ABSTRACT
Purpose: Studies have shown adolescent and young adult (AYA) participation in voting and other forms of civic engagement is associated with future optimism, increased life satisfaction and decreased health-related risk behaviors. Yet, AYA aged 18-24 are the least represented demographic at voting polls across the US. Recognizing voting and civic engagement may be an important health intervention for this population, we sought to determine factors associated with future voting intention (planning to vote in the next election) among AYA attending an urban adolescent clinic during the COVID-19 pandemic. Methods: We added four voting-related questions (Do you plan to vote in the next election? Did you vote in the last election? Are you registered to vote? Do you want to know how to register to vote?) to our pre-visit questionnaire distributed to all adolescent clinic patients ages 13-26 years. Both before and after the November 2020 election (i.e., July 2020 to March 2021), we collected 634 patient questionnaires;77% (N=487) were from patients who were age eligible to vote on November 3, 2020. We limited the current analysis to questionnaires from age eligible patients with complete responses of yes or no to all four voting questions (N=258). Using bivariate and multivariable logistic regression we examined associations between voting intention and the following factors: age, gender, race, registration status, voting in last election, and weeks to/from November 2020 election. Age was dichotomized to 17-21 vs. 22-26 years based on Locally Weighted Scatterplot Smoothing and race to Black vs. non-Black. This project was approved by the Johns Hopkins IRB. Results: Mean age was 20.7 years (SD=2.1);63.2% were 17-21 years. Sixty-five percent were female, 88% were Black, 73% were registered to vote, 48% voted in last election, and 76% had future voting intention. Mean weeks to/from November election was -1.26 (SD=10.2). In the adjusted model, older patients were nearly 70% less likely to declare future voting intention than younger patients (aOR=0.32, 95% CI=0.14-0.76);males were half as likely as females (aOR=0.45, 95% CI=0.21-0.96). Voting in the last election (aOR=18.63, 95% CI=5.51-62.97) and being registered to vote (aOR=6.12, 95% CI=2.82-13.27) predicted future voting intention. Future voting intention was not associated with race or weeks to/from November election in either the unadjusted or adjusted models. Conclusions: Our findings from a clinic sample of urban AYA point to a subgroup of youth who may be more vulnerable to disenfranchisement. The COVID-19 pandemic introduced new challenges for AYA voting and this study highlighted how providers might harness the health care visit to promote AYA voting. Registration status, one of the variables most strongly associated with future voting intention, is modifiable and easily evaluated during a healthcare visit. Future qualitative investigation will explore the differences in future voting intention by age and gender to identify other factors that may also be modifiable or addressed by adolescent providers in clinical settings. Sources of Support: Thomas Wilson Foundation (PI:Fields), NICHD T32HD052459 (PI:Trent).
ABSTRACT
Purpose: Adolescent health surveillance systems are critical for understanding patterns of marijuana use and generating data to evaluate changes in use following marijuana legalization and during the COVID-19 pandemic. The wording of survey questions may be misaligned with adolescents’ language about marijuana use and the ways they consume it. Our objectives were to compare terminology and prevalence of marijuana use between data from a local surveillance system and from a participatory research study. Methods: To understand marijuana use trajectories over the course of adolescence/young adulthood, we conducted the “Model Building with Adolescents on Peers, Partners, and Substance Use” (MAPPS) study. MAPPS was IRB-approved and included participatory group model building (GMB) with youth in Baltimore City. MAPPS participants were recruited from a health clinic and through community partners. Participants’ marijuana use was assessed with the eligibility screener, an enrollment survey, and through GMB exercises that were conducted over the course of four two-hour workshops. GMB exercises included structured activities with youth, including behavior over time graphs and documenting their mental models in real time. Two independent reviewers interpreted youths’ graphed estimates of marijuana use. Lifetime and past 30-day marijuana use prevalence estimates from MAPPS were compared to estimates from the Baltimore Youth Risk Behavior Survey (YRBS), which is conducted in partnership with CDC’s National YRBS program. Results: MAPPS participants (n=20) had an average age of 18;7 (35%) were male and 19 (95%) were Black. MAPPS participants almost exclusively used the terms weed and blunts for marijuana, whereas the Baltimore YRBS used the term marijuana, and mentioned that it was also called “pot, weed, or cannabis.” Results from MAPPS revealed several discrepancies between different assessments of marijuana use;100% reported lifetime use during GMB activities, whereas 50% (n=10) reported lifetime use on the eligibility screener and 60% (n=12) reported lifetime use on the enrollment survey. Collectively, MAPPS participants estimated that 86% of Baltimore 16-year-olds use marijuana, whereas data from the Baltimore YRBS indicate that 30.2% of eleventh graders report past 30-day use. MAPPS participants perceived that there was a high frequency of use among youth who use and explained that youth who “hit a blunt” off someone else, but who do not buy marijuana themselves, would be unlikely to self report as having used marijuana. Conclusions: Our participatory research with urban, Black youth suggests that the terminology they use for marijuana (i.e., weed, blunt) differs from terms used in local surveillance (e.g., marijuana, pot). We also found that they would consider prevalence estimates from surveillance studies to be underestimated because youth who consume peers’ blunts would not characterize themselves as having used marijuana. Therefore, surveillance questionnaires may be misestimating marijuana use due to discrepancies in terminology in questions versus in spoken language, and because collective use is not considered. Misestimations of use limit effective prevention programming, and bias studies that quantify changes in marijuana use following policy changes or during the pandemic. A more comprehensive understanding of patterns of marijuana use is an important step for improving surveillance, evaluation, and clinical assessment. Sources of Support: NIH K01DA035387.