What's in your milk?

Demko was 39 when she was diagnosed with a very aggressive breast cancer, and she's convinced that all the milk she drank played a part. "I'm not saying my history of breast cancer is related solely to dairy intake, but it's probably a contributing factor," says Demko. The potential link between cancer and birth control pills also scares her. Now, at age 50, there's no milk in her San Francisco refrigerator.

Demko is also watching out for her college-aged daughter, whom she's enlisted in her efforts to organize scientists and learn as much as possible about the potential link between milk and cancer. One summer, her daughter did an independent research project with a scientist at University of California, San Francisco, measuring insulin-like growth factor 1 (IGF-1) in dairy products. IGF-1 has been associated in some studies with increased height as well as cancer.

In October, Demko, a former corporate banker, helped raise funds to bring together about three dozen dairy researchers, from nutritional epidemiologists to dairy scientists, at a McGill and Harvard-sponsored conference. It was the first time such a diverse group of milk researchers had been brought together under one roof. Practically everyone in the room could agree on several things. First, cow's milk contains steroid hormones such as estradiol and testosterone, and peptide hormones such as IGF-1. Second, drinking milk has been shown to boost serum levels of certain hormones, particularly IGF-1, in humans. Third, high levels of certain hormones, particularly IGF-1, have been shown to increase the risk of certain cancers.

Some epidemiologists have connected those three dots and have suggested that cow's milk increases the risk of cancer. For example, large epidemiologic studies have appeared in major journals, reporting that prostate cancer - particularly aggressive forms - seem to be associated with dairy intake, and perhaps more strongly with total calcium intake. Such intake may double or triple the risk of aggressive prostate cancers, which kill about 2-3% of men.

The three days of presentations led Harvard epidemiologist Walter Willett to conclude that current US dietary guidelines, updated in 2005, were too bullish on milk. "I think it's not wise to recommend three [8-ounce] glasses per day for adults. Probably, a serving a day is OK; I don't see much reason that would be harmful. I'm concerned about two glasses a day, and three has a strong potential for harm."

Willett wasn't a member of the committee that created the 2005 guidelines, and two of those nutritionists who were - Pennsylvania State University's Penny Kris-Etherton and Purdue's Connie Weaver - say the committee reviewed the benefits, as well as questions of food safety, carefully. Willett's comments were, however, consistent with his previous statements on the subject, and were met with nods and agreement by many of the conference participants.

Whether you can connect the dots between milk consumption, IGF-1, and prostate cancer, however, as Willett did at the conference, is hardly an area of consensus.

Craig Baumrucker didn't grow up on a dairy farm. He's now a dairy scientist at Penn State, and he's exasperated by some of the claims being made against milk. Baumrucker, who attended the October conference and some of whose research is funded by dairy industry groups, doesn't connect epidemiologic associations, IGF-1, and cancer. "That mother's milk sustains and provides most needs for infants is without question," he says. "While it is known that milk's unique nutrition, such as special proteins and fats, [is] contributing some or all of this unique capacity, the suspected 'other factors such as hormones' are not clearly defined as contributing to these properties. Unless these known effects are controlled, evaluation of 'other factors' remains confounded."

IGF-1 is clearly on the list of "other factors" by those who say milk is harmful (see infographic). As its name suggests, it's known to promote growth, and some studies have linked high levels of it to cancer and even to higher rates of twins. While no one really disagrees that drinking milk is associated with higher serum levels of IGF-1, the mechanism is the subject of debate.

"It is very clear that it does seem, in multiple international studies, by independent people, with independent biases, that consumption of milk does raise your IGF-1 levels," says Michael Pollak, an oncologist who studies cancer risk and IGF-1 at McGill University and who was one of the organizers of the conference. When it comes to the mechanism, "historically, there have been two camps," says Pollak. "One says that all the IGF-1 is fully digested, so the hypothesis that you can absorb IGF-1 is kind of preposterous." More recently, there has been "controversial but serious evidence," says Pollak, that rather than existing only as free protein in the whey fraction of milk, IGF-1 is also intimately bound to casein and may therefore survive the gut intact.

"I think that well-meaning objective scientists can agree to disagree there," says Pollak. "But there is another game in town, which is the idea that milk consumption could increase IGF-1, but not through absorption. Rather, now there is the plausible hypothesis that IGF-1 may not be absorbed, but the milk may still raise the IGF-1 because it includes a growth hormone secretagogue." Pollak speculates that the secretagogue could be a GHRH-like substance, or ghrelin, or simply the presence of optimal amino acids, which would easily make it through the gut. He notes that one of the tests endocrinologists use to determine whether the pituitary is active is to inject arginine into the bloodstream, suggesting a role for such amino acids: "It could be that milk evolved not only to increase growth, but to stimulate growth factor production. It's still an unproven concept."

