Two recent studies have only added to the understandable public confusion around the safety of soy. First, a large, prospective epidemiological study in China reported that breast cancer survivors who consumed the most soy showed no greater mortality or relapse rate than those who ate the least soy. In fact, there was a statistically weak benefit for the soy lovers. On the other hand, researchers at the National Toxicology Program (NTP, a US Department of Health and Human Services program) issued a cautionary report and call for more research on the safety of soy-based baby formula, raising concerns that exposure to the concentrated soy extracts in the formulas may have detrimental effect on developing infants.

So what’s going on here? How can soy be safe for breast cancer patients, for whom relapse-free survival has been definitively tied to minimizing estrogen exposure, yet be unsafe for babies?

Whole Soy versus Soy Isolate

First, one point that gets lost in most consumer media reports is that there is a big difference, in terms of phytoestrogen exposure, between consuming soy in the form of whole bean products (soybeans, tempeh, and tofu) versus a soy extracts (soy protein isolate, for example). The baby formulas the the NTP researchers are investigating do not contain whole soy, but rather soy protein extracts or isolates, which may contain many times more genistein, the primary phytoestrogen in soy, than whole soy products contain. This matters because in laboratory animal studies, genistein as a stand-alone compound has been shown to cause adverse health effects, including predisposition toward some cancers and immunosuppression.

Reproduced from the UK's Daily Mail

In Asia, where soy products are regularly consumed, hormone sensitive cancer rates are generally lower, and there is certainly no evidence of widespread developmental problems. But this routine consumption of foods made from whole soybeans may prove irrelevant to the safety of soy protein isolates found in baby forumulas. In the quantities consumed, these formulas expose infants to concentrations of genistein that would normally require eating several pounds of tofu per day–a rather extreme quantity.

What the NTP report says

The NTP report called attention to the overall inadequacy of existing research into the health effects of soy-based infant formulas.  In a review of existing research, it explained that many studies were simply too small to result in definitive answers (see Size Matters). Others looked only for gross developmental changes, and failed to test for subtle changes in tissue development that might predispose or protect soy-fed infants from developing cancer, cause growth or hormonal irregularities, or cause cognitive and behavioral differences versus milk-fed peers. The NTP report itself doesn’t attempt to answer any of these questions — it simply makes the case that there are enough data to raise safety issues but not to provide answers. And, it calls for additional research, and for hearings to begin to define the parameters and goals of that research.

The issue is particularly crucial in areas of the world where there is heavy reliance on soy-based infant formulas for extended periods, whether for economic or nutritional reasons, or simply due to the mistaken belief that soy-based formulas are nutritionally superior to milk. What is well known is that fetal and infant exposures to harmful chemicals, or even normal human hormones in abnormal concentrations, can have permanent and harmful health effects later in life.

It’s possible that this debate will heat up as infant formula manufacturers get involved, not to mention the soybean growers’ industry and groups promoting a vegetarian lifestyle. But given the stakes for vulnerable infants, the NTP was right to raise the issue, and to begin the process of resolving it.

Soy for grownups

If the safety of soy protein isolates is an issue for infants, what about adults? There has been a similar mix of strong and weak science surrounding soy consumption in adults, with divided results on whether soy helps or harms the immune system, or affects hormone-dependent cancer risk, or bone mass, or menopausal symptoms.  Complicating matters, much of the media reporting of this research has failed to distinguish between whole soy foods and soy protein isolates or other processed products.

The task of sorting this messy body of research is not, to my knowledge, being undertaken by any one institution. In the meantime, a consensus of nutritionists seem to agree that whole soy foods (soybeans, tempeh, and tofu) are a heart-healthy substitute for dietary meat, and that limited soy consumption (loosely defined as twice a week) is safe for the survivors of breast and other hormone-dependent cancers, and that consumption of whole soy foods in moderation neither protects against nor promotes hormone-dependent cancers.  In contrast, soy protein isolate, genistien supplements, and even, in some instances, soy milk, have all been questioned in at least some research results, and should be consumed with caution, until conflicting data are resolved.

But for breast cancer patients and survivors, the evidence of the China study was strong enough that an oncologist at the Fred Hutchison Cancer Research Institute felt comfortable giving her blessing to regular consumption of whole soy foods (but not isolates or supplements).

Resources:

NTP Draft Brief on Soy Infant Formula (full text, PDF format)

China Study on Soy and Breast Cancer Survival (abstract only, from the Journal of the American Medical Association)

Thank you Gary Schwitzer for addressing absolute versus relative risk, despite the media’s (and some researchers’) viewing the Tamoxifen issue as proof of patients’ stupidity.

I’m going to start with the conclusion of the New York Times column that inspired this commentary, in which a doctor at Memorial Sloane-Kettering dismisses the Tamoxifen problem as something that will be solved by aromatase inhibitors, the new generation of anti-estrogen therapy. I have news for the oncology community: aromatase inhibitors won’t change a thing.

Women don’t take these pills, and doctors don’t understand why, because they each hold a very different view of their costs and benefits. I used the term costs rather than risks for a reason, because this has very little to do with women over-rating the rare fatal side effects (those are a factor, but they’re not the factor). To explain, let me start with a simple cost-benefit matrix since, unlike a risk tolerance matrix, it puts a different emphasis on probability:

A typical risk tolerance grid, which is what seems to drive oncology research, says, in a nutshell, that the bigger the harm, the more risk averse we become, and the more we’ll do to avoid it (the upper right hand corner in that kind of matrix is usually an angry red). But in a cost-benefit matrix, this quadrant merely represents big gains in exchange for big costs. If the gains minus the costs are a wash, most people would rationally be neutral; so, the difference between the gain and the cost, or the probability of achieving the gain (or incurring the cost) become crucial.

