Table 4C.

Associations between maternal serum ferritin concentrations at enrolment (Thurnham adjusted) and pregnancy and infant outcomes

	Model 1b			Model 2c
Variablesa	n	β (SE) or RR (95%CI)	P-value	n	β (SE) or RR (95%CI)	P-value
Pregnancy outcomes
Birth weight, g	242	32.1 (30.5)	0.29	242	32.3 (29.9)	0.28
Low birth weight (<2500 g), n (%)	242	0.95 (0.69, 1.31)	0.74	242	0.95 (0.69, 1.32)	0.77
Gestational age at birth, weeks	248	0.14 (0.14)	0.30	248	0.15 (0.14)	0.28
Preterm (<37 weeks), n (%)	248	0.83 (0.59, 1.16)	0.27	248	0.81 (0.57, 1.14)	0.22
Preterm and low birth weight, n (%)	242	1.04 (0.63, 1.74)	0.87	242	1.03 (0.61, 1.71)	0.92
Small for gestational aged, n (%)	241	0.93 (0.76, 1.12)	0.43	241	0.91 (0.74, 1.11)	0.35
Infant outcomes
Haemoglobin, g/dl	68	0.28 (0.23)	0.23	68	0.33 (0.23)	0.16
Anaemia (<11.0 g/dl), n (%)	68	0.77 (0.57, 1.03)	0.07	68	0.79 (0.60, 1.04)	0.09
Length, cm	217	0.47 (0.20)	0.02	217	0.45 (0.20)	0.02
Length-for-age z-score (LAZ)	215	0.17 (0.10)	0.07	215	0.16 (0.10)	0.10
LAZ < −2	215	0.85 (0.69, 1.06)	0.14	215	0.87 (0.69, 1.09)	0.22
Weight-for-age z-score (WAZ)	241	0.07 (0.07)	0.34	241	0.07 (0.07)	0.33
WAZ < −2	241	0.89 (0.61, 1.28)	0.52	241	0.89 (0.62, 1.29)	0.55
Weight-for-length z-score (WLZ)	177	−0.13 (0.14)	0.35	177	−0.12 (0.14)	0.37
WLZ < −2	177	1.02 (0.71, 1.48)	0.90	177	1.06 (0.74, 1.53)	0.74
BMI z-score (BMIZ)	214	−0.15 (0.11)	0.18	214	−0.14 (0.11)	0.18
BMIZ < −2	214	1.07 (0.78, 1.47)	0.67	214	1.09 (0.80, 1.50)	0.58
Ponderal Index, g/cm3	216	−0.0007 (0.0004)	0.06	216	−0.0007 (0.0004)	0.06
Head circumference, cm	217	−0.05 (0.12)	0.70	217	−0.05 (0.12)	0.68
Chest circumference, cm	217	−0.05 (0.14)	0.73	217	−0.06 (0.14)	0.70
Waist circumference, cm	217	−0.07 (0.20)	0.72	217	−0.05 (0.20)	0.78
Hip circumference, cm	217	−0.18 (0.27)	0.52	217	−0.20 (0.27)	0.46
Mid-upper arm circumference, cm	215	0.33 (0.22)	0.13	215	0.36 (0.22)	0.09
Biceps skinfold, mm	216	0.04 (0.07)	0.57	216	0.04 (0.07)	0.60
Triceps skinfold, mm	216	−0.15 (0.11)	0.16	216	−0.15 (0.11)	0.15

^aBaseline serum ferritin concentrations were adjusted for inflammation using Thurnham [21, 22] methods; natural log-transformed; beta (SE) is reported for continuous pregnancy outcomes; RR (95%CI) is reported for binary pregnancy outcomes

^bModels adjusted for gestational age of sample and vitamin B₁₂ intervention

^cModels adjusted for gestational age of sample, vitamin B₁₂ intervention, maternal BMI, standard of living index (SLI) >28, educational level, and ln(CRP)

^dSmall for gestational age (SGA) was defined by sex and gestational age <10th percentile using INTERGROWTH methods [17]