A Review of the Literature on the Cognitive Effects of Alcohol Hangover

Richard Stephens; Jonathan Ling; Thomas M. Heffernan; Nick Heather; Kate Jones


Alcohol Alcohol. 2008;43(2):163-170. 

In This Article


In the Introduction we highlighted executive function as likely to be susceptible to hangover effects, based on analogy with research into acute alcohol and sleep deprivation. However, only one study has assessed executive functioning during hangover and no effect was observed. This may be due to insensitivity associated with the manner of assessment - using a management simulation game rather than using well-validated neuropsychological testing. Future research should employ well-validated executive function tests.

Encouragingly, there is some consistency in the kinds of cognitive effect shown across the laboratory and naturalistic alcohol consumption studies. Verster et al. (2003) and McKinney and Coyle (2004) showed hangover-related memory decrements. The latter study also shares the finding of attention decrements with the laboratory-based studies of Roehrs et al. (1991) and Kruisselbrink et al. (2006). A convergent finding from differing methodologies is known as triangulation and is considered to be a sign of validity in psychological research (e.g. Howitt and Cramer, 2005). As more studies are carried out, there may be a continued trend of convergent findings, which could lead future reviewers to be able to reach more definitive conclusions. Nevertheless, we have identified some serious problems with the two methodological approaches used to study hangover effects.

Rigorous laboratory-based studies, where participants are blinded to alcohol consumption, have tended not to show effects of hangover on performance. This insensitivity may arise partly because the pharmacological model of drug action, where a certain drug dose is predicted to affect aspects of behaviour, may be of little relevance to normal drinking (Finnigan and Hammersley, 1992). This is because the typical laboratory-based controlled intake study ignores potentially important everyday aspects of drinking that are usually set by the drinker - for example, the number and types of drink consumed, the pace of consumption, whether food is also consumed, and the social setting. On the other hand, naturalistic alcohol consumption studies, which allow for all these factors, have tended to show effects of hangover on performance. However, as participants were unblinded in these studies, the significant results are likely to be contaminated by expectancy effects.

Finnigan and Hammersley (1992) reviewed all the acute alcohol studies published from 1980 to 1991 and this remains the most-up-to-date, comprehensive review published. They concluded that the research is 'ambitious rather than rigorous' (p. 74) and 'while ample evidence exists that alcohol is capable of impairing performance... it is premature to conclude that it invariably does so' (p. 74). The criticisms they raise with respect to acute alcohol studies may also be applied to the hangover studies reviewed here. In both literatures there are studies with flaws in basic experimental design, such as very small sample size and ineffective control of expectancy, for example by not including a placebo. Additionally, in both literatures the majority of publications describe laboratory-controlled alcohol consumption studies employing the pharmacological model of drug action. As outlined in the previous paragraph, this model may be of little relevance to normal drinking. Finnigan and Hammersley (1992) argued that: 'Natural intoxication may lead to more impairment, less impairment or different impairments' (p. 78). This is likely to be true for hangover effects as well as for acute intoxication.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: