Vaping-Specific Nicotine Dependence Across Days Among a National Sample of US Young Adults Who Vape

Abstract

Introduction

Vaping has increased among young adults (YAs), and nicotine dependence prevalence has increased among YAs who vape, in the US. Research on nicotine dependence symptoms (NDS) among YAs who vape remains limited, and accurately measuring NDS and their severity remains a challenge. To date, no research has considered day-to-day NDS, as previous research focuses on retrospective measures.

Aims and Methods

Daily data came from the Monitoring the Future Vaping Supplement; out of 1244 YAs, we analyzed data from 150 (M_age = 19 [SE = 0.44]; 71.2% non-Hispanic white; 57.8% male) who vaped nicotine at least once during 14 daily surveys (n = 882 nicotine vaping days). The vaping-specific Hooked on Nicotine Checklist measured NDS on nicotine vaping days. Dichotomous and count measures of NDS were used at the day level and aggregated to the person level. Descriptive analyses and multilevel regression models were used. Weights ensured generalizability and adjusted for attrition.

Results

Any NDS were reported on 64.5% of nicotine vaping days (average 1.93 symptoms); 43.3% of nicotine vaping days had more than 1 NDS. Results across persons (rather than days) showed any NDS on 53.8% of nicotine vaping days, and 45.4% YAs reported between 1 and 2 symptoms per vaping day; 23.3% reported 2 + symptoms on average. Multilevel models showed nontrivial variance across days and persons in NDS, and regular vaping at baseline predicted NDS.

Conclusions

Vaping-specific NDS varied across days and persons for any NDS and NDS severity. Future research should consider daily NDS fluctuations, identifying factors that confer risk to inform NDS prevention and treatment among YAs who vape.

Implications

NDS and its severity varied across days and persons, demonstrating the utility of daily data. The results were generally similar for experiencing any NDS and NDS severity, but there was slightly more variability across days for any NDS. Intervention and treatment strategies could utilize approaches that identify and address daily symptoms to help curb nicotine dependence. Future research and treatment efforts should also consider the immediate contexts and potential factors that explain particularly elevated NDS on a given day.

Introduction

Recent trends show that not only is vaping prevalence and frequency dramatically increasing among young adults (YAs) in the United States,^1,2 but also nicotine dependence prevalence and severity also show increases among those who vape.^2,3 Nicotine dependence is a key mechanism for continued vaping, which increases the risk for combustible cigarette use, other substance use, and exposure to potentially harmful toxins.^4–6 However, research on nicotine dependence among YAs who vape (YAV) is limited,^3,7,8 and accurately measuring and examining nicotine dependence and its severity for tobacco products such as vaping remains a challenge.⁹ Research on nicotine dependence among YAV is a public health priority, as it can inform both future research and practice focused on the YA vaping epidemic.

Following an emerging area of research that uses daily data to assess substance use such as vaping,^4,10,11 the current study examines daily reports of nicotine dependence symptoms (NDS) over a 2-wk period. NDS variability is not typically assessed across days, as most measures of NDS, and research on vaping in general, rely on retrospective reports of longer time intervals (eg, past-30 symptoms). Limited research on substances other than nicotine has focused on specific symptoms of dependence, such as craving; this research shows craving can vary across days, and this daily variability is consequential for substance use and health outcomes.^12,13 One study of 115 adults recruited via convenience sampling showed that lagged nicotine craving and dependence across days predicted later nicotine product use, regardless of nicotine product.¹¹ However, no extant research focuses on the examination of nicotine vaping and any NDS endorsement using daily reports among a national sample of YAs in the United States. More static measures of NDS (eg, past 30-day) are at the risk of inadequately capturing the dynamics of NDS, limiting our understanding of how best to evaluate and intervene to address vaping-specific NDS and its severity, especially among YAs.

This study addresses these gaps in research by examining vaping-specific NDS among YAV. Using a national sample of YAV in the United States, this study aimed to examine the distribution and variance of vaping-specific NDS across nicotine vaping days and persons using two specific research aims: research aim 1 (RA1) focused on the presence of any vaping-specific NDS, and research aim 2 (RA2) focused on NDS severity. For both RAs 1 and 2, we first assess the distribution of NDS across days and persons; we then examine variance in NDS at the day and person level and test whether demographic characteristics and vaping at baseline were associated with NDS frequency.

