• Sulamunn R M Coleman, Diann E Gaalema, Tyler D Nighbor, Allison A Kurti, Janice Y Bunn, and Stephen T Higgins.
• Vermont Center on Behavior and Health, University of Vermont, 1 South Prospect Street, MS 482, Burlington, VT 05401-1419, United States of America; Department of Psychiatry, University of Vermont, 1 South Prospect Street, MS 482, Burlington, VT 05401-1419, United States of America. Electronic address: sulamunn.coleman@uvm.edu.
• Prev Med. 2019 Nov 1; 128: 105853.

AbstractAlthough U.S. college graduates are at relatively low risk for smoking, 12-15% of U.S. smokers (~8 million people) are college graduates. Few studies have examined smoking risk among college graduates. To address that gap, the present study examined smoking risk among U.S. college graduates and those who did not graduate from college in a nationally representative sample of adults (National Survey on Drug Use and Health 2011-2017, n = 202,137). We examined smoking risk in association with well-established risk factors: alcohol abuse/dependence, drug abuse/dependence, mental illness, age, sex, race/ethnicity, and poverty status, using group contrasts and Classification and Regression Tree (CART) modeling. Smoking prevalence among U.S. college graduates and non-graduates was 10% and 26%, respectively. College graduates initiated any smoking and daily smoking at a later age and were lighter smokers than smokers who did not graduate college. Within college graduate and non-graduate groups, prevalence rates varied by orders of magnitude across different risk-factor profiles (ranges = 3-37% and 14-73% among graduates and non-graduates, respectively). Past year drug abuse/dependence was a robust predictor of smoking prevalence in both populations. For college graduates, past year alcohol abuse/dependence and mental illness were stronger predictors of smoking compared to those who did not graduate college, for whom race/ethnicity and age were stronger predictors. Overall, smoking risk increases to surprisingly high levels, even among college graduates, when select risk factors co-occur, particularly psychiatric conditions. Socio-demographic risk factors appear to be less robust predictors of smoking risk among college graduates relative to those who did not graduate college.Copyright © 2019 Elsevier Inc. All rights reserved.

24 keywords...

hide…