Pima Indian Contributions to Our Understanding of Diabetic Kidney Disease

Diabetes is the leading cause of kidney failure worldwide, and most cases of kidney failure attributable to diabetes occur in individuals with type 2 diabetes (1). Although its prominent role in kidney failure is widely accepted today, type 2 diabetes was previously believed to be associated with a benign kidney prognosis (2,3) and progressive diabetic kidney disease (DKD) was attributed largely to type 1 diabetes (4). This belief was based, in part, on early studies in Caucasians, who typically developed type 2 diabetes later in life and often succumbed to other diseases or complications of diabetes before the end-stage kidney complications of diabetes could develop. Even in populations at high risk for type 2 diabetes, such as American Indians, diabetes was often considered “an extremely mild disease” with no mention of kidney failure or other major complications (5).

Prospective epidemiologic studies in informative populations and large clinical cohorts were crucial to establishing the severe impact of type 2 diabetes on the major complications of diabetes. In this perspective, we describe a half century of studies in Pima Indians from the Gila River Indian Community, which contributed to critical changes in our understanding of kidney disease in type 2 diabetes. This American Indian population has one of the highest rates of type 2 diabetes in the world, with more than one-half of those ≥35 years old affected (6). We begin our review by exploring the early epidemiologic studies that identified and characterized the epidemic of DKD in this community and led to a significant expansion of our largely National Institutes of Health–supported research efforts related to DKD. We then discuss the studies of glomerular hemodynamic function and kidney structure that were initiated in response to our epidemiologic findings. These studies further expanded our knowledge of this disease and provided the framework for precision medicine studies to identify the mechanisms underlying this life-threatening diabetes complication. We conclude with examples of precision medicine studies conducted in this cohort that integrate the comprehensive clinical phenotype accumulated over decades with the glomerular hemodynamic data, quantitative morphometric analyses, and gene expression data from kidney tissue obtained from research kidney biopsies to illustrate the considerable scientific value derived from the long-term comprehensive characterization of this informative cohort.

Studies in the Pima Indians offer unique insights into the course, determinants, and mechanisms of kidney disease in type 2 diabetes because there are fewer comorbid conditions affecting the kidneys in this population than in others. This distinction is due largely to their younger age at onset of diabetes and their lower risk of hypertension and cardiovascular disease in comparison with other populations. Although we do not have a direct comparison of blood pressure in Pima Indians with type 2 diabetes and blood pressure in another population with type 2 diabetes, the prevalence of hypertension among Pima Indians without diabetes aged 25–74 years was 14% compared with 22% in the corresponding U.S. population (7), a difference attributed in part to low sympathetic nervous system activity and to lower β-adrenergic sensitivity than in Caucasians (8,9). Furthermore, the frequency of electrocardiographic evidence of coronary heart disease and of autopsy-proven myocardial infarction in Pima Indians ≥40 years of age, despite their high prevalence of diabetes, was one-half that found in Tecumseh, Michigan (10), and the incidence of fatal atherosclerotic coronary heart disease in the Pima Indians was less than one-half that found in the Framingham population after controlling for age, sex, and diabetes status (11). As a result, nearly all end-stage kidney disease (ESKD) encountered in this population is attributable to diabetes (12). In addition, the onset and duration of diabetes are known with greater precision than in other populations with type 2 diabetes because of serial testing.

Pima Indians from the Gila River Indian Community in Arizona were invited to participate in a longitudinal study of diabetes and its complications at intervals of approximately every 2 years between 1965 and 2007, regardless of health. A standardized examination was performed in community members ≥5 years old that included a medical history, physical examination, a 75-g oral glucose tolerance test after an overnight fast, and measures of urine albumin and protein excretion. Urine protein concentration was measured by dipstick in untimed urine specimens collected at each examination throughout the longitudinal study, and in those with a trace or more of protein the concentration was further quantified by the Shevky-Stafford acid precipitation method (13). Measurement of albumin concentration in the same urine specimens by nephelometric immunoassay was added in July 1982. Urine albumin and protein concentrations were expressed as ratios with urine creatinine concentration measured in the same sample.