When researchers broadcast juicy numbers about population risk and a risk-reducing drug, they’re almost always describing the bottom right-hand corner (huge benefit, small sacrifice), and almost as frequently understating the full costs as well as overstating the benefits. This New York Times column is a typical example. I can forgive anyone who read that column and came away thinking that taking Tamoxifen is a no-brainer and that patients are fuzzy headed at best.

Let me approach this from the perspective of breast cancer survivors, who are routinely prescribed Tamoxifen to prevent relapse. This is important. Breast cancer is relatively easy to knock out, but relatively hard to keep from coming back, and 80% of relapses are ultimately fatal. So, you could say that breast cancer patients have a lot to gain from Tamoxifen, and you would be forgiven for seeing their decision as another no-brainer. And yet, recent studies have found that fully 30% of breast cancer survivors abandon their Tamoxifen, or fail to take it as often as prescribed. Are they irrational? Hardly.

The Magnitude Question

For starters, the health researcher sound bytes you read in the media almost always ignore the magnitude of the risk. When you hear anyone describe anything as causing a percentage increase or decrease in anything, you should immediately ask “50% of what?” That’s the magnitude question. Tamoxifen doesn’t reduce cancer risk from 100% to 50% (that really would be a no-brainer!). It only halves the existing risk, and depending on the population, that can be anywhere from 50% to a single digit percentage risk. For the population of early-stage breast cancer patients who have undergone chemotherapy, the risk of relapse is (generalized) around 10%. So, Tamoxifen lowers their relapse risk from approximately 10% to approximately 5%. That’s the benefit side of the equation — a 50% improvement in an already smallish risk of a potentially catastrophic event.

Still, if you look at Tamoxifen from this perspective, it still looks pretty good — a significant (but not huge) reduction in the risk of a very bad outcome. That would put Tamoxifen somewhere in that upper left hand quadrant — where the majority of people would be expected to go for it.

Population Risk versus Individual Risk

But there a couple of key ingredients missing from this unfortunately typical analysis. First, as any breast cancer patient knows, there’s a big difference between population risk decrease and personal risk decrease. As an individual patient, you can take Tamoxifen ’til the cows come home and still relapse and die of your original cancer. That a given population of breast cancer survivors will have 50% fewer relapses if they all take Tamoxifen exactly as prescribed is zero guarantee that any individual woman will avoid relapse. Have you ever heard that sneaky fine print at the end of a commercial, “Individual results may vary“? Well, the same applies to anti-cancer drugs. And now you’re getting closer to how patients, rather than epidemiologists, view the benefits of Tamoxifen. The benefit that they, personally, will gain from Tamoxifen is simply not knowable.

Side Effects, and Side Effects of the Side Effect Drugs

Another thing that the sound bytes rarely address, but that the Tamoxifen package insert clearly states, is that Tamoxifen can cause debilitating strokes and death. This, like relapse, is an outcome squarely in that upper right hand corner of the risk matrix — something you’ll do a lot to avoid. Doctors and epidemiologists love to point out that the risk of relapse is much higher than the risk of a fatal clotting mishap, and they’re right. And yet, that doesn’t mean the risk of death from Tamoxifen can be ignored, especially when it’s time to swallow the pill every morning, or when contemplating that long flight to your in-laws for Thanksgiving. More commonly, women on Tamoxifen experience a variety of miserable side effects from the drug. Oncologists prescribe drugs to help with the side effects, typically anti-depressants or Neurontin or blood pressure medications or even estrogen, but all of these, too, have side effects.

So, when a woman decides whether or not to take Tamoxifen, she’s not applying the simple math of the epidemiologist: 50% reduction in relapse, minus 1% risk of death from uterine cancer, minus smallish risk of debilitating stroke = no brainer! In reality, especially for women who are already on the drug, and already suffering from intolerable side effects, the equation is a little more complicated:

50% reduction in relapse risk that may already be as low as 10%
– discounted by a 100% chance of side effect misery and low quality of life
– discounted by a 2-3% change of dying from the drug
– discounted by an 80% chance of bitterness if relapse occurs anyway
– discounted by an 85% chance of bitterness in the event of a Tamoxifen-induced stroke
– discounted by a 100% chance of bitterness in the event of Tamoxifen-induced uterine cancer

Bitterness. There’s that emotion thing. Any perfectly rational being would at this point say, “sure, but no one ever died of bitterness.” To this, most of the drop-out patients would respond, “true, but the risk of bitterness is a cost that, added to the cost of 5 years of misery, makes the magnitude-attenuated benefit not worth it.” Notice that it’s not a sound byte, and that we’re now in values territory.

What’s it worth to ya?

And to understand this values thing, you need only explore the side effects for a minute — not the relatively rare ones, like uterine cancer, but the exceedingly common ones, like severe hot flashes for example. Health pundits love to talk of hot flashes in the abstract, or as a minor nuisance. But to anyone who’s ever had a severe hot flash, like the kind you get on Tamoxifen, it’s like hearing a man dismiss labor pains. A bad hot flash is as disruptive and unpleasant as a sudden fit of vomiting or a convulsion. Severe hot flashes can eject you from deepest sleep into miserably fevered wakefulness twenty times a night, every night for months or years on end. They can become so intolerable that sufferers evade social gatherings, limit their activities, and withdraw from life. And what the New York Times column describes as “changes in vaginal discharge” often means vaginal dysfunction so severe that sex becomes torment. Breast cancer forums for younger women especially are overflowing with the shared misery of these side effects. In other words, Tamoxifen, and the aromatase inhibitors that are the only alternative, for a significant number of women who take them, have a huge impact on quality of life. And they are typically prescribed for a five year course.

