# News this Week

Science  10 Jun 2005:
Vol. 308, Issue 5728, pp. 1526
1. SCIENCE AND SOCIETY

# Smithsonian Gives Grudging OK to Film Backing ID Argument

1. Yudhijit Bhattacharjee

A film about the origins of the universe that makes a subtle argument for intelligent design (ID) has left the Smithsonian Institution's National Museum of Natural History with egg on its face. Museum officials say they are reluctantly hosting the upcoming event, even though it violates the museum's scientific and educational missions, because of an ironclad contract with the Discovery Institute, which is sponsoring the private screening. But after heavy criticism from its scientists and outsiders, the museum promises it won't happen again.

The controversy was triggered by a 26 May story in The New York Times that the Washington, D.C., museum would be co-hosting a film titled The Privileged Planet: The Search for Purpose in the Universe. The film is based on a book by Guillermo Gonzalez, an astronomer at Iowa State University in Ames, and Jay Richards, a philosopher at the Discovery Institute, the Seattle-based nonprofit organization that has been a leader of the ID movement. It presents findings to conclude that the suitability of Earth as a habitat for scientific observation is evidence that the universe was designed for human beings to discover its principles.

“It was a no-brainer,” says John Scofield, a Republican spokesperson for the spending committee that proposed a reshuffling of the VA's 2006 budget. Scofield was referring to the need to examine the growing incidence of post-traumatic stress disorder, substance abuse, and other serious mental health problems in soldiers returning from Iraq and Afghanistan. The committee chastised the VA for reportedly spending just 7% of its $784 million research budget on mental health and requested a jump to at least 20%. “We're not mandating” the increase, Scofield says, but the legislators warned VA officials that “any significant deviation from this goal shall be reported to the committee … with explanations as to why the goal was not met and remedies being put in place.” The House bill supplies only half of what the VA spends on research, all of it for direct research costs. (The rest, spent largely on salaries and overhead, is dispensed separately.) The legislators approved the level requested by the president,$393 million, which is down from $402 million this year. A VA spokesperson says about 10% of the agency's direct research funds goes to mental health. The Senate has yet to take up the 2006 VA spending bill, which passed the House on 26 May, and any differences must be reconciled before it becomes law. News of this proposed change swept through the VA community, fueling both apprehension and enthusiasm. “Not to invest this amount of money would frankly be pathetic,” says Thomas Horvath, a psychologist and chief of staff at the Michael E. DeBakey VA Medical Center in Houston, Texas, who says he advocated “frequently and loudly” for increased funding while overseeing mental health work in the VA's Washington headquarters from 1994 to 1999. But other VA researchers, including some in the mental health field, worry about congressional interference in funding priorities and the demand for rapid change. “Mental health is clearly an understudied, underfunded area in the VA,” says Alan Bellack, a psychologist at the Baltimore VA Medical Center and the University of Maryland School of Medicine. Still, he wonders whether the community is prepared to absorb such an increase and worries that other disciplines will be squeezed. The VA currently invests heavily in chronic diseases such as diabetes and cancer, traumatic injuries such as brain injury and amputation, and age-related problems such as dementia and Parkinson's disease, among other areas. The House proposal “will devastate VA research,” says a longtime VA scientist who requested anonymity. Legislators didn't specify what to cut. 3. PSYCHOLOGY # Survey Finds U.S. Mental Health Holds Steady 1. Constance Holden A nationwide psychological survey that mirrors one conducted in the early 1990s indicates that the mental health of Americans, which suffered a decades-long slide after World War II according to suicide rates and other statistics, hasn't gotten any worse over the past decade or so. Still, some 6% of the population at any given time have mental illnesses that are “seriously debilitating,” which makes the U.S. sicker psychologically than other developed nations, according to the survey conducted by the University of Michigan, Harvard University, and the National Institute of Mental Health (NIMH). Unlike physical ailments, which increase with age, neuro- psychiatric disorders generally hit young people, study directors noted at a press conference last week. As a result, said NIMH director Thomas Insel, mental illnesses are greater sources of disability and premature death than are chronic physical disorders. The new study, called the National Comorbidity Survey Replication, is based on household interviews of 9282 randomly selected adults in 35 states. Its earlier counterpart was the first to assess a nationally representative sample using standardized psychiatric terms. The new results, published this week in four papers in the Archives of General Psychiatry, reveal that over a lifetime, about 46% of the population falls prey to some sort of anxiety, mood, impulse-control, or substance-use disorder. And that's not counting complex psychiatric conditions such as schizophrenia, which are not amenable to a household survey. Many of the cases documented in the survey are mild, temporary, and never require treatment. However, noted Ronald Kessler of Harvard Medical School's Department of Health Care Policy, even mild cases may become more severe and “accumulate” if not treated early. Hence the high rates of comorbidity: 45% of the subjects diagnosed with one disorder also qualified for another. Depression and alcoholism go hand in hand, for example. The lag time between onset of a problem and treatment was 6 to 8 years for mood disorders and 9 to 23 years for anxiety disorders. “These new numbers raise the possibility that the stigma against treatment may be even greater than the stigma against the disorders themselves,” said Insel. Nonetheless, the past decade or so of mental health awareness campaigns and the availability of new drugs have paid off to some degree: 18% of those in the study reported getting some treatment in the prior year compared with 13% in the earlier survey. Still, the researchers found that less than one-third of those seeking help had “minimally adequate” care, as defined by guidelines agreed upon by groups such as the American Psychiatric Association. Lower education levels correspond with poorer mental health, but both blacks and Hispanics, who tended to have less education than the white people surveyed, reported less anxiety and depression. “Minorities in minority communities have particularly low rates,” said Kessler, who speculated that they have a “sense of belongingness that many other people don't have.” Despite continued inadequate treatment and long lag times in seeking help, there's a “sea change” occurring in the nation's mental health, added Kessler: “This is the first time we've been able to say there has not been a rise in mental disorders.” 4. PROTEOMICS # Canadian Global Database May Move to Singapore After Loss of Funding 1. Wayne Kondro 1. Wayne Kondro is a freelance writer in Ottawa. OTTAWA, CANADA—The world's largest free repository for proteomics data appears headed to Singapore from Toronto, barring an 11th-hour reprieve by Canadian funding authorities. At stake is the fate of the Biomolecular Interaction Network Database (BIND), which since its inception in 1999 has received$15.5 million from various Canadian agencies. It's an online database containing details of nearly 180,000 molecular interactions submitted by scientists from around the world. Last month its parent group, the Mount Sinai Hospital-based Blueprint Initiative, was forced to lay off half its 68 staffers, and its Canadian bank account will run dry on 30 June. Its future appears to lie in Singapore, which is providing $18.4 million over 5 years starting last summer for a nascent version of the database and has promised more in return for housing the entire database, says principal investigator Christopher Hogue. Hogue's troubles began after he asked Genome Canada earlier this year for$20.8 million over 4 years to continue running the database. The nonprofit corporation voted thumbs-down, citing what Genome Canada president Martin Godbout says were problems with its “management, budget justification, and financial plan.” Hogue says all of those problems—in particular, the requirement that BIND secure matching funding—stem from an unfortunate set of circumstances beyond his control. In particular, he says that a grant of nearly $10 million from an Ontario provincial program has been delayed because of a revamping of the program. Unannounced “rule changes,” he adds, precluded him from counting a component of the grant from the Economic Development Board of Singapore to create Blueprint Asia, a Singapore-based component of BIND, ostensibly on the grounds that international contributions aren't de rigueur. Godbout disagrees with Hogue's analysis. “No project failed the test only because of cofunding,” he says. The upshot, though, is that Hogue has been left scrambling for alternative resources to keep his remaining staff members from joining 33 ex-colleagues on the unemployment lines. And his prospects look no better than those of professional hockey resolving a contract dispute between owners and players and resuming play before next fall. Getting a second shot at funding from Genome Canada would first require the corporation to receive money next spring to hold a new competition. And although the revamped Ontario Research Fund last week issued a call for proposals, the deadline for submissions is not until 14 October. In addition, the fund cannot contribute more than one-third of the overall cost of a project. For BIND, that means other cash-strapped federal agencies would need to chip in to make up the difference. A move to Asia would compromise the country's nascent biotechnology sector, says Hogue, as well as its reputation as a reliable contributor to international science ventures. If no one steps forward, Canada will lose both trained bioinformatics experts and the scientific prestige that goes with hosting a global project, adds Francis Ouellette, director of the bioinformatics facility at the University of British Columbia in Vancouver. “It's sort of a waste of money to start a project and then basically throw it out and let somebody else reap the benefits,” he says. Hogue and others say that BIND's situation also points up the need for greater support for long-lived public data collections (Science, 8 April, p. 187). In the meantime, Hogue is unhappily mulling life in Asia and the fine print of proposed contracts with various Singapore agencies. “I'm a farm boy from Windsor, Ontario. It is my intent to stay in Canada,” he says. But his tone changes when he puts on his BIND administrator's hat. “I can't keep a global database operating without funds,” he says. “And I'm not going to shut the database down because, A, it's successful and, B, it's a much-needed resource.” 5. CLIMATE CHANGE # California Sets Goals for Cutting Greenhouse Gases 1. Erik Stokstad With the Bush Administration still looking for additional scientific evidence on climate change, states have led the way in proposing ways to reduce further warming. Last week, the high-profile Republican governor of the most populous state in the country weighed in, offering ambitious targets for curbing the state's emissions of greenhouse gases. “I say the debate is over,” Arnold Schwarzenegger announced at an annual world environmental festival. “We know the science. We see the threat. And we know the time for action is now.” California exerts a huge impact on the global environment. Its economy is the sixth largest in the world, and the state is the 10th largest emitter of greenhouse gases on the planet. “This is a potentially major political step,” says climatologist Stephen Schneider of Stanford University. He and others say the move could spur further action by other states—several in the Northeast are hammering out cap-and-trade systems, for example—and rekindle hopes abroad that the United States might eventually fall in line with the rest of the world on its policies to combat global warming. Speaking on 2 June in San Francisco at United Nations World Environment Day, Schwarzenegger argued for reductions of greenhouse gases on economic grounds. He cited threats such as the likelihood of reduced water supplies, rising sea level, and more agricultural pests. He also pointed to opportunities for state businesses to develop more environment-friendly technology. “It sends a real signal that … action on climate change is essential to maintaining a strong economy,” says Alden Meyer of the Union of Concerned Scientists. California companies might be able to trade emission credits with countries of the European Union, which has begun a cap-and-trade system. Although short on details, the executive order lays out three ambitious targets. It calls for lowering emissions to 2000 levels by 2010 and to 1990 levels by 2020. By 2050, the state's emissions would be 80% below the 1990 levels. The short-term targets are not as aggressive as those of the Kyoto treaty but are equivalent to a bill reintroduced last month by U.S. Senators John McCain (R-AZ) and Joseph Lieberman (D-CT). Although Schwarzenegger's executive order didn't mention how to achieve those reductions, he cited 2004 state regulations that require lower emissions from vehicles (which may become stalled in a court battle) and advancing the timetable to 2010 for generating 20% of the state's power from solar, wind, and other renewable sources. Michael Oppenheimer, an atmospheric scientist at Princeton University in New Jersey, says it would be feasible to achieve the 2010 target (which represents an 11% cut of emissions from today's levels) and the 2020 target (a 25% cut) by quickly adopting such green efforts. The deeper reductions by 2050 may require a cap-and-trade system for greenhouse gases similar to the one implemented by the European Union, he notes. Schwarzenegger has asked the state Environmental Protection Agency to examine options for such a system and report back in January. 6. EUROPEAN UNION # Researchers Lobby to Head Off Threatened Cuts 1. Gretchen Vogel BERLIN—High hopes among European researchers are turning to worry as political battles threaten to scuttle a planned budget boost and mar the launch of the long-sought European Research Council (ERC). In April, the European Commission proposed a doubling of the E.U.'s research budget, to €70 billion ($86 billion) between 2007 and 2013 (Science, 15 April, p. 342). The plan included €12 billion for a new ERC, which would fund basic research across Europe.