In 1988, informative subsets of individuals from this population with normal glucose tolerance, impaired glucose tolerance, newly diagnosed diabetes, or diabetes of at least 5 years’ duration who were at different stages of DKD and had no evidence of non-DKD were selected to undergo more detailed longitudinal studies of DKD (14–17). Many of these individuals continue to be followed. Glomerular filtration rate (GFR), renal plasma flow, and the permselective properties of the glomerular filtration barrier were measured serially during follow-up by the urinary clearance of iothalamate, p-aminohippurate, and neutral dextrans, respectively. High-performance liquid chromatography was used to measure the serum and urine concentrations of these filtration markers. Research kidney biopsies were also performed at various intervals during follow-up, and the kidney tissue underwent unbiased quantitative morphometry and, since 2003, compartment-specific gene expression profiling. Figure 1 shows a timeline for the epidemiology studies, kidney physiology studies, and kidney biopsies.

The epidemiologic studies described herein used data from the community survey, and the hemodynamic, structural, and molecular profiling studies are derived from the subset of 377 individuals who were invited to participate in the more detailed DKD studies. These studies were approved by the Institutional Review Board of the National Institute of Diabetes and Digestive and Kidney Diseases. Each participant provided written informed consent.

Early studies in the Pima Indians identified kidney disease as an important complication of type 2 diabetes, well before this type of diabetes was recognized to play a key role in CKD causation in other populations. Proteinuria was more common in participants with diabetes than in those without, and its frequency and severity were strongly associated with the duration of diabetes. Moreover, the incidence of ESKD was strongly associated with the presence of diabetic retinopathy (18). In the 1960s through the 1980s, nearly all of the threefold excess mortality associated with diabetes was found in those with proteinuria (19), and an autopsy study demonstrated that “intercapillary glomerulosclerosis,” the term then used to describe diabetic glomerulopathy characterized by nodular and diffuse mesangial expansion, was the dominant pathology in the kidneys among those with diabetes (20,21). A nephelometric immunoassay was established to measure urine albumin at low concentrations (22), replacing the expensive and technically challenging radioimmunoassay previously used to measure low urine concentrations of albumin. The use of this simple assay, combined with studies establishing the validity of a urine albumin-to-creatinine ratio as a measure of urine albumin excretion (23), contributed to the acceptance of the albumin-to-creatinine ratio as the standard screening and diagnostic method for evaluating DKD (24,25).

A cross-sectional study conducted in the Pima Indians (26) found that the prevalence of elevated albuminuria, defined by a single albumin-to-creatinine ratio ≥30 mg/g, was twice as high in participants with impaired glucose tolerance as in those with normal glucose tolerance, suggesting that even small elevations of plasma glucose concentrations prior to diabetes onset may affect the kidneys. Moreover, within the first 5 years of the onset of type 2 diabetes, 29% of Pima Indians had elevated albuminuria, a prevalence nearly four times as high as observed in those with normal glucose tolerance, and the proportion with elevated albuminuria after 20 years of diabetes was 86% (Fig. 2A). This confirmed that DKD is a frequent complication of type 2 diabetes in this population and that it is strongly associated with duration of diabetes. Further progression of DKD is common among Pima Indians with elevated albuminuria (23), with approximately one-half of those who develop macroalbuminuria (urine albumin-to-creatinine ratio ≥300 mg/g) progressing within 10 years to ESKD (18). Indeed, the cumulative incidence of ESKD in Pima Indians with type 2 diabetes was similar to that in the predominantly Caucasian patients with type 1 diabetes from the Joslin Clinic (12) (Fig. 2B) but far higher than in the predominantly Caucasian patients with type 2 diabetes from the Mayo Clinic. We attributed this difference to the younger age at onset of diabetes and the lower death rate from coronary heart disease in the Pima Indians than in the Caucasians with type 2 diabetes (18).

