Whereas our initial models tested the relationship between interdependent substance use behavior, they assumed that these effects are symmetric: that is, usage of one substance equally increases or decreases usage of another substance. In our next set of models, we relax this assumption and test whether usage of one substance increases behavior of another substance or decreases behavior , or both . These models were estimated separately as the combined model exhibited extreme collinearity. As shown in Table 3, there is a significantly positive creation function from marijuana use to drinking in both samples, implying that respondents’ marijuana use increased their odds of drinking initiation. Thus, one unit higher marijuana use made a nondrinker 62% and 60% more likely to start drinking rather than stay as a non-drinker at the next time point in Sunshine High and Jefferson High, respectively. On the other hand, the endowment function from marijuana use to drinking is not statistically significant at either school, implying that marijuana use does not affect the likelihood of stopping drinking behavior. The impact of marijuana use on smoking behavior differs across the two schools. We detect a statistically significant creation function in Sunshine High: a one unit increase in marijuana use increases the odds 62% that adolescent non-smoker will initiate smoking rather than stay as a non-smoker. There was no evidence of a statistically significant endowment function in Sunshine High. On the other hand, the pattern is reversed in Jefferson High with a statistically significant endowment function but a statistically insignificant creation function. Thus, in Jefferson High although marijuana use does not impact respondent’s likelihood of smoking initiation, one unit higher marijuana use made smokers 27% more likely to stay as smokers rather than quit smoking at the next time point.To understand the magnitude of these effects , we engaged in a small simulation study in which we omitted some of the effects from the SAB model shown in Table 2 and assessed the consequences for the level of substance use behavior in the schools. That is,grow rack we changed a particular parameter value from the one estimated in the model to zero, and then simulated the networks and behaviors forward 1000 times.

We then assessed the average level of smoking, drinking, and marijuana use in the network at the end of the simulation runs. To save space, we only present the results for Sunshine High; see S2 File for the Jefferson High results, which were similar.The highest level of smoking is observed when we set to zero the influence effect of friends on smoking behavior, as the percentage of non-smokers drops from 72% in the original model to 63%, and the percentage of heavy-smokers increases from 11% to 18% . The pattern was similar in Jefferson High, with analogous values of 48% to 42%, and 31% to 35%. This corroborates the findings in previous simulation research that peer influence has a protective effect on smoking and drinking adoption. The lowest levels of smoking are observed in the hypothetical scenario in which marijuana use has no effect on one’s own smoking behavior, as the percentage of non-smokers rises from 72% to 81%, and the percentage of heavy-smokers decreases from 11% to 5%. The analogous values in Jefferson High were 48% to 54%, and 31% to 25%. Regarding drinking behavior, we see that the effect of one’s own marijuana use is particularly important as setting this effect to zero results in a decrease in drinking behavior . In the scenario of no effect of marijuana use on drinking behavior the percentage of nondrinkers rises from 50% to 59% and the percentage of heavy drinkers falls from 13% to 7%. The analogous values in Jefferson High were 35% to 42% and 16% to 10%. It is notable that setting the influence effect of friends’ drinking on one’s own drinking behavior to zero reduces drinking somewhat . In Jefferson High, the number of heavy drinkers rises from 16% to 20%. For marijuana usage, very pronounced strong effects are observed for friends’ influence . Setting this influence effect to zero results in a sharp decrease in non-marijuana users from 62% to 47%, and a parallel large increase in heavy users from 19% to 32%. In Jefferson High, the analogous values were 61% to 43% and 18% to 33%. In sum, when the effect from marijuana use to cigarette use is turned off, more non-smokers and fewer heavy-smokers are expected in both schools. When the peer influence effect with regard to each substance use is turned off, fewer non-users and more heavy-users of each substance are expected in both schools.

In the scenarios in which we set other parameters to zero, the simulation results indicated that the substance use distribution was not altered in either school.Overall, our findings indicate some evidence of sequential substance use, as adolescent marijuana use increased subsequent smoking and drinking behavior in our two school samples. Whereas some existing research has found evidence that marijuana use leads to use of these substances, an important contribution of our study was simultaneously taking into account the substance use behavior of adolescents’ peer networks and other social processes occurring in networks. We found that marijuana use resulted in more smoking and drinking in both samples. Our findings are partially consistent with Pearson et al. , who found that that marijuana users smoked cigarettes more over time. Our findings are suggestive that marijuana use increases both alcohol and cigarette use. In addition, we made a distinction between whether interdependent substance use going from marijuana to cigarettes and alcohol results in initiation, cessation, or both. We found that marijuana use resulted in drinking initiation in both samples, and smoking initiation in Sunshine High. In contrast, marijuana use decreased the likelihood of smoking cessation in Jefferson High. Previous literature suggests that alcohol use is not a prerequisite for the initiation of marijuana use and the effect of alcohol use on the onset of marijuana use has declined while that of marijuana use on the onset of alcohol use has increased since 1965 , and our findings are consistent with this prior literature. Moreover, we tested cross-substance influence effects, which assessed whether the substance use behavior of one’s friends on a particular substance affected an individual’s own use of the other two substances. We found no evidence that such effects exist in our samples. We did, however, find peer influence effects for each specific substance, which is consistent with multiple past studies. Note, however, that whereas one implication is that having more friends who use marijuana, for example, results in greater marijuana use behavior on the part of the individual, another implication is that having more friends who do not use marijuana results in less marijuana use behavior. This relative symmetry of influence effects is sometimes overlooked when interpreting influence results, and our simulation results confirmed that this influence effect is in fact more likely to have a negative effect on substance use behavior.

These results are similar to an earlier simulation study that found that increasing the amount of peer influence in two high schools diminished school level smoking and drinking behavior . These results are consistent with theoretical insights from the Dynamic Social Impact Theory, which would predict that youth in friendship networks would adopt the same substance use behaviors through peer influence pathways, likely through social proximity and consolidation of youths’ attitudes and behaviors in adolescent networks. This highlights that the presumption that influence effects will always increase behavior is not necessarily accurate. In fact, we might expect that the dominant norms in a context will drive the direction of influence effects: in a school with little substance use, the greater number of non-users will push adolescents towards non-use, whereas in a school with high levels of substance adolescents are more likely pushed towards greater use. Given the complexity of our agent-based network models,greenhouse grow tables we demonstrated the relative magnitude of the effects by combining a small-scale simulation with a strategy in which we constructed hypothetical models that set certain key effects to zero and simulated the networks and behaviors forward. A key finding was that in a simulated world in which one’s own marijuana use did not affect smoking or drinking behavior, there would be a notable decrease in overall levels of smoking and alcohol usage in these schools, even controlling for the complexity of these models. We also saw that marijuana use operates as a mechanism between friends’ marijuana use and one’s own smoking and drinking behavior, as adolescents’ use of marijuana is impacted by their friends’ marijuana use, and this then affects the adolescent’s level of cigarette and alcohol use. Furthermore, one of the strongest effects detected was the influence effect of friends’ marijuana usage, as this has a particularly strong relationship to adolescents’ own marijuana use. Our findings highlight the importance of understanding interdependence in the use of multiple substances in adolescence, particularly those which operate through peer influence effects within friendship networks. Another notable finding was that depressive symptoms increased smoking behavior in Jefferson High. This high school has a relatively high average level of substance use compared to Sunshine High. Perhaps in a social milieu with a high average level of drug use, adolescents reporting higher levels of depressive symptoms may be more likely to display higher levels of cigarette smoking as compared to those who report lower level of depressive symptoms, given that past studies link depression and adolescent smoking. There are some limitations to note in this study. First, the time lags between the two sets of waves are not equal . Although it is preferable to have equal time periods, we performed a post hoc time heterogeneity test to ensure that the co-evolution of substance use behaviors and friendship networks was not significantly different across the three waves, or two time periods. Second, our SAB model specification is data intensive and can only be estimated for the two large schools among the 16 saturated schools in Add Health which are feasible for this type of analysis.

This limits generalizability and does not allow assessing why the interdependent effect from marijuana use to smoking is different across the two schools. Third, we had indirect information about marijuana use at time one, for a large percentage of the sample. Using this indirect information allowed us to avoid discarding a large amount of information at t1, however with a relatively small amount of potentially misclassified cases. Fourth, while the data are relatively old, we are aware of no evidence that the mechanisms of in person friendship formation, as captured in these Add Health network data,have changed significantly since the mid-nineties. In the current study, friendship networks were constructed through name generator items instead of real-time communication technology such as cell phone use. While future studies are needed to leverage existing technology such as cell phone usage for collecting adolescent social network data, these in person network data are likely still meaningful. Moreover, research suggests that cell phones help reinforce and reproduce existing social roles and structures rather than alter them. That said, future studies are needed to collect nationally representative contemporary data from US adolescents and investigate how the findings herein would be different if such technology was considered.Our findings have important implications for future studies. First, our findings suggest both feasibility and merit in exploring concurrent or sequential substance use behaviors across multiple time periods. Interdependence in substance use should be studied within one single model framework with multiple simultaneous on-going processes to reduce the risk of over-estimation of each process due to the auto correlation among them. Second, further explication of the interdependent effects from marijuana use to smoking and drinking is a useful direction for future research. Third, given smoking rates among adolescent youth have decreased significantly since the mid-1990s, more recent data are required to test whether our findings from these two Add Health large schools can be replicated in future research. Our findings also have practical implications for health behavior change interventions targeting adolescent substance use. Moreover, other research indicates that social networks can be leveraged for health behavior change interventions and may even be superior to non-network based interventions . Peer network based interventions targeting adolescent substance use might address the possibility that marijuana use increases alcohol and cigarette use.

It is also important to highlight the complexity of poly substance use in the context of a cross-sectional, retrospective study. Despite this, lifetime MA use disorder was retained in the multiple regression model, while the other substances did not. Due to limited data on participants who met criteria for a current substance use disorder or other measurements of current substance use parameters, our finding cannot speak to other potential factors associated with poly substance use that may explain differences in sleep between MA+ and MA− groups. Future studies to formally investigate poly substance use in more detail is needed to futher confirm our findings. In addition, we did not find associations between age, sex, or sexual orientation on sleep quality, which is contrary to well established literature on these topics . We suspect that the presence of other clinical risk factors for poor sleep, including those identified in this study , may be masking the detection of these variables traditionally known to impact sleep quality. There also remains the possibility that other unmeasured factors such as homelessness and/or SES may account for the observed relationship that MA was related to sleep in PWH that should be explored further in future studies. Lastly, the PSQI questionnaire is based on self-report, which is subject to recall and reporting bias. While there is merit in characterizing perceived sleep quality in vulnerable populations, as even the perception of poor sleep can influence mood and physical health , subjective measurements are just one facet of sleep quality and the inclusion of objective measurements such as actigraphy would enhance understanding of sleep in PWH and substance using populations. Importantly,cannabis dry rack the global PSQI score demonstrates strong sensitivity and specificity in distinguishing good from poor sleepers among the general population . While the sensitivity in detecting an insomnia diagnosis in PWH remains high , the specificity drops considerably .

This suggests that the PSQI may not just be capturing sleep quality in PWH and raises the question as to whether items such as “trouble staying awake during the day” or “trouble keeping enthusiasm” are purely a function of poor sleep or a result of HIV-infection, prescribed medications, and/or associated psychosocial factors. Studies investigating the quality of the PSQI sub-components in capturing sleep quality within PWH using factor analyses may be a natural next step for future research.For people with substance use disorders, denial of untoward consequences from their actions is common and can affect commitment to treatment. In 2019, 96% of untreated individuals with a substance use disorder in the previous year denied needing treatment.Psychodynamic approaches toward addiction encourage accountability and minimizing denial; and 12-step programs, such as Alcoholics Anonymous, target denial by encouraging clients to acknowledge that they have lost control over addictive behavior, with a focus on accountability-centered goals. Among participants who had poly substance misuse and attended Alcoholics Anonymous or Narcotics Anonymous, the number of days in attendance was associated with decreased self-deception measured in a followup assessment.The transtheoretical model of behavior change likewise posits that changing addictive behavior relies on a transition from lack of recognition that a problem exists to increased awareness and motivation to change.The rostral anterior cingulate cortex , which participates in self-related processing, including self-awareness, has been implicated in personal relevance of drug-related stimuli, as is the ventromedial prefrontal cortex, which contributes to decision making.In an fMRI study, denial of methamphetamine-related problems was negatively related to resting-state connectivity between the rACC and precuneus. Among participants who met diagnostic criteria for Methamphetamine Dependence ,denial of methamphetamine-related problems correlated negatively with overall cognitive function and with rACC connectivity to frontal lobe regions, including the precentral gyri, left ventromedial prefrontal cortex, and left orbitofrontal cortex.These data implicate the rACC and its connections in a person’s ability to acknowledge problematic aspects of their substance use.

