![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4749793%2Fce7530819a89de2ff5559e2ff6fc8c1e%2FFVC-Smoking_Status.png?generation=1597348209182236&alt=media\")
We can see that ex-smokers have higher FVC than patients who never smoked, it is just a coincidence because not having enough data to generalize?
\n
Thanks for reading, this is my first competition and I still have to learn a lot! 😄
", "messages": [ { "id": "969603", "postDate": "08/13/2020 19:54:07", "content": "We can see that ex-smokers have higher FVC than patients who never smoked, it is just a coincidence because not having enough data to generalize?
\n
Thanks for reading, this is my first competition and I still have to learn a lot! 😄
", "rawMarkdown": "We can see that ex-smokers have higher FVC than patients who never smoked, it is just a coincidence because not having enough data to generalize? \n\n\n\nThanks for reading, this is my first competition and I still have to learn a lot! 😄", "votes": null }, { "id": "969707", "postDate": "08/13/2020 21:47:04", "content": "I don't have the numbers next to me right now, but I think there might be one or several reasons for this to happen :
\n-the measure of the FVC might be biased. Smokers usually have more troubles breathing so they might be used to testing their FVC, unlike a non smoker. Tests are prone to a high variance, if you look at the curves.
\n-smokers have a lower life expectancy on average, which means that if they survive it implies that they potentially have a larger FVC than average (survivor bias)
\n-the sample might be built on that particular purpose
\n-it could be randomness but it seems very unlikely
\nThere are actually many potential reasons.
", "rawMarkdown": "I don't have the numbers next to me right now, but I think there might be one or several reasons for this to happen :\n\n-the measure of the FVC might be biased. Smokers usually have more troubles breathing so they might be used to testing their FVC, unlike a non smoker. Tests are prone to a high variance, if you look at the curves.\n\n-smokers have a lower life expectancy on average, which means that if they survive it implies that they potentially have a larger FVC than average (survivor bias)\n\n-the sample might be built on that particular purpose\n\n-it could be randomness but it seems very unlikely\n\nThere are actually many potential reasons.", "votes": null }, { "id": "969897", "postDate": "08/14/2020 03:36:00", "content": "Maybe these smokers have been compensating for their dead space by moving more air?
\nI saw the same phenomenon, but remember the same participants FVC is measured several times across ~60 weeks. Our sample size of individual current smokers is actually quite small. (2 women and 7 men)
", "rawMarkdown": "Maybe these smokers have been compensating for their [dead space](https://en.wikipedia.org/wiki/Dead_space_(physiology)) by moving more air? \n\nI saw the same phenomenon, but remember the same participants FVC is measured several times across ~60 weeks. Our sample size of individual current smokers is actually quite small. (2 women and 7 men)", "votes": null }, { "id": "974706", "postDate": "08/18/2020 02:47:25", "content": "Hi! @franarenas, I identified the same behavior on FVC, I attribute this to the small sample of the data for Currently Smokes
", "rawMarkdown": "Hi! @franarenas, I identified the same behavior on FVC, I attribute this to the small sample of the data for Currently Smokes", "votes": null }, { "id": "975026", "postDate": "08/18/2020 05:51:07", "content": "Best of luck
", "rawMarkdown": "Best of luck", "votes": null }, { "id": "1007777", "postDate": "09/12/2020 13:32:33", "content": "I was also quite surprised to see this.
\nI barely believed that it is a random effect, so I started to search for official statistics.
If one goes through several statistical studies on the influence of smoking on the respiratory functionality, one can see that they mostly show lower FVCs for smokers. However, there are some more or less recent studies which show different statistics - larger FVC for smokers.
\nThis effect, however, mostly relates to some \"lurking variable\", or some \"third effect\", which influences FVC apart from smoking.
For instance:
\nhttps://www.atsjournals.org/doi/pdf/10.1164/rccm.201512-2492OC
\n--> smoking doesn't associate with lower respiratory functionality for the people with childhood onset persistent asthma
https://biomedpharmajournal.org/vol11no2/effects-of-cigarette-smoking-and-age-on-pulmonary-function-tests-in-%E2%89%A5-40-years-old-adults-in-jordan/
\n--> number of years a person smoked is negatively correlated with the FVC, however, the more cigarets per day a person smokes, the higher FVC he/she has
and this funny statistics on university students:
\nhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC6709160/
\n--> cite \"Smokers reported significantly more wheezing and sputum production than nonsmokers, but exhibited better FVC and FEV1 values\". Moreover, those who smoke more cigarets per day, have better FVC than those who smoke less.
