(EA03-07-26) Impact of Changing Prevalence of Smoking, Alcohol, Overweight-Obesity, and Diabetes on Pancreatic Cancer Incidence in the United States, 1997–2022

Name of the scientific conference where the work was recently presented: ASPO

Date on which the work was previously presented at the scientific conference listed above: April 13, 2026

Objectives: Pancreatic cancer (PC) rates in the United States (US) have increased in the past two decades. We estimated the proportion of PC attributable to smoking, alcohol intake, overweight-obesity, and diabetes, and assessed how changes in the prevalence of these risk factors influenced US PC incidence trends during 2007 to 2022.

Methods: Data from the Pooling Project of Prospective Studies of Diet and Cancer ((18 US cohorts, n=2,558,335 participants, n=15,836 cases) were used to estimate multivariable-adjusted hazard ratios (HRs) for PC with smoking (former/current vs never), alcohol use ( >1 vs 0–1 drinks/day), overweight-obesity (BMI ≥25 vs < 25 kg/m²), and diabetes (yes vs no). Annual population attributable fractions (PAFs) were calculated using these HRs and risk factor prevalence estimates from the National Health Interview Survey from 1997 to 2012 (assuming a 10-year latency between exposure and pancreatic cancer incidence). PAFs were applied to US Cancer Statistics data to estimate average annual percent changes (AAPCs) in age-standardized PC incidence attributable and not attributable to each risk factor.

Results: From 2007 to 2022, PAF estimates for BMI ≥25 kg/m² (10.4% to 11.8%) and diabetes (4.0% to 6.5%) increased, while alcohol intake was stable (1.1%), and ever-smoking (15.2% to 12.8%) decreased. PC incidence significantly increased (AAPC, 0.83, 95% confidence interval 0.71, 0.95, P-value < 0.05) among people ≥50-years-old. Incidence of smoking-attributable PC declined (AAPC -0.29, P-value < 0.05) during 2007 to 2022. While PC incidence attributable to overweight/obesity, diabetes, and alcohol significantly (P-value < 0.05) increased (AAPCs, 1.62, 4.28, and 1.40, respectively), these risk factors did not substantially impact trends as the PC rates unrelated to these risk factors increased (AAPC, 0.72, 0.63, and 0.82, respectively; P-values < 0.05), were close to the overall PC rate and higher than the risk factor- attributable rates. The results were similar considering 5- and 15-year latency exposure prevalence and by sex and race-ethnicity.

Conclusions: While smoking, high BMI, diabetes, and alcohol intake contributed to PC incidence, trends in these risk factors did not explain recent increases in PC incidence.

Funding Sources: NCI Intramural Research Program of the NIH