ABSTRACT Introduction In response to the growing popularity of e-cigarettes globally, particularly among younger non-smokers, many countries have introduced restrictions around the flavours of e-cigarette products. In Australia, current legislation restricts e-cigarettes to only tobacco, menthol and mint but does not specify the flavouring chemicals that may impart these flavours. Methods A convenience sample of e-liquids (N = 81), pods (N = 16) and disposable e-cigarettes (N = 43) that were “tobacco”, “mint” or “menthol” flavoured were selected for analysis. Their chemical composition was determined using gas chromatography-mass spectrometry (GC-MS). Results Most tobacco flavoured samples contained ethyl maltol (86.7%), a caramellic flavoured chemical. While no flavouring chemicals known to impart a tobacco flavour were identified, the synthetic coolant WS-23 was absent or detected in lower concentrations compared to sweet e-cigarette products analysed previously. A coolant chemical (menthol, WS-23 or WS-3) was detected in all mint/menthol samples, although many also contained fruity (26.2%), vanilla (19.7%) and caramellic (19.7%) flavouring chemicals, often with no indication in the flavour name. Conclusions The composition of the tobacco flavoured samples was similar to prohibited sweeter e-cigarette products, with no common flavour chemicals identified which could be attributed to a tobacco flavour. The lack of coolants in tobacco flavoured e-cigarettes may increase their throat harshness, potentially aligning them more with the sensory experience of traditional cigarettes. While all mint/menthol samples contained coolants, fruity and sweet flavouring chemicals, characteristic of the products prohibited under the current legislation, were still detected in some samples.
ABSTRACT Introduction Abstinence outcome expectancies (AOEs), the anticipated consequences of drug cessation, predict key cessation outcomes such as withdrawal severity; however, the construct has yet to be extended to electronic nicotine delivery systems (ENDS). Methods We conducted a psychometric evaluation the Smoking Abstinence Questionnaire (SAQ) and a parallel vaping adaptation. A multigroup confirmatory factor analysis was conducted to test measurement invariance across people who smoke (n = 126) or vape (n = 193) nicotine, and construct validity was assessed by examining associations between AOE domains and a measure of vaping dependence. Results Results failed to support the original 10-factor structure of the SAQ; however, a modified 7-factor model demonstrated configural and metric invariance across nicotine product groups. Scalar invariance was not supported, as two items assessing the anticipated efficacy of nicotine replacement therapy and social benefits from quitting exhibiting noninvariant intercepts across groups. People who smoke reported higher latent mean scores for anticipated weight gain and changes in coffee consumption compared to people who vape. Further, the AOE domains demonstrated good internal consistencies and showed expected associations with vaping dependence. Conclusions Findings suggest that AOEs function similarly across cigarette and ENDS use, while identifying item-level and latent mean expectancies that differ between product groups. Implications This is the first study to extend the AOE construct to ENDS. Results suggest that many smoking-related abstinence expectancies extend to vaping, while some expectancies may differ by nicotine product.
Abstract Introduction Disability is consistently associated with higher smoking prevalence yet remains largely excluded from intersectional research. This study examined current cigarette smoking inequities among New York State (NYS) adults through an Intersectional Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy, incorporating disability as a core dimension alongside race/ethnicity, gender, age, and education. Methods This cross-sectional study analyzed a sample from the 2024 NYS Behavioral Risk Factor Surveillance System (N=38,061). Five social dimensions, including race/ethnicity, gender, age group, education, and disability status, defined 319 intersectional strata. Two-level logistic regression models estimated stratum-level variance, predicted prevalence, and interaction effects. Results Overall, 11.8% of participants reported current cigarette smoking. Approximately one-fifth of total variance in current cigarette smoking was attributable to between-strata differences, mostly explained by additive main effects, though residual variance remained. Predicted prevalence of current cigarette smoking varied across strata, ranging from 1.5% to 53.8%. The lowest-prevalence strata consisted of racial/ethnic minority college graduates without disability, while the highest-prevalence strata were predominantly non-Hispanic white or Other adults with less than high school education and disability. Ten strata exhibited significant interaction effects, departing from additive predictions. The null model median odds ratio (MOR: 2.44, 95% CI: 2.19, 2.68) revealed substantial between-strata heterogeneity, corresponding to over a twofold difference in odds when comparing two randomly chosen strata ordered by risk. Conclusions Cigarette smoking inequities in NYS were patterned by intersecting social positions, with substantial heterogeneity across intersectional strata. Disability appeared consistently among the highest-burden strata, and its combination with low education produced the greatest predicted smoking prevalence.
ABSTRACT Introduction Real-world smoking may be shaped by dynamic interactions between transient affective states and cognitive control processes. Ecological momentary assessment studies show that negative affect predicts smoking-related outcomes in daily life, however it is unclear whether momentary fluctuations in performance-based inhibitory control contribute to real-world smoking beyond affective state. Methods Adult light-to-moderate smokers (N = 77; ages 18–65) completed a 10-day ecological momentary assessment protocol on their smartphones, with three prompts per day. At each prompt, participants completed a Stop-Signal Task indexing inhibitory control and reported momentary affect on the Positive and Negative Affect Schedule; 45–105 minutes later they reported their actual cigarette consumption. Multilevel generalized linear mixed-effects models tested whether within-person fluctuations in inhibitory control and negative affect prospectively predicted subsequent smoking. The study and hypotheses were preregistered. Results Within-person increases in negative affect significantly predicted higher smoking rates at the subsequent assessment interval. Momentary inhibitory control was not significantly associated with subsequent smoking, and Bayesian model comparison favored the model without inhibitory control. Adding baseline psychological distress (Depression Anxiety Stress Scales-21) improved model fit, with higher distress predicting greater overall smoking, but it did not alter the within-person link between negative affect and smoking. Neither negative affect nor inhibitory control predicted craving. Conclusions In everyday life, transient negative affect is a more proximal predictor of smoking than momentary fluctuations in performance-based inhibitory control. More broadly, the findings highlight the value of intensive longitudinal designs for testing mechanistic models of smoking in naturalistic settings.
