A new Kiwi-led study has found a third of pups born to rats fed high doses of very rancid fish oil during pregnancy did not survive beyond two days - but experts say human mums-to-be shouldn't be worried about taking fish oil supplements.
Researchers at the Auckland University-based Liggins Institute wanted to investigate the health effects of "off" fish oil after an earlier study they did found most fish oil supplements sold in New Zealand were off to some degree.
Female rats in the latest study, which is published today in the American Journal of Physiology, were given either unoxidised fish oil, a highly oxidised fish oil, or water daily throughout pregnancy and almost 30 percent of baby rats born to mothers who had highly oxidised oil died within the first two days.
Giving pregnant rats the unoxidised fish oil did not increase mortality rates in their babies, indicating that the lethal effect on newborns came from the chemicals that omega-3 fatty acids break down into during oxidation.
Omega-3 fatty acids were known to be chemically fragile or "unstable", and can easily break down when exposed to natural conditions such as light, heat and oxygen.
The earlier study, published in Scientific Reports last year, tested 36 brands of fish oil supplements capsules and used the international industry standard tests of oxidation.
Eighty-three percent were oxidised beyond international recommended levels and how "off" they were had nothing to do with best-before date, price, or the country they came from.
Four studies from North America, South Africa and Europe have also uncovered high levels of oxidation in fish oil supplements.
"Once we discovered so many supplements were oxidised, we decided to focus on the health effects of oxidised fish oil during and after pregnancy," said Liggins research fellow Dr Ben Albert.
Pregnancy was a critical period when considering the safety of medications, and the same should apply to dietary supplements, he said.
"Chemicals that may be harmless to mothers could potentially disrupt developmental processes in the womb."
The study's lead author, Professor Wayne Cutfield, said the death rate was surprising.
"We'd expected some negative health effects on the rat offspring, but we didn't expect them to die."
Albert said was unclear exactly why the newborn rats died.
"Because we didn't expect them to die, we didn't design the study to look for reasons."
At weaning, the mothers given oxidised fish oil also had greater insulin resistance, which in humans can lead to type 2 diabetes.
Albert emphasised that the results of this study could not be directly applied to humans.
"Obviously, rats are not humans. Also, it's important to note that the fish oil dose we gave to the rats was higher than doses humans take, but this dose is commonly used in fish oil studies in rats.
"Also, the level of oxidisation was at least double what we detected in most products in our earlier study of fish oil supplements. Commonly, a more potent formulation is given to determine if there is any effect.
"Then, you design follow-up experiments to test different doses at different degrees, to see exactly when the effects start to show up.
"In future studies, we hope to examine what happens in pregnant rats when you vary how oxidised the fish oil is, and to understand exactly how the oxidised fish oil harms the baby rats."
Up to one in five New Zealand women take fish oil supplements during pregnancy, according to the latest estimate.
"While some women take fish oil during pregnancy to try to improve the development of their child's brain, there's no convincing evidence that this helps," Albert said.
Further, oxidised fish oil was unlikely to carry serious health risks in humans, Cutfield said.
"But at the moment, we just don't know the health risks to the unborn baby.
"And our first study showed it's not possible to know if fish oil is oxidised when we buy it. This suggests it may be wiser for women not to take fish oil supplements in pregnancy."
Further research was needed to assess any risk to humans, he said.
"In the meantime, pregnant women might consider eating fresh fish for omega-3 oils."
Responding to the new study, Professor Murray Skeaff of Otago University doubted there would be any relevance to the health of pregnant women who take fish oil supplements, "but curiosity demands an attempt to answer the question".
"The fish oils fed to the rats were oxidised under conditions so extreme as to bear almost no resemblance to fish oils consumed by human."
Nevertheless, the high mortality rate amongst pups born to rat mothers fed the oxidised oils proved that one or more compounds were produced in the fish oil during oxidation that were toxic to the rat, he said.
"What are these compounds, are they also harmful to humans, and are they present in commercially manufactured fish oil supplements?
"Unfortunately, there is nothing in the research by Cutfield that helps to answer these questions but surely they will try and find an answer."
Other researchers approached for comment by the New Zealand Science Media Centre questioned the study and the preceding paper, which had conflicted with data from the Omega-3 Centre in Australia.
"I was very concerned about the original paper, which they continually cite, and claim shows that most over the counter omega-3 supplements in New Zealand are oxidised," said Professor Lynette Ferguson, a nutritionist based at the Auckland Cancer Society Research Centre.
"Having done analytical work with these compounds myself, I am only too conscious that there are some technical difficulties in getting accurate results."
Dr Peter Nichols, a science advisor at the Omega-3 Centre and senior principal research scientist at Australia's CSIRO, said that following the 2015 study, the Australian Therapeutic Goods Administration performed follow-up analyses and all tested oils were not oxidised and omega-3 content met label claims.
"For this new study, the justification appears to be driven by the Scientific Reports paper, which remains in the strongest doubt and dispute," Nichols said.
"The new paper uses heavily oxidised oil that the New Zealand authors prepared. As Australian and New Zealand fish oils are not heavily oxidised, the study is seen as not relevant.
"The dose used is equated to 40 mL per day for a human consumer. This dose is seen as exceptionally excessive. Few consumers would be taking more than 5-10 g per day.
"The unoxidised oil actually and interestingly shows improvement in the new paper, versus the control treatment, although little is stated by the authors on this aspect."