The term fundamental motor skill was defined by Wickstrom (1977) as a basic motor activity for more advanced and highly specific activities such as running, jumping, and throwing, among others. Currently, the term fundamental motor skill reflects various terminologies that have been used in the literature such as motor proficiency, motor performance, fundamental movement ability, fundamental motor skill, motor skill and motor competence (Robinson et al., 2015).

Various terms have been used to describe fundamental motor skills, such as: gross motor skills (Pang and Fong, 2009; Mostafavi et al., 2013), fundamental motor patterns (Barnett et al., 2012), and fundamental movement skills (Staples and Reid, 2010; Barnett et al., 2015). Similarly, various terminologies have been commonly used to describe fine motor skills, such as: fine motor proficiency, fine motor accuracy, fine motor integration, manual dexterity, or fine motor coordination (Bruininks and Bruininks, 2005) or performance motor skills (Sortwell et al., 2022).

In view of the numerous terminologies found in the literature, the present systematic review is based on the terminologies used in the reviews published by Macdonald et al. (2018), Donnelly et al. (2016) and van der Fels et al. (2015), due to the fact that they present similar objectives. These reviews used the term Motor Skills and the underlying domains as gross motor skills and fine motor skills.

Motor skills (MS) refer to efficient and effective actions resulting from a learning process (Magill, 1984). According to movement control and precision, they are divided into two categories, gross motor skills (GMS) and fine motor skills (FMS). These two categories of motor skills are used in research to analyze the relationship between cognitive and motor development (Grissmer et al., 2010; Gentier et al., 2013; Raisbeck and Diekfuss, 2015; Van der Fels et al., 2015; Oberer et al., 2017; Macdonald et al., 2018; Haywood and Getchell, 2019).

GMS primarily use movements produced by large muscle groups. They include motor skills that imply movement of the body in space, such as walking, running, jumping, sliding, and postural skills, which refer to the ability to keep a controlled position or posture during a specific task or activity (they can be dynamic or static) and manipulative skills used to control objects in actions such as grabbing, hitting, absorbing, lifting, etc., with hands, feet or using other objects for that purpose (Ulrich, 2000; Grissmer et al., 2010; Lopes et al., 2013; D’Hondt et al., 2014; Magistro et al., 2015; Rudd et al., 2015; Chang and Gu, 2018; Haywood and Getchell, 2019). On the other hand, FMS can be defined as movements produced by small muscle groups that involve activities with great precision, implying two distinct capabilities, motor coordination, and visual integration, as well as the integration of both (Carlson et al., 2013). In this context, different types of FMS can be identified depending on the capabilities involved (Kimmel and Ratliff-Schaub, 2011; Carlson et al., 2013). A type of FMS consists of fine motor coordination (FMC), which refers to movements that involve oculus-manual coordination (eye-hand), manual dexterity, motor sequencing, and speed and precision, such as tracing, touching with fingers, building with Legos/blocks, moving coins from one place to another or inserting them into a slot, etc., which can also be called non-graphomotor skills (Davis and Matthews, 2010; Suggate et al., 2018).

Another type of FMS consists of visual motor integration or visuomotor integration (VMI) and spatial or visuospatial integration (VSI), which refers to the organization of small muscle movements of the hand and fingers through the processing of visual and spatial stimuli, more based on synchronized hand-eye movements (Sortor and Kulp, 2003; Carlson et al., 2013; Fuhs et al., 2014; Kim et al., 2018), typically writing, drawing, copying shapes, letters, or other stimuli (Beery and Buktenica, 1997; Grissmer et al., 2010; Roebers et al., 2014; Oberer et al., 2017) that can be called graphomotor skills (Davis and Matthews, 2010; Suggate et al., 2018). Figure 1 summarizes the categories of motor skills.

The performance of academic skills preferentially values two curricular areas, literacy and mathematics performance (Fernandes et al., 2016; Ribner et al., 2017). These two areas are considered prerequisites for performance in other subjects and, consequently, for academic success (Organization for Economic Cooperation and Development (OECD), 2016). However, mathematics plays an important role in the school curriculum and its development. In a modern, technological society, it is seen as a fundamental cognitive attribute, where successful early learning provides not only a framework for later learning (Duncan et al., 2007) but is also an indicator of future academic and professional success (Parsons and Bynner, 2005).

Mathematics is learned by children before school through numbers and quantities (Blair, 2002; McWayne et al., 2004) and the informal knowledge they acquire is often referred to as “basic numerical skills,” being a precondition for mathematical reasoning (Gersten et al., 2005; Jordan et al., 2006). Studies have shown that math skills in preschool education predict performance in reading, math, and science through grade 8 (children between 13 and 14 years old; Claesens and Engel, 2013).

