TABLE 4—
Effect When Not in Poverty |
Effect When in Poverty |
Differential Effecta of AUD/DUD When in Poverty |
||||
Variable | RD % (SE) | P | RD % (SE) | P | IC % (SE) | P |
Unadjustedb (Ref = neither disorder) | ||||||
Alcohol-use disorder only | 2.9 (0.7) | <.001 | 12.1 (3.1) | <.001 | 9.2 (3.2) | .005 |
Drug-use disorder only | 9.1 (3.0) | .004 | 28.1 (8.0) | <.001 | 19.0 (8.4) | .028 |
Both alcohol- and drug-use disorder | 10.0 (3.2) | .003 | 14.5 (6.4) | .027 | 4.5 (6.6) | .499 |
Adjustedc (Ref = neither disorder) | ||||||
Alcohol-use disorder only | 0.6 (0.5) | .257 | 3.6 (1.5) | .017 | 3.1 (1.6) | .052 |
Drug-use disorder only | 3.0 (1.8) | .092 | 11.4 (5.2) | .03 | 8.8 (5.4) | .112 |
Both alcohol- and drug-use disorder | 2.0 (1.6) | .233 | 2.2 (2.7) | .424 | 0.3 (3.0) | .918 |
Note. AUD/DUD = alcohol-use disorder/drug-use disorder; IC = interaction contrast; RD = risk difference.
Differential effect (i.e., IC) is the difference in RD (e.g., 9.2 = 12.1 – 2.9).
Unadjusted effects can be obtained by taking differences in incidence rates in Table 2.
Adjusted effects represent expected risk differences if each AUD/DUD by poverty strata was fixed to have all of the control variables to be equal.