But political tussles over member countries' contributions are threatening to shrink the whole of the commission's proposed €1.03 trillion budget by at least €150 billion. In a proposal put forward on 28 May, Luxembourg's Prime Minister Jean-Claude Juncker, whose country currently holds the E.U. presidency, said the main cuts would come from research programs as well as “structural funds,” which build roads and other infrastructure.

“It's very serious,” says Helga Nowotny of the Science Center Vienna, who is head of the European Research Advisory Board. Although ERC would still go forward even without the doubling, Nowotny says, a severely reduced budget will diminish its impact. Nowotny and her colleagues sent a letter on 6 June to more than 100 scientific and industrial leaders to lobby their governments to fund the full research proposal. The letter urges recipients to point out “the contradiction between what governments say in favor of research and how they act.”

E.U. Commissioner for Research Janez Potocnik says the financial decisions will be a “moment of truth for the E.U.” Potocnik was in Berlin on 2 June to try to persuade German leaders—some of the main holdouts in the budget battles—of the importance of research in the E.U. He told Science that European politicians say repeatedly that research and innovation should be the highest priority. But protecting subsidies and capping national contributions “turn out to have slightly higher priority.”

Some researchers are also concerned about an initial plan for the ERC circulated among the heads of European research councils at a meeting last month in Reykjavik, Iceland, says Ernst-Ludwig Winnacker, head of the German Research Foundation, the DFG. The plan seems to shift power away from a council of independent scientists to the staff of an “executive agency” who answer to the European Commission. Potocnik, however, says the worries are misplaced. All issues of substance, he says “will be decided by the scientific council. The commission will sign off” on the council's decisions.

“It's a matter of trust,” Nowotny adds. “Legally it is not possible to give €1 billion to a group of people who have not been elected or even appointed. It must be the commission who takes the ultimate responsibility. But the commissioner has always said he will be the guarantor for the autonomy of the ERC.” European scientists will be sure to remind him to keep his word.

7. MICROBIOLOGY

# Mosquito-Killing Fungi May Join the Battle Against Malaria

1. Martin Enserink

They already kill insects in fields, greenhouses, and gardens around the world. Now, a duo of fungi may also become a new weapon in the fight against malaria. In this issue of Science, two research groups report the results of lab experiments and field tests in Tanzania indicating that fungal spores can infect and kill adult Anopheles mosquitoes, the vectors of malaria parasites. Applied just like chemical pesticides, sprays containing the spores could be a new, environmentally friendly weapon against malaria, the researchers say.

“They have a pretty strong case,” says Christiaan Kooyman, who studies locust control using fungi at the International Institute of Tropical Agriculture in Cotonou, Benin. New control tools are necessary, Kooyman adds, because mosquitoes are increasingly becoming resistant to chemical pesticides. But whether the fungal strategy is technically or economically feasible remains to be seen, others caution. “I have seen plenty of false technological dawns” in vector control, says Jo Lines of the London School of Hygiene and Tropical Medicine.

That strains of the two fungi—called Beauveria bassiana and Metarhizium anisopliae—can kill mosquitoes didn't come as a surprise. Both species are used in agricultural biopesticide products, and so many different strains of each fungus exist that there's probably one to kill almost any insect species, Kooyman says. But no one had set such fungi loose on malaria mosquitoes until recently.

In 2003, one group, led by Bart Knols of Wageningen University and Research Centre in the Netherlands and the International Atomic Energy Agency in Vienna, published a lab study showing that spores of several fungi infected Anopheles gambiae when applied directly to the insects' bodies. Whereas pesticides kill overnight, these fungi grow slowly, often taking 10 or 12 days to kill.

As they report on page 1641, Knols's team has now tested this idea in the field. They suspended 3-m2 cloths impregnated with the fungus M. anisopliae from the ceilings of five traditional houses in a rural Tanzanian village, collected mosquitoes in the homes for 3 weeks, and kept the insects alive on glucose. Some 23% of female Anopheles gambiae mosquitoes became infected, shortening average life span by 4 to 6 days compared to controls from five untreated homes.

The study was much too small to detect an effect on malaria transmission and not designed to do so. But when the team modeled how such results would alter malaria transmission in a village if the cloths were applied year-round, they found that the number of infective bites for the average villager would fall from 262 to 64 annually. In order to make a dent in malaria cases and deaths, that number has to come down much more, to close to one bite per year. But that is feasible by upping the dose and spraying entire walls in many more houses, says Knols's collaborator Kija Ng'habi of the Ifakara Health Research and Development Centre in Tanzania.

On page 1638, a team led by Andrew Read of the University of Edinburgh and Matt Thomas of Imperial College London reports that the true effect of a fungus—in their case, B. bassiana—on malaria transmission may be even more pronounced than Knols's data suggest. In lab studies using Plasmodium chabaudi, a rodent malaria parasite, and a mosquito species called Anopheles stephensi, the group found that even in surviving mosquitoes, the fungus severely hampered the parasites' ability to develop and mature. “That looks like an important extra benefit,” says Wendy Gelernter, a biopesticide consultant at PACE, a company in San Diego, California. In addition, both teams have data suggesting that a fungal infection dampens mosquitoes' appetite for blood meals, making them less likely to pick up parasites in the first place.

Ken Neethling, production director for BCP, a South African company specializing in biopesticides, says his firm may explore the malaria biocontrol strategy commercially; others are interested as well, Thomas says. For now, both teams plan to tinker with the sprays' formulations to see if they can improve infection rates. One key problem: The spores start losing their infectiousness in a matter of weeks. If that can't be solved, the spray would have to be applied over and over. (Pesticides, in contrast, can last a year or longer.) That could be “a near-fatal flaw,” says Lines.

Still, these are problems well worth delving into, says Norbert Becker of the German Mosquito Control Association in Waldsee, Germany. As long as malaria kills more than a million people every year, he says, “every new strategy is appreciated.”

8. GENETICS

# In Voles, a Little Extra DNA Makes for Faithful Mates

1. Elizabeth Pennisi

Prairie voles are renowned for being faithful mates, but some individuals are more faithful than others. The difference may lie in their so-called junk DNA.

On page 1630, Elizabeth Hammock and Lawrence Young of Emory University in Atlanta, Georgia, report that fidelity and other social behaviors in male prairie voles seem to depend on the length of a particular genetic sequence in a stretch of DNA between their genes. The longer this repetitive sequence, or microsatellite, the more attentive males were to their female partner and their offspring. Those with shorter microsatellites neglected their mates and pups, at least to some degree.

Although there's no evidence that human infidelity or poor parenting stems from similar variations, Hammock and Young, as well as other researchers, have begun to explore whether microsatellites can account for behavioral differences between people and primates such as chimps and bonobos. The new study's results “will force us to think about these variations in so-called junk DNA and how [they] make for changes in behavior,” says Scott Young (who is not related to Lawrence Young), a neuroscientist at the National Institute of Mental Health in Rockville, Maryland.

Microsatellites are genetic stutters, usually just two or four bases long. There can be hundreds of these repeats in a row. They can befuddle the cell's DNA replication machinery, so the number of repeats within one may rise or fall from one generation to the next. And when they are in regulatory regions for genes, their changing lengths may affect the activity of those genes. This can have rapid evolutionary implications, Scott Young points out.

In the mid-1990s, researchers discovered a key microsatellite difference between prairie voles and their more promiscuous cousins, such as the meadow voles. Prairie voles have longer microsatellites near the gene encoding a receptor (V1aR) for the brain chemical vasopressin, and as a result they make more of the receptor than do meadow voles. This was the first clue that these sequences may influence social behavior. Last year, Young's team strengthened the connection when they caused meadow voles to emulate the faithful ways of prairie voles by adding extra copies of the V1aR gene to a portion of their brains (Science, 7 January, p. 30). “The vasopressin system is likely to be a major player in emotional and cognitive aspects of social bonding,” comments Rainer Landgraf, a neuroscientist at the Max Planck Institute of Psychiatry in Munich, Germany.

Now, Young and Hammock, originally one of Young's graduate students and now at Vanderbilt University in Nashville, Tennessee, have found that variations in V1aR-associated microsatellites among individual prairie voles influence expression of the gene and overall behavior. They paired and bred voles with long microsatellites and found that the resulting males spend more time licking and grooming their pups than did males with short microsatellites. They also placed males in cages with a female, allowing 18 hours for them to bond, then added a new female. Males with longer microsatellites spent more time with their partners than did those with shorter microsatellites. Taken together, the results “help create a picture of some of the building blocks that allow for the evolution of different levels of social behavior,” says Catherine Marler of the University of Wisconsin, Madison.

