Table 2. Fully adjusted associations between dietary pattern trajectories from (6 to 24 months) and IQ¹ at 8 and 15 years of age (n = 7652)².

		8 years									16 years
		FSIQ			VIQ			PIQ			FSIQ
Trajectory		β3	95% CI	P	β	95% CI	P	β	95% CI	P	β	95% CI	P
Healthy	Intercept	0.66	−0.20, 1.51	0.134	1.00	0.08, 1.92	0.033	0.03	−0.86, 0.93	0.939	0.75	0.00, 1.50	0.049
Healthy	Slope	1.07	0.17, 1.97	0.020	0.95	0.02, 1.88	0.046	1.03	−0.01, 2.06	0.051	0.49	−0.28, 1.26	0.208
Discretionary	Intercept	−0.92	−1.98, 0.14	0.089	−1.51	−2.56, −0.45	0.006	0.03	−1.20, 1.26	0.959	−0.29	−1.14, 0.56	0.503
Discretionary	Slope	−0.35	−1.03, 0.33	0.307	−0.52	−1.16, 0.12	0.110	−0.10	−0.94, 0.73	0.802	−0.73	−1.33, −0.14	0.017
Traditional	Intercept	1.03	−0.20, 2.26	0.100	0.97	−0.46, 2.19	0.196	0.98	−0.26, 2.23	0.121	0.09	−0.92, 1.09	0.860
Traditional	Slope	−0.19	−0.71, 0.33	0.466	−0.11	−0.64, 0.41	0.666	−0.20	−0.79, 0.39	0.495	−0.41	−0.77, −0.04	0.031
Ready-to-eat	Intercept	−3.83	−9.76, 2.11	0.205	−2.61	−9.38, 4.15	0.444	−4.42	−12.21, 3.38	0.262	−0.14	−5.87, 5.60	0.962
Ready-to-eat	Slope	0.32	−4.31, 4.95	0.891	2.42	−3.51, 8.35	0.414	−2.52	−8.53, 3.50	0.404	1.11	−3.10, 5.33	0.597

¹Abbreviations: FSIQ, full scale intelligence quotient; IQ, intelligence quotient; PIQ, performance intelligence quotient; VIQ, verbal intelligence quotient

²Eligible participants were children who had at least one measurement of IQ collected at either 8 or 15 years of age (n = 7652). IQ was measured using the Wechsler Intelligence Scale for Children at 8 years and the Wechsler Abbreviated Scale of Intelligence at 15 years of age. Incomplete IQ or missing covariable data was imputed by Multiple Imputation.

³Analyses show the beta-coefficient from multivariable linear regression analysis using generalized linear models and adjusted for the following covariables; sex, gestational age at birth, birth weight, ethnicity, singleton/twin, maternal age, parity, social class (according to standard UK classifications of occupation at the time of birth[17]), maternal education, other children, family income, maternal smoking, stimulation in the home environment (using an adaptation of the HOME questionnaire[18]) and maternal IQ (measured by WASI when the study child was 15 years of age). A one standard deviation change in Healthy dietary pattern trajectory from 6 to 24 months of age is associated with 0.66 (95% CI −0.20, 1.51) higher FSIQ scores at 8 years of age.