INTRODUCTION: Racial residential segregation has been shown to affect the absolute levels of racial disparities in a wide variety of health outcomes in the USA but it is not known whether changes in segregation also influence these racial health disparities. This study examines the relationship between changes in racial residential segregation over four decades (1980-2020) and trends in racial disparities in early mortality (under age 65) rates among non-Hispanic Black and non-Hispanic White persons across a wide range of health outcomes in 220 metropolitan statistical areas (MSAs) during the period 2001-2018. METHODS: Using the CDC WONDER Underlying Cause of Death database, we derived annual estimates of race-specific death rates and rate ratios for each MSA. We used latent trajectory analysis to examine the relationship between the level of segregation and changes in segregation over time in an MSA and trends in death rate disparities in that MSA. RESULTS: The trajectory analysis resulted in a linear, three group model in which trajectory Groups 1 and 2 had decreasing trends in the ratios of Black to White death rates over time while in Group 3, the disparity remained almost constant over time. Increases in the level of segregation in an MSA from 1980 to 2000 were significantly associated with the likelihood that the MSA was in Group 3 and experienced no improvement in racial health disparities in mortality over time. CONCLUSION: This paper provides new evidence that changes in segregation are related to trends in racial health disparities in mortality rates over time.