Evan Balaban, a neuroscientist at McGill University in Montreal, Canada, isn't convinced, however. He argues that, instead of simply showing correlations between microsatellite length and a behavior, the researchers should do transgenic experiments to establish that microsatellites were truly responsible for the different behaviors. Furthermore, “the behavioral effects are small,” Balaban adds.

Undeterred, Hammock and Young have already noted connections between V1aR microsatellites and primate behavior. Other researchers have associated the length of one of the four microsatellites in the human version of the gene with autism, a disorder of social interactions. In the chimp, this same microsatellite is 360 bases shorter, Hammock and Young note. But in bonobos, which are less aggressive than chimps and form more humanlike social bonds, the microsatellite is nearly identical to the human counterpart.

Even Balaban thinks such intriguing observations deserve follow-up. “Hopefully,” he says, “[this will] direct people's attention to studying the role that variation in the control of the regulation of genes plays.”

9. CELL BIOLOGY

1. Gretchen Vogel

At least one company says it is almost ready to try using human embryonic stem cells in patients. But several hurdles remain

Shortly before Congressman James Langevin cast his vote last month to relax federal rules on funding of stem cell research, the Rhode Island Democrat told his colleagues, “I believe one day I will walk again.” Langevin, who has been paralyzed since a gun accident at age 16, pleaded with his colleagues to vote with him. “Stem cell research gives us hope and a reason to believe. … We have a historic opportunity to make a difference for millions of Americans.”

With impassioned pleas like this, high-stakes battles in Congress, and billions of private and state dollars pouring into research on human embryonic stem (hES) cells, it often seems their therapeutic applications must be just around the corner. But a careful parsing of the claims from even the strongest advocates reveals the caveat “someday.”

How soon that someday might arrive is far from clear. Scientists are nearly unanimous that the study of hES cells will illuminate human development and disease. But whether the cells will actually be used to cure patients like Langevin is less certain. Cell therapies are more complicated than drugs, and hES cells, which have the potential to become any cell type in the body, carry special risks.

“The most sobering thing about [hES] cells is their power,” says neuroscientist Clive Svendsen of the University of Wisconsin, Madison, who works with both fetal and embryonic stem cells. The extreme flexibility and capacity for growth characteristic of ES cells makes them ideal for producing large quantities of therapeutic cells to treat, say, diabetes or spinal cord injuries. But these same traits also increase the risk that renegade cells could, as they have in animal studies, cause unwanted side effects, ending up in the wrong place or even sparking cancerous growth. “You have to learn to control that power in the dish” before thinking about putting the cells into patients, says Svendsen.

For that reason, most groups say they are at least five or, more likely, 10 years away from clinical trials. But one company is challenging that timeline. Geron in Menlo Park, California, says its animal studies suggest that stem cell therapy can be safe and might be effective for a select group of patients. The company hopes to start clinical trials of hES cells to treat spinal cord injuries as early as summer 2006. Already, the company is in discussions with the Food and Drug Administration (FDA), which is attempting to set safety standards for the field. Potential treatments with human ES cells face the same difficulties as all cell therapies, notes Malcolm Moos of FDA's division of cellular and gene therapies: There are few standardized techniques to measure the purity or potency of a cell population that would be delivered to a patient.

Most stem cell researchers view Geron's plans with hefty skepticism and caution that a premature rush to patients could seriously damage the already-controversial field. And it is far from clear whether FDA will allow the trial to proceed. But Geron, which funded the researchers who isolated the first hES cells in 1998, has several reasons to push ahead; the company holds a number of patents and exclusive licenses that give it more freedom—and more incentive—to develop possible products from hES cells. And whatever the outcome, scientists agree, Geron's ambitious plans will offer a test case of the hurdles scientists will have to overcome to prove that hES therapies are both safe and effective.