After a few months of these side effects, and with a more realistic appraisal of that “50% reduction”, is it really any wonder that a woman’s calculus of the benefit drifts over time? That drift is not a measure of a patients’ inability to do math. It’s a measure of our willingness to suffer chronically in exchange for a genuinely unknowable personal benefit. If you could guarantee a patient that she will not relapse if she takes Tamoxifen — a guarantee that it will keep her within the fortunate cured-for-life population — you would see a different calculus. But to pretend that Tamoxifen is tantamount to a life or death decision for any individual patient, or that abandoning the drug is lunacy, is itself irrational, or at least depressingly ignorant.

who doesn't abhor a vacuum?

I woke up this morning with a mammogram debate hangover. Enough with the new guidelines, already! But still, what do we do with the mammogram-shaped holes in our lives?

As it turns out, nature abhors a vacuum almost as much as I do, and there are already some interesting and, fingers crossed and lucky rabbit’s foot in hand, promising developments on the horizon. One of the most interesting is the recent research into microRNA. This type of protein molecule could lead to more accurate cancer screening than lousy mammograms or the crude bio-markers we have today, like PSA (for prostate cancer) and CA 15.3 (for breast cancer).

What’s miRNA?

Think back to your high school biology class… OK, too far? Then think back to the advent of biotechnology, and the massive Human Genome Project, and how companies like Genentech and Amgen were going to cure cancer and generate new organs and allow us to live forever. Heady days! Someone made a killing in the market, no doubt. As for the rest of us….

If you were anywhere but in a coma during the biotech bubble (about ten market bubbles ago), you heard a lot about genes and DNA. A light, quick, lemony refresher: DNA is a twisty molecule in the center (the nucleus) of the cells in your body, and it is organized into segments call genes. If you think of cells as little chemical factories that churn out proteins and other molecules all day long, then think of DNA as the ultimate molecule cookbook. DNA molecules are unzipped (DNA resembles a twisted ladder, and it “unzips” right up the middle of the rungs), and portions of it are copied. RNA molecules are the copies.

There are different types of RNA that serve different purposes, but the ones to care about right now are messenger RNA (or mRNA) and micro RNA (or miRNA). The mRNA molecules act as messages that cause parts of the cell to do something — to produce yet another protein, for example. MicroRNA molecules can intercept and bind to mRNA, changing its shape and making it ineffective. In essence, if you think of mRNA as a message, miRNA renders the message unreadable. Whether that’s a good thing or a bad thing depends on the nature of the message.

In a cancerous cell, there may be far too many mRNA messages that trigger growth. So, having miRNA molecules to intercept and destroy some of those messages would be a good thing. On the other hand, viral infections or DNA mutations can result in overproduction of miRNA molecules that turn off suicide messages — causing messed up cells to survive and grow into cancer. So, miRNA can act as the gas pedal or the brakes as far as cancer is concerned, depending on its function. There are hundreds of different miRNA molecules with hundreds of different functions, and researchers are looking at the overall mix of miRNA molecules in human tissues. What they’ve found is that the balance of different miRNA molecules can vary depending on what’s going on in your body.

Why I’m Wild About miRNA

This won't hurt a bit

A number of small studies (remember Size Matters? this is early stuff) have found that miRNA in the blood of test subjects could detect breast cancer, prostate cancer, diabetes, and other diseases. And, mixtures of miRNA appear to be a more consistent and direct measure of the presence of disease.

Existing biomarkers like PSA and CA 15.3 measure proteins that some tumor cells produce and some don’t, and which some healthy cells also produce. They can also be neutralized to some degree by the immune system until the number of tumor cells is extremely high. PSA and CA 15.3 can fail to rise even with significant cancer burden, and can be elevated in some cancer-free individuals. So, as screening tools, they stink. That’s why we’ve been stuck with breast self-exams and mammograms. The early miRNA research suggests that certain clusters of miRNA are a more specific and reliable test than either existing tumor markers (for breast and prostate) or mammograms, and contain more information about the characteristics of the cancer.

Cancer screening tests based on miRNA are in the pipeline for some cancers (including lung and colon), though not breast or prostate, yet. But it occurs to me that miRNA breast cancer diagnostics would be a fine way for the USPSTF (US Prevention Services Task Force — the folks who brought you the new mammography guidelines) to help fill the void they created. Perhaps they could recommend sugar-daddy funding for the research, or a dreamy tax break for the first ten companies to get their miRNA cancer detection kits through the FDA.

Some day, there may be a post-mammogram world where a simple annual blood test is all you need as an early screening tool. For now, stuck with prodding our girls and the occasional lousy mammogram. But hey, a girl can dream, can’t she?

References

MicroRNAs as Novel Biomarkers for Breast Cancer (full text)
H. M. Heneghan, N. Miller et al (2008)

miRNA as a biomarker for prostate cancer:
Circulating microRNAs as stable blood-based markers for cancer detection (full text )
P. S. Mitchell, R. K. Parkin, et al (2008)

Characterization of microRNAs in serum: a novel class of biomarkers for diagnosis of cancer and other diseases (full text)
Xi Chen, Yi Ba, et al (2008)

A truth serum for cancer — microRNAs have major potential as cancer biomarkers (full text)
LJ Chin and FJ Slack (2008)
A commentary of microRNA-as-biomarker research, explaining the recent history, and what’s yet to be discovered before effective clinical tests exist; be sure to read the article by Mitchell and Parkin, which addresses some of the questions raised here.

