The item read: “Marijuana is also called pot, weed, and grass. Are you planning to stop using marijuana?” . Participants were categorized as marijuana users if they indicated recent use on the staging item. All others were categorized as non-marijuana users. The following outcomes were assessed at 3, 6, and 12 months: 1) seven-day point prevalence abstinence, 2) smoking reduction, 3) presence of a quit attempt since the last assessment, and 4) stage of change for quitting smoking. Self-reported point prevalence abstinence and reduction were assessed with the item, “How many cigarettes have you smoked in the past 7 days?”. To measure point prevalence abstinence, responses were coded into abstinent in the past seven days or smoking . Reduction was calculated using baseline cigarettes per day, and coded into reduced or not reduced by at least 50% since baseline. Quit attempts were measured with, “Have you tried to quit smoking for at least 24 hours since your last Tobacco Status Project survey?” . Stage of change was measured using the Stages of Change Questionnaire , recoded into precontemplation, contemplation, preparation, or action/maintenance. Those in action/maintenance indicated that they had quit smoking. All outcomes were measured at each time point. Participants in the intervention group reported their perceptions of the intervention at treatment end by rating their agreement with 7 items. Items addressed whether the intervention was easy to understand, gave sound advice, gave participants something to think about, and helped them to be healthier, as well as whether they used the information, thought about the information,indoor grow rack and would recommend the intervention . Responses were coded as disagreement or agreement .
Engagement was measured by the number of Facebook comments an individual posted during the 90-day intervention, including comments on posts and during live counseling sessions . First, marijuana users and non-users at baseline were compared on baseline demographic and smoking characteristics. Second, differences in reported smoking outcomes between users and non-users during the follow-up period were analyzed using a series of generalized estimated equations with binary distributions and logit link functions for dichotomous variables and a multinomial distribution with a logit link function for the ordinal variable . Longitudinal analyses controlled for intervention group and adjusted for baseline stage of change , baseline average cigarettes per day, sex, alcohol use, and age participants began smoking regularly. The first two covariates were determined a priori and the latter were selected based on the observed baseline differences between marijuana users and non-marijuana users. Because all participants were smokers at baseline, longitudinal analyses only included data from the three follow-up points . Largely due to attrition, there were 493 missing data points across all three time points on the abstinence variable, 498 on the reduction variable, 489 on the quit attempts variable, and 502 on the readiness to quit variable. GEE analyses are relatively robust to missingness, and a participant’s data could still be included in the analyses if they were missing one or more time points. Third, chi-square tests for independence were used to compare marijuana users’ and non-marijuana users’ perceptions of the intervention. An independent-samples t-test was used to compare treatment engagement between marijuana users and non-marijuana users in the treatment group. Baseline participant characteristics are displayed in Table 1. Marijuana users were more likely to be male, more likely to drink alcohol, smoked fewer cigarettes per day, and began smoking cigarettes regularly at an older age than non-users. Associations between smoking variables and marijuana use at each follow-up time point are displayed in Table 2. Use of marijuana by young adult smokers was associated with a lower likelihood of reduced smoking and a lower likelihood of abstaining from smoking in the past seven days, as assessed over 12 months of follow-up.
Marijuana users and non-marijuana users did not significantly differ in likelihood of having made a quit attempt or readiness to quit smoking . Moreover, users and non-users did not significantly differ in their perceptions of the intervention or treatment engagement . This study showed longitudinal patterns of marijuana use, point-prevalence abstinence from smoking, and reduction in smoking among young adults participating in a digital smoking cessation intervention trial. Most importantly, results showed that young adult smokers who coused marijuana were less likely to reduce their cigarette smoking or to have been abstinent from smoking than were those who did not use marijuana; however, they did not differ in readiness to quit smoking or likelihood of having made a quit attempt. Although smoking marijuana in addition to cigarettes increases young adults’ likelihood of negative physical effects , smoking marijuana may make quitting cigarettes more difficult in part by perpetuating the habit of smoking. Quitting smoking requires breaking associations or cues between the behavior of smoking and other contextual factors . Young adults commonly use marijuana in conjunction with cigarettes . Thus, continuing to use marijuana may hamper cigarette smokers’ efforts to change their behavior. Indeed, results showed that marijuana users were less likely to have recently abstained from smoking or reduced their smoking over a 12-month period. On the other hand, marijuana use status was consistently unrelated to readiness to quit smoking at baseline and during the followup period. Moreover, users and non-users did not significantly differ in the likelihood of making a quit attempt over 12 months. Results are consistent with research showing that young adult marijuana users do generally view quitting smoking as important , but have less ability to follow through with a complete abstinence goal despite a desire to quit smoking . Overall, our finding that marijuana users are less likely to report recent abstinence or reduction in smoking is consistent with extant literature suggesting that marijuana users are less likely to be successful at quitting smoking .
