Physical Freezing in Children and Adolescents with Selective Mutism

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Unfortunately, the reviewer cannot recommend the publication of the article. However, the reveiwer encourages to publish the presented data on self-reported freezing in children with SM alone, compared to healthy controls. 

Comments for author File: Comments.pdf

Author Response

Please see attached document which contains the feedback and our responses

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Comments for authors:

This study, relying on parent reports and the Selective Mutism Questionnaire, provides valuable insights into the manifestations of autism and selective mutism in children. However, the absence of standardized diagnostic tools and gold standard validation raises concerns about potential diagnostic inconsistencies and biases.

1. The study does not specify the diagnostic criteria for autism; it is essential for the study to clarify the diagnostic criteria for autism, and the use of a gold standard diagnostic method is crucial.
2. Sample selection bias: This study relied on social media and relevant website advertisements for recruitment, which may lead to a biased participant sample. For example, the user base of social media may be skewed toward certain groups (such as parents), potentially neglecting families that do not have access to social media, thereby limiting the representativeness of the sample.
3. The study did not independently validate the structural validity of the SMQ nor specify whether factor analysis or other techniques were used to examine the construct structure of the questionnaire.
4. Criterion validity refers to the correlation between the questionnaire results and external standards, typically validated through comparison with other established tools or clinical diagnoses. This study primarily relies on existing literature support and has not independently validated criterion validity within the sample.
5. The diagnosis of autism and selective mutism (SM) primarily relied on parent reports and existing clinical information. Parent report bias is a significant limitation in such studies, particularly when information about children's behavior, symptoms, and diagnoses is gathered through parent reports.
6. Due to the low response rate of the scale, the conclusion should be more conservative.

 

Author Response

Please see attached document which contains the feedback and our responses

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Overall Evaluation

The manuscript reports an online parent–child survey examining physical “freezing” phenomena (throat and body) in children aged 5–16 years with selective mutism (SM), autism, both diagnoses, or neither. It combines parent-report SMQ data with a child survey focusing on subjective freezing sensations and perceived factors that increase speaking difficulty. The work addresses an underexplored topic (physical freezing in SM and SM+autism), uses a reasonably sized parent sample, and provides clinically relevant descriptive information.

Strengths include a clear focus, appropriate non-parametric statistics for skewed data, explicit acknowledgement of key limitations, and a well-structured discussion linking the findings to SM, autism, anxiety, and catatonia literature. However, the study design has substantial constraints (self-selected sample, high attrition, limited validation of diagnoses, very small child subsample, and lack of data on freezing in the autism-only group), and some aspects of the methods, analyses, and interpretation require further clarification or tightening. The description of the construction and interpretation of the “interference score” from the SMQ also needs to be more precise, and several figures and sections of the text would benefit from clearer wording and harmonized terminology.

General Comments

1. The manuscript addresses an important and relatively neglected clinical question: the role of physical freezing in SM and SM with co-occurring autism. The conceptual integration of SM, freezing, anxiety, sensory processing, and catatonia is a notable strength.

2. The parent sample is comparatively large for SM research, and the use of the SMQ (including its three domains) is appropriate. Internal consistency is reported and high. The additional “interference” questions are informative but need more explicit methodological description and justification of the way they are combined into a single score.

3. The child survey adds valuable first-person perspectives on freezing sensations and context factors that exacerbate speaking difficulty. However, the small number of respondents and the imbalance in diagnostic groups and sex greatly limit inferential power. These constraints are acknowledged but should be reflected even more clearly in how conclusions are framed.

4. The self-selected recruitment via SM-focused networks, high proportion of incomplete parent surveys, and absence of independent confirmation of autism/SM diagnoses restrict generalizability and may introduce selection bias. These issues are discussed in the limitations, but some statements in the Introduction and Conclusions still read as more general than the data justify.

5. The statistical methods are broadly appropriate, but the reporting of some analyses can be refined (e.g., exact p-values, effect sizes where feasible, and more consistent terminology when combining groups such as “SM only” vs “SM with or without autism”).

6. Presentation is generally clear and well organized, with a standard IMRAD structure consistent with MDPI format. Minor language issues (spelling, punctuation, and occasional ambiguous phrasing) remain and should be corrected in a thorough proofread. Characters such as “Heath” for “Health” and “beĴer” appear to result from encoding issues and should be standardized.

7. Figures are informative but in several cases the legends and captions do not fully explain group labels (e.g., “No” vs “Yes, Autism” in Figure 2B; “Responder – Boys” vs “Responder – Female” in Figure 5A). Clarifying these would improve readability, especially for readers not immersed in the SMQ literature.

Overall, with clearer methodological description, more cautious interpretation, and improved figure/text clarity, the manuscript could make a useful contribution to the literature on SM, freezing responses, and comorbid autism.