Serum MicroRNAs Are Promising Novel Biomarkers (full text)
Shlomit Gilad, Eti Meiri, et al (2008)
Demonstrates use of microRNA as accurate test for pregnancy phase.

Imagine that you’re a health consumer. It’s not that difficult, there are only a couple of details you need to muster: (1) that you’re concerned about breast cancer, the number 1 killer of women aged 15-59, and (2) that the medical experts have insisted, throughout your entire post-adolescent life, that mammograms do indeed save lives. So although the tests are a gawd awful humiliating misery, you subject yourself to them annually, or plan on doing so when you’re older.

Never mind!

Then one year, these same medical experts take another look at the data during a tense debate on medical costs and issue a Rosanne Rosannadanna-style “Oh. Never mind.” And this “never mind” happens in the same year in which other experts say, Hey, you know that BPA that we insisted is perfectly safe? Well we’re rethinking that. And, uh, sorry about possibly predisposing you or your children to cancer.

And by coincidence the mammogram Never-Mind also happened just a few years after those same medical experts did an epic flip flop on hormone replacement therapy (HRT). That Never Mind came after years of selling HRT as the new fountain of youth that would stave off hot flashes, sleep problems, alzheimers, heart disease, and would pretty much usher us into health nirvana well into old age.

Getting back to the mammography guidelines: last month, the radiology and breast imaging director at my regional cancer center (a National Cancer Institutes affiliated clinic) came out strongly in favor of women aged 40 to 49 continuing to have annual mammograms. There’s more than enough debate among the experts out there to justify any woman deciding she’d just as soon keep her options open for now.

So, the reason the scientists and mathematicians are confused is that they’re oversimplifying the issue. It’s not just about the tidy numbers or pretty graphs. It’s about consumer trust, and personal health risk, and individual choice. Consumers have learned, from the history of epic reversals and re-reversals by the experts and policy gurus, that these issues are a complicated, that one-size answers inevitably don’t fit, and that they boil down to a question of individual choice.

This brings me to my final point: women are not demanding mammograms, they’re demanding choice. Plenty of women are already foregoing annual mammograms and undoubtedly view the new guidelines as a welcome absolution from nagging doubt. But others quite reasonably fear that the health insurers will now phase out coverage of mammograms for women under 50. Women are afraid that insurers will rob women of the choice to get a mammogram, thanks to the guidelines. That’s what’s making women angry.

This has nothing to do with mammograms, everything to do with missing Gilda Radner:

A Poem by Gilda Radner (a.k.a. Roseanne Rosannadanna)

Doctors are whippersnappers in ironed white coats

who spy up your rectums and look down your throats

And press you and poke you with sterilized tools

And stab at solutions that pacify fools.

I used to revere them and do what they said

Till I learned what they learned on was already dead.

Here’s a super-important issue that this whole mammography debate is only glancing off: no one is offering anything to pre-menopausal women. Yes, I understand that the epidemiology shows little or no benefit. Yes, I understand that (1) the benefits of mammography were grossly oversold, and (2) the risks were grossly understated. Those were preventable mistakes, by the way (don’t get me started on the failure of our medical system to provide breast health specialists).

But, epidemiology aside, if you spend a month in the infusion center at any cancer clinic, you will meet women in their 30’s and 40’s with terminal breast cancer precisely because their doctors pooh-poohed the mammo (”you’re too young for breast cancer”). No amount of epidemiology or number crunching can get around the fact that, for some women, mammograms DO save lives, or at least breasts.

What the debate is mostly ignoring so far is this: that there’s a difference between public health recommendations and personal choice, but that in our current regulatory and insurance atmosphere, decisions in the public realm can obliterate choices in the second. Epidemiology is an important input into public health decision making, but is not the final answer in personal decision making, especially when our tools for identifying “high risk” women suck even worse than mammography. Are you Ashkenazi Jewish? No? Then no mammo for you. That’s what our individual risk assessment tools boil down to – BRCA. The cause of maybe 10% of breast cancers.

My two cents (okay, maybe this is four cents):

(1) There are still excellent reasons to get a mammo under the age of 49 (lost in the population data is the fact that some lives and breasts ARE saved), and

(2) doctors have been worse than abysmal at understanding and communicating the risks, and

(3) women under 50 should be allowed to choose and should be offered the information needed to choose intelligently, but

(4) the task force recommendations will likely eliminate this choice by leading insurers to drop pre-meno mammo coverage like a nuclear potato.

All of which makes me sad, and worried, that if the choice to obtain a mammogram is taken away from pre-menopausal women, that there will be more women like the ones I met at the cancer clinic.

Medicine has done an abysmally poor job at explaining the risks and benefits of mammograms to pre-meno women. But the answer to that failure isn’t to abandon this population altogether. That’s just swinging from one simplistic extreme to another.

The problem and source of controversy is two-fold: (1) no one really knows how much we’re exposed, and recent pilot studies indicate that it’s a lot more than the chemical industry claims, and (2) if you decide you would rather play it safe while more research is conducted, too bad. The industry and government have already decided for you that you’re going to get lots and lots of these chemicals in almost every product you buy.

Is this a cause for panic?  Better to say it’s a cause for getting off your comfy assumptions and getting involved. Too busy or confused?  Join the party, and here’s some motivation to help with that:

Where are these endocrine disrupters found? Here are just a few:

Atrazine: A weed killer widely used in US agriculture, and now widely found in our drinking water and food, it is linked to reproductive cancers, and chemically reverses the effect of some breast cancer treatment drugs. It has been banned in several European countries.

