semaglutide assoc with tobacco cessation

 A recent data-mining study found that smokers taking semaglutide had decreased tobacco use (see smoking cessation semaglutide AIM2024 in dropbox, or doi:10.7326/M23-2718)

 

Details:

-- 222,942 new users of diabetes medicines, including 5967 prescribed semaglutide versus seven other diabetes medications, were evaluated regarding three tobacco use disorder (TUD)-related healthcare measures: medical encounters for the diagnosis of TUD, smoking cessation medication prescriptions, and smoking cessation counseling, from 2017 to 2023

    -- these patients were identified in the TriNetX database, a global, federated, health research network providing access to de-identified and aggregated electronic health records of approximately 113 million patients and 64 large healthcare organizations covering diverse geographic regions, age, race, ethnicity, income and insurance groups, and clinical settings

        -- the TriNetX platform has been used to assess several other relationships between semaglutide and other GLP-1 receptor agonists (GLP-1RAs) with suicidal ideation, cannabis use disorder, and alcohol use disorder

    -- the seven medications compared with semaglutide included insulin, metformin, DPP-4 inhibitors, SGLT-2 inhibitors, sulfonylureas, thiazolidinediones (TZDs), and other GLP-1 receptor agonists (exenatide, liraglutide, albiglutide, and lixisenatide; the latter 2 are not available in the US)

-- prior to propensity matching, the insulin and semaglutide groups differed significantly by age, sex, race, diagnosis of obesity, some mental or behavioral conditions, and some cardiovascular condition conditions. There were also significant differences in prior smoking cessation medication prescriptions and counseling

-- after propensity matching, there were 5954 individuals in each  group: mean age 59, 51% female, ethnicity of Hispanic 4%/non-Hispanic 78%, 69% white/12% Black/4% Asian, adverse socioeconomic determinants of health 6%, problems related to lifestyle 21%

-- obesity as determined by BMI: the single largest group was BMI of 40 to 45 in 21%, BMI 45 to 50 in 12%, and the rest were mostly 7 to 10% up to a BMI of 60

-- depression in 35%, mood disorders 40%, anxiety disorders 43%, behavioral disorders 10%, the rest of the psychiatric disorders were less than 5%

-- alcohol use disorder 6%, opioid use disorder 6%, cannabis use disorder 3%, cocaine use disorder 3%

-- hypertension 84%, hyperlipidemia 69%, ischemic heart disease 32%, other forms of heart disease 42%, cerebral infarction 5%, cerebrovascular diseases 13%, cancer 45%, chronic pain 36%

-- prior tobacco use counseling 4%, smoking and tobacco use cessation counseling visits 6%, smoking cessation education 0.2%, drugs used in nicotine dependence 17% (nicotine replacement therapy 13%, varenicline 8%, bupropion 15%, nortriptyline 2%)

 

-- Major outcome: the three TUD-related health measures were assessed within a 12 month follow-up, comparing semaglutide to the seven other diabetes therapies

 

Results:

-- the sample size for new users of diabetic medicines ranged from 4231 for TZDs to 213,225 for insulin

 

-- Association of semaglutide with medical encounters for TUD diagnosis:

    -- strongest association was when compared to insulin, 32% decrease with semaglutide, HR 0.68 (0.63-0.74)

    -- weakness association (but still statistically significant) was when compared to other GLP-1RAs, with 12% decrease, HR 0.88 (0.81-0.96)

    -- the separation between the curves for most of the comparisons was month prominent within the first 30 days, continued to diverge more modestly until approximately 180 days, and then plateaued

        -- the curve for insulin, however, displayed a dramatic reduction in TUD diagnoses associated with semaglutide within the 30 days right after treatment initiation

    -- among patients who had not diagnosis of obesity, semaglutide was associated with  a lower risk of medical encounters for TUD diagnosis, ranging from a 40% decrease with insulin (RR 0.60) and TZDs (0.78), though the decrease in other GLP-1RAs it was a nonsignificant 15% decrease

    -- among patients who had a diagnosis of obesity, all relationships between semaglutide and the other diabetes meds were statistically significant and ranged from a 26% decrease with insulins (RR 0.74 (0.67-0.80)) to a 13% with other GLP-1RAs (RR 0.87 (0.79-0.96))

 

-- Association of semaglutide with smoking cessation medication prescriptions:

    -- strongest association was when compared to insulin, 68% decrease with semaglutide, HR 0.32 (0.28-0.38)

    -- weakness association was when compared to other GLP-RAs, 38% decrease, HR 0.62 (0.52-0.74)

        -- the separation of most of these curves was evident within 30 days and continued thereafter, though plateaued at 180 days

        -- similar associations were found in patients with and without obesity diagnosis

        -- this graph displays the associations of semaglutide effect in patients overall and without obesity

 

 

-- Association of semaglutide with smoking cessation counseling:

    -- semaglutide was associated with a lower risk of smoking cessation counseling compared with other medications, with HRs ranging from 0.69 to 0.85, but was statistically significant only when compared with insulins, metformin, and DPP-4 inhibitors

    -- in patients with no diagnosis of obesity, semaglutide was associated with decreased risk of smoking cessation counseling when compared to metformin and sulfonylureas; for those with a diagnosis of obesity, there was no statistically significant risk found with semaglutide

  

Commentary:

 -- as we all know, tobacco use is the most prevalent preventable risk factor for premature deaths, with an analysis for 1990-2019 finding an estimated 7.7 million annual deaths globally, stressing the importance of a successful TUD intervention

