Stress and Affect as Daily Risk Factors for Substance Use Patterns: an Application of Latent Class Analysis for Daily Diary Data

Using a Smartwatch App to Understand Young Adult Substance Use: Mixed Methods Feasibility Study

BACKGROUND

Young adults in the United States exhibit some of the highest rates of substance use compared to other age groups. Heavy and frequent substance use can be associated with a host of acute and chronic health and mental health concerns. Recent advances in ubiquitous technologies have prompted interest and innovation in using technology-based data collection instruments to understand substance use and associated harms. Existing methods for collecting granular, real-world data primarily rely on the use of smartphones to study and understand substance use in young adults. Wearable devices, such as smartwatches, show significant potential as platforms for data collection in this domain but remain underused.

OBJECTIVE

This study aims to describe the design and user evaluation of a smartwatch-based data collection app, which uses ecological momentary assessments to examine young adult substance use in daily life.

METHODS

This study used a 2-phase iterative design and acceptability evaluation process with young adults (aged 18-25 y) reporting recent alcohol or cannabis use. In phase 1, participants (8/15, 53%) used the data collection app for 14 days on their Apple Watches to report their substance use patterns, social contexts of substance use, and psychosocial risk factors (eg, affect). After this 14-day deployment, the participants completed a user experience survey and a semistructured interview to record their perspectives and experiences of using the app. Formative feedback from this phase informed feature modification and refinement of the app. In phase 2, an additional cohort (7/15, 47%) used the modified app for 14 days and provided feedback through surveys and interviews conducted after the app use period.

RESULTS

Analyses of overall app use patterns indicated high, consistent use of the app, with participants using the app for an average of 11.73 (SD 2.60) days out of 14 days of data collection. Participants reported 67 instances of substance use throughout the study, and our analysis indicates that participants were able to respond to ecological momentary assessment prompts in diverse temporal and situational contexts. Our findings from the user experience survey indicate that participants found the app usable and functional. Comparisons of app use metrics and user evaluation scores indicate that the iterative app design had a measurable and positive impact on users' experience. Qualitative data from the participant interviews highlighted the value of recording substance use patterns, low disruption to daily life, minimal overall burden, preference of platforms (smartphones vs smartwatches), and perspectives relating to privacy and app use in social contexts.

CONCLUSIONS

This study demonstrated the acceptability of using a smartwatch-based app to collect intensive, longitudinal substance use data among young adults. The findings document the utility of smartwatches as a novel platform to understand sensitive and often-stigmatized behaviors such as substance use with minimal burden.

Daily Stress, Drinking Motives and Alcohol Co-Use with Other Drugs

Background: Recent evidence indicates that alcohol and other substance co-use, compared to alcohol-only use, might be more closely associated with negative reinforcement processes, and thus more likely during periods of increased stress. The present study examined this possibility by using data from an intensive longitudinal (daily) study of college student drinkers (N = 1461, 54% women). We also examined individual differences in coping and enhancement drinking motives as predictors of alcohol and other substance co-use. Results: We used multilevel multinomial logistic regression to predict, relative to alcohol-only days, the likelihood of alcohol co-use with either cigarettes or marijuana, along with alcohol use with multiple substances and other substance-only use from daily interpersonal and academic stress, day-of-the-week, sex, and individual differences in coping and enhancement drinking motives. We found that, relative to alcohol-only, alcohol and marijuana co-use was more likely, and non-alcohol related substance use was less likely, on weekends. Alcohol and marijuana co-use was less likely, and other substance-only use was more likely, on days characterized by greater academic stress, whereas alcohol and cigarette co-use was more likely on days characterized by greater interpersonal stress. Individuals with higher levels of drinking to cope motivation were more likely to engage in alcohol and cigarette co-use, other substance-only use, and alcohol plus multiple substances, relative to alcohol-only. Individuals with higher levels of enhancement motives were more likely to engage in all types of alcohol and other substance co-use and other substance-only use relative to alcohol-only. Conclusions: Findings are discussed in terms of the complex nature of different patterns of co-use patterns when evaluating indicators of positive- and negative-reinforcement processes.

