Regional Variation and Socio-Economic Determinants of Suicide Mortality in Greece before and during Economic Crisis

Round 1

Reviewer 1 Report

Dear authors,

The study is interesting and easy to read but I have some recommendations and questions about it.

• Do not use acronyms in the abstract.
• Describe the acronyms the first time that you use them in the text.
• The material and methods section must inform at the beginning of the study type. 
• The material and methods section must include the next subsections: sample selection, study variables and data collection, data analysis, ethical aspects.
• How did you get all the information about the socio-economic and mental health variables? 
• I do not understand the change % in SDR in Table 1. How is it calculated?

Author Response

POINT-BY-POINT RESPONSE

Reviewer 1

We would like to thank the reviewer for constructive comments which helped us to improve the overall quality of our manuscript.

Comments and Suggestions for Authors

The study is interesting and easy to read but I have some recommendations and questions about it.

• Do not use acronyms in the abstract.

Thank you for the comment. We have revised the text, accordingly. (page 2, lines: 2-20)

• Describe the acronyms the first time that you use them in the text.

We have revised the text, accordingly (page 3, lines 10-11, page 4, lines 4,6,10).

• The material and methods section must inform at the beginning of the study type.

We agree with the reviewer. We have included a description of the study type in abstract and “methods” (Page 3; lines 31-32).

• The material and methods section must include the next subsections: sample selection, study variables and data collection, data analysis, ethical aspects.

Thank you for the suggestion. We have restructured the methods section according to the suggestion. Considering that, the study is not based on a “sample” but on datasets including the total number of deaths, instead of “sample selection” and “data selection” we used the term “Data selection”. In addition, the study design requires a separate mention on “period identification” (page 3; lines: 31 to page 5 line 29).

• How did you get all the information about the socio-economic and mental health variables?

The reviewer made a good point. In response to that, we have included all data sources in the “Data selection” paragraph of the revised manuscript (page 4; lines: 1-9).

• I do not understand the change % in SDR in Table 1. How is it calculated?

We would like to thank the reviewer for giving us the opportunity to further clarify this point. “Change of SDR” is the difference of SDR between of the two periods expressed as a percentage; in other words, the change of SDR (SDR₂ - SDR₁) divided by SDR₁. It is calculated as (SDRn₊₁-SDR_n)*100/SDR_n (Revised version; statistical analysis section; page 5; lines: 16-17).  We have also added this clarification as a note in table 1.

Author Response File: Author Response.pdf

Reviewer 2 Report

Overall comments:

The manuscript focus a relevant topic, being of pivotal importance in the context of the current pandemic.

However, I suggest that the authors follow the STROBE guidelines, in order to have a complete description of the study.

Specific comments:

ABSTRACT

1. Following STROBE guidelines, please indicate indicate the study’s design with a commonly used term in the title or the abstract.

2. Please indicate in the Background the period of the economic crisis. In a country’s history, we can find several crisis. Although we can anticipate which crisis the authors will address in the manuscript, only when we go to the methods description in the Abstract we are sure which is the period under consideration.

INTRODUCTION

1. The sentence “The impact mostly depends on the severity and duration of socio-economic effects…” is misleading. The authors are saying that the effects of the crisis depend on the effects themselves. Do the authors mean that the impact depend on the severity and duration of socio-economic conditions/context?

2. In order to better characterize the unemployment rate before and after/during the crisis, please provide information before 2010. As it is, I infer that from 2005 to 2010 the rate did not increase significantly which means that the effects if the crisis in terms of unemployment rate are not immediate. Do the authors have more information to better characterize these variations?

MATERIALS AND METHODS:

1. Please indicate early in this section the study design (it is only indicated in the Discussion)

2. Please provide more information regarding the study population and inclusion criteria. For example, which ages were considered (only in the Results we see that individuals with 15 years old or more were considered)

3. Please provide a clear definition of the variables considered, instead of just list them

4. Please provide a more detailed description of the statistical analysis performed

RESULTS

1. Is it possible to indicate the p-values in Table 1 and Table 2 (since they are presented in the text)? It will be easier to read immediately from the table.

2. Is it possible to put in bold the significant values in the tables?

DISCUSSION

1. The authors say “The magnitude of the differences observed in Greece and the fact that they exist already before the crisis…”. However, they only insufficiently address why this happens. Please present possible arguments to justify the differences before and during the crisis. It is only due to socio-economic or mental health determinants? Do the referred studies have limitations?

2. “This finding means that other not-investigated factors determine suicide patterns, underlining a needed for further research.” Can the authors raise some ideas regarding which possible not-investigated factors can be?