Even the skeptics say Geron chose a plausible target for the first trial, as spinal cord injuries may be significantly easier to tackle than diseases such as diabetes or Parkinson's (see sidebar, p. 1536). The trials would be based on work led by Hans Keirstead, a neuroscientist at the University of California, Irvine, who proved a persuasive spokesperson for the field during the campaign for California's Proposition 71, which provides $3 billion in funding for hES cell research. During last fall's campaign, Keirstead described his then-unpublished work, showing videos of rats with spinal cord injuries that had regained some mobility after injections of cells derived from hES cells. “I am extremely enthusiastic,” Keirstead says. “I am past the point of hope. In my mind the question is when. What we are seeing in these animal models is tremendous.” Keirstead and his colleagues, with funding and technical support from Geron, have developed a protocol that encourages hES cells to differentiate into cells called oligodendrocyte precursors. These cells can form oligodendrocytes, the cells that, among other functions, produce the protective myelin sheath that allows neurons to send signals along their axons. This sheath is often lost during spinal cord injuries. In a paper last month in the Journal of Neuroscience, Keirstead's team reported that these precursors, when injected into the spinal cord, could help improve recovery of rats that had suffered spinal cord injury. The cells aren't replacing injured neurons, Keirstead says, but are encouraging the natural healing process, presumably by restoring some of the myelination. Earlier studies in mice (Science, 30 July 1999, p. 754) showed that injecting mouse cells destined to form oligodendrocytes into injured or diseased animals could restore some myelination; Keirstead's team is the first to show that human ES cells can have similar effects. For newly injured rats, the results are promising. In animals that received oligodendrocyte precursors 7 days after their injury, the cells survived and apparently helped repair the spinal cord's myelin. Within 2 weeks, treated rats scored significantly better on standardized movement tests than control animals, which had received human fibroblasts or a cell-free injection. But when the researchers injected cells 10 months after the injury, they saw no effect—sobering news for people like Langevin suffering from old injuries. The cells survived but were apparently unable to repair the long-term damage. For that reason, Keirstead says, Geron's proposed clinical trial would target newly injured patients. The phase I trial, if it goes forward, will probably include only a handful of patients and, most importantly, Keirstead emphasizes, will not cure anyone. Its primary goal is to show that the treatment can be safe. “The public and scientists must realize that these are the first attempts,” Keirstead says. “No one is expecting them to cure. We are expecting them to treat, but we have no idea what the level of response is going to be.” ## Potential peril Proving safety is a tall enough order. In numerous animal studies, ES cells from mice and humans have proved difficult to control, differentiating into the wrong kind of cell, for instance, or migrating away from the injection site. In its spinal cord trial, Geron plans to inject ES-derived cells that can form just a single cell type, an approach that may circumvent some of these problems. For a full recovery, patients are likely to need new neurons as well as other support cells called astrocytes, but using precursors that differentiate into all three types of nerve cells can be problematic. In several rodent studies, partially differentiated mouse ES cells injected into the spinal cord have formed neurons, astrocytes, and oligodendrocytes and have helped animals recover from spinal cord injuries. But more recently, neural stem cells derived from adult animals—which also differentiate into the three cell types—have caused problems. As Christoph Hofstetter of the Karolinska Institute in Stockholm, Sweden, and his colleagues reported in Nature Neuroscience in March, neural stem cell treatments led to some recovery in rats' paralyzed hind legs, but the animals also developed a chronic pain sensitivity in their forelegs, which had been unaffected by the injury. In other experiments, preventing the formation of astrocytes seemed to eliminate the side effect, highlighting the importance of proper differentiation, Svendsen says. Perhaps the biggest worry is that hES therapies will spur tumor formation. One of the defining characteristics of ES cells is that they form disorganized tumors, called teratomas, when injected in undifferentiated form under the skin of immune-compromised mice. “The ES cell is basically a tumor-forming cell,” says neuroscientist Anders Bjorklund of Lund University in Sweden. “This aspect has to be dealt with seriously before the cells are applied in the clinic.” Even a benign tumor in the central nervous system would be serious, says Svendsen: “Any sort of growth in the spinal cord is not good news.” But Keirstead believes he has solved those problems. The key, he says, is a differentiation procedure that he claims produces cell populations in which 97% of cells express genes typical of oligodendrocyte precursors. “Teratomas are a real possibility if you put in naïve stem cells,” he acknowledges. “But that is the science of yesteryear. No one is even considering putting in any naïve ES cells.” Keirstead and his colleagues say in their paper that they found no evidence that their specialized cells formed astrocytes or neurons after injection. The team is also checking whether any of the injected cells leave the spinal cord. So far, Keirstead says, they seem to stay close to the site of injection. Keirstead's paper is promising, Svendsen says, but he's not convinced the work is ready for patients. “It didn't go into the detail you'd like to see before a clinical trial,” he says. The catch is that it's hard to be sure that a population of several million cells is free of any undifferentiated stragglers. To evaluate the risk of tumors, Keirstead and his colleagues are testing the differentiated cells in nude mice: animals bred to lack an immune system. If the animals live for a year without signs of teratomas, then Keirstead says he will feel confident that the cells are safe to try in humans. Several teams are making headway addressing another problem: possible animal contamination. To date, almost all human ES cell lines have been exposed to animal products. Cultured cells are often kept alive with fetal calf serum, for instance, and most hES cell lines have been grown on layers of mouse cells called feeder cells, which provide the key proteins that prevent ES cells from differentiating. These techniques have sparked worries that hES cell therapies could introduce exotic animal viruses into patients. In response, several teams, including Geron, have recently developed ways to grow new cell lines either on human feeder layers or without feeder cells at all. But the older cell lines have the advantage of being better characterized, says Geron CEO Thomas Okarma. That's why the company plans to use one of the original lines derived by James Thomson of the University of Wisconsin, Madison, in its first clinical trial. To reduce the risk of contamination, the company has been growing these cells for more than a year without any feeder cells. That may suffice for FDA, which has said that past exposure to animal cells does not disqualify ES cell lines from clinical use as long as certain safety standards are met. Okarma says Geron can demonstrate that its cells are uncontaminated. His claim is bolstered by a paper by another group published last week in Stem Cells. Joseph Itskovitz-Eldor of Technion-Israel Institute of Technology in Haifa and his colleagues tested five hES cell lines and several cultures of mouse feeder cells for signs of murine retroviruses, which lurk in the genome of all mouse cells. Although the team identified receptors for the so-called mouse leukemia viruses, they found no evidence that the virus had infected any of the human cells, even after growing on mouse feeders for years. Animal products still may pose a risk, says Itskovitz-Eldor. But the new work shows that “the cells can be tested, and we believe it will be possible to use them clinically.” More recently, researchers identified another potential downside to using mouse feeder cells. In February, Fred Gage and his colleagues at the Salk Institute for Biological Studies in La Jolla, California, reported that hES cells grown with mouse feeders expressed a foreign sugar molecule on their cell surface. Because humans carry antibodies to the molecule, the researchers suggested that it might tag the cells for destruction by the human immune system. If so, then any therapy created with existing cell lines was unlikely to succeed. But Keirstead, Okarma, and others now say that those concerns, widely reported, may have been overstated. Gage and his colleagues noted that the sugar gradually disappears once cells are removed from the feeder layers. Keirstead says that once cells are removed from mouse feeder layers for several months, the sugar disappears. Okarma adds that cells in Geron's feeder-free cultures have no sign of the foreign molecule. Finally, some scientists worry that ES cells might acquire harmful new mutations in culture, a common phenomenon with almost all cultured cells. Although ES cells “are probably 100 times more stable than adult stem cells in culture, they're not perfect,” cautions Mahendra Rao of the National Institute on Aging in Baltimore, Maryland. Such mutations would be particularly hard to detect ahead of time. ## Blazing a trail FDA, meanwhile, is trying to set safety standards for this burgeoning field. The agency announced in 2000 that cell therapies involving stem cells from embryos or adults would be regulated as drugs, not as surgical techniques. That means that researchers will have to meet certain standards of purity and potency. For most drugs, those standards are straightforward to set and easy to measure. Cellular products are much more complicated, says Moos, and it is less clear what sorts of measures will be used to evaluate cell populations. Geron is already working, with FDA advice, to gauge the risk of tumors. In ongoing studies, Okarma says, Geron scientists spike the differentiated cells with known amounts of undifferentiated hES cells to determine the threshold level that produces a teratoma in nude mice. He adds that the company has developed “extremely sensitive” assays to detect undifferentiated cells that could ensure cell preparations don't exceed that level. One of the major questions FDA must decide is whether cell therapies need to be tested in nonhuman primates before they enter humans. Keirstead is not convinced that transplanting cells into primates is a prerequisite for safety. “The question is, are we going to learn any more from putting human cells into monkeys than putting human cells into rats? I'm not sure we know.” FDA's Moos agrees: “Nonhuman primates are not the default choice and in fact are seldom the best choice.” For now, Moos says FDA will evaluate this issue, and others, on a case-by-case basis. Keirstead hopes that these efforts will ease the way for subsequent groups. The first clinical trial will leave behind a set of standards for other teams, he says: “Once you've done that, you've paved the road. You're no longer in 4 × 4s hacking your way through the jungle.” And even if the initial trial doesn't work, Keirstead believes it will advance the field. “An unsuccessful trial is extraordinarily important for moving forward. It's not something that patients and the media want to hear. But it's reality. An unsuccessful trial is scrutinized as much as a successful one.” Yet he remains stubbornly optimistic. “We can't eliminate the risk. But I can sleep at night because I do feel the risk is low and the chance of success is high.” Still, many scientists worry that Geron is moving too fast. They point to gene-therapy trials in which one young patient died of an unexpected immune reaction and others developed deadly leukemia. “Gene therapy is a paradigm that we can learn from,” says neuroscientist Douglas Kerr of Johns Hopkins University in Baltimore, Maryland. “They have actually induced harm in patients, and that has set back the field.” Others worry that the high-profile political debates may have already set expectations for the field too high, especially among patients who face devastating diseases. Honesty, not hype, is key, says stem cell researcher Bernat Soria of Miguel Hernández University in Alicante, Spain. “My experience is that when you are honest with patients, they are clever enough. The patients tell me, ‘We are not sure that there will be a solution for my disease, but please do the research anyway.’ People are aware that the road is going to be long. 10. CELL BIOLOGY # Still Waiting Their Turn 1. Gretchen Vogel Even enthusiasts agree that Geron's goal—to begin testing a human embryonic stem (hES) cell therapy in patients with spinal cord injury within a year—is a long shot. Prospects are more distant for using stem cells to treat other diseases, such as diabetes, Parkinson's disease, amyotrophic lateral sclerosis (ALS), and multiple sclerosis (MS). None is likely to reach the clinic for at least 5 or 10 years, most scientists in the field agree. And that's assuming abundant funding and faster-than-expected scientific progress. Some of the strongest advocates for hES cell research are those hoping to find a cure for type 1 diabetes. The driving force behind California's Proposition 71, Robert Klein, says, for example, that his primary motivation is to find a cure for his diabetic son. Diabetes kills the pancreas's β cells, which regulate the amount of insulin in the blood. Patients have to take frequent insulin injections and face many complications, including kidney failure and blindness. Replacing the missing cells could cure the disease. Initial trials using β-cell transplants from cadavers have shown promise, but side effects and the transplants' limited life span has dampened enthusiasm (Science, 1 October 2004, p. 34). And even if the therapy worked perfectly, each transplant requires cells from multiple cadavers. So researchers are looking for renewable sources of cells that could treat the millions of patients who might benefit. In theory, hES cells fit the bill nicely. In practice, however, although several groups have managed to coax mouse ES cells to differentiate into cells that make insulin, no one has yet managed to derive bona fide β cells from either mouse or human ES cells. One reason may be that unlike nerve cells or heart muscle cells, pancreatic cells are some of the last to develop during pregnancy. In mice, the cells appear on day 15 or 16, just a day or two before birth, and in humans, they appear in the 5th or 6th month. “If the road is longer, the possibility of getting lost is much higher,” explains Bernat Soria of Miguel Hernández University in Alicante, Spain, who has tried to produce β-like cells from both mouse and human ES cells. Fortunately, says Soria, the cells may not have to be perfect; several types of insulin-producing cells have helped alleviate diabetes symptoms in mice. But there is no leeway when it comes to safety. Diabetes is a chronic but not inevitably deadly disease, so any cell therapy must be safer and more effective than insulin shots. “We don't have a cure, but we have a treatment,” Soria says. “Despite the strong pressure we have from patients and families, the need for cell therapy is not as strong.” Scientists have already attempted to use cell therapies to treat Parkinson's disease, which attacks neurons in the brain that produce the neurotransmitter dopamine, leaving patients increasingly unable to move. In a handful of clinical trials in the last decade, physicians implanted dopamine-producing cells from fetal tissue—with decidedly mixed results. Whereas some patients showed significant improvement, others showed little or none. And some developed serious side effects including uncontrollable jerky movements. Scientists aren't yet sure what went wrong, although some suspect that patients may have received either too many or too few fetal cells, which are difficult to characterize in the lab. Dopamine-producing neurons derived from ES cells could provide an unlimited and well-characterized source of cells. And a trial in monkeys from a team at Kyoto University found that dopamine-producing neurons grown from monkey ES cells could improve animals' symptoms. But before ES-derived cells are tested in Parkinson's patients, scientists need to understand more about how the transplanted cells are behaving in the brain, says neuroscientist Anders Bjorklund of Lund University in Sweden. “The knowledge is just not good enough yet to justify any clinical trials” with hES cells, he says. Patients and doctors facing the nightmare of ALS may be willing to accept higher risks associated with early hES cell treatments. There is no effective treatment for this invariably fatal disease that kills motor neurons, and patients usually die within 5 years of a diagnosis. But “ALS is an order of magnitude harder than other diseases” to treat with cell therapy, says motor disease specialist Douglas Kerr of Johns Hopkins University in Baltimore, Maryland. Doctors still aren't sure what causes the disease, and even if scientists could coax stem cells to replace the lost motor neurons—“a pretty tall order,” Kerr says—any new neurons could be subject to the same deadly assault. More promising, he says, would be a cell or a mixture of cells that might somehow help slow the damage, but no one is sure what that might look like. Treating MS has similar challenges, says Hans Keirstead of the University of California, Irvine, who is working with Geron on its possible spinal cord injury trial. “We're much farther away from treating MS with stem cells,” he says. Like spinal cord injuries, the disease attacks the myelin sheath around nerve cells, and injected oligodendrocyte precursors have shown positive effects in animal models. But the human situation is more complicated, Keirstead says. Nerves damaged by MS are already surrounded by oligodendrocyte precursors, but something stops the cells from working. Indeed, Keirstead, who is relentlessly optimistic about the prospects of helping spinal cord injury patients, sounds much more sober about the prospects for other patients. “When I look at the work with Parkinson's, MS, and stroke, I think spinal cord injuries are very amenable to these strategies. The rest of the central nervous system is not.” 11. CELL BIOLOGY # Human Embryonic Stem Cells May Be Toxicology's New Best Friends 1. Gunjan Sinha 1. Gunjan Sinha is a freelance writer based in Frankfurt, Germany. Despite their glamorous reputation, human embryonic stem (hES) cells have a long way to go to prove worthy in treating disease (see main text). In fact, the most immediate medical benefit of these controversial cells could be better toxicology studies. Existing toxicity assays, performed on animals or immortal