Assessing the Impacts of COVID-19 and Social Isolation on Mental Health in the United States of America

Round 1

Reviewer 1 Report

General: The article raises an interesting research field concerning the effect that social isolation and COVID-19 infection and related death had on the prevalence of anxiety and depression in the general population of the USA in a state-by-state multiple time-series analysis. This is a very pertinent issue and the article deals well with this.

Introduction: The theoretical part is well developed. A clear stating and focusing of the argument is provided.

Methods: The data used in the study was collected and aggregated by the Delphi Research Group at Carnegie Mellon University in partnership with Facebook. The time series data described is non-stationary and co-integrated and thus was analyzed using Vector Error Correction Models (VECMs). Methods are appropriate and the fit between theoretical discussion and methodology is well formulated.

Results: Results are linked suitably to the other sections of the article. A well-organized and compelling discussion section is provided as well. The results are of interest for practice, public health policy and society more generally. The conclusions are linked to the hypothesis and background characteristics incorporated into the results.

Overall: The author(s) address(es) a significant research subject and presents interesting field material. This article could be a starting point for further research.

Quality of language: The quality of communication is good. The author(s) has/have paid attention to the clarity of expression and readability, such as sentence structure.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The approach is novel, and the results are interesting. I don't have the expertise to comment on the statistical methods, but the Discussion seems too generic and could be better aligned with the specific results (rather than just a ballpark accounting of the total numbers of positive or negative associations).

 Based on the Figures, the key results seem to be that COVID related incidence was most consistently related to anxiety than COVID-related deaths or social isolation (either measure) but primarily only in the more recent (2nd) lag, while incidence, deaths, and being homebound (less time spent with others) all were consistently related to depression symptoms but also only in the more recent (2nd) lag.  And that increases in incidence were consistently associated with both anxiety and depression symptoms in the mid-pandemic period (although only in a minority of states), but increases in deaths and social isolation had a mixed association with anxiety and depression symptoms that was positive in some states but negative in a comparable (or in one case, greater, i.e., anxiety related to increased amount of being homebound) number of states.  If my reading of the Figures is correct, then the authors over-state the findings by concluding that "incidence and death played significant roles in determining the prevalence of anxiety and depression in both time periods" (p. 15) and "social isolation did have a significant impact on anxiety and depression" (p. 16).  This could be more precisely stated in the Discussion, with commentary on the differences in association for anxiety vs. depression as well as the consistent findings for both.

 The most striking findings from the mid-COVID period (Time 2) were the widespread association of incidence with anxiety (perhaps suggesting the "weathering" effect of the extended pandemic), and the generally negative associations between increases in being homebound interpersonally (time w/others) with anxiety and deaths with being homebound (work outside home) with depression symptoms (perhaps suggesting that increased social contacts personally as things opened up after the early COVID surges was associated with worry about contamination, and increased work may have been associated with depression symptoms representing emotional exhaustion and burnout on top of the stressors of the year or more of COVID).

  The Figures would be more accurately labeled as "Association of COVID-19 with ..." than "Effect of COVID-19 on ..." (the latter incorrectly implies causality).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

I have enjoyed reading the paper and please see my attached feedback. I would make sure you ask at least 1-2 friends reading the paper thoroughly to make sure it is at a much better level, especially for minor language issues. I have pasted the feedback below as well in case the attached is missed.
……… feedback is below and also attached as a document

Comments for author File: Comments.pdf

It is quite good 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Excellent revision, much more specific discussion of the results.

Minor edit: "results" in line 404 should be "result"