Anxiety and alcohol: a relationship that may run both ways

Relief that can complicate the problem

For some people, drinking seems to reduce tension, inhibition or social fear for a while. This experience can turn alcohol into a rapid response to a difficult emotion: before a gathering, at the end of an exhausting day or when the mind will not stop anticipating danger.

Immediate relief, however, is not the whole story. As the effects wear off, disturbed sleep, restlessness and physical symptoms resembling anxiety may appear. Repeated heavy drinking, and withdrawal in someone who is dependent, can intensify those symptoms. A cycle can then develop: the person drinks to reduce distress, alcohol contributes to its return, and the need for relief reinforces further drinking.

This cycle occurs, but it does not explain every case. Some people experienced anxiety long before their drinking became problematic. In others, the symptoms appeared later. Family, social, traumatic and biological factors may also increase the likelihood of both problems. Research published in Neuropsychopharmacology on 13 July 2026 investigated that shared biological basis.

What the new study examined

The researchers used results from large genome-wide association studies. The multi-ancestry anxiety dataset included 620,190 people, of whom 84,577 were cases, while the problematic alcohol use dataset included more than one million participants.

'Problematic alcohol use' was a broad research category. It combined data on alcohol use disorder or dependence with scores measuring problematic behaviours and consequences. It did not describe everybody who drinks alcohol and was not a substitute for clinical assessment.

The team used several methods to estimate the overlap between genetic signals, identify shared regions and investigate possible influence in both directions. It also examined the tissues and cell types in which genes associated with those signals are expressed.

Substantial, but incomplete, overlap

The analysis estimated a moderate genetic correlation of 0.44. In the European-ancestry sample, approximately 52.5% of the estimated potential causal variants for the two traits were shared. Most of the shared variants pointed towards effects in the same direction, although some pointed in different directions.

This means that part of the genetic vulnerability associated with one problem is also related to the other. It does not mean that half of an individual's anxiety is alcohol-related or that 52.5% of people with anxiety will develop problematic alcohol use. The figures describe statistical overlap between sets of variants within datasets. They cannot be transferred directly to an individual.

The shared regions included genes involved in dopamine and cyclic AMP signalling pathways. Gene-expression analyses highlighted several tissues and brain regions, including the prefrontal cortex, anterior cingulate cortex, hippocampus, hypothalamus and amygdala. These findings help researchers develop biological hypotheses; they do not identify a single area responsible for anxiety and alcohol use.

What a bidirectional relationship means

The study used Mendelian randomisation, a technique that employs genetic variants to investigate whether an association is consistent with a possible causal influence. The analyses found signals in both directions: vulnerability to anxiety may be related to problematic alcohol use, while vulnerability to problematic use may be related to anxiety.

The word causal requires considerable caution here. Mendelian randomisation can come closer to a causal question than an ordinary correlation, but it depends on statistical assumptions. The variants may influence other traits, the categories under study are broad, and some of the overlap may reflect a general vulnerability to internalising or psychiatric problems.

The study therefore supports a reciprocal model, not an inevitable sequence. Pre-existing anxiety may dominate for one person, the effects of drinking for another, and shared factors may influence both in someone else.

Genes do not determine the outcome

A genetic association is not an individual prediction. Thousands of variants may make very small contributions, and their expression is combined with life experience, alcohol availability, stress, learning, social support and access to treatment.

Nor do the findings identify a medicine that can now be prescribed. The study located genes already targeted by existing drugs, but a gene–drug interaction does not show that the drug will treat co-occurring anxiety and problematic alcohol use. Some compounds could even worsen anxiety.

The immediate value of the research is different: it reinforces the need to ask about both problems. When somebody seeks help for anxiety, the role of alcohol is worth exploring. When they seek help for drinking, it is important to assess anxiety symptoms that may have existed beforehand or occur between drinking episodes.

Why reconstructing the timeline matters

It is not always possible to identify the starting point in a brief appointment. It can help to reconstruct when anxiety began, when drinking changed, what happens during periods with less alcohol and whether withdrawal symptoms occur.

This information may alter care. Anxiety that pre-dated drinking may need treatment in its own right. Symptoms that emerge when alcohol is reduced may require medical assessment, particularly if dependence is present: abruptly stopping can be dangerous and should not be attempted without professional advice.

When both conditions are present, addressing only one may leave in place a mechanism that maintains the other. Integrated care allows goals to be coordinated, risks to be monitored and contradictory advice to be avoided.