cell lines generated from animals, often poorly reflect human physiology. For example, chemicals or drugs tested on rodent livers produce the same toxicity in people only about 50% of the time, says Raimund Strehl of Cellartis, a Swedish biotech based in Göteborg. Animal research has also drawn the ire of protesters, particularly in Europe, where some countries, such as Germany, have made the general protection of animals an explicit part of their constitutions. As a result, pharmaceutical companies are looking for alternatives. Testing candidate drugs on, say, human liver or heart cells might help firms better weed out false leads or identify dangerous compounds before they enter clinical testing. To that end, some toxicologists already use immortal human cell lines or cells taken from cadavers, but both have shortcomings; immortal cells proliferate abnormally, and cadaver cells deteriorate in culture, for example. Human ES cells might offer a solution, but few published studies have compared these cells, or cells derived from them, to existing assays. “The utility of human ES cells as models for toxicology studies remains speculative, although there is real potential there,” says James Battey, who heads the Stem Cell Task Force at the National Institutes of Health in Bethesda, Maryland. The search for stem cell-based assays is particularly hot in Europe, fueled by the possibility of E.U. legislation that would require toxicity testing of some 30,000 existing chemicals—and substantially increase animal use. Several companies are exploring the use of undifferentiated and differentiated human or animal ES cells, says Thomas Hartung, head of the European Centre for the Validation of Alternative Methods (ECVAM), part of the European Commission, which is increasing its support of this work. The center has already validated one mouse ES cell assay developed by a German consortium, and Hartung says this and other mouse ES cell protocols could be adapted to hES cells as well. At least several companies are forging ahead on hES systems. Cellartis, which has established 30 different hES cell lines, is now collaborating with ECVAM scientists to turn such cells into cardiomyocytes—a much-sought-after cell in toxicology research. Geron in Menlo Park, California, has already derived human hepatocytes from its stem cell lines. The company is working with partners to further develop assays and intends to sell the liver cells in 2006 for toxicology studies, says David Greenwood, Geron's chief financial officer. And just last month, James Thomson of the University of Wisconsin, Madison, who originally isolated hES cells, announced that he and two colleagues were starting a company to generate hES cell-derived heart cells for drug testing. If any of these efforts succeed, hES cells may well help make drugs safer for patients—and spare some animals—long before they are used directly in the clinic. 12. VIROLOGY # Culture Systems for Hepatitis C Virus in Sight at Last 1. Jon Cohen The inability to grow this widespread pathogen in laboratories has delayed the development of more-effective drugs and a vaccine Since the discovery of hepatitis C virus (HCV) in 1988, researchers have made remarkable headway against this liver-destroying pathogen, which infects a staggering 170 million people around the world. Scientists have delineated in fine detail the interaction between HCV, the liver cells it targets, and the immune system. In many people, the drug cocktail of interferon-α and ribavirin eliminates the virus. Studies have also clarified that there are six major genetic families of HCV and that current treatments work better against some so-called genotypes than others. But a fundamental roadblock has stymied scientific progress: HCV has stubbornly refused to grow in laboratory cell cultures—until now. Two labs using overlapping but unique approaches this week published evidence online of cell culture systems that can grow relatively high levels of HCV; the virus, in turn, can establish new infections. HCV researchers are ecstatic. “This is really a great advance,” says Michael Gale Jr., who studies HCV at the University of Texas Southwestern Medical Center in Dallas. Gale notes that the new culture systems will finally enable researchers to study critical aspects of the viral life cycle such as cell entry, replication, and packaging into new virus particles, “each of which presents a novel drug target,” says Gale. More precise targets, in turn, may yield more effective drugs than interferon-α and ribavirin, which work by unclear, nonspecific mechanisms and fail in a substantial fraction of patients. They are expensive, too, and require a year of injections, so they are of little use in developing countries, where the disease is most prevalent. The culture systems also promise a better way to judge the power of experimental vaccines. This accomplishment is the result of a friendly but fierce competition between the two groups—one led by Frank Chisari of the Scripps Research Institute in La Jolla, California, and the other by Charles Rice of Rockefeller University in New York City—who shared many critical materials and ended up crossing the finish line neck and neck. As the Chisari team describes in the 6 June Proceedings of the National Academy of Sciences (PNAS) Early Edition, it designed the culture system by using an unusual isolate of HCV and manipulating an immortalized liver cell line. And Rice's group reports online in Science this week (www.sciencemag.org/cgi/content/abstract/1114016) similar results achieved by exploiting the same HCV isolate and a closely related cell line. Six years ago, Ralf Bartenschlager and colleagues at the Johannes-Gutenberg University in Mainz, Germany, laid the foundation for an HCV culture system in a report in Science. Bartenschlager's group engineered a small stretch of RNA that codes only for HCV's nonstructural proteins; specifically, the enzymes the virus uses to replicate. This “replicon”—so named because it carries all the information needed to copy itself—replicated to high levels when researchers artificially infected, or transfected, a cell line. But the system had two serious limitations: The replicon replicated efficiently in very few cells in culture, and it could not create a whole, infectious version of HCV. Bartenschlager (now at the University of Heidelberg) and, separately, Rice recognized that high-level replication might occur only in replicons that mutated and adapted to the cells. Both groups then demonstrated that by introducing mutations into the replicon, they indeed could vastly increase replication efficiency. Rice and co-workers further increased efficiency by improving the cell line. The replicons apparently only copied themselves in a subpopulation of the cells in culture. So the researchers transfected cells, identified the subpopulation that best supported the replicon, and then used interferon- to eliminate it. By “curing” the cultured cells, they created a new, optimized cell line. All the researchers needed next was to engineer a complete viral RNA that used the replicon with the adapted mutations as its backbone. “We hit the wall,” says Rice. These engineered HCVs replicated in the improved cell lines and also made proteins, but they did not form a new virus that could establish infections. For unknown reasons, the adaptive mutations in the nonstructural genes reduced the infectiousness of the complete virus. The race took an odd twist in 2001. Takaji Wakita of the Tokyo Metropolitan Institute for Neuroscience in Japan and his colleagues published a report in the Journal of Medical Virology describing an intriguing HCV isolate from an HCV-infected patient who developed fulminant hepatitis and then, oddly, cleared the virus. Wakita's group soon showed that a replicon made from the nonstructural regions of this virus worked in cell lines about as well as replicons that had adaptive mutations. Both the Chisari and Rice labs owe much of their success to clones of Wakita's isolate. “Without his clone, nothing would have happened,” says Chisari. Wakita first teamed up with Bartenschlager, but they had little success. Rice decided to create a chimeric virus that used Wakita's replicon for the nonstructural region and RNA from a different HCV from the same genotype for the rest of the genome. As Rice and co-workers show in their Science Express report, this chimeric virus produces high levels of infectious virus when transfected into the optimized cell line they had made earlier. Chisari collaborated directly with Wakita and used a complete clone of the virus isolated from the patient. “There's something special about that clone,” Chisari says. Rice also supplied Chisari with the optimized cell line, which Chisari's group attempted to improve one more time by transfecting and curing it again. In PNAS Early Edition, Chisari, Wakita, and co-workers report that when the clone was put into this cell line, it produced roughly equivalent levels of infectious HCV as did Rice's chimera. What is it that makes Wakita's isolate so special? “That's the main question in our laboratory,” says Wakita, who has another paper in press at Nature Medicine that describes how he and Bartenschlager ultimately succeeded, too. The secret must reside in the nonstructural region of that isolate, the only common part of the viruses used by each lab, says hepatitis researcher Robert Purcell of the National Institute of Allergy and Infectious Diseases in Bethesda, Maryland. “Eventually that will be sorted out because each gene will be sorted out,” says Purcell, who plans to put Wakita's clone into chimpanzees to better understand the relationship between the virus in vitro and in vivo. (Rice also plans to test his chimeric virus in chimps.) A key limitation of the current advance is that both Wakita's clone and Rice's chimera only contain RNA from one of HCV's six genotypes, which are geographically distributed around the world. “I think this is the long-sought-after culture system, but it's far from as good as we would like it to be,” says Chisari. “That's the next frontier,” agrees Rice. But for the time being, Chisari, Rice, and other hepatitis researchers at the front of the pack see far more opportunities than limitations. 13. PHYSIOLOGY # Science in the 'Death Zone' 1. Fiona Proffitt* 1. Fiona Proffitt is a freelance science writer in Oxford, U.K. A research team will scale Everest to investigate how a body copes with a lack of oxygen—and possibly learn why some patients do better than others in a respiratory crisis At 8848 meters, the summit of Mount Everest lies in the “death zone.” If a person were whisked straight up to this altitude without supplementary oxygen, he or she would lose consciousness within a minute or so and die soon after. The concentration of oxygen at the peak is only a third of that at sea level, putting it close to the limit for human survival. It wasn't until 1978 that Reinhold Messner and Peter Habeler achieved the remarkable feat of climbing Everest without bottled oxygen. Others have done it since then, but the ability of climbers to adjust to the lack of oxygen at extreme altitudes still puzzles physiologists—and tantalizes physicians. At a conference in London,* a group of mountaineering medics outlined their plans to investigate the physiological adaptation to hypoxia, or oxygen shortage, in a scientific expedition to Everest. Sister projects at sea level will see if the secret lies in our genes. The Everest team, from the Centre for Aviation, Space, and Extreme Environment Medicine (CASE) at University College London (UCL), says the research could also shed light on why some patients suffering from heart and lung disorders are more vulnerable than others. The UCL researchers believe their work may elucidate life-threatening conditions faced by some patients in intensive care units, whose blood oxygen levels may fall below 4 kilopascals (kPa). The normal pressure of oxygen in the blood is 12 to 14 kPa, and people usually lose consciousness when it falls below 5 kPa. Climbers summiting Everest without bottled oxygen are estimated to have blood oxygen levels below 4 kPa, says Hugh Montgomery, a UCL cardiovascular geneticist and research leader of the expedition. As intensive care consultants at UCL, Montgomery and Michael Grocott, who will lead the expedition overall, specialize in treating patients with acute respiratory distress syndrome (ARDS). With ARDS, blood oxygen levels fall dangerously low. Some patients pull through; others do not. Likewise, some climbers and high-altitude dwellers are able to cope with hypoxia; others cannot. Why the differences? Despite much research, says Montgomery, our understanding of the physiological mechanisms underlying the adaptation to hypoxia is “pretty primitive.” Attempts to explain human adaptability have focused on changes that enhance the delivery of oxygen to the tissues and cells, such as increases in breathing, heart rate, and red blood cell mass. But recent research shows that “something else is going on,” says Montgomery—and he contends that it probably has more to do with the biochemistry within cells than changes in oxygen delivery. Maximum oxygen consumption—the volume of oxygen consumed when a person is exercising as hard as possible—varies a lot among individuals. It is also often less than expected at higher altitudes, says Montgomery. Consider the 2003 study of 12 endurance runners led by physiologist Keisho Katayama of Nagoya University in Japan: The runners' oxygen consumption declined following 3 weeks' acclimatization to a simulated altitude of 4500 m, but they were still capable of doing the same amount of work on exercise tests. This suggests that the body may be able to tweak its metabolism to make cells burn leaner or even reduce oxygen demand by switching some cells off, Montgomery said at the conference. If this is true, perhaps ARDS sufferers are experiencing problems not only getting oxygen to their cells but getting their cells to use it more efficiently. It could also mean that it is unhelpful—even harmful—to give them supplementary oxygen. It may interfere with the body's self-protection mechanisms, warns Montgomery. Recent experiments bear this out: When immunophysiologist Michail Sitkovsky of the New England Inflammation and Tissue Protection Institute in Boston and colleagues gave oxygen to lung-damaged mice, it suppressed an anti-inflammatory mechanism normally triggered by hypoxia and caused further damage to the lungs, they report in the June issue of Public Library of Science Biology. Studying adaptations to hypoxia in sick people is problematic; once a patient is in respiratory distress, it's hard to ascertain what's normal. Why not instead see what happens to healthy people when their physiology is pushed to the limit? ask Grocott and Montgomery. They intend doing just that by climbing Everest in spring 2007. Along with remote medicine specialist Sundeep Dhillon, they will lead a team of six to 12 people—all medically trained, experienced climbers—up to the summit. The climbers will conduct physiological and mental performance tests on themselves at several stages along the way. About half the team will attempt the climb without supplementary oxygen. Others will test a new type of closed-circuit breathing apparatus that recycles exhaled air through a filter. The new design could provide a lightweight, compact alternative to conventional open-circuit systems used by patients with chronic respiratory diseases. This will not be the first time human physiological measurements have been made on Everest. In 1981 physiologist John West of the University of California, San Diego, and his team sampled lung oxygen and carbon dioxide (CO2) levels at the summit and took a blood sample at 6300 m. West and high-altitude physiologist Michael Ward, formerly of UCL, had earlier conducted the highest exercise tests, at 7440 m on neighboring Mount Makalu, as part of the 1960–61 Himalayan Scientific and Mountaineering (Silver Hut) Expedition, led by mountaineer Edmund Hillary. Still, the new expedition plans to collect unique data, both by working at higher altitudes and by using new technology, Grocott says. Some measurements—including those of blood oxygen and CO2 levels at the summit—will be groundbreaking. At several stages on the climb, the team will undertake cardiopulmonary exercise testing, using specially designed recumbent bicycles and devices that measure the amount of oxygen consumed and CO2 expelled with each breath. This will reveal whether there is an increased efficiency of oxygen use as climbers acclimate, how rapidly it happens, and what makes it happen. They'll test the idea that acclimatized people switch from burning fat to burning glucose, which uses less oxygen for a given energy output. Montgomery and anesthetist Monty Mythen of the Institute of Child Health at UCL are leading sister projects at CASE that aim to relate performance on cardiopulmonary exercise tests to changes in metabolism and genes suspected of being involved in hypoxia adaptation. For example, Montgomery would like to study high-altitude dwellers and elite climbers to look for variations in the angiotensin converting enzyme (ACE) gene. ACE increases levels of the peptide angiotensin II, which constricts blood vessels, and breaks down the polypeptide kinin, which dilates blood vessels. It also promotes fluid retention by increasing levels of the hormone aldosterone. In 1998, Montgomery and colleagues found that an ACE allele associated with lower production of the enzyme was linked to endurance performance in elite mountaineers and army recruits. Subsequent research by Montgomery and others linked the allele to endurance performance in elite athletes and, in 2002, they showed that ARDS sufferers are likely to have the deletion allele, which makes them less likely to survive. Although the physiological effects aren't entirely understood, clinicians may be able to help these patients by chemically mimicking the ACE-inhibitory effects of the insertion allele, says Montgomery. But not everyone is convinced that there is much to be gained from another scientific expedition to Everest or that the results can be applied to hypoxic patients at lower altitudes. “The mountain is a perfect model for the mountain” and not for a medical condition, says physiologist Robert Roach of the Colorado Center for Altitude Medicine and Physiology at Aurora. “There's very little to be learned [about ARDS and other conditions] from going to Mount Everest.” Roach argues that it will be hard to separate out the effects of hypoxia from those of other stresses. He and some other experts say it would be better to invest in studies of hypoxia under controlled conditions in low-pressure chambers. Low-pressure chambers have their drawbacks, counters West. Chambers make poor simulation models because people don't acclimate to low oxygen properly in them, “for reasons we don't quite understand,” he says. They're also expensive to run and hold a relatively small number of subjects, adds West. The Everest team expects to secure about$1.8 million for their high-profile trek from private sources. It would be impossible to secure this much for a chamber study, says Montgomery.