One of the most important clinical measurements, the diagnosis of a substance use disorder, involves clinical judgment, but self-reports are very important. Structured diagnostic interviews, such as the Structured Clinical Interview for DSM-IV or Mini-International Neuropsychiatric Interview , query self-reports of symptoms indicating craving, tolerance, withdrawal, and interference with activities of daily living. Although interview guidelines encourage the use of referral notes, records, and observations of friends and family,diagnosis often relies on interview with the client alone. In these interviews, denial of problems related to substance use is common and can alter diagnosis. This study sought to clarify how a diagnostic measure of Methamphetamine Dependence that relies on self-report is related to a participant’s denial of his or her addiction problem. Participants comprised a sample of 69 individuals who acknowledged enough symptoms on the SCID to meet criteria for the diagnosis of Methamphetamine Dependence. They also completed the University Rhode Island Change Assessment Scale , which assesses motivation for change by providing scores on 4 stages of change: Precontemplation, Contemplation, Action and Maintenance. The Precontemplation score measures the respondent’s denial that their drug problem warrants change,and is based on a transtheoretical model of addiction.In a prior study, the Precontemplation score was positively related to years of heavy methamphetamine use and arrests for drug offenses,supporting the notion that high scores reflect denial rather than the absence of problems. We hypothesized the Precontemplation score would correlate negatively with symptom severity, confounding the diagnosis.It is estimated that homelessness affects 3.5 million youth between the ages of 18 to 25 annually in the United States. Sexual and gender minority youth are over represented in homeless populations, with research indicating that between 30% and 40% of service-using homeless youth identify as SGM, within the context of approximately 6.4% of youth aged 18 to 29 identifying as SGM nationwide.

SGM youth include individuals who identify as lesbian, gay, bisexual, and transgender as well as gender queer, non-binary, agender, asexual, or another sexual or gender identity that is either or both non-heterosexual or non-cisgender . Previous studies examining pathways into homelessness among youth have repeatedly demonstrated that SGM youth are more likely to enter homelessness as a result of family members who are unaccepting of their gender identity and sexuality compared to heterosexual cisgender peers, demonstrating how SGM status is itself a risk factor for becoming homeless. In addition to disproportionately high representation among all unstably housed youth, SGM youth experiencing homelessness also face increased health risks compared to their heterosexual cisgender peers. With regards to mental health, lesbian, gay, bisexual, and transgender youth who are homeless are more likely to experience substance use and use a greater number of substances than heterosexual cisgender peers experiencing homelessness. Given these documented disparities, SGM homeless youth may be at higher risk for negative health outcomes related to substance use, such as HIV and viral hepatitis, which can further serve as a barrier to maintaining stable housing. Sexual and gender minority youth experiencing homelessness also report worse mental health outcomes, including increased suicidal ideation and more severe depressive symptoms, increased anxiety, and higher rates of post-traumatic stress disorder. One proposed model for conceptualizing these observed health disparities among SGM youth is the minority stress model, which acknowledges that SGM communities face an excessive burden of daily stigma and discrimination from living in a heterosexist, transphobic society, resulting in detrimental effects to their emotional, psychological, and physical health. These experiences of stigma and discrimination among SGM youth have been documented to occur in a wide variety of settings, including family rejection, homophobic bullying in community settings such as schools, and discrimination from clients and staff in emergency shelters. In addition to the violence and discrimination faced due to their sexual orientation or gender identity, SGM youth experiencing homelessness often hold multiple identities that place them into further marginalized groups, such as their racial and ethnic backgrounds. Youth of color, particularly Black youth, are more likely to experience homelessness compared to white peers. Furthermore, SGM youth of color face more difficult exits from homelessness compared to their white, heterosexual, cisgender peers.

Black LGBT youth experiencing homelessness are more likely to experience harassment from police and community members,trimming tray as well as increased sexualization and invisibility, which collectively make LGBT youth of color more vulnerable to various mental health disparities, such as increased substance use and prevalence of mood disorders. Policy agendas aimed at addressing the health disparities faced by SGM youth experiencing homelessness have emphasized the need to understand SGM youth as non-homogenous micro-communities with unique experiences, risk factors, and social environments. Previous studies of youth experiencing homelessness in San Francisco, the location of this study, have revealed high burdens of substance use and mental health conditions in line with national trends. In the San Francisco 2019 Homeless Unique Youth Count & Survey, one in five homeless individuals on a single night was under the age of 25. Of these unstably housed youth, nearly half identified as LGBTQ+. One in three surveyed youth reported ongoing drug or alcohol use, and 13% reported substance use as a cause of their homelessness. Mental health was another commonly reported cause of homelessness, with 30% of all San Francisco homeless youth indicating that their mental health was a contributing cause of homelessness. Symptoms of depression, PTSD, and anxiety among service-seeking San Francisco youth experiencing homelessness are correlated with increased prevalence of opioid and stimulant use, demonstrating the inter-connectedness of substance use and mental health outcomes. The high health burden of substance use and mood disorders is also tied to increased mortality among San Francisco youth experiencing homelessness, who experience mortality rates 10 times in excess of their stably housed, age-matched peers, with a majority of deaths resulting from substance use or suicide. While the disparities in the prevalence of substance use and negative mental health outcomes among SGM youth experiencing homelessness compared to heterosexual cisgender peers are well-described in the literature, a comparative understanding of downstream harms associated with use of specific substances between SGM and heterosexual cisgender youth experiencing homelessness remains poorly characterized. In this study, we employ a tool that quantifies the burden of negative impacts associated with use of a specific substance. Using a cross-sectional analysis of a racially diverse group of service-seeking youth experiencing homelessness ages 18 to 24 in a dense, urban environment, we examine risks of harmful use associated with specific substances among SGM youth experiencing homelessness compared to their heterosexual cisgender peers. Alongside substance use, we examined whether symptoms of depression, anxiety, and PTSD differed between SGM and heterosexual cisgender youth who sought services at a community-based site. We expected that SGM youth would exhibit greater health risks associated with substance use across all substances surveyed, as well as more severe symptoms of mental illness when compared with their heterosexual cisgender peers. All data were collected from a capacity-building initiative at a partnering multi-site, non-profit community based organization in San Francisco, California.Our study was designed as a cross-sectional investigation of youth aged 18 to 24 who utilized services from Larkin Street Youth Services, a community-based organization in San Francisco. Each year, Larkin Street serves 2,500 to 3,000 youth aged 12 to 24 years old. Clients include individuals who live outside or in a car, a shelter, a transitional living program, permanent supportive housing, a single-room occupancy hotel, a unit partially paid for using subsidies, or who are otherwise unstably housed. The organization also offers a wide array of programming, including emergency and transitional housing, basic needs services such as access to food, showers, laundry, and harm reduction supplies, educational and employment training programs, medical care, behavioral health services, case management, street outreach, and a youth leadership development program. Larkin Street also offers resources and programming specifically for SGM youth, and staff members undergo LGBTQ cultural competency trainings. In order for participants to qualify for our survey, they needed to be between 18 and 24 years of age and utilize services at Larkin Street. Recruitment strategies included posting flyers within the CBO’s residential and clinical spaces, referrals of participants from frontline workers, case managers, counselors, and group facilitators, and presentations at community housing meetings. Of note, a small subset of participants were surveyed at a service site that only serves youth living with a HIV diagnosis, but the majority of participants were recruited from Larkin Street’s other sites that serve all youth, including daytime drop-in centers, health clinics, and transitional housing spaces.

We examined correlations between number of observations provided by each participant and all other study variables to assess whether attrition across waves was associated with levels of any of our study variables of interest. While socioeconomic status was correlated with greater retention in the sample , no other study variables were associated with attrition across waves. Data were structured treating age as time, and were analyzed in the 5-year span from age 17 to age 21.1 Moreover, noting positive skew in our binge drinking outcome, we further employed robust estimator that provided more accurate and robust parameter estimation for models of non-normal data compared to the standard maximum likelihood estimation and alternatives . Latent curve models with structured residuals were used to address primary hypotheses and were conducted separately for depression and anxiety. A schematic of the LCM-SRs employed are provided in Figure 1. LCM-SRs are an integration of random effects and cross-lagged panel models aimed at estimating lagged effects in longitudinal panel data at the within-person level. In contrast to standard cross-lagged methodologies, LCM-SRs can estimate temporally lagged associations between depression and binge drinking that are partitioned of between-person differences in these associations. This is achieved by specifying a latent curve model to account for between-person differences in level and change in longitudinal data while estimating auto-regressive and cross-lagged effects of time-dependent residuals, which have a purely within-person interpretation. In these models, we evaluated model fit using the adjusted χ2 difference test, where a non-significant result was an indication of adequate model fit. We supplemented this test with a number of alternative fit indices , including scaled variants of the comparative fit index and root mean square error of approximation . We included socioeconomic status and sex at baseline as covariates to control for between-person demographic factors that may be influencing levels of internalizing symptoms and binge drinking.

Socioeconomic status was measured using a modified version of the MacArthur Sociodemographic Questionnaire . In sensitivity analyses,cannabis vertical farming we replicated each LCM-SR separately for males and females to explore whether effects differed across sexes.2 All analyses were conducted in R using ‘lavaan’ . A significance threshold of 0.05 or better was required across analyses. Analyses were supplemented with post-hoc power simulations conducted using the pwrSEM application to determine whether LCM-SRs employed were sufficiently powered to detect targeted autoregressive and cross-lagged effects. Power was computed using the Monte Carlo method following steps provided in Wang and Rhemtulla, 2021 and Hancock and French, 2013. Namely, we specified a population LCM-SR based on the standardized variances and covariances between variables included within our model. Assuming this covariance structure, we modified effect sizes for target parameters within these models across a range of magnitudes to assess the minimum effect size needed to detect each target parameter. Then, we generated 1,000 samples assuming 831 observations for this model, fit our LCM-SR to each, and recorded the proportions in which the target parameters were different from 0 based on a significance threshold of α = 0.05. Results are summarized in Table 2. Between-person effects indicated socioeconomic status was associated with greater growth in HED over time, and that women reported higher depression and anxiety and more growth in HED than men. Auto regressive path parameters indicated that HED was positively associated with levels at the subsequent time point, indicating moderate year-to-year stability over the five-year study window. Addressing within-person effects, binge drinking marginally predicted depression among females . No other effects were significant. We note that confidence intervals for within-person effects were narrow for the associations between depression and anxiety symptoms and subsequent HED , demonstrating high precision in these null associations in the sample. We examined lagged associations between adolescent binge drinking, anxiety, and depression symptoms over five years in a large, diverse, prospectively followed sample of participants from the NCANDA study. While the self-medication hypothesis suggested that binge drinking behaviors may follow the presence of internalizing symptoms as a means of coping with negative affect, other theories have suggested binge drinking may disrupt social and cognitive functioning and may serve as a metabolic insult that predisposes youth towards the occurrence of depression and anxiety.

One underlying mechanism of the impact of binge drinking on the development of depression and anxiety has been explained in the literature as related to systemic and CNS inflammation that develops with increasing alcohol use and leads to subsequent psychiatric symptoms . Alternatively, shared between-person factors, such as environmental and peer influences, may drive relations between these co-developing constructs. We used latent curve models with structured residuals to test these hypotheses concurrently. Given equivocal evidence that internalizing symptoms and binge behaviors are linked in adolescence, we supported our findings with post-hoc power analyses to determine whether the NCANDA sample was sufficiently powered to detect these effects if they were present in our target population. Results demonstrated trending evidence that binge drinking predicted subsequent depression and anxiety among females, and no direct evidence that internalizing symptoms predicted later binge drinking. This suggests that although binge behaviors may increase later internalizing risk for females, it is likely that binge behaviors may develop largely independently from internalizing factors across this developmental period . Since we found a small and unidirectional relation between binge drinking and adolescent depressive or anxiety symptoms, results provided slight evidence favoring substance induced anxiety and depression models of internalizing risk . As such, results highlight that efforts aimed at preventing early initiation and hazardous or harmful alcohol use may have significant downstream effects on incidence of internalizing symptoms for females within the adolescent period. By contrast, we found little direct evidence of a self-medication model when analyzed at the within-person level. Though links between internalizing symptoms and binge drinking have been observed for adults , present findings may therefore indicate that early-onset internalizing symptoms represent a different developmental pathway unique from adult-onset symptoms that confer elevated substance use risk , or relatedly, that stronger associations have been found with more severe levels or extended histories of substance use and internalizing disorder .