Specific Comments

Title, Abstract, and Keywords

1. Lines 1–3 (Title and authorship block): The article is labeled as “Article 1” and “FOR PEER REVIEW”; confirm that the final version will use the standard MDPI article header and remove any placeholder labels.

2. Lines 10–22 (Abstract): The design and sample are not fully specified. Explicitly stating that this is a cross-sectional, online parent survey with an embedded child survey, including total numbers of parent respondents and child respondents (N for each diagnostic group), would improve clarity and replicability.

3. Lines 15–22 (Abstract, results): The sentence “Findings suggest that throat and body freezing commonly occur in children with SM whether autistic or not” may overgeneralize given the small number of SM+autism child respondents and absence of freezing data for the autism-only child group. Wording could be moderated to emphasize that this was observed in this self-selected sample.

4. Lines 19–22 (Abstract, SMQ scores): The phrase “Lower SMQ scores tended to be related to greater frequency and extent of freezing” implies a statistical association; providing the correlation coefficient(s) and p-values for this relationship (if calculated) would support the statement, or otherwise it should be described more qualitatively.

5. Lines 23–24 (Keywords): Consider adding “survey” and “online questionnaire” or similar methodological terms, as well as “child report” or “self-report,” to increase discoverability.

Introduction

6. Lines 26–41: The opening paragraphs give a good overview of SM and its prognosis. Some references to ages and outcomes (e.g., “children diagnosed at age 7 or above appear to be more impaired four or more years later”) could include more precise information from the cited study (sample size, follow-up duration).

7. Lines 48–55: The list of additional behaviors associated with SM (sensory processing issues, developmental delay, problems using public bathrooms, eating difficulties, etc.) would benefit from clearer indication of which references correspond to which symptoms and whether these are consistently documented across studies.

8. Lines 56–87: The discussion of BI, SAD, autism, and SM is comprehensive but dense. Shortening or re-structuring this section (e.g., separating BI–SAD–SM links from autism–SM co-occurrence) could improve readability.

9. Lines 88–104: Several instances of “World Heath Organisation” should be corrected to “World Health Organization”.

10. Lines 131–137: The statement that physical difficulties “are barely mentioned” in SM research could be nuanced by acknowledging the few existing works that discuss motor or vocal inhibition and bodily symptoms, if any, to avoid suggesting complete absence of prior work.

11. Lines 138–152: The linkage to the defense cascade theory and BI research is appropriate. However, it might be helpful to clarify that the present study does not directly test these mechanistic models but uses them as a conceptual backdrop.

12. Lines 173–186 (catatonia description): This section is detailed and informative, but somewhat long relative to what is directly assessed in the study. A more concise summary emphasizing the overlap between catatonic symptoms and the freezing behaviors measured here could maintain focus.

13. Lines 212–223 (Hypotheses): Hypothesis 3 is difficult to parse (“with, in increasing order, autism alone, SM alone, or SM and autism”). Rephrasing this sentence to explicitly state the expected ordering of throat freezing across the four groups would reduce ambiguity.

Materials and Methods

14. Lines 227–238 (Recruitment): The recruitment strategy via SMiRA, NASEN, and social media is clearly described. It would be useful to specify approximate dates of data collection and whether any inclusion/exclusion criteria were applied beyond age and consent.

15. Lines 291–303 (Completion rates): A large proportion of partially completed or minimally completed surveys is described. Clarifying precisely how many parent responses were included in the analyses for each main outcome (SMQ, interference, diagnostic distribution) would help readers understand the effective sample sizes and possible attrition bias.

16. Lines 239–253 (Parent survey): The description of the SMQ is appropriate. The description of the “additional six questions” used for clinical purposes should state exactly how these were scored and combined to create the single “interference score” used later (e.g., averaging across items, handling of missing items, whether any items were reverse-scored).

17. Lines 255–269 (Child survey procedures): The procedures for parent support and child consent are clearly described. Indicating the approximate time needed to complete the child survey and whether any incentives were offered would give a fuller picture of participation conditions.

18. Lines 271–274: The statement “See the online supplement for copies of the questionnaires used” is important; ensure that the supplementary material includes both age-group versions and that item wording is clearly visible.

19. Lines 276–287 (Statistical analysis): The choice of Kruskal–Wallis and Spearman’s rank correlation is appropriate for skewed data. It would be useful to add brief information on (i) how missing data for individual SMQ items were handled, (ii) whether any multiple comparison adjustment was applied for the several group comparisons, and (iii) which software version was used for the analyses.

20. Lines 284–285: Home-schooled children are excluded from SMQ analyses due to missing school domain data. Consider stating explicitly whether their family and social domain data were also excluded, or whether an adjusted total score was computed; this is relevant because some of them are later discussed in relation to freezing behavior.