Materials and Methods

Participants and Procedure

Data came from the Monitoring the Future (MTF) Vaping Supplement (VS) collected from September 2020 to November 2020.^14,15 Participants were selected from a nationally representative sample of US 12th-grade students participating in MTF in spring 2019.¹⁶ In 2019, there were 13 713 MTF participants in 12th grade, and 7850 were eligible for the VS. Individuals were ineligible if they had been randomly selected for the MTF panel study¹⁷ or they did not provide valid contact information. Participants who reported vaping or other substance use in 12th grade were oversampled. Subsequently, 4358 eligible participants were randomly selected to participate in the MTF VS.^14,15 An incentive was provided for the completion of the baseline survey ($25) and up to an additional $60 for participation in the daily survey (participants were compensated based on the number of completed daily surveys). All participants provided consent, and the study was approved by the University of Michigan Institutional Review Board.

The MTF VS included two data collection phases. First, the MTF VS surveyed individuals about 1 year after the 12th-grade MTF survey (M = 19.5 years, SD = .44). Out of 4358 selected individuals, a total of 1244 participated in the 1-year follow-up survey. Second, participants who reported current nicotine or cannabis vaping at follow-up (N = 427; 34.3% of follow-up sample) were invited to complete 14 consecutive daily web-based surveys. While participants could have reported vaping other substances such as cannabis, to be included in this analysis, respondents had to report past 30-day nicotine vaping at follow-up (n = 309), participate in the daily surveys (n = 241), and report nicotine vaping at least once during the 14-day period (n = 150). Among participants in the 1-year follow-up survey, 935 did not report current nicotine vaping, 68 did not participate in the daily surveys after invitation, and 91 daily-survey participants did not report vaping during the daily surveys. Participants provided an average of 11 (of 14) days of data (SE = 0.51). Among the analytic sample, over half (54.2%) completed all 14 days, and 79.4% provided 7 or more days of data. Pair-wise deletion was used for any missing data (<5% missing). The data for this study are not publicly available, but code and syntax are available upon request.

Measures

Nicotine Dependence Symptoms

Nicotine vaping was assessed each day by asking participants, “Did you do any of the following on this day?,” and one response option was “use nicotine or tobacco.” The follow-up questions asked: “How did you use nicotine or tobacco?,” with a response option of “vaped nicotine.” Participants were asked about nicotine dependence on each day they vaped nicotine, using the six items from the Hooked on Nicotine Checklist tailored to vaping on a given day^7,18,19: Pertaining to the particular nicotine-vaping day, (1) “did you feel like you really needed to vape?,” (2) “did you have a strong urge to vape?”; and “when you had not vaped…” (3) “did you find it difficult to concentrate?,” (4) “feel more irritable?,” and “(5) feel nervous, restless or anxious?”; and finally, (6) “did you vape within 30 minutes of waking up in the morning?”. Each response option was yes/no.

A dichotomized measure of NDS was used to capture the presence/absence of any NDS on a given nicotine vaping day (1 = 1 symptom or more, 0 = no symptoms). According to theory and psychometric properties, a dichotomous measure of NDS accurately captures “loss of autonomy” over nicotine use.¹⁹ For RA2, a count measure of NDS was used (0–6) to assess NDS severity.¹⁹

Demographic and Baseline Vaping Measures

Demographic characteristics included race/ethnicity (non-Hispanic white, non-Hispanic Black, Hispanic, Asian, and another race/ethnicity), and biological sex (male, female). Vaping frequency at baseline captured regular vaping, measured by the number of days YAs vaped nicotine in the past 30 days (>5 days, ≤5 days).²⁰ Additional measures included binary (yes/no) indicators of other tobacco use (inclusive of cigarettes and every other product provided) and cannabis vaping across days.

Analytic Strategy

For RA1, descriptive analyses assessed the distribution of any NDS across nicotine vaping days and persons. For person-level descriptive analyses, NDS were aggregated from daily surveys, capturing the frequency of any NDS on nicotine vaping days among YAV. We then conducted two weighted multilevel regression models²¹: (1) a null model that assessed the intraclass correlation coefficient (ICC), capturing the percent of variance in any NDS across nicotine vaping days relative to persons, and (2) a model that tested whether race/ethnicity, sex, and frequency of nicotine vaping reported at baseline were associated with any NDS across nicotine vaping days. For RA2, identical descriptive analyses and multilevel models were used; however, the outcome was NDS severity. Logistic and negative binomial multilevel models were used for the respective outcomes. Stata v.17 was used for all analyses.

