Date of Award

January 2013

Document Type

Open Access Thesis

Degree Name

Master of Public Health (MPH)


School of Public Health

First Advisor

Linda M. Niccolai

Second Advisor

James L. Hadler


Background: Surveillance by the FoodNet program has shown that Salmonella, unlike other foodborne illnesses, has increased during the past fifteen years within the United States. Little work has been done to examine the role of socioeconomic factors in incidence of Salmonella. Neighborhood poverty level, measured as the percentage of people living below the poverty line, at the census tract level has been accepted as a useful indicator of socioeconomic status. This study utilized geographic information system (GIS) technology and census tract data to examine the variation in Salmonella incidence among census tracts in Connecticut. The objectives of this study were to examine: 1) the relationship between the incidence of all Salmonella serotypes and neighborhood poverty level (overall, by age group, and over time), and 2) whether the association between the incidence of Salmonella and neighborhood poverty level differs by serotype for the four most common serotypes seen in Connecticut from 2000-2011.

Methods: There were 5204 of 5484 (94.9%) reported cases of Salmonella from 2000-2011 that were geocoded and categorized into the appropriate poverty level. Census tracts were divided into four different levels of neighborhood poverty level representing the percentage of individuals living below the poverty line using information from the 2000 and 2010 censuses and the American Community Survey: 0 - 4.9%, 5 - 9.9%, 10 - 19.9%, and ≥20%. Age-adjusted incidence rates were calculated for each poverty level for the overall time period and for 2000-2005 and 2006-2011. We further stratified the data by age group (<5, 5-9, and >10 years old) and examined age-specific incidence rates overall and by time period. Age-adjusted incidence rates by neighborhood poverty level were also examined for the four most prevalent Salmonella serotypes in Connecticut for the overall time period and then for 2000-2005 and 2006-2011.

Results: There was a clear gradient for the association between neighborhood poverty level and incidence of Salmonella in Connecticut from 2000-2011; incidence increased with decreasing neighborhood poverty level. Using the 0 - 4.9% poverty group as the reference, the age-adjusted incidence rate ratios for 2000-2011 were 0.94 for the 5 - 9.9% group, 0.92 for the 10 - 19.9% group, and 0.81 for the ≥ 20% group. This trend persisted over time though there were differences according to age group and serotype. Those individuals younger than five years old and those with Salmonella heidelberg exhibited the opposite gradient (increasing incidence with increasing neighborhood poverty level). The gradient of lower incidence with increasing poverty level seen across all Salmonella for the entire 2000-2011 time period was also present for those with S. newport, S. enteriditis, and individuals greater than five years old. No association with neighborhood poverty level was seen for S. typhimirium.

Conclusions: Salmonella incidence overall for 2000-2011 increased with decreased neighborhood poverty. An exception was children less than five years for whom incidence increased with increased neighborhood poverty. Different Salmonella serotypes exhibited different trends in incidence related to neighborhood poverty level. We believe that these findings could be utilized in prevention efforts and in designing interventions geared toward specific populations by taking into account age, neighborhood poverty level, and the different trends for the four most prevalent Salmonella serotypes. Research into possible explanations for different incidence rates among serotypes by poverty level could provide additional insight.


This is an Open Access Thesis.