Total choline intake and working memory performance in adults with phenylketonuria

Table 4 Results of the multivariable regression analyses of working memory performance with energy-adjusted total choline intake and metabolic control^a

Model	N	Regression coefficient: choline intake (95% CI)^b	p-value^c	Regression coefficient: choline intake X Phe (95% CI)^d	p-value^c	Adjusted R²
1	40	0.04 (− 0.05, 0.1)	0.38	− 0.004 (− 0.02, 0.009)	0.50	0.36
2	30	0.10 (− 0.001, 0.19)	0.05	− 0.01 (− 0.02, 0.004)	0.16	0.44
3	30	0.11 (0.02, 0.20)	0.02	− 0.01 (− 0.02, 0.002)	0.10	0.52

^aAdjusted for intellectual ability (matrix reasoning) in model 1 (full sample, n = 40) and model 2 (reduced sample, n = 30); adjusted for intellectual ability and IDC from 6 to 11 years in model 3 (exploratory analysis with reduced sample, n = 30). Matrix reasoning (MR), IDC, and energy-adjusted choline intake were centered at sample means (MR = 10, IDC = 353 µmol/L, choline = 366 mg) prior to inclusion in the regression models. Most recent Phe was centered at 360 µmol/L
^bRegression coefficient represents the change in working memory performance per 100 mg/day increase in energy-adjusted total choline intake when Phe is optimal (360 µmol/L)
^cSignificance indicated at p ≤ 0.05; trends informative for future research indicated at p > 0.05 and ≤ 0.1
^dRegression coefficient represents the amount by which the association between energy-adjusted total choline intake and working memory performance changes based on a 50 µmol/L increase in most recent Phe concentration

ISSN: 1750-1172