Review Reports - Effects of LEGO<sup>®</sup>-Based Neurotherapy on Executive Functions in Children with Autism Spectrum Disorder: A Quasi-Experimental Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this manuscript. The topic is interesting and clinically relevant, as structured play-based interventions for supporting executive functions in children with ASD are an important area of applied neuropsychological research.

The manuscript has potential, but several methodological and reporting issues require clarification, particularly regarding the non-randomized design, high attrition, statistical reporting, group characterization, and interpretation of treatment effects. The conclusions should be made more cautious and better aligned with the limitations of the study design.

More detailed comments and suggested revisions are provided in the attached review file.

Comments for author File: Comments.pdf

Author Response

Responses reviewer 1

Major Issues:

The topic is clinically relevant and original, and the use of a structured LEGO®-based intervention for executive functions in children with ASD is interesting. However, the current manuscript should be clearly framed as a small exploratory pilot/quasi-experimental study rather than as evidence of intervention efficacy.

Response:

We sincerely thank the reviewer for this important observation. We agree that the original wording could have suggested a stronger level of causal inference than supported by the study design. In response, we revised the title and reframed the manuscript as a quasi-experimental study rather than a confirmatory efficacy trial.

The study design is highly vulnerable to selection bias. Participants were not randomized: families who agreed to participate formed the LBN group, whereas those who declined the intervention formed the control group (lines 128– 137). This procedure may have produced groups that differed in motivation, availability, family support, treatment expectations, and willingness to engage with therapy. Although blinded assessment is mentioned (line 135; lines 151– 153), assessor blinding does not compensate for non-random self-selection into treatment and control groups.

Response:

We sincerely thank the reviewer for this important methodological observation. We acknowledge that the absence of randomization and the voluntary allocation procedure increase susceptibility to selection bias and limit definitive causal inference. In response, we revised the Methods and Limitations sections to explicitly clarify the non-randomized nature of group assignment and reduced causal language throughout the manuscript. Additionally, blinded neuropsychological assessment procedures were implemented to reduce potential assessment bias, although we recognize that blinding does not eliminate the limitations.

The limitations section needs to be substantially strengthened. At present, it focuses mainly on small sample size (lines 405–412), which is insufficient. The discussion should explicitly address other major threats to validity, including high attrition, the passive control condition, lack of an active attention-matched control group, all-male sample, absence of ASD severity/IQ/language characterization, unequal medications and concurrent therapies, expectancy effects, and possible practice effects from repeated BANFE-3 testing. The manuscript reports that 60 patients initially agreed to participate, but only 21 completed the protocol: 11 in LBN and 10 in CTRL (lines 135-138). This means that approximately 65% of initially enrolled participants did not complete the study. Reasons for dropout, timing of dropout, and group-specific attrition are not reported, which is a major threat to validity.

Response:

The Limitations section (section 6) was expanded to explain the possible causes of low treatment adherence and participant attrition. These difficulties were primarily associated with adverse socioeconomic conditions, as the population served consisted largely of vulnerable families with limited financial resources. Many families were required to travel long distances, with transportation times ranging from approximately 2 to 3 hours to attend the sessions, which involved additional transportation expenses and a substantial investment of time. Furthermore, several parents reported working in physically demanding and economically unstable occupations, such as agricultural labor, factory or construction work, and domestic duties, which made it difficult to maintain consistent attendance in the therapeutic program.

Baseline characterization is insufficient. Table 1 reports sex, age, school grade, additional therapies, leisure activities, and medication (lines 223-236), but does not report ASD severity, IQ/developmental level, language level, socioeconomic status, baseline adaptive functioning, comorbidities, or standardized autism assessment scores. These variables are critical in pediatric ASD intervention studies.

Response:

We sincerely thank the reviewer for this important observation. We acknowledge that the baseline characterization was limited and that several clinical and developmental variables commonly considered relevant in pediatric ASD intervention studies were not systematically assessed within the scope of the present quasi-experimental study. However, the study did not include variables related to intellectual quotient (IQ) due to ethical, conceptual, and methodological considerations, as well as the fact that such variables are not central components of certain developmental neuropsychology explanatory models, such as the conceptual model proposed by Flores Lázaro JC, and Ostrosky-Solís F. Desarrollo neuropsicológico de lóbulos frontales y funciones ejecutivas. México: Editorial El Manual Moderno; 2012. Instead, inclusion criteria required participants to demonstrate basic executive-function-related abilities, specifically verbal fluency and the capacity to follow instructions, in order to ensure adequate participation in the intervention and assessment activities. Furthermore, we consider it important to highlight that one of the persistent challenges in neurodevelopmental disorders is the limited systematic assessment of executive functions, despite their potential relevance as an underexplored component for differential cognitive diagnosis. Finally, we emphasize the exploratory nature of the findings and the need for future controlled studies including broader clinical characterization and standardized assessments.

The two groups are small and unbalanced in clinically relevant characteristics. The CTRL group was older on average (10.27 ± 3.01 years) than the LBN group (8.7 ± 2.45 years), and pharmacological treatment appears unevenly distributed, e.g., methylphenidate: 10% in CTRL vs. 27.3% in LBN; risperidone: 0% vs. 9.1% (Table 1, lines 223-236). These potential confounders are not controlled statistically.

Response:

We appreciate the reviewer’s important observation and acknowledge that the small sample size and quasi-experimental design limited the possibility of achieving complete baseline equivalence between groups. In response, we revised the manuscript to more explicitly acknowledge these potential confounding factors as methodological limitations and reduced causal interpretation throughout the text. We also emphasized that the findings should be interpreted cautiously given the exploratory scope of the study and the absence of randomized allocation procedures.

The control condition is passive and not matched for therapist contact, attention, expectancy, structured play, or social interaction. Therefore, the observed differences cannot be attributed specifically to LEGO®-Based Neurotherapy; they may also reflect non-specific therapeutic attention, motivation, repeated structured engagement, or expectancy effects.

