1. Home
  2. About NIDDK
  3. Staff Directory
  4. Anne Elizabeth Sumner, M.D.

Anne Elizabeth Sumner, M.D.

Photo of Anne Sumner.
Scientific Focus Areas: Clinical Research, Epidemiology, Health Disparities, Social and Behavioral Sciences

Professional Experience

  • Endocrine, Diabetes and Metabolism Fellowship, Medical College of Pennsylvania, 1988-1990
  • Nutrition and Metabolism Fellowship, Hospital of the University of Pennsylvania, 1987-1988
  • Internal Medicine Residency, Reading Hospital and Medical Center, 1983-1986
  • M.D., University of Pennsylvania School of Medicine, 1982
  • B.A., Brown University, 1976

Research Goal

The purpose of our research is to design screening tests and early interventions to decrease the impact of the diabetes and heart disease epidemic occurring worldwide in African descent populations.

Current Research

Diabetes and heart disease are reaching epidemic proportions in people of African descent.  To reverse this trend and improve longevity and quality of life, clarity on the causes of cardiometabolic disease in people of African descent and improved screening is essential.  Diabetes and heart disease result from the combined effects of insulin resistance and relative beta- cell failure.  The earliest etiological and epidemiological studies were done in populations of white European descent such as Framingham Heart Study, the Nurse’s Health Study and the Physician’s Health Study.  This initial work demonstrated that insulin resistance and the associated hypertriglyceridemia were the major and earliest causes of cardiometabolic disease.  This research subsequently informed the development of screening tests for diabetes and heart disease.  Hence, the majority of screening tests for diabetes and heart disease focus on triglyceride (TG) levels and insulin resistance.  TG-based screening tests depend on the principle that elevated TG levels are an excellent marker of insulin resistance and should trigger intervention to prevent diabetes and heart disease.  TG-based screening tests include the Metabolic Syndrome, the TG/HDL-Ratio, and the Hypertriglyceridemic Waist.

In the first phase of our research, we determined that TG-based screening tests failed to detect early cardiometabolic disease in African-Americans and African immigrants to the United States.  The major reason for this failure is that in African descent populations, insulin resistance does not trigger hypertriglyceridemia.  Thus, TG levels are not a key marker of risk. These TG-based screening tests often diagnose people of African descent as normal at a time when intervention could have delayed or prevented disease.  Hence, TG-based screening tests represent a lost opportunity for early intervention and contribute to health disparities.  Once we made the observation that TG-based screening tests were not optimally beneficial in African descent populations, we used metabolic testing and mathematical modeling to identify the etiology.  Factors that have been uncovered so far include the following findings: African-Americans have high levels of the enzyme lipoprotein lipase, great sensitivity to insulin-induced free fatty acid suppression, and low levels of apoC-III and visceral adipose tissue.  

We are also exploring whether relative beta-cell failure has a greater impact than the combined effect of insulin resistance and obesity on the development of diabetes and heart disease in Africans and African immigrants.

Currently we are focusing on 3 areas: (1) identifying the best screening tests to detect pre-diabetes and diabetes in blacks; (2) the effect of sickle cell trait (SCT), HbC trait and G6PD deficiency on A1C’s performance of a diagnostic test for both prediabetes and diabetes; (3) the psychosocial determinants of diabetes and heart disease in African immigrants with specific attention to the effect of the stress and reason for immigration on risk for diabetes and heart disease in African immigrants.

Best screening tests to detect prediabetes and diabetes in Africans.

We have previously shown that in African immigrants, A1C had a diagnostic sensitivity for detection of abnormal glucose tolerance of only ~50%. But combining A1C with fasting glucose increased sensitivity to ~70%.

We have evaluated the diagnostic efficacy of two new nonfasting diagnostic markers of hyperglycemia, specifically glycated albumin (GA) and fructosamine. We found that glycated albumin performs better as a diagnostic test than fructosamine. In addition, we reported that the diagnostic efficacy of A1C combined with GA is ~70% this is as good as the sensitivity of A1C combined with fasting plasma glucose. The public health advantage of A1C combined with GA is that the detection of glycemic status, will not require fasting and therefore the opportunity for widespread screening for diabetes will be increased. Our next focus is identifying the African immigrants with hyperglycemia not detected by either A1C or GA and figuring out why.

