3 2 2' 3'
\ / \ /
C == C C -- C
/ \ // \\
4 -- C C -- C C -- 4'
\\ // \ /
C -- C C == C
/ \ / \
5 6 6' 5'
It's an elegant molecule, isn't it? Two phenyl groups — that is to say, benzene rings — joined together by a single bond. Benzene is a fascinating chemical — it was known for a long time before its structure was elucidated. Many different shapes were tried, before German chemist Friedrich August Kekule was inspired by a dream of a snake eating its own tail, and figured out its hexagonal shape. Those alternating single and double bonds are actually just a convenient notation — actually, benzene (and its derivatives) have on average one and a half bonds between each carbon atom. Six of the electrons are free to travel all around the ring. This is known as aromaticity.
But I digress. In the group of molecules this writeup concerns, those two phenyl groups are surrounded by a total of ten atoms, noted as positions 2 to 6 and 2' to 6'. Anywhere between one and ten of those positions are occupied by bromine atoms; the rest are hydrogens. The exact constitution is relatively unimportant; it's far easier to create a mix of these chemicals, knowing the average amount of bromine in the resulting mixture, than to precisely separate the different congeners. So the polybrominated biphenyls are a group of 209 separate chemicals altogether, which are only produced and found mixed together.
They were first used as fire retardants — notably stable chemicals and particularly known for not being combustible (much like their PCB cousins), they were mixed into plastics to reduce their flammability. It was eventually discovered that PBBs accumulate in the ecosystem — due precisely to that stability — concentrating further the further you travel up the food chain. Their exact effects on human health are still unknown, but what is known is mostly due to the massive exposure of much of Michigan's population in the early 1970s.
How the accident happened
Prelude
In September of 1973, a farmer named Rick Halbert in southwest Michigan began noticing problems in his herd of 400 dairy cattle. Milk production decreased; the cows seemed to be listless and have decreased appetite. Soon they began developing hematomas and abscesses, hair loss, and abnormal growth of their hooves. Halbert and his veterinarian examined the cattle and quickly ruled out common livestock diseases; no cause could be found for their illness. Halbert suspected something was wrong with their feed.
Halbert was feeding his cows in part on a high-protein feed supplied by Farm Bureau Services, the state's largest feed distributor. The company denied any problem with the feed it supplied, but Halbert decided to perform an experiment. He fed a dozen calves nothing but high-protein feed pellets. Within weeks, five of them died.
Halbert contacted the State Department of Agriculture; they agreed to perform a test, and fed mice the cattle feed. Two trials were performed, and all the mice promptly died. Farm Bureau Services continued to deny that there was a problem with their feed. Nevertheless, in early 1974 they began secretly conducting tests on the feed themselves, still denying to Halbert that any problem existed and failing to report their concerns to state government. Halbert was convinced that a problem existed, and eventually enlisted the help of scientists to analyze the feed and find out what was wrong with it.
In April 1974, mass spectrometry was performed on the feed, and a scientist at the USDA identified the contaminant in the feed. It was a chemical he had worked with — polybrominated biphenyl. The fire retardant should never have come into contact with animal feed, but apparently it had.
What went wrong
Michigan Chemical Corporation manufactured a range of chemicals for all sorts of different purposes. PBBs were sold to manufacturers who used them as fire retardants in their plastics. Michigan Chemical also produced magnesium oxide (MgO), a cattle feed additive that increased milk production. But the two chemicals should never have been confused.
The company sold a product called Firemaster BP-6, in the form of yellow pellets; magnesium oxide was white in color and granular in consistency. Not only that, but the two products were made and stored in separate buildings. Color coded bags were used to distinguish between different chemicals — bright red for the Firemaster, and blue for the MgO. It shouldn't have been possible for a mix-up to occur.
An inspection of the feed plant on 30 April helped to piece together the chain of events that led to the PBB contamination. A half-empty bag of another Michigan Chemical product, Firemaster FF-1, was found. This was an experimental product, in which the PBB was not in amber-colored pellets but ground into tiny grains; the granular PBB looked identical to MgO. The bag was neither red nor blue — it was a plain brown bag, and it lacked a proper label. The ingredients were not listed; the only labeling was the trade name, stenciled on top of the bag. Once opened, the label was torn and difficult to read.
In the spring of 1973, Michigan Chemical ran out of its usual color-coded bags, and used plain brown bags instead. None of them were labeled properly; many of the chemicals (including MgO) that Michigan Chemical manufactured were sold under multiple trade names, and only the trade names were printed on the brown bags. Further, when those poorly-labeled bags were accidentally shipped to Farm Bureau Services in May 1973, they were left with employees who had little job training and no technical knowledge. At least one employee couldn't read; others noticed that the bags were labeled with the wrong name, but they were told by their supervisor that it was merely a new name for the same MgO. Ultimately, it is estimated that between 500 and 1,000 pounds of PBB were mixed in with the cattle feed.
Aftermath
Initial response
At the time of the accident, little was known about the health impacts of PBBs. Toxicology studies in the early 1970s didn't seem to indicate any acute health impact from exposure to PBB — it wasn't a potent irritant, seemed nontoxic when ingested, and only mildly toxic when inhaled. Nevertheless, Michigan Chemical briefed its employees in 1971, explaining that prolonged exposure was potentially dangerous and that PBBs accumulated in the liver. Other evidence existed that PBBs might be dangerous; Dow Chemical had previously declined to manufacture PBBs because of concerns over toxicity and environmental impact, and because more extensive knowledge existed of the problems caused by the closely-related PCBs. So when the contamination was discovered, the health impacts were as yet unknown — but substantial suspicions existed.