Alternatively, consistent with the common-factor model, the effect sizes observed in this study may also imply minimal direct relation between these two conditions, such that associations observed between internalizing symptoms and binge behaviors may be better explained by common factors giving rise to each. For instance, a number of groups including Goodwin et al. found that after controlling for confounding factors , anxiety disorder was unrelated to all measures of substance use , and support by similar findings for confounding factors have been seen in a number of family and twin studies . In addition, several groups have proposed that genetic contributors and anxiety sensitivity render individuals susceptible to co-occurring anxiety and alcohol misuse ; similarly, common genetic factors have been identified as predisposing towards depression and alcohol dependence . While prior studies controlled for these between-person environmental and genetic factors directly, we analogously used a within-person design to partition factors that may be confounding associations. Employing this statistical approach provided evidence largely consistent with these prior findings, adding that we find only modest and unidirectional associations when sequenced over time. While the NCANDA study has major strengths of being a large, diverse, and prospectively followed sample, we note several limitations. First, episodic variation and measurement may be necessary to more accurately capture self-medication processes, such as in the application of weekly ecological momentary assessment methods ; the CDDR monthly binge drinking and Achenbach depressive symptom metrics from NCANDA may not have had the temporal sensitivity to capture these nuanced levels of association. Relatedly, some evidence suggests that while self-reported recall methods are reasonably accurate for adolescents and young adults at moderate levels , younger populations may under-estimate their alcohol consumption at higher levels of consumption . Thus, the present findings may reflect an underestimation of true binge drinking in the sample and may have tempered effect sizes observed in the present study. Second, it is also possible that other forms of substance use might have stronger links with internalizing pathology across this developmental period. Though our analyses focused primarily on testing theories of alcohol misuse in adolescence and young adulthood,drying cannabis examination of these effects across a wider range of substance outcomes may be a crucial extension of this research. Third, we note that several constraints were imposed on the estimated LCM-SRs in this study, including fixing cross-lagged and autoregressive parameters to equivalence across time.

As such, we encourage results to be replicated in future analyses, ideally in larger national samples of adolescent substance use such as the Adolescent Brain Cognitive Development Study . Fourth, the Youth and Adult Self-Report scales offer several advantages and limitations. Self-report intrinsically allows data to be gathered more easily since participants can describe their own symptoms, thus avoiding the need to meet or speak with a staff member. However, such reported symptoms can be less objective than reported symptoms from a trained, clinician staff member. Clinical scales such as the ASR/YSR capture a range of anxiety and depression symptoms that typically necessitate mental health intervention. This is both a strength and weakness in that we can capture lower level of symptoms, but are also gathering information on participants experiencing lesser symptoms than those having DSM-5 diagnostic criteria for a disorder such as major depressive disorder or generalized anxiety disorder. The ASR/YSR anxiety questionnaire as described in detail in the methods highlights that several different types of anxiety are probed in a mixed fashion, limiting generalizability and specificity to differentiate between types of anxiety disorders such as panic, specific phobias, separation, or generalized anxiety disorders using the results from this NCANDA study. Ferdinand et al. found that YSR anxiety scores predicted DSM-IV disorders only moderately while YSR depressive scores corresponded closely to DSM-IV major depressive disorder and dysthymia . Finally, very limited data was available for the NCANDA study for the context of each episode of binge drinking ; given social drinking factors may play a large role in predicting binge drinking engagement , drinking context could have an important impact on understanding our findings for the depression-binge drinking relationship. Our work provides an investigation of the interrelation between depression or anxiety and binge drinking in the large, diverse, and prospectively followed NCANDA sample. Our work is in line with other studies that have found minimal associations between depression or anxiety and binge drinking , and suggest that binge drinking may modestly predict later internalizing symptoms or that common factors may better explain links between these facets throughout adolescence. This may highlight the importance of simultaneous treatment of binge drinking and co-morbid depressive or anxiety disorders for both males and females. Future larger studies such as ABCD will be able to build on the early findings identified here. Mid-adolescence is a vulnerable developmental period for cigarette smoking uptake, the onset of mental health conditions, and the emergence of comorbid tobacco use and mental health problems . The over-representation of smoking among adolescents with mental health problems generalizes across various conditions , remains robust after controlling for confounders, and is mediated by theoretically-relevant factors suggesting a causal relation . The rapid emergence and appeal of novel tobacco and nicotine products such as electronic cigarettes raises the question as to whether the same adolescent subgroup with mental health problems is at risk for using these products . This is important to address because this population may be particularly vulnerable to nicotine addiction, given that neural plasticity during adolescence and neuropathology in psychiatric conditions can enhance the brain’s sensitivity to nicotine . E-cigarettes—electronic devices that deliver inhaled nicotine emulate the sensorimotor properties of conventional cigarettes—are gaining popularity among adolescents. According to 2014 estimates, past 30 day use of e-cigarettes is more common than conventional cigarettes among U.S. 8th- and 10th- graders, and many adolescent e-cigarette users have never tried conventional cigarettes . E-cigarettes may be an attractive alternative to conventional cigarettes among youth because of beliefs that they are less harmful, addictive, malodorous, and costly than conventional cigarettes . Furthermore, e-cigarettes come in flavors appealing to youth and may be easier to obtain than conventional cigarettes because of inconsistent enforcement of restrictions against sales to minors . Such factors may facilitate e-cigarette initiation in adolescents who would not otherwise smoke conventional cigarettes and may perhaps have fewer risk factors for smoking —including mental health problems.

Classification methods are often utilized to build toward categorical variables, however methods like neural networks are also designed for predicting continuous variables . Regression models are often used for the prediction of continuous measures or in the case of canonical approaches this can be with multiple dependent variables predicted simultaneously . Finally, in the case where there is no existent or optimal category or variable that the biomarkers seek to predict unsupervised approaches can be useful. With all these approaches variables can either be approached as linear or non-linear, although transformations and feature reduction approaches can mitigate these differences. It is important, regardless of approach, to understand the biological mechanisms being modeled by defining a model that best reflects the underlying systems to optimize prediction. Two key methods for statistical reduction of variables are selecting top ranking variables and creation of composite variables by factor or component-based analysis. Random Forest, as depicted in Figure 2, can be utilized to determine importance scores by evaluating the hierarchical functionality of a given variable as a bifurcator for optimizing classification . Random Forest is not alone in its utility to provide variable importance ranking but provides a nice mechanism for this analysis. The statistical creation of composite variables can be done through principal component analysis such that novel values are calculated for a set of variables that account for large swaths of variance with a single value vector . This can substantially increase the efficiency of a model and serve to highlight a robust latent feature. A summary of pain biomarkers discussed in this review article are provided in Table 1. Non-imaging pain biomarkers include opioid pain biomarkers: Beta-endorphin, B-cell opioid receptors, composite genetic, Mu-opioid receptor A118G polymorphisms, migraine opioid PET, and endogenous opioid function. Inflammatory pain biomarkers include cytokines, sICAM-1, cytokines related to back pain,cannabis grow racks cytokines related to peripheral neuropathy, substance P, and neuropeptides. Endocannabinoid pain biomarkers include: AEA in CRPS, 2-AG in optic neuromyelitis, AEA and 2-AG in headaches, ECB elements in multiple non-neuropathic pain conditions, ECB elements in endogenous opioid function, and ECB elements in gut-brain interactions.

There are pain biomarker genes related to neuropathic pain risk. MICRO-RNA dysregulation pain biomarkers are found in neuropathic pain, peripheral neuropathic pain, CRPS, migraine, and non-neuropathic pain conditions. Stress related pain biomarkers include allostatic load, Cortisol, DHEA, and allopregnanolone. Measuring saliva contains potentially particularly accessible pain biomarkers. Other pain biomarkers can be accessed via QST, skin conductance, pupil dilation, fatty acid pain biomarkers , neurotrophic factors, and serum neurotransmitters. Brain imaging pain biomarkers for measuring pain can be evaluated using three different MRI brain methods: gray matter structural imaging, white matter diffusion tensor imaging, and functional brain activation. Brain circuits related to pain mechanisms include an ascending brain circuit, a descending pain modulation circuit, the default mode circuit, the executive network brain circuit, and finally the salience network. Pain mechanisms in the brain can be measured via modulation in brain circuits: acute pain machine learning measures of chronic pain, pain rumination, pain mind wandering, placebo mechanisms, pain traits and states, and resilience. HIV peripheral neuropathy changes in the brain include reduced total cortical gray matter and reduced posterior cingulate cortex volume in particular, white matter degeneration, altered resting state networks, and aberrant expectation of pain relief. By focusing on a broad array of mechanisms and biomarkers, we can uncover important mechanistic connections and interactions across systems. Assessment and understanding in an appropriately comprehensive approach are challenging due to the vast and diverse literature and the complexity measurement. This review aims to facilitate navigation of this literature and the appropriate selection of biomarkers for future research. Health professional shortage areas are communities identified by the U.S. Human Resources and Services Administration in which there is a shortage of primary care health professionals.These shortages are accompanied by an absence of a consistent source of care, difficulty accessing care when needed, and a lack of outpatient preventative care, leading to increased hospitalizations.Multiple interventions have been attempted to increase access to care in HPSAs, including increased use of nonphysician providers. During the opioid epidemic, increasing access to naloxone furnishing has been viewed as critical in rural areas where opioid misuse is disproportionately high, including California’s Central Valley.

In the United States, pharmacists at community pharmacies are one of the most accessible points of care, with 90% of Americans living within 5 miles of a pharmacy.People seeking care have expressed interest in services at pharmacies not only because of ease of accessibility but also the availability of multilingual staff and extended hours that make it possible to access care on evenings and weekends.Previous studies have also shown that pharmacy-based care can extend services for patients in medically under served rural areas to reduce inappropriate prescribing, improve disease management, and enhance medication adherence and knowledge.In 2013, the California legislature passed SB 493, known as the Pharmacy Practice Bill, which expanded the role of pharmacists by giving them authority to furnish naloxone, hormonal contraception, nicotine replacement therapy, and travel medications, specifically prescription drugs and immunizations that are recommended by the Centers for Disease Control and Prevention to prevent or treat disease when travelling outside of the United States.California uses the term “furnish” to describe pharmacist-initiated prescription of medications.Expansion of pharmacist furnishing capabilities provides access to those in need, including people who use opioids.Past studies have sought to determine rates of pharmacist furnishing given its potential impact on access to care. However, these studies have focused on urban areas;previous studies of pharmacist furnishing of naloxone in California sampled primarily urban pharmacies ; previous studies on naloxone, hormonal contraception, and postexposure prophylaxis/preexposure prophylaxis furnishing were conducted in the San Francisco Bay Area only.As of the date of this study, there has been no prior research assessing furnishing rates in California’s Central Valley, a largely rural area, with a shortage of primary care physicians.However, understanding furnishing in these communities and those like it, particularly for naloxone, is critical given the disproportionate impact of the opioid epidemic in rural communities. For example, the age-adjusted rate of opioid-related overdose deaths in Fresno, one of the Central Valley’s largest counties, increased by 46%, from 48.6 per 100,000 residents in 2019 to 71 per 100,000 residents in 2020.This study sought to address this existing gap in research by assessing the extent of pharmacist furnishing, with a focus on naloxone, in the Central Valley. Research focused on the Central Valley due to the high potential impact of furnishing to increase access to care. It first assessed the extent of naloxone furnishing through a phone survey, then identified barriers and facilitators to implementation through interviews with a subset of furnishing pharmacists identified in the phone survey. We expected that rates of naloxone furnishing would be lower in disproportionately rural Central Valley pharmacies than in urban pharmacies evaluated in previous research, given the effects of high out-of-pocket costs in an area where people have lower incomes and social stigma surrounding opioid use disorders in more politically conservative communities.

