Conference abstract
A Review of Diabetes Prediction Equations in African Descent Populations
Pan African Medical Journal - Conference Proceedings. 2021:11(5).04
Jan 2021.
doi: 10.11604/pamj-cp.2021.11.5.1038
Archived on: 04 Jan 2021
Contact the corresponding author
Keywords: Diabetes, African-descent populations, prediction equations, review
Poster
A Review of Diabetes Prediction Equations in African Descent Populations
Margrethe F Horlyck-Romanovsky&,1, Regine Mugeni1, Jessica Y Aduwo1, Sara M Briker1, Thomas Hormenu1, Anne E Sumner1,2
1National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD, United States, 2National Institute of Minority Health and Health Disparities, National Institutes of Health, Bethesda, MD, United States
&Corresponding author
Background : Predicting undiagnosed diabetes is a critical step towards addressing the diabetes epidemic in African descent populations. Objective: To review diabetes prediction equations which detect diabetes in African-descent populations.
Methods: Using PubMed, Scopus and Embase databases, a scoping review identified studies of diabetes prediction equations developed, tested or optimized in African-descent populations.
Results: A total of 15 prediction equations were reported in 8 studies. The prediction equations were of two types. Prevalence prediction equations (n=10) detected undiagnosed diabetes and based on non-invasive criteria only. Non-invasive variables included demographic factors, blood pressure and measures of body size. Incidence prediction equations (n=5) predicted risk of developing diabetes and used both non-invasive and invasive variables. Invasive variables required blood tests and included TG, HDL, A1C and fasting plasma glucose. Prevalence prediction studies were conducted in the United States, Africa and Europe. Incidence prediction equation studies were conducted only in the United States. Area under the receiver operator characteristics curve (AROC) or its equivalent, the C-statistic, estimated equation performance. Discrimination levels were: 0.50 Conclusions: Overall, prevalence prediction equations are not sufficient to detect diabetes in African-descent populations. Incidence prediction equations appear to be effective because they include invasive variables. Adding invasive variables may be a way to increase prediction equation efficacy in African-descent populations.
Keywords: Diabetes, African-descent populations, prediction equations, review
A Review of Diabetes Prediction Equations in African Descent Populations
Margrethe F Horlyck-Romanovsky&,1, Regine Mugeni1, Jessica Y Aduwo1, Sara M Briker1, Thomas Hormenu1, Anne E Sumner1,2
1National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD, United States, 2National Institute of Minority Health and Health Disparities, National Institutes of Health, Bethesda, MD, United States
&Corresponding author
Background : Predicting undiagnosed diabetes is a critical step towards addressing the diabetes epidemic in African descent populations. Objective: To review diabetes prediction equations which detect diabetes in African-descent populations.
Methods: Using PubMed, Scopus and Embase databases, a scoping review identified studies of diabetes prediction equations developed, tested or optimized in African-descent populations.
Results: A total of 15 prediction equations were reported in 8 studies. The prediction equations were of two types. Prevalence prediction equations (n=10) detected undiagnosed diabetes and based on non-invasive criteria only. Non-invasive variables included demographic factors, blood pressure and measures of body size. Incidence prediction equations (n=5) predicted risk of developing diabetes and used both non-invasive and invasive variables. Invasive variables required blood tests and included TG, HDL, A1C and fasting plasma glucose. Prevalence prediction studies were conducted in the United States, Africa and Europe. Incidence prediction equation studies were conducted only in the United States. Area under the receiver operator characteristics curve (AROC) or its equivalent, the C-statistic, estimated equation performance. Discrimination levels were: 0.50 Conclusions: Overall, prevalence prediction equations are not sufficient to detect diabetes in African-descent populations. Incidence prediction equations appear to be effective because they include invasive variables. Adding invasive variables may be a way to increase prediction equation efficacy in African-descent populations.
Keywords: Diabetes, African-descent populations, prediction equations, review