This is a really unexpected result to me, especially the last study was done on young adults and not those, who went to the doctors with complaints, but just a sample of students. So here one can't associate the result with a theory that smokers more often do lung checks, thus their values are higher, as the non-smokers come only when they have bad complaints.
\nI still didn't find any reasonable biological explanation for that, but one perhaps should count with the fact that the FVC-(Smoking Status) correlation is not that straightforward thing. Also, we can never exclude some \"hidden effects\" and biases. Like, the ex-smokers might have also exercised sport more often in the recent years or whatever else.
", "rawMarkdown": "I was also quite surprised to see this. \nI barely believed that it is a random effect, so I started to search for official statistics.\n\nIf one goes through several statistical studies on the influence of smoking on the respiratory functionality, **one can see that they mostly show lower FVCs for smokers**. However, there are some more or less recent studies which show different statistics - larger FVC for smokers.\nThis effect, however, mostly relates to some \"lurking variable\", or some \"third effect\", which influences FVC apart from smoking.\n\nFor instance:\nhttps://www.atsjournals.org/doi/pdf/10.1164/rccm.201512-2492OC\n--> smoking doesn't associate with lower respiratory functionality for the people with childhood onset persistent asthma\n\nhttps://biomedpharmajournal.org/vol11no2/effects-of-cigarette-smoking-and-age-on-pulmonary-function-tests-in-%E2%89%A5-40-years-old-adults-in-jordan/\n--> number of years a person smoked is negatively correlated with the FVC, however, the more cigarets per day a person smokes, the higher FVC he/she has\n\nand this funny statistics on university students:\nhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC6709160/\n--> cite \"Smokers reported significantly more wheezing and sputum production than nonsmokers, but exhibited better FVC and FEV1 values\". Moreover, those who smoke more cigarets per day, have better FVC than those who smoke less.\n\nThis is a really unexpected result to me, especially the last study was done on young adults and not those, who went to the doctors with complaints, but just a sample of students. So here one can't associate the result with a theory that smokers more often do lung checks, thus their values are higher, as the non-smokers come only when they have bad complaints. \n\nI still didn't find any reasonable biological explanation for that, but one perhaps should count with the fact that the FVC-(Smoking Status) correlation is not that straightforward thing. Also, we can never exclude some \"hidden effects\" and biases. Like, the ex-smokers might have also exercised sport more often in the recent years or whatever else.", "votes": null } ], "comments": [ { "id": 969707, "author_name": "antonincadapeaud", "author_url": "", "post_date": "08/13/2020 21:47:04", "content": "I don't have the numbers next to me right now, but I think there might be one or several reasons for this to happen :
\n-the measure of the FVC might be biased. Smokers usually have more troubles breathing so they might be used to testing their FVC, unlike a non smoker. Tests are prone to a high variance, if you look at the curves.
\n-smokers have a lower life expectancy on average, which means that if they survive it implies that they potentially have a larger FVC than average (survivor bias)
\n-the sample might be built on that particular purpose
\n-it could be randomness but it seems very unlikely
\nThere are actually many potential reasons.
", "votes": null, "replies": [] }, { "id": 969897, "author_name": "ianzurutuza", "author_url": "", "post_date": "08/14/2020 03:36:00", "content": "Maybe these smokers have been compensating for their dead space by moving more air?
\nI saw the same phenomenon, but remember the same participants FVC is measured several times across ~60 weeks. Our sample size of individual current smokers is actually quite small. (2 women and 7 men)
", "votes": null, "replies": [] }, { "id": 974706, "author_name": "cv13j0", "author_url": "", "post_date": "08/18/2020 02:47:25", "content": "Hi! @franarenas, I identified the same behavior on FVC, I attribute this to the small sample of the data for Currently Smokes
", "votes": null, "replies": [] }, { "id": 1007777, "author_name": "gannadolinska", "author_url": "", "post_date": "09/12/2020 13:32:33", "content": "I was also quite surprised to see this.
\nI barely believed that it is a random effect, so I started to search for official statistics.
If one goes through several statistical studies on the influence of smoking on the respiratory functionality, one can see that they mostly show lower FVCs for smokers. However, there are some more or less recent studies which show different statistics - larger FVC for smokers.
\nThis effect, however, mostly relates to some \"lurking variable\", or some \"third effect\", which influences FVC apart from smoking.
