Background/Purpose: The reasons underlying substantial variation in lupus nephritis (LN) deaths by race/ethnicity remain largely unknown. There are no large population-based studies focused on the potential impacts of interactions between urbanization and various exposures, such as race/ethnicity, uninsured rates, and household income, on LN mortality. We hypothesize that the effect of urbanization on LN mortality varies according to the level of other confounding factors (effect modifier).

Methods: From the CDC-WONDER and the US Census Bureau database, we obtained data on LN deaths (total 8,899) by urbanization [large central metro, large fringe metro, medium metro, small metro, and nonmetro (micropolitan + noncore)],  race/ethnicity [Hispanic, non-Hispanic (NH) white (white), NH-black (black), NH-Asian/Pacific Islander (A/PI), and NH-American Indian/Alaska native (AI/AN)], uninsured rates (≥15% vs. < 15%), income groups (< $50,000 and ≥$50,000) from 1999 through 2019. We performed multiple logistic regression analyses by urbanization, adjusting for race/ethnicity, uninsured rates, and household income with interactions in the model to identify any potential effect modifications.

Results: We found the effect of urbanization was not constant for each confounder and varied according to certain levels of race/ethnicity and income groups. When stratified by race/ethnicity and income groups, whites in medium metro with a household income < $50,000 experienced greater odds of LN death (OR 1.2, 95% CI: 1.1–1.4) compared to large central metro areas. NH-blacks in large fringe metro (OR 0.87, 95% CI 0.77-1.00) and small metro areas (OR 0.60, 95% CI 0.41-0.88) with a household income ≥$50,000 experienced a protective effect, while mortality odds were nonsignificant for blacks in counties with household incomes < $50,000. Among Hispanics, a protective effect was seen in large fringe metro (OR 0.40, 95% CI 0.24-0.68) with a household income < $50,000 and in small metro (OR 0.30, 95% C 0.14-0.63) with a household income ≥$50,000. In A/PI individuals, a protective effect was seen in large fringe metro areas with income ≥$50,000, while those in medium metro areas with income < $50,000 experienced higher odds of LN death compared to those in large central metro areas. Lastly, counties with the median uninsured rate ≥15% had higher adjusted odds of death from LN than counties with uninsured rate < 15%, however, there was no interaction between uninsured rate at the county level with race/ethnicity and income.

Conclusion: Race/ethnicity and household income can modify the effect of urbanization on the risk of death from lupus nephritis. Overall, a median annual household income ≥$50,000 and residence in a large fringe metro (‘suburbs’) was consistently associated with lower odds of death from LN in racial/ethnic minority groups.

ACR Meeting Abstracts - https://acrabstracts.org/abstract/race-ethnicity-and-household-income-at-the-county-level-interacts-with-the-association-of-urbanicity-and-lupus-nephritis-mortality/