Table A4.

Mean and standard deviation of the estimates of the imputation methods using the PLoD missing mechanism, the three sample sizes, the three missing percentages, and four metabolites.

Sample Size	Missing Percentage	Metabolites/Imputation Method/Mean Estimate (SD)
		PC(32:2)	kNN-obs-sel	Succinylcarnitine	kNN-obs-sel	Glutamate	kNN-obs-sel	Urate	kNN-obs-sel
		MICE-pmm		MICE-pmm		MICE-pmm		MICE-pmm
n = 50	0%	−4.18 × 10−7		2.84 × 10−6		1.83 × 10−7		1.39 × 10−8
	15%	−4.32 × 10−7 (3.91 × 10−8)	−4.34 × 10−7 (3.35 × 10−8)	4.4 × 10−6 (2.56 × 10−7)	4.62 × 10−6 (1.79 × 10−7)	2.87 × 10−8 (1.09 × 10−9)	2.94 × 10−8 (1.29 × 10−9)	8.36 × 10−9 (2.53 × 10−9)	8.99 × 10−9 (2.19 × 10−9)
	30%	−4.22 × 10−7 (7.47 × 10−8)	−4.28 × 10−7 (5.19 × 10−8)	4.15 × 10−6 (4.4 × 10−7)	4.72 × 10−6 (2.83 × 10−7)	2.77 × 10−8 (2.07 × 10−9)	2.99 × 10−8 (1.99 × 10−9)	7.07 × 10−9 (3.7 × 10−9)	8 × 10−9 (3.14 × 10−9)
	60%	−3.52 × 10−7 (1.82 × 10−7)	−4.51 × 10−7 (1.32 × 10−7)	3.05 × 10−6 (1.18 × 10−6)	4.96 × 10−6 (1.19 × 10−6)	2.33 × 10−8 (4.89 × 10−9)	3.32 × 10−8 (3.59 × 10−9)	1.21 × 10−9 (7.1 × 10−9)	3.25 × 10−9 (7.47 × 10−9)
n = 150	0%	−3.64 × 10−7		1.53 × 10−6		2.89 × 10−8		9.58 × 10−9
	15%	−3.65 × 10−7 (1.26 × 10−8)	−3.58 × 10−7 (1.22 × 10−8)	1.51 × 10−6 (3.88 × 10−7)	1.72 × 10−6 (3.61 × 10−7)	1.64 × 10−8 (6.63 × 10−10)	1.67 × 10−8 (5.85 × 10−10)	9.18 × 10−9 (1.62 × 10−9)	9.08 × 10−9 (1.31 × 10−9)
	30%	−3.65 × 10−7 (1.96 × 10−8)	−3.52 × 10−7 (1.86 × 10−8)	1.49 × 10−6 (6.11 × 10−7)	1.86 × 10−6 (5.51 × 10−7)	1.61 × 10−8 (1.07 × 10−9)	1.7 × 10−8 (8.47 × 10−10)	8.91 × 10−9 (2.14 × 10−9)	8.25 × 10−9 (1.87 × 10−9)
	60%	−3.79 × 10−7 (3.83 × 10−8)	−3.74 × 10−7 (3.83 × 10−8)	1.66 × 10−6 (1.05 × 10−6)	2.11 × 10−6 (9.8 × 10−7)	1.5 × 10−8 (1.8 × 10−9)	1.8 × 10−8 (1.51 × 10−9)	5.83 × 10−9 (3.36 × 10−9)	3.91 × 10−9 (3.08 × 10−9)
n = 599	0%	−4.38 × 10−7		4.53 × 10−6		1.66 × 10−8		9.69 × 10−9
	15%	−4.2 × 10−7 (7.17 × 10−9)	−4.18 × 10−7 (8.61 × 10−9)	2.82 × 10−6 (1.59 × 10−7)	3.05 × 10−6 (1.5 × 10−7)	1.82 × 10−8 (3.37 × 10−10)	1.86 × 10−8 (3.1 × 10−10)	1.4 × 10−8 (6.57 × 10−10)	1.42 × 10−8 (5.63 × 10−10)
	30%	−4.23 × 10−7 (1.1 × 10−8)	−4.25 × 10−7 (1.52 × 10−8)	2.79 × 10−6 (2.52 × 10−7)	3.34 × 10−6 (2.34 × 10−7)	1.8 × 10−8 (5.37 × 10−10)	1.89 × 10−8 (4.81 × 10−10)	1.4 × 10−8 (9.58 × 10−10)	1.44 × 10−8 (8.97 × 10−10)
	60%	−4.39 × 10−7 (2.02 × 10−8)	−4.66 × 10−7 (3.56 × 10−8)	2.58 × 10−6 (5.62 × 10−7)	4.37 × 10−6 (6.13 × 10−7)	1.76 × 10−8 (9.99 × 10−10)	2.07 × 10−8 (8.04 × 10−10)	1.39 × 10−8 (1.64 × 10−9)	1.45 × 10−8 (1.57 × 10−9)

Estimates are the regression coefficient from the model BMI ~ age + sex + metabolite. Therefore, the estimates are the mean increase in BMI per 1 unit increase of the metabolite. The 0% rows are the estimates from the real data before amputing and imputing the missing values.