2009 | 2008 | 2007 | 2006 | 2005 | 2004 | 2003

Toy Tantrums - The Debate Over the Safety of Phthalates
January 30, 2006
Rebecca Goldin Ph.D
Alarming headlines trump serious discussion of safety issues

Recent political activity has focused the media's attention on phthalates (pronounced thal-ates). Or is that the other way around? A leading scientist on the issue, Dr. Shanna Swan, recently published an incendiary and scientifically irresponsible opinion piece (disguised as science) in the San Francisco Chronicle. Swan claims that her study suggests that "In-utero exposures to phthalates can lead to birth defects and genital malformations... in baby boys." In fact, the study Swan conducted is far from even considering this hypothesis.

Perhaps as a consequence of her self-promotional efforts, media coverage of Swan’s work has consistently given the public the impression that children are at risk when the body of research is far from conclusive. We laud the isolated cases of responsible coverage, such as Los Angeles Times and their description of the disagreement among scientists.

The stakes in this debate are high. Legislators in California debated and then narrowly rejected a proposal to ban phthalates in toys. Proponents say they will reintroduce the bill next year. Legislators in Maryland will soon debate a similar ban. Meanwhile, industry groups are up in arms, claiming that none of the studies claiming a risk from phthalates merit attention.

STATS has been following this issue for the past two years, and with the European Union’s ban on certain phthalates taking effect last week, we decided to see where the research currently stands and to pose the questions journalists need to ask and get answers to for the public to really know what their legislators plan to do on their behalf.

Phthalates are a collection of colorless oil-like chemicals that are used to make plastics more malleable in products ranging from toys to medical tubing. They are also used to make nail polish flexible and to prolong the scent of perfume. In a laboratory setting, very high doses of some, but not all, phthalates, given to rats produced developmental problems. Specifically, when pregnant rats were exposed to these high levels, their male offspring had genital deformities and fertility problems. They also had an abnormally short “anogenital distance”, which is the distance from anus to the base of the rat genitalia.

The question is whether exposure to phthalates in our environment (through plastics, food, cosmetics, or dust) could have the same effect on humans. Naturally, it would be unethical to subject humans to high doses of phthalates to see what might happen in a controlled setting. But scientists can conduct an epidemiological study, in which phthalate levels and biomarkers for fertility are measured to see if there might be any relationships. Specifically, this kind of study examines whether, in a randomly selected group of people, the people who have a higher exposure to phthalates also have a higher rate of fertility or genital developmental problems.

Last year, a study conducted on pregnant women and their male offspring to see if there were measurable affects from phthalates on male genital development was published in Environmental Health Perspectives. The study found that certain phthalates correspond to a shorter anogenital index (the anogenital distance adjusted for weight differences). The study did not examine the issue of genital defects, as there were no children with visible defects in the study. The study did not consider fertility issues, since the children were prepubescent.

The importance of what the study attempted to establish, and what it did not, was quickly lost in media coverage, which headlined scary titles such as “Study Links Plastics to Small Genitals” (Fox News) or “Chemicals in frequent use linked to defects in boys” (Sydney Morning Herald), “Chemicals in plastics harming unborn boys/ Scientists say chemicals have gender bending effect” (The Guardian), or “Chemical may harm sex organs of prenatal boys/Phthalate may interfere with male development, study finds” (MSNBC). These headlines misrepresent the issue, since there was no measure of “harm”, nor were there any defects.

The media was in part led astray by Dr. Shanna Swan, the lead author of the study. She has publicly taken the position that the correlation between phthalate exposure in utero and anogenital index proves that phthalates are causing reproductive harm, even though the study found neither actual genital defects nor fertility problems. Environmental groups have cited Swan in pushing for legislation to ban such phthalates in baby toys. Some claim that the scientific evidence shows that phthalates are harmful, while others claim that it is “Better to be Safe than Sorry”, an argument that loses water if you apply it to using our cars or drinking a cup of coffee.

In contrast, industry groups have taken the position that the study is without merit, citing flaws in measurement and analysis. Yet every journalist who follows the dollar wonders whether these criticisms are motivated by the financial consequences if consumers get too concerned about phthalates.

A Commission has been formed at the National Institutes of Health. Reputations are on the line. No one seems neutral. So what should legislators and parents and journalists make of this dispute?

The science
There are several important questions that need to asked when covering this story.

1. Is there a correlation between the anogential index and phthalate level, as the Swan paper claims? How rigorous is the science of the study itself?

