Pain severity change was calculated by subtracting the baseline PEG from the last recorded PEG; thus, negative values indicate a decrease in pain severity. To examine the impact of pain severity on changes in marijuana use, we constructed general linear models with change in pain severity as the outcome variable and change in marijuana use as independent variables. Residual diagnostics were performed to determine if linear model assumptions were satisfied. Opioid initiation and discontinuation outcomes: For the relationship between marijuana use and opioid initiation and discontinuation, we only considered marijuana use at baseline. This allowed us to explore the assertion that marijuana use facilitates tapering of opioids. We constructed multi-variable binary logistic regression models with opioid initiation and discontinuation as the outcome variables and marijuana use at the index visit as an ordinal variable . Due to small sample size and data separation, each logistic regression was performed using Firth’s penalized maximum likelihood estimation to reduce bias in the parameter estimates. We considered covariates at baseline that are potentially associated with these exposures and outcomes: age, race, gender, other substance use and mood symptoms. Viral load and CD4+ T-cell count were collected for descriptive purposes. Virologic failure was defined as plasma HIV RNA >1000 copies/mL at any time during the study period, without a repeated test within 30 days that found ≤1000 copies/mL31. Analyses were adjusted for CNICS site. All analyses were performed in SAS 9.4 . Most participants were from UAB and UCSD , consistent with prior studies from this cohort. Demographic and clinical variables are summarized in Table 1. Just over half of participants were aged >50 years. Most participants were male and nearly half identified as non-Hispanic Black. The median CD4+ T-cell count at the index visit was 582 cells/mm3, and only 16% of participants had a detectable viral load. Clinically significant symptoms of depression and anxiety were common,drying cannabis occurring in more than one third of the sample. Marijuana use varied by site, with the highest current and past marijuana use reported by participants recruited from Washington .

Overall, most participants reported no marijuana use in the past 3 months; 8% reported daily, 5% weekly, 3% monthly, and 13% 1–2 times in the past 3 months. Of participants who reported less than daily use, 11% reported increased use during follow-up. Of participants reporting having used any marijuana, 10% reported decreased use during follow-up. Median pain severity at baseline was 6.3/10 , and median change in pain severity during the follow-up was 0 . The most common chronic pain locations were low back and hands/feet. During the year prior to the index visit, 47% of participants were prescribed LTOT; 8% were initiated during the study period, and 10% were discontinued during the study period. Table 2 summarizes the analyses of the relationship between change in marijuana use and chronic pain severity during the study period. Among PLWH with chronic pain, neither increases nor decreases in marijuana use were associated with changes in pain severity. As described in Table 3, marijuana use at the index visit was not associated with either lower odds of opioid initiation or higher odds of opioid discontinuation. Notably, marijuana use at the index visit was associated with increased opioid initiation in the unadjusted analysis which did not achieve statistical significance in either the unadjusted or adjusted analyses, .In this study, we investigated potential benefits of marijuana use in PLWH. We did not find evidence that, among patients with chronic pain, marijuana use was associated with improvements in pain or reductions in opioid prescribing. This study adds to the evidence base from which HIV providers can draw when discussing marijuana use with their patients. Over the past several years, there has been a proliferation of research on the association between marijuana and health outcomes in the general population and in PLWH. Some studies have produced concerning findings, while other studies are more equivocal. For example, with regard to PLWH, studies suggest that marijuana use may be associated with sub-optimal HIV primary care visit adherence and cognitive impairment, but not with antiretroviral adherence, virologic suppression, or mortality.

Non–HIV-related harms of marijuana use include impaired driving, hyperemesis syndrome, cognitive impairment, psychosis, and other mood symptoms. Our findings suggest that these harms are not counterbalanced by benefits in terms of pain or reductions in opioid prescribing. We note that there are other conditions, such as multiple sclerosis and post-traumatic stress disorder, that are listed as an indication for medical marijuana in many states but have a similarly limited evidence base. Some states have legalized recreational and medical marijuana as a result of voter petitions or legislative mandates and not as a result of scientific inquiry. Similarly, indications for medical marijuana use published by states may not have depended on evidence, but rather community and political input. Our data suggest that at least in PLWH, this is putting “the cart before the horse”. Additional observational studies in populations with other chronic conditions will be essential to establishing which groups, if any, are most likely to benefit from medical marijuana. Our study has limitations. First, it was conducted in a clinical cohort of PLWH who are in clinical care – that is, they attend clinic visits, and are mostly virologically suppressed. Our findings may not be generalizable to other populations of PLWH who are not as engaged in care. While our analysis was longitudinal, it was drawn from only one year of follow-up data. A longer longitudinal study would address this issue and allow for more robust investigations of causal inference. Also, due to lack of reliable opioid dose data, we were only able to investigate initiation and discontinuation of opioid prescribing, rather than an increase or decrease, which may be an outcome more sensitive to change over a one-year period. Despite CNICS being one of the largest prospective cohort studies of PLWH, our sample size of PLWH with chronic pain, with or without changes in marijuana use, was small. It is not possible to definitively determine whether we were under powered to detect associations between marijuana use and pain/opioid outcomes, or whether these associations do not exist. Also, people with the heaviest marijuana use and people who did not use marijuana were excluded from some analyses, and sensitivity analyses could not be performed due to their small numbers. CNICS asks about “non-medical” use of marijuana, and participants’ interpretation of this question may vary. Individuals may use illicit marijuana to treat pain and other symptoms, and/or may not have access to medical marijuana in their state.

