Department of Psychology & Neuroscience, Boston College, Chestnut Hill, MA, 02467, USA.
Department of Psychology & Neuroscience, Boston College, Chestnut Hill, MA, 02467, USA.
Int J Drug Policy. 2024 Oct;132:104558. doi: 10.1016/j.drugpo.2024.104558. Epub 2024 Sep 3.
Our goal in this report was to quantify the degree to which opioid prescription rates and socioeconomic correlates of income inequality predicted overdose deaths in the 1055 U.S. Midwest counties. The study follows up a state-level analysis which reported that opioid prescription rates, social capital and unemployment explained much of the variance in opioid overdose death rates (Heyman, McVicar, & Brownell, 2019).
We created a data set that included drug overdose death rates, opioid prescription rates, and correlates of income inequality. Given that the variables of interest varied at the state and county level, multilevel regression was our statistical approach.
From 2006 to 2021, Midwest overdose drug deaths increased according to an exponential equation that closely approximated the equation that describes the increases in overdose deaths for the entire U.S. from 1978 to 2016 (e.g., Jalal et al., 2018). Retail opioid prescription sales increased from 2006 to 2012, but then declined so that by 2017 they were lower than in 2006. The regression analyses revealed that intergenerational income mobility was the strongest predictor of overdose deaths. The other consistently statistically significant predictors were opioid prescription rates, social capital, and unemployment rates. Together these predictors, plus pupil teacher ratios, single parent families, and attending college accounted for approximately 47 % of the variance in overdose death rates each year. In keeping with the decline in opioid prescription rates, the explanatory power of opioid prescription rates weakened over the course of the study.
Overdose deaths increased at a constant exponential rate for the years that it was possible to apply our regression model. This occurred even though access to legal opioids decreased. What remained invariant was the predictive strength of intergenerational income mobility; each year it was the predictor that explained the most variance in overdose deaths.
我们在本报告中的目标是量化阿片类药物处方率和收入不平等的社会经济相关因素对美国中西部 1055 个县的药物过量死亡的影响程度。该研究是对州级分析的跟进，该分析报告称，阿片类药物处方率、社会资本和失业率解释了阿片类药物过量死亡率的大部分差异（Heyman、McVicar 和 Brownell，2019）。
我们创建了一个包含药物过量死亡率、阿片类药物处方率和收入不平等相关因素的数据集合。鉴于感兴趣的变量在州和县两级有所不同，我们采用了多层回归作为统计方法。
从 2006 年到 2021 年，中西部地区的药物过量死亡人数按照一个指数方程增加，该方程非常接近描述 1978 年至 2016 年美国整体药物过量死亡人数增加的方程（例如，Jalal 等人，2018）。零售阿片类药物处方销售从 2006 年到 2012 年增加，但随后下降，到 2017 年，处方销售低于 2006 年。回归分析显示，代际收入流动性是药物过量死亡的最强预测因素。其他始终具有统计学意义的预测因素是阿片类药物处方率、社会资本和失业率。这些预测因素加上师生比、单亲家庭和上大学，每年解释了药物过量死亡率差异的约 47%。与阿片类药物处方率下降一致，阿片类药物处方率的解释力在研究过程中减弱。
在我们能够应用回归模型的几年中，药物过量死亡人数以恒定的指数速度增加。尽管获得合法阿片类药物的机会减少，但情况仍然如此。不变的是代际收入流动性的预测强度；每年它都是解释药物过量死亡差异最大的预测因素。
