背景:评估含糖饮料(SSB)税收通常依赖于模拟模型。我们以德国为例,评估了对SSB税收的反应的假设如何影响与2型糖尿病(T2DM)相关的预计体重变化以及随后的健康和经济影响。
方法:在主要分析中,我们使用边际价格弹性(PE)估算了德国按年龄和性别对SSB征收20%增值税时能量摄入的变化,并应用能量均衡模型预测体重变化.然后,我们量化了几个假设修改的影响:SSB自身PE调整为消费(M1)/基于替代荟萃分析(M2);SSB消费调整为漏报(M3);通过边际替代(M4a)或调整(M4b)交叉PE/卡路里变化的百分比(M4c)。我们还评估了替代税率为10%(S1)或30%(S2)的情况,包括果汁(S3)。我们计算了每个修改和情景的超重和肥胖率。我们模拟了对T2DM的影响,相关的医疗保健费用,和残疾调整寿命年(DALYs)在2011年德国成年人口的一生中使用马尔可夫模型。数据包括官方人口统计,全国调查,和荟萃分析。
结果:在德国征收20%的增值税可以使肥胖的男性和女性人数减少210,800[138,800;294,100]和80,800[45,100;123,300],分别。在人口的一生中,这将导致适度的T2DM相关的健康和经济影响(避免了76,700个DALYs[42,500;120,600];节省了23.7亿欧元[1.33;3.71]的成本).政策影响因修改而变化很大(均为DALYs):(M1)94,800[51,500;150,700];(M2)164,200[99,500;243,500];(M3)52,600[22,500;91,100];(M4a)-18,100[-111,500;68,300];(M4b)25,800与修改有关的政策影响的可变性类似于替代政策方案之间的可变性(全部避免了DALYs):(S1)26,400[9,300;47,600];(S2)126,200[73,600;194,500];(S3)342,200[234,200;430,400]。
结论:在SSB征税下预测的体重减轻对研究人员由于数据限制而经常需要的假设很敏感。因为这种可变性传播到对健康和经济影响的估计,在决策中使用结果时,应考虑由此产生的结构不确定性。
BACKGROUND: Evaluating sugar-sweetened beverage (SSB) taxation often relies on simulation models. We assess how assumptions about the response to SSB taxation affect the projected body weight change and subsequent health and economic impacts related to type 2 diabetes mellitus (T2DM) using Germany as an example.
METHODS: In the main analysis, we estimated changes in energy intake by age and sex under a 20% value-added tax on SSBs in Germany using marginal price elasticities (PE) and applied an energy equilibrium model to predict body weight changes. We then quantified the impact of several assumption modifications: SSB own-PE adjusted for consumption (M1)/based on alternative meta-analysis (M2); SSB consumption adjusted for underreporting (M3); substitution via marginal (M4a) or adjusted (M4b) cross-PE/as % of calorie change (M4c). We also assessed scenarios with alternative tax rates of 10% (S1) or 30% (S2) and including fruit juice (S3). We calculated overweight and obesity rates per modification and scenario. We simulated the impact on T2DM, associated healthcare costs, and disability-adjusted life years (DALYs) over the lifetime of the 2011 German adult population with a Markov model. Data included official demographics, national surveys, and meta-analyses.
RESULTS: A 20% value-added tax in Germany could reduce the number of men and women with obesity by 210,800 [138,800; 294,100] and 80,800 [45,100; 123,300], respectively. Over the population\'s lifetime, this would lead to modest T2DM-related health and economic impacts (76,700 DALYs [42,500; 120,600] averted; €2.37 billion [1.33; 3.71] costs saved). Policy impacts varied highly across modifications (all in DALYs averted): (M1) 94,800 [51,500; 150,700]; (M2) 164,200 [99,500; 243,500]; (M3) 52,600 [22,500; 91,100]; (M4a) -18,100 [-111,500; 68,300]; (M4b) 25,800 [-31,400; 81,500]; (M4c) 46,700 [25,300; 77,200]. The variability in policy impact related to modifications was similar to the variability between alternative policy scenarios (all in DALYs averted): (S1) 26,400 [9,300; 47,600]; (S2) 126,200 [73,600; 194,500]; (S3) 342,200 [234,200; 430,400].
CONCLUSIONS: Predicted body weight reductions under SSB taxation are sensitive to assumptions by researchers often needed due to data limitations. Because this variability propagates to estimates of health and economic impacts, the resulting structural uncertainty should be considered when using results in decision-making.