CNICS does not specifically query medical marijuana use. Another limitation is the inability to assess “intent”—that is, if medical providers and patients intended to use marijuana with a specific goal of reducing opioid dependency. Future clinical trials likely will focus on this as an intervention. Finally, we acknowledge that marijuana use is diverse in terms of route of administration and dose . Therefore, we were only able to consider marijuana use status and frequency. In conclusion, we did not find evidence that marijuana use in PLWH is associated with improved pain outcomes, or changes in opioid prescribing. This suggests that caution is warranted when counseling PLWH about potential benefits of recreational or medical marijuana. Further studies, including prospective trials of medical marijuana and large observational studies, are needed to understand what impact, if any, marijuana use can have on pain in PLWH.In the past few decades,cannabis curing the social and political atmosphere around marijuana use has led to increased divergence of public opinions and understandings about the issue. Variability in how common, how accepted, how practical, and even how lawful the use of marijuana is among communities and social groups frequently makes this a contentious issue. However, it is the very prominence and divisiveness of this issue that make it a timely topic of investigation. Marijuana use is an issue that may call to mind several considerations, making the matter complex and difficult to decisively judge. It can involve concerns about legality, safety, personal rights, social-cultural acceptability, and perhaps even morality. Some or all of these considerations may be involved in evaluations and judgments about marijuana use. While reasoning about marijuana use may undoubtedly be complicated for adults in the population who recognize the numerous facets involved in the issue, evaluations and judgments may be more complex for adolescents in the midst of a particularly transformative period of social and cognitive development. Though children begin to learn about their social worlds early on through their interactions and exchanges with their environments, it is not until adolescence they are able to incorporate a greater number of more complex components of their world into their thinking . In this way, reasoning becomes more complex and responses to social dilemmas more nuanced. The reasoning process does not, however, always produce clear and definitive conclusions. Because adolescents are still in the nascent stages of forming their understanding of social matters, a certain degree of opaqueness in their reasoning and evaluations is typical and expected. This can particularly be the case with regard to more complicated matters like drugs . Given the various features of the issue of marijuana use, as well as inconsistency in the “facts” and legislations on marijuana, this issue can be especially unclear for teenagers. Arguably, however, the multitude of factors that can be involved when forming judgments about marijuana use make investigating adolescents’ conceptualizations of the issue particularly instructive and revelatory of processes of adolescent reasoning. In the present study, adolescents’ evaluations and judgments about marijuana use were examined. The principal aim of this investigation is to shed light upon adolescents’ judgments about marijuana use, and to assess which particular informational assumptions their evaluations are based upon. The study also aimed to uncover the various considerations that teens find to be most salient and applicable to the matter. The present investigation is guided by Social Domain Theory , a framework regarding how children and adolescents think about their social world.

For over thirty years, this theory has provided a framework for investigating how children and adolescents construct their understanding of the world and the social cognitive processes they engage in when forming judgments about social matters. The key features of this theory and the value of using this framework for the investigation at hand are reviewed in the following sections.According to social domain theory, children come to understand their social world through their interactions with others, and in the process, construct different domains of social knowledge, such as the moral and conventional domains. The moral domain refers to concepts of justice, welfare, and rights that are obligatory, universal, and unchangeable. In contrast, the conventional domain refers to prescribed and generally accepted social norms and rules that are contextually determined. Though these domains may overlap in some cases , research has indicated that individuals distinguish between moral and conventional matters from a young age . The social domain framework also includes a third domain, referred to as the personal domain . The personal domain encompasses issues that are primarily related to concepts of the self, such as personal preferences, choices, and behaviors that do not directly affect others. An important distinction has been made between reasoning about issues that are personal matters and judgments about prudential matters . According to Tisak and Turiel , the prudential domain is similar to the moral domain in that it involves the issue of harm done unto persons. However, judgments about prudential issues focus on how particular actions impact the self, and in that sense are non-social and therefore lack the key social-interactional characteristic of moral issues. Research has shown that children and adolescents make distinctions among moral, conventional, personal, and prudential issues from a very early age, and understand and make judgments about their social worlds according to the domains .The literature in the field has consistently revealed the following criteria to be characteristic of issues evaluated as moral: judgment that the act is wrong, the wrongfulness of the act is not based on the existence of rules and is not contingent on rules , the wrongfulness of the act is not based on authority commands, so the act would be judged as wrong even if an authority states the act is acceptable, and the wrongfulness of the act is not based on common practice, so the act would be wrong even if it was an accepted practice among a group.