by Peter Waldman
[First in a Series]
For years, scientists have struggled to explain rising rates of some
cancers and childhood brain disorders. Something about modern living
has driven a steady rise of certain maladies, from breast and prostate
cancer to autism and learning disabilities.
One suspect now is drawing intense scrutiny: the prevalence in the
environment of certain industrial chemicals at extremely low levels. A
growing body of animal research suggests to some scientists that even
minute traces of some chemicals, always assumed to be biologically
insignificant, can affect such processes as gene activation and the
brain development of newborns.
An especially striking finding: It appears that some substances may
have effects at the very lowest exposures that are absent at higher
Some scientists, many of them in industry, dismiss such concerns. But
the new science of low-dose exposure is challenging centuries of
accepted wisdom about toxic substances and rattling the foundation of
Modern pollution restrictions aim to limit exposures to levels past
studies have found safe. For example, it's known mercury can cause
learning problems in children if it's above 58 parts per billion in the
bloodstream. Dividing 58 by 10 to provide a margin of safety, U.S.
regulators advise that children and young women not accumulate more
than 5.8 parts per billion of mercury, by limiting consumption of
certain fish such as tuna.
But what if it turned out some common substances have essentially no
safe exposure levels at all? That was ultimately what the U.S.
Environmental Protection Agency concluded about lead after studying its
effects on children for decades. Indications some other chemicals may
have no safe limits have led regulators in Europe and Japan to bar the
use of certain compounds in toys and in objects used to serve food. In
the U.S., federal scientists are devising new tests that could be used
to screen thousands of common chemicals to make sure they're safe at
extremely low exposures.
Using advanced lab techniques, scientists have found that with some
chemicals, traces as minute as mere parts per trillion have biological
effects. That's one-millionth of the smallest traces even measurable
three decades ago, when many of today's environmental laws were
written. With some of these chemicals, such trace levels exist in the
blood and urine of the general population.
Some chemical traces appear to have greater effects in combination than
singly, another challenge to traditional toxicology, which tests things
The human body is complex, and effects seen in tests on small
laboratory animals and in human cells don't necessarily mean health
risks to people. "The question is what do we do about these low levels
once we know they're there," says Steve Hentges of the American
Plastics Council, a trade association.
For their part, companies and industry groups have attacked low-dose
research as alarmist and are challenging the findings with scientific
studies of their own. Some industry studies have contradicted the low-
dose findings of university and government labs. One reason, says
Rochelle Tyl, a toxicologist who does rodent studies on contract for
industry groups, is that academics seek "to find out if a chemical has
an intrinsic capacity to do harm," while industry scientists try to
measure actual dangers to people.
The result is that low-dose research has sparked a number of heated
scientific and regulatory controversies:
-- Tiny doses of bisphenol A, which is used in polycarbonate plastic
baby bottles and in resins that line food cans, have been found to
alter brain structure, neurochemistry, behavior, reproduction and
immune response in animals. Makers and users of the chemical maintain,
citing a Harvard review of 19 studies, that the chemical is harmless to
humans at such levels.
-- Minute levels of phthalates, which are used in toys, building
materials, drug capsules, cosmetics and perfumes, have been
statistically linked to sperm damage in men and genital changes, asthma
and allergies in children. The U.S. Centers for Disease Control and
Prevention has detected comparable levels in Americans' urine.
Manufacturers say there is no reliable evidence that phthalates cause
any health problems.
-- A chemical used in munitions, called perchlorate, is known to
inhibit production of thyroid hormone, which children need for brain
development. The chemical has been detected in drinking-water supplies
in 35 states, as well as in fruits, vegetables and breast milk. The EPA
has spent years mulling what is a safe level in drinking water. The
Defense Department and weapons makers maintain it is harmless at much
higher doses than those that Americans ingest.
-- The weed killer atrazine has been linked to sexual malformations in
frogs that were exposed to water containing just 1/30th as much
atrazine as the EPA regards as safe in human drinking water. The
herbicide's main manufacturer, Syngenta AG, says other studies prove
atrazine is safe. The EPA favors more study.
With so much still unknown, regulators are proceeding on different
tracks in different countries. Japan's government designates about 70
chemicals as potential "endocrine disruptors" -- substances that may,
at tiny doses, interfere with hormonal signals that regulate human
organ development, metabolism and other functions. Japan has just
completed a $135 million research push on endocrine disruptors,
including setting up a national research center. The Japanese
government also has banned certain phthalates in food handlers' gloves
and containers, after detecting them in food. One manufacturer, Fujitsu
Ltd., has pledged to phase out its use of most suspected endocrine
disruptors over coming years.
