Cancer in Asian Americans in New Jersey: The influence of residential histories in Asian enclaves

Background: Asian Americans (AsA) are the fastest growing population in the United States (US) and cancer is the leading cause of death among AsA in the US. However, cancer incidence and outcomes vary across AsA subgroups. Residence in ethnic enclaves has been associated with both improved and adverse cancer-related outcomes, but findings have been inconsistent from studies of similar cultural groups living in ethnic enclaves in different geographic areas. Ethnic enclave residence may influence cancer outcomes through social, cultural, and structural mechanisms, including access to culturally concordant care, social support, and neighborhood resources. Objective and Specific Aims: This dissertation examined associations between residential history in Asian enclaves and cancer outcomes in Asian Indian or Pakistani, Chinese, Filipino, Korean, Vietnamese, and Japanese AsA New Jersey (NJ) residents. Specific aims were to: (1) examine incidence and trends of breast, thyroid, and lung cancers in AsA subgroups by nativity and sex; (2) characterize residential patterns in Asian enclaves in AsA diagnosed with breast, thyroid, or lung cancer; and (3) assess the association between residential history in Asian enclaves and survival for AsA diagnosed with female breast cancer and lung cancer. Methods: Cancer data were obtained from the New Jersey State Cancer Registry (NJSCR). We selected records from first primary diagnoses among Asian Indian or Pakistani, Chinese, Filipino, Korean, Vietnamese, and Japanese for invasive cancers of the breast (female), thyroid (female), and lung (male and female) from 2006-2017. For Aim 1, we used logistic regression and multiple imputation to impute missing nativity. Age-adjusted incidence rates (AAIR) were calculated per 100,000 using the 2000 US Standard Population. For Aim 2, probabilistic matching was used to determine match status between LexisNexis residential history data and patient addresses from the NJSCR patient contact studies and the registry data management system (SEER*DMS). Each residential census tract was assigned an Asian enclave status (quintiles, Q) based on definitions from the ENCLAVE study, and sequence analysis, using R package TraMineR, was applied to characterize time spent across Asian enclave quintiles. For Aim 3, residential histories from LexisNexis were used to classify whether individuals had ever (versus never) lived in an Asian enclave from diagnosis through the end of follow-up. Cox proportional hazards models estimated hazard ratios and 95% confidence intervals for all-cause and cancer-specific mortality by AsA subgroup and cancer site, adjusted for neighborhood socioeconomic status, insurance status, marital status, and sex (for lung only). Results: Filipino women had the highest breast cancer AAIR, while Korean women had the highest thyroid cancer AAIR. Lung cancer incidence was highest among Vietnamese men and Japanese women. AAIR for all cancers was higher among US-born compared to non-US-born AsA. In Aim 2, 91% of NJSCR addresses matched to an address in LexisNexis. Sequence analysis showed that the Japanese subgroup on average spent the most time in the lowest enclave neighborhoods (Q1) compared to other AsA race groups, whereas the Korean subgroup spent the most time in the highest enclave neighborhoods (Q5). In Aim 3, analysis of breast and lung cancer mortality showed that AsA who ever lived in an enclave neighborhood had similar risk of mortality as those who never lived in an enclave neighborhood. Conclusion: Cancer incidence rates varied by AsA subgroup and nativity and that residential patterns in Asian enclaves revealed heterogeneity between AsA subgroups. These findings suggest that disaggregating AsA data and the use of residential histories may provide additional insight into cancer patterns and neighborhood influences on health, while underscoring the need to interpret subgroup-specific findings within the constraints of available data and resources.