    -- the current meds do help some people, with varenicline being the best single agent overall, but the long-term success rate at 2 years is only in the 25-30% range: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5837420/

-- there are many anecdotal reports of patients on GLP-1 receptor agonists (GLP-1RAs) finding a significantly decreased interest in cigarettes (decreased desire and “cigarettes don’t taste so good”, per many of my patients on an array of GLP-1RAs as well as tirzepatide)

-- there are animal studies showing benefit, as well as a small study of 84 patients on exenatide plus nicotine patches vs just the patches that found 46.3% vs 26.8% smoking abstinence rates, with a post-cessation body weight being 5.6 pounds lower: https://pubmed.ncbi.nlm.nih.gov/33831213/

 

-- this current massive data-mining study did find that 3 potential surrogates for smoking cessation were much better with semaglutide use, both in obese and non-obese patients, than 7 other groups of the typical diabetes medications. Obesity was determined by BMI (a recent blog did find that other assessments of adiposity are clinically superior to BMI: https://gmodestmedblogs.blogspot.com/2024/09/body-roundness-index-is-better.html )

-- the presumed mechanism is that these GLP-1RAs reduce nicotine (and other substance) rewarding effects, as documented in animals and humans smokers; in addition, it has been shown that GLP-1 in mice stimulates the medial habenular projections to the interpeduncular nucleus and thereby enhances the adverse effects of nicotine, and the GLP-1RAs activate these neurons leading to decreased nicotine intake (see diabetes GLP1 inc adverse tobacco effects NatNeuro2017 in dropbox, or doi:10.1038/nn.4540)

    -- the same research group as in the data-mining article reported a lower risk for both incidence and relapse for cannabis use disorder, accessing the same the TriNetX platform: https://pubmed.ncbi.nlm.nih.gov/38486046/

    -- and there seems to be benefit in decreasing alcohol consumption in individuals with alcohol use disorder: https://gmodestmedblogs.blogspot.com/2023/09/glp-1-receptor-agonists-for-alcohol.html

    -- a detailed review article found benefit of GLP-1RAs for cocaine use in rats but the only one human study, a small one, did not; similar results for opiates: see https://bpspubs.onlinelibrary.wiley.com/doi/full/10.1111/bph.15677 , or https://doi.org/10.1111/bph.15677. There are some ongoing studies on GLP-1RAs and opiate and cocaine use disorders, so hopefully the benefits of GLP-1RAs will be clarified

-- my anecdotal finding (though I do have lots and lots of patients on GLP-1RAs) is that all of them work well for adiposity, and that the difference between GLP-1RAs is their actual potency, where semaglutide is one of the most potent when prescribed at high doses. It seems likely that the effects of GLP-1RAs for tobacco and other substance uses will track with GLP-1RA potency as well, again with semaglutide being at the head of the pack

 

-- one important difference between GLP-1RAs and other approaches leading to smoking cessation is that there is associated weight loss with the semaglutide (and the exenatide in the prior cited study) vs the more typical weight gain with the usual smoking cessation approaches. I have found this to be a very positive inducer for many patients to try GLP-1RAs who fear weight gain with other meds for smoking cessation:

    -- and, of course, the benefits of GLP-1RAs include dramatically improved glucose control (and, by the way, glucose control is actually impaired by smoking: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5429867/ ) in those with diabetes, as well as renal and cardiac protection

    -- by the way, there is also a research letter in JAMA oncology finding a decreased risk of colorectal cancer in patients with diabetes with and without overweight/obesity, also using the TriNetX platform (see dm GLP1 dec risk of colorectal cancer JAMAonc2024 in dropbox, or doi:10.1001/jamaoncol.2023.5573

 

Limitations:

-- one obvious limitation is that there are no specific data in the TriNetX database on tobacco consumption, including the total amount of tobacco consumed, the length of tobacco consumption, changes in tobacco use over time, or secondary exposure to tobacco

    -- there is an ongoing small placebo-controlled study of semaglutide for smoking cessation, which might clarify some of the issues with this current study

-- another major issue is that we do not know the dose of the semaglutide used. For example, in terms of effectiveness for weight loss or diabetes control, different GLP-1RAs themselves have different potencies, and semaglutide is one of the most potent ones, though has a range of doses. So, comparing high-dose vs low-dose semaglutide, for example, would likely favor high-dose semaglutide since it is by much more potent (but also not as well tolerated because of adverse effects related to GLP-1RA potency).

-- there was no placebo group in this study, which might lead to an overstatement of the effectiveness of semaglutide

-- there was no comparison of semaglutide to the other traditional meds for TUD, so hard to put the TUD benefit of semaglutide in perspective

 

So, not a great study since there was no direct information on actual tobacco use associated with semaglutide, only the 3 quite indirect measures  used, but:

-- this study does fit in and supports the observation that GLP-1RA may benefit an array of substance use disorders

    -- it does seem that per the understood mechanisms of action, GLP-1RAs may well have an important place in many substance use disorders, and some studies are currently underway

-- given the huge adverse effects of smoking from a global public health perspective, it would be very useful to add GLP1-RAs to the list of therapies, and that insurance companies include them in their covered medications…

    -- though, that being said, drug companies have dramatically, and unconscionably, increased the cost of semaglutide way beyond the actual cost of production by a factor of up to 1000 (https://gmodestmedblogs.blogspot.com/2024/04/drug-co-shenanigans-dm-meds.html )

 

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs:

 

 go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

 

if you would like to see the articles in this blog, please email me. 

 

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list