Are you thinking what I'm thinking? Defining what we mean by "polysubstance use."

The rise in drug overdoses and harms associated with the use of more than one substance has led to increased use of the term "polysubstance use" among researchers, clinicians, and public health officials. However, the term retains no consistent definition across contexts. The current authors convened from disciplines including sociology, epidemiology, neuroscience, and addiction psychiatry to propose a recommended definition of polysubstance use. An iterative process considered authors' formal and informal conversations, insights from relevant symposia, talks, and conferences, as well as their own research and clinical experiences to propose the current definition. Three key concepts were identified as necessary to define polysubstance use: (1) substances involved, (2) timing, and (3) intent. Substances involved include clarifying either (1) the number and type of substances used, (2) presence of more than one substance use disorder, or (3) primary and secondary substance use. The concept of timing is recommended to use clear terms such as simultaneous, sequential, and same-day polysubstance use to describe short-term behaviors (e.g., 30-day windows). Finally, the concept of intent refers to clarifying unintentional use or exposure when possible, and greater attention to motivations of polysubstance use. These three components should be clearly defined in research on polysubstance use to improve consistency across disciplines. Consistent definitions of polysubstance use can aid in the synthesis of evidence to better address an overdose crisis that increasingly involves multiple substances.

Daily Relations Among Alcohol and Cannabis Co-Use, Simultaneous Use, and Negative Consequences: A Day-Level Latent Profile Analysis

Objective

Concurrent and simultaneous cannabis and alcohol co-use confers risk for daily negative alcohol consequences. However, studies often treat co-use as a dichotomy, precluding examination of higher- and lower-risk co-use days. Additionally, little is known about specific alcohol consequences associated with daily co-use. Therefore, the current study 1) differentiated days based upon alcohol consumption, co-use, and simultaneous use, and 2) tested whether certain day-level use patterns conferred risk for daily alcohol consequences.

Methods

College student co-users (N=489) completed an online Timeline Followback, reporting daily alcohol consumption, negative alcohol consequences, concurrent cannabis and alcohol co-use, and simultaneous co-use (SAM) on drinking days over the past month. Day-Level Latent Profile Analysis differentiated days based upon drinking quantity, co-use, and simultaneous use, and tested whether patterns of use conferred risk for overall and specific negative alcohol consequences.

Results

Four day-level profiles emerged, including moderate consumption of alcohol-only days (57.5%), moderate consumption SAM use days (29.1%), higher consumption alcohol-only days (7.4%), and higher consumption SAM use days (6%). Higher consumption SAM use days were associated with more negative alcohol consequences than all other days; however, higher consumption SAM use days differed from higher consumption alcohol-only days in acute dependence symptoms. Higher consumption alcohol-only days were associated with more negative alcohol consequences than moderate consumption SAM days, particularly those that were action-oriented (i.e., dependence symptoms, blackout drinking, impaired control, risky behavior, social/interpersonal consequences).

Conclusions

Findings suggest that there are in fact lower-risk co-use days, and that links with unique negative alcohol consequences depend on levels of alcohol consumption and co-use.

Reasons for not drinking among young adults with simultaneous alcohol and cannabis use: A latent class analysis applied to daily diary data

Simultaneous alcohol and cannabis use (i.e., simultaneous use) is prevalent among young adults and often associated with negative consequences. Understanding reasons for not drinking (RND) may provide insight into a key intervention target for reducing negative consequences associated with simultaneous use. RND may vary on a day-to-day level, and multiple RND may be endorsed on a given day. Latent class analysis (LCA) of daily diary data is a nuanced approach that can identify complex patterns of daily RND as well as its day- and person-level covariates. The current study was a secondary data analysis of daily diary data from young adults who engaged in heavy drinking and recent simultaneous use (n = 154). We aimed to: (1) characterize daily RND, (2) use LCA to classify day-level patterns of RND, and (3) compare latent classes on same-day variables (i.e., positive and negative affect, day of the week), previous-day variables (i.e., substance use, intoxication level, consequences), and person-level characteristics (i.e., age, sex, baseline substance use frequency, simultaneous use motives). Participants completed up to 14 consecutive diaries. Multilevel LCA identified four classes of heterogeneous daily RND profiles. Daily RND classes significantly differed in terms of day of the week, previous day quantity of cannabis use, and several baseline variables (age, typical substance use, simultaneous use motives). Study findings offer preliminary support for heterogeneous RND classes among young adults engaging in simultaneous use and suggest multiple avenues for future research.