PCBs (polychlorinated biphenyls): Environmentally persistent chemicals used widely in manufactering and electronics. Most individuals tested have them in their blood, and you continue to absorb them from your food, water, and even your indoor environment. Because they have been used so long with so little regulation, they are now endemic in the US.

PBA (Bisphenol A): Widely used in plastics, including the lining in canned foods and some “green” products like SIGG water bottles, clear plastic baby bottles, and some plastic drinking water bottles, often at levels exceeding the controversial EPA exposure limits. The highest levels of BPA have been found in the blood of infants and children, the population most susceptible to BPA’s harmful effects.

Phthalates: You name it. These compounds are found in household products (detergents, plastic bags, food packaging), children’s toys, and the vast majority of personal care products (often disguised with the term “fragrance”).

Parabens: Preservatives found in almost every personal care product on the market, from soap to shampoo to moisturizers to baby products. Most importantly, they are in products labeled “all natural”, though most (but not all) products labeled organic omit them.

PBDE’s (polybrominated diphenyl ethers): Flame retardants that are so toxic they’re controversial even among firefighters’ unions, these compounds are now in your house, in carpeting, furniture, your mattress, and your children’s pajamas. They are persistent and accumulating in the environment, and in human breast milk, and in animal products including milk, meat, and fats.

Endocrine disrupters have been investigated for their links to disease for some time, but interest has recently increased when higher-than-expected quantities were found in drinking water and in human blood samples, and when environmental health scientists began to push for more research into the effects of multiple exposures. Most of the existing research on industrial chemical safety (in general, not limited to endocrine disrupting chemicals) is flawed in two important respects: (1) it was conducted by the manufacturer, with limited controls on the obvious conflict-of-interest, and (2) it is limited to the effects of individual chemicals, though it is a fact that we are exposed to a cocktail of chemicals every day that affect the health of our cells.

To quote a recent Endocrine Society Scientific Statement (which was adopted by the American Medical Association):

The evidence for adverse reproductive outcomes (infertility, cancers, malformations) from exposure to endocrine disrupting chemicals is strong, and there is mounting evidence for effects on other endocrine systems, including thyroid, neuroendocrine, obesity and metabolism, and insulin and glucose [balance].

So, as a consumer, you have to ask yourself: in the face of strong but not conclusive evidence, do you want to continue to expose yourself and your family to these chemicals, or would you rather play it safe and avoid exposure?  Does it matter to you that the American Medical Association agrees that these chemicals are contributing to cancer, birth defects, obesity, blood sugar control issues, and neurological defects? Regardless of your answer, or if you’re too young to have an answer, federal regulators and chemical manufacturers have already decided for you: you’re going to be exposed, like it or not. No one consulted you, you had no say in the matter.

Unhappy about this? Want to do something about it?  Are you already “going green” with your wallet? It turns out that simply voting with your wallet isn’t enough. For example, even women who shopped organic and stopped using any personal products containing “fragrance” still had excessive levels of phthalates in their blood, and were passing these chemicals right through to their babies. While I’m all for supporting genuinely organic products, and not just the phony “green” marketing label (go ahead and read the ingredients next time you pick up a product labelled “natural”), it’s not remotely enough. Using organic soap won’t protect you from the chemicals in your clothes, house, and food. Buying organic produce won’t protect you when the water they are irrigated with contains PCBs. Wearing organic cotton won’t protect you from the chemicals oozing out of your carpet pad, sofa, and bed.

There’s really only one way to solve this problem, and that’s to overhaul the way we safety check and control artificial and even naturally occurring chemicals. True, Congress has taken baby steps. The Consumer Product Safety Improvement Act of 2008 will ban phthalates in some children’s products. Some PBDEs have been banned but others have not. And how will those bans be enforced when less than 5% of all imported products are inspected? And even with perfect enforcement, getting phthalates out of kids’ toys doesn’t come close to getting phthalates out of your body and breast milk. Ditto for every chemical listed above.

If you would rather not be a guinea pig, and you would rather not turn your children into guinea pigs, there’s only one thing you can do that will make a difference: vote with your voice, and with your real vote, not just your wallet. Talk to people about this stuff, especially your elected officials. Contact your city, state, and federal governments and let them know how you feel. And because that’s a meaningless statement without a means of taking action, I’m adding a few links to organizations who make it easy. Know of another group you’d like to add to this list? Please email me, so we can expand this list!

As always, questions and comments are welcome!

Take Action, Get Involved, Add Your Voice:

These organizations all have a “take action” newsletter that will keep you in touch with government decisions affecting your exposure to endocrine disrupters and other chemicals. They each put consumer safety first, not economics or politics or jobs. If that causes economic “dislocations”, then let’s address them — subsidize, train, repurchase or reinforce our way into a healthy environment — but doing nothing just because “it’s complicated” is slowly poisoning us. The first step, though, is to start banning some of these chemicals. (Remember how eliminating CFC’s to fix the ozone hole was going to destroy the economy?)
Breast Cancer Fund
Safer States
Food and Water Watch
Safer Chemicals, Healthy Families
Campaign For Safe Cosmetics

Local consumer-interest groups devoted to environmental toxics exposure:
Washington Toxics Coalition
Silicon Valley Toxics Coalition

References:

Atrazine

Atrazine’s hormone-mimicking action reported here:
Atrazine-Induced Aromatase Expression Is SF-1 Dependent: Implications for Endocrine Disruption in Wildlife and Reproductive Cancers in Humans

The Breast Cancer Action interview that first exposed the corporate ties between atrazine production and aromatase inhibitor production:
A Hormonal Seesaw: The Atrazine and AI Connection

BPA (Bisphenol-A)

Human exposure to BPA and health effects:
Human Exposure to Bisphenol A

BPA exposures and research history summary:
Exposure of the US population to bisphenol A and 4-tertiary-octylphenol: 2003-2004

Flaws exposed in industry-sponsored BPA safety research:
Why public health agencies cannot depend on good laboratory practices as a criterion for selecting data: the case of bisphenol A

Medscape BPA exposure and health effects summary (free registration required):
Bisphenol A: A Scientific Evaluation: What Is the Magnitude of BPA Exposure?