The first step of data collection was a telephone survey of all pharmacies with the potential to furnish naloxone in the Central Valley,cannabis drying racks to identify overall furnishing rates. Four authors who were PharmD students, in collaboration with undergraduate researchers at the University of California Merced Nicotine & Cannabis Policy Center , first contacted all pharmacies that met inclusion criteria using the telephone number listed in the Board of Pharmacy license database. Using an existing screening question from previously published research on naloxone furnishing, upon initial contact an interviewer posed the question, “I heard that you can get naloxone from a pharmacy without a prescription from your doctor. Can I do that at your pharmacy?” Contact with each pharmacy was attempted up to 3 times. To identify potential interview contacts in the second step of data collection, interviews of furnishing pharmacists, each person at a pharmacy who that indicated it furnished naloxone was asked whether a furnishing pharmacist at the store would be interested in being interviewed for the study. If a pharmacist expressed interest, they received a cover letter, consent forms to sign by email or fax, and a list of interview questions. Researchers scheduled a time to interview after receiving this written consent. Pharmacies that did not furnish naloxone were not asked for interviews on the grounds that they would be unable to identify facilitators to furnishing naloxone at their store.Participants were interviewed in a semi-structured manner using an interview instrument used in previously published research to study furnishing of other medications and modified to address naloxone .This instrument included a list of questions, however each interview was conducted in a semi-structured format that allowed for a natural flow of discussion and gave participants opportunities provide additional information that may not have been specifically addressed in the prepared questions.Topics included the following: characteristics of the pharmacy and staff ; description of the furnishing process; perceptions regarding the effectiveness, advantages, disadvantages, facilitators, and barriers to furnishing; whether respondents also furnished other medications; and recommendations for reproducibility or improvement. Participants were interviewed via video or audio call except in one case, where responses were collected by e-mail. With permission, calls were recorded and transcribed. The interviewers took additional notes during and after the interview.The analysis began with calculation of descriptive statistics, including the percentage of pharmacies that furnished naloxone identified in the telephone survey. For interviews with the subset of naloxone furnishing pharmacists, descriptive analysis summarized the extent of furnishing for medications other than naloxone. Transcripts of each interview conducted, as well e-mail responses, were uploaded to Atlas.ti software for qualitative data analysis and deidentified by numbering each interview. Beginning with codes developed from past research on furnishing practices as a preliminary guide, as well as inductive methods to identify potential novel concepts, the investigators developed a code book classifying statements as referring to barriers or facilitators, then further subdivided them by type in Atlas.ti. Complete sentences were the minimum unit of analysis coded in the transcripts to identify common themes.To ensure validity and consistency across interviews and coding, each interview was conducted by a minimum of 2 researchers, and coding was completed simultaneously by all of the researchers who had conducted interviews. Disagreements were resolved by discussion until the group reached consensus. Transcripts, findings, and key quotations used to illustrate them were summarized in drafts circulated to the entire research team. Findings were triangulated based on reviews of previous studies of furnishing. Only findings identified as relevant by the group were included in the final analysis.The second step of data collection was interviewing furnishing pharmacists in the region for interviews about barriers and facilitators to furnishing. Among the contacted pharmacies that furnished naloxone, 8 furnishing pharmacists agreed to be interviewed. The stores where these pharmacists worked represented 5 of the 11 counties in the Central Valley . Of these, 5 were associated with a chain pharmacy, while the remaining 3 were independent. Although previous research on furnishing rates is limited, it suggests that naloxone furnishing is more common than furnishing of other medications. Interview participants were asked whether they also furnished other medications; as some of the factors that discourage or encourage furnishing may be consistent across medications. Six respondents indicated that the stores where they worked also furnished hormonal contraception, 3 respondents that their stores also furnished nicotine replacement therapy , and 1 that their store also furnished preexposure prophylaxis/post exposure prophylaxis. Respondents indicated that the pharmacies where they worked filled between 250 and 1000 prescriptions per day, averaging approximately 500. The time that respondents had held their positions ranged from 5 months to 20 years, and none had completed a residency. Results are provided in Table 2. With respect to barriers to furnishing, all interview participants listed cost to patients as the primary barrier. They noted that insurance did not necessarily cover naloxone, and when it did not, patients would not purchase it. As one stated, “The biggest barrier to this is first of all money. If it’s zero copay, they probably will take it. If there’s any copay, they’re just normally not going to pay for it.” . Other barriers to furnishing included time, cost, stigma, and lack of a shared language.

A microcomputer controlled the delivery of fluids, presentation of auditory and visual stimuli, and recording of the behavioral data. Rats were trained to self-administer 10% ethanol , 0.2% saccharin or water in 30 min daily sessions on a fixed-ratio 1 schedule of reinforcement, where each response resulted in delivery of 0.1 mL of fluid as previously described Briefly, for the first 3 days of training, water availability in the home cage was restricted to 2 h ⁄ day in order to facilitate acquisition of operant responding for a liquid reinforcer. During this time, rats were permitted to lever-press for a 0.2% saccharin solution. At this point, water was made freely available and saccharin self-administration training continued for another 3 days. The rats were then trained to self-administer ethanol by using a modification of the sucrose-fading procedure that used saccharin instead of sucrose . During the first 6 days of training rats were allowed to lever-press for a 5.0% ethanol solution containing 0.2% saccharin . Starting on day 7, the concentration of ethanol was gradually increased from 5.0 to 8.0% and finally to 10.0% , while the concentration of saccharin was correspondingly decreased to 0%. At the beginning of the saccharin-fading procedure a second but inactive lever was introduced. Responses at this lever were recorded during all training and testing phases as a measure of non-specific behavioral activation but they had no programmed consequences.At completion of the fading procedure, animals were trained to discriminate between 10% ethanol and water in 30 min daily sessions. Beginning with self-administration training at the 10% ethanol concentration, discriminative stimuli predictive of ethanol vs. water availability were presented during the ethanol and water self administration sessions, respectively. The discriminative stimulus for ethanol consisted of the odour of an orange extract , whereas water availability was signaled by an anize extract . The olfactory stimuli were generated by depositing six to eight drops of the respective extract into the bedding of the operant chamber. In addition,hydroponic grow table each lever-press resulting in delivery of ethanol was paired with illumination of the chamber’s house light for 5 s . The corresponding cue during water sessions was a 5 s tone .

Concurrently with the presentation of these stimuli, a 5 s time-out period was in effect, during which responses were recorded but not reinforced. The olfactory stimuli serving as S+ or S– for ethanol availability were introduced 1 min before extension of the levers and remained present throughout the 30 min sessions.The rats were only given ethanol sessions during the first 3 days of the conditioning phase. Subsequently ethanol and water sessions were conducted in random order across training days, with the constraint that all rats received a total of 10 ethanol and 10 water sessions.Reinstatement tests began the day after the last extinction session. These tests lasted 30 min under conditions identical to those during the conditioning phase, except that alcohol and water were not made available. Sessions were initiated by the extension of both levers and presentation of either the ethanol S+ or water S– paired stimuli. The respective discriminative stimulus remained present during the entire session and responses at the previously active lever were followed by activation of the delivery mechanism and a 5 s presentation of the CS+ in the S+ condition or the CS– in the S– condition. Animals were tested under the S+ ⁄ CS+ condition on day 1 and under the S– ⁄ CS– condition on day 2. Subsequently, reinstatement experiments were conducted every fourth day , in which AM404 was administered 30 min prior to the sessions. Responding at the inactive lever was constantly recorded to monitor possible non-specific behavioral effects.Pre-treatment with the anandamide transport inhibitor AM404 30 min prior to the ethanol self-administration session significantly reduced the operant response for ethanol in a dose-dependent manner . This effect was not due to a decrease in the reinforcing value of ethanol because progressive ratio experiments resulted in similar break points for animals treated with vehicle or AM404 . They were not derived or a motor inhibition induced by AM404 as the 2 mg ⁄ kg dose did not affect locomotion at the time of operant behavior testing . The effects were selective for ethanol because pre-treatment with AM404 did not modify operant responding for saccharin .

In addition, administration of AM404 did not alter food motivation and thus, food intake in rats deprived of food for 24 h . These results suggest that the pharmacological effects of the anandamide transport inhibitor are not related to a devaluation of the motivational state or a devaluation of motivational properties of natural reinforcers.In a subsequent experiment, we tested the efficacy of AM404 as a modulator of not only the operant responses for ethanol but also the operant responses elicited by the contextual stimuli associated with alcohol. As the highest dose tested resulted in significant inhibition of locomotion, we did not administer it in this context. Once a stable extinction baseline was observed, we induced relapse by presenting cues associated with ethanol delivery during training. Ethanol-related contextual stimuli elicited ethanol-seeking behavior, as operant responses induced by ethanol-associated stimuli were more intense and significantly higher than those observed on the last day of extinction . When AM404 was injected 30 min prior to cue presentation, it failed to alter the responses for ethanol seeking , indicating that anandamide uptake inhibition was not effective in preventing cue-induced relapse.The major finding of the present study is the demonstration that acute administration of the anandamide transport inhibitor AM404 reduce sethanol self-administration under an operant conditioning schedule. This compound does not affect the relapse induced by contextual cues associated with ethanol. The effects of AM404 seem to be selective for ethanol, as it was unable to suppress responding for other reinforcers, such as saccharin or food intake, suggesting that this effect is not related to a decrease in a general motivational state. This is confirmed by the lack of action of AM404 on the motivational properties of ethanol, as measured in the progressive ratio paradigm. This suppressive effect of AM404 on ethanol self-administration seems to be independent of the already known anandamide-induced motor impairment, as the lowest effective dose tested did not alter motor behavior in the open field. Moreover, the actions of AM404 were found to be independent of a potentiation of the sedative effects of ethanol.

Finally, neither experiments with cannabinoid CB1 receptor agonists nor with cannabinoid CB1 and CB2 receptor antagonists allowed us to obtain a direct pharmacological confirmation of the role of known cannabinoid receptors on the effects of AM404. The finding of a similar profile of effects using ACEA, a selective cannabinoid CB1 receptor ligand that shares the arachidonoyl moiety with both anandamide and AM404, suggests a common unknown target responsible for the effects of AM404 on ethanol self-administration. The lack of effects of WIN 55,212-2 and HU-210 at doses devoid of motor side-effects suggests that AM404 does not exert its actions through a CB1 receptor-mediated mechanism. AM404 was the first synthetic inhibitor of anandamide uptake and it has been shown to potentiate many effects elicited by anandamide in vitro and in vivo . As AM404 does not activate cannabinoid receptors , the effects of this drug were suggested to result from the elevation of endogenous anandamide levels . However, recent findings suggest that AM404 also directly activates the vanilloid VR1 receptor , complicating the identification of its mechanism of action on ethanol self-administration. However, the effect of AM404 was not reversed or enhanced by pre-treatment with the competitive vanilloid VR1 receptor antagonist capsazepine, indicating that the inhibitory action of AM404 is not mediated through VR1 stimulation and may be derived from other targets in the endocannabinoid system. Following this rationale we studied the involvement of the cannabinoid CB1 receptor, the natural target of anandamide. In order to confirm its participation we first studied whether the cannabinoid receptor antagonist SR141716A reversed the actions of AM404. This pharmacological test was complicated by the inhibitory actions of SR141716A on ethanol self-administration that precluded the observation of a reversal of the actions of AM404. A second strategy was to compare the actions of AM404 with those of selective cannabinoid CB1 receptor agonists belonging to three of the four main classes of cannabinoid agonists: eicosanoids ,flood tray aminoalkylindoles and classical cannabinoids . The effects of these compounds in ethanol self-administration are not similar to those of AM404. ACEA and WIN 55,212-2 reduced ethanol self-administration, although the component of motor inhibition of WIN 55,212-2 might be responsible for this effect. However, the classical cannabinoid receptor agonist HU-210 did not affect ethanol self-administration . We replicated this finding in a separate study in Marchigian Sardinian alcohol-preferring rats . These results indicate that the contribution of the CB1 receptors to AM404 cannot be supported. The similar profile of actions observed after systemic administration of either cannabinoid CB1 receptor agonists or antagonist seems to be challenging. It has been reported that both cannabinoid CB1 receptor agonists, such as tetrahydrocannabinol, CP55 940 and WIN 55,212-2, and cannabinoid receptor antagonist ⁄ inverse agonists, such as SR141716A, suppress operant behavior . These reports stress the pleiotropic spectrum of actions found after the interference with endocannabinoid signaling. The complex roles of the endocannabinoid system on the regulation of GABA and glutamate synapses throughout the brain circuits processing the appetitive ⁄ motivational properties of ethanol might explain these findings .

