Talk:Zolpidem

Boghog has pointed out that data from both observational and randomised clinical trial data found evidence of a link to cancer. I think the total number of patients studied is 1.8 million. I think we have to say something. Good counter points have been made that the research is not conclusive and otherwise has limitations and has not yet received more mainstream attention and publication in high impact journals. Surely the best solution is to summarise the evidence concisely but include the core limitations of the evidence/conclusions to avoid POV pushing or any misrepresentation. Thoughts?--Literaturegeek | T@1k? 14:00, 19 August 2018 (UTC)

Looking at all the refs, it would be UNDUE to include the risk of cancer as an actual adverse effect up in the medical section. It would be reasonable to discuss this in a research section, and to give the state of the art there, with all of its tentativeness. Jytdog (talk) 14:14, 19 August 2018 (UTC)

A research section that includes all the caveats definitely sounds like a good solution. Boghog (talk) 16:18, 19 August 2018 (UTC)

done, here. thoughts? Jytdog (talk) 18:33, 19 August 2018 (UTC)

Thanks for adding that. Some of the studies were controlled for confounding factors so I made a small change in the text to reflect that fact. Boghog (talk) 19:50, 19 August 2018 (UTC)

I think "some studies found, some studies didn't find" is wishy-washy and potentially misleading phrasing. 7 of the 8 studies in the meta-analysis found a correlation at the 95% level. (the remaining one looks like it just barely missed it, and it would be 8 for 8 at the 90% level). Why not just say that?

Same with with the "many of the studies failed to control for confounders like cigarette smoking and alcohol use." The meta-anlysis says "five studies adjusted tobacco smoking as a confounding factor, and four studies adjusted alcohol drinking as a confounding factor." Just say "3 of the 8 studies in the meta-analysis failed to adjust for tobacco smoking, which could bias the results either upward or downward depending on whether the hypnotic users smoked tobacco at a higher or lower rate than the control group."

I also think it is important to highlight the zolpidem-specific finding: "With regards to the type of hypnotics, zolpidem use showed the strongest risk of cancer..."

Finally, to the extent this isn't put in the adverse effect section (where I think it belongs, perhaps under a sub-header that identifies it as an area of ongoing research), there should be a internal link to it in that section. There is a lot of concern about the article not stating "Zolpidem causes cancer," but having an adverse effect section with no mention that large numbers of studies find an association between zolpidem and cancer, and a meta-analysis of them also finds a significant association, to me implies, incorrectly, this is not an area of significant concern many scientists are researching.

Do you all really think someone interested in the adverse effects of zolpidem would be more interested in the "diarrhea (1%)," in the second line of the section, but not be interested in a meta-analysis that links it with cancer? By all means, have a balanced presentation of the study, caveated as needed, but when I see in the adverse effects section "the most common side effects of long-term use included dry mouth (3%), allergy (4%), back pain (3%), flu-like symptoms (1%), chest pain (1%)" that says to me "The issue has been carefully studied and adverse effects quantified, and these are the only adverse effects out there." Declanscottp (talk) 03:56, 20 August 2018 (UTC)

Declan, does this edit address one of your main concerns?--Literaturegeek | T@1k? 13:57, 20 August 2018 (UTC)

It is an improvement, but doesn't really get to the central problem: there appear to be in excess of 100 published studies on the adverse effects of zolpidem, including many meta-analyses of them and at least one review article. People looking at "adverse effects" would have no way of knowing this. I think the article overall needs a lot of work, not just on the adverse effects section, and perhaps in a week or two after people have a chance to respond to my longer proposals, I can put in a more extensive revision that further changes can be based off of. Declanscottp (talk) 22:04, 20 August 2018 (UTC)

Declan, this drug actually causes diarrhea (to use the example you gave). We know this is true. Can you really not see the different between that, and a correlation with a bunch of confounders? The first is accepted knowledge; the second is a subject of research. Jytdog (talk) 20:07, 20 August 2018 (UTC)

I disagree the evidence that zolpidem causes diarrhea is stronger than it is for higher rates of falls in the elderly, higher all-cause mortality, higher rates of certain infections and cancers, and higher rate of overdose death. Declanscottp (talk) 22:57, 20 August 2018 (UTC)

Your stance is not supported by the breadth of high quality MEDRS refs. It is not OK to cherry pick refs, especially ones by self-acknowledged advocates, as you did here. That is not how we edit Wikipedia anywhere. Jytdog (talk) 23:13, 20 August 2018 (UTC)

You keep saying that, but then you don't respond to my detailed talk page comments. I am not cherry picking anything, I am citing the only meta-analyses in existence on these topics, as well a high quality review article from 2017. That seems to be as good as it gets on MEDRS.Declanscottp (talk) 23:32, 20 August 2018 (UTC)

If you mean the long section above where you cite a bunch of primary sources, those sources are irrelevant. We rely on MEDRS refs, and we listen to a bunch of them; not just the ones that we like. What MEDRS refs support the overall mortality cancer content? Jytdog (talk) 00:07, 21 August 2018 (UTC) (gah, un-distract Jytdog (talk) 01:45, 21 August 2018 (UTC))

In "that long section above" I argued in favor of including a secondary source, specifically "Zolpidem use and risk of fractures: a systematic review and meta-analysis Park, S.M., Ryu, J., Lee, D.R. et al. Osteoporos Int (2016) 27: 2935." I also described some primary sources to provide additional context and background to the secondary source. Declanscottp (talk) 00:38, 21 August 2018 (UTC)

This discussion is about what to say about cancer. It grew directly out of the section higher in the page, Talk:Zolpidem#Direct_quotation_from_a_meta-analysis_about_Zolpidem_and_cancer_reverted,_I_disagree. Everyone else in this section, is talking about cancer. You stayed on point in your first comment there, although you started to stray at the end; your next comment was completely off topic from cancer, and just above you were completely derailed. PMID 27105645 (the Ryu meta-analysis) says nothing about cancer or mortality; it says that it appears that zolpidem roughly doubles the risk of fracture; generally "139 cases of fracture occur for every 100,000 person-years not receiving zolpidem, and if we assume a 1.92-fold increased risk of fracture due to zolpidem, as determined in this study, an additional 127 cases of fracture can be expected for every 100,000 recipients of these drugs annually (the 1-year number needed to harm = 747)." To address this point. Sure we can add that. To put that in similar proportions to the percentages in the adverse effects section, the percentage of people taking the drug who can expect to have this sequella of impaired coordination is 3%. ((139+127)/100,000). Jytdog (talk) 01:25, 21 August 2018 (UTC)

If you want to talk about side effects generally, please open a new section on that, so we can focus on that. Jytdog (talk) 01:26, 21 August 2018 (UTC)

done Jytdog (talk) 01:29, 21 August 2018 (UTC)

Your math is wrong. 139+127/100,000 = .00266. x 100 = .266%. I've removed this statement from the article because I don't think it would have been included if the original calculation had been correct. Oro Temp (talk) 01:42, 6 April 2025 (UTC)