Response:

We sincerely thank the reviewer for this important observation. We agree that some mechanistic interpretations presented in the original version exceeded the scope of the data obtained in the present study. In response, we clarified that the control condition was passive in nature; however, as described in the descriptive variables, participants from both groups presented different social and family interaction contexts and dynamics.

In the case of the neurotherapy group, motivation to attend the sessions was consistently observed, likely because the activities focused on solving abstract problems through playful and game-based strategies. This play-based component represents a particularly motivating element for pediatric populations. In fact, several parents reported that their children actively requested to attend subsequent sessions, suggesting an adequate level of adherence and therapeutic engagement among those who completed the program.

Furthermore, we consider it important to emphasize that repeated and systematic participation constitutes a fundamental basis of neuropsychological rehabilitation processes. In this study, the activities did not consist of identical task repetition, but rather of a progressive increase in challenges and cognitive demands, adjusted according to the therapeutic objectives established for each subsequent session. This type of progressive approach has previously been reported in rehabilitation programs applied to other clinical conditions, such as epilepsy and cardiac disorders, as well as in LEGO®-based interventions aimed at strengthening social skills.

The term "neurotherapy" is not sufficiently justified. The intervention appears behavioral/cognitive and play-based, while no neurophysiological, neuroimaging, or biological outcomes were measured. The authorshave to justify the terminology more carefully or use a less mechanistic term such as LEGO®-based neurohabilitation/cognitive intervention.

Response:

We sincerely thank the reviewer for this important observation. We agree that the term “Neurotherapy” may be interpreted differently across clinical and neuroscientific contexts. In response, we revised the Introduction section to explicitly clarify that the intervention is framed within a neurohabilitation and neuropsychological rehabilitation approach aimed at stimulating executive and frontal-executive processes through guided cognitive, behavioral, and play-based activities. We also clarified that the intervention refers to experience-dependent cognitive activation and training rather than direct neuromodulatory or biological treatment. Additionally, speculative neurochemical and neurophysiological interpretations were removed from the Discussion section to ensure conceptual consistency with the quasi-experimental scope of the study.

Potential conflicts of interest require clearer disclosure. One author is Daniel B. LeGoff, whose work is central to LEGO®- Based Therapy, and the intervention is closely related to a recognizable therapy model. The funding statement also mentions support from the Cognitive Habilitation Foundation (lines 428-430). The conflict-of-interest statement currently says that no conflicts exist (lines 447-450), but the relationship between authors, intervention development, and institutional/foundation interests should be transparently described.

Response:

We sincerely thank the reviewer for this important observation. In response, we revised the Conflict-of-Interest statement to improve transparency regarding the relationship between the authors, the intervention framework, and institutional support. We clarified that no author received financial compensation related to the intervention evaluated in this study and that the funding sources had no role in study design, data collection, data analysis, manuscript preparation, or publication decisions.

Introduction

The introduction provides a generally useful background on ASD, executive functions, and LEGO®-based interventions. However, it is somewhat broad and should lead more directly to the specific rationale for targeting executive functions with LBN in ASD.

The distinction between LEGO®-Based Therapy (LBT) and LEGO®-Based Neurotherapy (LBN) is not sufficiently clear. Traditional LBT is introduced as a social-skill intervention (lines 72-78), whereas LBN is presented as targeting executive networks (lines 91-98). The authors have to explicitly explain what is novel in LBN compared with established LBT and which components are hypothesized to affect executive functions.

Response:

We sincerely thank the reviewer for this important observation. In response, we revised the Introduction section to more explicitly differentiate traditional LEGO®-Based Therapy (LBT) from the LEGO®-Based Neurotherapy (LBN) approach implemented in the present study. Specifically, we added a paragraph clarifying that, while traditional LBT primarily focuses on social interaction, communication, and collaborative play, the LBN framework additionally incorporates executive-function challenges integrated into LEGO®-based activities. These include planning, sequencing, sustained attention, working memory, inhibitory control, problem solving, and cognitive flexibility within a neurohabilitation-oriented framework. We also clarified that the distinction between LBT and LBN lies in the intentional incorporation of directed executive-function stimulation and experience-dependent cognitive training principles.

The statement that ASD diagnosis lacks objective biomarkers or neurocognitive assessments and that this may challenge diagnostic accuracy (lines 46-49) requires toning down or contextualized. ASD diagnosis is indeed behaviorally based, but standardized diagnostic instruments and clinical procedures exist.

Response:

We sincerely thank the reviewer for this important observation. In response, we revised the Introduction section to clarify that ASD diagnosis is currently based primarily on standardized behavioral and clinical assessments while acknowledging that ongoing research continues to explore complementary neurobiological and neurocognitive markers. The revised wording was intended to contextualize current diagnostic approaches without minimizing the validity of established clinical diagnostic procedures.

The introduction relies heavily on previous work by overlapping author groups and on prior LBN studies in other pediatric conditions (lines 87-105). This is relevant but may create circular justification. The authors should include a more balanced discussion of independent evidence, systematic reviews, and limitations of LEGO®-based interventions in ASD.

Response:

We understand the reviewer’s observation that several of the cited studies were conducted by related author groups. However, the currently available evidence regarding LEGO®-based interventions remains limited. Although both LEGO®-Based Therapy (LBT) and LEGO®-Based Neurotherapy (LBN) share an intervention approach grounded in play and game-based strategies, these approaches have pursued different therapeutic objectives. While LBT has primarily focused on strengthening social skills, LBN has oriented its interventions toward cognitive processes, particularly executive functions, attention, and memory.

Furthermore, to date, a specific line of LBN research directed toward ASD populations had not yet been developed, as this represents a relatively recent approach currently being explored in different clinical populations, including epilepsy, congenital heart disease, and Down syndrome. In this context, the objective of the present study was to replicate the intervention model and explore whether changes or improvements in cognitive abilities, particularly executive functions, could be observed, similarly to what has been previously reported in other neurohabilitation programs.

