The DSM-5 Craving Criterion Was Validated for Cannabis and Five Other Substances, With Moderate Craving More Predictive Than Severe | Research

The takeaway

In a study of 588 adults, the DSM-5 craving criterion was valid across cannabis, alcohol, tobacco, cocaine, heroin, and opioids, with moderate craving ("felt a very strong desire or urge") more strongly associated with clinical validators than severe craving ("could not think of anything else").

Psychiatrists, addiction diagnosticians, DSM researchers, cannabis use disorder clinicians.

The DSM-5 craving criterion is valid for cannabis just as for heroin and cocaine

What the researchers found

Across all six substances, craving was associated with most baseline validators. Moderate craving was more strongly associated with validators than severe craving and improved predictive validity for daily substance use. Including craving improved the validity and clinical relevance of DSM-5 SUD diagnoses across all substances.

Why it matters

The DSM-5 added craving as a new diagnostic criterion for substance use disorders but it had not been systematically validated across substances. This study confirms it belongs in the diagnosis and works consistently across different drugs including cannabis.

The numbers in context

N=588 adults. 6 substances assessed. 90-day daily electronic monitoring. Moderate craving more predictive than severe. Craving improved SUD diagnostic validity across all substances.

How the study worked

Validation study of 588 adults who engaged in binge drinking or illicit drug use and endorsed at least one DSM-5 SUD criterion. Assessed craving across 6 substances. Logistic regression estimated associations with validators. Electronic daily assessment tracked use for 90 days.

What this study cannot tell us

Sample required at least one SUD criterion, so results may not generalize to casual users. Self-reported craving may be interpreted differently across individuals. 90-day follow-up captures a limited window. Cannot determine whether craving causes continued use or vice versa.

How to read the evidence

Well-designed validation study with 90-day daily monitoring across 6 substances in a clinical sample.

When this study was published

Published in 2023.

The bigger picture

For cannabis specifically, validating the craving criterion matters because some have questioned whether cannabis produces "real" cravings comparable to harder drugs. This study shows the craving criterion works just as well for cannabis as for heroin or cocaine.

Questions still open

Why is moderate craving more predictive than severe craving?
Could craving assessments be used as a treatment monitoring tool for cannabis use disorder?

Common questions

Is cannabis craving a real clinical phenomenon?

Yes. This study validated the DSM-5 craving criterion for cannabis just as it did for alcohol, tobacco, cocaine, heroin, and opioids. Cannabis craving was consistently associated with clinical indicators of substance use disorder.

What level of craving is most concerning?

Moderate craving ("a very strong desire or urge to use") was actually more predictive of clinical problems than severe craving ("could not think of anything else"), possibly because severe craving is less commonly endorsed.

Read the original research

Validity of the DSM-5 craving criterion for alcohol, tobacco, cannabis, cocaine, heroin, and non-prescription use of prescription painkillers (opioids).

Psychological medicine, 53(5), 1955-1969

View on PubMed ↗Publisher’s paper ↗

Citation

Shmulewitz, D; Stohl, M; Greenstein, E; Roncone, S; Walsh, C; Aharonovich, E; Wall, M M; Hasin, D S. (2023). Validity of the DSM-5 craving criterion for alcohol, tobacco, cannabis, cocaine, heroin, and non-prescription use of prescription painkillers (opioids).. Psychological medicine, 53(5), 1955-1969. https://doi.org/10.1017/S0033291721003652