Keep up with the latest news and comment on genetically modified foods          

2013 articles

Golden rice myths

on .

Golden Rice in the field

Dr Michael Hansen debunks some of the many myths promoted about Golden Rice.


There has been a lot of misinformation and just plain propaganda about Golden Rice (GR) put out in the last week or so. It is stunning that so many are talking about this topic with out even a basic understanding of the real issues. Bottom line, there are clear unanswered questions on basic efficacy and safety of GR.

Here are some of the basic issues. First, both GR 1 and GR 2 (released in 2005) are Japonica rices. For those that don't know, there are basically two subspecies of Oryza sativa - the sticky, short-grained Japonica, and the non-sticky long-grained Indica. It's the Indica varieties that are grown submerged in paddies, while the Japonica is a dryland rice. The bottom line is that in the areas where people are people really are starving/have Vit A deficiency - India, Sri Lanka, Bangladesh, Philippines - the vast majority of the population eats Indica rices, not Japonica. As pointed out in the 2008 article in Science, “there’s a long way to go. . . Both the original Golden Rice, now called GR1, and GR2 were created with Japonica cultivars that are the scientist’s favorites but fare poorly in Asian fields. Researchers are now backcrossing seven GR1 and GR2 lines with the long-grained, nonsticky Indica varieties popular among Asia’s farmers” (see:

The reason that GR1 and GR2 are Japonica, is that it was more difficult to successfully engineer Indica varieties. IRRI knows this and so was trying to cross the GR2 with local, popular Indica varieties. IRRI had even stated that they would start field testing GR 8 Indica varieties in 2008 in Philippines, India, and Bangladesh. I was in Bangladesh in summer of 2009 and talked to the IRRI scientists there and they had not even seen a GR Indica variety. They told me there were promised seeds for a while, but still had seen nothing.

In the Philippines, the testing of a GR Indica variety began in October 2010; I believe it is a cross between GR2 and IR-64 (a popular Indica variety). At a 2011 hearing of in the Philippines Parliament, I testified and challenged the head of PhilRice GR project to present any data on the β-carotene levels in the GR Indica varieties, but the only data were from GR2. Also, Dr. Gerard Barry, head of IIRI's Golden Rice Project leader, who came to IRRI in 2003 after having worked for Monsanto for 20 years (1983-2003), was in the room but didn't say much.

I was in Bangladesh last August and visited BARI (Bangladesh Agricultural Research Institute). They did have GR2 Indica plants but they were in the greenhouse and no field trials had started. I was also in Vietnam last August and they too had GR2 Indica plants in the lab, but no field test.

So, there are no data on the β-carotene levels in GR Indica varieties, which is what the poor is South Asia actually eat.

Second, there are no real data on levels of β-carotene in rice over time. According to Golden Rice Humanitarian Board, “Because of their chemical nature - several conjugated double bonds - carotenoids are susceptible to light and oxidation.” Thus, β-carotene can break down in the presence of oxygen and light. So, the real question is what are the β-carotene levels in rice that has sat in storage at room temperature for month or two, similar to the local storage conditions for those who might grow this rice. Again, no studies have been done. It should be pointed out that when some GR1 was sent in 2001 to scientists in Germany to do bioavailability studies, the level of β-carotene was less that 1% of what is should have been; after cooking the level declined by 50%. Could the reason for such low levels of β-carotene in GR1 sent to Germany be due to the time delay from harvest to when the rice was tested in Germany? I'd bet the answer is yes. Again, no data have been presented on this issue.

As for bioavailability, there are the two studies - the controversial children-feeding study published last year and the earlier feeding study involving adults, published in 2009. Both of these studies have fundamental flaws, e.g. the experiment was designed in a way to give the best possible results. First, since levels of β-carotene may decline over time, it's interesting to look at the design of the feeding studies to see how long of a time there was between harvest and the study. For the 2009 study, the GR2 rice grains were harvested and then almost immediately put in storage at -70oC (which would delay any decline in β-carotene levels) until the feeding study began. In the 2012 study, the GR2 was harvested, allowed to dry for 3 days, and then put in storage at -70oC. So, GR2 rice used in the feeding studies was either immediately harvested or allowed to sit for 3 days before being put in storage, in part to prevent any decline in β-carotene levels. Second, the adult feeding study was manipulated to maximize absorption of β carotene. Since β-carotene is fat soluble, oil is needed to ensure absorption. So, the adult feeding trial consisted of a meal of 65 - 98 grams of GR plus 10 grams of butter. In other words, over 10% of the meal was butter, so of course absorption was good. What would the figure look like if the only part of the meal was GR without any butter? The poor in Asia will not be consuming rice with 10% butter.

So, the design of the feeding studies were biased toward producing the best possible result for GR, rather than designed for a more realistic situation, e.g. GR rice that has been stored at room temperature for a month or two and meals that don't contain over 10% butter by weight.

