Date of Award

January 2014

Document Type

Open Access Thesis

Degree Name

Master of Public Health (MPH)


School of Public Health

First Advisor

Amanda Durante


Background: Cardiovascular disease (CVD) is the leading cause of death in the United States and many of these deaths are preventable. Studies have shown that neighborhood-level characteristics may contribute to health outcomes, but no study has yet examined whether neighborhood crime contributes to early mortality from CVD.

Objective: We examined geographic trends in the association between neighborhood crime rates and premature mortality from coronary heart disease (CHD) using New Haven, CT USA as a model city.

Methods: Neighborhoods in New Haven were established by existing census tracts. CHD deaths were identified from the Connecticut Master Death Files and violent crime rates were calculated from the FBI Uniform Crime Reports. We conducted a global ordinary least squares (OLS) analysis and a geographically weighted regression (GWR) analysis to model average years of potential life lost (YPLL) by census tract.

Results: Out of 687 CHD deaths in the city of New Haven from 2005-2010, 319, or 46.4%, are considered premature. The OLS model accounted for 30.8% and the GWR model accounted for 48.6% of the variability in premature deaths from CHD. An increase of 10 violent crimes per 1,000 residents was associated with an average of 2.3 additional years of life lost (p=0.043), while holding other neighborhood factors constant. Moreover, the GWR model predicted a 7-fold disparity in premature CHD mortality across census tracts, ranging from 1.73 YPLL to 12.38 YPLL.

Conclusion: Our findings suggest that neighborhood violent crime rates may contribute to premature death from CHD. Modeling based on geographic variation is a powerful tool to enhance resolution of previously unidentified environmental factors contributing to preventable death from cardiovascular disease.


This is an Open Access Thesis.

Open Access

This Article is Open Access