Folate status of gastrointestinal epithelial cells is not predicted by serum and red cell folate values in replete subjects.

Localised folate deficiency has been implicated in colonic carcinogenesis and supplementation has been proposed for certain populations at risk. However, identifying those groups that may benefit is difficult as the relation between blood folate and gut epithelial cell values is unknown. The aim of this study was to define this relation. Epithelial cells mean (SEM) (sigmoid: 5.35 (0.56) x 10(6) cells, caecum: 6.6 (0.71) x 10(6) cells, duodenum: 4.0 (0.62) x 10(6) cells) were isolated from four endoscopic mucosal biopsy specimens (n = 25) by incubation with dithiothreitol (three hours) and EDTA (one hour). Lamina propria contamination was < 1%, with < 6% intraepithelial lymphocytes. Folate assay of isolates showed sigmoid colon folate content to be 20.1 (1.8) pg/micrograms DNA (10.2-46.6). In the same subject, caecal folate concentrations were lower (p < 0.01, n = 11) than sigmoid values, whereas duodenal isolates mirrored those of the sigmoid (19.4 (2.9) v 20.5 (3.2), n = 5). Sigmoid folate values were consistent over one to three weeks (n = 3). In a single case with blood folate deficiency, colonic values were normal. Serum folate and red cell folate correlated poorly with sigmoid epithelial cell folate content (r = 0.41, p = 0.063 and r = 0.17, p > 0.05 respectively). This study reports a modified ion-chelation isolation method for colonic biopsy specimens that yields large numbers of viable epithelial cells. Cell folate values remain constant with time though vary with intestinal region. The inability of serum or red cell folate values to predict those of the sigmoid epithelium suggests that they cannot identify those patients that might benefit from folate supplements.