An exploration of colorectal cancer incidence rates in North Dakota, USA, via structural equation modeling

Int J Colorectal Dis. 2019 Sep;34(9):1571-1576. doi: 10.1007/s00384-019-03352-9. Epub 2019 Jul 16.

Abstract

Purpose: The state of North Dakota has one of the highest incidence rates for colorectal cancer in the USA. Its high incidence rate, coupled with a large variation in incidence rates among counties within the state, makes North Dakota a "natural laboratory" in which to investigate environmental clues to colorectal cancer. We conducted a hypothesis-generating study to explore potential determinants of colorectal cancer in North Dakota.

Methods: We obtained county-specific incidence rates for North Dakota's 53 counties from the statewide cancer registry and corresponding data on county demographic, agricultural, and geophysical features from population-based sources. Candidate demographic/agricultural variables included median household income, population density, colorectal cancer screening rates, average farm size (in acres), and the percent of county fertilized. Geophysical variables included the uranium content of soil, residential radon levels, and source of drinking water (municipal or well water). Statistical analyses were performed via multivariate regression and structural equation modeling.

Results: Colorectal cancer incidence rates across North Dakota counties varied 3-fold. The structural equation model identified a significant role for well water use (p < 0.05). This finding is consistent with studies that implicate well water in colorectal cancer.

Conclusions: Well water contains several agents, e.g., bacteria, disinfection by-products, and nitrates that are potent colorectal carcinogens. Studies of well water use and colorectal cancer risk at the individual level in North Dakota are warranted.

Keywords: Colorectal cancer; Epidemiology; North Dakota; Structural equations.

MeSH terms

  • Colorectal Neoplasms / epidemiology*
  • Geography
  • Humans
  • Incidence
  • Latent Class Analysis*
  • Linear Models
  • North Dakota / epidemiology
  • Risk Factors