The European Union has banned some kinds of phthalates in cosmetics and
toys, and it is considering a ban on nearly all phthalates in household
goods and medical devices. The EU also is planning to require new
safety tests for thousands of industrial chemicals, many of which
already exist in people's bodies at trace levels. Industry, which would
have to bear the cost of proving countless current products safe, is
fighting the measures, calling them a massive unnecessary burden.
In the U.S., there are divisions within the government. The White House
plays down the issue, saying the low-dose hypothesis is unproved. But
many federal scientists and regulators at the EPA and Health and Human
Services Department are forging ahead with new methods for assessing
possible low-dose dangers. Legislatures in two states, California and
New York, are considering bills that would ban use of certain
phthalates in toys, child-care products and cosmetics, while a
California bill would restrict bisphenol A.
One of the early scientists to focus on possible low-dose risks was
biologist Theo Colborn of the World Wildlife Fund. Studying the decline
of certain birds, mammals and fish in the upper Midwest, Dr. Colborn
spotted some patterns: Species that struggled to survive in the
industrialized Great Lakes thrived in inland areas that were less
polluted. And some offspring in more-polluted regions had gender
abnormalities, such as feminized sex organs in males. She theorized
that trace amounts of chemicals in the environment were disrupting
Dr. Colborn and colleagues popularized low-dose concerns in a series of
conferences, articles and a best-selling 1996 book called "Our Stolen
Future." That year the EPA asked an outside advisory panel to consider
ways of screening industrial chemicals for hormonal effects, a process
In 2000, a separate EPA-organized panel, after reviewing 49 studies,
said some hormonally active chemicals affect animals at doses as low as
the "background levels" to which the general human population is
subject. The panel said the health implications weren't clear but urged
the EPA to revisit its regulatory procedures to make sure such
chemicals are tested in animals at appropriately small doses.
The EPA hesitated. It responded in 2002 that "until there is an
improved scientific understanding of the low-dose hypothesis, EPA
believes that it would be premature to require routine testing of
substances for low-dose effects."
The Bush administration's regulatory czar, John Graham -- administrator
of the Office of Information and Regulatory Affairs at the White House
Office of Management and Budget -- later publicly dismissed as unproven
the idea that the hormonal system could be disrupted by multiple low-
dose exposures to industrial chemicals. For the past two years, the
administration has proposed funding cuts for EPA research on suspected
endocrine disrupters, but Congress has kept the funding roughly level
at about $10 million a year.
Since the review panel met in 2000, scientists have published more than
100 peer-reviewed articles reporting further low-dose effects in living
animals and in human cells. These findings are generating some early
insights in the thorny process of translating laboratory data into
conclusions about human health.
One of the most provocative is that some hormonally active chemicals
seem to have more effects at extremely low exposures than at higher
ones. This challenges an axiom of toxicology stated by the Swiss
chemist Paracelsus nearly 500 years ago: The dose makes the poison.
Toxicologists traditionally derive risk by exposing rodents to
chemicals to find the lowest dose that leads to tumors, birth defects
or other readily observable effects. Regulators then divide the
highest "no-observable-effect" dose by an "uncertainty factor" --
anywhere from 10 to 1,000 -- to set a maximum human exposure they can
be confident is safe.
But now researchers have found chemicals that have hormonal effects on
lab animals and on human cells in much tinier amounts than their
standard no-observable-effect levels. And with some of these chemicals,
as the tiny doses given to animals are increased, the effects recede.
Then, at much higher levels, broad systemic impacts appear, such as
reduced body weight.
An example is bisphenol A, or BPA, the ingredient in polycarbonate baby
bottles and food-can linings. It evidently is widespread in the
environment. In the U.S., the CDC has found traces of it in 95% of
urine samples tested. In Japan, researchers have detected BPA in fetal
amniotic fluid and the umbilical cords of newborns.
Studying BPA in rats in 1988, the EPA concluded the lowest exposure
with an "observed adverse effect" was 50 milligrams a day per kilogram
of body weight. Dividing 50 by an uncertainty factor of 1,000, the
agency set a daily safe limit for humans of 0.05 milligrams of BPA per
kilogram of body weight. Since then, however, academic scientists in
several countries have done more than 90 studies that have found BPA
effects on animals and human cell cultures from exposures well below
The EPA used a relatively crude measure of the chemical's effects:
changes in rodents' body weights. The new studies looked at subtler,
hormone-related effects. Some studies found changes in rodents'
reproductive organs and brains at doses as low as 0.002 milligram per
kilogram of body weight per day. That is just one-25,000th the dose
that the EPA said was the lowest exposure having an observable adverse
Seeking to explain this pattern, scientists cite the endocrine system's
exquisite sensitivity. Animals and humans secrete infinitesimal amounts
of various hormones, such as estrogen, that trigger responses when they
occupy special receptors on the cells of various organs. BPA is among
numerous chemicals that can mimic estrogen by occupying cells' estrogen
receptors. When they do this at critical phases of development, the
chemicals can trigger unnatural biological responses, such as brain and
At higher doses, however, BPA and other endocrine disruptors -- instead
of triggering the unnatural responses -- appear to overwhelm the
receptors. That explains, scientists say, why some chemicals seem to
have more potent hormonal effects at very low doses than at higher ones.