Material and methods

The method section contains the basic elements of design, participants, instruments, intervention, and ethics, but several details needed for reproducibility and validity are missing.

A participant flow diagram is needed. The manuscript needs to clearly show how many were screened, eligible, consented, allocated to each group, started intervention, dropped out, completed posttest, and were analyzed. This is especially important because only 21 of 60 initially participating patients completed the protocol (lines 135-138).

Response:

We sincerely thank the reviewer for this important observation. In response, we incorporated a participant flow diagram into the revised manuscript in 2.2. Participants section to more clearly describe participant recruitment, eligibility, voluntary group allocation, follow-up procedures, attrition, and final sample included in the exploratory quasi-experimental analysis. The figure also clarifies the non-randomized nature of group allocation and the number of participants who completed the study protocol in both the LBN and CTRL groups.

The timing is unclear. The control group was reassessed six months later (lines 184-188), whereas the LBN group received 12-15 weekly 60-minute sessions (lines 188-191), which corresponds to roughly 3-4 months. The authors must clarify whether the pre-post interval was equivalent in both groups.

Response:

We sincerely thank the reviewer for this important observation. In response, we revised the Materials and Methods section in 2.4 General procedure section to more clearly describe the duration and follow-up procedures for both groups. We clarified that participants in the LBN group completed between 12 and 15 weekly intervention sessions, followed by biweekly and later monthly follow-up until the final neuropsychological assessment at six months. We also specified that participants in the CTRL group underwent a comparable six-month follow-up interval before the post-intervention evaluation.

The LBN intervention is described only in general terms (lines 193-207). A session-by-session table should be added, including targeted executive functions, materials used, task examples, progression criteria, therapist qualifications, fidelity/adherence monitoring, and whether parents or therapists were involved outside the sessions.

Response:

We sincerely thank the reviewer for this important observation. In response, we further clarified the Materials and Methods section to better describe the executive-function-oriented components incorporated into the LBN intervention, including planning, sequencing, working memory, inhibitory control, sustained attention, cognitive flexibility, and collaborative problem-solving activities through structured construction and programming tasks. We also clarified that activities progressively increased in complexity and were adapted according to participants’ cognitive performance and progress. Additionally, references to previous neurohabilitation studies conducted by our group were included to provide more detailed methodological descriptions of the therapeutic structure and intervention principles underlying the LBN approach.

The diagnosis of ASD is described as ICD-11 checklist-based (lines 123-141), but no standardized autism-specific instruments are reported, such as ADOS-2, ADI-R, CARS-2, SRS-2, or equivalent clinical severity measures. At minimum, the authors should report how diagnosis was confirmed and how ASD severity was characterized.

Response:

We sincerely thank the reviewer for this important observation. In response, we revised the Materials and Methods section in the 2.1 section to clarify that ASD diagnoses had been previously established by certified child psychiatrists at the Instituto Nacional de Pediatría according to ICD-11 clinical diagnostic criteria and multidisciplinary clinical evaluation procedures routinely implemented within the institutional electronic medical record system. We also clarified that standardized instruments such as ADOS-2, ADI-R, CARS-2, and SRS-2 were not systematically available for all participants because diagnoses had already been clinically established prior to study enrollment. Additionally, we noted that these instruments primarily represent structured behavioral assessment measures and that current transdisciplinary neurocognitive approaches increasingly seek to incorporate executive-function and cognitive measures as complementary tools for clinical characterization and differential diagnostic assessment in neurodevelopmental disorders.

The manuscript needs to report whether participants continued other therapies during the study and whether therapy frequency/intensity changed. Table 1 shows additional therapies (lines 223-236), but these are not controlled or discussed sufficiently.

Response:

We sincerely thank the reviewer for this important observation. In response, we revised the Materials and Methods section to clarify that participants who initiated the LBN program did not begin additional neuropsychological or behavioral interventions during the active intervention phase. We also clarified that concurrent therapies previously established before study enrollment remained clinically stable throughout the study period according to institutional follow-up records and parental reports in the 2.4 General Procedure section.

The LBN performance scale is described as having reliability 0.80-0.90 (lines 175-183), but the source of this reliability, its validation in ASD, interrater procedures, and scoring examples are not sufficiently documented.

Response:

We sincerely thank the reviewer for this important observation. In response, we clarified in the Materials and Methods section ( 2.3.2. section) that the LBN performance scale has been previously applied and reported in pediatric neurohabilitation studies involving children with congenital heart disease and epilepsy [31–35]. The reported reliability coefficients correspond to observational clinical performance domains used during intervention sessions. Although the scale has been previously used in pediatric neurohabilitation contexts, we acknowledge that large-scale independent validation specifically in ASD populations remains limited, which is consistent with the exploratory scope of the present study.

Statistical Analysis

ANCOVA is an appropriate analytical idea for pre-post controlled designs, but its use here is limited by the very small sample size (CTRL n = 10, LBN n = 11), high variability, and lack of reported assumptions.

Response:

We agree the ANCOVA should be considered with caution due to the small sample size; for this reason, we have included the gain score, i.e., the differences between pretest and posttest, to compare gain score between treatments by using non-parametric Wilcoxon rank-sum test. In addition, we have analyzed the assumptions behind the ANCOVA. We have included the information as Table S2 in Supplementary Materials.

The authors should report baseline equivalence between groups for all relevant demographic, clinical, and baseline BANFE-3 variables. Currently, Tables 2 and 3 provide pre/post means, but no direct baseline comparison or standardized differences are reported (lines 237-265).

Response:

The sample size is small, so it is difficult to have enough data to compare the Pretest and Posttest results by groups for demographic, clinical and baseline BANFE-3 variables. In Tables 2 and 3, we include a column summarizing the gain score, that is, the difference between the Pretest and Posttest, by including mean and SD.

