Polygenic risk score for predicting diabetic retinopathy in patients with type 2 diabetes: A twenty-year follow-up study

Table 1 Characteristics of the patients with type 2 diabetes in the discovery and replication cohorts, mean SD/n (%)

	Discovery cohort (n = 9670)		P value1	Replication cohort (n = 3665)		P value1
	T2D with DR (n = 4892)	T2D without DR (n = 4778)		T2D with DR (n = 2403)	T2D without DR (n = 1262)
Diagnosis age (years)	61.8 12.2	64.7 ± 13.6	< 0.001	59.2 ± 11.9	65.1 ± 13.3	< 0.001
Male	2611 (53.4)	2758 (57.7)	< 0.0012	1238 (51.5)	712 (56.4)	0.0052
Female	2281 (46.6)	2020 (42.3)		1165 (48.5)	550 (43.6)
HbA1c (%)	7.6 1.3	7.2 ± 1.0	< 0.001	7.8 ± 1.5	7.2 ± 1.1	< 0.001
BMI (kg/m2)	26.0 4.3	26.3 ± 4.4	0.001	26.3 ± 4.9	26.3 ± 4.4	0.931
Follow-up duration [months, median (min-max)]	24 (0-228)	108 (60-240)		36 (0-228)	108 (60-240)

¹Student t-test.

²χ² test.

HbA1c: Glycated hemoglobin A1c; BMI: Body mass index; DR: Diabetic retinopathy; T2D: Type 2 diabetes.

Table 2 The ten most significantly affected biological pathways and molecules related to diabetic retinopathy identified through ingenuity pathway analysis

Ingenuity canonical pathways	-log (P value)	Molecules
Assembly of RNA polymerase I complex	2.57	TAF1C
Rap1 signalling	2.48	PRKG1
BMAL1: CLOCK, NPAS2 activates circadian gene expression	2.26	KLF15
SIRT1 negatively regulates rRNA expression	2.22	TAF1C
Oxytocin in spinal neurons signaling pathway	2.14	PRKG1
Netrin-1 signaling	1.99	SLIT3
RNA polymerase I transcription	1.99	TAF1C
B-WICH complex positively regulates rRNA expression	1.97	TAF1C
Platelet homeostasis	1.75	PRKG1
Beta-catenin independent Wnt signaling	1.68	PRKG1

Rap1: Ras-associated protein 1; BMAL1: Brain and muscle ARNT-like 1; NPAS2: Neuronal PAS domain protein 2; SIRT1: Silent information regulator sirtuin 1; TAF1C: TATA-box binding protein associated factor, RNA polymerase I subunit C; PRKG1: CGMP-dependent protein kinase 1; KLF15: Kruppel-like factor 15; SLIT3: Slit guidance ligand 3.