Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Clin Psychiatry. Author manuscript; available in PMC 2009 November 16.
Published in final edited form as:
PMCID: PMC2777674

Cannabis withdrawal in the United States: a general population study

Deborah S. Hasin, Ph.D.,1,2,3 Katherine M. Keyes, MPH,1,3 Donald Alderson, M.S.,2 Shuang Wang, Ph.D.,1 Efrat Aharonovich, Ph.D.,2,3 and Bridget F. Grant, Ph.D., Ph.D.4



Although cannabis is the most widely abused illicit drug, little is known about the prevalence of cannabis withdrawal, its factor structure, clinical validity and psychiatric correlates in the general population.


National Epidemiologic Survey on Alcohol and Related Conditions participants were assessed with structured in-person interviews covering substance history, DSM-IV Axis I and II disorders, and withdrawal symptoms after cessation of use. Of these, 2,613 had been frequent cannabis users (≥3 times/week), and a cannabis-only subset (N=1,119) never binge-drank or used other drugs ≥3 times/week.


In the full sample and subset, 44.3% (se 1.19) and 44.2% (se 1.75), respectively, experienced ≥2 cannabis withdrawal symptoms, while 34.4% (se 1.21) and 34.1% (se 1.76), respectively, experienced ≥3 symptoms. The symptoms formed two factors, one characterized by weakness, hypersomnia, and psychomotor retardation, and the second by anxiety, restlessness, depression, insomnia. Both symptom types were associated with significant distress/impairment (p<.01), substance use to relieve/avoid cannabis withdrawal symptoms (p<.01), and quantity of cannabis use (among the cannabis-only users p<.05). Panic (p<.01) and personality disorders (p<.01) associated with anxiety symptoms in both samples, family history of drug problems with weakness symptoms in the subset (p=.01), and depression with both sets of symptoms in the subset (p<.05).


Cannabis withdrawal was prevalent and clinically significant among a representative sample of frequent cannabis users. Similar results in the subset without polysubstance abuse confirmed the specificity of symptoms to cannabis. Cannabis withdrawal should be added to DSM-V and the etiology and treatment implications of cannabis withdrawal symptoms investigated.

Keywords: cannabis, marijuana, withdrawal, general population, DSM-V


Cannabis is the most widely abused illicit drug in the United States. In the last decade, cannabis use disorders have become more prevalent, 1 the potency of cannabis available in the U.S. has strengthened 1,2 and the proportion of substance abuse patients whose primary drug is cannabis has increased. 3 Additionally, current cannabis use is projected to increase in the next decade among aging baby boomers. 4 Given these developments, an improved multidisciplinary understanding of this significant U.S. public health problem is essential to guide treatment and prevention efforts. An important aspect of the cannabis problem for which epidemiologic information is lacking is cannabis withdrawal.

For most diagnoses of substance dependence in DSM-IV, withdrawal is one of seven criteria that jointly reflect dependence as a combination of physiological and psychological processes. 5,6 In DSM-IV, withdrawal is defined as symptoms occurring after cessation of regular, heavy use of a substance. While withdrawal is not required to make a DSM-IV dependence diagnosis, withdrawal symptoms predict poorer prognosis among both alcohol- and drug-dependent individuals. 7,8 Withdrawal is also included in DSM-IV as a separate diagnosis, with specific withdrawal criteria for most substance categories (e.g., stimulants, sedatives).

In 1994, when DSM-IV was published, symptoms of cannabis withdrawal had been described in earlier studies and case reports. 911 However, the evidence for the validity of a cannabis withdrawal syndrome was considered too weak to include cannabis withdrawal in DSM-IV. Since then, evidence has accumulated from animal,12 inpatient laboratory,1315 prospective non-laboratory, 1618 clinical, 19,20 and other retrospective studies 21,22 indicating the occurrence of cannabis withdrawal symptoms, typically beginning within a few hours or days after cessation of cannabis use. Importantly, placebo-controlled laboratory studies show that administration of delta-9 THC suppresses cannabis withdrawal symptoms,23,24 indicating the pharmacological specificity of the cannabis withdrawal syndrome. Cannabis withdrawal symptoms commonly identified in these studies include anxiety, irritability, various types of physical discomfort, negative affect, and appetite and sleep changes.25 A recent study based on a convenience sample of non-treatment seeking cannabis users suggested that cannabis withdrawal symptoms clustered into two components, a physical and a psychological component.22

