Purpose

Risk factors for pancreatic cancer include racial/ethnic disparities and smoking. However, risk trajectories by smoking history and race/ethnicity are unknown. We examined the association of smoking with pancreatic cancer by race/ethnicity to generate age-specific incidence estimates by smoking history.

Methods

We modeled pancreatic cancer incidence by race/ethnicity, age, pack-years, and years-quit using an excess relative risk model for 182,011 Multiethnic Cohort participants. We tested heterogeneity of smoking variables and pancreatic cancer by race/ethnicity and predicted incidence by smoking history.

Results

We identified 1,831 incident pancreatic cancer cases over an average 19.3 years of follow-up. Associations of pack-years (p interaction by race/ethnicity = 0.41) and years-quit (p interaction = 0.83) with pancreatic cancer did not differ by race/ethnicity. Fifty pack-years smoked was associated with 91% increased risk (95% CI 54%, 127%) relative to never smokers in the combined sample. Every year quit corresponded to 9% decreased excess risk (95% CI 2%, 15%) from pack-years smoked. Differences in baseline pancreatic cancer risk across racial/ethnic groups (p < 0.001) translated to large differences in risk for smokers at older ages across racial/ethnic groups (65–122 cases per 100,000 at age 70).

Conclusion

Smoking pack-years were positively associated with elevated pancreatic cancer risk. Predicted risk trajectories showed a high impact of smoking cessation at < 65 years. Although we did not identify significant heterogeneity in the association of pack-years or years quit with pancreatic cancer risk, current smoker risk varied greatly by race/ethnicity in later life due to large differences in baseline risk.

中文翻译：

---

吸烟和戒烟改变效应导致胰腺癌风险过高：多种族队列研究

目的

胰腺癌的危险因素包括种族/民族差异和吸烟。然而，吸烟史和种族/民族的风险轨迹尚不清楚。我们按种族/民族检查了吸烟与胰腺癌的关联，以根据吸烟史生成特定年龄的发病率估计值。

方法

我们使用超额相对风险模型，针对 182,011 名多种族队列参与者，按照种族/族裔、年龄、包年数和戒烟年数对胰腺癌发病率进行了建模。我们按种族/民族测试了吸烟变量和胰腺癌的异质性，并根据吸烟史预测了发病率。

结果

我们在平均 19.3 年的随访中发现了 1,831 例胰腺癌病例。包年数（p相互作用，种族/民族 = 0.41）和戒烟年数（p相互作用 = 0.83）与胰腺癌的关联不因种族/民族而异。在合并样本中，与从不吸烟者相比，吸烟 50 包年的风险增加 91%（95% CI 54%、127%）。每戒烟一年，超额风险就会比吸烟年数降低 9%（95% CI 2%，15%）。不同种族/族裔组之间基线胰腺癌风险的差异 ( p  < 0.001) 转化为不同种族/族裔组中老年吸烟者风险的巨大差异（70 岁时每 100,000 人中有 65-122 例）。

结论

吸烟年数与胰腺癌风险升高呈正相关。预测风险轨迹显示 65 岁以下戒烟的影响很大。尽管我们没有发现戒烟年数或戒烟年数与胰腺癌风险之间的显着异质性，但由于基线风险存在巨大差异，当前吸烟者在晚年的风险因种族/民族的不同而存在很大差异。