The math skills children acquire in preschool education are important for developing a conceptual understanding of mathematics (Jordan et al., 2009; Siegler et al., 2012), as well as confidence in the ability to engage in activities that support analytical thinking, problem solving and reasoning and argumentation skills (Clements et al., 2004). However, teaching mathematics in preschool education must be related to children’s day-to-day interests, such as playing or exploring everyday situations (Silva et al., 2016).

It is recognized that pre-school children enjoy activities that develop their math skills (Ginsburg et al., 2006). However, most early childhood educators typically place greater emphasis on children’s social–emotional and literacy development and less attention to mathematics (National Research Council, 2009). Many early childhood educators avoid teaching mathematics because of their own negative experiences with it (Ginsburg et al., 2006; Clements and Sarama, 2007).

Fundamental learning for the development of mathematical competences in preschool education consists of two main areas: Numbers (1) and Geometry and Measurement (2). Each of these areas consists of different skills. The area of numbers is subdivided into the core of numbers (counting, cardinality, and identification of numbers), operations (addition and subtraction) and relations (comparing). Regarding the area of geometry and measurement, geometry consists of shapes and space, and measurements of length, area, and volume (National Research Council, 2009; Figure 2).

Numbers are abstractions that apply to a wide range of real and imaginary situations. These do not exist in isolation, but constitute a system of relations and operations by which they can be compared, added, subtracted, multiplied, and divided. These relationships apply to a wide variety of problems (National Research Council, 2009). On the other hand, geometry and measurement provide systems for describing, representing, and understanding the world. Geometry is the study of shapes and spaces (two-dimensional-2-D and three-dimensional – 3-D). Measurement is about determining the size of object shapes (National Research Council, 2009). In this sense, preschool children need to learn the following math skills (National Research Council, 2009):

Core of numbers (cardinality, counting, and identification);

Cardinality – Children learn the concept of cardinality when they understand that adding an object means counting to the next number (Sarnecka and Carey, 2008).

Counting – Counting means listing the count numbers in order, usually starting at 1. It is a way of making a 1 to 1 correspondence between each object (Wynn, 1992).

Number Identification – It is the ability to associate a written number (e.g., 5) with a verbal word (e.g., five; Jordan et al., 2009).

Relationships (comparing);

Higher Institute of Education and Sciences of the Douro, Sports Department, Penafiel, Portugal, Penafiel, Portugal.

Comparing – represents the comparison of quantities of groups of objects using words such as “more/bigger,” “less/smaller” and “equal.” A basic way to compare two quantities of objects is by direct correspondence. If a child has a noticeably larger set of black beads compared to a set of white beads, the child identifies which group has the larger amount of beads. Thus, these skills can be developed separately from other basic math skills, such as counting and cardinality and number identification, because it is not necessary to know the exact number of objects in each group to successfully compare two groups (Traverso et al., 2021).

Operations (addition and subtraction);

Addition and subtraction – refer to basic arithmetic skills such as adding and subtracting and are used to relate quantities. Children are only prepared to develop these skills when they understand the concepts of cardinality and counting. These skills prepare children to develop more complex arithmetic skills such as multiplication and division (Barth et al., 2008; Canobi and Bethune, 2008).

Geometry (shapes and space).

Shapes – Shape is the basic way children learn to name objects (Jones and Smith, 2002). Children have an innate and implicit ability to recognize and match shapes (Anderson, 2000).

Space – Space includes two main skills: spatial orientation and spatial visualization of images. Spatial orientation involves knowing where you are and how to move around in the world (Gelman and Williams, 1997). Children learn words such as “beside” and “between.” Later, they learn words referring to frames of reference, such as “in front of,” “behind.” The words “left” and “right” are learned much later, and are a source of confusion for several years (Gopnik and Meltzoff, 1986). In these early years, children can also learn to analyze a route through a space (Wang and Spelke, 2002). Spatial visualization of images is about understanding and performing imagined movements of 2-D and 3-D objects. This requires being able to create a mental image and manipulate it through a close relationship between these two cognitive abilities. Spatial visualization of images has been positively associated with the construction and composition of shapes (Sarama et al., 1996).

Measurement (length, area, and volume).

Length Measurement – quantifies the distance between points in object or space.

Area Measurement – is a quantity of 2-D surface area that is contained within a boundary.

Volume Measurement – volume introduces even more complexity by the addition of a third dimension (3-D), presenting a significant challenge to students’ spatial structuring (Curry and Outhred, 2005).

One way to more formally assess children’s understanding of measurements is through comparison tasks (Mullet and Paques, 1991).