Encouragingly, marijuana users and non-marijuana users participating in the digital smoking cessation arm of the intervention did not differ in their perceptions of the intervention or their engagement in it. This suggests that young adults who use marijuana were receptive to the content and digital platform of the smoking cessation intervention. Future intervention content could highlight the negative effects continued marijuana use may have on quitting smoking, and could serve as a resource for young adults who want to quit using one or both substances. The variables most strongly and consistently associated with smoking outcomes over time were baseline stage of change for quitting smoking and marijuana use. Both should be assessed to inform treatment efforts with young adult smokers. Strengths of this study include multiple smoking-related outcomes, a 12-month longitudinal design, and a focus on young adults . This study had a few notable limitations. First, outcomes were self-reported. Our group has previously demonstrated the reliability and validity of young adults’ online self-reported tobacco and marijuana use , as well as the accuracy and limited bias of self-reported point prevalence abstinence in the present sample . Therefore, we opted to use self-reported abstinence,indoor farming equipment which had a much higher response rate. Second, current marijuana use was categorized into use versus non-use. It is possible that the relationship between marijuana use and smoking outcomes differs by heaviness of marijuana use, which our survey item did not assess. Although past research has shown that readiness to avoid marijuana use is significantly correlated with past 30 day marijuana use , future research should include a more detailed measure of marijuana use. The measure of alcohol use was similarly nonspecific, and a more detailed measure may yield different results. Moreover, up to twice as many of the participants indicated being abstinent for 7 days at each follow-up than identified as being in action/maintenance for having quit smoking. This was especially true of participants who were not using marijuana concurrently, as reflected in the significant difference in point prevalence abstinence and reduction between marijuana users and non-marijuana users. This finding may be due to the sample including non-daily smokers, and/or the young adult age of the participants. Based on self-report, 5-10% of the sample refrained from smoking for at least one week, yet were not committing to quitting. Future research could include more nuanced measure of marijuana use and measures of smoking specific to non-daily cigarette smokers. For the first time in several decades—and concomitant with the rise in opioid use, misuse, and dependence—life expectancy has declined in the United States, and life expectancy gains have stalled in Canada. Consistent with these global estimates, in an accompanying paper for the Special Issue, Astrid Guttmann and colleagues analyzed 2002–2016 national data from the United Kingdom and Canada to identify women who likely used opioids during pregnancy and demonstrated markedly elevated mortality rates over up to 10 years of follow-up. The elevated rates were particularly striking for mortality due to avoidable causes like unintentional and intentional injuries. Using 1998–2014 data from a large sample of primary care practices in the UK, John MacLeod and colleagues show that coprescription with benzodiazepines was highly prevalent among patients receiving opioid agonist and partial agonist treatment and that coprescription was strongly associated with drug-related poisonings. This study adds to the relatively thin evidence base about the potential hazards of benzodiazepine coprescription in the setting of opioid agonist treatment. Although opioid agonist treatment should not be with held from patients concurrently taking benzodiazepines or other central nervous system depressants, these studies suggest a need for vigilance by healthcare professionals providing care for such patients to minimize the risk of overdose or death.
Coprescription of alprazolam may warrant particularly heightened scrutiny, however, given that it is the short-acting benzodiazepine most frequently involved in drug overdose deaths.The elevated mortality risks facing people with opioid use disorders are attributable to a complex web of interrelated structural and psychological causes. The concept of the “risk environment” may be useful to reference here, given its focus on the interplay between various structural factors that increase vulnerability to morbidity and mortality. The study by Zehang Li and colleagues provides an example of the use of spatiotemporal data to characterize one aspect of the risk environment. Applying a Bayesian space–time model to emergency medical services dispatch data on suspected heroin-related overdose incidents from Cincinnati in 2015–2019, the investigators identified significant spatial heterogeneity in the distribution of these calls, with strong associations with features of the built environment and temporal spikes corresponding to local media reporting. Analyzing 2005–2016 claims data, Yu-Jung Wei and colleagues identified more than 200,000 adults with new claims related to opioid use disorder or overdose. They found that, by the end of the study period, nearly one-half had filled no opioid prescriptions in the 12 months prior to an incident opioid use disorder diagnosis or overdose. Among those who had filled opioid prescriptions, nearly three-quarters were prescribed a mean daily dose lower than the threshold needed to trigger most risk stratification algorithms. Also noteworthy is the analysis of 2015–2016 data from the US National Survey on Drug Use and Health by Joel Hudgins and colleagues. These authors found that approximately 1 in 20 adolescents and young adults reported either past-year opioid use disorder or past-year non-medical use of prescription opioids and that three-quarters of those reporting non-medical use of prescription opioids had obtained them from outside the healthcare system. These estimates are generally consistent with trends identified in similar, previously published analyses of NSDUH data. Thus, although opioid prescribing patterns undoubtedly played a significant role in how opioid use disorders came to be so highly prevalent and asymmetrically distributed in the US, a public health response that focuses solely on prescribing behavior is likely to be ineffective in reducing the number of fatal and nonfatal opioid overdoses.For people with existing opioid use disorders, opioid agonist treatment is known to reduce mortality. Monica Malta and colleagues add to this evidence base with a systematic review showing a wide range of health and prosocial benefits of opioid agonist treatment for people with opioid use disorders who are incarcerated or have recently been released. Opioid agonist treatment may have important collateral health effects as well. Analyzing data from a 3-country cohort of people who inject drugs, Charles Marks and colleagues found that people who inject drugs and who receive opioid agonist treatment are approximately half as likely to assist others in initiating injection drug use. They then developed a deterministic, dynamic transmission model of initiation into injection drug use, ongoing drug use, and cessation of drug use.