Background
Methods
Systematic literature search
Search | Search terms | Databases and timeframes |
---|---|---|
Search 1 N = 131 | Subject OR Title (cannabis or marijuana or marihuana) AND Subject OR Title (“affective disorder” or “anxiety disorder” or anxiety) NOT (mouse or mice or rat or rats) | PsycInfo (1806-March 2013) |
Search 2 N = 168 | Subject OR Title (cannabis or marijuana or marihuana) AND Subject OR Title (“affective disorder” or “anxiety disorder” or anxiety) AND Keyword (misus* or abus* or depend* or "harmful use" or "harmful usage") NOT (mouse or mice or rat or rats) | Medline (1950-March 2013) |
N = 256 | Total N from both searches excluding duplicates |
Study selection
-
anxiety/no anxiety in cannabis users/non-users or cannabis users with/without CUD or
-
cannabis use/no use or cannabis use with/without CUD in cases with anxiety/no anxiety;
Data extraction
Study; Country | Design* | TotalN(Nin this study**) | Sample type | Sampling method | Anxiety assessment; diagnostic system | Anxiety diagnosis (timeframe) | Anxiety prevalence | Cannabis assessment; diagnostic system | Cannabis use; CUD (timeframe) | Cannabis use/CUD prevalence |
---|---|---|---|---|---|---|---|---|---|---|
Agosti et al., 2002; NCS, USA [26] | Cross-sectional | 8098 | General population | Probability | CIDI DSM-III-R | AD (current) | – | CIDI DSM-III-R | Use (past month); CD (lifetime) | – |
Beard et al., 2006; NoRMHS, Australia [27] | Longitudinal | 9191 (1013) | General population | Random | CIDI ICD-10 | PD, SP, OCD (at T1) | 15% (at T1) | CIDI ICD-10 | CUD (at T0) | CUD 3% (at T0) |
Brook et al., 1998; Upstate NY, USA [28] | Longitudinal | 975 (T3 = 745 T4 = 698) | Adolescents- adults | Random | DISC DSM-III-R | SA, OAD, SAD (at T4) | – | DISC DSM-III-R | Never- ≥ weekly (at T3) | Use (at T3): 56% never 9% ≥ weekly |
Brook et al., 2001; Colombia [29] | Longitudinal | 2226 | Adolescents | Random | HSC | AD (at T1, T2) | – | Interview | Non-regular- regular (≥monthly; lifetime) | – |
Buckner et al., 2008; Oregon, USA [30] | Longitudinal | T1 = 1709 T4 = 816 | High school students | Random | K-SADS DSM-III-R | SAD (at T1) | 2% (at T1) | LIFE, SCID-I/NP DSM-IV | CD (at T4) | CD 6% (at T4) |
Buckner & Schmidt, 2008; USA [31] | Cross-sectional | 337 (214) | Undergraduate students | Random | SIAS | Equivalent to SAD | 19% (scores in clinical range) | Questionnaire | Never- frequent (≥weekly; lifetime) | 31% never 32% frequent |
Buckner et al., 2012; USA [32] | Cross-sectional | 343 (200) | Adult tobacco smokers | Random | SIAS | Equivalent to SAD | – | Questionnaire | Never- current (daily; past month) | 19% never 39% current use; 13% current daily |
Cascone et al., 2011; Switzerland [33] | Cross-sectional | 110 | Adolescents in schools/psycho-educational unit | Convenience | STAI-Y B | Trait anxiety | – | ADAD DSM-IV-TR | Use (past month) CD (past 12 months) | Past month: No CD: 94% none CD: 66% daily |
Chabrol et al., 2005; France [34] | Cross-sectional | 212 | High school and college students | Random | STAI A | State anxiety | – | Questionnaire DSM-IV | Past 6 months: None- > daily; CD | 46% none 23% > daily |
Chabrol et al., 2008; France [35] | Cross-sectional | 248 | High school students | Random | STAI A | State anxiety | – | Interview | None- use ≥1× (past 6 months) | 76% none 24% users |