The literature has pointed out a positive association between mathematics and motor skills (Macdonald et al., 2018) and, according to the theory of “Embodied Cognition” (Embedded Cognition), cognition emerges from the “coupling” (embodied relationship) of the individual with the physical and social context as a result of sensorimotor activity (Smith, 2005). According to this paradigm, body movement causes changes in neural networks, stimulating significant gains in cognition. Within this paradigm, several investigations have highlighted the importance of movement in cognition, particularly in the performance of math skills such as abstract cognitive representations in general, and in the improvement of basic numerical representations in particular (Link et al., 2013). As proposed by Fischer et al. (2018), numbers are embodied concepts and not abstractions dissociated from sensory experiences. In addition, the theory assumes that certain cognitive and motor areas of the brain are activated simultaneously when solving mathematical problems (Fischer and Brugger, 2011).

Much research has been produced in an attempt to demonstrate the association between MH and cognition, but few have considered different categories of motor skills (gross/fine) and different categories of cognitive skills (executive functions/academic success in reading and mathematics; Oberer et al., 2017). Depending on the different categories of variables studied, the results differ and it is not possible to reach conclusive data (Magistro et al., 2015; Veldman et al., 2019). Furthermore, each study analyzed different motor skills and academics, performed them in different ways, or evaluated populations with different characteristics. All these aspects contribute to a disparity of results in this area (Magistro et al., 2015; Veldman et al., 2019). It is, therefore, a complex area of study that does not allow consensual conclusions (Escolano-Pérez et al., 2020). In this sense, it is considered necessary to study the associations between the different categories of specific motor skills and academic skills, in order to contribute to the success of children’s learning.

The subject of mathematics was selected for this study for the following reasons: first, because it is a “universal language” (across all countries); second, because it is the subject in which many school-age children have difficulties in learning, a problem with incidence ranging from 3% to 7% (Shalev et al., 2005; Swanson et al., 2009; Devine et al., 2013, 2018). These difficulties occur more than expected (Koponen et al., 2018; Willcutt et al., 2019) and are already observed at preschool age (Desoete et al., 2012; Devine et al., 2018); third, in light of the priority for children to develop basic numeracy skills upon entry to grade 1 (Rimm-Kaufman et al., 2000; Verdine et al., 2014; Duran et al., 2018; Cameron et al., 2019; Nesbitt et al., 2019; Escolano-Pérez et al., 2020); fourth, it is a strong predictor of future academic success (Parsons and Bynner, 2005; Carr and Alexeev, 2011; Fuchs et al., 2016).

Considering the importance of mathematics in future school results, knowing which motor skills can contribute to improving mathematical performance will help educators to select and program more appropriate strategies for teaching and learning math skills.

In this sense, the present systematic review study aims to identify in children with typical development who attend preschool education, the different categories of motor skills that are associated with math skills and the instruments used in investigations carried out with this objective.

It was hypothesized that motor skills positively influence mathematical performance.

A search was carried out in electronic databases, by two reviewers (PF, PF), according to the PRISMA protocol (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) to identify relevant studies (Moher et al., 2009). The databases searched were ERIC, PubMED, SciELO, Scopus, and Web of Science. The research was conducted on January 15, 2022 and the period was restricted to the last 10 years. The research made no restrictions regarding language. Although the last 5 or 10 years may be well-defined periods for research (Sampaio and Mancini, 2007; Donato and Donato, 2019), they may not, however, be the most appropriate. Thus, we conducted a search in the databases used in this study on systematic reviews published on the subject under study and found that the vast majority were published from 2012 onwards. Thus, the last 10 years were chosen instead of the last 5 years. EndNote was the program used to manage bibliographic references. Keywords were used to identify the relevant literature according to the objective of the study. The research was performed according to the abstract of the articles and the phrase used to detect them in all databases was: ((motor AND (motor AND (proficiency OR competency OR skill* OR development OR ability OR performance OR gross OR fine)) AND (“academic performance” OR “academic achievement” OR “academic grids” OR math* OR numeracy) AND (child* OR preschool*)).

After removing duplicates, the titles and abstracts of the remaining studies were read, by two reviewers independently (PF, PF), and selected with reference to the predefined inclusion and exclusion criteria to assess their potential eligibility for this systematic review. Studies in which the abstract clearly indicated that they would be ineligible for inclusion were immediately eliminated, however, those in which there was some doubt as to their eligibility were kept.

Subsequently, the full texts of these articles were obtained to assess eligibility for inclusion in this review by the two reviewers (PF, PF) and in case of doubt the studies were reevaluated together, following inclusion and exclusion criteria. This research followed the PICOS criteria (population, intervention, comparison, outcome, study; Moher et al., 2009): P, preschool children with typical development; intervention, motor skills diagnosis; C, relationship between motor skills and mathematics; O, evidence of the influence of motor skills on mathematical performance; S, cross-sectional or longitudinal studies, with or without intervention, in any language and in any publication format (articles and or papers).

