Table 2. Algorithm to estimate risk for incident type 2 diabetes mellitus using total points for the categorical model with logistic regression analysis in the 6-year follow-up cohort in Beijing and Shanghai, China, 2005-2011.

Items	Reference value (Wij)	βi	P value	βi(Wij-WiREF)	Pointij = βi(Wij-WiREF)/B a
Sex		0.2271 ( = B)	0.027
Male	0 (W1REF)			0	0
Female	1			0.2271	1
Body mass index (kg/m2)		0.0533	0.001
<24	21.4 (W2REF )			0	0
24 to <28	25.8			0.2345	1
≥28	29.2			0.4157	2
Fasting plasma glucose (mmol/L)		0.9624	<0.001
<5.6	5.1 (W3REF )			0	0
5.6 to <6.1	5.8			0.6437	3
­ ≥ 6.1	6.4			1.2511	6
HbA1c (%)		1.3375	<0.001
<5.5	5.3 (W4REF )			0	0
5.5 to <5.9	5.7			0.5350	2
≥5.9	6.1			1.0700	5
Hypertension (%)		0.2420	0.019
No	0 (W5REF )			0	0
Yes	1			0.2420	1
C-reactive protein (mmol/L)b		0.1104	0.019
<0.39	-2.12c			0	0
0.39 to 0.99	-0.49			0.1800	1
≥1.00	0.65			0.3058	2
Maxim total point					17
Total point	0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
6-year risk of type 2 diabetes, %	0.20 0.24 0.29 0.34 0.39 0.44 0.50 0.56 0.61 0.66 0.71 0.76 0.80 0.83 0.86 0.89 0.91 0.92

^aThe points are rounded to the nearest integer.

^bC-reactive protein was tertiled in the current sample.

^cThe reference figure was the median of natural log transformed values in each C-reactive protein category.