Albuminuria Risk in Hispanic Populations: Not So Black and White

For over a century, epidemiologic and clinical studies have relied on self-report for race and ethnicity. Subjects have mostly been classified into over-simplified categories such as black, white, and Hispanic. In this issue of Circulation: Cardiovascular Genetics, Peralta et al. report the results from an investigation of the association between albuminuria and ethnicity and race, measured by both self-report and genetic admixture analysis, which provides an estimate of each subject’s unique ancestral heritage (Peralta citation).

The authors chose to analyze albuminuria as the major phenotype because it is an important marker of chronic kidney disease (CKD) and a risk factor for progression of CKD (1, 2). Furthermore, albuminuria has long been recognized as a major risk factor for cardiovascular morbidity and mortality. Mogensen discovered over 25 years ago that albuminuria is associated with increased all-cause mortality in patients with type 2 diabetes (3). Since then, it has been shown that among non-diabetic subjects with hypertension, microalbuminuria is associated with cardiovascular and cerebrovascular events (4), as well as CKD risk (5). In healthier populations, albuminuria is also a risk factor for mortality (6), and for mortality, myocardial infarction, and end-stage renal disease at all levels of glomerular filtration rate (GFR) (7). The link between albuminuria and increased morbidity and mortality is not well understood, but may reflect endothelial dysfunction among multiple vascular beds (8).

The importance of microalbuminuria, typically defined as a urine albumin to creatinine ratio between 30 and 300 mg/g, is debated because microalbuminuria often reverts to normoalbuminuria, and the associated risk for morbidity and mortality is significantly less than for macroalbuminuria (4, 9-11). For this reason, microalbuminuria is conventionally viewed as a risk factor for cardiovascular and chronic kidney diseases only when persistently present. However, such logical application of microalbuminuria as a screening test may not be so straightforward, since the HOPE study demonstrated that albuminuria, even in the upper normoalbuminuria range, confers risk for cardiovascular events and mortality among diabetic, as well as non-diabetic subjects with a history of cardiovascular disease (12).

Specific populations are known to be at increased risk for albuminuria, which is likely due to a combination of environmental, cultural, and genetic factors. The prevalence of albuminuria in the non-diabetic NHANES III population was higher among blacks and Mexican Americans compared to whites, especially those with an estimated GFR less than 60 mL/min/1.73 m²(13). Similar results were seen in the Kidney Early Evaluation Program (KEEP) where American Indians/Alaska Natives, Asians, African Americans, and Hispanics all had increased albuminuria compared to whites (14). Evidence for genetic regulation of albuminuria includes a Finnish study of diabetic probands and their first degree relatives, which found the heritability of albumin excretion rate to be approximately 30% (15). Another study of 96 families found that urinary albumin excretion was heritable among both diabetic and non-diabetic relative pairs (16). Finally, the non-muscle myosin heavy chain gene, MYH9, which has been linked to multiple different renal diseases, is associated with albuminuria in hypertensive African Americans (17).

To assess the association between albuminuria and both genetic ancestry and country of origin, Peralta et al. examined over 1,400 self-reported Hispanic and white control subjects enrolled in the Multi-Ethnic Study of Atherosclerosis (MESA) (Peralta citation). The study measured albuminuria, assessed country of origin, and tested for specific single nucleotide polymorphisms (SNPs), known as ancestry informative markers (AIMs). These AIMs permitted the quantification of each subject’s European, Native American, and African heritage. Consistent with previous studies, Hispanics were more likely to have significant albuminuria compared to whites. Extending these findings with genetic admixture analysis, the investigators found that among self-reported Hispanics, Native American ancestry was associated with increased albuminuria, while European ancestry was associated with lower levels of albuminuria. However, neither association was significant in fully adjusted models.

The investigation broadens our understanding of the link between ethnicity and albuminuria in two important ways. First, by examining both ethnicity and country of origin, Peralta et al. found that the increased risk for albuminuria among Hispanics was dependent on country of origin. Mexican and Central American Hispanics and Dominican Hispanics had significantly higher levels of albuminuria than whites. On the other hand, Puerto Rican Hispanics had similar levels of albuminuria as white Puerto Ricans, though the absence of association between albuminuria and Puerto Rican ancestry could be due to a lack of statistical power. The second interesting finding was the variability in ancestry by country of origin. The prevalence of Native American ancestral markers varied from 11% among Dominican Hispanics to 41% among Mexican and Central American Hispanics, while African ancestry was lowest (7%) among Mexican and Central American Hispanics and highest (39%) among Dominican Hispanics. European ancestry was highest among Puerto Rican Hispanics, a finding that may partly explain the lack of a significant difference in albuminuria between Hispanics and whites of Puerto Rican origin. These results merit consideration in planning future studies that aim to examine the relationship between ethnicity and disease – self-reported ethnicity is likely an oversimplification and may not be sufficient for phenotype associations with ancestral genotypes.

The study by Peralta et al. has a number of strengths. MESA was specifically designed to evaluate subclinical cardiovascular disease, including albuminuria, in a multi-ethnic cohort. The investigators purposefully selected Hispanics and a comparison group (non-Hispanic whites). In addition, the AIMs were chosen expressly to differentiate Caucasian and Native American ancestry. Using a sophisticated genetic admixture analytical strategy, the investigators discovered broad ancestral heterogeneity within Hispanic subgroups, which seems to impact significantly on albuminuria.

A few limitations need to be considered, however. First, the study enrolled relatively few subjects with diabetes (12%), which is a major risk group for complications associated with albuminuria. It is therefore not clear that results can be extrapolated to diabetic patients. Second, although albuminuria is an accepted surrogate marker for cardiovascular and chronic kidney diseases (18), these more important phenotypes were not included in this study. Third, AIMs were not evaluated in the white control group, who may also have complex ancestries. Finally, only 9.5% of subjects had microalbuminuria, and even fewer (2.2%) had more significant macroalbuminuria. Therefore, the genetic associations are based mainly upon albuminuria within the normal range. Although results derived from a more extensive array of albuminuria values would have strengthened the conclusion, as previously mentioned, albuminuria below the microalbuminuria threshold has been associated with increased risk of cardiovascular events in both diabetic and non-diabetic populations (12).

Despite the intriguing findings from this report, the clinical ramifications are not entirely clear. While the study confirms that Hispanics are at greater risk for albuminuria, there are no data to suggest that screening this high-risk group will ultimately reduce the incidence of cardiovascular or renal disease. As genetics technologies become more comprehensive, less expensive and thereby more amenable to development of personalized databases, it is likely that compilation of one’s risk alleles, including ancestral markers, may be part of such databases. In the interim, this study suggests that it may be helpful to attribute risk using more easily attainable information, such as country of origin, which may function as a crude surrogate for ancestral genetic variants.

In contrast to clinical applications, the current report has clearer implications for research. Future studies should not consider Hispanics to be a homogenous group, but should collect data on country of origin and even AIMs, if possible. If intervention trials for albuminuria in Hispanic populations are to be contemplated, the study by Peralta et al. supports oversampling of subjects of Mexican, Central American and Dominican origin, with the goal of enrichment for Native American and African risk alleles. From a basic science perspective, a major goal of genetic studies has been the detection of specific risk/protective genes. The identification and characterization of disease-associated ancestral variants might therefore lead to such genes, with potentially important implications for both disease pathophysiology and improved clinical risk stratification. Until such data are generated, the findings from this study are merely broad statistical associations, but certainly a step in the right direction. As techniques such as genetic admixture analysis become more prevalent, we can anticipate more nuanced investigations regarding the associations between race, ethnicity, and disease, which should permit the field to move beyond simple black and white categorization.