Atlas Research Digest: The damage arrived where the plastic did not
In sheep, dietary plastic damaged the testis without a particle ever reaching it, and a missing bacterial metabolite carried the injury instead. Also: a first look inside human ovarian follicular fluid, and a warning that some detections may not be real.
Nearly all the effort in this field over the last few years has gone into one question: where do the particles end up. Blood, placenta, lung, semen, testis. Underneath that search sits a model so obvious nobody bothers to state it, which is that a particle travels somewhere and does its harm where it lands. A group working in sheep has now produced real testicular injury without ever finding a particle in the testis, and the way they established the route is the most useful thing here.
Male Hu lambs were fed polystyrene microplastics at 75 or 150 mg per kg of diet for 120 days, then followed for a further 180 days with no exposure at all. The injury was real and it outlasted the dosing: seminiferous architecture disrupted, germ cell apoptosis raised, and single cell sequencing showing spermatogonia failing to differentiate with the Nrf2 antioxidant response suppressed underneath. Then the authors went looking for the particles in testicular tissue, and did not find them. What they found instead was in the rumen, where exposure had cut back the butyrate producing bacteria, leaving both ruminal butyrate and circulating beta hydroxybutyrate depressed. To test whether that was actually the route, they supplemented exposed lambs with sodium butyrate and recorded the order in which things came back. Ruminal butyrate recovered first, then the circulating metabolite, then systemic inflammation and oxidative status. The testis was last, with Nrf2 signalling restored. A rescue that recovers in the sequence a mechanism predicts is a much stronger claim than a correlation. The cautions are real. A rumen is not a human gut, and the doses were chosen by the experimenters rather than measured in anything living freely. But the shape of the result travels further than the species does. The organ that was damaged was not the organ that was exposed.
An exposure that ends, and a susceptibility that does not
The sheep work separates exposure from injury in space. A second study separates them in time, across a generation. Gestating mice were given 1.25 mg of 50 nanometre polystyrene beads by mouth each day from the fifteenth day of gestation, continued through lactation, and stopped at weaning. The offspring were never dosed directly. On the measures anyone would check first they looked fine: no effect on the survival of the mothers, a normal sex ratio in the litters, and oral tolerance to ovalbumin intact. Then at postnatal day 63, with the exposure long over, the researchers induced colitis, and the disease was significantly worse in both sexes. In males, weathered particles did more damage than pristine ones, which is the more realistic of the two comparisons, since nothing in the environment stays pristine. What this describes is not injury. It is a susceptibility installed early that stayed invisible until something later came along to test it.
A review of microplastics and inflammatory bowel disease sets out what would have to be true for any of that to matter in people: particles downregulating tight junction proteins such as ZO-1 and claudin-1, altering endocytosis, and shifting the microbiome. Its own verdict on the state of that case is the part worth carrying away. The evidence linking these particles to intestinal inflammation is still largely associative, and no causal relationship in humans has been confirmed.
A new compartment, and an argument about whether we can see it
Particles have now been reported in the fluid that surrounds a human egg. Twenty seven women undergoing IVF provided follicular fluid, processed under contamination control and read by confocal micro Raman spectroscopy against validated polymer libraries. Twenty two of the 27 samples contained microplastics, one to three particles in each. Polypropylene was the most frequent, followed by polyethylene, polyamide, polystyrene and polyethylene terephthalate (PET). The procedural blanks run alongside the samples came back clean, and in this particular literature that is the detail giving the rest of it weight. The authors call it a pilot and make no claim about egg quality or fertility, because the work cannot support one.
Ovaries, uteruses and testes recovered during routine spay and neuter procedures gave a second group access to reproductive tissue in 100 cats, some household pets and some feral. Particles turned up in 71 percent of the animals overall, in every one of the household females, and in only 36 percent of the feral males. The detail that keeps the study honest is what those particles were. Of the ones confirmed by infrared spectroscopy, only a quarter were plastic at all. Just over half were non plastic anthropogenic fibres such as rayon, and a fifth were cellulose. Ninety seven percent of everything recovered was fibre rather than fragment. PET was the commonest plastic, which points back at packaging, including the packaging that cat food arrives in.
