Table 1 Demographic and vitals statistics across identified T2D cases, controls, and T2D subtypes in AoU and MGB datasets. T2D total represents the aggregate across all three identified subtypes. Age, HbA1c, BMI are calculated by subgroup median on the date of T2D diagnosis. Sex, race, ethnicity, and income features are static.
From: Opportunistic screening of type 2 diabetes with deep metric learning using electronic health records
AoU | Green T2D subtype (nā=ā172) | Yellow T2D subtype (nā=ā833) | Red T2D subtype (nā=ā496) | |
|---|---|---|---|---|
Sex | Female | 63.40% | 62.20% | 56.20% |
Male | 34.30% | 34.90% | 41.40% | |
Race | Asian | 4.70% | 2.30% | 1.80% |
Black | 20.30% | 27.30% | 27.40% | |
White | 57.60% | 48.30% | 53.80% | |
Ethnicity | Non-Hispanic | 81.40% | 77.60% | 84.20% |
Hispanic | 14.50% | 17.80% | 11.60% | |
Age at diag | Median | 58 | 57 | 57 |
HbA1c at Diag | Median | 6.7 | 6.5 | 6.4 |
BMI at diag | Median | 31.4 | 34.1 | 36.2 |
eGFR at diag | Median | 72 | 63 | 60 |
MGB | Green T2D Subtype (nā=ā408) | Yellow T2D Subtype (nā=ā631) | Red T2D Subtype (nā=ā131) | |
Sex | Male | 51.50% | 54.40% | 36.60% |
Female | 48.50% | 45.60% | 63.40% | |
Race | Asian | 1.00% | 1.30% | 1.50% |
Black | 10.80% | 11.10% | 13.70% | |
White | 80.40% | 76.90% | 73.30% | |
Ethnicity | Non-Hispanic | 88.2% | 90.0% | 89.3% |
Hispanic | 2.0% | 1.7% | 1.5% | |
Age at diag | Median | 57.8 | 55.9 | 56.1 |
HbA1c at diag | Median | 6.8 | 6.7 | 6.6 |
BMI at diag | Median | 31.8 | 34.0 | 36.1 |
eGFR at diag | Median | 59.5 | 60 | 60 |