Disentangling the temporal relationship between alcohol-related attitudes and heavy episodic drinking in adolescents within a randomized controlled trial
Research Report
Summary and key findings
Should alcohol prevention work focus on changing young people’s attitudes or their behaviour? A new study adds further weight to the argument for focusing on changing what young people do rather than how they think about alcohol.
The study followed pupils in 105 schools in Scotland and Northern Ireland who were taking part in an alcohol prevention programme. Over 12,000 pupils, with an average age of 12.5 years at the start, completed surveys about their attitudes and behaviours around alcohol on four occasions over a period of two and a half years.
The researchers could not see that changes in the children’s attitudes could in any way predict changes in their alcohol consumption behavior. It was more likely that the attitudes they expressed reflected their previous alcohol consumption behavior. Having consumed alcohol had a greater impact on the children’s future attitudes than the other way around.
Previous research has shown a connection between attitudes and behavior around alcohol. Those who consume more alcohol generally have a more positive attitude towards alcohol and vice versa.
It has also been shown by scientific studies that children’s attitudes change with age and are influenced by their friends’ and parents’ attitudes towards alcohol and whether they are exposed to alcohol marketing. However, according to the researchers, the link between the child’s own attitudes and behavior is not as well explored.
In the study, the researchers were able to see that the attitude a child indicated at the beginning of the study played a role in future attitudes. Most of the children did not change their behavior during the study.
However, those who tested positive for binge alcohol use also tended to continue consuming alcohol. It was also more likely that these children changed their attitudes towards alcohol than it would be possible to use past attitudes to predict who would start engaging in binge alcohol use.
Negative attitudes towards alcohol therefore only seem to set the stage for a future negative attitude towards alcohol until external factors change the young person’s behavior, according to the study authors. When behavior and attitude do not match, a cognitive dissonance arises that is easiest to resolve by changing one’s attitude, the researchers say.
This is why external factors and the possibility of changing one’s behavior should be the focus of alcohol prevention work.
For example, it was seen that students from schools with lower socio-economic status had more positive attitudes towards alcohol and were more likely to engage in binge alcohol use. Consuming alcohol was also more widespread in the Scottish schools in the study than in the Northern Irish schools. The researchers hope that the study will be a step towards developing more effective prevention programs in the future.
Abstract
Background and aims
Within many alcohol prevention interventions, changes in alcohol-related attitudes (ARA) are often proposed as precursors to changes in alcohol use behaviour. This study aimed to measure the longitudinal relationship between alcohol-related attitudes and behaviour during the implementation of a large-scale prevention trial.
Design and setting
This study was a two-arm school-based clustered randomized controlled trial. A total of 105 schools in Northern Ireland and Scotland participated in the Steps Towards Alcohol Misuse Prevention Programme (STAMPP) Trial.
Participants
A sample of 12,738 pupils (50% female; mean age = 12.5 years at baseline) self-completed questionnaires on four occasions (T1–T4). The final data sweep (T4) was 33 months post baseline.
Measurements
Individual assessments of alcohol-related attitudes and heavy episodic alcohol use were made at each time-point. Additional covariates included location, school type, school socio-economic status and intervention arm. Estimated models examined the within-individual autoregressive and cross-lagged effects between alcohol-related attitudes and heavy episodic alcohol use across the four time-points (Bayes estimator).
Findings
All autoregressive effects were statistically significant for both alcohol-related attitudes and heavy episodic alcohol use across all time-points.
Past alcohol-related attitudes predicted future alcohol-related attitudes.
Similarly, past heavy episodic alcohol use predicated future heavy episodic alcohol use.
Autoregressive effects for heavy episodic alcohol use were larger than those for alcohol-related attitudes at all time-points. In the cross-lagged effects, past heavy episodic statistically significantly predicted more positive alcohol-related attitudes in the future except for the initial T1–T2 path. In contrast, past alcohol-related attitudes did not predict future heavy episodic alcohol use across any time-points.
Conclusions
Changes in alcohol-related attitudes were not a precursor to changes in heavy episodic alcohol use within the Steps Towards Alcohol Misuse Prevention Programme (STAMPP) Trial in Scotland and Northern Ireland.
Rather, alcohol-related attitudes were more likely to reflect prior alcohol consumption status than predict future status. Heavy episodic alcohol use status appears to have a greater impact on future alcohol attitudes than attitudes do on future heavy episodic alcohol use.
Meaning of the findings
These findings are much more consistent with a cognitive dissonance and rationalization model of the relationship between alcohol-related attitudes and alcohol use behaviour than theoretical models that proposed attitudes as the precursors of behaviours.”
Percy A, Padgett RN, McKay MT, Cole JC, Burkhart G, Brennan C, et al. Disentangling the temporal relationship between alcohol-related attitudes and heavy episodic drinking in adolescents within a randomized controlled trial. Addiction. 2024. https://doi.org/10.1111/add.16721
The findings suggest a high degree of within-individual consistency over time in both adolescent self-reported alcohol-related attitudes and heavy episodic alcohol use. Prior attitudes and prior alcohol use behaviours were the main predictors of future attitudes and behaviours, respectively.