How BPA exposure predisposes infants to breast cancer later in life:
Does breast cancer start in the womb?

Environmental Working Group – Tips for Avoiding BPA Exposure (insufficient, but slightly better than nothing):
Bisphenol A: Toxic Plastics Chemical in Canned Food: Consumer tips to avoid BPA exposure

PBDEs (polybrominated diphenyl ethers)

PBDE Exposure levels in Adults and Children, and Health Effects:
http://www.ehponline.org/members/2008/116-5/focus.html
PBDEs and Your Health

Developmental harm from in utero exposure to PBDEs in rats:
Developmental Exposure to Low-Dose PBDE-99: Effects on Male Fertility and Neurobehavior in Rat Offspring

Exposure to PBDEs through food:
Red Meat and Poultry source of PBDEs

Environmental Working Group – Tips for Avoiding PBDE Exposure (insufficient, but slightly better than nothing):
Reducing your exposure to PBDEs in your home

Washington State flyer on reducing PBDE exposure (insufficient, but slightly better than nothing):
How Can You Reduce Your Exposure to PBDEs?

And by the way, don’t be fooled by fake consumer groups set up by chemical industry PR firms:
Citizens for Fire Safety Exposed

Phthalates

Effects of phthalate exposure in animals:
Perinatal Exposure to [Phthalates] Alters Sexual Differentiation of the Male Rat

Human exposure to phthalates through household products and some drugs may exceed EPA threshholds:
More concerns about phthalates: Effects may be cumulative

Possible link between phthalate exposure and ADHD symptoms in school-aged children:
Phthalate Exposure Linked To ADHD

Phthalates may adversely affect emotional/psychological development of boys:
Pilot study relates phthalate exposure to less-masculine play by boys

This article discusses phthalates’ impact on the psychological development of boys, and emphasizes that “shopping green” ain’t enough to protect yourself:
Toxins in toiletries harming unborn kids

How much would you have in common with this test subject?

The next time a health news report is causing you excessive worry (or manic optimism), when they’ve dodged your goo tool, and perhaps even your size matters tool — when some editor is rubbing his hands in glee convinced that they’ve got you this time because the study involved people, and lots of them — your should instead turn up your nose and sniff, “Yes, dahling, but what kind of people?”

This isn’t simply about channelling Leona Helmsley, although that can be fun. What I’m getting at is that the nature of test subjects matters as much as their number. The quality of a human study is based in part on how the participants were selected, and their characteristics. It’s not so much whether the participants are normal (whatever that means), as whether they are like you.

You’ve heard a version of this in the red-wine-lowers-heart-disease studies (and ensuing debate about the healthfulness of moderate drinking). One objection aimed at those studies is that selectively looking at red wine drinkers to judge heart disease risk amounts to asking the wrong question of the wrong group of people (and assigning the wrong credit). People who tend to choose red wine, the argument goes, might also have other tendencies, like eating their vegetables, or playing tennis at the club all day instead of working on the toxic chemical mop-up crew, or avoiding fast food like the plague. In other words, the kind of people studied may not represent the average person, or more importantly, you.

This is a huge factor in analyzing cancer risk and prevention studies, and even treatment studies, and is commonly missed by the media. Shockingly few media reports bother to mention one of my favorite people-separating factors: namely, whether the study participants were pre-menopausal or post-menopausal. For a hormone-driven cancer study, it’s hard to find a more relevant characteristic than hormone levels, and yet this detail is commonly omitted.

The kind of people studied is also a huge issue for most lifestyle-risk reports. Spend thirty seconds on Google and you will be overwhelmed by the sheer number of cancer risk and lifestyle studies out there. There are volumes upon volumes of them. On the topic of heart disease risk, we’ve all been whipsawed by the good-fat-bad-fat and good-carb-versus-bad-carb routines. And breast cancer research is a particularly rich example.

We are the beneficiaries of some very large, very well-designed population studies, notably the Nurses’ Health Study and Women’s Health Initiative. One of the breast cancer risk factors identified early on was obesity; similarly, a tofu-rich diet was identified as a protective factor. But over time, repeated analyses produced conflicting findings for these factors. The most settled explanation (at present) seems to be this: that the age at which you are exposed matters. For example, one theory is that a tofu-rich diet significantly lowers breast cancer risk only if it is consumed during childhood. Obesity may be a risk factor only for post-menopausal women. In other words, the kind of people for these factors is defined by age group.

Any number of factors can affect the validity of a study population, so much so that it would be impossible to generalize. But age, gender, family health history, personal history, nationality, pregnancy history, activity level, geographic region, exposure, lack of exposure… all of these can be important factors that make any given piece of health news either relevant or irrelevant to you.

Tool #3 is simply a reminder to ask, when reading a report or even evaluating studies your doctor mentions: who were the subjects and how were they selected? Were they selected in a way that ensures they are like you in the ways that matter for the benefit or risk in question? Or, if they were chosen for a certain outcome (breast cancer), were they chosen in a way that includes you (menopausal status, ethnicity, age, pregnancy history, hormone use) or is relevant to the topic being studied? If the answer is “No” to any version of the “what kind of people” question, then you’re rolling the dice if you make any Big Decisions based on those studies, even if they were big, and expensive, and well-done in every other way.

