Table 3. Association of Each Substance Variable With Transitions Between Stages of Lifetime Illicit Drug Use, Use Disorders, and Remission, Adjusted for All Sociodemographic Variablesa.
| Model | Transition 1: Commencing Use | Transition 2: Use to Abuse, With and Without Dependence | Transition 3: Use to Dependence | Transition 4: Remission From Abuse, Without Dependencee | Transition 5: Remission From Dependencee | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Odds Ratio, (95% CI) | χ21 | P Value | Odds Ratio, (95% CI) | χ21 | P Value | Odds Ratio, (95% CI) | χ21 | P Value | Odds Ratio, (95% CI) | χ21 | P Value | Odds Ratio, (95% CI) | χ21 | P Value | |
| Total sample size, No.b | 90 022 | 23 073 | 23 073 | 2088 | 1167 | ||||||||||
| 1 | |||||||||||||||
| Per 10% of age cohort already using drugsc | 1.31 (1.29-1.33) | 1389.01 | <.001 | 1.11 (1.07-1.16) | 24.84 | <.001 | 1.10 (1.03-1.18) | 7.16 | .007 | 1.65 (1.54-1.77) | 197.06 | <.001 | 1.65 (1.44-1.88) | 53.87 | <.001 |
| 2 | |||||||||||||||
| Per 10% of age cohort already using alcoholc | 1.51 (1.49-1.54) | 2043.51 | <.001 | 1.13 (1.07-1.19) | 21.85 | <.001 | 1.08 (0.99-1.18) | 2.94 | .09 | 1.44 (1.30-1.61) | 44.00 | <.001 | 1.67 (1.12-2.50) | 6.22 | .01 |
| 3 | |||||||||||||||
| Highest level of individual’s alcohol involvementd | |||||||||||||||
| Use | 4.64 (4.39-4.90) | 3087.76 | <.001 | 1.50 (1.26-1.80) | 19.60 | <.001 | 1.39 (0.99-1.96) | 3.53 | .06 | 1.39 (1.11-1.72) | 8.61 | .003 | 1.94 (1.19-3.18) | 6.95 | .008 |
| Abuse | 10.78 (9.38-12.40) | 1120.27 | <.001 | 5.52 (4.40-6.92) | 219.64 | <.001 | 3.80 (2.50-5.78) | 39.03 | <.001 | 1.36 (1.07-1.72) | 6.38 | .01 | 2.18 (1.32-3.62) | 9.14 | .003 |
| Dependence | 12.81 (10.29-15.94) | 522.38 | <.001 | 6.48 (4.94-8.50) | 182.37 | <.001 | 6.33 (4.12-9.73) | 71.06 | <.001 | 1.49 (1.12-1.99) | 7.59 | .006 | 1.76 (1.09-2.84) | 5.30 | .02 |
| Remission | 4.08 (3.20-5.21) | 128.18 | <.001 | 2.59 (1.78-3.78) | 24.48 | <.001 | 2.01 (1.20-3.37) | 7.08 | .008 | 2.25 (1.67-3.02) | 28.88 | <.001 | 3.28 (1.99-5.40) | 21.70 | .001 |
| Joint test of all 4 indicators, χ24 (P value) | NA | 3251.41 | <.001 | NA | 465.08 | <.001 | NA | 150.16 | <.001 | NA | 30.29 | <.001 | NA | 42.15 | <.001 |
| 4 | |||||||||||||||
| Type of drug(s) already used by individual | |||||||||||||||
| Cannabis | NA | NA | NA | 3.10 (2.53-3.81) | 116.54 | <.001 | 2.53 (1.76-3.63) | 25.02 | <.001 | 1.68 (1.36-2.08) | 22.50 | <.001 | 1.65 (1.20-2.27) | 9.56 | .002 |
| Cocaine | NA | NA | NA | 3.22 (2.79-3.72) | 255.96 | <.001 | 3.54 (2.83-4.42) | 123.08 | <.001 | 1.10 (0.96-1.26) | 1.98 | .16 | 0.88 (0.73-1.06) | 1.74 | .19 |
| Other | NA | NA | NA | 2.62 (2.28-3.01) | 184.01 | <.001 | 3.25 (2.61-4.04) | 112.36 | <.001 | 0.80 (0.71-0.89) | 16.55 | <.001 | 0.92 (0.75-1.12) | 0.67 | .41 |
| Joint test of all 3 indicators, χ23 (P value) | NA | NA | NA | NA | 696.02 | <.001 | NA | 379.52 | <.001 | NA | 43.05 | <.001 | NA | 12.99 | .005 |
| 5 | |||||||||||||||
| Individual used ≥2 drug types | NA | NA | NA | 5.17 (4.66-5.73) | 976.45 | <.001 | 5.99 (5.02-7.16) | 390.67 | <.001 | 0.86 (0.76-0.98) | 5.45 | .02 | 0.91 (0.74-1.13) | 0.74 | .39 |
Abbreviation: NA, not applicable.
All transitions are based on weighted person-year data. Each model was estimated for all transitions with only the 1 substance variable entered at a time as factors of each transition controlling for person-year age groups; sex; education level; major depressive episode; broad bipolar disorder; number of anxiety disorders; and survey (transition 1); all controls specified for transition 1 as well as age tertile of commencing alcohol use (transition 2); all controls specified for transition 2 as well as drug abuse (transition 3); and all controls specified for transition 2 as well as speed to transition from use to disorder and years with disorder (transitions 4 and 5).
Indicates the total unweighted number of respondents included in model conditioning on initial stage.
Percentage (divided by 10) of ±5-year specific cohort who had used the substance by the prior person-year.
Alcohol transition variables are mutually exclusive; only the highest alcohol level ever having been met (use < abuse < dependence < remission) is indicated for each person year.
Individuals with a missing age at onset of remission were excluded from the model (n = 147 for remission from abuse and n = 104 for remission from dependence).