Lower Threshold
Even though the test has been a standard part of blood chemistry panels for decades, the tools for uncovering its value have only existed for a few years. “The availability of big data and the robust database in the VA allowed us to look at large numbers of patients who had continuity of care” to find correlations, Rhee said.
The researchers retrospectively analyzed the association of outpatient RPG levels with a subsequent diagnosis of diabetes using data from the VA Informatics and Computing Infrastructure Corporate Data Warehouse. The study included 942,446 veterans who did not have a diabetes diagnosis during a baseline year between 2002 and 2007, had at least three RPG tests during that year and had at least one primary care visit at the VA every year for the next five years.
The average age was 63.4 ± 12.4, 96% were male, 82.7% were white, 81.3% were current or former smokers, and mean body mass index was 28.9 ± 5.4. While the researchers did not look at diabetes screening rates in the cohort, “almost all would meet the criteria for screening,” Rhee observed.
Within five years, 94,559 veterans received an initial diagnosis of diabetes.
Among those who developed diabetes, the median RPG level was 112 mg/dl, with 90% and 95% percentiles at 135 and 145 mg/dl, respectively. Among veterans who did not develop diabetes, the median RPG level was notably lower at 101 mg/dl with the 90% percentile at 114 and the 95th at 120 mg/dl.
The analysis showed that multiple readings of 115 mg/dl or 130 mg/dl indicated a high risk of developing diabetes in the next few years. Two or more tests with RPG at or above 115 mg/dl predicted a new diagnosis of diabetes within a year with specificity of 77% and sensitivity of 87%. Within three years, nearly a quarter of veterans with test results in this range developed diabetes.
Two or more tests with RPG at or above 130 mg/dl had a 93% specificity and 59% sensitivity for prediction of diabetes within 12 months and 95% sensitivity for a diagnosis within three years, which puts the results on par with a fasting plasma glucose test of 126 mg/dl or greater.
“In Year One, if you have RPG of 130 to 139 mg/dl, one year later you have a 45-fold increased odds of having incident diabetes compared to someone with levels less than 110,” Rhee said. At Year Three, the higher test levels confer a 24-fold greater risk and a Year Five, the risk remains 17.5 times higher than someone with RPG levels below 100 mg/dl.
RPG levels proved a better predictor of future diabetes than age, sex, race/ethnicity, body mass index and cardiovascular disease risk factors, and adding those factors to RPG only slightly increased its predictive value, the researchers found.
Based on the study, Rhee hoped that, if providers “look at RPG levels that happened to be part of screening panels and see that a patient had at least two above certain thresholds, it would prompt the primary care provider to go forward with screening.”
The simplicity of the tool could help identify more veterans at risk of diabetes in time to intervene before they develop the disease. “Early identification is very important,” Rhee said. “Good diabetes control early on helps prevent diabetes complications down the road. It tends to slow progressive worsening of diabetes, reduce complications, keep patients from going on insulin so soon.”
- Rhee MK, Ho YL, Raghavan S, Vassy JL, Cho K, Gagnon D, Staimez LR, Ford CN, Wilson PWF, Phillips LS. Random plasma glucose predicts the diagnosis of diabetes. PLoS One. 2019 Jul 19;14(7):e0219964. doi: 10.1371/journal.pone.0219964. eCollection 2019.
- Liu Y, Sayam S, Shao X, et al. Prevalence of and Trends in Diabetes Among Veterans, United States, 2005-2014. Prev Chronic Dis. 2017;14:E135. Published 2017 Dec 14. doi:10.5888/pcd14.170230.