Adolescent simultaneous use of alcohol and marijuana by trends in cigarette and nicotine vaping from 2000 to 2020

BACKGROUND

Simultaneous alcohol and marijuana (SAM) use is associated with adverse consequences for youth. While SAM use is overall declining among youth, prior studies indicate increasing marijuana use among US adolescents who ever used cigarettes, suggesting possible moderation of the alcohol-marijuana relationship by cigarette use.

METHODS

We included 43,845 12-th grade students participating in Monitoring the Future data (2000-2020). A 5-level alcohol/marijuana measure was used, including past-year SAM, alcohol-only, marijuana-only, non-simultaneous alcohol and marijuana, or no use. Multinomial logistic regressions estimated associations between time periods (categorized based on sample size: 2000-2005, 2006-2009, 2010-2014, 2015-2020) and the 5-level alcohol/marijuana measure. Models adjusted for sex, race, parental education and survey mode and included interactions of time periods and lifetime cigarette or vaped nicotine use.

RESULTS

While overall SAM among 12th graders decreased from 23.65% to 18.31% between 2000 and 2020, SAM increased among students who never used cigarettes or vaped nicotine (from 5.42% to 7.03%). Among students who ever used cigarettes or vaped nicotine, SAM increased from 39.2% in 2000-2005-44.1% in 2010-2014 then declined to 37.8% in 2015-2020. Adjusted models controlling for demographics indicated that among students with no lifetime cigarette or vaped nicotine use, students in 2015-2020 had 1.40 (95% C.I. 1.15-1.71) times the odds of SAM, and 5.43 (95% C.I. 3.63-8.12) times the odds of marijuana-only (i.e., no alcohol use) compared to students who used neither in 2000-2005. Alcohol-only declined over time in both students who ever and never used cigarettes or nicotine vape products.

CONCLUSION

Paradoxically, while SAM declined in the overall adolescent US population, the prevalence of SAM increased among students who have never smoked cigarettes or vaped nicotine. This effect arises because of a substantial decline in the prevalence of cigarette smoking; smoking is a risk factor for SAM, and fewer students smoke. Increases in vaping are offsetting these changes, however. Preventing adolescent use of cigarettes and nicotine vaped products could have extended benefits for other substance use, including SAM.

Daily Stress Processes as Potential Intervention Targets to Reduce Gender Differences and Improve Mental Health Outcomes in Mid- and Later Life

The current study examines daily stress processes as risk factors for comprised mental health in midlife and later life, specifically for gender differences in depression risk. Using data from the Midlife in the United States (MIDUS) study and the National Study of Daily Experiences (NSDE), we examine (1) gender differences in depression; (2) the prospective effects of differential exposure and affective responses on 10-year depression status; (3) gender differences in daily stress-depression links. Furthermore, we explore whether the protective factor of help-seeking behavior moderates the effects of daily stress on depression. Participants included 1289 (mage = 55; SD = 12; range = 34-83; 56% female) individuals who completed the second waves of MIDUS and the 8-day NSDE daily diary protocol and participated in the third wave of MIDUS approximately 10 years later. Respondents completed assessments of depression and their seeking assistance from a psychiatrist, mental health professional, counselor, or religious leader. Covariate-adjusted logistic regression analyses revealed increased odds of depression among women compared to men, but no significant gender difference after taking daily stress into account. Higher levels of stressor exposure, negative affect, and affective reactivity were associated with increased odds of depression for both men and women. Compared to those who did not engage in help-seeking behavior, those who did had significantly greater odds of depression, and there were asymmetric patterns of daily stress effects across groups. These findings highlight differential exposure, negative affect, and affective responses to daily stress as potentially accessible intervention targets for reducing stress in daily life and mitigating longer-term depression risk during mid- and later life.