After reading the articles in full, some were excluded because they did not meet all the inclusion criteria. Eligible studies were kept and included for methodological and subsequent quality assessment, data extraction, discussion, and conclusions.

An additional search was conducted in the references of the articles included in the review in order to add relevant articles. After the search, no referenced articles met all the criteria for inclusion in this review.

The methodological quality of the studies was assessed by two independent reviewers (EC and IM). In case of disagreement, the studies were jointly reassessed until a consensus was reached regarding the final score.

The Methodological Quality Checklist for studies based on Observational Methodology (MQCOM), intended for studies using observational methodology, was used to assess the methodological quality of the studies. This instrument was firstly designed by Chacón-Moscoso et al. (2018) that determined the primary criteria/dimensions to take into account when reporting research using observational methods and developed a list of metrics to quantify them (Anguera and Hernández-Mendo, 2015; Portell et al., 2015; Chacón-Moscoso et al., 2016). Recently this instrument was reduced to 16 questions (Chacón-Moscoso et al., 2019), allowing to identify the main methodological quality items needed to conduct studies based on observational methodology and offer the results as a useful tool for authors conducting studies and reviewers making publication decisions.

The MQCOM is composed of 16 items divided into 11 criteria/dimensions: Criterion 1 – Delimitation of objectives is composed of 3 items (Item 1. Reference to observational methodology, specifying whether observation is direct or indirect; Item 2. Delimitation of study objectives; Item 3. Theoretical framework referenced). Criterion/dimension 2 – Observational design, consists of 3 items [Item 4. Observation unit criteria (idiographic: study units are formed by one or more participants if there is a stable link between them; nomothetic: two or more study units); Item 5. Temporal criteria (punctual: one or two observation sessions; follow-up: more than two observation sessions); Item 6. Dimensionality criteria (one-dimensional: one level of response; multidimensional: two or more levels of response)]; Criterion/dimension 3 – Participants/observation units, consists of 1 item [Item 7. Clear specification of inclusion and exclusion criteria for observation units (reasons why some units were chosen in the study and others were not)]. Criterion/dimension 4 – Observation instruments, consists of 2 items [Item 8. Adequacy of the observation instrument (combination of field format with category system, field format, category system, or scale of estimation); Item 9. Codification manual with definition of the categories/behaviors and specification of dimensions (in multidimensional designs)]; Criterion/dimension 5 – Software use, consists of 1 item [Item 10. Software used to register data (SDIS-GSEQ v. 4.2.1./GSEQ 5, LINCE, MATCH VISION STUDIO, Transana, other: specify), control data quality (SDIS-GSEQ v. 4.2.1./GSEQ 5, LINCE, HOISAN, GT, SAS, other: specify)], and analyze data (SDIS-GSEQ, HOISAN, THEME v. 6, R, SAS, other: specify); Criterion/dimension 6 – Data, consists of 1 item [Item 11. Specification of data type as sequential/concurrent (sequential data: behaviors that cannot overlap and belong to a single dimension; concurrent data: behaviors that can co-occur and belong to several dimensions) and event-based/time-based (event-based: the primary parameter used in the record is order of events; time-based: the primary parameter is duration)]; Criterion/dimension 7 – Specification of parameters, consists of 1 item (Item 12. Type of parameters according to given use); Criterion/dimension 8 – Observational sampling, consists of 1 item (Item 13. Delimitation of sessions: clear establishment of criteria (temporal, behavioral, or mixed) for the beginning and the end of sessions within the observation period and of criteria for acceptance of sessions: between-sessions constancy, within-sessions constancy, or temporary disruptions); O Criterion/dimension 9 – Data quality control, consists of 1 item [Item 14. Between-observer reliability (agreement between the records of different observers)/within-observer reliability (agreement between the records of the same observer at two time points)]; Criterion/dimension 10 – Data analysis, consists of 1 item (Item 15. Type of data analysis performed); Criterion/dimension 11 – Interpretation of results, consists of 1 item (Item 16. In the discussion section).

According to Chacón-Moscoso et al. (2019) items are rated between zero and one. The value zero represents, does not comply; The value one represents, meets the criteria; The value 0.5 represents, partially complies. Items 3 and 15 are exceptions, as they still have different intermediate quotations between zero and one (0, 0.33, 0.67, and 1). Item 11 has a rating of nine, not applicable. The minimum value of the MQCOM is zero points and the maximum is 16 points (Chacón-Moscoso et al., 2019).

Data were independently extracted by two reviewers (PF, PF) and cross-checked, and controversial issues were discussed based on the original text to determine the final outcome. The extracted information included study characteristics: author, year, type of study and country where the study was carried out; main characteristics of the sample (number, sex and age); aim of the study, main results and conclusions; specific components of motor skills associated with different mathematical competences; instruments used to assess motor skills and mathematical competences.