Weights were used to adjust for the complex survey design of the MTF VS study,¹⁴ oversampling based on vaping and other substance use, and survey nonresponse. Weights allowed estimates to be representative of the national 12th-grade MTF sample. Supplemental analyses were conducted to determine how the number of nicotine vaping days was associated with reported NDS across nicotine vaping days, and whether results changed after removing nicotine vaping days that also involved (1) other tobacco use and (2) cannabis vaping.

Results

The analytic sample was 71.2% non-Hispanic white, 1.9% Black, 9.5% Hispanic, 2.4% Asian, and 15.0% reported another race/ethnicity, and 57.8% was male. At the day level, 37.7% of days involved nicotine vaping, and 2.7% of nicotine vaping days included other tobacco use and 15.5% included cannabis vaping. Any NDS was reported on 64.5% of nicotine vaping days (SE = 2.3). The average frequency of nicotine vaping days with an NDS at the person level was 53.8% (SE = 5.3); most frequent percentages were 0% and 100% (31.3% and 31.8% of sample, respectively), as shown in Supplementary Figure 1. The ICC from multilevel models was 0.83, indicating 17% of variance in any NDS on a given nicotine vaping day existed across days relative to persons. Shown in Table 1, Black YAs had lower log-odds of experiencing an NDS (log-odds = −3.72, SE = 1.55; p = .017) and Hispanic YAs had higher log-odds (log-odds = 2.66, SE = 0.09; p = .006), compared with non-Hispanic white YAs. Regular vaping at baseline was positively and strongly associated with experiencing an NDS on a given nicotine vaping day (log-odds = 4.54, SE = 1.03; p < .001). YAs reporting vaping more than 5 days had an NDS on 60.7% of nicotine vaping days compared with only 16.7% for YAs who reported less frequent vaping.

Table 1.

Multilevel Regression Models for Any and Number of Nicotine Dependence Symptoms (NDS)

		Any NDS		Number of NDS
Variables		Coef.	(SE)	Coef.	(SE)
Race/ethnicity
Non-Hispanic White (reference)
Non-Hispanic Black		−3.72	(1.55)	−1.64	(0.87)
Hispanic		2.66	(0.97)	0.77	(0.42)
Asian		−2.65	(1.57)	−1.48	(1.03)
Another race/ethnicity		0.11	(0.90)	0.11	(0.52)
Sex
Female (reference)
Male		−0.65	(0.71)	−0.11	(0.36)
Regular vaping at basline (vs. not regular)		4.54	(1.03)	1.82	(0.39)

Notes. Days N = 878; Individual N = 148; weights were used for all estimates.

All statistically significant (p < .05) relationships are bolded.

Multilevel logistic regression models were used for any NDS, and multilevel negative binomial regression models were used for number of NDS.

For multilevel logistic regression models, coefficients represent log-odds. For multilevel negative binomial regression models, coefficients represent log of expected count.

Results for RA2 showed that the mean number of NDS across nicotine vaping days was 1.93 (SE = 0.12). Even though experiencing only 1 NDS was common (20.8% of nicotine vaping days), 43.7% of nicotine vaping days had more than 1 NDS. The number of symptoms above one were relatively evenly distributed across nicotine vaping days (2 = 11.6%, 3 = 9.2%, 4 = 6.2%, 5 = 3.9%, and 6 = 12.9%). At the person level, YAs experienced 1.48 (SE = 0.24) symptoms per nicotine vaping day on average. Supplementary Figure 1 shows that most YAs either reported no NDS (31.3%) or between 1 and 2 symptoms (45.4%) per day; 23.3% reported more than 2 symptoms on average. ICC for the NDS count measure was 0.94, indicating 6% of variance in the number of NDS experienced on a given nicotine vaping day existed across days relative to persons. There was a markedly higher count of NDS for regular vaping at baseline compared with less frequent vaping (coeff. = 1.82, SE = 0.39; p < .001).

Supplemental analyses showed that there was an average of 6.5 nicotine vaping days per person (SE = 0.66). The two most frequently reported number of nicotine vaping days were 1 and 14 (mean = 60.9%; range = 7.1%–100.0%; median = 50%). Since most participants provided 14 days of data, no significant association was found between days with data and NDS outcomes. There was a significant association between number of nicotine vaping days and NDS frequency (coeff. = 0.03, p = .002) and severity (coeff. = 0.13, p = .008) across persons. YAs with more nicotine vaping days reported a higher frequency and severity of NDS. Results generally remained unchanged after removing nicotine vaping days that included other tobacco use and days that included other tobacco use or cannabis vaping.

