Table 5.
Unadjusted and adjusted odds ratios (95% confidence interval) for estimated risk of alcohol abuse and alcohol dependence in young adulthood by level of childhood depressed mood among youth reporting alcohol use based on multinomial regression analyses1. Data are from an epidemiologic sample originally recruited at the time of entry into first grade classrooms in 1985-1986 from an urban mid-Atlantic public school system (n=2311).
| Childhood depressed mood level | Estimated Relative Risk of Alcohol Abuse Observed during the YA Follow-up Interval |
||||
|---|---|---|---|---|---|
| Total | Males | Females | African-American | Other race-ethnicity | |
|
Unadjusted | |||||
| None | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Low | 0.96 (0.61-1.52) p=0.86 | 0.89 (0.47-1.69) p=0.72 | 0.98 (0.47-2.04) p=0.95 | 0.90 (0.47-1.73) p=0.75 | 1.56 (0.93-2.62) p=0.09 |
| High | 0.53 (0.25-1.16) p=0.11 | 0.62 (0.24-1.60) p=0.32 | 0.63 (0.21-1.95) p=0.42 | 0.52 (0.18-1.53) p=0.24 | 0.62 (0.23-1.65) p=0.34 |
|
Adjusted | |||||
| None | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Low | 0.94 (0.57-1.53) p=0.79 | 0.86 (0.47-1.60) p=0.64 | 1.01 (0.53-1.93) p=0.98 | 0.78 (0.40-1.53) p=0.48 | 1.51 (0.88-2.57) p=0.13 |
| High | 0.58 (0.27-1.22) p=0.15 | 0.60 (0.29-1.25) p=0.17 | 0.58 (0.19-1.74) p=0.33 | 0.53 (0.20-1.45) p=0.22 | 0.70 (0.24-2.08) p=0.53 |
|
Estimated Relative Risk of Alcohol Dependence Observed during the YA Follow-up Interval3 | |||||
|
Unadjusted
| |||||
| None | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Low | 1.56 (1.01-2.43) p=0.047 | 1.29 (0.73-2.27) p=0.38 | 2.26 (1.07-4.78) p=0.03 | 1.60 (0.95-2.69) p=0.08 | 2.23 (0.79-6.26) p=0.13 |
| High | 1.82 (1.20-2.76) p=0.005 | 2.00 (1.23-3.23) p=0.005 | 2.56 (1.32-4.96) p=0.005 | 2.21 (1.49-3.29) p<0.001 | 1.42 (0.58-3.47) p=0.45 |
|
Adjusted2 | |||||
| None | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Low | 1.50 (0.89-2.52) p=0.13 | 1.29 (0.67-2.47) p=0.45 | 2.27 (1.04-4.95) p=0.04 | 1.35 (0.75-2.42) p=0.32 | 2.14 (0.71-6.48) p=0.18 |
| High | 2.07 (1.31-3.26) p=0.002 | 2.20 (1.37-3.52) p=0.001 | 2.55 (1.16-5.62) p=0.02 | 2.46 (1.68-3.61) p<0.001 | 1.38 (0.47-4.09) p=0.56 |
One multinomial regression model was used to assess alcohol abuse and alcohol dependence as separate outcomes, referenced to drinkers without either disorder.
The multivariable model was adjusted for sex and race (when applicable), age, caregiver history of heavy drinking and/or illegal drug use, peer drinking, neighborhood disadvantage, conduct problems, intervention status, subsidized lunch, and clustering within schools. In order to avoid biases resulting from nonrandom deletion of participants within our prospective study cohorts, and to keep the representativeness of the original sample intact thereby stabilizing the number of participants contributing information to the regression analyses, we created separate categories for missing information for all variables in the model.
See text and appendix for clarification of ‘clinically significant’ alcohol dependence syndrome under observation in this study.