RESULTS & CONCLUSIONS

Conclusions are generally a bit overstated, it would be better to keep them closer to study findings (eg. "study suggests a link between this tar-stained fingers and addictive behavior or concomitant high alcohol consumption"

GENERAL COMMENTS

In abstract:

1. Conclusion is vague and needs to be modified, maintaining the focus on the study findings.

2. Giving OR and 95% CI for non-significant results in Abstract/Results is unnecessary; other more meaningful results could be presented instead.

3. Abstract/Objective: objective was to investigate association between tobacco stains and: (1) tobacco related disease and (2) behavioral and environmental characteristics. Currently objective 2 suggests only harmful alcohol use (", or of harmful alcohol use, independent ...") so this should be expanded to clarify to the reader what was the scope of the study.

Article summary

1. First key message is unclear -- should be rephrased so that it is informative even for a reader who did not read the whole paper

2. Additional limitation is that study only includes only hospital population.

Introduction

1. Paragraph 2, line 9: "...and is a reliable marker of concurrent disease" -- not clear

Methods

1. Give full name for DSM-IV criteria

2. Details on occupation were collected, but this variable was not used in the present analysis (other than unemployment)?

3. power is calculated for 50 case-control pairs, but 49 were used in the analysis. This should be commented on.

Discussion

1. clarify what does "A tar stained finger gives clue for a tobacco related conditions four times out of five" mean.

2. lack of an association with other investigated tobacco related illnesses could arise from insufficient power of the study

3. To address - Could peripheral circulation affect tar deposition or removal, or rate of skin cell regeneration?

4. At its current form conclusion that "yellow staining is rather a proxy of the consequences of tobacco: its presence in itself seems to identify high/risk smokers due to the tobacco exposure, rather then a specific increased susceptibility" is overstating the results. Either explain better or keep conclusions closer to results. Results show association to PAD -- avoid generalization to association with tobacco related illnesses as only PAD was significant
   Table 4. It seems 50% of cases have harmful alcohol use and 67% are on psychotropic medication. Are these overlapping? Since association between harmful alcohol use is one of the main take home messages of the study, maybe additional analyses could clarify the associations.