Consistent with previous studies that linked TB and poverty
[12–14], this study found higher rates of TB among U.S.-born and foreign-born persons living in areas identified by the U.S. Census 2000 as having low income, high crowding, less education, and high unemployment. However, this SES gradient was much less steep among the foreign-born, as evidenced by the large differences in TB rates across SES levels among the U.S.-born, and the comparatively smaller differences among the foreign-born. Compared to U.S.-born persons, rates of TB in foreign-born persons were high at all SES levels for all variables.
Two previous studies linked TB surveillance data with ZIP Code-level measures of SES from the U.S. Census 1990. One study, based on data from New Jersey, reported that foreign-born persons with TB resided in areas of higher SES than U.S.-born persons with TB
. The other study, which linked national TB surveillance data to ZIP Code SES measures, found that controlling for SES reduced racial disparities in TB rates more among U.S.-born than foreign-born persons
The findings of this study support the hypothesis that TB rates among the foreign-born are more strongly influenced by experiences in their countries of origin than by their environments in the United States
[5–7]. This study found that TB rates among foreign-born persons from the five most common countries of origin reflected the TB rates in their respective countries, both in relative magnitude and rank order: persons from the Philippines and Vietnam had the highest rates, followed by persons from India and China, with persons from Mexico having the lowest rates
In addition, although the crowding variable had the strongest association with higher rates of TB in both foreign-born and U.S.-born populations, the association was weaker in the foreign-born. Household crowding can be considered both an indicator of low SES and a mechanism for TB transmission
[12, 20–22]. The weaker association between crowding and TB rates in foreign-born persons suggests a smaller role for recent transmission. Studies based on the genotyping of Mycobacterium tuberculosis strains have also concluded that TB among foreign-born persons is less frequently related to recent transmission than among the U.S.-born
[6, 23, 24].
The association of SES with TB rates differed by country of origin. Rates among persons from Asia showed a relatively strong gradient with declining SES. In contrast, differences in TB rates by SES were much smaller among Mexicans, whose TB rates were highest in the least crowded quartiles and lowest in the most crowded. This discordance could be partly explained by differences in family structure, as other research has shown that foreign-born persons from Mexico often have larger households than other foreign-born populations, reflecting a higher proportion of married-couple families and a lower proportion of householders age 65 and older
. On the other hand, perception of crowding varies by culture; compared to Anglo-Americans and African-Americans, persons of Mexican and Vietnamese heritage have been shown to have a lower perception of crowding at every level of household density, regardless of income
. Future studies could explore the association of TB rates, SES, and country of birth in different regions of the United States, based on immigrant settlement patterns, which might shed light on these differences
This study has a number of limitations, including the use of census ZCTA data based on ZIP Code, the only local geographic indicator available from the national TB surveillance database. ZIP Codes are used primarily by the U.S. Postal Service for delivering mail and are not necessarily reflective of neighborhoods; they are larger than census tracts and may encompass socioeconomically diverse areas
. Previous studies have found that SES measures at the levels of census tract and census block are more consistent measures of health disparities than SES measures at the ZIP Code-level
. In addition, U.S. Census 2000 ZCTAs are not always equivalent to ZIP Codes, so it is possible that some of the NTSS case ZIP Codes are incorrectly linked with ZCTA census data. Finally, while almost 96% of TB records of foreign-born cases had ZCTA matches, only about 75% of TB records of U.S.-born cases had matches; this could lower the rate estimates and affect the rate ratios in unknown ways. These discrepancies should be a strong stimulus for public health surveillance systems to switch from ZIP Codes to the collection of block group or census tract information.
An additional limitation is that measures of SES were calculated overall by ZCTA, not stratified on nativity, because this level of detail was not available. To the extent that this caused differential misclassification of SES, this could account for some or all of our findings.
A final limitation is the use of a static population denominator to estimate average TB rates over the ten-year period considered. These years were chosen to account for changes in the population before and after the 2000 U.S. Census.
The findings that (1) ZIP Code SES has a smaller impact on TB rates among foreign-born persons, and (2) this impact varies by country of origin, have implications for local TB prevention and control and for future research. Prevention and research efforts that focus only on identification and treatment of LTBI among low-SES persons or in low-SES areas will overlook or underserve some foreign-born populations. For example, TB control programs use enablers and incentives to promote treatment completion, which has historically been low
. To date, most studies of ways to improve treatment completion have been done in low-income groups and few have focused on foreign-born persons
[30–36]; it is not known whether the impact of incentives and enablers varies by SES or country of origin. In addition, previous research has suggested that, depending on country of origin, some foreign-born populations are more likely than U.S.-born persons to receive TB care exclusively from private physicians
. Others have identified neighborhood disadvantage as an independent predictor of health care access, even after controlling for individual-level SES
. Such differences in access and care seeking highlight the need for collaboration between health departments and private providers, particularly with regard to the diagnosis and treatment of LTBI among the foreign-born.