Offering to be your own guinea pig, while climbing Everest and experiencing hypoxia, is not for the faint-hearted. But as West says: “Science is full of educated guesses. There's nothing like making the measurement.”

• *KnO2wledge: Lessons Learnt from Life at the Limits, University College London, 27 April.

14. HIGH-ENERGY PHYSICS

# KEK Researchers Catch Glimpse of Outlandish Particles

1. Charles Seife

Enigmatic results are forcing physicists to consider that their collision data may bear the footprints of an unexpected newcomer and a long-sought hybrid

“When you have eliminated the impossible,” said Sherlock Holmes in The Sign of Four, “whatever remains, however improbable, must be the truth.” High-energy physicists are now coming to improbable conclusions about two recently discovered particles known as the X(3872) and the Y(3940).

“If you take the results at face value, it looks like the X(3872) is something new,” says Stephen Godfrey, a physicist at Carleton University in Ottawa, Canada. Scientists have been unable to reconcile the properties of the particle and the Y(3940) with simple predictions from standard theories about how quarks and gluons make composite objects. They suspect that the X(3872) may be the first member of a new class of particles not predicted by theory, and the Y(3940) might be quarry physicists have hunted for years: the hybrid meson.

Scientists working on the Belle experiment at the KEK laboratory in Tsukuba, Japan, caught the first glimpse of the X(3872) in 2003. They were using an accelerator to smash together electrons and positrons, producing vast numbers of B mesons—moderately heavy particles—which then decay into lighter particles. By studying the products of such decays, scientists at Belle and at the rival BaBar experiment at the Stanford Linear Accelerator Center are figuring out fundamental properties of the building blocks of matter (Science, 22 August 2003, p. 1026). But in the KEK data, there was a curious spike.

In a certain type of data plot, particles created in the decay of B mesons show up as a peak in the graph. When studying B mesons that had decayed into a combination of particles known as J/ψs and pions, the Belle researchers noticed that there appeared to be a peak at 3872 MeV—evidence of a new species roughly four times as massive as the proton. This was a problem.

“If it was a standard particle, theorists told us that there should be a strong decay” into a gamma ray and another type of particle, says Stephen Olsen, a physicist at the University of Hawaii, Manoa, and member of the Belle collaboration. But the new particle didn't seem to decay in that manner.

Baffled, the experimenters took a closer look. Theorists had predicted about a dozen as-yet-unidentified but otherwise ordinary particles around 3872 MeV, so they had to eliminate those candidates before they could conclude that the mystery particle was something entirely new. “There's a lot of slop in the models,” says Godfrey. “[The X(3872)] was not quite what was expected, but one should be a bit conservative and ask what are the conventional states that it could be.”

By studying the ways that the new particle decays and other characteristics, such as the angles at which the decay products fly apart, the scientists crossed all the standard particles off the list. “They seem definitely to have seen something inconsistent” with conventional particles, says Estia Eichten, a physicist at Fermi National Accelerator Laboratory (Fermilab) in Batavia, Illinois.

So what is this new particle? The most likely possibility, Olsen says, is that two mesons known as D0 and D0*, each made up of a charm quark and an up antiquark, cleave together and make a more complex object—much as a proton and a neutron can stick together to form a nucleus of heavy hydrogen. Eichten agrees that the sticky-meson hypothesis explains the properties of the X(3872) quite well.

The Y(3940), a slightly heavier mystery particle found by the Belle team, also fails to decay as theory says it should. “It's supposed to [almost] always decay to a D0 and D0*,” says Olsen. Instead, the particle gives rise to a J/ψ and other products. This strange decay pattern, like that of the X(3872), is leaving physicists scratching their heads. “It may be a completely new kind of particle,” says Olsen. Although conventional particles haven't yet been ruled out, he adds, this time physicists “don't have such a big list to track down” before concluding that the Y(3940) is something completely new.

If so, its newness is quite different from that of the X(3872). Unlike the X(3872), whose graph shows a nice clean signal with a sharp and narrow peak, the Y(3940) has a broad peak that's less distinct. A two-meson amalgam doesn't explain that signature and other properties. But a long-predicted but never-seen creature does: the hybrid meson.

A meson can be thought of as two quarks attached by a rubber band; the band is the strong-force “glue” that holds the meson together. “But say you pluck that elastic band so that it's vibrating,” says Godfrey. “It's a higher, excited state of the meson. You can calculate the predicted mass of these states.” This is a hybrid meson, and its excited glue would make it behave like a cross between ordinary quarky matter and matter that's made up of glue. A hybrid meson would have the unusual decay properties observed in the Y(3940), but it's too early to rule out more mundane explanations.

Physicists around the world are trying to follow up on these discoveries. Experiments at Fermilab have also spotted the X(3872) and are looking for the Y(3940). Belle physicists are working hard to rule out conventional explanations and bolster the case for the unconventional ones—and to chase down a handful of other strange particles that have been cropping up. “There's a few others we're looking at that we haven't reported yet,” says Olsen. Although it's still too early to declare that physicists have ventured into uncharted territory, the game is definitely afoot.

15. # From Dearth to Deluge

1. Eliot Marshall

The charge that women were being excluded from clinical studies led to the Women's Health Initiative; it produced a flood of data and controversy

At a congressional hearing 15 years ago, male leaders of the U.S. biomedical world endured a grilling about sex and science. Specifically, on 18 June 1990, Representative Henry Waxman (D-CA) asked the heads of the various National Institutes of Health (NIH) what they were doing to enforce a 3-year-old mandate to include women in clinical trials. (The NIH director's job was vacant at the time.) One by one, the witnesses acknowledged that they didn't have much to report. It was a pivotal moment, say advocates for women's health studies. It attracted the media and reduced an arcane debate about disease prevention and sex differences into a simple theme: Women were being excluded.