The accumulating evidence has developed a consensus on the validity of cannabis withdrawal and the need to include it in DSM-V. 22, 25, 26 However, the epidemiology of cannabis withdrawal in the general population remains entirely unknown. This information is important for several reasons. First, if cannabis withdrawal occurs only among a small minority of frequent cannabis users in the general population, then its public health significance is limited. However, if withdrawal is common among frequent cannabis users, this is important knowledge to disseminate to clinicians and to the general public. Second, some previous studies were based on clinic patients, who are often polysubstance abusers, so the withdrawal symptoms they attributed to cannabis may have been due to other drugs. Third, many studies of untreated cannabis users have been small (e.g., samples of 30, 16 12, 17 and 18, 18) potentially leading to unstable results. Fourth, while laboratory studies have provided the fundamental data identifying the cannabis withdrawal syndrome, it is possible that individuals willing to spend several days in an inpatient setting or to visit a laboratory every day for over a month may differ from the broader underlying population of frequent cannabis users in a manner that affects the study results.27 Without conducting research on cannabis withdrawal in a representative sample of frequent cannabis users, the extent to which sample biases have affected our understanding of cannabis withdrawal cannot be known. To our knowledge, such a study has never previously been conducted. Therefore, using data from a large, nationally representative survey that included frequent cannabis users, we investigated: (1) the prevalence of cannabis withdrawal symptoms; (2) their factor structure; (3) whether cannabis withdrawal is accompanied by significant distress or impairment, or use of another substance to relieve/avoid withdrawal (important indicators of clinical significance), and (4) the demographic characteristics, psychiatric and cannabis use history associated with cannabis withdrawal. We addressed these questions in two groups. The first included all frequent cannabis users in the survey (n=2,613), regardless of use of other substances. Results from this group maximized representativeness, but left the possibility open that withdrawal symptoms due to other substances were erroneously attributed to cannabis. The second group, a subset of the first (N=1,119), never abused other substances. This group served the important function of ensuring that results on withdrawal symptoms attributed to cannabis were not actually due to another substance.



Respondents were participants in the National Epidemiologic Survey of Alcohol and Related Conditions (NESARC), a nationally representative United States sample of 43,093 civilian non-institutionalized participants aged 18 and older. The sampling frame is detailed elsewhere. 28,29 The research protocol, including informed consent procedures, received full ethical review and approval from the U.S. Census Bureau and U.S. Office of Management and Budget. The National Institute on Alcohol Abuse and Alcoholism (NIAAA) sponsored the study, conducted in 2001–2002 by the U.S. Bureau of the Census. Young adults, Hispanics, and African-Americans were oversampled. The overall response rate was 81%. Data were adjusted for oversampling and nonresponse. The sample was then weighted to reflect the U.S. population from the 2000 Decennial Census. We included NESARC participants that used cannabis ≥3 times/week during their period of heaviest use. We further analyzed the subset of this sample that never drank ≥5 drinks/occasion or used other drugs ≥3 times/week (N=1,119) to ensure that withdrawal symptoms were due to cannabis rather than another substance.


Professional interviewers from the Census Bureau experienced with health surveys administered the in-person interviews, after ten days of centralized training supervised by NIAAA. Regional supervisors re-contacted a random 10% of all respondents to verify interviews. Further, 2,657 randomly selected respondents were re-contacted and re-asked 1 to 3 sections of the interview, providing NESARC reliability data. 30 In the few cases with uncertain accuracy, data were discarded and a supervisor repeated the interview.


All variables were assessed with the NIAAA Alcohol Use Disorder and Associated Disabilities Interview Schedule-DSM-IV (AUDADIS-IV), 30, 31 a computer-assisted interview with builtin skip, logic, and consistency checks designed for lay interviewers. Detailed withdrawal symptom information was obtained in the DSM-IV substance dependence module. Reliability and validity (including psychiatrist reappraisals) of substance dependence diagnoses and symptom items was good to excellent in U.S. and international clinical and general population studies. 3035

All withdrawal symptoms listed in DSM-IV were asked across all substances. Structured questions covered these symptoms after cessation of use, i.e., the morning after, or within the first few days. Eighteen symptoms were included in this study: anxiety, insomnia, vivid or unpleasant dreams, hallucinations, restlessness, shaking, depressed mood, hypersomnia, psychomotor retardation, and several symptoms of physical discomfort including feeling weak or tired, bad headaches, muscle cramps, runny eyes or nose, yawning, nausea, sweating, fever, and seizure.

Clinical significance was addressed with two variables. One was significant distress or impairment due to the cannabis withdrawal symptoms in social, occupational, or other important areas. The second was use of another substance to avoid or alleviate the withdrawal symptoms. We investigated the relationship of cannabis withdrawal to demographic characteristics (gender, age, race/ethnicity, marital status and education) and to psychiatric and drug use history, including lifetime primary diagnoses of major depressive, generalized anxiety and panic disorders, since these share symptoms with cannabis withdrawal. Substance-induced disorders were excluded. We also included personality disorders assessed in the NESARC (avoidant, dependent, obsessive-compulsive, paranoid, schizoid, histrionic and antisocial). The reliability and validity of these diagnoses are fair to excellent, 31, 3638 comparing favorably to other research diagnostic interviews. Cannabis history variables included duration of period of heaviest use (≥1 one year vs. others), usual number of joints smoked during period of heaviest use (≤1 vs. 2 [the median] or more), and age at onset of use <16 years. Period of heaviest use was dichotomized as <52 weeks (1st quartile) vs. ≥52 weeks. Additional predictors included family history of drug problems (parent or sibling positive).