Considering the methodological heterogeneity among the studies and instruments for the collection of data related to motor skills and mathematics and quality of the studies, it was not possible to carry out a meta-analysis.

The total search result in the databases produced 2,909 studies (PubMed: n = 860; Scopus: n = 1,004; Web of Science: n = 778; SciELO: n = 35; ERIC: n = 232), of which 965 were automatically excluded because they were duplicates and, after reading the title and abstract, 1,776 studies were excluded, thus resulting in a total of 168 studies potentially suitable studies. After reading these studies in full, 150 were excluded because they did not meet all inclusion criteria and, so, only 18 were included. After selecting the studies included for the review, a manual search of their bibliographic references was performed, from which no study was obtained, either because they did not meet all the eligibility criteria or because they were already included in the selection. The flowchart according to PRISMA methodology is shown in Figure 3.

Of the 18 studies included in this review, the following procedures were performed: Methodological quality assessment (3.2); Analysis of the main characteristics of the studies (3.3); Association between motor skills and the performance of mathematical competences (3.3); Identification of the instruments used in the studies to assess motor skills and mathematical performance (3.4).

With regard to the studies’ rating, observed through MQCOM, on a scale from 0 to 16 points, the studies with the lowest rating obtained 12 points (Suggate et al., 2017; Fischer et al., 2018, 2020; Clark et al., 2021) and the highest rating was obtained by the study of Escolano-Pérez et al. (2020) (Table 1). Regarding the methodological characteristics, all 18 studies make reference to the methodology used (item 1), use observation units with criteria (item 4), fit the observation instruments to the study (item 8), the data and parameters are specified in the study (items 11 and 12), and all studies indicate having performed some inferential analysis to analyze the data. It is noteworthy that all studies perform a good interpretation of the results in the discussion section (item 16), except for the studies of Brock et al. (2018) and Osorio-Valencia et al. (2017) which partially met this criterion. Among all the studies, only one does not partially meet the theoretical framework (item 4) nor does it specify the coding manual with definition of the categories/behaviors and specification of the dimensions (Osorio-Valencia et al., 2017). The major limitation of the studies was the non-use of software for data collection, since only the study by Escolano-Pérez et al. (2020) met this criterion and the study by Clark et al. (2021) partially met it. Probably the design of the studies did not require the use of specific software for data collection, analysis, and interpretation.

The 18 studies included in this review were observational studies, 11 (61%) longitudinal and 7 (39%) transversals, which identified associations between motor skills and mathematical performance, using validated and reliable instruments for this purpose, in typically-developed children of both sexes, who attended preschool education. Regarding their country of origin, the majority of studies, 10 (56%), were conducted in the United States (Dinehart and Manfra, 2013; Verdine et al., 2014; Manfra et al., 2017; Brock et al., 2018; Duran et al., 2018; Kim et al., 2018; Cameron et al., 2019; Nesbitt et al., 2019; Greenburg et al., 2020; Clark et al., 2021), followed by three from Germany (17%; Suggate et al., 2017; Fischer et al., 2018, 2020). The remaining five studies were carried out in different countries, namely Singapore (Khng and Ng, 2021), Spain (Escolano-Pérez et al., 2020), South Africa (De Waal, 2019), Mexico (Osorio-Valencia et al., 2017), and Northeast Pacific (Becker et al., 2014; Table 2).

Regarding the sample size, the studies presented a minimum of 33 (Clark et al., 2021) and a maximum of 34.491 (Greenburg et al., 2020), with a total of 43.447 participants. All studies included children of both sexes. Regarding teaching frequency, most studies, 12 (67%), exclusively included children from preschool education, seven transversals (Becker et al., 2014; Suggate et al., 2017; Fischer et al., 2018; De Waal, 2019; Fischer et al., 2020; Clark et al., 2021; Khng and Ng, 2021) and five longitudinal (Verdine et al., 2014; Osorio-Valencia et al., 2017; Kim et al., 2018; Cameron et al., 2019; Escolano-Pérez et al., 2020). The remaining six studies (33%) were longitudinal and included children from preschool education and later years (Dinehart and Manfra, 2013; Manfra et al., 2017; Brock et al., 2018; Duran et al., 2018; Nesbitt et al., 2019; Greenburg et al., 2020; Table 2). Regarding age, in studies that included only preschool education, children were aged between 2.7 (Fischer et al., 2018) and 6.8 (Kim et al., 2018), and in studies that involved children from preschool education and beyond, the minimum age was 4 years and the maximum was up to the 5th grade (children between 10 and 11 years old; Greenburg et al., 2020).