The threshold for microalbuminuria (also referred to as moderate albuminuria [urine albumin-to-creatinine ratio ≥30 mg/g]) was selected because it was approximately the 95th percentile of the urine albumin-to--creatinine ratio in a “healthy” subset of Pima Indians aged ≥15 years who had normal glucose tolerance, took no medicines, had no known kidney or cardiovascular diseases, had a normal serum creatinine concentration (<1.5 mg/dL), and had normal blood pressure—defined at the time as <160 mmHg systolic and <95 mmHg diastolic (26). The threshold for macroalbuminuria (also referred to as severe albuminuria) was arbitrarily defined by the albumin-to-creatinine ratio corresponding approximately to dipstick-detectable (≥1) proteinuria that was traditionally taken to represent “clinical proteinuria” in past studies of DKD (26).

Prior to the widespread use and acceptance of renal replacement therapy in the community, DKD was the leading cause of death in Pima Indians with diabetes, thus emphasizing the critical need to investigate its characteristics and determinants (19,27). Once renal replacement therapy became readily available in the community and deaths from kidney disease per se diminished, cardiovascular disease became the leading cause of death in those with DKD (27), just as it is in other populations with diabetes. The prominence of type 2 diabetes as a major cause of ESKD was subsequently confirmed in other populations as the incidence of type 2 diabetes increased worldwide and the onset of type 2 diabetes occurred at earlier ages than in the past (1,28,29).

Advances in therapeutic management of diabetes and its complications in the Pima Indians (30) combined, in recent years, with a higher proportion of those with diabetes of long duration (31) have profoundly influenced the clinical course of DKD in this population. An increasing proportion of cases of DKD are characterized by low glomerular filtration rate without an elevation in urine albumin excretion, a finding we attribute, at least in part, to the increased use of antihypertensive medicines that suppress albuminuria (32). Also, potentially as a result of this effect, the rate of GFR decline has slowed, leading to a reduction in the incidence of ESKD, particularly among those ≥45 years old (31) (Fig. 2D).

Although the role of hyperglycemia in the development of diabetes complications is undisputed (18,33), high blood pressure was not initially recognized as a risk factor for DKD but was often considered a consequence of DKD (34,35). Some of the early studies establishing the association between high blood pressure and the risk of DKD in type 2 diabetes were conducted in the Pima Indians (21,33,36–38) (Fig. 2C). In addition to these widely recognized risk factors for progressive DKD, the extensive clinical characterization of the Pima Indians was crucial for identifying additional DKD risk factors, including low birth weight (39), intrauterine exposure to diabetes (40,41), periodontal disease (42), and exposure to persistent organic pollutants (43). These factors could also affect kidney disease risk in other populations. Familial aggregation of DKD in type 2 diabetes in Pima Indians (44) inspired extensive studies to identify the genetic determinants of DKD (45,46). Comparative studies of DKD in various populations worldwide, including the Pima Indians, also provided early evidence of the ethnic differences in DKD prevalence (47).

Diabetes in Pima Indian children and adolescents was first identified in the mid-1960s with the initiation of the longitudinal population study. Studies confirmed that this youth-onset diabetes was entirely type 2 (48), as it was characterized by ongoing insulin secretion (49) and lack of insulin dependence (6), absent or low levels of islet cell and GAD antibodies (50,51), and absence of strong linkage or association with maturity-onset diabetes of the young loci (52,53). The prevalence of youth-onset type 2 diabetes in the Pima Indians, defined as onset before 20 years of age, has increased in recent years, due in part to a growing incidence of exposure to intrauterine diabetes (54,55) and to an increasing prevalence and severity of obesity in childhood and adolescence (55–57). These trends have also led to an increase in the frequency of diabetes complications, including DKD, in midlife (58) and to a fivefold greater risk of diabetic ESKD between the ages of 25 and 54 years among Pima Indians with youth-onset type 2 diabetes compared with risk for those with older-onset diabetes (59) (Fig. 2E). An alarming rise in youth-onset type 2 diabetes is now reported in populations worldwide (29,60).