2. If there is such a correlation between phthalate exposure and anogenital index, does this have impact on important issues such as genital development and fertility?

3. How and why are we exposed to phthalates?

4. What should be done?

The Correlation Between Anogenital Index (AGI) and Phthalate Levels
Forget fertility issues. Forget deformed penises. Scientists are not in agreement about the hard science in Swan’s study, specifically, the claim that the sons of pregnant mothers exposed to (certain) phthalates exhibit smaller anogenital index than the average, and more significantly, the claim that phthalate exposure in utero is associated with reduced penile volume and delayed testicular descent.

The merits
This is one of the first studies to try to measure the impact of these chemicals on humans. The study appeared in a peer-reviewed scientific journal, which means that some (anonymous) scientists deemed the study worthy of publication. Peer-review is one of the gold standards of scientific research. The study did use sophisticated measures and statistical analyses standard in epidemiological studies. And the authors were careful to describe the difficulties in conducting such research. The researchers who did the measurements were trained, minimizing inaccuracy. We believe that Swan does indeed show a correlation between AGI and phthalates found in the urine of the pregnant women, though our analysis suggests the correlation is weaker than claimed.

Only recently have there been any attempts to assess the impact of phthalates on humans. Swan's study is carefully done, uses reliable measurements, and contributes to this important effort.

The weaknesses
Other scientists have added their own complaints in a letter to the Environmental Health Perspectives.

An important point they make is that no one has established the “normal range” of AGD or AGI in boys. We can’t tell whether the “smaller AGI” found in phthalate-exposed boys is outside of the normal range. If, for example, we found that eating red meat correlates with being taller, our level of concern would change depending on how dramatic the effect is. If “taller” means that red-meat eaters sport an extra inch, we might not blink an eye. But if “taller” means that red-meat eaters are a full head above the crowd, we might suspect nature gone awry. Swan’s study cannot lay claim to abnormal development, based on this measurement, without evidence that the range of observed values are indeed abnormal.

For those who read the literature, it should be noted that norms for anoscrotal distance have been established, this being the distance between the anus and the base of the scrotum. However, as Swan mentions, this distance does not correlate with phthalate exposure in utero.

The Anogenital Index doesn’t sell newspapers, but genitals do!
One of the big claims in press coverage of phthalates is that baby boys may have deformed genitalia thanks to phthalates. Not one baby in the study had any abnormal genital development. But does the relationship between phthalates and anogenital index mean that there is an increased risk for fertility problems?

One of the most important claims made by the Swan study concerned undescended testicles. In human males, the testicles typically descend in utero around seven to eight months of gestation. Swan’s study claims that short AGI correlates with an increased rate of (partially) undescended testicles at measurement. If this were true, it would be a significant finding. Undescended testicles are a risk factor for infertility and testicular cancer.

Swan originally claimed that she correlated AGI below the 25th percentile to incomplete testicular descent. In a letter to Environmental Health Perspectives, Swan acknowledged that this relationship was not statistically significant, but she also showed that the trend (lower AGI and more incomplete testicular descent) is statistically significant.

The importance of this is that Swan has justified AGI as an important biomarker for male human development. This suggests that phthalates do merit further attention.

However, we should still be careful about making firm, causal conclusions. For example, Swan's study did not adjust for premature birth in measuring testicular descent. Since premature babies are much more likely to have undescended testicles, this may have skewed the results. Nonetheless, it is important to follow up with future studies.

Sex and Fertility
Another question that requires further investigation is whether those boys exposed in utero to higher levels phthalates are more likely to have sexual or fertility problems. This may be answered in the next fifteen years, but it is so-far not known.

The evidence to support the view that boys will suffer more problems comes from rat studies in which very high doses of phthalates resulted in reduced AGI and fertility problems. However, there are some rat studies which suggest that lower doses of phthalates reduce AGI but do not lead to fertility problems. If humans follow the rat pattern, they also may have no fertility issues caused by these chemicals. If they do not follow the rat pattern (some toxicologists believe that humans process and discard phthalates better than rats), then the whole comparison is moot. The issue with undescended testicles may play a role, since this is also correlated with reduced fertility.

What do other researchers say about phthalates?
The National Institutes of Health has put together several expert panels to evaluate the literature and research on phthalates; these reports can be read here. They evaluated each of seven phthalates separately in 2000, and recently updated their review for one phthalate, DEHP. Ironically enough, the panel downgraded their level of concern for DEHP in healthy pregnant women, in part based on the fact that Swan’s study found no correlation between this particular phthalate and small AGI.

