Table 3.

Association of polygenic risk score (PRS) for schizophrenia and bipolar disorder with MMSE and educational attainment.

	Ordinal education (n = 105,710)		MMSE (n = 27,005)		MMSE < 24 (n = 27,005)		MMSE change (n = 6194)
PRS	aOR (95% CI)c	P	Beta (95% CI)a	P	aOR (95% CI)a	P	Beta (95% CI)b	P
Model 1: independent effect
SCZ	1.044 (1.031–1.057)	<0.001	−0.069 (−0.115, −0.023)	0.003	1.066 (1.008–1.127)	0.025	−0.045 (−0.135, 0.046)	0.334
BPD	1.021 (1.009–1.033)	0.001	0.054 (0.009, 0.010)	0.020	0.917 (0.867–0.970)	0.003	0.067 (−0.023, 0.158)	0.146
Model 2: SCZ & BPD shared
SCZBPD vs. CONT	1.053 (1.041–1.065)	<0.001	−0.023 (−0.067, 0.020)	0.294	1.019 (0.967–1.075)	0.477	0.038 (−0.049, 0.124)	0.395
Model 3: SCZ-specific
SCZ vs.BPD	0.986 (0.975–0.998)	0.017	−0.037 (−0.080, 0.006)	0.092	1.015 (0.963–1.069)	0.589	−0.125 (−0.210, −0.040)	0.004
Model 4: split effect
SCZ concordant	1.030 (1.013–1.047)	0.001	−0.101 (−0.164, −0.038)	0.002	1.088 (1.007–1.175)	0.033	−0.062 (−0.189, 0.064)	0.335
SCZ discordant	1.011 (0.996–1.026)	0.142	0.035 (−0.021, 0.090)	0.221	0.987 (0.922–1.056)	0.707	−0.019 (−0.129, 0.090)	0.730
BPD concordant	1.022 (1.006–1.038)	0.007	0.073 (0.013, 0.133)	0.017	0.918 (0.853–0.989)	0.024	0.082 (−0.038, 0.202)	0.180
BPD discordant	1.001 (0.987–1.015)	0.894	0.049 (−0.005, 0.102)	0.073	0.950 (0.890–1.014)	0.122	−0.004 (−0.108, 0.101)	0.947

^aAdjustment for gender, age, birth cohort, education level, TWB (chip version) and first 20 population stratification dimensions.

^bAdjustment for gender, age, birth cohort, education level, duration of follow-up, TWB (chip version) and first 20 population stratification dimensions.

^cAdjustment for gender, age, birth cohort, TWB (chip version), and first 20 population stratification dimensions.