Evaluation of methods for assessing visual function of infants

Purpose One of the more crucial questions concerning infant supplementation of docosahexaenoic acid (DHA) is related to the role of this fatty acid in visual development. However, the optimal method for assessing visual function of infants has not been established and is complicated by the lack of a *gold standard.* Method We have evaluated the test/retest reliability of transient pattern visual evoked responses (VER) latency and amplitude, acuity card procedure and sweep VER by studying 18 four month old infants. While not commonly used, the transient VER does reflect photoreceptor, synaptic, myelin and cortical function of several brain layers. DHA has been shown to affect the photoreceptor outer segments and brain development. Results Looking at percent difference between test and retest as well as SD, transient VER latency was lowest (3%, 7% SD). The other methods were markedly more variable: sweep VER (2%, 22% SD), acuity card (8%, 20% SD) and transient VHP amplitude (7.5%, 39% SD) Average coefficients of variation showed a similar trend: transient latency 8%, sweep VER 15%, acuity cards 30% and transient amplitude 53%. Furthermore there was a poor correlation among the various assessment tests. Conclusions While all methods provide valid and reliable data for a group, these assessment methods are less valid for estimating visual acuity/function of an individual subject. The poor correlations suggest that each test may be assessing a different aspect of vision, support none.