ANCOVA assumptions are not reported: linearity between baseline and posttest scores, homogeneity of regression slopes, normality of residuals, homoscedasticity, and influence/outlier diagnostics. These assumptions are especially important with n = 21 and wide standard deviations.

Response:

Below we show the analysis of the ANCOVA assumptions for the different variables analyzed. In all variables the assumptions of ANCOVA are met: linearity, homogeneity of regression slopes, normality assumption of residuals, homoscedasticity, and there are not outliers neither influence observations. We have included the information as Table S2 in Supplementary Materials.

Table S2: Model diagnostic metrics and statistical tests for the ANCOVA assumptions.

ANCOVA assumption Statistical Test

Subtotal OMC

Subtotal APC

Subtotal DLC

Total Executive Functions

Reset linearity test

R = 0.355

df1=2, df2=16

p-value = 0.706

R = 0.447

df1=2, df2=16

p-value = 0.647

R = 0.842

df1=2, df2=16

p-value = 0.449

R = 0.169

df1=2, df2=16

p-value = 0.846

Rainbow linearity test

Rain = 2.25

df1=11, df2=7

p-value = 0.145

Rain = 1.526

df1=11, df2=7

p-value = 0.195

Rain = 0.529

df1=11, df2=7

p-value = 0.834

Rain = 0.566

df1=11, df2=7

p-value = 0.809

Homogeneity of regression slopes

F = 0.190

df1=1, df2=17

p-value = 0.669

Homogeneity

F = 0.778

df1=1, df2=17

p-value = 0.390

Homogeneity

F = 3.570

df1=1, df2=17

p-value = 0.0760

Homogeneity

F = 0.980

df1=1, df2=17

p-value = 0.336 Homogeneity

Shapiro-Wilk normality test

S = 0.938

p-value = 0.196

S = 0.939

p-value = 0.206

S = 0.913

p-value = 0.063

S = 1.000

p-value = 0.330

Levene Homoscedasticity test

F = 0.197

df1=1, df2=19

p-value = 0.662

F = 0.426

df1=1, df2=19

p-value = 0.522

F = 0.011

df1=1, df2=19

p-value = 0.918

F = 1.0

df1=1, df2=19

p-value = 0.330

Influence / outlier diagnostics

No outliers

The reporting of statistical results is insufficient. The authors should provide F values, degrees of freedom, exact adjusted p-values, adjusted mean differences, 95% confidence intervals, and effect sizes such as partial eta-squared or standardized mean differences. Figures with stars alone are not sufficient.

Response:

Below we show the analysis of the ANCOVA, and the effect size measures eta squared and epsilon squared. We have included the information as Table S1 in Supplementary Materials.

Table S1: ANCOVA and effect sizes for the Effect of LBN in children with ASD as evaluated by the BANFE-3 test.

Statistical Results

Subtotal OMC

Subtotal APC

Subtotal DLC

Total Executive Functions

F values

F = 3.220

df1=1, df2=18,

p-value = 0.089

F = 0.019

df1=1, df2=18,

p-value = 0.891

F = 26.974

df1=1, df2=18,

p-value = 0.00006

F = 12.181

df1=1, df2=18,

p-value = 0.0026

Estimated marginal means

Post-Pre = 1.790

p-value = 0.089

Post-Pre = 0.139

p-value = 0.891

Post-Pre = 5.190

p-value = 0.00006

Post-Pre = 3.49

p-value = 0.0026

Effect size: eta squared

95% CI

Eta2 = 0.05

(0.00, 1.00)

Magnitude= small

Eta2 = 0.0

(0.00, 1.00)

Magnitude= small

Eta2 = 0.50

(0.29, 1.00)

Magnitude= large

Eta2 = 0.34

(0.08, 1.00)

Magnitude= large

Effect sizes: epsilon squared

95% CI

Epsilon2 = 0.09

(0.00, 1.00)

Magnitude= small

Epsilon2 = 0.04

(0.00, 1.00)

Magnitude= small

Epsilon2 = 0.53

(0.28, 1.00)

Magnitude= large

Epsilon2 = 0.40

(0.12, 1.00)

Magnitude= large

The manuscript uses multiple outcomes: OMC, APC, DLC, and total executive functions, plus two analytical approaches (ANCOVA and gain-score Wilcoxon tests). The primary outcome should be prespecified, and correction for multiple comparisons should be transparent across all tests, not only described generally as Bonferroni correction (lines 209-219 and Figure 3 caption lines 307-312).

Response:

We have included the corresponding tables with the results of the applied tests in Supplementary Material document, in particular, the ANCOVA results in Table S1, the diagnosis of the ANCOVA metrics and assumptions in Table S2, and the Gain score results including the Effect size in Table S3.

The gain-score analysis should report medians, interquartile ranges, W/U statistics, exact p-values, and rank-biserial correlation or another non-parametric effect size. Figure 4 reports only significance symbols and general interpretation (lines 314-334).

Response:

For the gain score analysis, we have reported for values and details about statistical test and p-values for the Wilcoxon rank sum test, and also we include effect size measures. We have included the information as Table S3 in Supplementary Materials.

Table S3: Effect of LBN on the gain score of children with ASD, as evaluated by the BANFE-3 test.

