. 2021 Jun 16;114(4):1286–1294. doi: 10.1093/ajcn/nqab193

TABLE 3.

Vitamin B-12 biomarkers in relation to PPI drug usage and atrophic gastritis characterized by vitamin B-12–fortified food intake (n = 3298)¹

	Fortified food consumption²
	Nonregular				Regular
	n = 1449				n = 1849
	Controls³	PPI users	Atrophic gastritis		Controls³	PPI users	Atrophic gastritis
	n = 797	n = 518	n = 134	P value⁴	n = 968	n = 698	n = 183	P value⁴
Age, y	73.0 (72.4, 73.5)^a	75.8 (75.1, 76.5)^b	73.1 (71.7, 74.4)^a	<0.001	72.6 (72.1, 73.2)^a	75.5 (74.6, 76.1)^b	74.7 (73.5, 75.8)^b	<0.001
Sex, n (% male)	284 (36)	167 (32)	44 (33)	0.422	334 (35)	216 (31)	60 (33)	0.312
BMI, kg/m²	27.7 (27.4, 28.1)	27.7 (27.3, 28.2)	28.3 (27.4, 29.1)	0.619	27.2 (26.9, 27.5)^a	28.4 (28.1, 28.8)^b	27.7 (26.9, 28.4)^ab	<0.001
Serum total vitamin B-12, pmol/L	257 (248, 265)^a	253 (241, 264)^a	196 (175, 216)^b	<0.001	277 (269, 285)^a	275 (265, 285)^a	256 (237, 276)^b	0.022
Serum holoTC, pmol/L	58.3 (56.2, 60.4)^a	53.5 (50.7, 56.4)^b	43.1 (38.0, 48.3)^c	<0.001	66.1 (64.2, 68.0)^a	62.7 (60.4, 65.1)^b	52.3 (47.7, 56.8)^c	<0.001
Plasma homocysteine, µmol/L	15.2 (14.7, 15.6)^a	15.8 (15.2, 16.4)^a	17.7 (16.6, 18.8)^b	0.001	13.6 (13.3, 14.0)^a	13.8 (13.4, 14.2)^a	15.3 (14.5, 16.1)^b	0.003
cB12 indicator, n (% deficient)⁵	166 (21)	136 (26)	67 (50)	<0.001	98(10)	115 (16)	55 (30)	<0.001

Data presented are adjusted mean (95% CI) unless otherwise indicated. cB12, combined B12; holoTC, holotranscobalamin; PPI, proton-pump inhibitor.

Participants were classed as nonregular (0–4 portions/wk) or regular (≥5 portions/wk) consumers of vitamin B-12–fortified foods. The most commonly eaten fortified foods were ready-to-eat breakfast cereals (providing added vitamin B-12 levels of 1.6–2.5 µg/100 g, equating to an intake of 0.5–0.8 µg/30-g serving), and to a lesser extent specific brands of fat spreads (providing 2.5–5 µg/100 g, equating to an intake of 0.25–0.5 µg/10-g serving). One participant had missing data for fortified food consumption and could not be classified, and therefore was not included in this analysis.

Participants without atrophic gastritis and nonusers of PPI drugs.

⁴

P < 0.05; analysis via χ² for categorical variables or ANCOVA for continuous variables (adjusted for age, sex, BMI, RBC folate, creatinine, alcohol units per week) on log-transformed data where appropriate with Bonferroni post hoc tests. Values across a row without a common superscript letter are significantly different.

⁵

cB12 indicator was calculated using serum total vitamin B-12, serum holoTC, plasma homocysteine, serum folate, and age to provide a combined indicator value. A value ≤ −0.5 was defined as deficient (20).