Can lack of improvement in vitamin A status indicators be explained by little or no overall change in vitamin A status of humans?

Changes in vitamin A status using conventional indicators, i.e., serum and breast milk retinol concentrations and the modified relative dose response test, following a vitamin A intervention have not always been shown. A simplified model to predict calculated changes in vitamin A status after intervention is described. The model shows that changes in indicator values cannot be expected if the change in vitamin A status is only marginal. A critical review of several papers using vitamin A status assessment indicators was undertaken. Assumptions that included current knowledge concerning vitamin A absorption and metabolism were made and applied to the data. Based on current recommended daily allowances for women and children, one cannot necessarily expect a change in indicators if an overall change in vitamin A status was not achieved. Thus, when designing vitamin A intervention studies, the following parameters should be considered if applicable to the population to be enrolled in the study: average body weight, estimated liver weight, amount of vitamin A administered, estimated loss in breast milk and study duration.