Set against both of those is a review arguing that much of the detection literature underneath them cannot carry the weight now being put on it. Quantitative confirmation that intact particles cross into tissue remains limited. Fluorescence based results account for a large share of what has been published, and they are vulnerable in ways that are easy to miss. Dye can leach out of the particle and stain the tissue around it. Tissue autofluoresces on its own. A particle resting on a surface can read as a particle inside it. The biomarkers usually reported alongside a detection, oxidative stress and inflammation, are too unspecific to pin on plastic in particular. What the authors propose is a tiered approach beginning with polymer specific confirmation of the particle itself. Read next to the follicular fluid study, this is less a contradiction than an explanation of why that study went to the trouble of Raman spectroscopy and clean blanks.
How much of it is in the air indoors
Indoor air comes out far dirtier than outdoor air in a meta analysis of 27 studies across 16 countries, 21 of them carrying usable numbers. In moderately polluted indoor settings the mean was 1,380 particles per cubic metre against 102 outdoors. Before anyone rearranges a house around that, look at the heterogeneity the authors themselves report. The I squared runs above 99 percent, which means the pooled studies disagree with one another almost completely. That is why they present their own figure as a descriptive summary rather than an estimate anyone should lean on. It also sits awkwardly next to a synchronised indoor and outdoor sampling campaign already in the Atlas, which measured an indoor to outdoor ratio of about 0.85 in one coastal city, meaning marginally cleaner inside. That campaign carries the limitation the detection review was warning about, since its particles were screened by Nile Red fluorescence and only 14.6 percent were confirmed by Raman. Two results this far apart are not a finding about buildings. They are a finding about how unsettled the measurement still is.
One exposure route here is unusual, because it is prescribed. Clear orthodontic aligners are worn around 22 hours a day for one to two years, increasingly by children. Laboratory work shows both thermoformed and 3D printed aligners shedding micro and nanoplastics under oral, gastric and mechanical conditions, with the directly printed devices releasing more and larger particles. What has never been measured is the particle burden in an actual wearer. The distance between a device known to shed in a beaker and a dose nobody has quantified in a person is this whole field in miniature, with the difference that here the exposure is handed to someone by a clinician.
What holds up across a lot of experiments
The most useful piece of synthesis is also the least dramatic. A preregistered systematic review gathered 185 preclinical studies across C. elegans, zebrafish, Drosophila, cultured cells and rodents, and scored each for methodological quality rather than simply counting up results. Two findings were consistent across every one of those models: impaired survival and disrupted development. Behavioural effects on anxiety, memory, learning, sociability and locomotion recurred throughout, and a number of the studies converged on disruption of the serotonergic system. Agreement across organisms that share almost nothing is hard to dismiss as an artefact of any single model. It is also still evidence about worms, flies, fish and rodents.
Two further reviews fill in the picture at the level of the cell, and puncture one commercial assumption. A synthesis of in vitro work finds apoptosis recurring as the shared endpoint across gut, liver, lung, brain and reproductive cell models, with the particular pathway differing by organ. And a review of 27 fish studies examined the plastics sold as biodegradable, among them polylactic acid and polyhydroxyalkanoate. They do not fully break down in water. They accumulate in the intestine, liver and gills, and produce multi organ toxicity at times comparable to or worse than conventional plastics. Whatever problem a biodegradable label solves, it is not this one.
The honest summary has not moved far. Mechanisms are being described with increasing precision, at doses the experimenters chose, and nobody has yet measured what a person actually takes in and then shown what that quantity does. What has moved is the question worth asking next. If a missing bacterial metabolite can carry an injury from the gut to an organ the particles never reach, then finding no plastic inside a damaged organ is not a test of whether plastic damaged it. The work in sheep does not show that this happens in people. It does show that a negative result in the obvious place is not the reassurance it appears to be.
Sign in to start a discussion.