This protocol compares diabetes (DM) and heart disease risk in African-Americans (AA) and African immigrants (AI). The AA cohort is known as TARA for: Triglyceride and Cardiovascular Risk in AA. The AI are known as the: Africans in America cohort. The sample of AA participating is representative of the AA population in the US because the prevalence of obesity (43%), prediabetes (preDM) (22%) and hypertension (21%) is similar to NHANES data. However, no national data on diabetic or cardiac health of AI exists. To correct this deficiency the Africans in America cohor was established. To determine the prevalence of DM and preDM, we are relying not just on fasting glucose but we are also performing oral glucose tolerance tests (OGTT) and measuring A1C levels and glycated albumin levels. In performing these tests, we discovered that the prevalence of preDM and hypertension is twice as high in AI than AA men. In addition, the rate of undiagnosed DM was 7% in AI men vs <1% in AA men. In contrast the rate of hypertension, DM and preDM are similar in AI and AA women. Identifying the reasons for why AI men are less healthy than AA men is a major focus of research in this protocol. To improve the metabolic health of AI men, it is essential to learn why preDM, DM and hypertension is occurring more often in AI men than AA men even though AI men are less obese, less often smokers and more often married. As a next step we are examining the effect of stress of immigration on AI health. We measure cardiometabolic stress with the Allostatic Load Score (ALS). The immigration related factors we are examining are: age of immigration, duration of stay in US and reason for immigration. We discovered that when the reasons for immigration are work, study or asylum/refugee, metabolic stress measured by ALS is higher than when the reasons for immigration are associated with keeping families together such as reunification programs, or diversity lottery for self and immediate family. Hence reason for immigration is key medical history. In addition, we are studying if A1C as a diagnostic test for DM, can replace the OGTT. A1C is a hemoglobin dependent test and AI have a high prevalence of sickle cell trait (SCT) (i.e. 10 to 40%) and hemoglobin C (HbC) trait (i.e. 15% in West Africa). Therefore, before widespread use of A1C as a diagnostic test for DM is instituted in Africa, validation is necessary. When we evaluated diagnostic efficacy of A1C in AI, we found that the diagnostic sensitivity for the detection of DM was only 30% and for preDM was 60%. Then we did additional analyses, we proved that the poor efficacy of A1C could not be accounted for by potential confounders such as anemia or nutritional deficiencies (B12 or folate) or sickle cell trait. We were able to prove that our participants were not anemic or nutritionally deficient. Hence, these factors are not responsible. When the Africans are divided according to SCT status, the sensitivity of A1C was 50% in both groups. Therefore, A1C is not be an ideal test in Africans and SCT is not be a compromising factor. This important information was just published in Frontiers in Endocrinology. The next conundrum is the effect of HbC trait on A1C efficacy. We found in the small sample number of AI with HbC trait, that the diagnostic efficacy was < 10%. We believe we are the first to study the influence of HbC trait on the diagnostic efficacy of A1C. Most studies either exclude people with HbC trait from analyses or combine them with people with SCT. Therefore, if there is an independent effect of HbC trait, if is not detectable. We need a larger sample size of people HbC trait in order to definitive prove whether HbC trait represents yet another reason why A1C is a suboptimal diagnostic test in Africans. Glucose 6 phosphate dehydrogenase (G6PD) deficiency is another factor that might affect the diagnostic efficacy of A1C. G6PD deficiency is more common in African than European descent populations. A recent GWAS analyses predicted that G6PD deficiency is associated with A1C levels that are 1% lower than in the absence of G6PD deficiency. If this is true, it would be another factor lowering the efficacy of A1C as a diagnostic test in African descent populations. This GWAS study needs to be confirmed with actual metabolic data. We are collecting the data to do this analysis. To improve the diagnostic efficacy of screening tests for DM, we are examining alternatives such as glycated albumin (GA) as a single test or in combination with A1C or fasting glucose. We have found that fasting glucose combined with A1C has a diagnostic sensitivity of 70%. But since obtaining a fasting sample can be problematic, we tested the combination of A1C and GA. We have found that in the detection of DM and preDM, the combination of A1C and GA had a ensitivity of 72%. This is important because GA is an inexpensive, easy test to set up and run. We are in the process of establishing collaborations with physicians several African countries to assess the efficacy of A1C and GA. The relationship of body size to cardiovascular and DM risk is another area of investigation. In our cohorts, the mean body mass index (BMI) in AA is 30.6 kg/m2 but only 26.4 kg/m2 in AI. BMI is a mathematical method used to correct weight for height. Due to the broad range of BMI in the participants in this cohort, it is possible to evaluate the relationship of body size to insulin resistance, a major factor in the development of DM, and heart disease. We have found in AA men a waist circumference (WC) of 102 cm predicts both insulin resistance and obesity. This is in agreement with the National Cholesterol Education Program values for whites. But in AI men, insulin resistance occurs at a much lower WC, specifically 91 cm. This difference between AA and AI men suggests that a single WC of risk does not apply to all African descent populations. In AA women a WC of 96 cm predicted both insulin resistance and obesity and this WC of risk was similar in AA, AI, Black South African and West African women. Therefore, among populations of African descent, there may be less variation in women than men. However, as the WC of risk is 88 cm in white women, there is a large difference by race. Guidelines which screen for disease might be more effective if this was better appreciated and understood. Elevated TG and low HDL are considered hallmarks of insulin resistance. However, while elevated TG is a marker of insulin resistance in whites, we have shown that TG is not a marker of insulin resistance in AA. Results from TARA were so impressive that the hypothesis that TG was not a marker of insulin resistance in AA was subsequently tested in NHANES data (1999-2001). In this dataset of whites, AA and Mexican Americans (MA), the fact that TG was not a marker of insulin resistance was confirmed. However, TG was a powerful marker of insulin resistance in whites and MA. Altogether this research on race differences in the relationship of TG to insulin resistance again demonstrates that to detect disease at time when intervention can affect outcome, there is a need to develop ethnic-specific guidelines. Recently the TG/HDL ratio at a level of >3.0 has been suggested to be a marker of insulin resistance. This is well established in whites. After demonstrating the TG/HDL ratio did not work in AA, we tested the ratio in white South African women, Black South African women and West African women. While the ratio effectively predicted insulin resistance in white women, it did not African descent women. Hence findings related to insulin resistance in whites are not universally applicable. In summary this protocol is dedicated to research which defines relevant risk factors and prevents through early diagnosis DM and heart disease in people of African descent globally.
Showing the most recent 10 out of 49 publications
