TY - JOUR
T1 - The causal nature of the association between neighborhood deprivation and drug abuse
T2 - A prospective national Swedish co-relative control study
AU - Kendler, K. S.
AU - Ohlsson, H.
AU - Sundquist, K.
AU - Sundquist, J.
PY - 2014/9
Y1 - 2014/9
N2 - Background Risk for drug abuse (DA) is strongly associated with neighborhood social deprivation (SD). However, the causal nature of this relationship is unclear. Method Three Swedish population-based cohorts were followed up over 5 years for incident registration of DA in medical, legal or pharmacy records. In each cohort, we examined the SD-DA association, controlling carefully for individual socio-economic status (SES) with multiple measures, in the entire sample and among pairs of first cousins, paternal and maternal half-siblings, full siblings and monozygotic (MZ) twins discordant for SD exposure. The number of informative relative pairs ranged from 6366 to 166 208. Results In all cohorts, SD was prospectively related to risk for incident DA. In relative pairs discordant for SD exposure, the SD-DA association was similar to that seen in the entire population in cousins, half-siblings, full siblings and MZ twins. Eliminating subjects who were residentially unstable or had DA in the first two follow-up years did not alter this pattern. When divided by age, in the youngest groups, the SD-DA association was weaker in siblings than in the entire population. Conclusions Across three cohorts, controlling for individual SES and confounding familial factors, SD prospectively predicted risk for incident DA registration. These results support the hypothesis that the SD-DA association is in part causal and unlikely to result entirely from personal attributes, which both increase risk for DA and cause selection into high SD environments. At least part of the SD-DA association arises because exposure to SD causes an increased risk of DA.
AB - Background Risk for drug abuse (DA) is strongly associated with neighborhood social deprivation (SD). However, the causal nature of this relationship is unclear. Method Three Swedish population-based cohorts were followed up over 5 years for incident registration of DA in medical, legal or pharmacy records. In each cohort, we examined the SD-DA association, controlling carefully for individual socio-economic status (SES) with multiple measures, in the entire sample and among pairs of first cousins, paternal and maternal half-siblings, full siblings and monozygotic (MZ) twins discordant for SD exposure. The number of informative relative pairs ranged from 6366 to 166 208. Results In all cohorts, SD was prospectively related to risk for incident DA. In relative pairs discordant for SD exposure, the SD-DA association was similar to that seen in the entire population in cousins, half-siblings, full siblings and MZ twins. Eliminating subjects who were residentially unstable or had DA in the first two follow-up years did not alter this pattern. When divided by age, in the youngest groups, the SD-DA association was weaker in siblings than in the entire population. Conclusions Across three cohorts, controlling for individual SES and confounding familial factors, SD prospectively predicted risk for incident DA registration. These results support the hypothesis that the SD-DA association is in part causal and unlikely to result entirely from personal attributes, which both increase risk for DA and cause selection into high SD environments. At least part of the SD-DA association arises because exposure to SD causes an increased risk of DA.
KW - Causal inference
KW - co-relative design
KW - drug abuse
KW - social deprivation
KW - socio-economic status
UR - http://www.scopus.com/inward/record.url?scp=84904974936&partnerID=8YFLogxK
U2 - 10.1017/S0033291713003048
DO - 10.1017/S0033291713003048
M3 - Article
C2 - 25055172
AN - SCOPUS:84904974936
SN - 0033-2917
VL - 44
SP - 2537
EP - 2546
JO - Psychological Medicine
JF - Psychological Medicine
IS - 12
ER -