An additional important and unique of our work is that we are evaluating the reproducibility of the OGTT, as well as fasting plasma glucose (FPG), A1C, GA and fructosamine in the diagnosis of hyperglycemia in African immigrants. Our preliminary results are that the results of the OGTT as a diagnostic test are very reproducible for the detection of diabetes but not prediabetes. For both FPG and A1C, the diagnostic sensitivity is very low and that this low sensitivity does not improve with repeated testing. Data on the reproducibility of GA and fructosamine is pending.

Effect of SCT, HbC trait and G6PD deficiency on the diagnostic efficacy of A1C

Heterozygous hemoglobinopathies such as sickle cell trait (SCT) and HbC trait have levels of HbA which are lower than normal by 25 to 50%. Therefore, in the presence of either SCT or HbC trait A1C might not correctly reflect glycemia. While there is some published data, much of it conflicting on the effect of SCT on A1C as a diagnostic test, there is no published data on the effect of HbC trait. This is because most studies either exclude people with HbC trait or combine them with SCT. Therefore, the separate effect of HbC trait on A1C is not known.  In our study we have 14 people with HbC trait, 50 percent of whom had either prediabetes or diabetes. In this small sample the diagnostic sensitivity of A1C was 0% and the specificity was 100%. We need to continue to recruit, discover HbC trait on electrophoresis and evaluate. If this finding holds, it will be important in areas of the world where the prevalence of both diabetes and HbC trait are high to have diagnostic alternatives to A1C.

G6PD Deficiency was recognized by genetic studies to be associated with low A1C levels even in the presence of hyperglycemia. This finding needs to be tested clinically. Therefore, we sought permission to test this and even though the number of African immigrants who have been evaluated to date are few, this appears to be true. Consecutive testing of African immigrants who have G6PD deficiency should yield the answer.

Psychosocial determinants of diabetes and heart disease in African immigrants

We have demonstrated that the reason for immigration affects biologic stress measured by allostatic load. When reasons for immigration were work or seeking asylum/refugee status, biological markers of metabolic stress was significantly higher than if the reasons for immigration were study, family reunification or diversity lottery.  Going forward, we are evaluating the degree to which biologic stress related to reason for immigration translates into increased risk for both diabetes and heart disease. 

Applying our Research

My research will improve cardiometabolic health through better screening and early intervention.

Need for Further Study

Overall, there is great public health significance to our work with African descent populations. Our research should lead to results that convert undiagnosed diabetes into diagnosed diabetes and better prevention of diabetes and its complications including cardiovascular disease in African descent populations worldwide.

We need to explore and identify universal screening tests including test meals for diabetes and heart disease.

Select Publications

A1C Underperforms as a Diagnostic Test in Africans Even in the Absence of Nutritional Deficiencies, Anemia and Hemoglobinopathies: Insight From the Africans in America Study.
Briker SM, Aduwo JY, Mugeni R, Horlyck-Romanovsky MF, DuBose CW, Mabundo LS, Hormenu T, Chung ST, Ha J, Sherman A, Sumner AE.
Front Endocrinol (Lausanne) (2019) 10:533. Abstract/Full Text
Improved Detection of Abnormal Glucose Tolerance in Africans: The Value of Combining Hemoglobin A(1c) With Glycated Albumin.
Hobabagabo AF, Osei-Tutu NH, Hormenu T, Shoup EM, DuBose CW, Mabundo LS, Ha J, Sherman A, Chung ST, Sacks DB, Sumner AE.
Diabetes Care (2020 Oct) 43:2607-2613. Abstract/Full Text
View More Publications

Research in Plain Language

The specific goal of this research program is the identification of screening tests that are effective in detecting individuals of African descent who are at high risk for developing diabetes. Screening tests which work well create the opportunity to intervene when both disease prevention and minimization of complications can be optimized. Due to the magnitude and multiplicity of racial and ethnic differences, the Section on Ethnicity and Health is a major advocate for evaluating screening tests in the populations in which they are going to applied. The African descent populations with whom the Section is currently working include: African Americans, African immigrants to the United States and Africans living in sub-Saharan Africa.

Last Reviewed July 2023