Cheung et al., 2010; CAMH, Canada [36] | Cross-sectional cycles of 2001–2006 survey | 14531 (13478) | General population | Probability | GHQ12 ≥4/12 symptoms | AMD (past 12 months) | 9% (past 12 months) | Interview | None- daily (past 12 months) | – |
Cougle et al., 2011; NCS-R, USA [37] | Cross-sectional | 5672 | General population | Stratified probability | CIDI DSM-IV | PTSD (lifetime) | 7% (lifetime) | CIDI | Never- use ≥1× (lifetime) | 42% users (lifetime) |
Crum et al., 1993; ECA, USA [38] | Longitudinal | 18572 (577) | General population | Probability | DIS DSM-III | OCD (past 12 months) | 1% (past 12 months) | DIS | No use- use ≥6× (past 12 months) | 84% no drug use 9% cannabis |
Degenhardt et al., 2001; NSMHWB, Australia [39] | Cross-sectional | 10641 | General population | Stratified random | CIDI DSM-IV | SAD, AP, PD, GAD, OCD, PTSD (past 12 months) | 6% (past 12 months) | CIDI DSM-IV | Past 12 months: No use- use ≥5×; CD | Past 12 months: 5% users 2% CD |
Degenhardt et al., 2010; VAHCS, Australia [40] | Longitudinal | 1943 (1520, wave 1–8) | High school students | Stratified random | GHQ12 > 2 symptoms (at 24) | AMD (at 24) | 21% (at 24) | Interview | None- weekly+ (past 6 months at 15–17) | 34% users (at 15–17) |
Degenhardt et al., 2013; VAHCS, Australia [41] | Longitudinal | 1943 (1756, wave 1–9) | High school students | Stratified random | CIDI ICD-10 | GAD, SAD, PD, AP (past 12 months at 29) | 11% (at 29) | CIDI ICD-10 | None/<weekly- weekly + (past 6–12 months at 15–29); CD (past 12 months at 29) | Lifetime: 67% none/ <weekly 2% weekly+ 4% CD (at 29) |
Fergusson et al., 1996; CHDS, New Zealand [42] | Longitudinal | 1265 (927) | Adolescents (16 years) | Stratified | DISC/DIS DSM-III-R | GAD, OAD, SA (at 15–16) | 9% (at 15–16) | Interview | None- use (past 12 months at 15–16) | 20% users (at 15–16) |
Hayatbakhsh et al., 2007; MUSP, Australia [43] | Longitudinal | 7223 (3157) | Adult children of mothers in study | Convenience | YASR (resembles DSM-III-R) | AMD (at 21) | – | Interview | Never used drugs- frequent ≤ daily (past month) | 12% frequent |
Lamers et al., 2006; USA [44] | Cross-sectional | 41 (30) | General population | Convenience | BAI | BAI scores | – | Questionnaire urine screen | None (past 12 months)- use ≥10× (lifetime) | 50% non-users 50% users |
Low et al., 2008; USA [45] | Cross-sectional | 632 | Adolescents in primary care | Convenience | PRIME-MD DSM-IV | PD, GAD, AD (past 1–6 months) | 7% (past 1–6 months) | PRIME-MD DSM-IV | CA (past 6 months) | 6% CA |
Martins & Gorelick, 2011; NESARC, USA [46] | Cross-sectional | 43093 | General population | Stratified random | AUDADIS DSM-IV | PD, AP, SP, GAD (lifetime) | 17% (lifetime) | AUDADIS DSM-IV | CA + CD (lifetime) | – |
McGee et al., 2000; DMHDS, New Zealand [47] | Longitudinal | 1037 (891) | Adolescents (at 15) | Convenience | DISC DSM-III | Internalising disorders (AMD; at 15) | 12% (at 15) | Interview | None- use ≥1 (past 12 months) | 14% users (at 15) |
NPMS, UK; appendix, Moore et al., 2007 [10] | Longitudinal | 8580 (1578) | Adults (at 16–74) | Random | CIS-R ≥ 12 | AMD | 11% CIS-R ≥12 | Interview | Use: no/yes (lifetime); CD (past 12 months): no/yes | 16% users 2% CD |