The contamination of Farm Bureau Services' high protein feed occurred in May 1973; the first signs of its impact didn't appear until Rick Halbert noticed in September. Even after that, it took quite some time before the cause of his cattle's illness was established; all in all, nearly a year passed before investigators figured out that the cattle feed was contaminated.
Once the feed contaminant was identified, it became obvious that a far more massive problem existed than just the death of Rick Halbert's cattle. Federal and state agencies began to work to establish the extent of the disaster and to investigate the potential impacts upon the health of Michigan's residents. Immediately, scientists and government agencies began working to figure out safe exposure levels and how many animals and people might be impacted.
Investigation
The first step in evaluating the magnitude of the contamination happened when the FDA set an "action level" of 1.0 ppm in milk and milk products, two weeks after the contaminant was identified; establishing an action level doesn't require public debate or even scientific certainty. It's a tool to permit quick action on the issue, but doesn't purport to set a level below which exposure to a chemical is safe. The USDA chose 1.0 ppm as the action level for meat a month later; in November 1974, this action level was reduced to 0.3 ppm, and eventually the state legislature reduced it further to 20 ppb.
Testing commenced quickly, and farms were quarantined if their meat or milk had contamination levels above 1.0 ppm. If farm products were contaminated below that level, it was assumed that the cattle that produced them were healthy. The problem was that more objective measures of sickness were impossible — while cattle that were heavily exposed to PBBs developed visible illness, it was recognized that far lower levels of exposure could potentially cause health problems in humans down the line. But no one knew how low those levels of exposure had to be in order to guarantee public safety.
The Michigan Department of Health began to perform screenings on farm workers in July 1974; next a short-term epidemiological study began comparing 300 people, divided into people exposed and unexposed to PBBs. The FDA conducted surveys for health problems in families on quarantined farms. However, while it appeared that people were indeed suffering health problems as a result of PBB exposure, neither agency was forthcoming with the results of investigation.
Action
Beginning in July 1974, state government passed a law establishing a disposal site for contaminated animals and permitting state agencies to condemn and destroy affected animals. But fearing lawsuits, the state simply established disposal facilities and provided for the destruction of livestock, without forcing farmers to do so. Most farmers did, though, as they had no market for quarantined livestock; thus many farmers faced the financially disastrous condition of being left without their entire livestock herds. The animals were buried in Kalkaska County, in north-central Michigan.
By early 1975, pressure began to rise to remove all food products with detectable PBB contamination from the market, but FDA research didn't seem to indicate that low levels of exposure were a health threat. Critics pointed out, though, that these FDA studies were not conducted properly, as most of the people in the control groups also had detectable blood levels of PBBs. At the same time, many farmers were claiming that animals with exposure levels even below the revised 0.3 ppm action level were ill. The public was growing increasingly upset at the state's handling of the problem. Hearings were conducted in 1976 in order to collect accounts of illness associated with PBB contamination. State bureaucrats blamed the farmers for "poor management" — essentially claiming that apparent symptoms of PBB exposure were actually caused by the farmers. Two weeks after the hearings ended, a group of farmers organized a march in Lansing, which culminated in the dumping of PBB-contaminated cattle carcasses on the steps of the state capitol.
In the end
By the end of 1975, 28,900 cattle had been destroyed; buried with them were 865 tons of animal feed and 27 tons of various dairy products. As investigations into the disaster continued, it became clear that beef and milk were not the only products affected. Feed for other animals was contaminated when it was mixed on the same machinery; carcasses of cattle not fit for sale were rendered to yield protein supplements in other animal feed. In addition to the cattle and milk, 5,920 pigs, 1.5 million chickens, and nearly 5 million eggs were destroyed.
Testing in the late 1970s showed that about 97% of Michigan residents had measurable levels of PBB contamination; further, testing over time revealed that these levels did not decline: the chemical accumulated in the body and remained in adipose tissue. Testing further indicated that most breast milk was contaminated as well, which meant that even those of us who weren't born at the time of the disaster likely have been exposed. Many farms remained contaminated — buildings and soil still retain testable levels of PBB in some areas, and the disposal sites are still heavily contaminated.
Farmers whose livestock was contaminated above 0.3 ppm were eventually compensated by insurance companies, but no compensation existed for farmers whose animals had lower levels of exposure. In February 1977, the first court case by farmers affected by low-level PBB toxicity was filed, and it became both the longest and the most expensive court battle in Michigan history. Sixteen months after it began, the court found in favor of Michigan Chemical and Farm Bureau Services; some other farmers received small out-of-court settlements. A lawsuit filed in October 1977 on behalf of people with health problems caused by PBB never made it to trial.
The long term impacts of the exposure of virtually all of Michigan's residents to PBBs remain unknown. Very little was known for certain until well after the end of the crisis; PBB was eventually linked to joint pain, neurological problems, reduced liver function, and altered immune systems. More recent research shows that daughters born to women with high levels of PBB contamination begin menstruating earlier. While fears of an eventual dramatic increase in cancer rates never materialized, those with the highest levels of PBB were shown in 1998 to have significantly higher rates of lymphoma and cancers of the digestive tract. It remains to be seen if more long term epidemiological research will indicate other problems.
References
Michael R. Reich. "Environmental Politics and Science: The Case of PBB Contamination in Michigan." American Journal of Public Health. 1983.
David Dempsey. "Thirty Years Later: The Lessons of PBBs." Lansing City Pulse. 2003. (http://www.lansingcitypulse.com/031001/health/)
"PBBs in Fire Retardant Associated with Early Menstruation in Michigan Girls Whose Mothers were Exposed in 1973." National Institutes of Health. 2000. (http://www.nih.gov/news/pr/dec2000/niehs-08.htm)
"ToxFAQs for Polybrominated Biphenyls (PBBs)." Centers for Disease Control. 2004. (http://www.atsdr.cdc.gov/tfacts68.html)