Data Analysis

Prevalences and standard errors were estimated using SUDAAN to adjust for the complex sample design. 39 Weighted correlations were produced in SAS. Weighted exploratory factor analysis (EFA) was conducted using Mplus 40 using Promax rotation, allowing correlated factors. EFA results were used to construct observed cannabis withdrawal symptom count variables. To examine how indicators of clinical significance and aspects of predictors related to cannabis withdrawal as dependent variables, weighted negative binomial regression models were conducted with STATA. 41 Demographic variables were included as needed based on association with clinical significance indicators and outcome variables. The regression coefficient for a predictor is the log ratio of the change in means in outcome for a unit change in the predictor. Therefore, the coefficients were exponentiated to provide ratio estimates interpreted similarly to an odds ratio.


Sample and subsample characteristics

Frequent cannabis users comprised 6.58% (N=2,613) of the NESARC. Most used cannabis 5–7 days a week during heaviest use (Table 1). Over half (57.2%) were diagnosed with DSM-IV cannabis abuse, while 16.2% were diagnosed with DSM-IV cannabis dependence (≥3 of 6 dependence criteria, not including withdrawal). Mean age of first cannabis use was 16.3 (range, 5–53), and total mean years from first to most recent use of cannabis was 12.5 (range, <1–53 years). Mean period of heaviest use was 5.2 years. Over two-thirds smoked >1 joint/day on days they smoked during their period of heaviest use; mean joints smoked/day was 3.9 (s.e., 0.1). About one-fifth had primary major depression, 8.8% had generalized anxiety disorder, 12.6% had panic disorder and 39.4% had personality disorders.

Table 1
Characteristics of NESARC participants using cannabis ≥3 times/week during heaviest period

The subsample of frequent cannabis-only users comprised 2.69% of the NESARC (N=1,119). Compared to polysubstance frequent cannabis users, these 1,119 cannabis-only users were more likely to be female, Hispanic, younger, unmarried, and better educated; less frequent cannabis users; with lower prevalence of cannabis use disorders, panic disorder and personality disorders (p<.02 – <.001).

Prevalence of cannabis withdrawal symptoms

The most prevalent symptoms (Table 2) were feeling weak/tired, hypersomnia, yawning, psychomotor retardation, anxiety, and depressed mood. In the full sample of frequent cannabis users, 57.7% experienced ≥1 cannabis withdrawal symptom, 44.3% experienced ≥2 symptoms, and 34.4% experienced ≥3 symptoms. Among the subset of frequent cannabis users that did not abuse other substances, 59.4%, 44.2%, and 34.1% experienced ≥1, ≥2, and ≥3 cannabis withdrawal symptoms, respectively. Thus, a large proportion of frequent cannabis users experienced one or more of these symptoms, with little difference seen between those that used and did not use other substances. From a different perspective, among the entire NESARC sample, the prevalence of being a frequent cannabis user that experienced ≥3 cannabis withdrawal symptoms was 2.1% (SE=0.1). All symptoms were tested for differences in prevalence by age, sex, ethnicity and education. Only scattered symptoms were found to differ by these characteristics; in light of the number of tests run and exploratory nature of the tests, the few significant results did not appear to contribute substantially to our understanding of these symptoms.

Table 2
Lifetime prevalence of cannabis withdrawal symptoms among all participants using cannabis ≥3 times/week during heaviest use, and the subset using that often without other substance use potentially leading to withdrawal

Association of cannabis withdrawal symptoms with cannabis dependence criteria and clinical characteristics

Among all frequent cannabis users, the correlation between the number of cannabis withdrawal symptoms and the number of DSM-IV cannabis dependence criteria (range, 0 – 6) was 0.54 (p<.001), while the corresponding correlation among frequent cannabis-only users was 0.64 (p<.001), indicating strong associations. To place these results in context, the corresponding correlation for alcohol withdrawal symptoms and number of DSM-IV alcohol dependence criteria (range, 0 – 6) among those ever drinking ≥5 drinks/day was 0.58, and the corresponding correlation for cocaine withdrawal symptoms and number of DSM-IV cocaine dependence criteria (range, 0 – 6) among those using cocaine ≥3 times/week was 0.76 (p<.001). Thus, the association of withdrawal symptoms to other dependence criteria was similar for cannabis and alcohol, and only slightly weaker than the corresponding association for cocaine.

