Dear Sir:

The article by Tang et al (1) represents a valuable contribution to the discussion about the potential nutritional effect of β-carotene-containing rice. With the use of an impressive methodology, the authors provide some evidence for an uptake of β-carotene from rice in humans.

Given the enormous effort and elegance in the scientific methods used the study, however, the study fail to provide significant data for the evidence of a nutritional benefit from Golden Rice because of its biomathematical deficits. This would not be of great concern if the data were used to justify the statement that β-carotene can be absorbed from rice. Unfortunately, the data are used to advertise for the suggested benefits of the technology of genetically modified organisms in populations who may not be able to qualify the study results and conclusions drawn. Apart from this larger, principal, discussion, there are 2 critical questions regarding the data presented in the study.

In Table 3, the authors present their main findings. The table is reproduced here with the addition (in the lower section in bold type) of the median and the magnitude of the SD and the mean-median difference and its magnitude (calculated from the data provided by the authors in the upper lines). The means and SDs on the basis of 5 probands with large interindividual variability is a weak basis for far-reaching nutritional conclusions. The SDs of the results range from 29% to 51% of the mean in a nonnormally distributed data set. Therefore, the statement “our analysis showed a very efficient bioconversion of β-carotene to vitamin A” is based on 2 of 5 values above the median in Table 3. Even considering the limited amount of intrinsically labeled β-carotene-containing rice available—with ≈20 μg β-carotene/g rice—it is to be questioned why the research group did not choose a more homogenous study population at least in terms of the variables of age, sex, and nutritional and vitamin A status, at the start.

A second question concerns why the authors did not use a dietary approach more similar to the diets of the individuals who were suggested to benefit from the consumption of this β-carotene-containing rice. One of the arguments used for advertising Golden Rice is that the people at risk of vitamin A deficiency have such poor diets that other sources of β-carotene and vitamin A are not accessible to them. Because diet definitely has an effect on the bioavailability of β-carotene from any β-carotene-containing food, the choice for a study diet that included meat, oil, and nuts, which does not represent a poor diet, is of concern. Therefore, the results of the study do not much help us in preventing vitamin A deficiency in populations at risk. The argument of a better conversion rate with β-carotene-containing rice may at best be interpreted as follows: This rice is to be considered as one means of providing β-carotene besides the known vegetables and algae and in absence of animal-derived dietary sources of vitamin A. The suggested superior conversion rate alone does not solve all intrinsic nutritional, medical, and social problems of the “Golden Rice approach” in preventing vitamin A deficiency.

More research in the prevention of vitamin A deficiency is required, and animal studies in piglets may be an appropriate model to investigate the different approaches of supplementation, fortification, natural β-carotene from the diet, and nutrient-oriented plant breeding before humans are further exposed to studies that obviously do not address potential health risks.