Patton et al., 2002; VAHCS, Australia [48] | Longitudinal | 1943 (1601, wave 1–7) | High school students | Stratified random | CIS-R ≥ 12 at 21 | AMD (at 21) | 16% (at 21) | Interview | None- < weekly (past 6 months at 15–17) | 59% users (lifetime) |
Roberts et al., 2007; TH2K, USA [49] | Cross-sectional | 4175 | Adolescents | Probability | DISC DSM-IV | AP, GAD, PD, SAD, PTSD (past 12 months) | 7% (past 12 months) | DISC DSM-IV | CUD (past 12 months) | 3% CUD (past 12 months) |
Swift et al., 2008; VAHCS, Australia [50] | Longitudinal | 1943 (1520, wave 1–8) | High school students | Stratified random | CIS-R > 11 | AMD (at 15–17) | – | Interview, CIDI DSM-IV | Past 12 months at 24: None- weekly+; CD | 28% weekly+ at 24 who used at 15-17 |
van der Pol et al., 2013; CanDep + NEMESIS-2, Netherlands [51] | Cross-sectional | 1324: D+: 252 N2: 1072 | General population/ ‘coffee shop’ users (18–30) | Stratified random; convenience/chain-referral | CIDI DSM-IV | SAD, PD, GAD, AP (past 12 months) | 8% (past 12 months) | CIDI DSM-IV | No CD (group N2; none or <3×/week use)- CD (D+; use ≥3×/week) (past 12 months) | 16% CD (past 12 months) |
Van Laar et al., 2007; NEMESIS, Netherlands [52] | Longitudinal | T0: 7076 T2: 4848 | General population | Probability | CIDI DSM-III-R | PD, AP, SAD, SP, GAD, OCD (3-year incidence, T0-T2) | 6% (at T0-T2) | CIDI DSM-III-R | No use- use >5× (lifetime at T0) | – |
Wittchen et al., 2007; EDSP, Germany [53] | Longitudinal | T0: 1395 T3: 1019 (1310) | General population | Random | CIDI DSM-IV | PD, GAD, AP, SAD, SP, SA, OCD, PTSD (at T0) | 23% (at T0) | CIDI DSM-IV | Use: no/yes (lifetime); CUD (lifetime) | Lifetime: 54% use 13% CUD |
Zvolensky et al., 2006; CSHS, USA [54] | Cross-sectional | 4745 | General population | Stratified random | DIS DSM-IV-TR | PA (lifetime) | 6% (lifetime) | DIS DSM-IV-TR | Lifetime: No use- use ≥5×; CD | Lifetime: 25% users 1% CD |
Zvolensky et al., 2010; NCS-R, USA [55] | Cross-sectional | 5672 | General population | Stratified probability | CIDI DSM-IV | PD (lifetime) | 6% (lifetime) | CIDI | None- use ≥1× (lifetime) | 42% users (lifetime) |
Study/Name (Part 1: anxiety diagnoses) | Cannabis use vs. no use | CUD vs. no CUD (or no use) |
OR(95% CI) Anxiety/no anxiety in user/non-user |
OR(95% CI) Anxiety/no anxiety in CUD/no CUD | Data location in study |
ORadjusted for | |
---|---|---|---|---|---|---|---|
Agosti et al., 2002; NCS, USA [26] | Current AD in lifetime CD who used within past month vs. no CD | 2.6 (1.5-4.5) | Text p. 646 | No information | |||
Beard et al., 2006; NoRMHS, Australia [27]1
| T0 CUD (vs. no CUD) to T1 AD | .78 (.18-3.30) | Table three | Unadjusted (comorbidity with other diagnoses possible) | |||
Brook et al., 1998; Upstate NY, USA [28] | T4 AD to T3 use vs. no use | 1.16 (1.00-1.35) | Table two | Demographics, prior AD | |||
Brook et al., 2001; Columbia [29]2
| T1 AD to T2 regular vs. non-regular use | .94 (.86-1.03) | Table one | Demographics, cannabis use at T1 | |||
T1 regular vs. non-regular use to T2 AD | 1.48 (1.09-2.01) | Table two | Demographics, AD at T1 | ||||
Combined: AD or regular use: T1 vs. T2 | 1.18 (.94-1.48) | ||||||
Buckner et al., 2008; Oregon, USA [30] | T1 SAD to T4 CD vs. no CD | 4.88 (1.43-16.64) | Text p. 235 | Gender, T1 anxiety, conduct, mood, alcohol use disorders, T1 CUD excluded | |||
