If you are among those people who think that the association between anxiety and alcohol use disorders is old stuff, then brace yourself as what comes next is likely to change your mind.
I could start by emphasising the severe consequences that these conditions may lead to in terms of both health and socio-economic expenses (Hoffman D. L. et al., 2018; Bouchery E. E. et al., 2011), but references exist for this reason. Instead, I will focus on highlighting the inconsistencies of the evidence.
The need for a systematic review is very much justified when one considers the many (and rather conflicting) theories on whether anxiety precedes, thus causes, alcohol use or vice versa. For example, research suggests that anxiety triggers alcohol consumption as a form of alleviation (self-medication hypothesis; Khantzian, E. J., 1990), whereas other evidence shows that anxiety acts as a protective factor against alcohol consumption (Pardini D. et al., 2007).
It is fair to assume that there is still no general agreement on the directionality of this association. That is why this recent systematic review (Dyer M. L. et al. 2019) focused on investigating the relationship between anxiety and alcohol use. The paper also includes a meta-analysis which helped set the bases to identify those individuals who are more likely to engage in frequent and heavy alcohol consumption.
For the purpose of this review, both PRISMA and MOOSE guidelines were used. While the former is usually followed when reporting randomised trials, it is a helpful tool to report any research, due to it being methodical and comprehensive; the latter, however, is more appropriate when reporting observational studies. By taking advantage of the features of both tools, the authors could rely on a more robust methodology.
The inclusion criteria for the prospective studies were:
- English language peer-reviewed publication;
- Anxiety exposure measured either in children (<10 years) or adolescents (10 to <18 years);
- Alcohol outcome(s) measured at least six months later than exposure;
- Longitudinal design.
Exclusion criteria applied to studies where post-traumatic stress disorder was present, alcohol initiation was the only outcome, and statistical analyses were inappropriate for the purpose of the review.
PubMed, Scopus, Web of Science, and PsycINFO were used to identify the studies; relevant study characteristics were extracted and a quality assessment was performed.
Because of the heterogeneity of the studies, the authors grouped exposure and outcomes in the following way:
- Alcohol use categories: drinking frequency/quantity, binge drinking, and alcohol use disorder (AUD);
- Anxiety categories: generalised anxiety disorder (GAD), internalising disorders, obsessive compulsive disorder (OCD), panic disorder, separation anxiety disorder, social anxiety disorder, specific phobias, and miscellaneous anxiety categories.
A total of 51 studies made it into the review, while the meta-analysis included only 3 studies.
The narrative synthesis revealed 97 associations with anxiety, which I have summarised in the following table:
|Number of associations||Positive associations||Negative associations||Equivocal association||Unclassifiable associations|
|Alcohol consumption (collectively)||97 (total)||32 (33%)||17 (18%)||25 (26%)||23 (24%)|
|Drinking frequency/ quantity||37||9 (25%)||8 (21%)||9 (24%)||11 (30%)|
|Binge drinking||14||3 (21%)||4 (28%)||2 (14%)||5 (36%)|
|Alcohol use disorders (AUD)||46||20 (44%)||5 (11%)||14 (30%)||7 (15%)|
Both the narrative synthesis and meta-analyses confirmed that no positive associations were found between generalised anxiety disorder (GAD) and alcohol use disorder (AUD). Conversely, there was some evidence for a positive relationship with AUD and all the remaining anxiety categories.
The results showed no negative associations between collective alcohol consumption and OCD, panic disorder, separation anxiety and specific phobias, while there were mixed positive and negative associations for miscellaneous anxiety, social anxiety, GAD, and internalising disorders.
Finally, both drinking frequency/quantity and binge drinking were inconsistently associated to anxiety.
- Although there seems to be some evidence that anxiety in childhood and adolescence is prospectively associated with later alcohol use disorder (AUD), the same is not true for generalised anxiety disorder (GAD).
- The evidence for drinking frequency/quantity and binge drinking is, for now, contradictory.