The drama had been carefully choreographed, recalls Phyllis Greenberger, executive director of the Society for Women's Health Research (SWHR) in Washington, D.C. In 1989, its leaders discovered that a government panel co-chaired by NIH official Ruth Kirschstein had found that “there was a dearth of women enrolled in clinical trials,” as Kirschstein recalls. The panel had also instructed U.S. health agencies to insist that grantees recruit more women—or explain why they didn't intend to.

“That was the hook” on which to hang an indictment, says Greenberger: “Here was a mandate that said they were supposed to be doing something—and they weren't doing it.” Meanwhile, SWHR's co-founder, NIH obstetrician-gynecologist Florence Haseltine, says that she used her own money to hire the lobbying team of Marie Bass and Joanne Howes to push the campaign, leading to a congressional audit and the 1990 hearing.

Although women were not systematically excluded from trials, later analysts have found (see sidebar) that they were kept out of some famous studies of heart disease in the 1980s. One of the male-only studies—with the acronym MR FIT—examined the benefits of a careful diet and exercise. Another, the Physicians' Health Study by Harvard Medical School in Boston, Massachusetts, found that middle-aged doctors who took a small dose of aspirin every other day had a significantly lower risk (44% less) of heart attack. For efficiency, it had enrolled 22,071 men (who were more likely than women to have heart attacks and yield data) but not one woman. Critics said the results were irrelevant for half the population.

16. # Clinical Trials: Keeping Score on the Sexes

1. Eliot Marshall

Biomedicine used to have a bias against including young women in clinical trials, says Ruth Kirschstein, a former director of the National Institutes of Health (NIH)—partly because of the thalidomide disaster. This drug, given to pregnant women to stop nausea in the 1950s and early 1960s, caused thousands of birth defects. After it was withdrawn, regulatory agencies directed that young women should be kept out of clinical trials to protect fetuses they might be carrying. That attitude lived on into the 1980s—long after new tests and testing methods had made it easy to identify early pregnancy and avoid risks, Kirschstein says. In 1987, a government panel she co-chaired found that U.S. agencies were slighting women's health and ordered that more women be included in research trials (see main text). Some trials, particularly big ones looking at heart disease, were designed to focus on male patients. Too often, and with too little evidence, results based on males were extrapolated to females, the review found.

Some independent analysts argue that, aside from the big heart studies, the imbalance was never that great. But since the initial complaints, both sides agree, NIH has tipped the balance to include more women. NIH's own evaluation of grants found in 1987 that only 13.5% looked at diseases unique to women. But then, only 6.5% were on diseases unique to men (Science, 11 August 1995, p. 766). In 1990, a General Accounting Office (GAO) study ordered by Congress found that NIH was being “very slow” in carrying out its own goals of recruiting more women and investigating physiological differences between the sexes. A sample of 50 grant submissions on diseases that affect both sexes, GAO found, included 20% that didn't mention the subjects' sex, and “some” that excluded women but didn't explain why. In a 2000 audit, GAO found that NIH had made “significant progress.” Peer reviewers reported that 94% of grant proposals in 1997 met the standards for including women, and “more than 50% of the participants in clinical research studies that NIH funded” were women.

As it became clear that women were not being excluded from the new crop of trials, critics shifted ground. For example, in a paper released on 10 May, the Society for Women's Health Research (SWHR) in Washington, D.C., said that NIH is not doing enough to get researchers to run clinical trials in a way that will bring to light physiological differences between men and women. The report by Viviana Simon, Sherry Marts, and other members of the SWHR staff found that “a very small percentage” of all indexed NIH grants between 2000 and 2003 (about 3%) were awarded to study sex differences. It also found that the richest institutes—such as those dedicated to cancer, heart disease, and infectious diseases—scored low in the study's index of funding research on sex differences. Now that women are being included in trials, SWHR argues, researchers should be doing more to learn how they differ from men.

17. # Gender in the Pharmacy: Does It Matter?

1. Jocelyn Kaiser

Studies of how women's and men's bodies process drugs have turned up mostly minor differences. But some drugs may be less or more effective in women or cause more side effects, and other variations may await discovery

In 1989, a 39-year-old woman blacked out while she and her husband were eating dinner. He rushed her to the Naval Medical Center in Bethesda, Maryland, where tests showed that her heart had a dangerous irregular rhythm that can lead to cardiac arrest. Doctors were puzzled: The woman was taking a popular antihistamine, Seldane, overdoses of which had caused abnormal heart rhythm, yet she was taking the recommended dose. The doctors consulted Louis Cantilena, a clinical pharmacologist at the hospital, who in turn called a colleague at the U.S. Food and Drug Administration (FDA). Looking through FDA's adverse events database, FDA staffers found two dozen cases of arrhythmias from Seldane, or terfenadine—the majority in women. It was one of the first red flags that researchers might have been missing sex differences in responses to drugs.

Combing through data on other medications, FDA and researchers realized that at least nine drugs could cause potentially fatal heart arrhythmias in women, especially when prescribed with certain antibiotics. By 2001, FDA had pulled four of these drugs off the market, including Seldane. “There's no way to know how many, but there were deaths,” says Raymond Woosley, then a pharmacologist at Georgetown University Medical Center in Washington, D.C., who began studying the problem.

The drug withdrawals fueled an argument made by advocates for women's health: Sex differences in responses to drugs had been missed because women were not always included in clinical trials, or if they were, the data were not broken down by sex. That has changed considerably in the past dozen years, after the National Institutes of Health brought more women into clinical trials and FDA rescinded a 1977 rule that excluded women of childbearing age from early trials (see p. 1570)—with positive results, advocates say. Earlier this year, for example, researchers reported in the New England Journal of Medicine (NEJM) that aspirin—which protects men against heart attack but not stroke—has exactly the opposite effect in women.

Yet sex differences in drug responses remain controversial. Concerns center on two aspects: how quickly drugs are metabolized and absorbed, and how they affect the body once they're in the bloodstream. Although studies have found many differences in how women and men process drugs, these changes are less worrisome than expected. Differences in how safe and effective a given blood level of a drug is for a man or woman are probably bigger issues, many experts agree. This is harder to study, however, and so far only a few clear-cut examples have emerged. That leaves some experts skeptical that sex will matter much in the long run; genetic variation among individuals, especially of different ethnicity, may dominate, they say. “Gender is not the major concern that we thought it would be,” says Leslie Benet, a pharmacologist at the University of California, San Francisco (UCSF).

But others counter that drug researchers have barely scratched the surface. Despite prodding, clinicians still don't always analyze data on women separately, and more research—and better research tools—may yet reveal more serious gender differences, they say. Even subtle sex differences may be important in an era of personalized medicine. “When it's all done, we still find sex is a factor that keeps coming out,” says clinical pharmacologist and cardiologist Janice Schwartz of UCSF.

## Less than expected

As far back as 1932, researchers noticed that female rats could be knocked out with half the dose of barbiturates needed for male rats of the same size. But such differences were largely ignored until 1993, when FDA reversed course on including women in trials. Until then, trial results were dominated by “the cult of the typical 70-kilogram male,” says Sherry Marts, vice president of scientific affairs for the Society for Women's Health Research in Washington, D.C.

That began to change, slowly. The 1993 FDA guideline explicitly urged drug companies to look for sex differences in drug processing, or pharmacokinetics. Researchers examined old animal data and new human evidence suggesting that males and females differ in the activity of liver enzymes that metabolize drugs, particularly the cytochrome P450 enzymes, whose ex-pression is mod-ulated by sex hormones. One such enzyme, CYP3A4, is involved in metabolizing more than half of all therapeutic drugs; women clear some CYP3A4 enzyme drugs more quickly than men do and thus may need a higher dose to get the same effect. (The difference probably also involves women's lower liver levels of a protein called P-glycoprotein that shunts the drug out of cells in which CYP3A4 processes it.)

Women are also smaller on average than men are, they may absorb drugs more slowly, and their kidneys filter excreted drugs out more slowly. Because women tend to have more body fat, fat-soluble drugs stay in their bodies longer. All this means a woman who swallows the same number of pills as a man may end up with a larger or smaller level in her blood.

In many trials, however, differences in women's responses to a drug disappear if the dose data are simply adjusted for body weight or surface area. The gender differences in processing drugs that remain appear to be relatively minor, says pharmacologist Bernd Meibohm of the University of Tennessee, Memphis. The best evidence is an FDA study of the 300 new drugs reviewed from 1995 to 2000, more than half of which provided sex data. Only for 11 drugs were pharmacokinetic differences greater than 40%, and none resulted in separate dosing instructions for women—indicating the difference wasn't important to the clinical outcome, notes Margaret Miller, science program manager for FDA's Office of Women's Health. Pharmacologist Gail Anderson of the University of Washington, Seattle, is not convinced, pointing out that in negotiating labels with FDA, companies would resist dosing for subpopulations—or even for body weight—because it makes it harder to market the drug.

Metabolism differences in women do matter for drugs that must be given in very precise doses, such as the blood thinner warfarin and cancer drugs and immunosuppressive drugs. But doctors already carefully tailor doses of those drugs to the individual, notes Benet. “The bottom line is, [sex differences in pharmacokinetics] doesn't seem to make a major difference,” agrees Meibohm.

## The female heart

A potentially much bigger problem is the difference in how men's and women's bodies react to the drug once it's reached the bloodstream. Known as pharmacodynamics, this property of drugs is harder to measure: Gauging improvement in depression, for example, is trickier than detecting the blood level of a chemical. But disparities for some classes of drugs have emerged.

Probably best-established are differences in responses to medications that can affect the heart's rhythm, such as terfenadine. These include some antihistamines, antibiotics, antiarrhythmics, and antipsychotics. Woosley and others showed that these drugs share the ability to block potassium channels in the heart, which in turn can affect the heart's rhythm. Two-thirds of reported arrhythmias from these drugs occur in women; they are especially vulnerable because the female heart has a longer “QT interval,” the time it takes to recharge between beats. More than 30 marketed drugs are known to cause arrhythmias, notes Woosley, president of the C-Path Institute in Tucson, Arizona. The University of Arizona lists these drugs on a Web site (http://www.qtdrugs.org/).

