Table 3. Linear regression models of MTHFD1L rs11754661 for rumination score as an outcome variable, separately in Budapest and Manchester.

	Budapest						Manchester
	Predictor variable	N	Beta	s.e.	t	P	Predictor variable	N	Beta	s.e.	t	P
Additive	MTHFD1L rs11754661	862	0.158	0.054	2.915	0.004	MTHFD1L rs11754661	1258	0.095	0.046	2.049	0.041
	Age	862	−0.004	0.002	−2.417	0.016	Age	1258	−0.008	0.002	−4.693	<0.001
	Gender	862	0.220	0.034	6.409	<0.001	Gender	1258	0.311	0.037	8.456	<0.001
Dominant	MTHFD1L rs11754661	862	0.157	0.055	2.828	0.005	MTHFD1L rs11754661	1258	0.107	0.050	2.120	0.034
	Age	862	−0.004	0.002	−2.395	0.017	Age	1258	−0.008	0.002	−4.700	<0.001
	Gender	862	0.220	0.034	6.419	<0.001	Gender	1258	0.310	0.037	8.437	<0.001

PLINK linear regression equations were constructed with the predictor variables displayed in the rows. Additive and dominant models were run, separately in Budapest and Manchester; all with A as the minor allele. (Recessive models have not been run because of low number in AA groups.) Significant findings for MTHFD1L rs11754661 as a predictor variable are marked with bold.