烟草控制政策与牙齿缺失:一项双重差分研究
本研究将美国各州卷烟税、零售价格和室内禁烟政策数据与1995-2012年15轮BRFSS个体数据(N=3,840,519)合并,采用协变量调整的双向固定效应线性与泊松模型及异质性稳健双重差分估计。结果显示卷烟税每增加1美元使当前吸烟率降低0.35个百分点(95%CI:-0.56,-0.13),戒烟率增加0.51个百分点(95%CI:0.19,0.82);但税收和价格对牙齿缺失无稳健影响,室内禁烟也无稳健效果。
要点
- 卷烟税每增1美元,当前吸烟率降0.35个百分点,戒烟率升0.51个百分点。
- 卷烟税与价格对牙齿缺失无稳健影响,完全无牙与税收无关。
- 室内禁烟对吸烟和牙齿缺失均无稳健效果,部分点估计为正。
核心结论
- 卷烟税和定价是减少吸烟的有力杠杆,但未能在18年窗口内证明对牙齿缺失的人群层面益处。
- 室内禁烟对吸烟和牙齿缺失均无稳健效果。
- 持续税收暴露与严重牙齿缺失的提示性关联值得在更长周期评估。
研究设计与样本
- 研究合并州政策数据与1995-2012年15轮行为危险因素监测系统个体数据,总样本N=3,840,519。
- 采用协变量调整的双向固定效应线性与泊松模型估计风险差和风险比,并辅以异质性稳健双重差分估计。
- 平行趋势假设通过实证和因果图评估。
主要发现
- 卷烟综合税每增加1.00美元,当前吸烟率降低0.35个百分点(95%CI:-0.56,-0.13),戒烟率增加0.51个百分点(95%CI:0.19,0.82)。
- 持续暴露下吸烟减少持续存在(10年平均税收:-0.47个百分点;95%CI:-0.95,0.01),零售价格显示类似效果。
- 完全无牙与税收在所有协变量调整模型中无关(当年税收:0.08个百分点;95%CI:-0.14,0.31)。
- 缺失≥6颗牙仅显示弱且无统计学意义的保护关联,累积暴露最强(10年平均税收:-0.44个百分点;95%CI:-1.17,0.30)。
- 餐馆和商场/杂货店室内禁烟与更低吸烟或更少牙齿缺失无关,点估计小且部分为正(如餐馆禁烟与当前吸烟:+0.11个百分点;95%CI:-0.34,0.55),商场/杂货店禁烟未通过阴性对照测试。
临床怎么用
- 口腔科医生可支持卷烟税和定价作为控烟策略,但不应预期其短期内在人群层面减少牙齿缺失。
- 对患者进行烟草控制宣教时,应聚焦吸烟对口腔健康的直接危害而非仅依赖税收政策。
- 室内禁烟政策对牙齿缺失的防护作用缺乏证据,临床建议需个体化。
局限性
- 研究仅覆盖18年窗口,可能不足以观察牙齿缺失的长期效应。
- 持续税收暴露与严重牙齿缺失的提示性关联未达统计学意义,需更长周期验证。
- 室内禁烟部分点估计为正且阴性对照测试失败,提示潜在偏倚。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
This study estimated the effects of state cigarette taxes, retail cigarette prices, and indoor smoking bans on smoking behavior and tooth loss among US adults. State policy data were merged with individual data from 15 cycles of the Behavioral Risk Factor Surveillance System (1995-2012; N = 3,840,519). Covariate-adjusted 2-way fixed effects linear and Poisson models estimated risk differences and risk ratios for contemporaneous and cumulative (10-y average) exposures, complemented by heterogeneity-robust difference-in-differences estimators; the parallel trends assumption was assessed empirically and with causal diagrams. A $1.00 increase in the combined cigarette tax reduced current smoking by 0.35 percentage points (95% CI: -0.56, -0.13) and increased quitting by 0.51 percentage points (95% CI: 0.19, 0.82); the smoking reduction persisted under sustained exposure (10-y average tax: -0.47 percentage points; 95% CI: -0.95, 0.01), and retail prices showed similar effects. In contrast, taxes and prices showed no robust effect on tooth loss. Complete edentulism was unrelated to taxes in every covariate-adjusted specification (current-year tax: 0.08 percentage points; 95% CI: -0.14, 0.31), and loss of ≥6 teeth showed only a weak, statistically nonsignificant protective association that was strongest for cumulative exposure (10-y average tax: -0.44 percentage points; 95% CI: -1.17, 0.30). Indoor smoking bans (restaurant and mall/grocery) were not associated with lower smoking or less tooth loss; point estimates were small and several were positive (e.g., restaurant ban and current smoking: +0.11 percentage points; 95% CI: -0.34, 0.55), and the mall/grocery ban failed a negative control test. Cigarette taxation and pricing remain powerful levers for reducing smoking, but a population-level benefit for tooth loss could not be demonstrated within this 18-y window, and indoor smoking bans showed no robust effect on either outcome; the suggestive association between sustained tax exposure and severe tooth loss warrants evaluation over longer horizons.
引用
- Baumeister SE, Chaffee BW, Holtfreter B, et al. Tobacco Control Policies and Tooth Loss: A Difference-in-Differences Study. J Dent Res. 2026. PMID: 42924910. doi:10.1177/00220345261489486
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。