For instance:
\nhttps://www.atsjournals.org/doi/pdf/10.1164/rccm.201512-2492OC
\n--> smoking doesn't associate with lower respiratory functionality for the people with childhood onset persistent asthma
https://biomedpharmajournal.org/vol11no2/effects-of-cigarette-smoking-and-age-on-pulmonary-function-tests-in-%E2%89%A5-40-years-old-adults-in-jordan/
\n--> number of years a person smoked is negatively correlated with the FVC, however, the more cigarets per day a person smokes, the higher FVC he/she has
and this funny statistics on university students:
\nhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC6709160/
\n--> cite \"Smokers reported significantly more wheezing and sputum production than nonsmokers, but exhibited better FVC and FEV1 values\". Moreover, those who smoke more cigarets per day, have better FVC than those who smoke less.
This is a really unexpected result to me, especially the last study was done on young adults and not those, who went to the doctors with complaints, but just a sample of students. So here one can't associate the result with a theory that smokers more often do lung checks, thus their values are higher, as the non-smokers come only when they have bad complaints.
\nI still didn't find any reasonable biological explanation for that, but one perhaps should count with the fact that the FVC-(Smoking Status) correlation is not that straightforward thing. Also, we can never exclude some \"hidden effects\" and biases. Like, the ex-smokers might have also exercised sport more often in the recent years or whatever else.
", "votes": null, "replies": [] }, { "id": 975026, "author_name": "deenihero", "author_url": "", "post_date": "08/18/2020 05:51:07", "content": "Best of luck
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "969603": "We can see that ex-smokers have higher FVC than patients who never smoked, it is just a coincidence because not having enough data to generalize? \n\n\n\nThanks for reading, this is my first competition and I still have to learn a lot! 😄", "969707": "I don't have the numbers next to me right now, but I think there might be one or several reasons for this to happen :\n\n-the measure of the FVC might be biased. Smokers usually have more troubles breathing so they might be used to testing their FVC, unlike a non smoker. Tests are prone to a high variance, if you look at the curves.\n\n-smokers have a lower life expectancy on average, which means that if they survive it implies that they potentially have a larger FVC than average (survivor bias)\n\n-the sample might be built on that particular purpose\n\n-it could be randomness but it seems very unlikely\n\nThere are actually many potential reasons.", "969897": "Maybe these smokers have been compensating for their [dead space](https://en.wikipedia.org/wiki/Dead_space_(physiology)) by moving more air? \n\nI saw the same phenomenon, but remember the same participants FVC is measured several times across ~60 weeks. Our sample size of individual current smokers is actually quite small. (2 women and 7 men)", "974706": "Hi! @franarenas, I identified the same behavior on FVC, I attribute this to the small sample of the data for Currently Smokes", "975026": "Best of luck", "1007777": "I was also quite surprised to see this. \nI barely believed that it is a random effect, so I started to search for official statistics.\n\nIf one goes through several statistical studies on the influence of smoking on the respiratory functionality, **one can see that they mostly show lower FVCs for smokers**. However, there are some more or less recent studies which show different statistics - larger FVC for smokers.\nThis effect, however, mostly relates to some \"lurking variable\", or some \"third effect\", which influences FVC apart from smoking.\n\nFor instance:\nhttps://www.atsjournals.org/doi/pdf/10.1164/rccm.201512-2492OC\n--> smoking doesn't associate with lower respiratory functionality for the people with childhood onset persistent asthma\n\nhttps://biomedpharmajournal.org/vol11no2/effects-of-cigarette-smoking-and-age-on-pulmonary-function-tests-in-%E2%89%A5-40-years-old-adults-in-jordan/\n--> number of years a person smoked is negatively correlated with the FVC, however, the more cigarets per day a person smokes, the higher FVC he/she has\n\nand this funny statistics on university students:\nhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC6709160/\n--> cite \"Smokers reported significantly more wheezing and sputum production than nonsmokers, but exhibited better FVC and FEV1 values\". Moreover, those who smoke more cigarets per day, have better FVC than those who smoke less.\n\nThis is a really unexpected result to me, especially the last study was done on young adults and not those, who went to the doctors with complaints, but just a sample of students. So here one can't associate the result with a theory that smokers more often do lung checks, thus their values are higher, as the non-smokers come only when they have bad complaints. \n\nI still didn't find any reasonable biological explanation for that, but one perhaps should count with the fact that the FVC-(Smoking Status) correlation is not that straightforward thing. Also, we can never exclude some \"hidden effects\" and biases. Like, the ex-smokers might have also exercised sport more often in the recent years or whatever else." }, "source": "meta" }