Disparities in Spinal Muscular Atrophy-Related Mortality in the United States, 2018–2023

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript repeatedly refers to the “post-treatment era,” yet provides limited background on the specific disease-modifying therapies, their approval and implementation timelines, and the evidence demonstrating meaningful improvements in survival outcomes. Although the study defines 2018–2023 as the “post-treatment era,” the Introduction does not sufficiently describe the specific therapies, their approval or uptake timelines, or the empirical evidence supporting survival benefits. To justify the temporal framing of the study, a more explicit and systematic discussion of these therapies and their population-level impact is warranted.

While the selection of Hispanic individuals as the reference group is statistically acceptable, readers may question the rationale for this choice. In addition, the finding that non-Hispanic White individuals exhibit the highest mortality rates requires interpretation beyond differences in disease prevalence alone.

The Methods section provides insufficient discussion of key analytical limitations, including the inability to perform multivariable analyses and the structural constraints of the data that preclude ascertainment of treatment status.

Additionally, the ICD-10 code is incorrectly reported as “G12,8” in the Methods section and should be corrected.

The analysis is largely descriptive in nature, and care should be taken to ensure that the interpretation of results—particularly with respect to the causes of observed disparities—does not imply causal relationships.

The discussion of data suppression and reliability thresholds (e.g., counts <20) is appropriate; however, transparency would be improved by explicitly stating in the Results section whether any strata were excluded due to data suppression.

Author Response

Reviewer 1, Comment 1:
The manuscript repeatedly refers to the “post-treatment era,” yet provides limited background on the specific disease-modifying therapies, their approval and implementation timelines, and the evidence demonstrating meaningful improvements in survival outcomes. Although the study defines 2018–2023 as the “post-treatment era,” the Introduction does not sufficiently describe the specific therapies, their approval or uptake timelines, or the empirical evidence supporting survival benefits. To justify the temporal framing of the study, a more explicit and systematic discussion of these therapies and their population-level impact is warranted.

Thank you for the insightful comments. We have expanded the introduction to describe and report that timeline of the therapies’ approval between 2016-2020. Additionally, we clarified in the methods sections that the CDC WONDER multiple causes of death files are extracted by two periods: 1999-2020 and 2018-2023. Therefore, we extracted the 2018-2023 data as the database is structured in this way and it represents the post-treatment initiation and approval period. Lines: 42-52.

 

 

Reviewer 1, Comment 2:

While the selection of Hispanic individuals as the reference group is statistically acceptable, readers may question the rationale for this choice. In addition, the finding that non-Hispanic White individuals exhibit the highest mortality rates requires interpretation beyond differences in disease prevalence alone.

Thank you for the great comment. For the comparison analysis, we used the groups with the lowest AAMR as reference group and for race/ethnicity comparison, the Hispanic individuals carried the lowest AAMR. Additionally, Higher age-adjusted mortality rates from spinal muscular atrophy in white individuals likely reflect higher disease incidence in this population due to underlying genetic differences in SMN1 carrier frequencies and copy number variations, rather than differences in disease severity or access to care. We cited these genetic studies in our discussion. Lines: 178-191.   

 

Reviewer 1, Comment 3:

 The Methods section provides insufficient discussion of key analytical limitations, including the inability to perform multivariable analyses and the structural constraints of the data that preclude ascertainment of treatment status. Additionally, the ICD-10 code is incorrectly reported as “G12,8” in the Methods section and should be corrected.

Thank you for the comment. We agree that due to the constraints of the data and rarity of the condition (low overall counts), we are limited in conducting a multivariable analysis. The mortality rates are however already “age adjusted”. We also corrected the error in the ICD code. Lines: 69.

 

Reviewer 1, comment 4:

The analysis is largely descriptive in nature, and care should be taken to ensure that the interpretation of results—particularly with respect to the causes of observed disparities—does not imply causal relationships. The discussion of data suppression and reliability thresholds (e.g., counts <20) is appropriate; however, transparency would be improved by explicitly stating in the Results section whether any strata were excluded due to data suppression.

Thank you for the insightful comments. We have added this to the limitation section emphasizing the caution advised in interpretation. We also added to the results section which strata were not reportable due to suppressed counts. Lines: 215-219.

