Community Level Correlates with Adolescent Smoking - University of Minnesota, School of Public Health
Smoking prevalence varies across communities, but factors
contributing to differences in smoking by community are not well
understood. To learn more about
community effects on smoking prevalence, we examined how community level
characteristics based on census data relate to adolescent smoking using data
from the Minnesota Adolescent Community Cohort (MACC) study.
The MACC study utilized a unique sampling frame. It divided the state into 129 group-level
units based on geographical and political (GPU) boundaries thought to reflect
local tobacco control environments. Then
60 GPUs were selected by stratified random sampling. From each selected GPU, 60 adolescents ages
12-16 were randomly selected to establish the cohort needed to study the effect
of the tobacco control environment on smoking trajectory. Smoking status information was collected from
each participant every 6 months. Tobacco control information for each GPU (such
as school policies, ordinances and their enforcement, and media exposure) was
collected annually.
Preliminary results from baseline data show considerable
variability in census-derived measures among GPUs. The mean across all coefficients of variation
(CV) for census measures was 39%. Levels of CV ranged from 6% (%
English-speaking) to 66% (% with graduate degree). Adolescent past month smoking also varied
significantly by community (CV=41%). The
proportion of adolescents who smoked in the past 30 days was statistically
significantly correlated with GPU-level characteristics, including % urban
(-0.26), % less than high school degree (0.26), % bachelor's degree (-.041), %
graduate degree (-0.40), median housing value (-0.36) and median income
(-0.30).
These findings suggest that community-level characteristics
may play an important role in adolescent smoking.
For more information, contact: Lindsey
Fabian, MPH
Phone: 612-626-9991
Email: fabian@epi.umn.edu
Obj 1
Rev 05-24-07
|
