Nicotine dependence and smoking quantity are both robustly associated with the CHRNA5-A3-B4 gene cluster in the 15q25 region, and SNP rs16969968 in particular. The purpose of this paper is to use structural equation modeling techniques (SEM) to disentangle the complex pattern of relationships between rs16969968, nicotine quantity (as measured by the number of cigarettes an individual smokes per day; CPD) and nicotine dependence (as measured by the Fagerström Test for Nicotine Dependence; FTND). CPD is an indicator, but also a potential cause, of FTND, complicating the interpretation of associations between these constructs and requires a more detailed investigation than standard GWAS or general linear regression models can provide. FTND items and genotypes were collected in four samples, with a combined sample size of 5,373 respondents. A mega-analysis was conducted using a multiple group SEM approach to test competing hypotheses regarding the relationships between the SNP rs16969968, FTND and CPD. In the best fitting model, the FTND items loaded onto two correlated factors. The first, labeled “maintenance,” assesses the motivation to maintain constant levels of nicotine through out the day. The second was labeled “urgency” as its items concern the urgency to restore nicotine levels after abstinence. We focus our attention on the “maintenance” factor, of which CPD was an indicator. The best fitting model included a negative feedback loop between the Maintenance factor and CPD. Accordingly, the motivation to maintain higher levels of nicotine increased the quantity of nicotine consumed, which subsequently decreases the maintenance motivation. The fact that the Maintenance-CPD feedback model fits the data best implies that there are at least two biological pathways that lead from rs16969968 to smoking behaviors. The model is consistent with a supply and demand system, which allows individuals to achieve a homeostatic equilibrium for their nicotine concentration.

尼古丁依赖和吸烟量都与CHRNA5-A3-B4 密切相关15q25 区域的基因簇，特别是 SNP rs16969968。本文的目的是使用结构方程建模技术 (SEM) 来解开 rs16969968、尼古丁量（以个人每天吸烟的数量；CPD）和尼古丁依赖（通过Fagerström 尼古丁依赖测试；FTND）。CPD 是 FTND 的一个指标，但也是一个潜在原因，它使解释这些构造之间的关联变得复杂，并且需要比标准 GWAS 或一般线性回归模型所能提供的更详细的调查。在四个样本中收集了 FTND 项目和基因型，总样本量为 5,373 名受访者。使用多组 SEM 方法进行了一项大型分析，以测试有关 SNP rs16969968、FTND 和 CPD 之间关系的竞争假设。在最佳拟合模型中，FTND 项目加载到两个相关因素上。第一个，标记为“维持”，评估在一天中保持恒定尼古丁水平的动机。第二个被标记为“紧迫性”，因为它的项目涉及在禁欲后恢复尼古丁水平的紧迫性。我们将注意力集中在“维护”因素上，其中 CPD 是一个指标。最佳拟合模型包括维护因子和 CPD 之间的负反馈循环。因此，维持较高水平尼古丁的动机增加了消耗的尼古丁量，这随后降低了维持动机。维护-CPD 反馈模型最适合数据的事实意味着至少有两条生物途径从 rs16969968 导致吸烟行为。该模型与供需系统一致，该系统允许个人实现其尼古丁浓度的稳态平衡。