Make a tax-deductible donation today

Measuring risky drinking: An examination of the validity of different episodic drinking thresholds in predicting alcohol-related harms



Dr Michael Livingston, Centre for Alcohol Policy Research


In 2009 the National Health and Medical Research Council released revised low-risk drinking guidelines, which suggested that Australians drinking five or more standard drinks on a particular occasion were putting themselves at risk of harm (1). These guidelines were heavily criticised (2, 3), despite being broadly consistent with guidelines specified internationally. In the research literature too, studies of episodic drinking typically use a measure based on the frequency of drinking five or more standard drinks (a definition which itself varies based on the standard units being used). While this threshold clearly defines drinking behaviour with a range of risks and negative consequences, there has been limited research outside of United States college-based studies to determine its appropriateness.


This report explores this issue in two ways. Firstly, using the 2010 National Drug Strategy Household Survey data, a variety of different drinking thresholds are examined, to determine the demographic and attitudinal predictors of episodic heavy drinking at various levels (i.e. 5+, 11+ and 20+ drinks). There are significant variations in prevalence rates depending on the threshold used:

  • 42% of drinkers report drinking 5+ drinks on a monthly basis
  • 15.8% of drinkers report drinking 11+ drinks on a monthly basis and
  • 5% of drinkers report drinking 20+ drinks on a monthly basis.

There are substantial variations in heavy drinking prevalence based on demographic and other factors. Generally speaking, across all three definitions, heavy episodic drinking is more common amongst males, young adults, people who have never married, those living in regional Australia, people who smoke, people who use illicit drugs, people who started drinking at younger ages and people who drink beer or pre-mixed spirits.

Across the three different definitions of risky drinking, broadly similar relationships were identified, suggesting that the same factors are associated with heavy drinking regardless of how it is defined.

The second phase of this study examined 15 risky-drinking thresholds (based on volume and frequency) and their relationship with three self-reported alcohol-related problems and behaviours to try to determine the most appropriate definition of ‘risky-drinking’. The most appropriate risky drinking threshold identified varied depending on the mode of analysis (i.e. on the goodness of fit measure used) and on the type of outcome being considered (e.g. thresholds were generally higher for injury than for self-reported risky behaviour). In general, risky drinking thresholds of seven or fewer drinks provided the best balance between sensitivity (ability of a threshold to correctly identify people likely to experience harm) and specificity (the ability of a threshold to correctly identify people not likely to experience harm).


These findings support the continuing use of a risky-drinking definition of five or more drinks, based on the Australian drinking guidelines.

Recent research papers

FARE continues to fund and undertake research that contributes to the knowledge-base about alcohol harms and strategies to reduce them.

This research is used to inform our approach to evidence-based alcohol policy development, ensuring that the solutions we are advocating for are informed by research. FARE’s research is also often quoted by governments, other not-for-profit organisations and researchers in public discussions about alcohol, demonstrating that FARE is seen as a leading source of information.

Join our community

Will you join the community taking action on alcohol?

Join our community

Fill out the form below to receive regular updates & resources.

Join our community

Get updates & resources straight to your inbox