Research Article

Identification of Potential Type II Diabetes in a Large-Scale Chinese Population Using a Systematic Machine Learning Framework

Table 1

Characteristics of variables.

VariablesDiabetes ()Nondiabetes () value

Age (years)<0.001
BMI (kg/m2)<0.001
Waist circumference (cm)<0.001
Systolic pressure (mmHg)<0.001
Diastolic pressure (mmHg)<0.001
Ethnicity, (%)<0.001
 Han50,691 (70.38)331,413 (64.93)
 Uygur10,864 (15.08)95,913 (18.79)
 Kazak1147 (1.59)18,893 (3.70)
 Hui8126 (11.28)52,838 (10.35)
 Mongolian76 (0.11)1214 (0.24)
 Other nationalities1123 (1.56)10,140 (1.99)
Gender, (%)<0.001
 Male34,641 (48.09)239,875 (47.00)
 Female37,386 (51.91)270,536 (53.00)
Physical activity, (%)<0.001
 Yes26,239 (36.43)154,585 (30.29)
 No45,788 (63.57)355,826 (69.71)
Drinking status, (%)<0.001
 Yes15,944 (22.14)102,852 (20.15)
 No56,083 (77.86)407,559 (79.85)
Drinking amount (g)<0.001
 ≥1706687 (9.30)39,479 (7.73)
 <17065,240 (90.70)470,932 (92.27)
Smoking amount (cigarettes)10 (8-20)10 (7-20)<0.001
Smoking status, (%)<0.001
 Yes10,683 (14.83)63,920 (12.52)
 No61,344 (85.17)446,491 (87.48)
Dietary ratio, (%)<0.001
 Meat based2849 (3.96)13,554 (2.66)
 Meat balanced66,603 (92.47)482,864 (94.60)
 Vegetarian based2575 (3.58)13,993 (2.74)
Sugar loving, (%)<0.001
 Yes940 (1.31)4560 (0.89)
 No71,087 (98.69)505,851 (99.11)
Oil loving, (%)<0.001
 Yes2722 (3.78)13,068 (2.56)
 No69,305 (96.22)497,343 (97.44)
Salt loving, (%)<0.001
 Yes4261 (5.92)20,896 (4.09)
 No67,766 (94.08)489,515 (95.91)
Fatty liver, (%)<0.001
 Yes22,331 (31.00)52,800 (10.34)
 No49,696 (69.00)457,611 (89.66)
Hypertension, (%)<0.001
 Yes29,937 (41.56)112,348 (22.01)
 No42,090 (58.44)398,063 (77.99)

Median (IQR). Abbreviation: BMI: body mass index.