Loren Cordain, a professor of health and exercise science at Colorado State University, says it's an insulinogenic effect rather than due to IGF-1 being absorbed. Baumrucker says he just doesn't know the mechanism and would like to see proper studies. "How do you account for nutritional status versus hormones?" he asks. He says one way to do that would be blinded studies that compare outcomes of people who drink normally constituted milk with those who drink milk from which IGF-1 or other proteins have been removed.

In the absence of such research, he's fairly unimpressed with the idea that absorption of IGF-1 plays a key role. "We're secreting these things all the time," he says of growth factors and steroids, at higher levels than milk could ever provide. Pasteurization reduces protein hormones by 10% to 15%, he points out, and ultrahigh-temperature processing, which makes it possible to keep milk unrefrigerated and is gaining ground in Europe, removes even more.

Ultimately, it may not matter for human health as to which mechanism is at work. "It's almost immaterial whether the strategy that nature devised was to absorb a hormone or to find a way to stimulate hormone production," says Pollak, who also isn't sure IGF-1 can be blamed for increased cancer risk. Whatever effect IGF-1 consumption is having, "it remains a very minor component of interpersonal differences," he says. The genetic component is quite strong, he notes. "You could have a non-milk drinker with an IGF-1 level of 200, and a milk drinker with an IGF-1 level of 120. But the reason you cannot connect the dots is that while milk raises IGF-1 levels, there are so many other factors that are raising levels."

Edward Giovannucci, who has studied milk consumption and prostate cancer as part of Willett's group, says the magnitude of the effect is relatively modest compared to the difference in prostate cancer risk. "Perhaps 2-3 glasses of milk per day may increase IGF-1 levels by 10%," he says. "Whereas in our data, we see a relative risk of fatal prostate cancer of 2-3. I think it could be part of the explanation, but it doesn't seem to account for the whole difference."

"Just because it's in milk, is there a biological effect? Show me a phenotype," says Baumrucker. "There's something about milk, but what the hell is it?"

If you ask Baumrucker's colleague, Ron Kensinger, that "something about milk" isn't estradiol, either. Kensinger, another dairy scientist at Penn State, has measured estradiol levels in the milk of the 206 cows in the Penn State dairy research herd. The highest -17B levels, in cows that are late in pregnancy (greater than 141 days) was 3 picograms per milliliter. The test, he says, is highly sensitive, with a limit of detection of 0.7 pg/ml. Translated to a glass of raw whole milk - in other words, ignoring the effects of pasteurization and homogenization - that's 330 picograms. Drinking the US recommended dietary allowance of three glasses would add up to one nanogram per day. Compare that, says Kensinger, with the approximately 14 micrograms per day a prepubertal male produces, or the 24 micrograms per day a woman in the late stages of pregnancy produces. "So it's 1/1000th of the amount," he says.

Measurement of hormones, however, is not fool-proof. Saskia Sterk, of the EU Community Reference Laboratory in Brussels, said at the conference that current methods may be missing a decent percentage of many hormones, and her lab is working on better methods. Moreover, if IGF-1 really were getting stuck in the casein fragment, it wouldn't be measured by traditional methods, which look only at milk's whey fraction.

Even if the measurements are low, Kensinger isn't convinced that much of the hormones or growth factors in milk are bioactive in humans anyway. "Look at the data on birth-control pills," he says. "Only 30 to 50% of the steroid hormones in those pills crosses the gut to get to the bloodstream." Oral delivery just isn't very efficient, he says, and anything that does get to the bloodstream after passing through the gut hits the liver's first-pass metabolism, which decreases bioavailability by as much as 80%.

His data agree with epidemiologic studies of breast cancer and milk consumption. Scientists had hypothesized that the two would be related, either because of IGF-1 or estradiol in milk and the hormonal sensitivity of many breast cancers. Despite Demko's belief that milk has something to do with her breast cancer, however, epidemiologic studies have shown no relationship between the two, according to Pollak and Jianjun Zhang, an epidemiologist at the University of Arkansas for Medical Sciences, who has published on prostate, breast, and colorectal cancer, and milk. "I'm not willing to accept that hormones in milk are a bad thing," says Kensinger, "and maybe the opposite is true."

Connie Weaver lived on a small farm with "a diverse portfolio of activities: primarily chickens, but some dairy cows," until she was 8 years old. The Purdue University nutritionist, who was part of the committee that created the 2005 nutrition guidelines, is more worried about osteoporosis, and the calcium in milk that could prevent it, than she is about hormones. "Usually when you consume a protein, it's degraded in the digestive tract, so the chances that it's going to be absorbed intact isn't likely," she says. "That doesn't mean we shouldn't research the area and check out all the possible risks, but I don't think there's much evidence of harm at this point."

