Baye Assefa Mulu, Fenta Teferi Gedif, Karuranga Suvi, Nnakenyi Ifeyinwa Dorothy, Young Ekenechukwu Esther, Palmer Colin, Pearson Ewan R, Ulasi Ifeoma Isabella, Dawed Adem Y
Department of Pharmacology and Clinical Pharmacy, College of Health Sciences, Addis Ababa University, Addis Ababa, Ethiopia.
Department of Pharmaceutics and Social Pharmacy, College of Health Sciences, Addis Ababa University, Addis Ababa, Ethiopia.
Front Endocrinol (Lausanne). 2025 Mar 25;16:1501383. doi: 10.3389/fendo.2025.1501383. eCollection 2025.
Early diabetes screening is critical in sub-Saharan Africa (SSA), where the prevalence is increasing, yet a large proportion of cases remain undiagnosed. This study aimed to evaluate the performance of fasting plasma glucose (FPG) in screening diabetes and/or prediabetes compared to the 2-hour plasma glucose (2-h PG)-level in SSA.
Data from a population-based, cross-sectional diabetes screening survey involving 1550 individuals in Butajira, Ethiopia, and Enugu state, Nigeria were analyzed. Fasting plasma glucose and a 2-hour 75-g oral glucose tolerance test (OGTT) were utilized for diabetes screening. In addition, we determined and plotted the receiver operating characteristic curve for FPG against the reference standard 2-h PG to evaluate the screening tool's sensitivity and specificity.
The mean (SD) age of the study participants was 44.5 (± 16.43) years, with men comprising 50.4% of the cohort. Among 1550 individuals analyzed, 4.6% and 16.8% demonstrated diabetes and prediabetes, respectively, as identified by either FPG or 2-h PG. The agreement between FPG and 2-h PG in identifying diabetes and prediabetes was moderate, with kappa statistic of 0.56 (95% CI, 0.51 - 0.61; p<0.0001) for diabetes and 0.45 (95% CI, 0.40 - 0.50; p<0.0001) for prediabetes. FPG failed to detect 34.1% of all prediabetes and 44.4% of all diabetes cases. The sensitivity of FPG in identifying diabetes cases was 44.3% at a cut-off 126 mg/dL with a specificity of 99.3%. We identified the optimal FPG cut-off for detecting newly identified diabetes cases using 2-h PG to be 105 mg/dL associated with a sensitivity and specificity of 67.2% and 94.0%, respectively.
FPG was able to correctly identify 99.3% of individuals with no diabetes but a significant percentage of diabetes cases would have remained undiagnosed if only FPG had been utilized instead of the 2-h PG. The use of 2-h PG test is recommended to diagnose diabetes in older individuals, females and non-obese persons who would be missed if tested by only FPG. Lowering the cut-off value for FPG to 105 mg/dL substantially increases the identification of individuals with diabetes, thus improving the effectiveness of FPG as a screening test for type 2 diabetes.
在撒哈拉以南非洲地区（SSA），早期糖尿病筛查至关重要，该地区糖尿病患病率不断上升，但仍有很大比例的病例未被诊断出来。本研究旨在评估空腹血糖（FPG）在筛查糖尿病和/或糖尿病前期方面与2小时血糖（2-h PG）水平相比的性能，研究对象为SSA地区人群。
分析了来自埃塞俄比亚布塔吉拉和尼日利亚埃努古州的一项基于人群的横断面糖尿病筛查调查的数据，该调查涉及1550名个体。采用空腹血糖和2小时75克口服葡萄糖耐量试验（OGTT）进行糖尿病筛查。此外，我们绘制了FPG相对于参考标准2-h PG的受试者工作特征曲线，以评估筛查工具的敏感性和特异性。
研究参与者的平均（标准差）年龄为44.5（±16.43）岁，男性占队列的50.4%。在分析的1550名个体中，通过FPG或2-h PG鉴定，分别有4.6%和16.8%的个体患有糖尿病和糖尿病前期。FPG和2-h PG在鉴定糖尿病和糖尿病前期方面的一致性中等，糖尿病的kappa统计量为0.56（95%CI，0.51 - 0.61；p<0.0001），糖尿病前期为0.45（95%CI，0.40 - 到0.50；p<0.0001）。FPG未能检测出所有糖尿病前期病例的34.1%和所有糖尿病病例的44.4%。FPG在识别糖尿病病例时，在截断值为126mg/dL时的敏感性为44.3%，特异性为99.3%。我们确定使用2-h PG检测新发现的糖尿病病例的最佳FPG截断值为105mg/dL，其敏感性和特异性分别为67.2%和94.0%。
FPG能够正确识别99.3%的无糖尿病个体，但如果仅使用FPG而不是2-h PG，仍有相当比例的糖尿病病例会未被诊断出来。建议使用2-h PG试验来诊断老年个体、女性和非肥胖个体的糖尿病，这些个体若仅通过FPG检测会被漏诊。将FPG的截断值降低到105mg/dL可大幅增加糖尿病个体识别率，从而提高FPG作为2型糖尿病筛查试验的有效性。