Cascone et al., 2011; Switzerland [33] | Past 12 months CD (vs. no CD) predicted with STAI-Y B | 1.02 (.97-1.08) | Table five | Withdrawal coping, social problems, recruitment context | |||
Cougle et al., 2011; NCS-R, USA [37]3
| Lifetime: PTSD to use vs. no use | 2.45 (1.70-3.52) | Table one | Demographics, lifetime alcohol/nicotine dependence/abuse | |||
Crum et al., 1993; ECA, USA [38]4
| Past 12 months: OCD to use vs. no use | 1.54 (.78-3.04) | Table one | Excluded past or baseline OCD cases | |||
Degenhardt et al., 2001; NSMHWB, Australia [39] | Past 12 months: AD to use vs. no use | Past 12 months: AD to CD vs. no use | .88 (.60-1.29) | 1.40 (.84-2.37) | Table four | Demographics, other substance use, neuroticism | |
Degenhardt et al., 2013; VAHCS, Australia [41] | AD at 29 to ≥ weekly vs. no use at 15-29 | AD at 29 to CD at 29 vs. no CD | 3.2 (1.1-9.2) | 2.2 (1.1-4.4) | Table one, three | Demographics, alcohol/substance use at 29, adolescent anxiety/depression | |
Fergusson et al., 1996; CHDS, New Zealand [42] | AD at 16 to use vs. no use at 15 | 1.2 (.5-2.8) | Table three | Demographics, substance use and dependence, anxiety/depression, other mental health problems at 15 | |||
Low et al., 2008; USA [45] | Past 6 months: CA (vs. no CA) to AD | 1.4 (.4-4.7) | Table four | Demographics, sampling site, depression | |||
Martins & Gorelick, 2011; NESARC, USA [46] | Lifetime: CUD (vs. no CUD) to AD | 3.2 (2.0-5.2) | Table four | Demographics | |||
Roberts et al., 2007; TH2K, USA [49] | Past 12 months: AD to CUD vs. no CUD | .9 (.4-2.1) | Table six | Concurrent (past 12 months) mood, conduct, ADHD disorders; alcohol, other substance abuse/dependence | |||
van der Pol et al., 2013; CanDep + NEMESIS-2, Netherlands [51] | Past 12 months: AD to CD (vs. no/non-frequent use) (D + vs. N2 groups) | 1.12 (.48-2.63) | Table two; Authors | Demographics, childhood adversity, tobacco (past month), frequent alcohol, cocaine, ecstasy use (past 12 months) | |||
van Laar et al., 2007; NEMESIS, Netherlands [52] | AD (T0-T2) to T0 use vs. no use | 1.18 (.71-1.97) | Table four | Demographics, neuroticism, childhood trauma, lifetime: alcohol, other SUDs, psychotic symptoms, AD | |||
Wittchen et al., 2007; EDSP, Germany [53] | Lifetime use vs. no use to T0 AD | Lifetime CUD vs. no CUD to T0 AD | 1.5 (1.1-2.1) | 1.7 (1.1-2.5) | Table four | Gender | |
Zvolensky et al., 2006; CSHS, USA [54] | Lifetime: use vs. no use to PA | Lifetime: CD vs. no CD to PA | .89 (.63-1.30) | 2.1 (1.1-4.3) | Text p. 482 | Demographics, other substance use | |
Zvolensky et al., 2010; NCS-R, USA [55]3
| Lifetime: PD to use vs. no use | 1.70 (1.33-2.17) | Table one | Demographics, lifetime alcohol, nicotine, illicit substance abuse/dependence | |||
NCS-R combined3
| Lifetime: PTSD + PD to use vs. no use | 2.04 (1.50-2.78) | |||||
Study/Name (Part 2: anxiety severity scores)*
|
Non-user
M (SD); N
(timeframe)
|
Cannabis User
M (SD); N
(timeframe)
|
CUD
M (SD); N
(timeframe)
|
OR
(
95% CI
) Anxiety/no anxiety in user/non-user
|
OR
(
95% CI
) Anxiety/no anxiety in CUD/no CUD
|
Location in study
|
Scale (study exclusionary criteria)
|