The study conducted by Greenburg et al. (2020) was included in this review as the authors distinguished in their research two specific aspects of FMS (FMC and the integration of motor information with visual (VMI) and spatial (VSI) information), addressing in this study these specific motor skills only as VSI and not VMI skills (Table 2).

Regarding the study of the associations between motor skills and mathematics, considering the main conclusions from the 18 studies, only in the study carried out by De Waal (2019) whose objective was to determine if there were correlations between motor skills and the academic performance in preschoolers aged 5–6 years old, there was a moderate and significant correlation between a specific component of GMS, namely balance skills (dynamic and static) and mathematical performance (0.46 > r > 0.23; p < 0.05). On the other hand, FMS have been reported to be associated with mathematical performance in 17 studies (Table 2).

When analyzing the distribution of studies that associated mathematical performance with FMS, namely FMC and VMI, it was found that eight studies reported VMI (Becker et al., 2014; Verdine et al., 2014; Brock et al., 2018; Duran et al., 2018; Cameron et al., 2019; Nesbitt et al., 2019; Escolano-Pérez et al., 2020; Khng and Ng, 2021), four studies FMC (Suggate et al., 2017; Fischer et al., 2018, 2020; Clark et al., 2021) and five studies reported both FMS (FMC + VMI; Dinehart and Manfra, 2013; Manfra et al., 2017; Osorio-Valencia et al., 2017; Kim et al., 2018; Greenburg et al., 2020; Figure 4).

Among the VMI, FMS was the one that stood out the most among the 18 studies, having been reported in 13 (72%), and FMC was associated with mathematical performance only in nine (50%). Thus, among FMS, VMI (more than FMC) was more frequently associated with mathematical performance, and is still reported as an important factor for the diagnosis of learning disabilities in mathematics (Khng and Ng, 2021). However, both VMI and FMC were mentioned as strong predictors of mathematical performance in the present (VMI: Dinehart and Manfra, 2013; Becker et al., 2014; Duran et al., 2018; Khng and Ng, 2021; FMC: Dinehart and Manfra, 2013; Suggate et al., 2017; Fischer et al., 2020) and in the future (VMI: Verdine et al., 2014; Manfra et al., 2017; Osorio-Valencia et al., 2017; Brock et al., 2018; Cameron et al., 2019; Escolano-Pérez et al., 2020; Greenburg et al., 2020; CMF: Manfra et al., 2017; Osorio-Valencia et al., 2017; Greenburg et al., 2020).

In the frequency analysis of the association between MS and math skills (Figure 5 and Tables 3, 4), regarding GMS, namely balance, it was associated with the following math skills: counting, measurement (length), shapes and spatial relations (De Waal, 2019). Regarding FMS, both FMC and VMI were associated with all math skills (Figure 5; Tables 3, 4). However, counting was the math skill that was most frequently reported to be associated with FMS. It was associated with FMC in all nine studies (100%; Dinehart and Manfra, 2013; Manfra et al., 2017; Osorio-Valencia et al., 2017; Suggate et al., 2017; Fischer et al., 2018; Kim et al., 2018; Fischer et al., 2020; Greenburg et al., 2020; Clark et al., 2021) and with VMI (Dinehart and Manfra, 2013; Verdine et al., 2014; Manfra et al., 2017; Osorio-Valencia et al., 2017; Duran et al., 2018; Kim et al., 2018; Escolano-Pérez et al., 2020; Greenburg et al., 2020; Khng and Ng, 2021) in nine studies (53%). One can highlight the fact that measurement skills (length), shapes (Dinehart and Manfra, 2013; Manfra et al., 2017; Kim et al., 2018; Greenburg et al., 2020), cardinality (Osorio-Valencia et al., 2017; Fischer et al., 2018; Kim et al., 2018; Fischer et al., 2020), and comparing (Osorio-Valencia et al., 2017; Fischer et al., 2018, 2020; Clark et al., 2021) were reported in four studies. Only in one study number identification (Kim et al., 2018), addition, and subtraction (Suggate et al., 2017) were reported to be associated with FMC. Regarding VMI, except the math skill spatial relations, which was only reported in a single study (Kim et al., 2018), all other skills were reported in four or more studies. Thus, in six studies the number identification skill was reported (Verdine et al., 2014; Duran et al., 2018; Kim et al., 2018; Nesbitt et al., 2019; Escolano-Pérez et al., 2020; Khng and Ng, 2021), addition and subtraction (Becker et al., 2014; Brock et al., 2018; Duran et al., 2018; Cameron et al., 2019; Nesbitt et al., 2019; Khng and Ng, 2021), and measurement (length; Dinehart and Manfra, 2013; Manfra et al., 2017; Duran et al., 2018; Kim et al., 2018; Nesbitt et al., 2019; Greenburg et al., 2020). In five studies, the cardinality (Verdine et al., 2014; Osorio-Valencia et al., 2017; Duran et al., 2018; Kim et al., 2018; Escolano-Pérez et al., 2020) and shapes skills (Dinehart and Manfra, 2013; Manfra et al., 2017; Duran et al., 2018; Kim et al., 2018; Greenburg et al., 2020) were reported. The comparing skill was reported in four studies (Verdine et al., 2014; Osorio-Valencia et al., 2017; Nesbitt et al., 2019; Khng and Ng, 2021).