As an example, we have recently described that intracerebral injections of SR141716A only affect ethanol selfadministration in rats when the CB1 antagonist is infused in the prefrontal cortex but not in the hippocampus or dorsal striatum . Moreover, in this study, local blockade of fatty acid amidohydrolase, the main enzyme that degrades anandamide, enhances ethanol self-administration when injected into the prefrontal cortex. However, we cannot exclude additional targets such as noncloned cannabinoid-like receptors on which anandamide and WIN 55,212-2 may act. Thus, the present study stressed the need to clarify the growing complexity of endocannabinoid pharmacology, especially in the field of motivated behaviors. Although the present results exclude VR1, CB1 and CB2 receptors as the targets of the effects of AM404, we cannot exclude the contribution of endocannabinoids elevated by AM404 to the present actions, especially because the endocannabinoid system has been recently implicated in the neuroadaptations that occur during acute alcohol exposure, alcohol dependence and abstinence. Several studies have documented that endocannabinoid transmission is acutely inhibited by ethanol and becomes hyperactive during chronic ethanol administration, as revealed by the increase in the levels of endocannabinoids and the down-regulation of CB1 receptors . Thus, it is tempting to imagine that those compounds that increase endocannabinoid transmission, such as AM404, might be useful in reducing operant responses for ethanol. With the precautions derived from the non-CB1 profile of the effects of AM404, we propose that the increased levels of endogenous cannabinoids occurring during chronic ethanol administration contribute to facilitate the action of AM404; the neuroadaptations in the central nervous system associated with chronic ethanol intake lead to an increase in anandamide levels and this event could enhance the action of AM404 acting through the increased endogenous anadamide. However, we also demonstrate that the acute administration of AM404 was not able to suppress the relapse response for ethanol, i.e. the reinstatement of ethanol responding induced by the presentation of contextual cues associated with ethanol after a period of extinction. The differential response to AM404 in self-administration and relapse conditions may have a neuropharmacological basis in the recently described changes in endocannabinoid levels after chronic ethanol exposure . A possible explanation for these differences may reside in the probable alterations induced by chronically consumed ethanol in the functionality of the receptor systems mediating the central effects of ethanol that sustain ethanol-drinking behavior in rats. These neuroadaptation processes might result in a decreased potency and efficacy of the ligands. The increased levels of anandamide observed during ethanol consumption may return to basal levels or even disappear and thereby AM404 could not be acting in such a situation.This hypothesis is supported by the results obtained recently by Gonzalez et al. who showed that the levels of endocannabinoids underwent significant changes in reward-related areas during relapse, showing the lowest values in this phase.

Similar to findings in studies involving alcohol and brain injury, substance abuse was associated with poorer neuropsychological and functional outcomes . Literature reviews also support this finding, with findings indicating that almost 40% of TBI patients had a positive toxicology screen, or had reported using drugs, with marijuana use accounting for more than half of the drug use . Similar to the large percentage of missing data for alcohol screen, the variable presence of other drugs also had a large percentage of missing data . This is important to consider, as a large percentage of missing data may cause bias. Yet, in this study, even with the large percentage of missing data, the presence of other drugs was found to have a negative influence on TBI severity as indicated by lower GCS scores compared to those who did not have other drugs present on admission. It is important to consider that both alcohol and drug use at the time of injury can confound GCS assessment in trauma patients. Although findings from this study corroborate findings from TBI literature examining substance use, it may be judicious to acquire GCS scores after any intoxicating substances have worn off, perhaps hours or even up to a few days post injury. The GCS score is often assessed numerous times in a trauma patient’s hospital stay, however, the NTDB data set does not include other GCS scores, only the first one on arrival at the hospital. Finally, the large percentage of missing data for both alcohol screen result and presence of other drugs should be considered and addressed. Because blood alcohol and drug measurements in emergency departments are likely biased towards intoxicated and incoherent patients. This can help explain the large percentage of missing data when it comes to these two variables. As mentioned previously, clinicians often will forget to draw a blood sample for alcohol and or drugs, and even if they do, these results may not be entered into the medical record or the registry in a timely and accurate manner. These variations in practice create a large proportion of missing data as it relates to alcohol and toxicology screens performed and documented. For purposes of this study,vertical grow rack alcohol screen results were imputed, but as helpful as imputation can be to an analysis, it can also misrepresent the actual number of participants with a positive alcohol result thereby biasing the results.

Participants with a known history of substance abuse were found to have slightly higher GCS scores when compared to patients who did not. For every participant who had a history and a diagnosis of substance abuse, GCS scores increased by .075 units. Higher GCS scores indicated better neurological function and a less severe TBI. The study by Nguyen et al. and Leskovan et al. explore the relationship between marijuana use, and alcohol, on mortality. The effect of marijuana on TBI severity is far less studied than alcohol, though preclinical studies have shown that the presence of marijuana is associated with some neuroprotective effects, including attenuated cell apoptosis, alleviation of cerebral edema, and improved cerebral blood flow . Further studies are needed to investigate the effects of marijuana on TBI severity alone, not when combined with alcohol or other substances. These findings cannot be discussed without addressing the issue of missing data. Variables that influence GCS scores and TBI severity, such as alcohol screen result, sex, presence of drugs, history of cancer, history of mental and personality disorder, and history of alcohol abuse all had some element of missing data. All the aforementioned variables had less than 6% of the data missing, with some of them having less than 1% missing data . Similarly, history of comorbid conditions all had less than 3% missing data. The two variables that had a large percentage of data missing were the presence of THC and the presence of other drugs . Despite the missing data, both those variables were found to have a statistically significant influence on GCS scores, hence, TBI severity. Though statistically significant, the validity of those findings should be cautiously interpreted within the context of such large percentage of missing values for these hypothesized explanatory variables. One of the leading causes of injuries resulting in TBI incidence are collision related, such as motor vehicle or motorcycle crashes. Furthermore, almost half of the US states have legalized marijuana for medical use with some states allowing recreational use of marijuana. Therefore, collision type mechanism of injuries was examined to see if there was any mediating influence on TBI severity in the presence of THC.

It was determined that motor vehicle collisions did not influence, or mediate, the relationship between THC and TBI severity. However, motorcycle collisions suggested a partial influence on TBI severity. This was an expected result as studies have shown that head injuries are the leading cause of death in fatal motorcycle crashes . It is therefore not surprising to see that GCS scores were reduced when motorcycle collisions were examined for mediating influences on TBI severity in the presence of THC. In one study by Steinemann et al. , THC positivity among road traffic collisions in one US state tripled, with the number of THC positive patients presenting to the highest-level trauma center doubling. However, this data should be interpreted cautiously within the context of such large percentages of missing values for hypothesized explanatory variables. Finally, it is important to note the surprising finding that only 22 participants were found to have been involved in a motor vehicle collision, and only 16 were involved in a motor cycle crash. In the original data set, only 16,324 of 997,970 were involved in a motor vehicle collision, and 12,826 of 997,970 were involved in a motor cycle collision. In 2015, the CDC reported that more than 2.3 million people presented to the emergency department with motor vehicle-related injuries. Because not every single motor vehicle collision warrants a trauma activation or for the patient to be seen by a trauma surgeon, the number represented in the trauma registries would be much less. Hence, this may somewhat explain the lower numbers presented in the 2017 NTDB data set . Several limitations of this study should be noted. Primarily, this study was a retrospective cohort study, therefore it may be missing potentially relevant data. Retrospective cohort studies,though time efficient and cost effective, can be limited due to the nature of data collected. Missing data on several important predictor variables represents another drawback. The patient population in this study was heavily skewed towards moderate and severe TBI patients from one year of available data. A more evenly distributed sample over a longer time period with a larger number of moderate and severe TBI patients would provide more sensitive analyses. The retrospective nature of this study limits the conclusions that can be determined as the methodology was not able to ascertain any measure of acute versus chronic marijuana use. Urine toxicology screens, such as those used in the ED, detectable levels of THC can be present for up to 4.6 days after the last noted use for individuals who do not use marijuana frequently, or up to 15.4 days after last use for those who are frequent users .

Therefore,vertical farming racks the presence of marijuana at the time of exposure may not correlate with recent use. Timing of exposure may be a factor and is an important limitation in this study. Additionally, study findings are based on patients with TBI that have had a urine THC test performed. Since not all patients with moderate or severe TBI were tested for the presence of THC, bias is thus introduced. There was a large percentage of study participants who were not tested or had missing test results for THC . Consequently, a more accurate analysis of THC prevalence and association was not possible as there was no way to determine which of those cases that were not tested or had no results documented were positive for THC. It is important to note that despite there being a small percentage of THC prevalence, this study reflects only one year worth of data, from 2017, and that establishing previous prevalence rates for comparison from the NTDB cannot be calculated. This is because the presence of THC was never abstracted nor documented in the data set prior to 2017. Future studies examining prevalence rates for a series of years is warranted. Observational research has been shown to provide mis-estimations of the outcome of interest. Data analyzed from the NTDB is extracted from various trauma registries across the United States and Canada. Each hospital employs its own registry abstractors who input the data collected from the electronic medical record into the registry which then feeds into the NTDB. This is an important limitation as the documentation and accuracy of data inputted may be inaccurate, incomplete, or inconsistent. This can result in information bias. Furthermore, systematic under reporting of data by participating hospitals can result in selection bias and create an inconsistent database. An example of this was the lack of consistency in the measurement and documentation of blood alcohol levels at time of hospital admission, and the missed opportunities for urine testing. This contributed to a large percentage of missing data which may have also introduced informational bias. Additionally, this variation in reporting results in incomplete data, as seen in this study, as well as conflicting data. There were two occasions where participants were documented as having not being tested for any substances yet were each found to have been positive for THC and/or cocaine. Outcomes of such practices and variations between trauma registries leads to a lack of confidence regarding data accuracy and resulting analyses. Traumatic brain injury is a significant public health concern and a leading cause of death and disability. Many TBI patients have substance use exposure at the time of injury. This study aimed at examining the relationship between marijuana exposure at the time of injury and TBI severity in moderate and severely injured TBI patients. The study findings are timely as the number of states legalizing marijuana for both medical and recreational use increases. This retrospective cross-sectional design study analyzed a large data set retrieved from the National Trauma Data Bank of patients with traumatic brain injury and the association between the presence of THC and brain injury severity, as defined by the GCS score. This is the first known study to examine the presence of THC at the time of injury and its effect on brain injury in a large demographic from a national dataset. The NTDB dataset captures 65% of all trauma hospitals capture; so, with some confidence the claim can be made that moderate and severe TBI, in this data set, are representative of the TBI population in North America. This study found a smaller prevalence rate of THC presence in a purposive sample of TBI patients, but further studies are needed to estimate more accurate prevalence rates now that future datasets from the NTDB will delineate the types of substances tested. This will also allow for larger datasets to be analyzed which may yield different results. As is, the current dataset is not sufficient to establish strong analyses due to the large percentage of missing data, inconsistencies within the data itself, and limited to one dataset as previous datasets did not have the necessary drug information needed for analysis. Despite the limitations inherent to retrospective studies and to databases such as the NTDB, findings from this study suggest an important link between the presence of a positive THC results and GCS score, hence TBI severity. Only one research study at the time of when the systematic literature review for this present study was done investigated the effects of THC presence in TBI patients and its influence on mortality. To date, there has been one identified study that investigated the influence of marijuana on TBI mortality . When examining the differences between participants who tested positive for THC and those who did not, it was found that GCS scores were lower for those who tested positive, indicating a more serious TBI. Additionally, participants who had a had a current diagnosis, or history of, cancer or substance abuse, were more likely to have tested positive for THC. This study found that the presence of THC was significantly associated with lower GCS scores and a potentially more severe TBI; this relationship was significant without controlling for other predicting variables.

The treatment phase of identified erroneous data involves correcting, deleting or leaving the error unchanged . For purposes of this study, if impossible or missing values are observed, they will have to be deleted, as there would be no way of correcting that value related to the retrospective and secondary nature of the data. For data points that are true extremes, further examination on the influence of these data points, individually and collectively, on analysis will be made prior to determining whether or not that data point will be deleted or left unchanged . It is important to deal with missing data because missing data can create bias. First, an exploratory analysis will be performed to look at frequencies or percentages of missing data, and to help identify how much data is missing. Next, an analysis of the mechanisms, or types, of missingness will be performed to identify whether the missing data is missing completely at random , missing at random , or not missing at random using statistical tests, such as Little’s test for MCAR. Following this, an analysis for patterns of missingness will be performed using a missing pattern value chart. There are two patterns that may be potentially observed: 1) a monotone pattern where data is missing systematically, or 2) an arbitrary pattern where data are missing at random . While the analyses are not definitive, they can bring attention to blatant anomalies in the missingness of data and help to make decisions on the missing data handling procedures.There are a variety of methods that can be utilized to deal with missing data. The type of method utilized will depend on the percentage of missing data present and cannot be specified beforehand. Simple methods, such as list wise or pairwise deletion are helpful when the percentage of missing data is less than 5%. Listwise deletion, also known as complete-case analysis,vertical growing systems removes all data for a case with one or more missing values. In other words, that case is omitted completely.

