Table 1.

The association of polygenic risk score (PRS) for schizophrenia and bipolar disorder with the corresponding diseases in 106806 unrelated participants form the Taiwan Biobank data. Participants with BPD were excluded when predicting SCZ, and vice versa.

		Schizophrenia 176 cases & 105919 controls			Bipolar disorder 695 cases & 105919 controls
PRS	# variants	OR	p-value	R2a (%)	OR	p-value	R2a (%)
SCZ	872,662	2.22	<0.001	4.32%	1.18	<0.001	0.23%
SCZ (EAS)	865,712	1.81	<0.001	2.39%	1.18	<0.001	0.23%
BPD	871,342	1.51	<0.001	1.14%	1.24	<0.001	0.39%
SCZBPD vs. CONT	867,670	1.63	<0.001	1.64%	1.24	<0.001	0.38%
SCZ vs.BPD	863,006	1.15	0.074	0.12%	0.92	0.024	0.06%
SCZ concordant	531,978	2.18	<0.001	4.10%	1.22	<0.001	0.33%
SCZ discordant	334,201	1.55	<0.001	1.32%	0.99	0.742	0.00%
BPD concordant	531,983	1.69	<0.001	1.86%	1.27	<0.001	0.48%
BPD discordant	334,204	0.94	0.369	0.03%	1.08	0.041	0.05%

^aIncrease in Nagelkerke pseudo R² when adding the PRS into the model including gender, age, batch version, and 20 population stratification dimensions.