Resources:

For an encyclopedic, and myth-busting, list of cancer risk factors, visit the Risk Fact Sheets compendium at the National Cancer Institute online:

For the full Young Frankenstein Abby Normal brain quote, check out wikiquote.

The full article can be downloaded here, but I’ll provide a few highlights:

The study (sponsored by the NIEHS, or National Institute of Environmental Health Sciences) looked at several different organochlorine pesticides, of which the most well known is DDT. The link between DDT and cancer has been investigated in the past, with mixed results. The NIEHS study addressed several of the limitations of previous studies, including (drum roll, and link to Health News Tool #3), the kind of people studied. Among other flaws, past studies involved pools of individuals who were not representative of the general public (for example, farm workers, who had a greater-than-average exposure) or used measures that failed to adequately reflect exposure to the pesticide.

Should Mr. Yuk appear on blood transfusions?

In addition to addressing these weaknesses, the NIEHS study looked at the association between prostate (and breast) cancer for several pesticides individually, and at different exposure levels (using actual blood concentrations as measure of exposure). What they found was a distinct association between blood levels of each pesticide and development of prostate cancer. The higher the blood levels, the greater the prevalence of prostate cancer. Not all pesticides were the same — some showed a weaker link, but for some pesticides the prevalence of prostate cancer increased three-fold at the highest blood concentrations.

Now, I don’t know about you, but I can’t recall the last time anyone even used the acronym “DDT” in a sentence. DDT was banned nearly 40 years ago in the United States, yet continues to persist in our environment, to pollute our food and water, and quite possibly to promote cancer, despite decades of wrangling over its “cost-benefit profile”. Not only does it frustrate me that our health is still being threatened by pesticides abandoned more than a generation ago, but it makes me view the current debates over everything from Bisphenol A to Round-Up Ready Soybeans with a more-than-jaundiced eye.

In that moment I realized that I had absolutely no basis for choosing between them. I flashed on recent incidents, one at each hospital, in which a patient was killed through stupid and preventable error (for example, by injection of improperly labeled cleaning fluid instead of MRI contrast agent). And I recalled at least one near-fatal horror story for each hospital from friends and colleagues. I wasn’t thrilled about trusting my life to either, and certainly had no basis for choosing between them. So there I was, in not a little pain and with an ever-so-gradually collapsing lung, at a loss for choice in what is supposed to be a thriving, consumer-choice-enhancing free market for health care services.

Conservatives love to romanticize the free market, and there are plenty of reasons to do so. One of those reasons is consumer choice and another is competition. But, as with everything flowing from the mouths of politicians and pundits, a whole universe of nuance is lost. And in the case of health care choice, there’s one Great Big Glaring lost nuance: information access.

The High, Hidden Cost of Information Hoarding

The one thing conservative pundits fail to mention about our free market is that information, like coffee or diamonds or medicine, is a valuable commodity. Companies tend not to give it away for free, at least not usually, and especially not when it makes them look bad versus the competition. The very information you most want as a consumer is left in private hands to disclose, or bury, as strategy dictates. So, as a consumer, you not only have to pay for goods and services with your time and money, you also have to pay for information with your time and money.

Its the information-hoarding aspect of the current United States market for health services that bothers me. It makes it crystal clear that consumer choice, at least on the big issues, is fairly meaningless. I can choose my doctor, and I can choose my hospital and my insurer, but the basis on which I make those decisions is as shaky and ephemeral as a castle built on quicksand. There is no meaningful comparison or rating by which to choose your doctor, your hospital, or even your insurance benefits (think you know what your benefits are? I’d guess you haven’t suffered a serious illness yet).

I’ve seen the very high-level hospital “scorecards”, which rate hospitals on heart attack survival, for example. But they tell me nothing about hospital’s oncology outcomes, or maternity practices, or their philosophy about flu shots, or the number of false negatives from their testing labs, or the number of preventable, fatal staph infections their patients suffer each year. The one, pitifully inadequate resource for patient choice, aside from anecdote or meaningless local popularity contests (”Seattle’s best surgeon” or “Houston’s best hospital” — best for what??), is the ability to search indeterminately stale public records to see if a disciplinary action has been filed against a doctor. There are a handful of online attempts to provide consumer information (ratemds.com, vitals.com, healthgrades.com, and others), but they are messy, voluntary, and beyond spotty. I have yet to find a rating for a single one of my doctors in any of them.

We have freedom of choice, but choice is meaningless without information, and the information is too costly to obtain. In the case of a new lip-plumping lip gloss, the information about whether it works or not is small: a few bucks and a few days. But in the case of hospital safety or a physician’s judgment, the cost can literally be your health or even your life.

Going Dutch

I was mulling this problem when I came across this fascinating interview with the Dutch health minister in the New York Times. According to the minister, the Dutch government establishes prices for health services, and consumers can choose from among just-shy-of a dozen insurance plans offering a range of basic coverage and a variety of premium options. Because the prices are controlled, the conservative “think” tank The Heritage Foundation would argue that the Dutch have socialized, or nationalized, or centrally-planned health care and that such approaches remove all patient choice. Yet Dutch patients have more insurance options to choose from than I do, and have a better level of service (at least as measured in convenience factors like after-hours consultations) than we have in the US.