Among all frequent cannabis users, those with ≥3 cannabis withdrawal symptoms reported a median duration of 105 weeks, compared to 157 weeks in those who experienced <3 symptoms (non-parametric comparison of medians, p<0.01). Those reporting 3 cannabis withdrawal symptoms had higher quantity of cannabis use versus those with <3 symptoms (mean 4.3 joints versus 3.6 joints; t=2.6, p=0.01). There was no statistically significant difference in the mean age of onset between those reporting ≥3 symptoms vs. <3 symptoms (17 vs. 18, p=0.10). Those with ≥3 symptoms were more likely to have a family history of drug problems than those with <3 symptoms (χ2 =8.3, df=1, p<0.006). Results were essentially the same in the subsample of cannabis-only users, except that those with ≥3 symptoms had a significantly younger age of onset (15.9 vs. 17.4, p=0.04).

Factor analysis

In the 2,613 frequent users, the final model (Table 3) included 12 symptoms. The fit statistics indicated that a two-factor solution had the best model fit (x2=35.12, df=23, p=0.17; RMSEA=0.01). The eigenvalues over 1.0 also indicated two factors; the first factor, termed “Weakness,” and the second termed “Anxiety/Depression.” The item loadings on the factors indicated that the symptoms within each factor comprised a homogeneous set that measured the factor well, with the Weakness factor measured by weakness, hypersomnia, psychomotor retardation, and yawning (factor loadings 0.54–0.99) and the Anxiety/Depression factor measured by anxiety, depression, sweating, nausea, muscle aches, restlessness, tremors and insomnia (factor loadings 0.52–0.88). The two factors were correlated (0.64). In the cannabis-only subsample of 1,119, a similar two-factor model (Table 3) was obtained (x2=15.3, df=16, p=0.50, RMSEA = 0.046). The Weakness factor was measured by the same symptoms found in the larger group (factor loadings 0.65–0.88), and the Anxiety/Depression factor was measured by the same symptoms excepting nausea, which was not retained (factor loadings 0.55–1.00). The two factors were correlated (0.59).

Table 3
Exploratory factor analysis: factor loadings

Clinical significance of cannabis withdrawal symptoms

In examining the clinical significance of the weakness and anxiety/depression cannabis withdrawal symptoms, distress/impairment showed a strong and statistically significant relationship to the symptom count for both types of withdrawal symptoms in the full sample of frequent cannabis users (Table 4) and the subsample (Table 5). Furthermore, a strong, significant relationship was found between use of a substance to relieve/reduce cannabis withdrawal and both the weakness and anxiety/depression withdrawal symptoms in the full sample of frequent cannabis users (Table 4) and the subsample (Table 5).

Table 4
Negative Binomial Regression Results, All Frequent Cannabis Users (N=2,613)
Table 5
Negative Binomial Regression Results, ‘Cannabis Only’ Frequent Users (N=1,119)

In examining the relationship of demographic characteristics, psychiatric and cannabis use history to cannabis withdrawal symptoms in the full sample of frequent cannabis users (Table 4), gender was associated with both the weakness symptoms and the anxiety/depression symptoms. Duration of heaviest use (less than one year) was significantly related to both types of withdrawal symptoms. Major depressive disorder was associated with anxiety/depression symptoms but was unrelated to weakness symptoms. Generalized anxiety disorder was unrelated to cannabis withdrawal symptoms, panic disorder was associated only with anxiety/depression withdrawal symptoms, and personality disorder was associated with both types of withdrawal symptoms.

In examining the relationship of psychiatric and cannabis use history to cannabis withdrawal symptoms in the subsample of frequent ‘cannabis only’ users (Table 5), age was associated with both anxiety/depression symptoms and weakness symptoms, while gender was associated with weakness symptoms only. Quantity of cannabis was significantly related to both types of withdrawal symptoms. Duration of heaviest use (less than one year) was associated with anxiety/depression withdrawal symptoms but not weakness symptoms, while family history of drug problems was associated with weakness but not anxiety/depression withdrawal symptoms. Major depressive disorder and personality disorders were associated with both types of withdrawal symptoms, generalized anxiety disorder was not associated with either type, and panic disorder was associated with only the anxiety/depression withdrawal symptoms.


The results of this study indicate that cannabis withdrawal symptoms are highly prevalent among frequent cannabis users in the U.S. general population. The most common withdrawal symptoms were feeling weak/tired, hypersomnia, sleeping, yawning, depression and feeling nervous or anxious, similar to symptoms found in volunteers in inpatient laboratory studies 13, 14, 26 and in patients.20, 19 Further, cannabis withdrawal, consistently identified by 11 symptoms, differentiated into two factors, one characterized by weakness and the other characterized by anxiety and depression. The findings are based on the largest and most representative sample of frequent cannabis users studied to date.