Buckner & Schmidt, 2008; USA [31] | 23.6 (11.6); 105 (lifetime) | 22.2 (13.1); 109 (≥weekly frequent use; lifetime) | .81 (.50-1.33) | Table one | SIAS (none) | ||
Buckner et al., 2012; USA [32] | 21.3 (15.5); 66 (lifetime) | 21.7 (13.7); 134 (past month) | 1.05 (.62-1.80) | Table one; Authors | SIAS (high-risk suicidal behaviour, psychotic) | ||
Chabrol et al., 2005; France [34]5
| 35.3 (11.1); 98 (past 6 months) | 37.1 (10.3); 114 (past 6 months) | 38.1 (9.8); 44 (past 6 months) | 1.36 (.83-2.22) | 1.61 (.84-3.07) | Table two Table three | STAI A (none) |
Chabrol et al., 2008; France [35] | 38.3 (12.4); 189 (past 6 months) | 42.9 (13.3); 59 (past 6 months) | 1.94 (1.14-3.30) | Table one | STAI A (none) | ||
Lamers et al., 2006; USA [44] | 4.0 (4.3); 15 (past 12 months) | 2.9 (1.9); 15 (lifetime) | .55 (.15-2.03) | Table three | BAI (alcohol, drug dependence, schizophrenia, depression, antisocial behaviour, psychoactive drug use) | ||
Study/Name (Part 3: anxiety + depression, AMD)
|
Cannabis use vs. no use
|
CUD vs. no CUD (or no use)
|
OR
(
95% CI
) AMD/no AMD in user/non-user
|
OR
(
95% CI
) AMD/no AMD in CUD/no CUD
|
Location in study
|
OR
adjusted for
| |
Cheung et al., 2010; CAMH, Canada [36] | Past 12 months: AMD to daily use vs. no use | 2.05 (1.18-2.93) | Table two | Demographics, alcohol misuse | |||
Degenhardt et al., 2010; VAHCS, Australia [40]6
| AMD at 24 (wave 8) to weekly + use vs. no use past 6 months at 15–17 (wave 1–6) | .88 (.55-1.40) | Table two | Demographics, adolescent: AMD, alcohol, nicotine use | |||
Hayatbakhsh et al., 2007; MUSP, Australia [43] | AMD at 21 to frequent (past month) vs. never used drugs (lifetime) | 2.1 (1.1-4.0) | Table four | Demographics, no other illicit drugs, maternal and adolescent: AMD, alcohol, nicotine use | |||
McGee et al., 2000; DMHDS, New Zealand [47]7
| Internalising disorders at 15 to use vs. no use past 12 months at 15 | 2.45 (1.41-4.25) | Table five | Unadjusted (adjusted OR could not be used because 95% CI were not reported) | |||
NPMS, UK; appendix, Moore et al., 2007 [10] | AMD (CIS-R ≥ 12) to ever use vs. no use | AMD (CIS-R ≥ 12) to CD vs. no CD | .8 (.4-1.6) | .9 (.2-3.6) | p. IV | Excluded if baseline CIS-R ≥ 12, demographics, other drugs, alcohol, nicotine use | |
Patton et al., 2002; VAHCS, Australia [48]6
| AMD (CIS-R ≥ 12) at 21 (wave 7) to < weekly use vs. no use past 6 months at 15–17 (wave 1–6) | 1.4 (.94-2.0) | Table three | AMD at 15–17, alcohol use, parental demographics | |||
Swift et al., 2008; VAHCS, Australia [50]6
| Weekly + use vs. no use (past 12 months at 24, wave 8, who used cannabis at 15–17, waves 1–6) to AMD (CIS-R > 11) at 15–17 (at 3/6 waves of wave 1–6) | CD vs. no CD (past 12 months at 24, wave 8, who used cannabis at 15–17, waves 1–6) to AMD (CIS-R> 11) at 15–17 (at 3/6 waves of wave 1–6) | 2.0 (1.0-3.8) | 1.4 (.71-2.9) | Table four | Demographics, adolescent: maximum level of cannabis use, nicotine and alcohol use, antisocial behaviour | |
VAHCS combined6
| AMD at 15–24 to at least < weekly use at 15–17 (vs. no use) | 1.35 (.80-2.27) |
Meta-analysis
-
OR > 1: 1.5 (small), 2.5 (moderate), =4 (large)
-
OR<1: 0.67 (small), 0.4 (moderate), =0.25 (large)
-
OR=1 and/or 95% CI crossing the line of no effect (OR= 1): no association between cannabis use (or CUD) and anxiety.