Mr. Hentges of the American Plastics Council says studies show BPA is
harmless at the tiny levels to which humans are exposed. In 2001 the
plastics council agreed to pay Harvard's Center for Risk Analysis, part
of the Harvard School of Public Health, $600,000 to review BPA studies.
The 10 panelists found "no consistent affirmative evidence of low-dose
BPA effects" on the basis of 19 studies that were selected by April
2002 for review.
However, many more BPA studies kept coming out, and when the center
published its report last fall, three of the 10 panelists declined to
be listed as authors. "There are other papers published after the 'cut-
off' date that the panel did not review that may have altered their
conclusions," says one of the three, Paul Foster of the National
Institute of Environmental Health Sciences. A fourth, Claude Hughes of
Quintiles Transnational Corp., a pharmaceutical consulting firm, signed
but made the same point in a journal commentary criticizing the report
and calling for a new EPA risk assessment. The Harvard risk center's
executive director, George Gray, acknowledges that a "torrent of new
papers on BPA" may have made it impossible for the panel to review
everything by its deadline.
The plastics council's Mr. Hentges says his group reviews all studies
on BPA and believes none have changed the basic conclusion of the
Harvard report. "We continue to believe that the weight of evidence
indicates BPA poses no risk to human health," he says.
Environmental chemicals don't exist in isolation. People are exposed to
many different ones in trace amounts. So scientists at the University
of London checked a mixture. They tested the hormonal strength of a
blend of 11 common chemicals that can mimic estrogen.
Alone, each was very weak. But when scientists mixed low doses of all
11 in a solution with natural estrogen -- thus simulating the chemical
cocktail that's inside the human body today -- they found the hormonal
strength of natural estrogen was doubled. Such an effect inside the
body could disrupt hormonal action.
"In isolation, the contribution of individual [estrogen-like chemicals]
at the concentrations found in wildlife and human tissues will always
be small," wrote the scientists, led by Andreas Kortenkamp, who directs
research on endocrine disruptors for the EU. But because such compounds
are so widespread in the environment, the researchers concluded, the
cumulative effect on the human endocrine system is "likely to be very
To test chemicals, toxicologists traditionally dose animals with a
single substance and then dissect them. But this method can't spot the
subtle effects associated with today's multiple exposures to low-dose
chemicals, says John Bucher, of the National Institute of Environmental
Now he and his boss, Christopher Portier, are revamping the federal
government's National Toxicology Program, which sets standards for how
chemicals are tested. Over about seven years, they hope to develop a
series of lab tests that will ultimately screen some 100,000 industrial
compounds, individually and in mixtures, for biochemical "markers" such
as effects on specific genes.
The chemicals then will be ranked by mechanism of action and suspected
toxicity, and assigned priorities for further study. "It's taken us 25
years and $2 billion to study 900 chemicals," Dr. Portier says. "If
this works, we can study 15,000 in a year."
Scientists have found effects on rodents from steadily smaller
some chemicals, such as Bisphenol A, used in food-can linings and
Daily dose in milligrams per kilogram of body weight
-- 1988 (EPA's predicted safe dose for humans): 0.05 mg
-- 1997 (Linked to enlarged prostate in male mice): 0.002 mg
-- 1999 (Linked to early puberty in female mice): 0.0024 mg
-- 2003 (Linked to altered sperm in male rats): 0.0002 mg
Source: Scientific articles
In the Laboratory
Studies have linked some common chemicals with toxic effects, though not
necessarily at levels to which humans are exposed:
CHEMICAL: Bisphenol A (BPA)
WHAT IT'S IN: Polycarbonate plastic bottles and food-can linings
WHAT IT'S LINKED TO: Altered brain, behavior and sex organs in rats
CHEMICAL: Dibutyl Phthalate (DBP)
WHAT IT'S IN: Cosmetics, shampoos, pills, nail polish, plastic toys
WHAT IT'S LINKED TO: Gene and hormone changes in rodents; genital
abnormalities in human infants
CHEMICAL: Diethylhexyl Phthalate (DEHP)
WHAT IT'S IN: Polyvinyl chloride building products, food packaging,
WHAT IT'S LINKED TO: Birth defects in mice; pre-term birth in human
infants; early puberty in girls
WHAT IT'S IN: Drinking water in 35 states, fruits, vegetables, breast
WHAT IT'S LINKED TO: Brain and behavior changes in rats; thyroid
Sources: Scientific articles; U.S. Environmental Protection Agency