3. “However, in our study, no association was found with SDR of overall mental disorders, organic and non-organic psychoses and alcohol dependence syndrome. The lack of association may be due to the low levels of mortality of these conditions.” Why? How can the authors infer that this can be due to the low levels of mortality of the conditions? Can you compare/support the findings with the results from other studies?

Final remarks:

Is advisable that the manuscript goes through a thorough revision since there are some typos and some sentences need some clarification. For example:

- Financial crisis resulted to a dramatic and unprecedented in Europe deterioration of 160 the socio-economic conditions in the country. (Delete “in Europe”)

- A systematic review of the association between unemployment and suicide found that unemployment might increase the risk of suicide by 250% within the first 5 years, the rates remaining elevated by 20% even after 12-15 years (revise the English)

I recommend major reviews before the acceptance of this paper.

Author Response

POINT-BY-POINT RESPONSE

Reviewer 2

We would like to thank the reviewer for constructive comments which helped us to improve the overall quality of our manuscript.

Comments and Suggestions for Authors

Overall comments:

The manuscript focus a relevant topic, being of pivotal importance in the context of the current pandemic.

However, I suggest that the authors follow the STROBE guidelines, in order to have a complete description of the study.

Thank you for the suggestion. Responding to that, we applied the STROBE guidelines. Below, we provide a summary checklist of their application in our study.

Summary STROBE checklist of the study

Specific comments:

ABSTRACT

1. Following STROBE guidelines, please indicate the study’s design with a commonly used term in the title or the abstract.

We agree with the reviewer. We have included a description of the study type in abstract (page 2, line 6) and in “methods” (page 3; lines 31-32).

2. Please indicate in the Background the period of the economic crisis. In a country’s history, we can find several crisis. Although we can anticipate which crisis the authors will address in the manuscript, only when we go to the methods description in the Abstract we are sure which is the period under consideration.

We have modified the text of abstract, accordingly (“after the 2008 financial crisis”) (page 2, line 2).

INTRODUCTION

1. The sentence “The impact mostly depends on the severity and duration of socio-economic effects…” is misleading. The authors are saying that the effects of the crisis depend on the effects themselves. Do the authors mean that the impact depend on the severity and duration of socio-economic conditions/context?

Thank you for the comment. The expression is really misleading because of the lack of a word. The correct expression is: “The impact on suicidality is mostly on the severity and duration of socio-economic effects” and we have revised the text, accordingly (page 3, lines 5-6).

2. In order to better characterize the unemployment rate before and after/during the crisis, please provide information before 2010. As it is, I infer that from 2005 to 2010 the rate did not increase significantly which means that the effects if the crisis in terms of unemployment rate are not immediate. Do the authors have more information to better characterize these variations?

We recognize this omission, which was mainly due to space saving reasons. Responding to the remark, we modified table 3 to include average values of the independent variables. (page 19, column 2-3).

Regarding to unemployment rate, it was 7.8% in 2008, 9.6% in 2009, increased to 12.7% in 2010 and then it climbed to 17.9% in 2011 and finally to 24.4% in 2012. Thereafter, it remained at similar levels until 2016. As it is shown in figure 1, the suicide mortality rate displayed a significant increase since 2011, without any lag in relation to unemployment. However, it is difficult to provide all this information in the paper. In response to the comment, we have included in table 3 the average values of all independent variables, to provide the reviewer with some information (page 19). In addition, we have included a supplementary file including two tables and figures with all annual values of (a) the socio-economic and (b) mental health-related variables. (Supplementary material, Appendix S1-S2) We hope that these two additions provide adequate information on the developments. 

MATERIALS AND METHODS:

1. Please indicate early in this section the study design (it is only indicated in the Discussion)

Thank you for the comment, we have included the study design in the revised version of the manuscript. (Page 3; lines 31-32).

2. Please provide more information regarding the study population and inclusion criteria. For example, which ages were considered (only in the Results we see that individuals with 15 years old or more were considered)

The comment is very helpful. Although in the methods section it is clarified that data included all deaths due to suicide by age, in table 3 we included only the rates of the age groups over 15 years. That was because of the small number of deaths recorded in this age group. Now, that point has been clarified, by adding the following note in table 3. “*The age group of 0-14 years was not included in the table, since only four deaths of this age have been recorded during the entire 2011-6 period” (page 19, table note).

3. Please provide a clear definition of the variables considered, instead of just list them.

Thank you for the comment. We added definitions of GDP, DHI and unemployment rate, as well as the measurement unit of the psychotropic drug consumption (page 4; lines 28-32, page 5, lines 1, 9-10).