Although the introduction and widespread use of highly efficacious medicines to control blood pressure, reduce hyperglycemia, and block the renin-angiotensin system have reduced the incidence of ESKD in the Pima Indians (31) and in other populations (61), the increasing epidemic of type 2 diabetes in youth and the rapidly rising prevalence of type 2 diabetes worldwide (29) could reverse these encouraging trends (62). Identifying these early trends in high-risk populations, such as the Pima Indians, is critical to developing timely and successful prevention and management strategies that can benefit populations worldwide.

Glomerular hemodynamic function in DKD has been studied extensively in the Pima Indians. GFR, measured by the urinary clearance of iothalamate, was, on average, 14% higher in individuals with impaired versus normal glucose tolerance and continued to increase after the onset of diabetes, reaching a plateau in those with microalbuminuria (17,63) (Fig. 3A). GFR generally remained high but stable in individuals with diabetes with normoalbuminuria or microalbuminuria, even with long-standing diabetes, but declined with the appearance of macroalbuminuria (17). During long-term follow-up, an early decline in kidney function, defined by an average GFR loss of ≥3.3% per year over ∼4 years, was associated with a nearly fivefold increased risk of ESKD during subsequent follow-up of up to 17.8 years, but a decline in GFR predictive of ESKD was strongly dependent on the simultaneous progression to macroalbuminuria (64). A reduction in the ultrafiltration coefficient (K_f) was the major determinant of the decline in GFR associated with the progression from microalbuminuria to macroalbuminuria (65).

The size-selective properties of the glomerular filtration barrier were investigated in Pima Indians using dextran sieving studies and pore theory. Fractional clearances of neutral dextrans of graded sizes showed that normoalbuminuric participants with diabetes for <3 years had a loss of size restriction for larger dextrans compared with a normal glucose tolerance group. This manifested as a selective increase in the sieving coefficients of large and nearly impermeant dextrans at the high-radius end of the sieving curve (16) (Fig. 3B). Pima Indians with type 2 diabetes of long duration and macroalbuminuria exhibited a further loss of size restriction for larger dextrans compared with a long-term normoalbuminuric group with type 2 diabetes (Fig. 3C), suggesting an evolving defect in size selectivity that was not apparent in those with microalbuminuria (15). These findings suggest that the passage of macromolecules across the glomerular barrier occurs through several mechanisms. In early DKD, increased GFR, changes in tubular reabsorption of filtered proteins, and alterations in charge-based restriction at the glomerular filtration barrier may be responsible for the appearance of microalbuminuria (66), whereas a defect in size selectivity is responsible for higher levels of albuminuria, suggesting an underlying structural defect emerging with increasing albuminuria (see below) (15).

Enhanced trans-glomerular trafficking of protein may contribute to the progressive decline in GFR by stimulating glomerular cells to produce extracellular matrix, which eventually obliterates the glomerular capillary (17). Exposure to high concentrations of proteins may also induce the production of proinflammatory and profibrotic factors by tubular cells, leading to tubular atrophy and interstitial fibrosis (67). Elevated intraglomerular pressure itself, although not directly measurable in humans, may also mediate progressive DKD. Higher baseline GFR, a potential surrogate for elevated intraglomerular pressure, was associated with a more rapid decline in GFR in a cohort of Pima Indians (83% of whom had type 2 diabetes) during a median follow-up of 9.1 years, with adjustment for possible confounders with a linear mixed regression model (68). In addition, estimates of intraglomerular pressure and afferent and efferent arteriolar resistance derived from the Gomez equation (69,70) suggested that an unstable intraglomerular hemodynamic trajectory characterized by either elevated intraglomerular pressure followed by a rapid decline or a rapid change in intraglomerular arteriolar tone was strongly associated with incident ESKD, independent of GFR, during a median follow-up of 17.5 years (Fig. 3D) (71). Of note, participants with progressive DKD typically exhibited one or the other of these unstable intraglomerular hemodynamic trajectories, but not both, suggesting two distinct hemodynamic mechanisms of DKD progression. Those with unstable intraglomerular hemodynamic pressures were more likely to be younger, more obese, and have a shorter duration of diabetes, higher GFR, and lower HbA_1c and albumin-to-creatinine ratio than those with unstable arteriolar tone.

