“Side Effects” Observed Within ~1–90 Days After Cannabis Onset: Epidemiological Estimates for the United States, 2004-2014

Karl C. Alcover*, Christopher Thompson, James C. Anthony

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


Background and Objectives: Among newly incident cannabis users, fewer than 4% develop cannabis dependence syndromes within 12 to 24 months, but earlier cannabis side effect (SE) incidence and early syndrome formation remains understudied. We estimate cannabis SE incidence within ~1 to 90 days after first use, and estimate odds ratios (OR) for SE pairs to quantify potential syndromic “running together.”. Methods: Each year, 2004 to 2014, the US populations under study consisted of noninstitutionalized civilian residents age 12-years-and-older, sampled for National Surveys on Drug Use and Health. Computerized self-interviews identified 3710 newly incident cannabis users and asked about SEs. Analysis-weighted year-specific SE and SE-SE pair incidence proportions were estimated. Analysis-weighted odds ratio (OR) estimates quantify SE-SE co-occurrences, judged as “greater than chance” when lower bounds (LB) of 95% confidence intervals (CIs) exceed 1.0. Meta-analysis is used to summarize and to check reproducibility. Results: Illustrative of estimates on 17 SEs, roughly 1/2 of cannabis initiates experienced “wanting or trying to cut down or stop using,” but 80% of these cut back, such that less than 7% had symptom-like inability to cut back or stop. An estimated 25% had “spent a lot of time getting or using cannabis.” The SE-SE paired estimate is 2.8% for those who had spent a lot of time and also had wanted/tried to cut down (95% CI = 2.0, 4.0). OR estimation suggests no syndromic co-occurrence of this SE-SE pair (OR = 0.9; 95% CI = 0.5, 1.6). In contrast, “cannabis causing serious problems at home/work/school” and “continuing to use despite physical problems” were more rare, but had strong SE-SE co-occurrence (OR = 14.8, LB = 2.4). For 78 of 136 possible SE-SE pairs, meta-analysis LB estimates exceeded the 1.0 threshold. Conclusion and Scientific Significance: In these US community samples of cannabis initiates studied soon after first use, the vast majority were free of individual SE experiences counted toward Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition cannabis use disorder diagnoses. For a minority, SE-SE pairs might indicate syndrome formation. These epidemiological meta-analysis estimates might find use in refinement of clinical practice guidelines for fast effective syndrome screens when primary care patients are seen soon after cannabis onset, akin to clinical screening guides designed for newly incident drinkers (Am J Addict 2019;00:1–9).

Original languageEnglish
Pages (from-to)465-472
Number of pages8
JournalAmerican Journal on Addictions
Issue number6
StatePublished - 1 Nov 2019
Externally publishedYes


Dive into the research topics of '“Side Effects” Observed Within ~1–90 Days After Cannabis Onset: Epidemiological Estimates for the United States, 2004-2014'. Together they form a unique fingerprint.

Cite this