Results

21. Lines 291–303 (Parent survey response): The description of completion rates is helpful. For transparency, reporting a CONSORT-style or STROBE-style flow description (possibly in a figure or table) indicating numbers at each stage (invited, started, provided child data, completed SMQ, etc.) would strengthen the methodological transparency.

22. Table 1 (Country of residence): Table 1 is clear. Consider reporting percentages to one decimal place consistently, or using integers only, to improve visual consistency.

23. Lines 303–310 (Number and ages of children): When reporting ages and school stages, it would help to explicitly define the age bands corresponding to “infant”, “junior”, and “secondary” years to assist non-UK readers.

24. Table 3 (Diagnosis by age): Table 3 is informative. The row “Percentage of Total SM who are also Autistic” could be more clearly labeled as “Percentage of children with SM in each age band who also have autism”.

25. Lines 320–327 (Parental view vs diagnosis): The description of Table 4 is helpful. Consider adding a short interpretive sentence highlighting the main pattern (e.g., proportion of children perceived as underdiagnosed, frequency of suspected dual diagnoses) to guide the reader.

26. Lines 333–341, Figure 2: The description of SMQ scores across diagnostic groups is good. It would be helpful to specify the exact p-value for the overall Kruskal–Wallis test and whether any post-hoc pairwise comparisons were performed (e.g., SM only vs autism only vs dual diagnosis vs no diagnosis), including which groups differed significantly.

27. Lines 351–359, Figure 3: When introducing the interference results, the text currently combines “children with a diagnosis of SM, either alone or with autism” into a single group (median 2.33). Given that elsewhere the dual-diagnosis group is distinguished, indicate explicitly whether here the SM+autism and SM-only groups were pooled and why this was done.

28. Figure 3 (stacked bar chart): The color scheme and legend (Never/Seldom/Often/Always) are clear, but the figure caption could more explicitly state that the percentages within each bar sum to 100% and that the x-axis categories correspond to family, school, social domains and perceived bother for child/parent.

29. Figure 4 (scatter plot of SMQ vs interference): The figure is informative. Reporting the Spearman rank correlation coefficient and 95% CI in the text (not only on the figure) would make this result more accessible.

30. Lines 368–383 (Child survey sample description): The small size and diagnostic distribution of the child sample are appropriately noted. It would be useful to explicitly note that only one child with autism only participated and did not provide freezing data, which is a key limitation for interpreting freezing in autism-only children.

31. Figure 5 (Panels A–D):

• Panel A: The label “Non-responder” plus “Responder – Boys” and “Responder – Female” is somewhat confusing. Clarifying that this panel shows the proportion of children who did or did not complete the child survey within each diagnostic group, stratified by sex, would help.

• Panels B–C: Group labels “No”, “Yes, Selective Mutism”, “Yes, both” should be defined clearly in the caption.

• Panel D: The legend “Boy/Girl” is clear; however, specify that only children who reported at least some freezing are included in the body-part analyses if that is the case.

32. Table 5 (Responders vs non-responders): The table provides helpful context. The footnote on the single child with both diagnoses who responded but could not have a full SMQ score due to home-schooling is important; consider clarifying whether this child was included in any “total SMQ” analyses.

33. Table 6 (Child survey responses by sex): This table is rich in information. To enhance readability, consider grouping related variables (diagnosis, SMQ/interference, freezing symptoms, body parts affected, emotional responses) with subheadings within the table or re-ordering variables to match the flow of the Results text.

34. Lines 415–433 (Children’s feelings when expected to speak): This qualitative content is valuable. To strengthen the presentation, consider organizing these feelings and triggers into a small set of thematically labeled categories (e.g., performance fears, sensory discomfort, interpersonal pressure) and indicating which were most frequently mentioned.

Discussion and Conclusions

35. Lines 447–456 (Timing of diagnoses): The suggestion that SM may delay autism diagnosis is plausible and interesting. It may be helpful to add a brief caution that this conclusion is based on cross-sectional age patterns in a self-selected sample rather than longitudinal diagnostic trajectories.

36. Lines 452–457 (Proportion of SM children who are autistic): The estimate that “nearly a quarter” of children with SM are also autistic is clearly derived from this sample. Emphasizing in the same sentence that this is a proportion within this specific convenience sample (recruited largely from SM networks) would prevent misinterpretation as a population prevalence estimate.

37. Lines 479–521 (Diagnostic challenges and parental views): This section is well argued. When suggesting the importance of parental views, it may also be useful to caution that parental perceptions of under- or over-diagnosis could be influenced by their own mental health and knowledge about SM/autism; this does not undermine their value, but provides balance.

38. Lines 495–517 (Differences in speaking behaviours): The comparison with Ludlow et al. is appropriate. It would be informative to briefly speculate on possible reasons for the differing patterns in the school domain between the two studies (e.g., sample composition, severity, educational context).