At the sample level, mean alcohol-related attitudes remained relatively constant over time, although this undoubtedly masks a degree of individual variation. As participants aged, the influence of prior attitudes on future attitudes increased in line with most psychological theories of learning.
While there is inherent growth in the rate of heavy episodic alcohol use over time, most students maintained the same heavy episodic alcohol use status during the 4 years of the study (i.e. remained within the non-heavy episodic alcohol use group). Even when students made a transition into heavy episodic alcohol use, they were also likely to remain consistent in this newly acquired behaviour in future. This long-term consistency in adolescent alcohol use is well documented.
Even among emergent adolescent alcohol users, alcohol consumption can be thought of as both a state and a trait, where the stable trait for higher risk alcohol use is largely determined by dispositional decision-making characteristics, while state fluctuations in alcohol consumption were largely driven by more situational decision-making factors (e.g. peers, access and location). The trait-like stability of alcohol consumption is generally maintained into and throughout adulthood.
When the researchers examined the reciprocal relationships between alcohol-related attitudes and heavy episodic alcohol use, it was past heavy episodic alcohol use that shaped future attitudes rather than attitudes driving behaviours.
Within adolescents, alcohol-related attitudes appear to lag heavy episodic alcohol use. This lag has also been observed in prior alcohol and smoking studies. These findings are much more consistent with a cognitive dissonance and rationalization model of the relationship between alcohol-related attitudes and alcohol use behaviour than theoretical models that proposed attitudes as the precursors of behaviours.
With non-alcohol consumers, it is likely that alcohol use status remains highly consistent with alcohol-related attitudes until the transition to heavy episodic alcohol use, when attitudes then evolve to support the newly acquired alcohol use status (i.e. become more pro-alcohol).
Therefore, negative attitudes are likely to support continued negative attitudes towards alcohol (via the autoregressive pathway), until situational factors create an environment conducive to a change in alcohol use status (engaging in heavy episodic alcohol use).
The resultant cognitive dissonance arising from the mismatch between alcohol-related attitudes and heavy episodic alcohol use appears more likely to be resolved by adopting more pro-alcohol attitudes than desisting from future heavy episodic alcohol use (as binge alcohol use demonstrates strong autoregressive effects within the model).
Any causal link between these two processes appears predominantly in the direction of alcohol use behaviour impacting upon alcohol-related attitudes, and not the reverse.
This finding is in stark contrast to the body of existing research supporting models such as the Theory of Planned Behaviour that propose causal attitudes to behaviour relationships. However, the majority of studies testing models such as the Theory of Planned Behaviour have failed to include reciprocal paths for the influence of behaviour on changes in attitudes and intentions, as was modelled in this study. These two processes are developmentally intertwined, but prior research has tended to view the inter-relationship from one direction only, with alcohol-related attitudes the predictor and alcohol use behaviour the outcome. Permitting the relationship between alcohol-related attitudes and behaviour to be cross-lagged appears to fundamentally change the observed relationships, where alcohol use behaviour is the predictor of alcohol-related attitudes.
It is worth noting that, in addition to the observed autoregressive and cross-lagged effects, alcohol-related attitudes and heavy episodic alcohol use were both shaped by the wider school environment inhabited by young people. In particular, a higher level of poverty within the school (as indexed by the proportion of pupils on free school meals) was linked with more positive alcohol-related attitudes and higher heavy episodic alcohol use.
This relationship between school-level SES and alcohol consumption has been noted in some but not all research. School-level heavy episodic alcohol use was also higher within schools in Scotland and within mixed-sex schools. These effects may be due in part to the peer alcohol use norms established within certain types of school.
Given the literature described in the Introduction, it is unsurprising that reducing positive attitudes towards alcohol is a significant component in many alcohol prevention interventions aimed at adolescents. However, as evidenced in this study, the relationships between alcohol-related attitudes and alcohol-related behavioural outcomes are complex.
Evaluations of many alcohol education programmes have shown larger impacts upon attitudes than on alcohol use behaviours. However, studies undertaking mediation analysis of trial outcomes have failed to show alcohol-related attitudes as a significant mediator of intervention effects on heavy episodic alcohol use or life-time alcohol consumption.
While it is hard to see future prevention interventions not attempting to reduce positive alcohol-related attitudes among pupils, together with attempting to reduce their engagement in heavy episodic alcohol use and subsequent exposure to alcohol-related harms, there is little evidence presented here to support alcohol-related attitudes being considered as either a relevant secondary outcome (together with primary consumption outcomes such as binge alcohol use) within an evaluation trial or as a key mediator of behavioural outcomes within an intervention logic model.
The study findings have significant implications for prevention theory, and also the design and evaluation of alcohol prevention interventions. They challenge the commonly held belief that attitudes towards alcohol drive alcohol use behaviours. On the contrary, the results of this study suggest that attitudes are much more of a product of behaviour than vice versa. Interventions aiming to prevent adolescent risk-taking through promoting negative attitudes towards the behaviour may be misguided. Future theory and practice should reflect the observed temporal relationship between alcohol attitudes and alcohol use behaviour.