Overall, the phthalate story is a complicated one, in which animal studies are well-established and human studies are all but nonexistent. There is evidence that our exposure level is about 1000 times lower than the level of exposure at which rats display observable effects. However, since primates are different from rats, we cannot conclude that phthalates are either safe or unsafe based on the animal studies.

Perhaps the most important finding of the NIH panel concerns the way we are exposed to phthalates. It is not, as the media might have you believe, through toys and cosmetics.

How are we exposed to phthalates?

Cosmetics have taken a big hit, as has the toy industry, from activist-driven campaigns (such as those conducted by the Coalition for safe Cosmetics and the Public Interest Research Group over the past year).

But according to an NIH review done in 2000, the biggest source of exposure to phthalates is food. Food constitutes approximately 85-90 percent of phthalate exposure in adults, mostly through meat and fish. For infants, depending on whether a baby is breast or formula fed, the rate is 44-60 percent from food, with the remaining amount in both groups almost entirely attributed to dust. There have been studies about how easily phthalates from baby toys leech into a saliva-solution, but no one has determined whether this is typically ingested (and, if so, how much) by babies.

DINP is not used in pacifiers and baby bottles. But it is used in soft vinyl toys.

In February 2003, the Consumer Product Safety Commission (CPSC) rejected a Public Interest Reseach Group (PIRG) petition to ban plastic containing the phthalate DINP in toys for children aged five and under. The CPSC’s decision that there was no demonstrable risk was based on studies showing the rate that DINP migrates from vinyl toys and on the way children “mouth” toys.

Total mouthing time for babies 3-12 months old is about 10 minutes per hour, including pacifiers, bottle nipples, parts of their own bodies, and any other device (designed for sucking or not). For older babies, the numbers go down. Pacifiers constitute most of babies' sucking time (about a third) and their own body parts are preferred next. Soft vinyl toys containing DINP were sucked on for under 11 seconds per hour, or under 5 minutes a day. Even the most avid suckers (in the 99th percentile) were chomping at their DINP toys for at most 12 minutes per day. A baby would have to suck for about ten times as long before they could consume enough DINP to have any potential adverse effects.

And in August of the same year, the European Union’s Institute for Health and Consumer Protection produced a risk assessment report for DINP, which concluded that:

“The end products containing DINP (clothes, building materials, toys and baby equipment) and the sources of exposure (car and public transport interiors, food and food packaging) are unlikely to pose a risk for consumers (adults, infants and newborns) following inhalation, skin contact and ingestion.

“The indirect exposure via the environment is unlikely to pose a risk to humans following the main route of exposure, the oral route… As combined exposure of adults is almost exclusively related to occupational exposure, the overall assessment indicates no concern for adults. For infants, combined exposure which is mainly related to exposure from toys and via the environment is not considered of concern.”

But legislators, activists and the media are still worrying about chewable toys, even though not all phthalates are used in toys, nor were all the phthalates reviewed in Swan’s studies correlated with shorter AGI. For example, the phthalate DEHP has been under close scrutiny because it has such a remarkable effect on rats, and it is found in infant toys that could be mouthed. Yet this particular phthalate was not correlated with shorter AGI in infant boys according to the Swan study. Another phthalate – DBP – has been found to cause problems in rats, and is also associated with smaller AGI according to the Swan study, yet it is not typically used in infant toys.

The phthalate that is used most in baby toys, DINP, was not considered by Swan’s study, and does not cause reproductive health problems for rodents, according to the expert NIH panel in 2000. However, DINP has been shown to cause other problems in rats at high doses, and the NIH panel recommended scientific studies to evaluate risk to humans at current (at that point, undetermined!) levels of exposure. Since then, scientists have begun to concur that the risk of toxic effects from DINP is small or nonexistent, as noted by the Institute for Health and Consumer Protection.

What should we do about it?
Overall, these finding suggest that legislation restricting the use of phthalates in toys is going to have virtually no effect on phthalate exposure. Neither will attempts to limit phthalates in cosmetics. Such measures will not reduce the main sources of our exposure to phthalates (since our main source is food). It will not reduce our exposure (or our children’s) to the particular phthalates that are correlated with short AGI even if we buy into the claim that this is a worthy goal.

Rebecca Goldin Ph.D is Director of Research for STATS. For media or other inquiries, please contact Matthew Felling at


Technorati icon View the Technorati Link Cosmos for this entry