39. Lines 523–534 (Life interference): The proposed use of life interference scoring as an additional discriminator between diagnostic groups is an interesting idea. However, given that these questions are not validated as a separate scale, the text might benefit from more cautious wording (e.g., describing this as exploratory and suggesting future validation work).

40. Lines 535–561 (Freezing behaviour): The discussion rightly notes that no conclusions can be drawn about the relative contribution of autism vs SM to freezing in the dual-diagnosis group. This limitation could be highlighted slightly earlier when the main results are summarized, to avoid over-attribution.

41. Lines 567–583 (Implications for treatment): The recommendation to prioritize reduction of stress and pressure, and to consider sensory processing and physical freezing in treatment planning, is clinically meaningful. It may be useful to distinguish more clearly between implications supported by the current data (e.g., reported triggers like pressure to speak, fear of mistakes) and those extrapolated from broader theory (e.g., catatonia/defense cascade mechanisms).

42. Lines 589–660 (Limitations): The limitations section is appropriately comprehensive. To align with the cautious interpretive stance, consider explicitly adding that the strong associations between SMQ scores and freezing/interference may partly reflect shared method variance (all parent-report) and that the child-report data, while valuable, are too sparse for robust subgroup comparisons.

43. Lines 668–683 (Conclusions): The conclusion that “physical freezing of both the throat and other parts of the body may be an important factor in preventing speech and movement in children with SM” is supported by the descriptive data but should be framed as preliminary and based on self-report in a small child subsample. Similarly, statements advising that autism “should also be considered a possibility and investigated when children present with SM symptoms” are reasonable but might be softened by noting that formal diagnosis requires comprehensive assessment.

Language, Style, and Formatting

44. Throughout (e.g., lines 30, 45, 93, 182, etc.): Several words appear with typesetting artifacts (e.g., “beĴer”, “Heath”). A full proofread and correction of such encoding issues is recommended.

45. Abbreviations (lines 724–725, 852–855): The list of abbreviations is helpful. Ensure that all abbreviations used in the text (e.g., BI, DD, NDD) are introduced at first use in the main text and appear in this list.

46. References (lines 726–855): The reference list is extensive and appears current. A careful check for consistency in formatting (e.g., capitalization, journal names, use of DOIs) against the journal’s reference style is recommended.

Author Response

Please see attached file

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Please see attached file.

Comments for author File: Comments.pdf

Author Response

Please see attached document 

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

accept

Author Response

No changes suggested, reviewer considered the manuscript suitable for publication

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript has improved in clarity, structure, and methodological transparency. The additional explanations regarding the SMQ interference score, recruitment characteristics, and the limitations of the child survey meaningfully strengthen the paper. The discussion is now more balanced, and speculative interpretations are more clearly separated from data-driven observations.

The remaining issues are minor and relate primarily to wording precision, consistency in terminology, and ensuring that conclusions consistently reflect the descriptive nature of the child-report data.

 

Specific Comments

1. Abstract (Lines ~16–22)
   While the wording has been moderated, the abstract still risks implying a level of generalizability that exceeds the child survey data. Phrases describing freezing as commonly occurring would benefit from explicitly anchoring this observation to the self-selected sample and small child subsample.

2. Methods – Interference Score (Lines ~344–346)
   The description of how the interference items were scored and combined is clearer. For completeness, consider explicitly stating that this score is exploratory and not a validated subscale at first mention, rather than only later in the Discussion.

3. Results – Group Terminology (Figures 3–5; Lines ~330–353)
   Terminology for diagnostic groupings is largely consistent, but occasional shifts between “SM with or without autism” and explicit four-group comparisons remain. A final pass to ensure uniform group labels across text, figures, and captions would improve readability.

4. Child Survey Results (Lines ~403–411)
   The limitation regarding the absence of freezing data from the autism-only child group is now stated, which is appropriate. Consider moving this clarification slightly earlier in the subsection to frame the results before interpretation.

5. Discussion – Mechanistic Interpretation (Lines ~576–585)
   The discussion of catatonia, freezing, and the defense cascade is thoughtful but still somewhat speculative. Although caveats are included, tightening the language to further emphasize that these models are conceptual frameworks rather than explanations tested in the present study would strengthen this section.

6. Conclusions (Lines ~721–730)
   The conclusions are largely appropriate. One or two sentences could be further softened to consistently reflect that child-reported freezing findings are preliminary and descriptive, particularly when referring to physical freezing as an “important factor.”

7. Language and Style
   Most previously noted language and encoding issues have been corrected. A final proofread focusing on sentence length and minor stylistic smoothing—particularly in the Discussion—would further improve fluency.

Author Response

We have used the comment function in word to highlight the changes

Author Response File: Author Response.pdf