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript presents a nationwide, population-based analysis of spinal muscular atrophy (SMA)–related mortality disparities in the United States (2018–2023) using the CDC WONDER database. The topic is timely and relevant, particularly in the context of disease-modifying therapies and newborn screening expansion. The study provides useful descriptive epidemiological insights into sex-, race/ethnicity-, region-, and age-based disparities. However, several methodological clarifications, interpretative refinements, and framing adjustments are required to strengthen the rigor, transparency, and impact of the work. Many of the conclusions are hypothesis-generating and should be framed as such, given the inherent limitations of death-certificate–based data.

1. The manuscript defines 2018–2023 as the “post-treatment era.” While reasonable, this choice requires stronger justification.
2. The authors need to clarify why 2018 was selected as a uniform starting point and discuss whether treatment penetration was sufficiently widespread across age groups and regions during this period.
3. The authors should explicitly acknowledge that adult SMA populations may not uniformly reflect post-treatment outcomes, particularly in earlier years of the study window.
4. SMA cases were identified using ICD-10 codes G12.0, G12.1, G12.8, and G12.9. Codes G12.8 and G12.9 may capture unspecified or non-5q motor neuron diseases, introducing potential misclassification bias. Please discuss how inclusion of these codes may affect specificity and whether sensitivity analyses excluding G12.8/G12.9 were considered.
5. The authors need to clarify whether SMA was treated as an underlying cause of death or included as any contributing cause, and how this impacts interpretation of mortality rates.
6. Interpretation of Mortality Disparities Without Prevalence Adjustment. The analysis relies exclusively on mortality data without accounting for underlying SMA prevalence across demographic groups.
7. The authors need to emphasize the limitations resulted from higher age-adjusted mortality rates (AAMR), particularly among NH White individuals, may reflect higher disease prevalence rather than higher case fatality.
8. This limitation should be emphasized more clearly in both the Results interpretation and Discussion.
9. Conclusions should avoid implying differential lethality without prevalence-adjusted denominators.
10. The study conducts multiple subgroup comparisons across sex, race/ethnicity, region, and age. Please clarify whether any adjustment for multiple comparisons was considered or why it was deemed unnecessary. While the analysis is descriptive, readers may over-interpret p-values without this context.
11. The reported male–female difference in AAMR is statistically significant but modest in absolute terms. The Discussion should better contextualize the clinical and public-health significance of this difference.
12. The authors need to clarify why crude mortality rates rather than age-adjusted rates were emphasized for age-group analyses.
13. The authors need to ensure consistent use of terms such as mortality, mortality rate, and age-adjusted mortality rate throughout the manuscript.

Author Response

Reviewer 2, Comment 1:

This manuscript presents a nationwide, population-based analysis of spinal muscular atrophy (SMA)–related mortality disparities in the United States (2018–2023) using the CDC WONDER database. The topic is timely and relevant, particularly in the context of disease-modifying therapies and newborn screening expansion. The study provides useful descriptive epidemiological insights into sex-, race/ethnicity-, region-, and age-based disparities. However, several methodological clarifications, interpretative refinements, and framing adjustments are required to strengthen the rigor, transparency, and impact of the work. Many of the conclusions are hypothesis-generating and should be framed as such, given the inherent limitations of death-certificate–based data.

1. The manuscript defines 2018–2023 as the “post-treatment era.” While reasonable, this choice requires stronger justification.

Thank you for the insightful comments. We have expanded the introduction to describe and report that timeline of the therapies’ approval between 2016-2020. Additionally, we clarified in the methods sections that the CDC WONDER multiple causes of death files are extracted by two periods: 1999-2020 and 2018-2023. Therefore, we extracted the 2018-2023 data as the database is structured in this way and it represents the post-treatment initiation and approval period. We added this clarification on the methods section. Lines: 63-65.

 

 

Reviewer 2, Comment 2:

The authors need to clarify why 2018 was selected as a uniform starting point and discuss whether treatment penetration was sufficiently widespread across age groups and regions during this period.

Thank you for the comment. We have clarified the starting point of the study in the methods section and expanded the introduction to clarify and detail the timeline of approval of the treatments. However, this population-level data focuses on all-cause mortality data and does not contain information on receipt of treatment or distribution of treatment across regions. Lines:63-65, and 42-52.  