When analyzing data from studies exclusively involving children who attended preschool, it was found that all math skills were associated with both FMC and VMI: counting – in six studies – (Osorio-Valencia et al., 2017; Suggate et al., 2017; Fischer et al., 2018; Kim et al., 2018; Fischer et al., 2020; Clark et al., 2021) and cardinality (Osorio-Valencia et al., 2017; Fischer et al., 2018; Kim et al., 2018; Fischer et al., 2020) and comparing – (Osorio-Valencia et al., 2017; Fischer et al., 2018, 2020; Clark et al., 2021) in four studies – were the most frequently associated with FMC. Regarding VMI, the math skills most frequently associated with it were counting (Verdine et al., 2014; Osorio-Valencia et al., 2017; Kim et al., 2018; Escolano-Pérez et al., 2020; Khng and Ng, 2021) and number identification (Verdine et al., 2014; Kim et al., 2018; Nesbitt et al., 2019; Escolano-Pérez et al., 2020; Khng and Ng, 2021), both in five studies, and cardinality (Verdine et al., 2014; Osorio-Valencia et al., 2017; Kim et al., 2018; Escolano-Pérez et al., 2020), comparing (Verdine et al., 2014; Osorio-Valencia et al., 2017; Nesbitt et al., 2019; Khng and Ng, 2021), and addition and subtraction (Becker et al., 2014; Cameron et al., 2019; Nesbitt et al., 2019; Khng and Ng, 2021) in four studies.

These results suggest that counting, cardinality, and comparing were the math skills most associated with FMC; and counting, number identification, cardinality, comparing, and addition and subtraction with VMI. However, the math skill of counting was the most commonly associated with FMS (FMC and VMI), as of the 17 studies that analyzed this math skill, it was associated with FMS in 13 (76%; Tables 3, 4).

In the 18 studies included, 10 instruments were used to assess motor skills (Table 4).

In the only study that associated GMS with mathematical performance (De Waal, 2019), the Kinder kinetics Screening test was used (Pienaar et al., 2016).

To assess FMS, nine instruments were used. To exclusively assess FMC, three instruments were identified: Grooved Pegboard Test (GPT, Strauss et al., 2006; Clark et al., 2021); Movement Assessment Battery for Children, 2nd edition (MABC-2, Henderson et al., 2007; Fischer et al., 2020); Battery designed to provide an estimate of children’s fine motor skills in preschool (BEFMS, Martzog, 2015). To exclusively assess VMI, four instruments were identified: The Brigance Inventory of Early Development III — Standardized (IED III, French, 2013; Khng and Ng, 2021); Test of Visual-Motor Integration, 6th edition (VMI, Beery and Beery, 2010; Becker et al., 2014; Verdine et al., 2014; Brock et al., 2018; Cameron et al., 2019; Escolano-Pérez et al., 2020); Copy Design Task (CDT, Osborn et al., 1984; Nesbitt et al., 2019); NEuroPSYchological assessment battery, 2nd edition (NEPSY, Korkman et al., 1998; Duran et al., 2018). To simultaneously assess FMC and VMI, three instruments were identified: Learning Accomplishment Profile-Diagnostic, 3rd edition (LAP-D, Nehring et al., 1992; Dinehart and Manfra, 2013; Manfra et al., 2017; Greenburg et al., 2020); (NEPSY, Korkman et al., 1998; Kim et al., 2018); Peabody developmental motor scale, 2nd edition (PDMS-2, Folio and Fewell, 2000; Osorio-Valencia et al., 2017; Table 4).

For the assessment of mathematical performance, 16 instruments were identified among the 18 studies (Table 4).

Only in one study it was identified: Mathematics Sharing Stories, Cwikla, 2014 (Clark et al., 2021); Dot Counting Task, Clark et al., 2021 (Clark et al., 2021); Test of basic instrumental aspects: reading, writing and numerical concepts (PAIB-1, Galve-Manzano et al., 2009; Escolano-Pérez et al., 2020); Finger-based number representations (Wasner et al., 2015; Fischer et al., 2020); Number tasks (Nosworthy, 2013; Fischer et al., 2020); Foundations for learning: Assessment Framework Grade R (De Waal, 2019); McCarthy Scales of Children’s Abilities (MSCA, McCarthy, 2004; Osorio-Valencia et al., 2017); Numerical skills (Dowker, 2008; Suggate et al., 2017); Nonfinger-based numerical skills and Finger-based numerical skills (Crollen et al., 2011; Suggate et al., 2017); Wechsler Individual Achievement Test, 3rd edition (WIAT-III, Wechsler, 2007; Verdine et al., 2014).