Discussion

Despite the prevalence of vaping among YAs in the United States,^1,2 there remains a significant gap in our understanding of vaping-specific nicotine dependence.^3,7,8 In particular, how NDS are experienced across days is unknown. Using a national sample of YAs in the United States, we found that reporting at least 1 NDS was common across nicotine vaping days and YAs, and a clear bimodal distribution was found, indicating YAs who vape nicotine tend to report NDS either frequently or not at all. There was nontrivial variance in experiencing an NDS across nicotine vaping days, but more variance existed across persons than days. This was also the case for NDS severity, but more variance across persons was observed for NDS severity compared with any NDS. It is important to note that there was noticeable variability in NDS severity in terms of nicotine vaping days with 2 or more symptoms of dependence (and persons who experienced 2 + symptoms). In sum, YAs in the United States who vape nicotine are experiencing symptoms of nicotine dependence, such as withdrawal and craving, and the presence and severity of these symptoms varies across persons and days. A greater understanding of what explains particularly high levels of dependence is needed for intervention and treatment.

More static measures that retrospectively assess longer time periods (eg, past 30 days) do not capture the dynamic nature of NDS at the daily level. Both frequency and severity of NDS should be considered a process that varies across time periods as short as days, and findings support the importance of assessment timing^11–13 when it pertains to evaluating vaping-specific nicotine dependence among YAs.

Compared with non-Hispanic white YAs, Hispanic YAs had a higher frequency and Black YAs had a lower frequency of NDS, while no differences across demographics were found for NDS severity. Current literature on NDS and vaping among Hispanic and Black YAs in the United States is limited, and further research is needed to replicate findings and monitor NDS among different races/ethnicities. Vaping frequency at baseline exhibited important differences, as YAs who reported vaping regularly exhibited a higher frequency and severity of NDS across vaping days.

Research, prevention, and treatment efforts should consider NDS across days and individuals. Intervention and treatment strategies could utilize approaches that identify and address risk and protective factors that explain variations in nicotine dependence to support treatment. Future research and treatment efforts should also consider the immediate contexts and factors that are associated with daily fluctuations in NDS. It is possible certain vaping devices, nicotine levels, or other factors such as social contexts may influence NDS across days.

There were limitations to the current study. Participants were recruited from high schools (those absent or who dropped out were not included), and attrition occurred during the study. However, weights were used to adjust for attrition and to ensure the generalizability of estimates. NDS was only examined on vaping days, but future research should consider NDS experiences on nonvaping days, as well as how symptoms vary within a day.

Conclusion

This research addressed the need to better understand vaping-specific NDS by assessing daily measures of NDS among YAs in the United States. The presence and severity of NDS varies across days and persons, and these findings have implications for research and practice. Identifying ways to better evaluate NDS is important for prevention, interventions, and future research focused on the YA vaping epidemic. Understanding daily fluctuations in NDS could inform strategies that target immediate indicators of addiction. Future research should include the acknowledgment of the variability in NDS symptoms over time in order to identify salient intervention targets.

Supplementary material

Supplementary material is available at Nicotine and Tobacco Research online.

Author contributions

Michael Parks (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Investigation [equal], Methodology [equal], Validation [equal], Visualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), Brooke Arterberry (Investigation [equal], Methodology [equal], Validation [equal], Writing—original draft [equal], Writing—review & editing [equal]), and Megan Patrick (Conceptualization [equal], Data curation [equal], Funding acquisition [equal], Investigation [equal], Project administration [equal], Resources [equal], Supervision [equal], Validation [equal], Writing—original draft [equal], Writing—review & editing [equal])

Declaration of Interests

The authors declare no conflicts of interest.

Funding

Data collection and manuscript preparation were supported by research grants from the National Institute on Drug Abuse (R01DA001411 and R01DA016575). The study sponsors had no role in the study design, collection, analysis or interpretation of the data, writing of the manuscript, or the decision to submit the paper for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the study sponsor.

Data Availability

The data from the Monitoring the Future Vaping Supplement study are not publicly available. Stata syntax and code are available upon request.