Associations of Preconception Exposure to Air Pollution and Greenness with Offspring Asthma and Hay Fever
, Thomas Halvorsen 7, Joachim Heinrich 3,11, Ole Hertel 10, Gerard Hoek 12, Mathias Holm 13, Kees de Hoogh 14,15, Christer Janson 16, Andrei Malinovschi 17, Alessandro Marcon 6, Torben Sigsgaard 18, Cecilie Svanes 1,2,†and Ane Johannessen 2,†Round 1
Reviewer 1 Report
The present paper analysed if greenness and air pollution exposure in parents' childhood affect offspring asthma and hay fever. It is current and interesting field of study. The paper analysed a consistent sample and the authors concluded that the parental exposure to air pollution appears to influence risk for asthma and allergies in future offspring.
Some minor revisions are required to improve the paper:
- The authors must explore the literature review developed in this field in the Introduction section;
- In the end of the Introduction section I suggest to authors insert a brief paragraph about the paper structure.
- The authors must explore the limitations of the work done and give some insights to future studies in this field. The conclusion section is so short.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Summary
This study aimed to investigate the effect of paternal exposures to greenness and air pollution on their offspring’ asthma and hay fever risk. It used a potentially-sufficient sample of 1106 parents with 1949 offspring in select cities in Norway and Sweden. The authors found evidence of maternal exposure to air pollution during childhood as a risk factor for offspring asthma and hay fever. Although this study posed an interesting research question, and appears to have the data to answer the question, its methods, results, and discussion are lacking in details, organization, and synthesis of findings.
Major points
Exposure assessment. These appeared to have been averaged over long time periods. For outcomes such as offspring hay fever and asthma, it is difficult to be convinced that averaged exposures over a long period would be the appropriate exposure measure. It would also be appropriate for exposures to be summarized in the main text.
Air pollution modeling: do you have any metrics on the accuracy of the “back-and-forth” extrapolation of your air pollution estimates?
Statistical analysis. A table or better organized description of what statistical models are run are needed. Which exposures are included in which models? How are they different from each other? Were there analyses stratified by offspring sex? Is stratification by country sufficient? Was stratification by center considered?
Were interactive effects between exposures considered?
Modeling with exposures as tertiles. What are the values for these tertiles? It is unclear what “low”, “medium”, and “high” levels of exposures mean, and how they relate to the current air pollution limit values, which the authors conclude as being too high. Details about these limit values were lacking; it is unlikely these limit values would average over similar periods as the exposure windows in this study.
Table 2 cannot be read in full.
More details about the mediation analysis, its assumptions, and interpretation are needed. Why were results for only select few of the pollutants shown? The authors mention on line 314 that this study conducted multiple testing with many combinations of exposures and outcomes. Limited display of results in Table 3 support the possibility that the associations shown were spurious.
Table 3 finds that for asthma, medium exposures had larger magnitude effects on odds ratio (I assume, since a logistic model was run) for asthma. Why do you think this is? Please discuss.
Response rate to surveys were ~40%. How does this bias your results? What is the potential for selection bias in your study?
Please discuss the possibility of other sources of bias. Were your results confounded?
From Results onwards, the prose was not well written. Lines 197-199 show extreme carelessness. Paragraph usage is questionable, with new paragraphs separating sentences that can easily be in one paragraph (e.g. limitations paragraph).
Can you explain why results are different for the air pollutants you investigated? What are the different potentials of each pollutant for epigenetic effects, which are proposed as the possible explanation for your findings?
Minor points
Grammatical editing is needed (e.g. line 252)
Sex and gender were both used. Please make sure to use one term consistently or to note why one is used in some areas but not others.
The result on line 48: (OR 2.15, 95%CI 1.28-3.61, OR 2.19, 95%CI 1.32-3.64) cannot be found in the main text or main tables
Directed Acyclic Graphs in Supplemental Material come with no explanation on what the different colors on the graph mean.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
1, "Figure 1. XXXXX” is not directly under the title. It is the same for Figure 2.
2, Table 2 can be seen only partially.
3, It seems that Table 3 can be seen only partially (Table 3a only has information about PM10, while table 3b have PM10, PM2.5 and O3).
4, Where is the evidence that support "paternal high BC exposure reduced the risk of offspring early onset asthma relative to low exposure"?
5, The discussion should be improved. For example, line 263-264, "The observed association in the present study between O3 and hay fever supports the belief in a similar paternal pattern for air pollution exposures." There should be one or more citation for this sentence.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
The manuscript is of great value to air pollution investigations. It is a significant attemp to anwers the question about long term exposure to air pollutants below european and WHO guidelines and the relationship abouit the state of the environment and environmental health.
However, such great value becomes an issue if authors do not pay attetion to the following recommendations:
-Introduction must be complemented after an intensive literature reveiw on air pollution exposure
-Results are interesting but there is not detailed explanation or description of them. I would suggest that it may be beneficial for this study to include comparison data when applicable, for example by finding similar research to contrast the results. The way the results are presented fit very much an informative report and less a scientific paper.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Thank you for addressing my comments.
Please edit the following.
1. Line 191-2. WHO and EU recommendations. Specify averaging times.
2. Lines 367-370. Please rewrite. Current form is too colloquial.
Author Response
Thank you for addressing my comments.
Please edit the following.
Point 1: Line 191-2. WHO and EU recommendations. Specify averaging times.
Response 1: Thank you for suggesting this. We have added the averaging times for the WHO and EU recommendations as follows (line 190-192): “Only annual mean values for PM2.5 and PM10 exceeded the WHO-recommendations in some centres (PM2.5 for parents 0-18 years in Umea, Uppsala and Bergen; parents 0-18 years and offspring 0-10 years in Gothenburg; PM10 for parents 0-18 years in Uppsala and Gothenburg). No annual mean exposures exceeded the recommended EU-values (Table S12).”
Point 2: Lines 367-370. Please rewrite. Current form is too colloquial.
Response 2: We agree with you and have rephrased the paragraph (line 347-353): “Second, in the current study we tested numerous exposures for associations with the outcomes. This multiple testing can increase the possibility of more false positive findings due to type 1 error (Patel CJ et al, J Epidemiol Community Health, 2014). However, due to a relatively low sample size we rather believe there may be an under-estimation rather than over-estimation of the associations in our analyses.”