On the safety side, there is the basic issue of levels of retinoic acid and other retinoids. Retinoic acid (RA) is a potent teratogen. Indeed, RA is the active ingredient in an acne medication that will not be prescribed to women of childbearing age. Any way, as many know, the golden color of GR was an unintended effect. The original design of GR involved inserting two genes that code for enzymes (phytoene synthase and crt-1) that would transform a precursor (geranylgeranyl-PP) into lycopene, as the scientists were trying to create a red rice. Red rice is naturally found in South Asia, particularly southern India and Sri Lanka, but the red color (lycopene) is found in the outer layers of the rice so that when it is milled, the lycopene is largely removed. The scientists were trying to create a red rice where the lycopene is in the endosperm and so would not be milled away. Unknown to the scientists, the rice turned on an endogenous gene, which coded for lycopene cyclase, which transforms lycopene into β-carotene. Hence, the unexpected yellow color.

Β-carotene is transformed into retinal in the presence of the enzyme oxygenase. If retinal is reduced, it forms retinol, aka Vit A; if retinal is oxidized, it forms retinoic acid, a potent teratogen. When I pointed out at the Philippine House of Representatives that GR experiment lead to an unexpected increase in β-carotene, and that they should look at RA levels, since there are only two steps in a metabolic pathway between β-carotene and RA, and since trying to engineer biosynthetic pathways can cause all sorts of unintended effects, the IRRI scientist could produce no data on RA levels, much less the levels of other retinoids. He argued that people have been eating foods such as carrots, that are high in β-carotene levels (higher than the levels of GR), for hundreds of years, yet there's no evidence of a big problem with birth defects. I had to point out that people and the food they eat have long co-evolutionary history. If there had been varieties of carrots that did have high RA levels that lead to birth defects, those carrot varieties would tend not to be used over time.

There's one other issue that is integral to the fight over GR, especially for the Philippines, and that is the fact that Vit A deficiency (aka VAD) is no longer a big public health issue. According to WHO, if more than 20 percent of the vulnerable population (children under 5 years old and pregnant women), it is considered a severe public health concern. The Micronutrient Forum considers that 15% VAD among the vulnerable population is the cut off for public health significance. If you look at the data for the Philippines for children under 5, the level of VAD in 1993, 1998 and 2003 were 35%, 38% and 40.1%, respectively (see: For pregnant women, the data for 1993, 1998, and 2003 were 16.4%, 22.2%, and 17.5%, respectively; for lactating women, the figures are 16.4%, 16.5%, and 20.1%, respectively. So, at the height of when GR was developed, there was a severe problem with VAD in Philippine children.

The data on VAD levels in 2008 show a remarkable decline. For children 5 or younger, only 15.2 have VAD, while the figures for pregnant and lactating women are 9.5% and 6.4%, respectively. In other words, dramatic declines in VAD over a 5 year period, to the point where it's just above the threshold of what would be considered of public health significance.

Why the reason for the big decline? Beginning in 2000 or so, there was a campaign of Vit A supplementation as well as food fortification. Indeed, when I was in the Philippines in 2011, there were advertisements on the side of buses for Jollibee's (local chain selling chicken, sort of like KFC) new item - chicken tenders with bits of carrot and Vit A added. So, those programs obviously had a big impact. Yet, IRRI still wants to develop GR2 Indica rices for use in the Philippines and other South Asian countries. I haven't been able to get similar figures on VAD levels in India or Bangladesh over time, and could use some help with that.

I won't go into the potential environmental issues, although a number have been identified by others.

Bottom line, even if there had been no push back from NGOs, GR would still not be on the market due to the technical issues, e.g. getting the engineered traits crossed into Indica rices that people will actually eat. And we still don't have the most basic of data on the GR2 Indica varieties, such as β-carotene levels at harvest and after one to two months or more of storage; presence of other metabolites, e.g. RA and other retinoids. Also, since there was a push throughout Asia in the 2000s to push Vit A supplementation and food fortification, it would be interesting to see if the VAD levels have declined in other countries, as they have in the Philippines. After seeing the Philippine data on VAD levels over time, I simply asked how many millions have been spent on GR with nothing to show for it so far, while other methods, e.g. supplementation and food fortification, have been shown to work, but got no real response from the other side. The ultimate answer is to realize that VAD is a symptom of poverty - someone who is so poor they can only afford rice and virtually nothing else - and that rather than treat the symptom, one should treat the cause - poverty. Also, food diversification via sustainable agriculture and land reform are the longer-term answers since many local crops foods have high β-carotene levels.

Hope this information is useful.



Michael Hansen, Ph.D.
Senior Scientist
Consumers Union
101 Truman Ave.
Yonkers, NY 10703