Gain-scores

Subtotal OMC

Subtotal APC

Subtotal DLC

Total Executive Functions

Median (1^st, 3^rd quartiles)

CTRL: 0.0

(-28.0, 6.0)

LBN: 8.0

(-0.5, 26.5)

CTRL: 0.0

(-13.0, 14.0)

LBN: 8.0

(-20.0, 1.5)

CTRL: 0.0

(-1.0, 1.0)

LBN: 17.0

(-12.0, 22.0)

CTRL: 0.0

(-4.0, 4.0)

LBN: 8.0

(-9.0, 26.0)

Wilcoxon rank sum test with continuity correction

W = 35.5

p-value = 0.1794

W = 68

p-value = 0.3784

W = 9

p-value = 0.0013

W = 15

p-value = 0.0051

Wilcoxon rank sum test

W = 35.5

p-value = 0.1682

W = 68

p-value = 0.3597

W = 9

p-value = 0.0012

W = 15

p-value = 0.0046

Effectsize

effsize = 0.301

n1=10, n2=11

magnitude = moderate

effsize = 0.200

n1=10, n2=11

magnitude = small

effsize = 0.709

n1=10, n2=11

magnitude = large

effsize = 0.619

n1=10, n2=11

magnitude = Large

R effectsize

95% CI

0.76

(-0.10, 1.63)

-0.17

(-1.1, 0.76)

1.43

(0.79, 2.07)

1.22

(0.48, 1.95)

Effectsize rc rank-biseral correlation

95% CI

-0.35

(-0.70, 0.13)

0.24

(-0.26, 0.63)

-0.84

(-0.94, -0.61)

-0.73

(-0.89, -0.39)

The statement that ANCOVA reduces the sample-size requirement (lines 405-411) is potentially misleading. ANCOVA can improve precision when assumptions are met and baseline-outcome correlations are adequate, but it does not resolve the problems of self-selection, attrition, confounding, and very small sample size.

Response: Thank you by your observation. We agree with you that results are needed based on experimental designs with substantial sample sizes, adequate patient selection, minimizing or avoiding attrition, and conducting clinical trials that minimize confounding factors. We used ANCOVA because it allows to improve the precision of results despite the small sample size. We have modified the paragraph.

Original paragraph

Nevertheless, the main limitation of this study is the small sample size. Although large samples are generally needed to ensure statistical significance, the use of ANCOVA can reduce this requirement. Additionally, previous studies suggest that sample sizes between 8 and 20 participants may be sufficient to detect significant differences [47]. Similarly, more recent studies indicate that under certain conditions, even small samples (~10 subjects) may be adequate, depending on the effect size R-squared [48]. These results should be interpreted with caution and require validation in studies with larger samples.

New paragraph

The principal limitation of this study is the small sample size, which may limit the generalizability of the findings and reduce statistical power. Although large samples are generally needed to ensure statistical significance, the use of ANCOVA can improve the precision of estimations, when assumption behind ANOVA are met, and correlations between baseline and outcome are adequate. Related to ANCOVA, previous studies suggest that sample sizes between 8 and 20 children may be sufficient to detect significant differences [49] and more recent studies indicate that under certain conditions, even small samples (~10 subjects) may be adequate, depending on the effect size R-squared [50]. These results should be interpreted with caution and require validation in studies with larger samples.(pages 13-14)

The software reporting is outdated and imprecise. R 3.4.1 and RStudio 0.99.902 are old (lines 217-218), and the statement "R package version 4.1.1" is unclear. The package name and version should be specified, presumably ggplot2.

Response: Thank you by your observation. We have updated the R version and specified the library used for graphical representations.

Results

The descriptive results are clearly presented at a basic level, but more complete numerical reporting is needed.

Response: Thank you by your observation. We have included different Tables with numerical reports in Supplementary Material document, in particular, the ANCOVA results are in Table S1, the diagnosis of the ANCOVA metrics and assumptions are in Table S2, and the Gain score results including the Effect size are in Table S3.

Table 1 should include p-values or, preferably, standardized differences for baseline comparability. Given the tiny sample, p-values alone would be uninformative; effect-size-style balance diagnostics would be more useful.

Response:

Thank you. We think that in Table 1, where only descriptive statistics are presented to identify the characteristics of the population, it is not necessary to obtain p-values. We are not performing any formal tests to compare the population. We are only describing the characteristics of the population in the study. We compare the Pretest and Posttest in the following subsections, and we include Tables with details of statistical test in Supplementary Material document.

Tables 2 and 3 show mean ± SD values, e.g., DLC increased from 65.7 ± 18.5 to 83.3 ± 27.2 and total executive functions increased from 66.4 ± 20.7 to 81.9 ± 30.6 in the LBN group (lines 248-254). However, these descriptive improvements should not be overinterpreted without adjusted effect sizes and confidence intervals.

Response:

Thank you. We have revised the wording of the paragraphs so that they are not interpreted as a cause-and-effect relationship, but rather as purely descriptive (pages 7-8 in 3.1 section). Moreover, we understand that Tables 2 and 3 are descriptive statistics. We compute the statistical tests and effect sizes in the following subsections.

Figure 1 is difficult to interpret. It shows multiple colored individual lines for LBN performance by session (page 7, lines 266-287), but the legend is unclear, participant profiles are not defined, and the figure shows 12 sessions although the methods state that participants completed 12-15 sessions (lines 188-191).

Response: Thank you by your observation. We have updated the figure, including all the sessions of participants, the legends of participants identifier (Id).

Figure 2 is visually useful but potentially misleading because regression lines and 95% confidence intervals are shown for groups of only 10 and 11 participants (lines 288-298). The plots should be supplemented with exact numerical ANCOVA results.

Response: Thank you by your observation. We include the exact numerical ANCOVA results in Tables S1 and S2 in Supplementary Material document.

Figure 3 requires clearer labeling and exact statistics. The right-side plots report very strong significance symbols, including **** p ≤ 0.0001 and ** p ≤ 0.005 (lines 307-312), but the manuscript does not provide the corresponding adjusted mean differences, p-values, confidence intervals, or effect sizes.

Response: Thank you by your observation. We include the exact numerical ANCOVA results in Tables S1 and S2 in Supplementary Material document.

The text states that statistically significant differences were found in DLC and total executive functions (lines 299-304), but it is not fully clear whether this refers to within-group change, between-group adjusted posttest comparison, or both. This distinction must be made explicit.