Patterns of opioid use behaviors among patients seen in the emergency department: Latent class analysis of baseline data from the POINT pragmatic trial

INTRODUCTION

The nation's overdose epidemic has been characterized by increasingly potent opioids resulting in more emergency department (ED) encounters over time. ED-based opioid use interventions are growing in popularity; however, they tend to treat people who use opioids as a homogenous population. The current study sought to understand heterogeneity among people who use opioids who encounter the ED by identifying qualitatively different subgroups among participants in an opioid use intervention clinical trial at baseline and examining associations between subgroup membership and multiple correlates.

METHODS

Participants were from a larger pragmatic clinical trial of the Planned Outreach, Intervention, Naloxone, and Treatment (POINT) intervention (n = 212; 59.2 % male, 85.3 % Non-Hispanic White, mean age = 36.6 years). The study employed latent class analysis (LCA) using five indicators of opioid use behavior: preference for opioids, preference for stimulants, usually use drugs alone, injection drug use, and opioid-related problem at ED encounter. Correlates of interest included participants' demographics, prescription histories, health care contact histories, and recovery capital (e.g., social support, naloxone knowledge).

RESULTS

The study identified three classes: (1) noninjecting opioid preferers, (2) injecting opioid and stimulant preferers, and (3) social nonopioid preferers. We identified limited significant differences in correlates across the classes: differences existed for select demographics, prescription histories, and recovery capital but not for health care contact histories. For example, members of Class 1 were the most likely to be a race/ethnicity other than non-Hispanic White, oldest on average, and most likely to have received a benzodiazepine prescription, whereas members of Class 2 had the highest average barriers to treatment and members of Class 3 were the least likely to have been diagnosed with a major mental health illness and had the lowest average barriers to treatment.

CONCLUSIONS

LCA identified distinct subgroups among POINT trial participants. Knowledge of such subgroups assists with the development of better-targeted interventions and can help staff to identify the most appropriate treatment and recovery pathways for patients.

Intersectional stigma subgroup differences in unhealthy drinking and disordered marijuana use among Black and Latino cisgender sexual minority young men

BACKGROUND

We investigated associations of intersectional stigma subgroups with alcohol and marijuana use among Black and Latino sexual minority young men. Subgroups included Minimal Stigma (low to no stigma), Select Social Stigma (occasional stigma in social relationships), Multiform Heterosexism (internalized and interpersonal heterosexism from family/friends), Multiform Racism (racism across diverse contexts), Compound Stigma (frequent, ubiquitous racism and heterosexism).

METHODS

Cohort of Black and Latino cisgender sexual minority young men (n = 414; baseline ages 16-25) surveyed semiannually 2016-2019. Generalized estimating equations integrated with latent class analysis modeled linear and quadratic age effects and association of stigma subgroups with past 6-month alcohol use, marijuana use, unhealthy drinking, and marijuana use disorder symptoms.

RESULTS

All past 6-month substance use peaked between ages 21-23 years old. Across all ages and relative to Minimal Stigma, odds of drinking were higher in every subgroup and highest in Compound Stigma (OR=2.72, 95% CI 1.17-6.35); unhealthy drinking was higher in every subgroup and highest in Multiform Heterosexism (β = 3.31, 95% CI 1.92-3.89); marijuana use disorder symptoms were higher in most subgroups and highest in Compound Stigma (β = 1.30, 95% CI 0.76-1.85). Marijuana use odds did not differ among groups.

CONCLUSION

By examining intersectional stigma subgroups, we identified subgroups for whom substance use was elevated during a development period when use tends to be highest. Young men experiencing stigma patterns characterized primarily by heterosexism or heterosexism together with racism may be especially at risk for developing unhealthy drinking behaviors and marijuana use disorder symptoms.