Strengths and limitations
Despite the lack of a conclusive answer to the question “does anxiety precede the development of alcohol use?”, I personally believe in the power of systematic reviews to provide an excellent (and often much needed) overview of a specific research question. This is particularly true when we consider that this is the largest systematic review ever done on the prospective associations between anxiety and alcohol use, suggesting that we can rely on its comprehensiveness.
Though large, one may wonder whether its methodology was appropriate. Among all the limitations already acknowledged by the authors on the paper, i.e. English language limitations, presence of unclassifiable associations, exclusion of studies with more intricate statistical models, impossibility of inferring causality, low power due to small samples in some studies, and use of broad measures of internalising behaviour (e.g. depression vs anxiety), the latter is the one that, in my opinion, may have compromised the validity of the paper most due to measurement bias. However, the authors addressed this limitation by re-running the analyses excluding internalising associations, which resulted in overall unchanged findings.
Implications for practice
Among the listed future research directions, I was astonished to read that, while anxiety seems to be positively associated with alcohol use disorder (AUD), there was not enough evidence to claim the same about drinking frequency/quantity and binge drinking. This becomes especially relevant when we think about the existence of a continuity ranging from binge drinking to AUD. A systematic review investigated whether alcohol consumption can lead to subsequent alcohol problems and stressed the need for high-quality prospective cohort studies, which are likely to result in better preventive strategies (McCambridge J. et al., 2011). With this in mind, it is imperative that future studies examining this association use an adequate sample, adjust for potential variables that may affect the association, and measure anxiety and alcohol use in a consistent way.
Furthermore, the authors emphasise the multi-dimensionality of mental health disorders, particularly regarding different anxiety symptoms and potential moderators. Specifically, the presence of genetic and environmental variables makes it even harder to explore the relationship between exposures and outcomes. But how can researchers be consistent while trying to be as inclusive as possible at the same time? This thought gave me a headache for quite a long time, and I still cannot find a suitable answer. However, what I do know is that the best currently known way to get a grasp of all the research out there on a specific topic is to rely on systematic reviews, so I think it makes sense to start from here.
Conflicts of interest
Dyer, M. L., Easey, K. E., Heron, J., Hickman, M., & Munafò, M. R. (2019). Associations of child and adolescent anxiety with later alcohol use and disorders: a systematic review and meta-analysis of prospective cohort studies (PDF). Addiction, 114(6), 968-982. doi: 10.1111/add.14575
Hoffman, D. L., Dukes, E. M., & Wittchen, H. U. (2008). Human and economic burden of generalized anxiety disorder (PDF). Depression and anxiety, 25(1), 72-90. doi: 10.1002/da.20257
Bouchery, E. E., Harwood, H. J., Sacks, J. J., Simon, C. J., & Brewer, R. D. (2011). Economic costs of excessive alcohol consumption in the US, 2006 (PDF). American journal of preventive medicine, 41(5), 516-524. doi: 10.1016/j.amepre.2011.06.045
Khantzian, E. J. (1990). Self-regulation and self-medication factors in alcoholism and the addictions. Similarities and differences. Recent developments in alcoholism: an official publication of the American Medical Society on Alcoholism, the Research Society on Alcoholism, and the National Council on Alcoholism, 8, 255-271. [Europe PCM Abstract] Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/2185521
Pardini, D., White, H. R., & Stouthamer-Loeber, M. (2007). Early Adolescent Psychopathology as a Predictor of Alcohol Use Disorders by Young Adulthood (PDF). Drug and alcohol dependence, 88, S38-S49. doi: 10.1016/j.drugalcdep.2006.12.014
Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) website: http://www.prisma-statement.org/
Stroup, D. F., Berlin, J. A., Morton, S. C., Olkin, I., Williamson, G. D., Rennie, D., … & Thacker, S. B. (2000). Meta-analysis of observational studies in epidemiology: a proposal for reporting. Jama, 283(15), 2008-2012. doi: 10.1001/jama.283.15.2008
McCambridge J, McAlaney J, Rowe R. Adult consequences of late adolescent alcohol consumption: a systematic review of cohort studies (PDF). PLoS Med 2011;8:e1000413. https://doi.org/10.1371/journal.pmed.1000413