The largest early clinical study of cannabis effects, a study of 53 men hospitalized for 21–42 days, indicated that in the week following cessation of cannabis use, symptoms commonly reported included sleep disturbance (89%), restlessness (89%), and irritability (62%).9 In more recent prospective outpatient and inpatient studies, 1318 no cannabis withdrawal symptom emerged in all studies. However, symptoms commonly reported were anxiety (after low dose13) 14, 16, 18 and symptoms of physical discomfort. 1618 Irritability was also commonly found, 1318 although in one study 13 only in the low consumption condition; in a second study 14 only in the high consumption condition; and in a third study, evidence was only “moderate.”17 The anxiety/depression withdrawal symptoms found in these studies corresponded well to our second factor, providing important consistency between this large general population sample and previous prospective outpatient and inpatient studies. Weakness withdrawal symptoms were also found in lab studies 13, 14 and in a clinical study of adolescents, especially males. 20 While fatigue was not found during cannabis withdrawal among cannabis users in their home environments, 17 the other studies in conjunction with the present findings indicate that these symptoms merit further investigation. Irritability and anxiety may receive greater clinical consensus as regular features of cannabis withdrawal because they are subjectively and clinically striking compared to fatigue and related symptoms. Varying emergence of weakness and depression/anxiety symptoms as part of cannabis withdrawal may also relate to varying levels of Δ9-THC in cannabis and varying levels of other cannabinoids that, while weaker than Δ9-THC, may interact with Δ9-THC to alter its effects. 42

Our results also indicated support for the validity of the cannabis withdrawal syndrome in the general population, as clinical distress/impairment and using to relieve or avoid symptoms were strongly related to both depression/anxiety and weakness in both samples. We did not find that age of onset of cannabis use predicted cannabis weakness or depression/anxiety withdrawal symptoms in either sample. This could be due to restricted variance of this variable, as early age of onset is highly associated with frequent cannabis use. 43, 44 We also found that shorter duration of cannabis use predicted more withdrawal symptoms. This could be because shorter-term users did not have tolerance to cannabis, as might be expected of longer-term or dependent users. These validators should be further addressed in future laboratory studies to establish more specific mechanisms. Finally, weakness/fatigue cannabis withdrawal symptoms in our study were associated with a family history of drug problems in the subset of frequent cannabis only users, while the depression/anxiety withdrawal symptoms were not. Further evidence of the familial nature of these two types of symptoms would contribute important information towards understanding their etiology.

Budney and colleagues showed that psychopathology measured on symptom scales was related to cannabis withdrawal 19 in patients seeking treatment for marijuana problems, but no previous studies addressed the relationship of cannabis withdrawal to DSM-IV disorders. We studied primary major depression, generalized anxiety and panic disorder because their symptoms partially overlap with cannabis withdrawal symptoms. We studied personality disorders because this has not been done before and the high prevalence of personality disorders in “healthy” volunteers for research studies 27 may have influenced earlier findings. The association of primary panic disorder or major depression with cannabis depression/anxiety withdrawal symptoms suggests possible common vulnerability, meriting further investigation. The association of personality disorders with depression/anxiety but not weakness withdrawal symptoms in the subsample suggests further study of weakness symptoms in participants screened for personality as well as Axis I psychopathology, as weakness withdrawal symptoms may emerge more among individuals without personality disorders, and prior laboratory and prospective studies screened for Axis I psychopathology only.

Study limitations are noted. First, withdrawal symptoms were measured via retrospective structured self-report rather than observation. However, the coherence of our findings on the anxiety withdrawal symptoms with earlier studies adds strength to the aggregated literature on this aspect of cannabis withdrawal, while the finding on the weakness factor merits further investigation. Second, the symptoms analyzed included those listed for other substances in DSM-IV. While multiple symptoms of anxiety, physical discomfort, and negative affect were included, irritability was not. As noted above, a number of studies show irritability as a cannabis withdrawal symptom. While our factors might have been altered if we had included irritability, irritability tends to co-occur with anxiety, depression, and insomnia1, 7, 22 suggesting that the addition of irritability would not result in a major restructuring of the factors. However, future epidemiologic studies on cannabis withdrawal should clearly include irritability. Third, we did not have precise measures of the amount of Δ9-THC ingested and cannabis potency and efficiency of smoked self-administration can vary considerably. Most of those in our sample appear to have been heavily exposed to cannabis during their period of heaviest use, but we cannot precisely quantify the relationship of cannabis ingested with the emergence of withdrawal symptoms on abstinence. Fourth, information is lacking on whether those in the sample who used cannabis 3–4 times a week used cannabis on consecutive days. However, these respondents constituted a minority of the sample, and results of the factor analysis were virtually identical when the 3–4x/week users were removed from the sample. Fifth, it is possible that individual symptoms we identified might be related to other conditions, e.g., that the individual symptom of weakness might reflect postural hypertension, a direct effect of cannabis use. 45 However, our wording of the question on the timing of the symptoms relative to cannabis use (“the morning after, or in the first few days after”) in addition to the empirical clustering of symptoms found reduces the possibility that the full syndrome reported is a conglomeration of individual symptoms resulting from other diverse conditions.

