Skip to main content

Table 2 Hazard ratios from survival analyses predicting rate of progression to age of first drink, first heavy drinking, remission, and relapse, and age of first alcohol dependence, remission, and relapse

From: Transitions in alcohol use over time: a survival analysis

Model/variables

To first drink

Heavy drinking

Alcohol dependence

To first HD

To remission

To relapse

To first AD

To remission

To relapse

1738 event, 31 censored

1293 event, 442 censored

932 event, 363 censored

157 event, 775 censored

640 event, 1904 censored

524 event, 117 censored

68 event, 456 censored

Model 1: Clinical risk

N = 1158

N = 1141

N = 1288

N = 675

N = 1113

N = 506

N = n/a

 Psychiatric risk

  Nicotine dep

1.19**

1.30***

–

–

1.40**

–

–

  Externalizing

1.41***

1.14***

0.89**

1.28†

1.31***

0.89†

–

  Internalizing

–

1.06

1.16**

–

1.38***

–

–

 Family history

  Father

1.13†

1.15†

–

–

1.30**

–

–

  Mother

–

–

–

–

–

–

–

  Twin

1.28***

1.39***

–

1.54†

2.22***

0.79†

–

Model 2: individual predictors

N = 925

N = 931

N = 715

N = 517

N = 900

N = 334

N = 279

 Personality traits

  Extraversion

–

–

0.97***

–

–

–

0.95

  Openness

1.01

–

–

1.05†

–

–

–

  Neuroticism

0.98**

–

–

–

1.02†

–

–

  Agreeableness

0.97**

1.00

1.03**

0.95†

0.99

1.03**

–

  Conscientiousness

–

–

1.02**

–

–

–

–

 Drinking motives

  Enhancement

1.02†

1.05***

0.97**

–

1.04**

–

–

  Social

1.02

–

0.97**

1.10***

–

–

–

  Coping

–

–

0.97**

–

1.05***

0.95***

1.15***

  Conformity

–

–

1.05**

–

–

–

0.86**

 Alcohol expectancies

  Risk and aggression

1.08***

1.09***

–

1.12

1.14***

–

–

  Tension reduction

–

1.10**

0.99

–

–

–

–

  Sociability

1.05

–

–

–

–

–

–

  Sexuality

–

–

–

0.78†

–

–

–

  Liquid courage

–

–

–

0.74**

–

–

–

  Cog and Beh impair

0.96

0.96

1.14***

–

–

1.00

–

  Self-perception

–

–

–

–

1.07

–

–

  1. †p < .05; **p ≤ .01; ***p ≤ .001
  2. ns not significant. Variable was included in the model as it was a significant predictor in earlier models, but was not significant in the final model
  3. – Not included in the model as the variable was not a significant predictor of the outcome in a previous model where only the variables within the specific subdomain were included at the same time
  4. HD heavy drinking, AD alcohol dependence. Externalizing antisocial personality disorder and drug dependence. Internalizing Major depression, panic, post-traumatic stress, and generalized anxiety disorders. Cog and Beh Impair cognitive and behavioral impairment
  5. All models also included a control for the previous transition and demographic variables (age, income, and education level). Missing data for ages or other variables creates slight differences in numbers of cases counted towards events across models