Kidney structural injury associated with type 2 diabetes has been characterized extensively in the Pima Indians with the use of unbiased quantitative morphometry (14,65,72,73) on tissue obtained from kidney biopsies. Since these biopsies were performed solely for research, and were not performed for a clinical indication, the selection of participants was not biased toward unusual cases. Tissue was examined by light and electron microscopy, and the lesions of diabetic glomerulosclerosis paralleled those seen in other populations (4). Importantly, the structural damage exhibited in these biopsies is exclusively attributable to diabetes (73) and is comparable with the damage described in type 1 diabetes (74,75) (Fig. 4). Importantly, the kidney lesions observed in the Pima Indians are more homogenous than those reported in other populations with type 2 diabetes, due in large part to younger age, less hypertension, and lower frequency of cardiovascular disease at the onset of DKD among Pima Indians than in other populations (76–78). This permits a pristine characterization of DKD in type 2 diabetes that is not preceded by kidney injury unrelated to diabetes.

Podocyte injury plays an essential role in the development and progression of DKD, and early biopsy studies in the Pima Indians were instrumental in characterizing this role. Participants with longer-term type 2 diabetes and microalbuminuria at the time of kidney biopsy had 20% fewer podocytes per glomerulus, and those with macroalbuminuria had 40% fewer podocytes per glomerulus, than Pima Indians with normoalbuminuria (73). Moreover, those with fewer podocytes per glomerulus were more likely to develop increasing albuminuria (79). Pima Indians with microalbuminuria who underwent a second kidney biopsy had a 35% decline in the number of podocytes per glomerulus, on average, during the 4 years between kidney biopsies, indicating that podocyte detachment was a key contributor to early progressive DKD (65). With the glomerular hypertrophy that accompanies the onset of diabetes, podocytes extend their foot processes to maintain coverage of the expanded glomerular basement membrane, a process that may adversely affect their functional integrity (73). Sustained mechanical stress, and the combined effects of increased passage of plasma proteins across the glomerular filtration barrier and intraglomerular hypertension, may ultimately lead to the podocyte detachment we observed on serial kidney biopsies.

Abnormalities in kidney structure are present early in the course of DKD, typically before the onset of clinically detectable kidney disease (73,80,81), and in the Pima Indians they strongly predicted the decline in kidney function (82). Major structural predictors of declining kidney function included mesangial expansion; increased global glomerulosclerosis percentage, mean glomerular volume, and podocyte foot process width; reduced glomerular filtration surface density; and a loss of endothelial fenestrations (82). When study participants underwent another kidney biopsy an average of 9.3 years later, nearly all these structural parameters worsened, and the structural changes correlated with decline in GFR and increase in albuminuria. At this early clinical stage of DKD, however, the correlation of change in kidney structure with change in albuminuria was stronger than with the change in GFR, particularly in those with normal or elevated GFR at baseline (64,72).

The accumulated clinical observations of the past half century in this cohort of Pima Indians together with >4,000 direct measures of kidney function spanning three decades and >300 protocol kidney biopsies provide unparalleled opportunities for data integration to further extend our knowledge of the mechanisms underlying the development and progression of DKD and related complications using the tools of precision medicine. Several examples of mechanistic studies conducted in the Pima Indians that used this approach are described below (83–85).