A disadvantage when using listwise deletion is that it can reduce the sample size. On the other hand, pairwise deletion, also known as available-case analysis, aims at minimizing the loss of other potential data incurred with listwise deletion. Pairwise deletion still uses that case when analyzing other variables with non-missing values; it just excludes that one value with a missing data. However, pairwise deletion does have its disadvantages in that most software packages use the average sample size across analyses which can create over or underestimation. If the percentage of missing data is greater than 5%, then more advanced methods of dealing with missing data can be utilized, such as imputation. Imputation methods will depend on the pattern of missingness identified and the type of variable requiring imputation . In patterns where missing data is systematic or monotone, methods such as regression, predicted mean matching or propensity scoring are helpful. In patterns where missing data is arbitrary or at random, methods such as multiple imputation using maximum likelihood regression methods to predict missing values based on observed values and sensitivity analyses that simulate the results based on a range of plausible values can be used. Aim 1. For Aim 1, the objective is to determine the prevalence of marijuana exposure in patients with moderate or severe TBI. Analyses will be conducted using the Statistical Package for the Social Sciences software. The proportion of TBI patients who have marijuana present on admission will be reported. Unadjusted prevalence will be determined through a 2×2 table. Prevalence rates will be calculated for total number of TBIs. Aim 2. For Aim 2, the objective is to determine the correlates associated with the presence of marijuana exposure at the time of injury. The correlates included in Aim 2 will be also collected for the sample of participants without marijuana exposure at time of injury. Measures of central tendency, including range, means, proportions and standard deviations will be calculated. These basic summary statistics will be calculated for continuous variables and binary categorical variables . Continuous variables will be plotted to assess for normality; tests to assess for normality will include kurtosis and skewness.

If data is normally distributed, then parametric statistics will be utilized. If data is not normally distributed, then non-parametric statistics will be utilized. Frequency distributions, including numbers and percentages, will be generated for each of the categorical variables/correlates; scatterplots will be created so that outliers can be identified. All correlate variables presented in table 6 will be examined; all the variables but one are categorical variables. Categorical variables will be mapped against presence of marijuana exposure and TBI severity to determine if significant differences are present across each of the categories. Tests to determine significant differences across categories include chi-square test or Fisher’s exact test based on the data. Variables that are identified as significant will be used as covariates in the adjusted prevalence rates. The variable of age is a continuous variable. The literature suggests that the relationship between age and drug exposure is not linear so we will test this relationship in this study. For this study a bar plot graph plotting age against marijuana exposure will be used to determine if a linear relationship exists. If there is not a linear relationship, the variable will be categorized. Correlates that are identified as significant will become covariates in the adjusted prevalence analysis. Prior to the adjusted prevalence analysis, these covariates will be examined for multi-collinearity. Aim 3. For Aim 3, the objective is to determine the relationship between marijuana exposure at the time of injury, the mechanism of injury, and TBI severity. The null hypothesis is that a relationship between marijuana at the time of injury, the mechanism of injury, and severity of TBI does not exist. As illustrated in the conceptual framework , mechanism of injury is considered a mediating variable; it potentially mediates the relationship between marijuana exposure at time of injury and TBI severity . First an estimate of the effect between marijuana exposure and TBI severity will be obtained without the mediator variable of mechanism of injury. To test for mediation, several regression analyses will be conducted that include the mediator variable and significance of the coefficients will be examined in each step to assess for direct and indirect effects. First, I will test for a direct relationship between marijuana exposure and TBI severity. Assuming there is a significant relationship between the two variables, I will then conduct an analysis to determine if marijuana exposure affects mechanism of injury. Assuming there is a significant effect, I will then conduct an analysis to determine if mechanism of injury affects TBI severity, and whether the mediation effect is complete or partial . To determine if the mediation effect is statistically significant I will use either the Sobel test or bootstrapping methods .

All analyses will be conducted unadjusted and then adjusted for covariates and confounders identified a priori and via aim 2 . The analyses will use logistic regression modeling because the dependent variable, TBI severity, is a dichotomous variable with only two choices, moderate or severe TBI. While TBI severity can be considered a continuous variable if using the number scoring of the GCS scale, a binary variable will be used as it is easier to interpret for clinicians using a numerical score: clinicians treat not on subtle degrees of TBI severity,pruning cannabis but whether it is a moderate or severe one based on GCS threshold cut-offs. Dummy variables will be used to input non-binary categorical variables into the analysis. However, with the predicted large sample size, and understanding the potentially significant confounding effects of certain variables such as other drugs, I hope to create binary variables for each drug listed in the NTDB database . But if this is unable to be done another approach would be to code all drug use into 3 categories: a value of 0 assigned for ‘no drug use’, a value of 1 for ‘stimulants’ only . Observational studies offer valuable methods for studying various problems within healthcare where other study design methods, such as randomized controlled designs , may not be feasible or even unethical. High quality observational studies can render invaluable and credible results that positively impact healthcare when studying clinically relevant topics in patient populations of interest to practicing clinicians. Despite this, observational studies can be subject to a few potential problems within the design and analytical phases rendering results highly compromised. Potential problems that will be encountered in this study design are selection bias, information bias and confounding. Possible countermeasures to address these problems will be discussed in this section. A potential problem regarding selection bias is present in the current study. The target study population is comprised of a purposive sample of patients registered in the NTDB. The NTDB is a centralized national trauma registry developed by the American College of Surgeons with the largest repository of trauma related data and metrics reported by 65% of trauma centers across the U.S. and Canada. The main advantage to utilizing such a registry for this study is that it constitutes the largest trauma database in the U.S. Furthermore, the NTDB allows for risk-adjusted analyses which can be important when evaluating outcomes in trauma . Despite its incredible potential in informing trauma related research, the selection of participants from the NTDB is not without its own biases. The reporting of data into the NTDB is done on a voluntary basis by participating trauma centers, rendering a convenience sample that may not be representative of all trauma patients, and may also not be representative of all trauma centers across the U.S. . This creates the problem of selection bias. Furthermore, the NTDB is subject to the limitations of selection bias is that it includes a larger number of trauma centers with typically more severely injured patients potentially under representing patients with milder traumatic injuries and injury scores . Additionally, patients who may be traumatically injured and who are not admitted to a participating trauma center will not be included in the NTDB, nor will trauma patients who died on scene before being transported. Another consideration to note is that participating hospitals may differ in their criteria of which patients to include in the database, specifically patients who are dead on arrival or those who die in the Emergency Department . This discrepancy in inclusion and exclusion criteria between hospitals regarding specific injuries makes representative comparisons potentially difficult.

Lastly, it is important to mention that large databases such as the NTDB are subject to missing data or disparate data. This is often due a multitude of factors, a few of which various demographic data points, test results and other key information, such as procedures, that may not be documented in the health record and therefore omitted in the database . Missing data often contributes to information bias; however, it can also contribute to selection bias because one of the methods in dealing with missing data is excluding participants for which data is missing thereby creating potential selection bias. Missing data may undermine the ability to make valid inferences, therefore, steps will be taken throughout the design and operational stages and methods within this study to avoid or minimize missing data. Methods to reduce information bias that can lead to selection bias will be discussed in the analysis section of this paper. Due to the methods by which data are collected and inputted into the NTDB, potential problems are encountered in terms of data accuracy. Under reporting of variables obtained from the NTDB has often been noted as a problem due to the reliability of data extraction by participating hospitals . The data is self-reported and often inputted by staff dedicated to data collection. A major variance between participating hospitals is that hospitals with more resources are more likely to have dedicated staff to data collection. This can lead to informational bias in those hospitals that are more compliant in reporting data metrics when compared to others that are not. For example, hospital data registries that have incomplete data on complications may appear to deliver better care than hospitals that consistently record all complications.

Unlike MWH, WWH demonstrated a global impairment profile with spared verbal recognition. Consistently, previous findings regarding memory impairment among PWH found this impairment to be more dependent on frontal and subcortical structures with relatively normal memory retention but impaired memory retrieval . Even in the female-specific profile of relative weakness in learning and memory, recognition was less impaired compared to learning and recall. We can only speculate as to why the sparing of recognition in the global impairment profile was specific to WWH and to verbal vs. visual memory. It is possible that, in the context of cognitive impairment in HIV, the female advantage in verbal memory may be most salient for the least cognitively taxing memory component, recognition performance, and this advantage is not fully adjusted for in our demographically corrected T-scores. Despite the heterogeneity in cognitive profiles by sex, the sociodemographic/clinical/biological factors associated with these cognitive profiles were similar for MWH and WWH suggesting that, although the same factors confer increased vulnerability to cognitive dysfunction, the adverse effects of these factors impact brain function differently in men and women. In both MWH and WWH, WRAT-4 had the greatest discriminative value of profile class followed by HIV disease variables , depressive symptoms, age, race/ethnicity and years of education. WRAT-4 scores have been consistently identified as an important determinant of cognitive function among PWH, with lower WRAT-4 scores conferring risk for cognitive impairment . WRAT-4 performance may be particularly salient in this population, given that reading level may reflect education quality, above and beyond years of education, especially in lower socioeconomic populations because of the many factors impacting education quality . Additionally,vertical farming systems reading level is associated with health outcomes including hospitalizations and outpatient doctor visits and, thus, may be a proxy for bio-psychosocial factors underlying general health . HIV disease variables were also strong determinants of cognitive profiles in both men and women.

Aside from some instances of a shorter duration of HIV disease relating to more cognitive impairment in WWH and in the total sample, the more biologically-based HIV disease variables were associated with cognitive impairment in the expected direction; higher current and nadir CD4 count and lower viral load were protective against cognitive impairment. It is curious that the global weakness with spared verbal recognition profile in women was associated with more severe HIV-related variables yet with shorter duration of HIV infection. We speculate that the shorter HIV infection in WWH may reflect CNS effects of untreated and/or early course HIV infection. Alternatively, the self-reported shorter duration of infection may not have been accurate, to the extent that WWH lived longer with untested/undetected infections. Findings are consistent with a wealth of literature relating proxies of HIV disease burden and severity to cognitive function and suggests that, even in the era of effective ART when viral suppression is common, HIV disease burden can have adverse effects on the brain possibly due to poor penetration of ARTs into the CNS, ART resistance, poor medication adherence , and/or the establishment of viral reservoirs in the CNS reservoir . In line with hypotheses of mental health factors relating to cognitive impairment profiles more strongly in women, current diagnosis of MDD was a predictor of cognitive profiles only among WWH. Although the prevalence of a current or lifetime diagnosis of MDD did not differ between WWH and MWH, MDD was an important risk factor of demonstrating Global weaknesses with spared verbal recognitioncompared to the profile demonstrating only Weakness in motor function . Our work indicates that HIV comorbid with depression affects certain cognitive domains including cognitive control, and that these effects are largest in women. Specifically, WWH with elevated depressive symptoms had 5 times the odds of impairment on Stroop Trial 3, a measure of behavioral inhibition, compared to HIV-uninfected depressed women, and 3 times the odds of impairment on that test compared to depressed MWH. In a recent meta-analysis, small to moderate deficits in declarative memory and cognitive control were documented not only in individuals with current MDD but also in individuals with remitted MDD, leading to the conclusion that these deficits occur independently of episodes of low mood in individuals with “active” MDD .

Together these lines of work suggest that MDD would exacerbate cognitive difficulties in PWH, particularly in the cognitive domains of declarative memory and cognitive control in WWH. Our study has limitations. Although we were adequately powered within both WWH and MWH , the magnitude of power was discrepant by sex considering that women represented 20% of our sample. Larger-scale studies in WWH only are currently underway. The generalizability of our findings also warrant additional study as the profiles identified here may not represent the profiles among all PWH. Due to the unavailability of data, we were unable to explore certain psychosocial factors as potential determinants of cognitive profiles. Our analyses were cross-sectional which allows us to identify determinants associated with cognitive profiles but precludes us from determining the temporal relationships between these factors and cognitive function. Although many of the related factors may be risk factors for cognitive impairment, reverse causality is possible with some of the factors resulting from cognitive impairment . Additionally, interpretation of the machine learning results should be done with care as RF is an ensemble model that is inherently non-linear in nature. This means that the importance and predictive power of every variable is specified in the context of other variables. This can lead to situations where an important predictive variable in the RF model has no significant difference in the overall comparison but has dramatic differences when included with other variables in the model. As such, this model should be interpreted as hypothesis-generating and identifies variables in need of further investigation. Lastly, because our study was focused on sex differences in cognitive profiles within PWH, we did not include a HIV-seronegative comparison group. Thus, we cannot determine the degree to which HIV contributes to sex differences in cognitive profiles. However, the independent HIV-related predictors does suggest that HIV has a role. Despite these limitations, we selected RF over linear models such as lasso and ridge regression because RF models had more predictive power and higher accuracy in this data compared to the linear models, even linear models with tuning parameters such as ridge and lasso that can used for feature selection. The results from these models mirror the P-values for the univariate comparisons , which is expected since analysis of variance and t-tests are also linear models. Moreover, RF models are more optimal for handling missing data, the inclusion of categorical predictor variables, and the use of categorical outcome measures which was the case in the present study.

