The effects of area-level deprivation on colorectal cancer incidence at the small area-level in Pennsylvania from 2008 to 2017

Abstract

Background

Colorectal cancer (CRC) is the third most diagnosed cancer in the United States. Area-level deprivation increases CRC risk, but traditional indices and analyses have limitations. Weighted Quantile Sum (WQS) regression and hierarchical Bayesian approaches offer better alternatives for highly correlated indicators and addresses spatial dependencies and reliability issues. The purpose of this study is to identify the most explanatory area-level neighborhood deprivation indicators and investigate its association with CRC incidence using advanced spatiotemporal methods.

Methods

Analyzing 34,250 CRC cases from Pennsylvania between 2008 and 2017, we constructed an area-level neighborhood deprivation index using WQS from 39 block group and census tract level demographic, social, economic, and housing indicators from the US Census Bureau’s American Community Survey five-year pooled estimates. Census tract was used when block group data was unavailable. Spatiotemporal modeling, using hierarchical Bayesian methods, assessed the effect of age, sex, area-level neighborhood deprivation, healthcare access, CRC screening, and rurality on the risk of block-group level CRC incidence.

Results

For the neighborhood deprivation index, we identified nine statistically-significant area-level economic, demographic, and housing-related variables (p < 0.05). Of these, the total count of housing units, median household income, and proportion of the population ages 25 years of older not graduating high school contributed 61 % of the total weight of the index. Area-level neighborhood deprivation significantly predicted CRC risk, with a 1.33-fold increase in incidence for each one-unit increase of the index, adjusted for block group age- and sex-distribution. Access to healthcare, CRCS adherence, and rurality were not significantly associated with the incidence of colorectal cancer.

Conclusion

A WQS-developed area-level neighborhood deprivation index may be useful in identifying small geographic areas at highest risk of CRC incidence. Further research is needed to determine whether key deprivation indicators can direct public health interventions with implications for policy and resource allocation tailored to regional risk profiles.