A transcriptomic analysis of kidney tissue in Pima Indians with early DKD and in Caucasians with more advanced DKD identified JAK/STAT signaling as one of the most highly regulated pathways in the glomerular and tubulointerstitial compartments of persons with early and progressive DKD compared with control subjects (83) (Fig. 5). Glomerular JAK2 mRNA levels were higher in early DKD and lower in more advanced disease, whereas tubulointerstitial JAK2 levels were higher in more advanced disease. These findings illustrate that enhanced JAK2 expression, which was previously associated with enhanced fibrosis in DKD, corresponds temporally with the progression of DKD from a disorder characterized by initial glomerulopathy to one involving tubulointerstitial fibrosis in more advanced disease. These DKD pathway maps provided part of the rationale for testing the JAK1 and JAK2 inhibitor baricitinib as a potential therapy in advanced DKD (ClinicalTrials.gov identifier: NCT01683409). In a small randomized clinical trial conducted elsewhere, baricitinib significantly reduced the level of urine albumin excretion in a dose-dependent manner, compared with placebo, after 6 months of treatment, with a sustained benefit noted 4 weeks after discontinuation of baricitinib (86). In addition, reduction of blood and urine biomarkers of JAK-STAT activation predicted from these gene expression profiling studies preceded the reduction in albuminuria. This study illustrates the potential value of tissue-level molecular analysis in DKD in the identification of therapeutic targets and their associated biomarkers (85), and it closes the loop on precision medicine by demonstrating its potential for patient stratification and targeting therapies.

Because early kidney structural injury is a strong predictor of progressive DKD that precedes its clinical manifestations (82,87), we sought to characterize the pathophysiologic mechanisms of DKD by examining the link between transcriptional regulation and structural damage in early DKD in Pima Indians, using the precision medicine approach illustrated in Fig. 6 (85,88). The structural lesion most strongly associated with transcriptional regulation was cortical interstitial fractional volume (VvInt), the fraction of cortical tissue made up of interstitium and a measure of tubulointerstitial fibrosis associated with progressive DKD. VvInt-related transcripts showed significant enrichment for inflammatory mechanisms (Fig. 7), and associations of these VvInt-linked genes with long-term increases in albuminuria and declines in GFR demonstrated the relevance of these early-stage processes to clinical progression of DKD. The concordant regulation of a significant segment of the VvInt-associated transcripts in Caucasians with more advanced DKD supports the relevance of the identified molecular mechanisms in other populations (85). Inflammation plays an important role in progressive DKD, and this study linked inflammatory mechanisms with a specific structural lesion. Further precision medicine studies may help us link other biomarkers of inflammation found in Pima Indians and in other populations (89–92) with specific molecular pathways and patterns of structural injury and permit us to identify therapies that target these pathways within affected individuals.

With the arrival of the coronavirus disease 2019 (COVID-19) pandemic, the comprehensive characterization of the Pima Indian DKD cohort provided the impetus to explore how COVID-19 might disproportionately affect individuals with DKD. As illustrated in Fig. 8, we performed a side-by-side analysis to identify shared ACE2-associated mechanisms in DKD and COVID-19–associated kidney disease (84). The ACE2 receptor was selected because it is the primary cell-entry receptor for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus responsible for COVID-19. For definition of the expression pattern of ACE2, cell populations obtained by single-cell RNA sequencing analysis of kidney biopsies from Pima Indians with DKD and transplant living donor biopsies were clustered based on the transcriptional profiles of individual kidney cells. The clusters derived from >110,000 cells represented the entire spectrum of kidney cell types found along the nephron. Comparison of these single-cell RNA profiles in kidney biopsies from Pima Indians with DKD with those from healthy living donors showed that ACE2 expression was predominantly localized to proximal tubular epithelial cells, a cell-specific localization confirmed by in situ hybridization. Notably, ACE2 expression levels were unaffected by exposure to renin-angiotensin-aldosterone system inhibitors in the DKD cohort (Fig. 8). Molecular network analysis linked ACE2 expression to innate immune response and viral entry machinery. Similar cellular programs were observed in ACE2-positive proximal tubular epithelial cells obtained from urine samples in hospitalized patients with COVID-19, suggesting a consistent ACE2-coregulated proximal tubular epithelial cell expression program that may interact with the SARS-CoV-2 infection processes. The upregulation of viral infection pathways in DKD could explain the higher susceptibility of patients with diabetes to severe COVID-19 (84).

