Commentary on Pennay <i>et al</i> . (2015): Measuring the effects of alcohol mixed with energy drinks on intoxication with real‐world data

The experimental literature has shown that energy drinks enhance stimulation and lessen the sedating effects of alcohol 1, 2, but they do not appear to counteract the impairing effects of alcohol on psychomotor performance 1, 3. With evidence that people who consume alcohol mixed with energy drinks (AmEDs) tend to underestimate how intoxicated they really are 1, 4, there is growing concern among alcohol researchers and public health advocates that AmED consumers are drinking alcohol at a faster pace and over a longer period of time, resulting in higher levels of intoxication and a greater likelihood of alcohol-related harm. However, given ethical issues in experimental research related to administering alcohol and caffeine in quantities that mirror actual use in real-world drinking situations [see 5], we rely upon non-experimental research methods to illuminate the AmED–intoxication link. Building on work by Thombs et al. 6, the paper by Pennay and colleagues 7 is one of the first to examine AmED consumption in social drinking settings using objective measures of intoxication [i.e. breath alcohol concentration (BrAC)]. They found a significant positive association between AmED use and BrAC. Unsurprisingly, however, this association became non-significant when the number of drinks and other variables were taken into account. Thus, a simple interpretation is that people who consume AmEDs end up drinking more alcohol, possibly mediated by the longer duration of drinking made possible by the stimulating effect of energy drinks. The paper adds an interesting twist, however, with analysis of the interaction between AmED use and duration of drinking in relation to BrAC. Among those with shorter drinking durations, the consumption of AmED was linked to higher BrAC, whereas among those with longer drinking durations this association was no longer present. The interpretation of this interaction is unclear, however, because the time at which AmEDs were consumed is unknown. For example, if AmEDs were consumed early in the evening, these results could reflect higher consumption early in the evening in the AmED group compared with the non-AmED group, with a reduction in drinking as the effects of AmEDs wore off. Additionally, although time of recruitment was included in the regression model, it may not account fully for potential confounding with drinking duration, as it was a dichotomous measure (i.e. before versus after midnight), while the recruitment period ranged from 9 p.m. to 5 a.m. These interesting analyses highlight an important question. That is, how do energy drinks influence the blood alcohol concentration (BAC) curve depending on timing and dosage of both energy drinks and alcohol, taking into consideration the duration of the drinking episode? Perhaps the consumption of AmEDs early in the evening will have little influence on BrAC recorded at the end of the night, but its use later in the evening will have a fairly substantial effect. As with most non-experimental studies, Pennay and colleagues did not collect detailed information about the timing of AmED use and alcohol-only use. However, young people's use of substances, including the combination of different substances, appears to be quite carefully timed. For example, Hunt et al. 8 noted the careful sequencing of different drugs in order to gain or sustain ‘a certain type of intoxication’ (pp. 512) among polydrug users, all within the context of a ‘pursuit of hedonism’. Although Hunt's work focused on the use of illicit drugs, their finding of planned and strategic intoxication appears to parallel hedonistic motivations for AmED use 9, with qualitative evidence showing that young people consume AmEDs in the early part of the night as an initial ‘boost’ and then again later on when they begin to get tired 10. To capture clearly these patterns of use with real-world data and thereby map out more accurately the effects of energy drinks on the BAC curve, more precise data on the timing and dosage of AmEDs and alcohol are needed. While survey approaches such as diary methods are useful for capturing event-level associations 11, the application of novel techniques such as those used by Kuntsche and colleagues 12, where hourly data are collected in real time via cellphones, will reduce inherent recall problems in retrospective accounts 13 and provide data on timing and level of use. Such approaches might be augmented with new mobile telephone applications such as standard drink calculators (capturing exact drink sizes, product names, etc.), as well as applications for estimating BAC. Although the validity of these applications will need to be developed and tested further, given that the majority of young adults have smartphones 14, data collection, as well as interventions 15, employing such devices may be the way forward. With hourly assessments of AmED and alcohol use it would be possible to determine whether or not energy drink consumption early in the evening promotes later drinking (i.e. consistent with evidence of a priming effect shown in the laboratory 16), or whether heavier use and lengthy drinking duration promotes subsequent energy drink use 10. Assessments of subjective intoxication could also be collected and tested in relation to actual or estimated BACs, confirming laboratory evidence that consumption of AmEDs contributes to underestimates of actual intoxication 1. Moreover, a within-subjects design with data collected within individuals across multiple drinking occasions would allow for a more powerful analysis, controlling for unmeasured individual characteristics, such as sensation-seeking and risk-taking propensity, which have been found to be related consistently to AmED use 17-19. With these and other novel approaches, we may achieve more accurate estimates of the link between AmED use and intoxication with real-world data. I am grateful to Kathryn Graham for her helpful editorial suggestions. None.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.