Logistic regression analysis of drug use among individuals who left home once a week or less
Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | Model 6 | Model 7 | Model 8 | Model 9 | Model 10 | |
Painkiller | Stimulant | Sedative | Marijuana | Cocaine | Club drug | Hallucinogen | Heroin | Inhalant | Meth | |
OR (p value) | OR (p value) | OR (p value) | OR (p value) | OR (p value) | OR (p value) | OR (p value) | OR (p value) | OR (p value) | OR (p value) | |
Age | 0.97 (0.002) | 0.93 (<0.001) | 0.96 (<0.001) | 0.96 (<0.001) | 0.93 (<0.001) | 0.92 (<0.001) | 0.94 (<0.001) | 0.93 (<0.001) | 0.94 (<0.001) | 0.94 (<0.001) |
Female | 0.30 (<0.001) | 0.27 (<0.001) | 0.31 (<0.001) | 0.45 (0.01) | 0.26 (<0.001) | 0.30 (0.001) | 0.33 (0.002) | 0.35 (0.01) | 0.24 (<0.001) | 0.32 (0.002) |
Minority* | 1.88 (0.04) | 1.86 (0.08) | 1.69 (0.08) | 1.03 (0.91) | 1.42 (0.36) | 1.75 (0.14) | 1.73 (0.14) | 2.09 (0.06) | 2.08 (0.07) | 1.74 (0.15) |
Married or cohabit | 1.13 (0.71) | 1.65 (0.23) | 1.23 (0.53) | 1.15 (0.66) | 1.96 (0.16) | 2.05 (0.13) | 1.61 (0.28) | 2.67 (0.04) | 1.81 (0.22) | 1.87 (0.18) |
Household with children | 2.94 (0.001) | 2.56 (0.02) | 1.98 (0.04) | 2.13 (0.01) | 1.78 (0.19) | 2.03 (0.11) | 3.01 (0.01) | 2.21 (0.08) | 3.47 (0.01) | 3.12 (0.01) |
Work fulltime | 1.22 (0.55) | 0.98 (0.96) | 0.84 (0.58) | 1.29 (0.41) | 0.95 (0.91) | 1.13 (0.77) | 1.01 (0.99) | 1.12 (0.78) | 1.19 (0.69) | 1.22 (0.63) |
College and above | 1.28 (0.43) | 1.82 (0.12) | 1.76 (0.07) | 0.79 (0.43) | 1.43 (0.39) | 1.28 (0.55) | 1.21 (0.64) | 1.48 (0.36) | 1.56 (0.30) | 1.11 (0.80) |
Family network | 1.01 (0.89) | 1.02 (0.76) | 1.01 (0.80) | 1.02 (0.68) | 1.07 (0.33) | 1.06 (0.40) | 1.05 (0.51) | 1.08 (0.27) | 1.13 (0.10) | 1.12 (0.12) |
Friendship network | 1.08 (0.13) | 1.19 (0.002) | 1.09 (0.06) | 1.10 (0.03) | 1.21 (0.002) | 1.21 (0.003) | 1.20 (0.003) | 1.18 (0.01) | 1.14 (0.06) | 1.16 (0.02) |
Trauma symptoms | 1.04 (0.046) | 1.02 (0.34) | 1.04 (0.047) | 1.02 (0.41) | 1.05 (0.06) | 1.05 (0.10) | 1.03 (0.21) | 1.05 (0.09) | 1.04 (0.15) | 1.03 (0.25) |
Depression | 1.19 (0.18) | 1.41 (0.02) | 1.19 (0.15) | 1.30 (0.03) | 1.09 (0.60) | 1.27 (0.15) | 1.38 (0.045) | 1.31 (0.11) | 1.44 (0.04) | 1.30 (0.11) |
Anxiety | 0.99 (0.71) | 0.99 (0.90) | 1.02 (0.57) | 0.97 (0.50) | 1.01 (0.86) | 0.97 (0.53) | 0.96 (0.39) | 0.96 (0.49) | 0.94 (0.30) | 0.98 (0.72) |
Pain on most days | 2.80 (<0.001) | 3.11 (0.002) | 1.48 (0.19) | 2.45 (0.002) | 3.31 (0.01) | 4.67 (<0.001) | 5.11 (<0.001) | 4.12 (0.001) | 5.61 (<0.001) | 4.31 (<0.001) |
Constant† | 0.10 (0.004) | 0.08 (0.01) | 0.12 (0.01) | 0.29 (0.08) | 0.04 (0.001) | 0.04 (0.002) | 0.03 (<0.001) | 0.02 (<0.001) | 0.01 (<0.001) | 0.02 (<0.001) |
Nagelkerke R2‡ | 43.10% | 55.10% | 39.40% | 36.40% | 53.10% | 55.80% | 54.90% | 55.00% | 58.00% | 54.20% |
AUC§ | 0.853 | 0.91 | 0.835 | 0.821 | 0.917 | 0.925 | 0.915 | 0.921 | 0.93 | 0.92 |
*Minority status (or protected class) individuals are defined as African American, Hispanic American, Native American Indian, Asian/Pacific American or female by the Civil Rights Act of 1991 (Public Law 102–166, USA). Individuals who are disabled, defined by the American with Disabilities Act of 1990, 42 USC § 12101), are also categorised as minority. These categories are used in the US Census.
†Also known as the intercept or β₀, which represents the log-odds of the outcome when all predictor variables are equal to zero.
‡This is a pseudo R² statistic used to assess the goodness of fit of a logistic regression model. Traditional R² (in linear regression) does not apply to logistic regression, so Nagelkerke R² provides an adjusted measure of how well the independent variables explain the variability in the dependent variable.
§AUC, which The AUC quantifies the probability that a randomly chosen positive instance is ranked higher by the model than a randomly chosen negative instance. It ranges from 0 to 1, where: AUC=0.5 suggests no discrimination (random guessing), AUC<0.5 indicates worse-than-random classification and AUC=1 represents perfect classification.
AUC, area under the curve.