RF models also account for the complexity in the data that can arise from multi-collinearity often seen in large feature sets. In conclusion, our results also suggest that sex is a contributor to the heterogeneity in cognitive profiles among PWH and that cognitive findings from MWH or male-dominant samples cannot be wholly generalized to WWH. Whereas, MWH showed an unimpaired profile and even a cognitively advantageous profile, WWH only showed impairment profiles that included global and more domain-specific impairment,cannabis grow room which supports previous findings of greater cognitive impairment in WWH than in MWH . Although the strongest determinants of cognitive profiles were similar in MWH and WWH including WRAT- 4, HIV disease characteristics, age and depressive symptoms, the direction of these associations sometimes differed. This suggests that the effects of certain biological, clinical, or demographic factors on the brain and cognition may manifest differently in MWH and WWH and that sex may contribute to heterogeneity not only in cognitive profiles but in their determinants although studies with larger numbers of WWH areneeded to more definitively test these hypotheses. It is important to detect these differing cognitive profiles and their associated risk/protective factors as this information can help to identify differing mechanisms contributing to cognitive impairment and whether these mechanisms are related to HIV disease, neurotoxic effects of ART medications, and/or comorbidities that are highly prevalent among PWH . Given the longer lifespan of PWH in the era of effective antiretroviral therapy, cognitive profiling will also inform aging-related effects on cognition in the context of HIV and perhaps early clinical indicators of age-related neurodegenerative disease. By identifying cognitive profiles and their underlying mechanisms, we can ultimately improve our ability to treat by tailoring and directing intervention strategies to those most likely to benefit. Overall, our results stress the importance of considering sex differences in studies of the pathogenesis, clinical presentation, and treatment of cognitive dysfunction in HIV. Traumatic brain injury is a significant public health concern as it is a leading cause of mortality, morbidity and disability in the United States. According to the World Health Organization, TBI is expected to become the third leading cause of death and disability in the world by 2020. In the United States TBI contributes to a third of all injury-related deaths. The leading causes of injuries resulting in TBI prevalence are traffic related, such as motor vehicle crashes, or non-traffic related, such as falls. Notably, up to 51% of all TBI patients have substance use exposure at the time of injury. Substance use includes alcohol and drugs such as marijuana. Current existing research suggest that in general, substance-exposed patients may have worse TBI outcomes, including greater rates of mortality and severity of injury. Research has also shown that substance use exposed TBI patients suffer worse functional outcomes, which can result in socioeconomic burden to patients and the nation at large. This healthcare burden has been calculated to be approximately $76.5 billion in 2010 alone. There is a substantial body of research elucidating the role alcohol plays in injuries that lead to TBI prevalence and outcomes. Specifically, alcohol use results in impairments such as diminished motor control, blurred vision, and poor decision making, which has been shown to increase the risk of traffic related injury.

This research has been used to create public health policies and prevention programs that have made a significant health impact, such as reducing the number of alcohol-impaired drivers. Other substances have not been as well studied. For example, marijuana is a drug that despite being federally and legally regulated, remains the most widely used drug in the U.S. Marijuana use has been shown to result in similar cognitive impairments as alcohol use, such as lack of coordination, inability to pay attention, and decision-making abilities, suggesting marijuana users are similarly at increased risk for TBI. There is some indirect evidence of this, in that it has been shown that marijuana users in general are about 25% more likely to be involved in a motor vehicle crash and that the older adult marijuana users have a greater risk for falls. However, concrete data linking marijuana exposure at time of injury and TBI prevalence and severity is scarce. Adding to the concern, national surveys on drug use and health have documented an increase in individual daily marijuana use over the last 5 years. As the number of states legalizing marijuana for both medical and recreational use increases, it is imperative to resolve the ambiguity within the research available regarding the relationships between marijuana exposure at time of injury, mechanism of injury, and TBI prevalence and severity. This study found that the presence of THC was significantly associated with lower GCS scores and a potentially more severe TBI, but this relationship was significant without controlling for other predicting variables. Furthermore, a significant relationship was found between GCS scores, age, and blood alcohol levels at the time of presentation in the ED. Older participants were found to have higher GCS scores, indicating a less serious brain injury. Study participants who had higher blood alcohol levels were found to have lower GCS scores, indicating a more serious brain injury. Age and higher blood alcohol levels were found to be associated, with higher blood alcohol levels noted in younger patients. A linear regression showed different results when examining the relationship between the presence of THC and GCS scores, hence TBI severity. When controlling for all other variables, the presence of THC was not found to be an independent predictor of TBI severity.

In the oldest age decade, the H+/D− group had the highest positive psychological factors, suggesting an important relationship between these positive psychological factors and being able to live a relatively long, non-depressed life as a person living with HIV. Hence, positive psychological factors may be protective for PLWH. Individuals’ subjective health ratings may provide valuable insight to their overall well-being, as previous studies have shown an association between reported worse health ratings and an increased risk of mortality . This finding may also reflect a potential “survivor effect” given that these older individuals have had HIV for longer and as long-term survivors, may view living with HIV more positively compared to prior expectations. This study has strengths in its multi-cohort design methodology that allows us to examine the combined effects of HIV and depression on HRQoL across age cohorts; there are also some limitations, however. For example, we were not able to address questions regarding the onset of depressive symptoms in relation to HRQoL or the positive psychological factors. For instance, depression may lead to less resilience and grit or vice versa. Like prior studies , we found a higher proportion of elevated depressive symptoms among PLWH, and individuals with elevated depressive symptoms reported lower HRQoL and positive psychological factors. There may be other factors related to depression and acquiring HIV not captured by our present variables that may account for the difference in depressive symptoms by HIV status. Another limitation is the small sample size per group, especially within the H−/D+ group. Furthermore, the sample, particularly the within the PLWH groups, was predominantly male and these results may not be generalizable to females. However, within the United Sates the majority of middle-aged to older PLWH are male; thus, our study cohort is similar to the broader characteristics of PLWH in the U.S. . Given the negative consequences of depression in PLWH, it is important to identify those in greatest need of treatment.

Prior work has highlighted the usefulness of cognitive behavioral therapy for depression treatment among PLWH,rolling grow benches even in those with advanced HIV disease . Furthermore, meta-analytic work has shown psychotherapeutic interventions reduce depressive symptoms in PLWH, which in turn may lead to improved psychiatric and medical outcomes . With this said, older PLWH are less likely to be engaged in behavioral health treatment for depression than younger PLWH, highlighting the need to address underlying factors contributing to the lack of adequate mental health treatment among older PLWH . However, increasing or improving positive psychological factors may provide one potential avenue to mitigate depressive symptoms.Neisseria gonorrhoeae and Chlamydia trachomatis are the two most common bacterial sexually transmitted infections worldwide, estimated to have caused 87 and 127 million infections, respectively, in 2016. Men who have sex with men are disproportionately affected by STIs, including N. gonorrhoeae and C. trachomatis. Infections by N. gonorrhoeae and C. trachomatis can increase the risk of HIV transmission and acquisition, mediated through ulceration and mucosal inflammation. Extragenital chlamydia and gonorrhea infections are common among MSM and are of public health importance. Recent rectal gonorrhea or chlamydia infections have been associated with increased risk for HIV acquisition. Pharyngeal N. gonorrhoeae infections are also important, as they can serve as a reservoir for antimicrobial resistance. Extragenital infections are commonly asymptomatic and screening is necessary to make a diagnosis. The U.S. Centers for Disease Control and Prevention recommends at least annual screening for rectal and pharyngeal infections among sexually-active MSM. The World Health Organization guidelines also support periodic screening for rectal and urethral infections among MSM. Data regarding extragenital N. gonorrhoeae and C. trachomatis infections are primarily from high-resource settings. A recent meta-analysis of STIs in PrEP users found nearly one in four had chlamydia, gonorrhea, or syphilis at PrEP initiation. However, few reports from low resource settings were included in that meta-analysis, highlighting the need for additional data from these settings. In low-resource settings, there are significant infrastructure and cost barriers that limit the widespread availability of diagnostic tests needed to screen for extragenital N. gonorrhoeae and C. trachomatis.

Understanding the burden of gonorrhea and chlamydia in low-resource settings is also important for HIV prevention, as it can often be an entry point into HIV pre-exposure prophylaxis programs that are being scaled up worldwide. In Vietnam, the 2013 HIV/STI Integrated Biological and Behavioral Surveillance sampled 1587 MSM across the country and found a 5% prevalence of urethral chlamydia and <3% of urethral gonorrhea. That report found a 10% prevalence of rectal chlamydia and <3% of rectal gonorrhea, but oropharyngeal testing was not performed. Aside from that report, data regarding the prevalence and risk factors for extragenital chlamydia and gonorrhea infections among MSM in Vietnam are scarce. A better understanding of the prevalence and correlates of N. gonorrhoeae and C. trachomatis infections among MSM in Vietnam is needed to effectively plan for STI screening, diagnosis, and prevention programs in the setting of limited resources, especially in the context of the rapid scale-up of HIV PrEP programs. The objectives of this study were to determine the baseline prevalence of urethral, rectal, and pharyngeal N. gonorrhoeae and C. trachomatis infections within a cohort of HIVnegative MSM in Hanoi, the capital and second-largest city in Vietnam, and to examine the factors associated with N. gonorrhoeae and C. trachomatis infections. Between July 2017 and April 2019, MSM were recruited to participate in the Health in Men -Hanoi study, a prospective, observational cohort designed to investigate the prevalence and incidence of HIV and STIs, as well as the social and behavioral characteristics within this population. Participants were recruited from concurrent HIV and STI surveys among MSM that utilized time-location sampling, respondent-driven sampling, and internet-based sampling methods. Recruited individuals presented to the Sexual Health Promotion Clinic at Hanoi Medical University where informed consent and study enrollment were completed. Cohort inclusion criteria were: assigned male sex at birth, aged ≥ 16 years, having oral or anal sex with another man or transgender woman in the prior 12 months, living in Hanoi continuously for the prior 3 months and without a plan to move in the next two years, and serologically confirmed to be HIV-negative at baseline. At the time of the study, no participants were enrolled in a PrEP program, as PrEP was not available in Vietnam. Data collected at baseline in the sub-sample of HIV-negative MSM were used for this study.

Socio-demographics, substance use, sexual practices, history of STIs, and history pertaining to HIV counseling, testing, treatment, and care services, were collected through audio computer-assisted self-administered interviewing . Group sex was defined as more than one partner in a sexual encounter in the prior six months. Participants were asked about any rectal and genitourinary symptoms in the prior 6 months. Rectal symptoms were classified as any of the following: dyschezia, pruritis, bleeding, discharge, or ulcers. Genitourinary symptoms were classified as any of the following: dysuria, discharge, bleeding, pruritis, or ulcers. All participants received client-centered HIV and STI risk-reduction counseling. Urine samples, rectal swabs, and pharyngeal swabs were collected using cobas PCR urine sample kits and cobas PCR female swab collection kits and were tested for N. gonorrhoeae and C. trachomatis by NAAT on the cobas 4800 CT/NG v2.0 system . Blood was collected for HIV testing and was performed on the ARCHITECT HIV Ag/Ab Combo . Serologic testing for syphilis was done using the Architect Syphilis TP assay , with positive samples undergoing rapid plasma reagin testing and Treponema pallidum hemagglutination , as indicated . All participants with a positive NAAT for C. trachomatis or N. gonorrhoeae were considered to have an infection. Test results for C. trachomatis or N. gonorrhoeae were classified as missing if a specimen was not available for testing or if the testing had inconclusive results. Those with a positive T. pallidum-specific antibody and a measurable RPR were considered to have a syphilis infection. Descriptive statistics were applied to socio-demographic, behavioral,drying cannabis and clinical data. Predictive logistic regression modeling was used to evaluate factors associated with N. gonorrhoeae and C. trachomatis infections separately and the combined outcome of having either infection. Variables for consideration were selected a priori using an approach that included variables based on biologic basis, as well as known risk factors and confounders. The variables included in the bivariate analyses were: age, education, income, ATS use for sex, group sex, meeting sexual partners via mobile apps, prior diagnosis of STIs, and genitourinary orrectal symptoms. Symptom status was dichotomized for the logistic regression models. All variables in the bivariate analyses were also included in the multivariate analysis, with the exception of any substance use in the prior 3 months and amphetamine-type stimulantuse in the prior 3 months, which were excluded from the multivariate analysis due to high collinearity with ATS use to enhance sexual performance in the prior 6 months. Records with missing variable data were excluded from the logistic regression models. All data analyses were done using R version 3.61. There were 1498 participants in the baseline survey. Nine did not have any samples for N. gonorrhoeae and C. trachomatis testing and were excluded from the analysis. Among the remaining 1489 participants, the median age was 22 years . Income in the prior month was less than 5 million VND for 40.5% of participants and 30.8% had completed university education. Substance use in the prior 3 months was reported by 8.3% of participants and 6.5% reported using ATS to enhance sexual performance in prior 6 months. Among those reporting anal sex in the prior 6 months, 32.1% had insertive sex, 30.0% had receptive sex, and 29.5% had both.