Long-term follow-up of a population carefully selected to epitomize a specific disease of interest provides an ideal opportunity for discoveries that can lead to paradigm shifts in our understanding of that disease. The Pima Indians are such a population, as they have contributed to a transformation in our understanding of kidney disease in type 2 diabetes. During a half century of follow-up, they have shown us that type 2 diabetes is indeed a major cause of DKD, that numerous modifiable risk factors are associated with its progression, that significant glomerular hemodynamic perturbations appear with the onset of type 2 diabetes and contribute to the initiation and progression of DKD, and that structural kidney injury precedes the development and predicts the progression of clinically apparent disease. Secular trends in disease progression and the emergence of new groups of individuals, such as children with type 2 diabetes who are at increased risk for DKD, have foreshadowed similar trends in other populations worldwide, providing clinicians and investigators with important opportunities to develop and test new management tools to address these adverse trends before they become entrenched. Selected clinical contributions of studies in the Pima Indians are summarized in Fig. 9.

Many findings about DKD first reported in the Pima Indians have been confirmed in other populations, and likewise those initially identified in other populations have been corroborated in the Pima Indians, illustrating both the considerable scientific value and the generalizability of this work. In addition to the scientific value of this research, we believe that it contributed to better kidney outcomes in the Pima Indians (30,31). Indeed, the shared belief in the benefits of this research helped establish the long-term partnership with study participants that made this work possible. Through this successful partnership we collected an unprecedented amount of clinical, physiologic, morphometric, genomic, and transcriptomic data, which permitted us to build and integrate the multilayered data sets needed to conduct state of the art precision medicine studies of DKD (93). New mechanistic insights identified in the Pima Indians and confirmed in other populations with use of these tools have greatly expanded our understanding of this complex disease and will undoubtedly influence our approach to its management in the coming years.

Acknowledgments. The authors acknowledge the significant contributions of David J. Pettitt, Robert L. Hanson, Lois I. Jones, Enrique Diaz, Jillian Loebel, Roselene Lovelace, Bernadine Waseta, and Camille Waseta (National Institute of Diabetes and Digestive and Kidney Diseases, Phoenix, AZ), Bryan D. Myers and Kristina L. Blouch (Stanford University School of Medicine, Stanford, CA), George W. Williams and Gerald J. Beck (The Cleveland Clinic Foundation, Cleveland, OH), Meda E. Pavkov (Centers for Disease Control and Prevention, Atlanta, GA), Petter Bjornstad (University of Colorado School of Medicine, Aurora, CO), Pierre J. Saulnier (University of Poitiers, INSERM CIC1402, Poitiers, France), Viji Nair (University of Michigan, Ann Arbor, MI), and Ann Plamer, Frida Maiers, and Zour Yang (University of Minnesota, Minneapolis, MN).

Funding. This research was supported by the Intramural Research Program at the National Institute of Diabetes and Digestive and Kidney Diseases, by the American Diabetes Association (Clinical Science Award 1-08-CR-42) to R.G.N, by an Interagency Agreement with the Centers for Disease Control and Prevention (16FED1604631) to R.G.N., and by National Institute of Diabetes and Digestive and Kidney Diseases contracts (N01-DK-62285 and N01-DK-72291).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.