Response: Thank you by your observation. We clarify the comment, we add the following sentence:

BANFE-3 scores obtained at baseline and post-intervention were compared between the CTRL and LBN groups. Descriptive analyses revealed an increase in the LBN group in all evaluated functions, with the exception of APC; i.e., the estimated marginal means increase for the LBN group in comparison to the CTRL group, except for APC, where they are similar. In addition, the results show an increase in the scores of the experimental LBN group compared to those of the CTRL group, with statistically significant differences in the DLC area and total executive functions; i.e., their estimated 95% confidence intervals are displayed without overlapping, and having large effect sizes (Figure 3, and Tables S1 and S2 in Supplementary Material.(page 10)

Figure 4 shows significant gain-score differences for DLC and total executive functions (lines 314-334), but the results should include exact Wilcoxon statistics and effect sizes in the text or a table.

Response:

Thank you by your observation. We have included details about the gain score in Table S3 in Supplementary Material document.

Discussion

The discussion is generally coherent and clinically interesting, but it overstates the implications of a small, self-selected, quasi-experimental pilot study.

Response: Thank you by your observation. The data interpretation as a quasi-experimental study was applied and not as cause effect.

Causal language should be reduced throughout. The findings support an association between participation in LBN and improvement in selected BANFE-3 indices, not definitive evidence that LBN caused executive-function improvement.

Response: Thank you by your observation. Causal language was minimized, and the data were interpreted as improvements in cognitive function, avoiding generalizations or definitive conclusions. Additionally we added limitations section.

The mechanistic paragraph on glutamatergic, dopaminergic, and cholinergic modulation (lines 388-400) is speculative and unsupported by the study data. No neurotransmitter, genetic, neuroimaging, electrophysiological, or biomarker outcomes were collected. This paragraph has to be removed or rewritten as a very cautious future-research hypothesis.

Response: Thank you by your observation The paragraph was removed

The conclusion should be softened. The phrase "positive effects" (lines 413-419) should be replaced with wording such as "preliminary results suggest potential benefit" or "LBN was associated with improvements in this small pilot sample."

Response: Thank you. Conclusions were modified.

Overall Assessment

This manuscript addresses an original and clinically relevant topic. The use of LEGO®-based structured activities to support executive functioning in children with ASD is interesting, and the inclusion of a control group and blinded neuropsychological assessment are important strengths. The observed improvements in dorsolateral cortex-related BANFE-3 scores and total executive function scores are promising. However, the manuscript is not yet strong enough as an efficacy article. The study is limited by non-random self-selection into groups, very high attrition from 60 initial participants to 21 completers, small and unbalanced groups, insufficient clinical characterization, passive control condition, incomplete intervention reproducibility, underreported statistics, and overinterpretation of neurobiological mechanisms. The paper should be reframed as a preliminary exploratory pilot study, with substantially improved methodological transparency and statistical reporting. Recommendation: Major revision. The manuscript has potential, but it requires substantial revision before publication. The authors should tone down causal and mechanistic claims, provide a full participant-flow diagram, clarify timing and allocation, expand intervention details, improve statistical reporting, correct references/DOIs, and present the findings as preliminary rather than confirmatory.

Response: Thank you by your observation. The manuscript was modified taking your observations.

Minor issues

The Abstract states only that a pre-post intervention study was conducted (lines 33-35), but the Methods later describe a control group (lines 132-138). The design should be described consistently in the Abstract.

Response: Thank you by your observation. The abstract was modified

Use consistent terminology: "children", "patients", "participants", "CTRL", "control", "treatment", "intervention", "LBN", and "LEGO®-based therapy" are used interchangeably.

Response: Thank you by your observation. The terms were corrected

Correct typographical errors in author affiliations and institutional names, e.g., "Subdireccón" and "Departmento" (lines 13-15).

Response: Thank you by your observation. The words were corrected.

Table 1 contains "Secundary"; this should be "Secondary" (line 227).

Response: Thank you by your observation. The word was corrected

Table 2 contains "Subototal APC"; this should be "Subtotal APC" (line 334).

Response: Thank you by your observation. The word was corrected

The note below Table 3 contains "Dorso. lateral cortex"; this should be "dorsolateral cortex" (line 339).

Response: Thank you by your observation. The word was corrected

Several figure labels should be corrected: "Lego® B-N" should be standardized as "LBN"; "Dorsolateral"/"Dorsolateral Cortex" should be spelled consistently; "emmean test" should be written clearly as estimated marginal means.

Response: Thank you by your observation. We have corrected the labels of figures. "Lego® B-N" is "LBN"; "Dorsolateral" is "Dorsolateral Cortex"; "emmean" is defined as “estimated marginal mean”.

The sentence "We hypothesized that LBN affects excitatory neurotransmitter and acetylcholine and dopamine modulation" (lines 398-400) is grammatically unclear and scientifically unsupported by the current data.

Response: Thank you by your observation. The paragraph was eliminated. Effectively the information is unsupported by the data.

The Institutional Review Board statement appears incomplete: "approved by the Institutional Research and Ethics Committee of the. INP" (lines 433-436). Remove the misplaced period and ensure consistency between approval number 2022/045 and protocol code 45/2022.

Response: Thank you by your observation. The statement was corrected. The protocol number was corrected in the Review Board statement and in 2.1. Study design, settings and ethical considerations section.

The manuscript should consistently use MDPI reference style, including journal abbreviations, DOI formatting, capitalization, page ranges, and access dates.

Response: Thank you by your observation. The format references and numbering were corrected.

Reference [16] appears to merge two different articles into one reference: Ikeda et al. 2018 and Hill 2004 are listed in the same item (lines 486-489). These should be separated into two independent references, and all subsequent citation numbering should be checked.

Response: Thank you by your observation. The correct reference was added

Reference [26] appears questionable as written. The cited item is described as "LEGO-based therapy for autism spectrum disorder: A systematic review" in Autism 2014, 18, 795-806 (lines 506-508), but the well-known systematic review with a similar title is Narzisi et al., Brain Sciences 2021, 11, 702. The authors should verify the bibliographic details and DOI.