Advantages of the study are also noted. The similarities between the full group and the subset suggested that withdrawal symptoms of cannabis were not being confused with other substances in the full group. First, the large and representative sample afforded a new vantage point on cannabis withdrawal. Further, the subset of respondents who did not abuse other substances provided an important opportunity to confirm that cannabis withdrawal symptoms reported by the larger group were not due to other substances they had used. Second, weighted factor analysis of the symptoms provided an organized quantitative method of arriving at a set of items that measured two aspects of cannabis withdrawal well. While replication of the factor structure is clearly needed, this study is the first to provide information from such analyses. Third, the measures included in the study enabled us to address important historical and psychiatric correlates of the two factors, contributing new and valuable information on clinical significance and associated aspects of psychiatric history.

In demonstrating a syndrome, some variability in its presentation does not negate its existence. For example, few would now argue against the existence of an alcohol withdrawal syndrome, despite well-recognized variability in its clinical features. 46 While such variability in the presentation of cannabis withdrawal may relate to the amount of the active substance ingested, it is also likely to be related to individual variation in the pharmacokinetics and pharmacodynamics of the psychoactive elements in cannabis. A better understanding of these processes appears important in understanding the development of cannabis disorders and their treatment. Additionally, we found gender and age effects on the number of depression/anxiety and weakness symptoms reported. Demographic differences in the presentation of withdrawal symptoms are not unique to cannabis, 47 but should be further investigated in future studies.

The findings of this study have several implications. First, the prevalence of cannabis withdrawal symptoms among frequent cannabis users in the U.S. general population indicates a problem national in scope that is not limited to patients or individuals who willing to volunteer for inpatient laboratory studies. Given the projected increase in cannabis use in the coming decade, 4 the established increase in potency of the cannabis in general distribution in the U.S., 2 and evidence that higher potency increases the reward properties of cannabis, 27 the prevalence of cannabis withdrawal among users highlights the importance of understanding such withdrawal better. Second, the findings suggest that clinicians should be aware of the potential for cannabis withdrawal symptoms in frequent cannabis users, even if the primary drug is a different substance. Due to the common use of marijuana among patients in substance abuse treatment, the occurrence of cannabis withdrawal may need closer attention than previously believed. Post-hospital marijuana use predicts poor outcome of dependence on other substances, 48 and cannabis withdrawal symptoms may contribute to such poor outcome. Coverage of cannabis withdrawal symptoms in research evaluations can provide more information on this point, potentially assisting in identification of patients at risk for poor outcome due to these symptoms. Third, the study findings further support addition of cannabis withdrawal to DSM-V and ICD-11. Our analyses did not directly investigate a valid threshold, but a 3-symptom threshold would be consistent with other withdrawal categories in DSM-IV and would identify about one-third of all frequent cannabis users. Lastly, the findings suggest that an understanding of the etiology of these symptoms would be worthwhile. Such an understanding should contribute to the development of better treatments for patients with cannabis use disorders, an increasing public health problem.

Supplementary Material


The National Epidemiologic Survey on Alcohol and Related Conditions was conducted and funded by the National Institute on Alcohol Abuse and Alcoholism with supplemental funding by the National Institute on Drug Abuse. This study was, in part, funded by the Intramural Program of the National Institutes of Health, National Institute on Alcohol Abuse and Alcoholism. This research was supported in part by grants from the National Institute on Alcoholism and Alcohol Abuse (K05 AA014223) and the National Institute on Drug Abuse (RO1 DA018652) and support from New York State Psychiatric Institute. The views and opinions expressed in this report are those of the authors and should not be construed to represent the views of any sponsoring organization, agencies, or the U.S. government. We thank Valerie Richmond, MA for manuscript preparation and editorial assistance.