 

Reviewer 2, Comment 3:

The authors should explicitly acknowledge that adult SMA populations may not uniformly reflect post-treatment outcomes, particularly in earlier years of the study window.

Thank you for the great comment. We have added this important point to the limitation section. Lines: 215-218.

 

Reviewer 2, Comment 4:

SMA cases were identified using ICD-10 codes G12.0, G12.1, G12.8, and G12.9. Codes G12.8 and G12.9 may capture unspecified or non-5q motor neuron diseases, introducing potential misclassification bias. Please discuss how inclusion of these codes may affect specificity and whether sensitivity analyses excluding G12.8/G12.9 were considered.

Thank you for the great comment. The ICD codes listed on the CDC WONDER reflect the ones used on death certificates. The motor neuron disease code (G12.2) is different than the ones used for SMA, and this MND code was excluded from this analysis. Only codes specific to SMA were included. When divided by the two suggested ICD codes (G12.0 and G12.1), there were 174 deaths across the period, which is not possible to stratify and analyze accurately. The other two codes (G12.8 and G12.9) carried 647 total deaths.  

 

Reviewer 2, Comment 5:

The authors need to clarify whether SMA was treated as an underlying cause of death or included as any contributing cause, and how this impacts interpretation of mortality rates.

Interpretation of Mortality Disparities Without Prevalence Adjustment. The analysis relies exclusively on mortality data without accounting for underlying SMA prevalence across demographic groups.

Thank you for the insightful comment. As mentioned in the methods section, the multiple-cause-of-death files were used to extract the data including SMA as both contributor and underlying cause of death. Notably, the total number of SMA-related deaths was 821, with 565 deaths showed SMA as underlying cause of death. We added this data point to the main results section. Lines: 90-91. Additionally, the data from the CDC WONDER/death certificates do not contain prevalence data. 

 

Reviewer 2, Comment 6:

The authors need to emphasize the limitations resulted from higher age-adjusted mortality rates (AAMR), particularly among NH White individuals, may reflect higher disease prevalence rather than higher case fatality. This limitation should be emphasized more clearly in both the Results interpretation and Discussion.

Thank you for the great comment. We have discussed the role played by higher prevalence in NH White individuals including genetic prevalence studies and carrier studies in our discussion section. Lines: 176-188.

 

Reviewer 2, Comment 7:

Conclusions should avoid implying differential lethality without prevalence-adjusted denominators.

Thank you for the comments. We have revised the conclusion to avoid implying differential case fatality or lethality between the comparison groups given that we are unable to obtain prevalence-adjusted denominator. Lines: 224.  

 

Reviewer 2, Comment 8:

The study conducts multiple subgroup comparisons across sex, race/ethnicity, region, and age. Please clarify whether any adjustment for multiple comparisons was considered or why it was deemed unnecessary. While the analysis is descriptive, readers may over-interpret p-values without this context.

Thank you for the great comment. The mortality rates are age-adjusted in this study and rates are standardized to the US population. A multivariable analysis would not be possible due to low counts when stratified by year of death, age group, sex, race/ethnicity, and regions, simultaneously. SMA is a rare condition and hence counts are understandably low.

 

 

\Reviewer 2, Comment 9:

The reported male–female difference in AAMR is statistically significant but modest in absolute terms. The Discussion should better contextualize the clinical and public-health significance of this difference.

Thank you for the comments. We have expanded on this point in our discussion section. Lines: 174-176.

 

Reviewer 2, Comment 10:

The authors need to clarify why crude mortality rates rather than age-adjusted rates were emphasized for age-group analyses. The authors need to ensure consistent use of terms such as mortality, mortality rate, and age-adjusted mortality rate throughout the manuscript.

Thank you for the comments. When stratified by age groups, the CDC WONDER software does not allow age-adjusted rates because the crude mortality rates already reflect age specific rates. We have revised the manuscript to be as consistent as possible in the terms used. However, it is essential for readers to differentiate crude rates versus age-adjusted rates. Lines: 69-71.

 

 

 

   

 

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Your detailed responses to the reviewers’ comments are greatly appreciated.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have responded to the comments appropriately.