It was identified in two studies: Test of Early Mathematics Ability – 3rd edition (TEMA-3, Ginsburg and Baroody, 2003; Duran et al., 2018; Khng and Ng, 2021); Test for diagnosing basic math skills (TEDI-MATH, Kaufmann et al., 2009; Fischer et al., 2018, 2020); KeyMath-3 Diagnostic assessment (KeyMath3, Connolly, 2007; Duran et al., 2018; Kim et al., 2018).

The Learning Accomplishment Profile-Diagnostic, 3rd edition (LAP-D, Nehring et al., 1992; Dinehart and Manfra, 2013; Manfra et al., 2017; Greenburg et al., 2020) was identified in three studies.

The Woodcock-Johnson Tests of Achievement (WJ-III, Woodcock et al., 2001; Becker et al., 2014; Brock et al., 2018; Duran et al., 2018; Cameron et al., 2019; Nesbitt et al., 2019) was identified in five studies.

Thus, of the seven instruments used to assess VMI, the VMI test (Beery and Beery, 2010) was the most used, as it was reported in five studies. To assess FMC, the most commonly used instrument was the LAP-D (Nehring et al., 1992) in three studies (Table 4).

The main characteristics of the instruments used in the assessment of VMI require tasks of copying geometric figures, letters, numbers or objects, using a sheet of paper and pencil. To assess FMC, the main characteristics of the instruments used allowed the assessment of dexterity and manual speed through tasks that demand the handling and manipulation of objects such as threading beads, placing coins, building blocks, turning cylinders, folding paper or cutting it with scissors.

It is essential to recognize that there were also some limitations in this review. First, it is important to point out that there is an evident gap in the literature of studies that report the association between gross motor skills and mathematical performance in preschool children. Second, there is the considerable heterogeneity of instruments used across studies to assess motor and math skills, making it difficult to compare and clearly interpret the results across studies. Third, the fact that some included studies report covariates that may influence the results, such as demographic factors (for example, socioeconomic status) and cognitive factors (such as executive function and its components). Covariates reported by each eligible study were not discussed as they were beyond the scope of this review.

In order to allow a more accurate comparison of results between studies in the future, researchers should consider consistent use of valid, reliable, and homogeneous standardized instruments. In addition, studies should control demographic, cognitive and physical confounding factors. Finally, as students with neurodevelopmental disorders attend regular schools, future investigations should also examine the relationships between motor and math skills in this population to inform possible forms of intervention.

As it is known that children with FMS difficulties may present negative mathematical performance and given the importance of math performance in future school results, early identification of these difficulties will help educators to design better strategies for teaching math skills. However, although the main instruments reported in the literature accurately assess FMS, these instruments tend to be expensive, time-consuming to administer, require individualized assessments, demand a lot of training and experience, and are usually administered by experts. Thus, the characteristics of most instruments for the evaluation of FMS reveal limitations in the view of our kindergartens’ reality. In this way, their administration in the classroom context by educators can be very conditioned taking into account the characteristics of most classes, which are very numerous and with very extensive curricula.

Due to its importance for academic success, it is considered that educators should carry out an assessment of the development of FMS to their students, in order to detect possible problems associated with mathematical performance. Eventually, if the child shows development problems in FMS, he/she should be referred for a specialized evaluation by a technician with qualifications for this purpose. For this reason, there is an urgent need for a new instrument to evaluate FMS in preschool education children with the ability to adjust to the reality of our kindergartens, that is, one that requires less administration time, does not require much experience or specialized training, has the possibility of being administered to the group/class in a classroom school context, requires few material resources and produces results that can be easily interpreted, classified and associated with mathematical performance. In this sense, this instrument can be a starting point for the early detection of FMS deficits and, consequently, the referral of the child for a new reassessment by a qualified professional. Furthermore, if the child has problems at this level, he or she may benefit from a timely intervention by a specialist and, consequently, prevent/reduce his/her difficulties in mathematical performance.

However, despite FMS being significantly associated with math performance, these skills also require the involvement of GMS. In this sense, it would be important to know which GMS can have the most influence on FMS so that a possible motor intervention program is more efficient and more likely to be successful.

For future research, we suggest the identification of instruments to assess fine motor skills in preschool children, with characteristics that allow their administration by the educator in the classroom context.