Response: Thank you by your observation. The correct references is Narzisi et al., 2021 in Brain Sciences. The reference was corrected

Reference [42] is incomplete: it contains only a dissertation repository link without authors, title, year, institution, document type, or DOI/stable bibliographic information (lines 541-542).

Response: Thank you by your observation. The reference was corrected (now reference 45)

DOI formatting is inconsistent across references, alternating between "https://doi.org/...", "doi:", "DOI:", and plain DOI notation, e.g., references [21-22], [41], [45-52] (lines 498-565). MDPI style should be applied consistently.

Response: Thank you by your observation. The format was corrected

References [53] and [54] should be completed with DOI information where available. For example, the Borm et al. ANCOVA sample-size paper has a DOI, and the Bujang et al. ANCOVA sample-size article is commonly cited with DOI 10.2427/12117 (lines 565-570).

Response: Thank you by your observation. The doi was added in the reference of Borm et al., 2007 (now reference 49) and in Bujang et al., 2017 (now reference 50).

Several references contain typographical or formatting problems, e.g., "Snadny,, K." (line 547), inconsistent capitalization, incomplete journal titles, and inconsistent punctuation. The entire reference list should be checked against the journal style.

Response: Thank you by your observation. Each reference was reviewed and revised as necessary.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Manuscript ID: brainsci-4323131

Title: “Effects of LEGO®-Based Neurotherapy on Executive Functions in Children with Autism Spectrum Disorder”

Thank you for the opportunity to review this manuscript. The study is original and has significant clinical value. However, some major revisions are needed before the study is ready for publication.

General comment: Children on the Autism Spectrum are not patients. Avoid using medical labels to characterize individuals with autism. They say little about people as individuals and tend to reinforce stereotypes of disabled people as ‘patients’ or unwell.

It is advisable to use inclusive language. Replace the phrase “pediatric patients with ASD” with children with ASD throughout all your manuscript.

Introduction

Page 3. Lines 101-105. The obesity and the congenital heart disease are organic, not neurodevelopmental disorders. Please, rephrase and correct accordingly.

Methods

Page 4. Lines 152-154. Transfer lines 152-154 in the general procedure subsection. Move them exactly after line 187.

Page 4. The authors should describe the place where the intervention sessions took place and the qualifications of the clinician(s) who performed them.

Page 4. Lines 176-183. Who performed the evaluation of each session? Were they clinicians with significant clinical experience and education at play therapy? Were they trained on LBN? Did the researchers audio-record the play sessions with LBN?

Page 4. The authors report that they evaluated inter-rater reliability. Who was the second rater? What qualifications he/she had? Was he/she blind to the first’s clinician assessment?

Results

Page 5. Lines 223-230. Please transfer lines 223-230 and table 1 in the Methods section and specifically in the subsection of the participants. Put all these information that refers to participants after line 151.

Page 5-Line 230-233. Delete these lines. The authors have presented this information in the Methods section.

Page 6-Line 238. Rename this subheading as “3.1. Cognitive evaluation of participants”

Page 6-Line 266. Rename this subheading as “3.2. Performance of LBN in children with ASD”.

Page 7-Line 288. Rename this subheading as “3.3. Effects of LBN on children with ASD”.

Discussion

Page 11-Line 370. A citation is missing.

Page 11-Lines 388-400. Your assumptions are not based to evidence. Your study DID NOT measure excitatory neurotransmitter or acetylcholine or dopamine. Please, delete these lines.

Author Response

Title: “Effects of LEGO®-Based Neurotherapy on Executive Functions in Children with Autism Spectrum Disorder”

Thank you for the opportunity to review this manuscript. The study is original and has significant clinical value. However, some major revisions are needed before the study is ready for publication.

It is advisable to use inclusive language. Replace the phrase “pediatric patients with ASD” with children with ASD throughout all your manuscript.

Response: We appreciate this important observation. We agree on the importance of using inclusive and respectful language when referring to individuals with autism spectrum disorder (ASD). In response to this comment, the manuscript was revised and the term “pediatric patients with ASD” was replaced with “children with ASD

Introduction

Page 3. Lines 101-105. The obesity and the congenital heart disease are organic, not neurodevelopmental disorders. Please, rephrase and correct accordingly.

Response: We appreciate this observation and agree that obesity and congenital heart disease are not neurodevelopmental disorders. In response to this comment, the term “neurodevelopmental” was replaced with “neurological conditions” throughout the manuscript to more accurately reflect the nature of the conditions described principally to the end of the Introduction section.

Methods

Page 4. Lines 152-154. Transfer lines 152-154 in the general procedure subsection. Move them exactly after line 187.

Response: The requested transfer has been implemented. Lines 152–154 were moved from the general procedure subsection and are now located in the end of 2.4 section.

Page 4. The authors should describe the place where the intervention sessions took place and the qualifications of the clinician(s) who performed them.

Response: We appreciate this observation. In response to this comment, a more detailed description of the setting where the intervention sessions took place was added to the manuscript, along with information regarding the therapists’ training and prior experience in conducting the interventions. The information is in 2.4. General procedure section

Response: We appreciate this observation. In response to this comment, the description of both the evaluators and the therapists conducting the sessions was expanded in the 2.4 section. The manuscript now clarifies that the assessments were performed by an experienced neuropsychologist trained in the administration of BANFE-3, while the intervention sessions were conducted by therapists previously trained in cognitive-behavioral approaches and neuropsychology. In addition, it was specified that the sessions were recorded and later analyzed using a session performance scale in order to monitor protocol adherence and document each participant’s difficulties and achievements throughout the intervention.

Page 4. The authors report that they evaluated inter-rater reliability. Who was the second rater? What qualifications he/she had? Was he/she blind to the first’s clinician assessment?