Condomless anal intercourse in the prior 6 months was reported by 57.6% of participants. Anal sex with two or more partners in the prior month was reported by 31.8% of participants. Group sex in the prior 6 months was reported by 24.9% of participants. Over half of participants reported meeting sexual partners via websites or mobile apps in the prior 6 months. There were 841 participants who did not have genitourinary or rectal symptoms in the prior 6 months. There were 235 participants with a prior diagnosis of chlamydia, gonorrhea, or syphilis. The prevalence of syphilis was 18.3% . There were 1378 participants included in the analyses of factors associated with N. gonorrhoeae, C. trachomatis, or either N. gonorrhoeae or C. trachomatis infection, excluding those with missing variable data . In the multi-variable analysis of the combined N. gonorrhoeae or C. trachomatis outcome, those aged 25-34 years had lower odds of infection compared to those with ages 16-24 years . This was largely contributed to by C. trachomatis infection . Other independent factors associated with having either N. gonorrhoeae or C. trachomatis infections included having two or more recent sex partners , condomless anal intercourse in the prior six months , which was driven by C. trachomatis , and meeting sexual partners via mobile apps or the internet , which was driven by N. gonorrhoeae . Genitourinary or rectal symptoms in the prior 6months and group sex were associated with infections in bivariate analysis, but not in the multivariate model. A prior STI diagnosis and ATS use to enhance sexual performance were not associated with any infections in the multi-variable models. .In this study of young, HIV-negative MSM in Hanoi, Vietnam, we found a high prevalence of N. gonorrhoeaeand C. trachomatisinfections with more than one in four participants having one of these infections at baseline. Rectal infections occurred in 73.9% of those with chlamydia and 70.5% of gonorrhea infections occurred in the oropharynx. Limiting testing to the urethral site would have missed nearly three-quarters of C.trachomatis or N. gonorrhoeae infections within this cohort, as 27.4% of infections occurred in the urethra. Half of all persons with chlamydia or gonorrhea were asymptomatic, and reporting genitourinary or rectal symptoms were not associated with infections, highlighting the need for routine screening in this population. Prior surveys of urethral chlamydia or gonorrhea in Vietnam found a similar prevalence of C. trachomatisand N. gonorrhoeae , compared to the overall urethral prevalence of 7.1% and 1.3%, respectively, we reported here. While data on extragenital chlamydia and gonorrhea within Vietnam are very limited, surveys from Ho Chi Minh City, Hanoi, and Nha Trang including urethral, rectal, and pharyngeal testing among HIVnegative male sex workers, many of whom are MSM, found a high overall prevalence of N. gonorrhoeae, up to 29%, and up to 17% for C. trachomatis, although data stratified by anatomical site were not reported.

The potential applicability of the interest group mechanism identified in this paper across policy domains also has implications for fundamental models of lawmaking in American politics. Standard models conceive of lawmakers as primarily driven by the preferences of the median voters in their districts, which are generally taken as exogenous . Alternative perspectives suggest that lawmakers are primarily responsive to the pressures of organized interests seeking to advance policy goals, and moreover, that the ability of competing groups to influence politics is structured by the existing policy-scape . Findings presented here support the notion that existing policy, in part by shaping interest group capacities, affects congressional representation. This paper therefore provides quantitative empirical grounding for the difficult-to-test arguments in favor of the policy-focused approach— and one empirical framework for scholars working in this vein.Wealthy foundations have taken on increasingly prominent roles influencing education policy in the U.S. This paper uses a mix of qualitative and quantitative evidence to study the drivers and implications of the engagement of major foundations in the politics of charter schools. I show that states that adopted favorable charter laws, in addition to empowering charter schools as political actors, also drew wealthy foundations into the charter policy space by enabling them to make investments in developing new schools. Foundations later sought to protect those investments, leveraging strategic grant-making to drive the growth of a pro-charter advocacy network with national scope. Findings underscore the importance of state policy experimentation in catalyzing new interest group coalitions,commercial racks with implications for ideas about policy reform in American federalism.

In recent years, contests over policies governing charter schools have generated some of the most hard-fought battles in state politics. In 2016, Massachusetts voters rejected a ballot initiative that would have lifted the state’s cap on charter schools to allow 12 new schools each year after a $33 million campaign—at that point the most expensive in the state’s history. A few years later, in 2019, on the other side of the country, California Governor Newsom signed legislation adding restrictions to new charter schools after a big-money campaign pitting teachers unions against charter advocates. That teachers unions and other incumbent organized interests in the K-12 education sector would resist charter schools makes good sense. Teachers unions are some of the most active and well-resourced organized interests in American politics, particularly at the state and local levels where most education policy is made . Teachers at charter schools are much less likely to be unionized , so the rise of charter schools poses an acute threat to their continued strength. And while funding formulas vary across the states, broadly speaking, the more students enroll in charters the less funding is available to district schools, so the growth of charter schools also threatens union jobs in the long run. What is somewhat more surprising is the emergence of a well-resourced pro-charter advocacy coalition battling to defend and expand chartering. This coalition often includes charter schools themselves, who also are sometimes able to drum up grassroots support among the parents of their students. But, as of 2017, charter schools only enrolled about 6 percent of all public-school K-12 students . Even large charter networks like The Knowledge is Power Program do not have the resources to go toe-to toe with teachers unions in the political sphere. And charter school parents are usually lower income people of color—not a group seen as particularly powerful in American politics. More fundamental to the pro-charter political coalition than the schools themselves are wealthy philanthropists and the advocacy groups they fund.

For instance, Great Schools, which spent $23.6 million in 2016 to try to raise a cap on the number of charter schools in Massachusetts was bank-rolled primarily by the Walton family and Michael Bloomberg . Indeed, existing research has documented how the coordinated engagement of wealthy foundations has been fundamental to the emergence of a pro-charter coalition of interest groups combining a national scope with local on-the-ground presence . This paper traces the emergence and growth of this pro-charter coalition and studies its implications for the politics of education. I argue that the rise of the pro-charter education coalition depended fundamentally on early policy victories during a particular “window of opportunity” for the charter school movement. Advocates took advantage of the broad attention to education reform in the 90’s and early 2000’s to pass “charter laws” across a wide range of states. These laws provided a legal framework for new charter schools to be authorized. I show that, even though a majority of states adopted charter laws in this period, charter sector growth depended fundamentally on a smaller set of states with highly pro-charter policies. This growth, I argue, was essential for building a broader political coalition supported by foundations. In the 90’s and early 2000’s, foundations’ primary role was to provide financial and technical support to charter schools to get up-and-running. But the involvement of these foundations in directly supporting schools and other charter operations planted the seeds for subsequent political engagement. As charter schools grew and came under increasing pressure from hostile teachers unions, foundations recognized that the continued growth and viability of the charter school sector depended not just on their operational support— but also on the development of a pro-charter political coalition. Drawing on data submitted to the IRS by non-profit organizations , I document a shift in foundation grant-making towards greater political advocacy. Elite interviews suggest that key foundations recognized the importance of building political capacity through grant-making to defend earlier investments in the charter movement. The consequences of the rise of this foundation-funded, nationally scoped, political coalition have been profound.

Exploring several mini-cases, I show how foundation-funded groups have been fundamental to efforts to expand charter schools to new locales—and seek to defend charter schools in places where they have gained a foothold. This analysis has implications for our understanding for how reforms challenging incumbent vested interests can unfold over time. As Finn, Manno, and Wright write: “Aside, perhaps, from mayoral control, chartering is by far the most significant manifestation of structural and governance innovation in public education…” . What is interesting about this case for the literature on public policy reform is that, unlike other durable reforms , the advent of charter schools—except in some extreme cases like New Orleans —has largely failed to dislodge incumbent education interests. While charter school policy reforms have, to an extent, politically empowered charter schools and charter networks themselves,greenhouse rolling benches these interests have been less important to the broader pro-charter coalition than foundations. More so than generating their own interest group supports by conferring benefits , early charter laws changed the politics by drawing previously sidelined political actors—in this case, foundations—into the charter coalition. The role of philanthropists in politics is a growing and important topic of study in political science . With greater inequality concentrating wealth at the top of society, foundations have developed ever-greater financial resources . In addition, a growing cadre of living donors have sought to leverage strategic grant-making and political engagement to accelerate structural change by driving policy shifts . But this paper shows the relationship also goes in the other direction: how foundations engage in politics is shaped by prior policy decisions through policy feedback dynamics . The paper unfolds as follows. I first provide background on the growth of charter schools in the U.S. and discuss the importance of state policy decisions for the charter school sector. I then trace the emergence of a pro-charter political coalition, highlighting the role of state experimentation with charter laws in building this coalition. I proceed to present several minicases that underline the importance of this pro-charter political coalition to expanding and defending charter laws. Finally, I discuss implications for understandings of policy reform over time in American federalism and conclude. Laws allowing for the establishment of public charters schools were adopted in 40 states in the 90’s and early 2000’s. The first to adopt was Minnesota, which passed its charter law in 1991. The federal government also adopted new charter school policy in this period. The Federal Charter School Program, initiated in 1994 by amendments to the Elementary and Secondary Education Act, directed critical funding to support the growth of charter schools in states that allowed them . The expansion of charter schools generally coincides with greater choice in K-12 education. Where charters have become established, parents can opt to send their children to either publicly funded charter schools or district schools tuition-free. Charter schools are publicly funded, but privately operated. Governance from authorizers under state jurisdiction, versus local school districts, generally allows them greater autonomy than traditional public schools Charter schools’ political momentum came in part from renewed attention to education policy in the 80’s and 90’s. Several reports were published in the early 1980’s highlighting major issues in the American K-12 education system.

The most famous of these was A Nation at Risk , which famously claimed that: “Our society and its educational institutions seem to have lost sight of the basic purposes of schooling, and of the high expectations and disciplined effort needed to attain them” . The report’s call for politicians to pay greater attention to education was heeded, even as the analysis underpinning its key findings were later disputed . In the 1980’s, the states and the federal government experimented with a wide range of education reforms ranging from teacher certification standards to more standardized testing to school-based management. Most of the reforms adopted in this period operated within the highly bureaucratic system established by progressives in the early 20th century. Indeed, new policies on standards and testing were designed to further bureaucratize and centralize the education system. These types of reforms, Chubb and Moe argued in their influential Politics, Markets, and America’s Schools, were destined to fail, since they failed to address the institutional problems underlying K-12 education’s woes. The most important factor determining a school’s performance, they proposed , was its level of autonomy. And a top-down bureaucratic management structure was anathema to holding schools accountable while maintaining school autonomy. Market control, versus democratic control, they argued, would allow for greater school autonomy and, as a result, improved academic performance. Chubb and Moe thus pushed for an alternative set of reforms aimed at decentralizing the education system, instilling choice, and leveraging market competition to achieve improvements. Similar ideas were also being promoted on the left side of the political spectrum. In 1988, University of Massachusetts professor Ray Budde released Education by Charter: Restructuring School Districts.Budde advocated for allowing innovative teachers to apply for special charters to create new programs, thus devolving authority down to teachers and enhancing their autonomy. American Federation of Teachers president Al Shanker latched onto the chartering concept and promoted it as a way for teachers and their unions to maintain their central role in the face of seemingly inevitable education reforms. Chartering thus emerged in this period as a “middle-path” between the highly rigid existing system and a privatized system of vouchers promoted by those on the far right of the political spectrum . Policy entrepreneurs first took chartering from concept to law in the state of Minnesota. The effort was led by Joe Nathan, a former Minnesota teacher who had written a book promoting the charter school concept and then worked for the National Governors Association’s education reform group commissioned by Lamar Alexander and Bill Clinton. Nathan partnered with Ted Kolderie from Citizens League, a moderate “good government” Minnesota think-tank, and former State Senator Ember Reichgott Junge to develop and enact a bill that would put in place a process for schools to apply for charters to operate independently of school districts. The Minnesota bill was ultimately supported by a minority of the Democratic party , but by enough Republicans to pass. Bipartisan support within the “window of opportunity” generated from attention to education reform was critical to overcoming opposition from teachers unions and school boards in Minnesota, and later, elsewhere . Contrary to Al Shanker’s hopes, charter laws generally did not establish a role for teachers unions in the chartering process, instead generally specifying that new charter schools could operate outside of negotiated collective bargaining contracts.