There was a recent recall of so-called ‘radioactive shrimp’ that were potentially contaminated with cesium-137 (Cs-137). But contamination isn’t an all-or-nothing affair, so you might wonder exactly how hot the shrimp were. As it turns out, the FDA’s report makes clear that the contamination was far below the legal threshold for Cs-137. In addition, not all of the recalled shrimp was definitely contaminated, as disappointing as all of this must be to those who had hoped to gain radioactive Super Shrimp powers.

After US customs detected elevated radiation levels in the shrimp that was imported from Indonesia, entry for it was denied, yet even for these known to be contaminated batches the measured level was below 68 Bq/kg. The FDA limit here is 1,200 Bq/kg, and the radiation level from the potassium-40 in bananas is around the same level as these ‘radioactive shrimp’, which explains why bananas can trigger radiation detectors when they pass through customs.

But this event raised many questions about how sensible these radiation checks are when even similar or higher levels of all-natural radioactive isotopes in foods pass without issues. Are we overreacting? How hot is too hot?

Healthy Radiation, Normal Radiation

Ionizing radiation from nuclear sources forms both an unavoidable and an essential part of food safety. The practice of food irradiation involves exposing food to gamma rays in order to destroy anything that is still alive in it, like bacteria and other potentially harmful microorganisms. Much like heating with pasteurization and similar practices that aim to wipe out these microorganisms, this can render food safe for consumption for much longer than would otherwise be possible.

Whereas food irradiation does not actually introduce radioactive isotopes to the foodstuffs, these isotopes can still enter prospective food in other ways. Long before these infamous Indonesian shrimp – likely prawns – found themselves post-mortem on their way to the US, these critters were either happily galivanting about in the Pacific Ocean or less happily stuck in a shrimp farm, doing all the things that pre-mortem shrimp do. This includes consuming a lot of shrimp food, starting with plankton and moving up to worms, bivalves and other crustaceans as they mature.

All of these food sources along with the water that they live in contain some level of radioactive isotopes, ranging from the uranium-238 that’s plentiful in seawater, to tritium from atmospheric sources, and manmade isotopes like cesium-137 from nuclear weapons testing. Most isotopes, including Cs-137, do not bioaccumulate: in humans Cs-137 has a biological half-life of about 70 days . This suggests that this particular batch of whiteleg shrimp ingested some kind of relatively Cs-137-rich food shortly before harvesting.

The Pacific Ocean area was a particularly prolific area when it came to nuclear weapons testing, with of the worldwide approximately 2,121 tests so far the US and France detonating a significant number in the Pacific. Tests such as the 15 megaton Castle Bravo experiment featuring the ironically named SHRIMP device, which significantly raised the amount of carbon-14 (C-14), Cs-137, and strontium-90 (Sr-90) in the region. Due to its bioaccumulating nature, Sr-90 with its 29-year half life poses a particular risk, while C-14 with its 5,700-year half life is generally deemed of no consequence, on par with the normal intake of potassium-40 as both isotopes behave in a very similar way in the body.

Although the amount of Cs-137 from these tests has reduced significantly due to natural radioactive decay, this provides one potential path through which these and many other isotopes from both manmade and natural sources can find themselves inside small crustaceans prior to their untimely demise at the hand of bipedal primates with an appetite for seafood.

The one question that remains here is how we can know that a certain amount of an isotope per kg of foodstuff is too much for human consumption. How dangerous is the radioactive potassium-40 in bananas really?

Setting Limits

We earlier listed the FDA’s 1,200 Bq/kg as the limit for the Cs-137. A radioactive source rates one Becquerel if it undergoes one disintegration event per second, and dividing this by the weight gives a rough measure of radiation density. But all decay biproducts aren’t created equally. If we look at the FDA guidance documents pertaining to radionuclides in imported food, we can see that this listed limit pertains to the so-called Derived Intervention Levels (DILs), superseding the older Levels of Concern (LOCs). The same document lists the DILs for other isotopes, including:

• Sr-90 at 160 Bq/kg.
• Iodine-131 at 170 Bq/kg.
• Cs-134 + Cs-137 at 1,200 Bq/kg.
• Pu-238 + Pu-239 + Am-241 at 2 Bq/kg.