Most interestingly, the Dutch government is instituting quality indicators (in other words, a rating system) for hospitals, giving patients a meaningful basis for choice. In other words, the Dutch health minister gets that it’s not just a question of how prices are set. He gets that a free market is not the only way to have competition or meaningful choice. Most importantly of all, he understands that without informed consumers, neither choice nor competition is meaningful or effective. Without access to information, the potential benefits of a “free” market system are just too costly.

I’m not advocating the Dutch system, though it certainly has some positives, and helps the Dutch exceed the U.S. in both healthy life expectancy and avoiding preventable deaths. I am advocating a systematic, centralized information resource for consumers. Since our health services market is a highly regulated hybrid market, why not regulate it a little more, in the direction of meaningful competition and consumer choice? Why not require and enforce detailed reporting by hospitals, clinics, CDC field personnel, and even insurers — and make that information available to consumers? Yes, that would cost money. But so, too, does hiding information about the worst and best service providers. The next time your Senator lectures you on the dangers of stifling the free market, ask him or her about the dangers of stifling outcome-based competition, patient choice, and meaningful reform through information hoarding. Ask him or her about the costs of caveat emptor, or “buyer beware”, in the health care market. The next time your Senator warns that health reform will mean tax hikes, ask your Senator when (s)he intends to cut the hidden tax burden imposed by information hiding.

When a research study sets out to really prove or disprove something, it enlists hundreds or thousands of participants. This makes the study very, very expensive. A small army of health professionals, administrative staff, software engineers, and research assistants are needed to train, treat, and track that many participants. Why go to all that expense, all the grant writing and organizing and stress and strain? Because it’s the only way to rule out chance or, as epidemiologists like to call it, sampling error
.

What, exactly, is sampling error? There’s a superb example in the Harry Potter movies, in the form of Bertie Bott’s Every Flavour Beans. For example, at the end of Sorcerer’s Stone, Dumbledore admits that he stopped eating them after having once encountered a vomit-flavored bean. The thing with Every Flavour Beans is that they come in every flavor. So if you try just a few, you could easily conclude that (1) they’re the most delicious things on earth, or (2) they’re the most noxious substance under the sun, or (3) neither of the above. Because your sample is so small, all you would really be reporting is your personal experience, which can be very different from what the beans are really like. This is sampling error, and it’s why size matters. And the more important and complicated and ambiguous the question, the more size matters.

Sampling error is problematic enough when only one or two factors are involved (a vote for or against, a preference for toffee versus earwax flavor), but when a large number of variables are involved (think of the array of side effects that accompany most new drugs), then sample size deserves huge weight.

Instead of jelly beans, let’s say you’re trying to decide whether to take oral contraceptives. If you ask ten different women what side effects they experienced, you may well get ten different sets of side effects. Some of those answers will be common (weight gain, possibly), and others may have no overlap and range from minor (mild headaches) to severe (cancer). Given ten different answers, how do you decide what the risks truly are?

You need to expand your sample size. If you were to ask ten thousand women the same question, then your sample size would give you a far better sense of the probability of your developing any given symptom. If six thousand women out of those ten thousand experience a given side effect, you know something very different than if only three, or even six, of your ten friends experienced it.

Here’s a chart from the Pew Research Center, a non-partisan opinion polling agency. This chart is about sampling error in the context of opinion polls, not medicine, so you can ignore the specific numbers. What’s important here is the concept, and the trend. This chart is a diagram of why size matters.

Notice that the sample size, or number of participants, increases as the red line moves to the right. And, the red line drops (the margin of error decreases) as the red line moves the right. So, the bigger the study, the better the evidence.

Margin of Error

Margin of Error is the technical version of “size matters.” Margin of error means the percentage by which the results might vary from reality. A margin of error of 10% means that the study finding, whatever it was, is really “plus or minus 10%”. This can be a huge margin or a narrow margin, depending on the study finding.

For example, if your study showed a 10% risk of some outcome, with a 10% margin of error, it means the actual risk is not 10%, but somewhere between 0% and 20%. On the other hand, if the study found a 60% risk, then the actual risk is somewhere between 50% and 70% (the error, relative to the conclusion, is less of an issue, but still an issue).

So, margin of error is a handy short-hand measure of reliability (the quality of the study data, of which size is a big factor). And because of this, you can also think of it as a short-hand measure of significance (whether the study is useful for Big Decisions or only water cooler conversation).

In medical research, there’s a related concept called the “Confidence Interval” or “CI”. Like the margin of error, this is a loose measure of how reliable and significant the study conclusions are. Since CIs are one of the primary tools for evaluating the import of a medical study, and there are some interesting nuances involved, I’m going to devote a separate blog post to them. For now, bear in mind that, as with margin of error, smaller is better for confidence intervals, and that in general, the bigger the study, the smaller the confidence intervals.

Duration is Also Size

Finally, if the study involves a chronic condition like hypertension, or coronary artery disease, or cancer, or obesity, then the study has to be big in another way: Duration.

Duration is a huge issue for cancer preventative drugs or risks. Ditto weight loss products or sunblocks or supplements. It would be meaningless to put a large number of people on a diet or a new drug and watch them for only a few months. Even if the study population showed a benefit (or risk), and the effects could be tied closely to the drug or risk factor, there would be no way to know whether the effect was lasting. For most of us, lasting is what counts. Would you spend $10,000 on a pill that was going to help only for a little while? If “help” means prolonging your life, the answer might be yes. But if “help” means weak evidence of benefit in exchange for potentially serious side effects, the answer might well be no thanks!

So, the next time someone’s trying to get you all worked up over a new stay-young-forever concoction, or a cancer prevention pill, instead of tearing your hair or lining up at the clinic for the Pill-du-Jour, you can calmly say “Sorry, but your puny sampling size just doesn’t do it for me”.