Response: We appreciate this observation. In response to this comment, the manuscript was revised to clarify that all therapists had training in cognitive-behavioral therapy and neuropsychology, as well as specific training in the LBN intervention and in the use of the session evaluation scale. In addition, it was specified that the primary evaluator conducted the clinical and diagnostic assessments, while the therapists, although trained in the implementation of the sessions, were not involved in those evaluations in order to preserve the standardization and objectivity of the procedure.

Results

Response: The requested modification has been implemented. The corresponding text was relocated and is now located in the end of 2.2 Participants section in methods section (lines 218-229).

Page 5-Line 230-233. Delete these lines. The authors have presented this information in the Methods section.

Response: The requested deletion has been implemented. Lines 230–233 were removed from this section and the information as incorporated in the end of 2.1. Study Design, Settings, and Ethical Considerations in Methods section.

Page 6-Line 238. Rename this subheading as “3.1. Cognitive evaluation of participants”

Response: We appreciate this observation. In response to this comment, the subtitle was changed to “3.1. Cognitive Evaluation of Participants” to improve the clarity and accuracy of the section.

Page 6-Line 266. Rename this subheading as “3.2. Performance of LBN in children with ASD”.

Response: We appreciate this observation. In response to this comment, the subtitle was changed to “3.2. Effects of LBN on Children with ASD” in order to more accurately reflect the content of the section.

Page 7-Line 288. Rename this subheading as “3.3. Effects of LBN on children with ASD”.

Response: We appreciate this observation. In response to this comment, the subheading was changed to “3.3. Effects of LBN on Children with ASD” to improve the clarity and consistency of the manuscript organization.

Discussion

Page 11-Line 370. A citation is missing.

Response: Thank you by your observation. The citation was added (line 450)

Page 11-Lines 388-400. Your assumptions are not based to evidence. Your study DID NOT measure excitatory neurotransmitter or acetylcholine or dopamine. Please, delete these lines.

Response: Thank you by your observation. The paragraph was eliminated

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have substantially improved the manuscript and addressed most of the major concerns raised in the previous review. The revised version now frames the study more appropriately as a small exploratory quasi-experimental study, clarifies the non-randomized allocation procedure, adds a participant flow diagram, expands the limitations section, improves the distinction between LEGO®-Based Therapy and LEGO®-Based Neurotherapy, and provides additional statistical information in the supplementary material.

Only minor revisions are still required before publication. First, the manuscript should maintain cautious wording throughout and avoid implying confirmed intervention efficacy, given the non-randomized design, very small final sample, high attrition, passive control condition, and limited ASD clinical characterization. Second, the key statistical outcomes should be summarized more clearly in the main text, not only in the supplementary tables. Third, the description of the intervention should be made slightly more reproducible, especially regarding session structure, progression criteria, and fidelity monitoring. Finally, the manuscript requires careful language and formatting correction, including minor errors in the abstract, figure captions, terminology, and repeated or awkward phrases.

Overall, the revision is much stronger, and the manuscript may be suitable for publication after these minor corrections.

Author Response

Comments reviewer 1

Response:

Thank you by your valuable comments

Response:

We thank the reviewer for this important observation. We carefully revised the manuscript to ensure that the language remains appropriately cautious and consistent with the exploratory quasi-experimental nature of the study.

Second, the key statistical outcomes should be summarized more clearly in the main text, not only in the supplementary tables.

Response:

Thank you by your valuable comments The summary of the tables was added in 3.3 section (lines 375-391) and 3.4 section (lines 405-409)

Third, the description of the intervention should be made slightly more reproducible, especially regarding session structure, progression criteria, and fidelity monitoring.

Response:

Thank you for this suggestion. We expanded Table 2 and the Methods section to provide a clearer description of the intervention structure, activity progression, and performance monitoring procedures.

Finally, the manuscript requires careful language and formatting correction, including minor errors in the abstract, figure captions, terminology, and repeated or awkward phrases.

Response:

We thank the reviewer for this observation. The manuscript was carefully revised for language, formatting, terminology consistency, figure captions, and overall readability. Minor errors, repeated expressions, and awkward phrasing were corrected throughout the text.

Overall, the revision is much stronger, and the manuscript may be suitable for publication after these minor corrections.

Response:

Thank you for your valuable feedback, which helped improve the manuscript

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review the revised manuscript entitled: “Effects of LEGO®-Based Neurotherapy on Executive Functions in Children with Autism Spectrum Disorder”. The authors have addressed most of my previous concerns.

However, some minor corrections must be performed before the manuscript is ready for publication.

General comment

The Introduction and discussion sections are well-organized.

Methods

Line 261. Study 35 refers to obesity. Citation is inaccurate.

Line 310. A full-stop after the word “scores” is missing.

Results

Line 323-333. In these lines, the authors present “mean” scores. This is not apparent in their text; therefore, it should be corrected.

Lines 344-355. These lines provide an explanation about the observed data and should be transferred to the discussion section. In the Results section, only the presentation of data is permitted.

Author Response

Comments reviewer 2

However, some minor corrections must be performed before the manuscript is ready for publication.

Response:

Thank you by your valuable comments

General comment

The Introduction and discussion sections are well-organized.

Response:

Thank you by your valuable comment

Methods

Line 261. Study 35 refers to obesity. Citation is inaccurate.

Response:

Thank you by your valuable comment. The references were corrected (line 261)

Line 310. A full-stop after the word “scores” is missing.

Response:

Thank you by your valuable comment. The full stop was added (line 312)

Results

Line 323-333. In these lines, the authors present “mean” scores. This is not apparent in their text; therefore, it should be corrected.

Response:

Thank you by your valuable comment. The paragraph was corrected (lines 323-333)

Lines 344-355. These lines provide an explanation about the observed data and should be transferred to the discussion section. In the Results section, only the presentation of data is permitted.

Response:

Thank you by your valuable comment. The paragraph was eliminated of the 3.2 section in results and added in the Discussion (lines 478-489).

Author Response File: Author Response.pdf