1. Compton WM, Grant BF, Colliver JD, et al. Prevalence of marijuana use disorders in the United States: 1991–1992 and 2001–2002. JAMA. 2004;291:2114–2121. [PubMed]
2. ElSohly MA, Ross SA, Mehmedic Z, et al. Potency trends of delta9-THC and other cannabinoids in confiscated marijuana from 1980–1997. J Forensic Sci. 2000;45:24–30. [PubMed]
3. Substance Abuse and Mental Health Services Administration (SAMHSA) Office of Applied Statistics, Treatment Episode Data Set (TEDS), Highlights 2002, National Admissions to Substance Abuse Treatment Services DASIS Series S-22, DHHS Publication number (SMA) 04-3946. Rockville MD: 2004.
4. Colliver JD, Compton WM, Gfroerer JC, et al. Projecting drug use among aging baby boomers in 2020. Ann Epidemiol. 2006;16:257–265. [PubMed]
5. Hasin D, Hatzenbuehler ML, Keyes K, et al. Substance use disorders: diagnostic and statistical manual of mental disorders, fourth edition (DSM-IV) and international classification of diseases, tenth edition (ICD-10) Addiction. 2006;101(Suppl 1):59–75. [PubMed]
6. Edwards G. The Alcohol Dependence Syndrome: a concept as stimulus to enquiry. Br J Addiction. 1986;81:171–183. [PubMed]
7. Hasin D, Paykin A, Meydan K, et al. Withdrawal and tolerance: their prognostic significance in DSM-IV alcohol dependence. J Stud Alcohol. 2000;63:431–438. [PubMed]
8. Schuckit MA, Daeppen JB, Danko GP, et al. Clinical implications for four drugs of the DSM-IV distinction between substance dependence with and without a physiological component. Am J Psychiatry. 1999;156:41–49. [PubMed]
9. Jones RT, Benowitz N, Bachman J. Clinical studies of cannabis tolerance and dependence. Ann N Y Acad Sci. 1976;282:221–239. [PubMed]
10. Nowlan R, Cohen S. Tolerance to marijuana: heart rate and subjective “high. Clin Pharmacol Ther. 1977;22:550–556. [PubMed]
11. Rohr J, Skowlund S, Martin T. Withdrawal Sequelae to Cannabis Use. Int J Addict. 1989;24:627–631. [PubMed]
12. Lichtman AH, Martin BR. Marijuana withdrawal syndrome in the animal model. J Clin Pharmacol. 2002;42(11 Suppl):20S–27S. [PubMed]
13. Haney M, Ward AS, Comer SD, et al. Abstinence symptoms following smoked marijuana in humans. Psychopharmacol (Berl) 1999;141:395–404. [PubMed]
14. Haney M, Ward AS, Comer SD, et al. Abstinence symptoms following oral THC administration to humans. Psychopharmacol (Berl) 1999;141:385–394. [PubMed]
15. Hart CL, Ward AS, Haney M, et al. Comparison of smoked marijuana and oral Delta(9)-tetrahydrocannabinol in humans. Psychopharmacol (Berl) 2002;164:407–415. [PubMed]
16. Kouri EM, Pope HG. Abstinence symptoms during withdrawal from chronic marijuana use. Exp Clin Psychopharmacol. 2000;8:483–492. [PubMed]
17. Budney AJ, Hughes JR, Moore BA, et al. Marijuana abstinence effects in marijuana smokers maintained in their home environment. Arch Gen Psychiatry. 2001;58:917–924. [PubMed]
18. Budney AJ, Moore BA, Vandrey RG, et al. The time course and significance of cannabis withdrawal. J Abnorm Psychol. 2003;112:393–402. [PubMed]
19. Budney AJ, Novy PL, Hughes JR. Marijuana withdrawal among adults seeking treatment for marijuana dependence. Addiction. 1999;94:1311–1322. [PubMed]
20. Crowley TJ, Macdonald MJ, Whitmore EA, et al. Cannabis dependence, withdrawal, and reinforcing effects among adolescents with conduct symptoms and substance use disorders. Drug Alcohol Depend. 1998;50:27–37. [PubMed]
21. Wiesbeck GA, Schuckit MA, Kalmijn JA, et al. An evaluation of the history of a marijuana withdrawal syndrome in a large population. Addiction. 1996;91:1469–1478. [PubMed]
22. Copersino ML, Boyd SJ, Tashkin DP, et al. Cannabis withdrawal among non-treatment-seeking adult cannabis users. Am J Addict. 2006;15:8–14. [PubMed]
23. Haney M, Hart CL, Vosburg SK, et al. Marijuana withdrawal in humans: effects of oral THC or divalproex. Neuropsychopharmacol. 2004;29:158–170. [PubMed]
24. Budney AJ, Vandrey RG, Hughes JR, et al. Oral delta-9-tetrahydrocannabinol suppresses cannabis withdrawal symptoms. Drug Alcohol Depend. 2007;86:22–29. [PubMed]
25. Budney AJ, Hughes JR, Moore BA, et al. Review of the validity and significance of cannabis withdrawal syndrome. Am J Psychiatry. 2004;161:1967–1977. [PubMed]
26. Haney M. The marijuana withdrawal syndrome: diagnosis and treatment. Curr Psychiatry Rep. 2005;7:360–366. [PubMed]