To her, that means the benefits of consuming milk far outweigh the risks. Multiple studies of osteoporosis and consumption of milk and calcium have shown a protective effect. Cohort studies also have shown a weak protective effect from milk against colorectal cancer, about a 15% reduced risk. A randomized, controlled trial of dairy products supports that idea, although case control studies have not confirmed the results. "We went through all this literature evidence, and the studies were, on a lot of levels, showing up to three servings of dairy a day were associated with better bones, and decreased risk of fracture, insulin resistance, and stroke. The only randomized controlled trial of calcium supplementation, she says, actually found a protective effect.

"I think most people agree, calcium intake is important," she says. "It's confusing to me why it's so controversial about the source. It has become a political football instead of looking at the science somehow. If you could provide all the nutrients in milk some other way, I don't have a problem with that." She just doesn't know how realistic that is.

For Pollak, the most fascinating part of the story is figuring out what has happened to dairy cows over the past century. On average, cows are producing six times the milk they did in 1900. There have been changes to what cows are fed; Monsanto has introduced recombinant bovine somatotropin (see "Dairy Economics,"); and there is some question as to whether cows are being milked longer into their pregnancies, which could alter milk hormone levels, although the data are controversial. Most of the increased milk production has been due to selective breeding. Today, most of the nine million dairy cows in the United States were sired by just a handful of bulls. "In the guided natural selection that has occurred in the dairy industry, classic animal husbandry has led to a huge increase in the milk production per animal," says Pollak. "It's very, very impressive. The fascinating gap in knowledge for me is that we don't really understand which genes were selected for."

"We speculate that the milk production is regulated hormonally," he says. "So it leaves open the possibility that the selection was for endocrine variants." It's very likely that levels of lactose and other major constituents of milk would be similar over the years. But, says Pollak, "that leaves open the question, have any of the hormonal microconstituents in milk been affected?"

When he first started looking into the issue, Pollak figured he could just pick up a 1982 textbook on the subject, then a 1992 textbook, and then a 2002 textbook to determine what had happened to these microconstituents. He soon found out that wasn't the case. The issues with hormonal assays decades ago were even worse than Sterk's evidence for how bad they are now, so the data are either useless or nonexistent.

Milk researchers would love to have an analog to the blood banks that have allowed infectious disease researchers to trace the presence and mutation history of various viruses. Such resources are very limited, however, says Pollak. So, they've turned to places where unchanged herds of cow-like animals live, such as Mongolia. There's also a herd of dairy cattle in Minnesota, which has been left to breed without selective pressure from humans. Getting access to their milk is difficult, says Willett, who has requested and been denied samples in the past.

Willett didn't grow up on a dairy farm, but his father was a physiologist who did the first transplant of a fertilized ovum in a dairy cow. Willett is quite concerned about the studies showing an increased risk of prostate cancer among those who drink milk.

Zhang, however, says those studies have a major limitation: The cohorts are physicians and nurses, two highly educated populations. That could mean that what people read about milk and cancer could affect their responses to surveys. "People should be very cautious about extrapolating the results," he says. "It would be more meaningful to have some results from the general population. They've made a big contribution to the scientific literature, but we need to recognize the limitations."

"One of the complexities of making dietary recommendations is that you have to eat something, and you're always replacing something," says Giovannucci. "Part of the issue is, what are you going to drink instead of the milk? Two glasses of milk are likely to be better than two glasses of Coca-Cola. In that context, milk is probably healthful."

"Even doing research in this area, I'm trying to be cautious about what recommendations to make, and about how concerned people should be," says Giovannucci. "I think certainly calcium should be much less of a concern than obesity and smoking and lack of exercise. We do need to find out whether, at least for some men, recommending calcium is doing more harm than good." He says that for men, 700-1000 mg of calcium per day is probably reasonable.

How the evidence is presented can be confusing to the public, says Clement Adebamowo, who studies milk consumption and acne (see "Milk and Human Health) in Nigeria. "I think it's important that we don't oversell the evidence or the story. At best, it is an evolving story."

"People shouldn't throw out all their milk," says Adebamowo. "Certainly they should continue to take dairy and milk, but in moderate quantities, not more than three glasses as some ads say. One family I visited with, the wrestling coach at their son's school, recommended intake of 100% milk protein - whey - formula. It's those kinds of things that are bothersome. We really don't know the long-term implications of those kinds of diets."