instead, outcomes may vary according to contextual moderators such as SES, linguistic experience and distance between languages, comorbidity profiles, notably ADHD and ASD, and developmental timing. By identifying these moderating factors, the present review aims to reframe the debate from one of uniform to one of contextual variability.

Several methodological limitations may help explain the inconsistencies and the lack of comparability within the bilingual ADHD literature. These include inadequate control of SES, which independently influences vocabulary development and EFs (Calvo and Bialystok, 2014; Morton and Harper, 2007) as well as influencing the magnitude of reported bilingual effects (Lowe et al., 2021), lack of consensus regarding how SES should be operationalized, and imprecise characterization of participants' language backgrounds. Understanding the independent contribution of these factors to bilingualism-related outcomes is crucial for providing evidence-based guidance to caregivers, doctors and therapists, who often recommend rearing children with neurological disorders in monolingual environments due to concerns about language delay (Alfano et al., 2025). To refine theory and guide practice, future studies should consider the following steps. The factor of SES should be re-examined and considered a multidimensional construct consisting of income, education and occupation. This shift is necessary in order to better understand the influence, if any, each component has on a study's focus (Antonoplis, 2023). Failure to adequately control for SES increases the risk that ADHD symptoms, EF performance and performances on receptive vocabulary can be attributed to factors other than SES (Bialystok et al., 2017; Bruce et al., 2006; Kouvava et al., 2022; Morton and Harper, 2007; Sharma et al., 2022). Furthermore, controlling for SES decreases the likelihood of bilingual effects being wrongly attributed to language status rather than SES itself (Morton and Harper, 2007). This moderating factor needs to be kept in mind so as to avoid false attributions to bilingualism (Lowe et al., 2021), which may further contribute to difficulties in replicating bilingual advantage effects (Dick et al., 2019; Gunnerud et al., 2020; Lehtonen et al., 2018; Lowe et al., 2021; Ware et al., 2020). On this matter, we also point out the limitation that some of our cited studies which displayed positive effects of bilingualism contain low precision estimates (Andreou et al., 2020; Peristeri et al., 2024). There would also be the question as to whether effects of linguistic distance (Gallo et al., 2024; Perovic et al., 2023; Radman et al., 2021) would have emerged differently if participant groups were larger in size. To enhance ecological validity, we likewise advocate for larger sample sizes (Lehtonen et al., 2018; Lowe et al., 2021) and commend the collaborative approach taken by authors over the years, as recruiting sufficient suitable participants for any study is no easy feat.

Future studies should also take care to measure bilingual experience multidimensionally. De Cat et al. (2022b; p. 112) underscore that there is currently “an absence of consensus on how to quantify bilingualism”. Relevant dimensions include, among others, education and literacy, language input quality and language mixing practices. Stricter assessments of participants' language profiles by incorporating a Q-BEx questionnaire or the LEAP-Q (De Cat et al., 2022a; Marian et al., 2007) can shed light on the potential effects of more fine-grained components of bilingualism. Accounting for these heterogenic factors between bilinguals can illuminate the potential effects of more fine-grained components of the bilingual experience on specific EFs (Hartanto and Yang, 2020), while mitigating uncertainties that lower accuracy rates and longer reaction times in bilingual ADHD groups may have stemmed from the general knowledge of additional languages (Mor et al., 2015). Given that effects of bilingualism develop and change over a period of time (Gallo et al., 2024), longitudinal studies better highlight these emerging observations, a factor which cross-sectional study designs may miss. Emerging longitudinal studies are also accentuating time-dependent effects in neurodevelopmental populations, showing that bilingual effects may surface only at certain developmental stages (Peristeri et al., 2024). This underscores the need for time-sensitive designs in ADHD research to identify when and to what degree bilingual exposure may scaffold cognitive trajectories. Lastly, to add to the emergence of bilingual effects, we recommend that future studies explicitly note the language of testing. This transparency enhances comparability between studies as bilingual effects on EFs have been noted to be more pronounced when an experiment is conducted in a participant's L1 in comparison to their L2 (Grundy and Timmer, 2017).

In addressing participants' language profiles, the effect of LD on EFs is still an open question with benefits found to surface from linguistically distant language pairings (Perovic et al., 2023) and linguistically similar language pairings (Gallo et al., 2024; Radman et al., 2021). However, whether these compensatory effects extend to EFs in bilinguals with ADHD has not yet been explicitly examined. Furthermore, with recent evidence suggesting a bilingual advantage for those with ASD (Peristeri et al., 2024), there is the potential that ADHD with comorbid ASD may profit from such a pairing. This investigation would be particularly relevant for the inattentive and combined presentations of ADHD which appear to be more prone to comorbid ASD (Grzadzinski et al., 2016; Zablotsky et al., 2020). To that end, in recognition of the complex relationship between ADHD and ASD and the effects they pose upon the traits of an individual's primary disorder (Mullane et al., 2009; Sprenger et al., 2013), research designs should treat both ADHD and ASD symptoms as continuous rather than categorical variables. This dimensional approach to neurodevelopmental conditions exemplifies the broader methodological shift needed across all variables in bilingual ADHD research, moving from binary classifications toward nuanced, multidimensional frameworks that capture the true complexity of linguistic diversity.

Together, these observations point to a clear agenda for future research. First, SES should be modeled as a multidimensional construct (income, education, occupation) rather than a single index, allowing researchers to isolate which components interact with bilingualism in ADHD. Second, bilingual experience should be quantified with standardized multidimensional tools such as Q-BEx or LEAP-Q to capture proficiency, frequency, context of use, and code-switching practices. Third, longitudinal designs spanning key developmental windows, from early schooling to adolescence, are crucial to reveal age-dependent and time-sensitive bilingual effects that cross-sectional snapshots miss. Fourth, linguistic distance should be explicitly measured, using genetic and structural metrics, to test whether distant or close language pairings differentially recruit control mechanisms in ADHD. Finally, ADHD and ASD should be treated as spectra rather than categorical groupings, acknowledging symptom continuities that may modulate bilingual effects on cognition and communication. Overall, this shift from binary comparisons toward multidimensional modeling may help future research clarify the conditions under which bilingual experience is associated with cognitive, linguistic, and behavioral outcomes in ADHD.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Pui Fong Kan, University of Colorado Boulder, United States

Reviewed by: J Bruce Morton, Western University, Canada

Inmaculada Mendez-Freije, University of Oviedo, Spain

Footnotes

The Q-BEx (Quantifying Bilingual Experience) is a questionnaire designed with input from researchers, teachers and speech and language therapists (SLTs) as to areas of language experience that should be measured. It is an online tool that can be shared by a link (URL) with its current availability in 33 different languages. The questionnaire exists in a comprehensive form, however, it is customisable to accommodate various research aims for both children and adults. Structure dependent, its administration can last from approximately 10 minutes to 45 minutes. It is free to use and accessible on the Q-BEx website which also lays out customisation steps.

The LEAP-Q (Language Experience and Proficiency Questionnaire) was designed to gather self-reported proficiency and language experience data to assess one's proficiency in a language acquired in addition or alongside their first language. It is freely available in 45 different languages, as well as for various dialects, on the Bilingualism and Psycholinguistic Research Laboratory website.