-
anxiety/no anxiety vs. cannabis use/no use,
-
anxiety/no anxiety vs. CUD/no CUD (or no use),
-
anxiety + depression/no diagnosis vs. cannabis use/no use.
Sensitivity and moderator analyses
Publication bias analyses
Results
Anxiety vs. use/no useN= 15 (Figure2) | Anxiety vs. CUD/no CUD (or no use)N= 13 (Figure3) | Anxiety + depression vs. use/no useN= 5 (Figure4) | ||
---|---|---|---|---|
Overall mean weighted
OR
|
OR (95% CI) p
two-tailed
| 1.24 (1.06-1.45) .006* | 1.68 (1.23-2.31) .001* | 1.68 (1.17-2.40) .004* |
Heterogeneity statistics
|
Q(df)
p
two-tailed
I
2
| 30 (14) .009* 53% | 55 (12) <.0001* 78% | 8 (4) .091 50% |
Which studies, removed one at a time from the analysis, remove the significance of the overall mean weighted OR? | None | None | Cheung et al., 2010 [36] | |
Which studies, added to all previous studies one at a time, remove the significance of the overall mean weighted OR? | Degenhardt et al., 2001 [39] | NPMS, 2007 [10] | ||
Rosenthal’s Fail-safe
N
for
p > .05
|
N studies needed to remove the significance of the overall mean weighted OR (N studies missing for every study in meta-analysis needed to remove the significance of the overall mean weighted OR)
|
N=49 (49/15 = 3) |
N=105 (105/13 = 8) |
N=17 (17/5 = 3) |
Duval and Tweedie’s Trim-and-Fill analysis
| Funnel plot symmetrical? N studies missing on either side of the overall mean weighted OR
| YES | YES | YES |
N=0 |
N=0 |
N=0 | ||
Begg and Mazumdar Rank Order Correlation
|
τ
| .13 | -.19 | -.30 |
p
two-tailed
| .488 | .360 | .462 | |
Egger’s regression
| intercept | .25 | 1.73 | −4.05 |
p
two-tailed
| .748 | .005* | .415 |
Subgroup of studies | Anxiety vs. cannabis use/no use | Anxiety vs. CUD/no CUD (or no use) | |||||
---|---|---|---|---|---|---|---|
Nstudies |
OR(95% CI) |
p
two-tailed
|
Nstudies |
OR(95% CI) |
p
two-tailed
| ||
Overall mean weighted
OR
| 15 | 1.24 (1.06-1.45) | .006* | 13 | 1.68 (1.23-2.31) | .001* | |
Subgroup analyses (Figure S3) | |||||||
ORs adjusted for confounders | Yes | 12 | 1.24 (1.04-1.47) | .014* | 10 | 1.66 (1.17-2.37) | .005* |
No | 3 | 1.28 (.78-2.08) | .326 | 3 | 1.86 (1.10-3.15) | .021* | |
Yes vs. No |
Q(1) = .04; p = .848 |
Q(1) = 8.96; p = .003* | |||||
Clinical diagnosis of anxiety (based on DSM/ICD) | Yes | 9 | 1.29 (1.04-1.61) | .021* | 11 | 1.87 (1.43-2.44) | <.001* |
No | 6 | 1.17 (.93-1.48) | .186 | 2 | 1.14 (.78-1.66) | .509 | |
Yes vs. No |
Q(1) = .30; p = .586 |
Q(1) = 36.91; p < .001* | |||||
Meta-regression
|
N
studies
|
slope
|
slope p
two-tailed
|
N
studies
|
slope
|
slope p
two-tailed
| |
Predictor: Year of publication Outcome: weighted OR
| |||||||
15 | .02 | .299 | 13 | -.005 | .901 |