Before the clinical onset of xerophthalmia, mild vitamin A deficiency leads to early keratinizing metaplasia and losses of mucinsecreting goblet cells on the bulbar surface of the conjunctiva of the eye. These functional changes on the ocular surface can be detected by microscopic examination of PAS-hematoxylin stained epithelial cells obtained by briefly applying a cellulose acetate filter paper strip (Hatchell and Sommer, 1984; Natadisastra et al., 1987; Wittpenn et al., 1986) or disc (Keenum et al., 1990) against the temporal conjunctivum. An alternative approach involves transferring cell specimens from the filter paper to a glass slide before staining and examination (Carlier et al., 1991). Specimens are classified as normal or into degrees of abnormality, depending on the density and distribution of stained normal epithelial cells, goblet cells, and mucin “spots” (contents of goblet cells). Vitamin A status is defined by target tissue cellularity, integrity, and function, which, unlike biochemical measures, if compromised may take several weeks to normalize following vitamin A repletion (Keenum, 1993). In spite of that, there is an association between the prevalence of conjunctival impression cytology (CIC) abnormality and serum retinol and RDR test results (Sommer and West, 1996). Although CIC is used for assessment, there are few data that relate CIC status to dietary vitamin A intake in the United States, other well-nourished populations, or malnourished populations. As a result, CIC was not selected as the functional indicator for the EAR for vitamin A.

There is sound evidence for a role of vitamin A in the maintenance of both humoral antibody responses and cell-mediated immunity. In experimental animals, both nonspecific immunity (Butera and Krakowka, 1986; Cohen and Elin, 1974) and antigen-specific responses, including delayed-type hypersensitivity (Smith et al., 1987), blastogenesis (Butera and Krakowka, 1986; Friedman and Sklan, 1989), and antibody production (Carman et al., 1989, 1992; Pasatiempo et al., 1990; Ross, 1996; Stephensen et al., 1993), have been shown to be altered by a deficiency of vitamin A or enhanced by vitamin A supplementation. The number and cytotoxic activity of natural killer cells (Dawson et al., 1999; Zhao et al., 1994) is reduced in vitamin A deficiency, although responsiveness to activation is maintained.

Several human studies have linked impairment in immunity to