What these isotopes have in common is that they are generally only produced by artificial sources, while omitting a very common natural isotope like potassium-40 (K-40) which only forms the third-largest source of natural background radiation after thorium-232 and uranium-238. Since K-40 is readily present in soil and anywhere else that other potassium isotopes are present, it’s practically unavoidable to consume significant amounts of K-40 each day, regardless of whether you’re a crustacean, plant or mammal.

K-40 is both a beta and gamma emitter, with approximately 140 grams of it present at any given time in a 70 kg adult human body, where it is responsible for an approximate constant 4,000 Bq of radiation.

Despite the long half-life of 1.248 billion years, K-40’s prevalence makes up for this sluggish nuclear decay rate, with around 4,000 of such disintegration events happening inside an adult human body each second, as a K-40 nucleus decays into either argon-40 or calcium-40 via gamma or beta decay respectively.

We can contrast this with Cs-137’s 30 year half-life and somewhat similar decay into barium. Nearly 95% of Cs-137 nuclei decay into the metastable barium-137m via beta decay, before decaying into the stable barium-137 via gamma decay. The remaining 4% decay immediately via beta decay into this stable nucleus.

The much shorter half-life and primary gamma decay route make Cs-137 significantly more radiologically active than K-40. Yet while more gamma radiation may sound worse, one has to remember that the biological impact for radiation exposure once ingested is flipped around. For example, while the very powerful alpha radiation is luckily stopped by the top layers of our skin and dissipates its energy mostly in dead skin cells, you don’t want the same to happen to living cells like the inside of your lungs or various other soft issues, with alpha radiation absolutely cooking the nearest layers of cells.

This is where gamma decay ironically helps to distribute the radiation exposure from Cs-137 somewhat, while also complicating the comparison with K-40, as that isotope decays mostly via beta decay and thus can potentially do more damage per event to local tissue as beta radiation does not travel as far through the body.

Overabundance Of Caution

The American Nuclear Society (ANS) article on the “contaminated shrimp” event probably puts this event in best context. Normally shrimp from the Pacific region contains some level of Cs-137, but these recent batches caught the attention at the importing ports due to a 100x higher level of Cs-137 than normally seen. That sounds like a problem, but it only places the shrimp roughly in line with bananas.

A 2023 study performed in Poland found that of animal products produced in that country, cattle muscle tissue showed Cs-137 levels up to 23.5 Bq/kg (wet weight), sheep nearly 50 Bq/kg, and in wild game animals some muscle tissue scored well over 4,000 Bq/kg. All of which place these commonly consumed animal tissues well above the typical value for Indonesian shrimp, and either in the ballpark or significantly above that of the ‘contaminated’ shrimp.

Threshold Models

Government regulations pertaining to radiation exposure are most often based on the linear no-threshold (LNT) model, which extrapolates down from very high radiation doses where we can measure the damage more easily. But it does so linearly, making the assumption that ten multiple small doses, even if they are spread out over time, are equivalent to one exposure that is ten times as strong.

Recent studies have suggested that below 100 mSv there are no observable effects, which suggests that a model that incorporates a threshold might make more sense for radiological contamination of food.

The National Academy of Science report on low levels of radiation from 2005, on which most of the US regulations are based, at the time rejected the threshold model due to insufficient evidence. They also cite studies where very small doses are claimed to have negative effects on children still in the womb, suggesting that the lower threshold may not be uniform across different populations.

Even if a lower threshold does exist, and there is an increasing push by scientists for moving past the LNT, establishing the exact value for this threshold is difficult. Below a certain dosage, there just isn’t significant epidemiological data. You cannot prove a negative: “below this level there will be no increased risk of cancer”. One can only say that no excess risk was detected in this or that particular study.

Add in sensitivity about manmade radioisotopes in drinking water, food, and anything else that is sold or presented to the public, and most governments take the LNT approach even if it is likely to be very conservative. And that, in short, is why we got a ‘radioactive’ shrimp recall, when eating that banana might arguably be more hazardous for you.