27. Bunce SC, Noblett KL, McCloskey MS, et al. High prevalence of personality disorders among healthy volunteers for research: implications for control group bias. J Psychiatr Res. 2005;39:421–430. [PubMed]
28. Grant BF, Moore TC, Sheppard J, et al. Source and Accuracy Statement: Wave 1 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism; 2003. [Accessed on 8/20/2007]. Available at:
29. Grant BF, Stinson FS, Dawson DA, et al. Prevalence and co-occurrence of substance use disorders and independent mood and anxiety disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) Arch Gen Psychiatry. 2004;61:807–816. [PubMed]
30. Grant BF, Harford TC, Dawson DA, et al. The Alcohol Use Disorder and Associated Disabilities Schedule: reliability of alcohol and drug modules in a general population sample. Drug Alcohol Depend. 1995;39:37–44. [PubMed]
31. Grant BF, Dawson DA, Stinson FS, et al. The Alcohol Use Disorder and Associated Disabilities Interview Schedule-IV (AUDADIS-IV): reliability of alcohol consumption, tobacco use, family history of depression and psychiatric diagnostic modules in a general population sample. Drug Alcohol Depend. 2003;71:7–16. [PubMed]
32. Canino GJ, Bravo M, Ramirez R, et al. The Spanish Alcohol Use Disorder and Associated Disabilities Interview Schedule (AUDADIS): reliability and concordance with clinical diagnoses in a Hispanic population. J Stud Alcohol. 1999;60:790–799. [PubMed]
33. Cottler L, Grant B, Blaine J, et al. Concordance of DSM-IV alcohol and drug use disorder criteria and diagnoses as measured by the AUDADIS-ADR, CIDI and SCAN: results of a cross-national study. Drug Alcohol Depend. 1997;47:195–206. [PubMed]
34. Hasin D, Grant B, Cottler L, et al. Nosological comparisons of alcohol and drug diagnoses: a multisite, multi-instrument international study. Drug Alcohol Depend. 1997;47:217–226. [PubMed]
35. Hasin D, Carpenter K, McCloud S, et al. The Alcohol Use Disorder and Associated Disabilities Interview Schedule (AUDADIS): reliability of alcohol and drug modules in a clinical sample. Drug Alcohol Depend. 1997;44:133–141. [PubMed]
36. Hasin D, Goodwin R, Stinson F, et al. The epidemiology of major depressive disorder: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Arch Gen Psychiatry. 2005;62:1097–1106. [PubMed]
37. Grant BF, Hasin DS, Stinson FS, et al. The epidemiology of DSM-IV panic disorder and agoraphobia in the United States: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry. 2006;67:363–374. [PubMed]
38. Grant BF, Hasin DS, Stinson FS, et al. Prevalence, correlates, co-morbidity, and comparative disability of DSM-IV generalized anxiety disorder in the USA: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Psychol Med. 2005;35:1747–1759. [PubMed]
39. Research Triangle Institute. Software for Survey Data Analysis (SUDAAN), Version 8.1. Research Triangle Park, NC: Research Triangle Institute; 2002.
40. Muthen BO, Muthen LK. Mplus, Version 3.13. Los Angeles, CA: Muthen & Muthen; 2005.
41. StataCorp. STATA/SE, Version 9.1. College Station TX: StataCorp; 2005.
42. Ameri A. The effects of cannabinoids on the brain. Prog Neurobiol. 1999;58:315–348. [PubMed]
43. Baumeister SE, Tossman P. Association between early onset of cigarettes, alcohol and cannabis use and later drug use patterns: an analysis of a survey in European metropolises. Eur Addict Res. 2005;11:92–98. [PubMed]
44. Ellickson PL, D’Amico EJ, Collins RL, et al. Marijuana use and later problems: when frequency of recent use explains age of initiation effects (and when it does not) Subst Use Misuse. 2005;40:343–359. [PubMed]
45. Gorelick DA, Heishman SJ, Preston KL, et al. The cannabinoid CB1 receptor antagonist rimonabant attenuates the hypotensive effect of smoked marijuana in male smokers. Am Heart J. 2006;151:754.e1–754.e5. [PubMed]
46. Saitz R. Introduction to alcohol withdrawal. Alcohol Health Res World. 1998;22:5–12. [PubMed]
47. Pomerleau CS, Tate JC, Lumley MA, et al. Gender differences in prospectively versus retrospectively assessed smoking withdrawal symptoms. J Subst Abuse. 1994;6:433–440. [PubMed]
48. Aharonovich E, Liu X, Samet S, et al. Post-discharge cannabis use and its relationship to cocaine, alcohol, and heroin use: a prospective study. Am J Psychiatry. 2005;162:1507–1514. [PubMed]