4. Please provide a more detailed description of the statistical analysis performed.

We have added a separate paragraph summarizing statistical analyses performed (pages 5, lines: 14-22).

RESULTS

1. Is it possible to indicate the p-values in Table 1 and Table 2 (since they are presented in the text)? It will be easier to read immediately from the table.

Thank you for the comment. Undoubtedly, it was an omission and a loss of information. Now p values have been added in both tables, with the needed clarifications. (page 17, line 4, page 18, line 3)

2. Is it possible to put in bold the significant values in the tables?

We have revised the tables, accordingly. (pages 17,18)

DISCUSSION

1. The authors say “The magnitude of the differences observed in Greece and the fact that they exist already before the crisis…” However, they only insufficiently address why this happens. Please present possible arguments to justify the differences before and during the crisis. It is only due to socio-economic or mental health determinants? Do the referred studies have limitations?

This is a good comment, which motivated us to further discuss regional variation. Responding to that, (a) we mentioned in brief the findings of reviews on suicide risk factors (page 8; lines: 1-5), (b) we performed some more correlation analyses, focusing only on the pre-crisis regional variation (2001-2010) the results of which are mentioned in discussion (page 8, lines: 6-14) and (c) we discussed about other possible determinants (page 8, lines: 6-14, page 9, lines 13-14, page 10, lines: 1-5, 13-15). All these additions have been included in the second paragraph of the discussion section (page 8; lines: 1-15 and page 9; lines 13-14, page 10, lines: 1-5, 13-15).

2. “This finding means that other not-investigated factors determine suicide patterns, underlining a needed for further research.” Can the authors raise some ideas regarding which possible not-investigated factors can be?

Thank you for pointing it out. This point was partly addressed together with the previous comment, where we discussed the risk factors of suicidality. At this point, we included a short mention to what already was discussed. In addition, in the next paragraph of “discussion”, where the mental health-related factors are discussed, we included an additional parameter, which has been described as “deaths of despair” (page 10;lines: 1-5).

3. “However, in our study, no association was found with SDR of overall mental disorders, organic and non-organic psychoses and alcohol dependence syndrome. The lack of association may be due to the low levels of mortality of these conditions.” Why? How can the authors infer that this can be due to the low levels of mortality of the conditions? Can you compare/support the findings with the results from other studies?

We agree that it is a reasonable question. Therefore, in appendix S2, we included data on the annual rates of the mental disorder mortality, which support their low levels in Greece. The problem is that, due to low mortality, in some regions and/or years, the number of events may be zero or the fluctuation of the rates may be misleading in the small regions, and finally the observations may be insufficient to produce statistically significant results in correlation and regression analyses. However, looking at the numbers of table 3 and appendix S2, we find out that after 2012, a significant increase in mental disorder mortality was recorded in the country. Is it associated with the increase in suicide mortality? Because of small numbers, we cannot give an answer based on statistical results.

Now, responding to the comment, we included some more clarifications on that, as follows:

In some regions and years, the number of events may be zero or very small, and therefore, the corresponding observations misleading and inappropriate to produce statistically significant results in correlation and regression analyses. However, looking at the numbers of table 3 and appendix S2, we find out that after 2012, a sharp increase in mental disorder mortality was recorded. This increase is may be associated with the rise in suicide mortality, but it cannot be statistically supported (page 9, lines: 30-33, page 10, lines: 1-2).

Final remarks:

Is advisable that the manuscript goes through a thorough revision since there are some typos and some sentences need some clarification. For example:

- Financial crisis resulted to a dramatic and unprecedented in Europe deterioration of the socio-economic conditions in the country. (Delete “in Europe”)

Thank you for the correction. (page 7, line 19).

- A systematic review of the association between unemployment and suicide found that unemployment might increase the risk of suicide by 250% within the first 5 years, the rates remaining elevated by 20% even after 12-15 years (revise the English)

Thank you. It was rephrased. (page 7, lines: 21-24).

I recommend major reviews before the acceptance of this paper.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear authors,

You have include the next information in line 61 "we applied a quasi-experimental approach". In mi opinion it would be better to indicate a "case-control study approach" because you did not include an intervention.

Kind regards

Author Response

Thank you for your comment

Case control studies refers to individual data while ecological studies to population data.

Our study is based on population data obtained from the Greek Statistical Authority. Consequently, our study has an ecological design

Reviewer 2 Report

The authors properly addressed all my previous suggestions. I have no further questions. I only detected some minor spelling mistakes: For example, in line 27 of the abstract, there are two commas and two periods. So, I recommend an overall revision of the manuscript.

After this, the manuscript is satisfatory to be published.

Author Response

Thank you for your comment we revise the manuscript accordingly