Women also appear to respond differently to drugs for treating or preventing cardiovascular disease. The latest example is the study of aspirin reported in the 31 March issue of NEJM of results from the Women's Health Study, which was launched in 1993 after protests that a previous study had included only men. Whereas men were protected from heart attacks but not stroke, women 45 years or older who took low-dose aspirin for 10 years had no fewer heart attacks but a 17% lower rate of stroke. The results may involve differences in physiology in women, such as smaller coronary arteries than men have and lower lipid levels before menopause.

Sex differences also seem to come into play in the reaction to opiates. The strongest evidence comes from a series of studies in the mid-1990s led by Jon Levine's group at UCSF, which looked at how men and women respond to drugs known as kappa-receptor opiates after wisdom tooth surgery. The drugs worked much better on women and caused fewer side effects than opiates such as morphine that have a different target, mu receptors. Male and female rodents also respond differently to opiates.

But other groups haven't yet replicated the UCSF results, and in a controlled lab setting—for example, with volunteers subjected to mild heat and muscle pain—the same sex differences aren't always observed, cautions pain researcher Roger Fillingim of the University of Florida, Gainesville. The discordant studies may reflect factors such as the type of pain or opiate drug dose, Fillingim says. “I just don't think we have enough data” to know which conditions result in sex differences, he says.

The jury is still out on antidepressants as well. A fairly large study led by psychiatrist Susan Kornstein of Virginia Commonwealth University in Richmond and published in 2000 in the American Journal of Psychiatry reported that women responded better to selective serotonin reuptake inhibitors (SSRIs); men got more help from tricyclics, which target receptors for serotonin and other neurotransmitters. Not all subsequent studies have found these differences, however.

Many agree that important sex differences are yet to be discovered. For example, researchers noticed only in 2002 that an old drug for heart failure, digoxin, raised the death rate in women by 4% in an earlier trial, possibly because they received too high a dose, notes Schwartz. The painkillers known as COX-2 inhibitors, one of which was withdrawn from the market last year because of side effects, are also under the microscope. Concerns were raised after Garrett FitzGerald's group at the University of Pennsylvania in Philadelphia found that blocking the COX-2 enzyme in mice also hinders estrogen's protective effects against cardiovascular disease, suggesting that giving COX-2 drugs to young women could put them at higher risk for heart attacks and stroke (Science, 19 November 2004, p. 1277).

Pinning down sex differences should become easier with new biomarkers—such as brain imaging—that enable researchers to measure disease and other endpoints, such as pain, more objectively. “My guess is we're going to find a lot more gender differences,” Woosley says.

Sex differences are also coming up in the context of pharmacogenomics: genetic differences, often tied to a single polymorphism, or mutation, that affect an individual's disease susceptibility, say, to heart disease or response to a drug. Although these mutations are usually not carried on the X chromosome and so are independent of sex, they can be modulated by sex hormones. For example, researchers recently found a polymorphism that makes redheaded women—but not men—more responsive to opiates, notes Fillingim, a co-author (Science, 16 July 2004, p. 328).

Even if most sex differences in drug responses aren't dramatic, they will feed into the cost-benefit tradeoff for a drug—all part of the personalized medicine equation, says Miller of FDA. “The most important message is to look for differences in treatment response by gender,” says Kornstein, editor-in-chief of the Journal of Women's Health. She and others say their colleagues in drug research are listening, but not everyone is on the same page yet.

18. # Sex and the Suffering Brain

1. Constance Holden

Researchers are seeking biological reasons for the widespread gender differences in the prevalence and symptomatology of mental disorders

It's easy to start a fight about whether there are gender differences when it comes to mental skills, but there's little debate that patterns of mental illness and disorders vary between the sexes. Women, for example, are more likely to get depressed (see table). Men are more severely afflicted by schizophrenia. Females have more anxiety. Males exhibit more antisocial behavior. Most alcoholics and drug addicts are male; females have more eating disorders. Even suicide has a gender bias. Females make more attempts; males are more successful.

Although culture helps shape how the two sexes express mental problems, some differences persist across cultures and across time, says psychiatrist Kenneth Kendler of Virginia Commonwealth University in Richmond. And that suggests a role for biology. In fact, says Thomas Insel, head of the National Institute of Mental Health (NIMH) in Bethesda, Maryland, “it's pretty difficult to find any single factor that's more predictive for some of these disorders than gender.”

Talking about sex differences has long been taboo in some quarters—“People hear ‘sex differences’ and think you're talking about individuals, not populations,” says Insel. “It's critical to remember there's a huge amount of variation within a population and overlap between populations.” But neuroscience research, especially the explosion in brain imaging, has produced data that are hard to ignore. “Every time you do a functional MRI on any test, different parts of the brain light up in men and women,” says Florence Haseltine, a reproductive endocrinologist at the National Institute of Child Health and Human Development (NICHD) in Bethesda, Maryland. “It's clear there are big differences.” Understanding them will have “tremendous implications” for treatments of brain diseases and injuries, says Viviana Simon, director of scientific programs for the Society for Women's Health Research in Washington, D.C.

Most mental disorders are complex and resist the hunt for specific genes, yet family and twin studies have demonstrated significant heritability for them. These disorders interact with brain differences between the sexes that arise from genes on the X and Y chromosomes and from the bath of gonadal hormones that soak fetal brains early in gestation. Sex hormones are far-reaching in their powers, notes Insel. “They are sort of master transcription regulators; they affect hundreds of downstream genes. … There's no question these are big players in mental disorders.” Those sex-related changes are sort of early filters, influencing the expression of underlying disorders in different ways, says psychologist Elizabeth Susman of Pennsylvania State University, University Park.

No one has managed to draw an unbroken line from prenatal development to adult behavior. But some researchers are now trying to tease apart just what aspects of brain anatomy and chemistry can help account for the gender skewing in mental disorders. “We're just at the beginning of trying to examine these differences,” says Cornell University endocrinologist Margaret Altemus. Some studies are contradictory, and there is still more known about animals than about humans.

## Affective disorders

Epidemiologic studies show that women are more vulnerable than are men to most disorders that affect the emotions. These include major depression and a host of anxiety-related conditions, such as generalized anxiety disorder, panic disorder, posttraumatic stress disorder, and phobias.

Anxiety and depression are very closely related: Eye-blink tests reveal that a strong “startle response” is a good predictor for both. Negative experiences can trigger both anxiety and depression in vulnerable people. Those feelings involve the activation of multiple neurotransmitter and hormonal systems, including stress mechanisms that are heavily influenced by sex hormones.

The human stress response basically has two components: the autonomic nervous system that causes raised heartbeat, sweaty hands, and gut churning; and the slower-responding hypothalamic-pituitary-adrenal (HPA) axis. The HPA axis involves a cascade of hormonal events that are normally counteracted by release of the stress hormone cortisol.

Because female depression rates start to rise during puberty, notes psychologist Laura Stroud of Brown University in Providence, Rhode Island, researchers posit that hormones play a role in women's vulnerability to affective disorders. Animal and human research has shown that sex hormones affect stress responses in different ways. Just the right amount of estrogen is required for emotional balance, says NICHD endocrinologist George Chrousos. Some women go into “withdrawal” and hence depression when levels drop. But too much estrogen can overactivate the HPA axis, also resulting in depression, he says. Testosterone, on the other hand, may protect against stress and depression through its damping effect on HPA reactivity. This is illustrated by a study appearing last month in Neuropsychopharmacology. Psychiatrist David Rubinow of NIMH and colleagues suppressed sex hormone production in 10 men and stimulated their HPA axes with corticotropin-releasing hormone (CRH). They found that testosterone replacement “significantly blunted” the cortisol response to CRH.

Other research shows that fear, a powerful stressor, may activate stress responses more in females than in males. Data from rats and mice “overwhelmingly indicate that females show more intense fear responses than males,” says Jaak Panksepp of the Medical College of Ohio in Toledo. Testosterone appears to reduce males' reactions to pain, he adds. Human evidence is beginning to accumulate. In a study in the journal Emotion in 2001, researchers at the University of Florida (UF) observed the reactions of 50 women and 45 men to pictures of distressing things, such as car crashes and mutilations. They found that women had more extreme autonomic reactions as gauged by heart rate, skin conductance, and the startle response. “Women are more reactive on average in everything involving negative stimuli,” says UF psychologist Margaret Bradley.

Brain-imaging studies are supplying a wealth of new data. A group at Westmead Hospital in New South Wales, Australia, recently completed a study (in press at Neuroimage) showing that females exhibited more widespread activation of the amygdala, the seat of the fear response—corresponding with rapid heart rate and sweating—than males did in reaction to pictures of people with fearful facial express-ions. In another study, in press at Neuro-report, a team headed by David Silbersweig of Cornell University's Weill Medical College in New York City found that normal females show more reaction to stress—in this case, anticipating a pain on the wrist—in the subgenual prefrontal cortex, a key region linked with anxiety and depression.

Although the evidence is more ambiguous, gender differences have also been seen in the HPA axis, the other component of the stress machinery. Observing 50 young volunteers, half of them women, Brown's Stroud found that in an “achievement” test, in which subjects must make a speech and perform a subtraction task in front of a panel of judges, men secreted higher levels of cortisol. But in a “social rejection” test, in which trained confederates made the subjects feel excluded in brief interactions, women's cortisol levels were higher, her team reported in the journal Biological Psychiatry in 2002.

NICHD's Chrousos contends that the HPA axis is slightly more reactive in females, as shown in a study in which young women responded with significantly higher levels of cortisol to a CRH challenge. Chrousos speculates that a highly tuned HPA axis, clearly evident in rodents and primates, is an evolutionary adaptation to help mothers protect their young.

## Aggression and impulsivity

Sex hormones are also implicated in aggression-related gender differences, notes Kendler. There is abundant evidence, he says, that men are more prone to expressing unhappiness through an “externalizing pathway” of physical behavior that includes drinking, drug abuse, and violence, whereas women are more likely to “internalize,” leading to depression and disorders such as anorexia.

The pattern of male externalizing, like sex differences in general, becomes more pronounced during puberty, when the hormones are flowing. “We know, from primate studies, that testosterone is directly related to aggression,” says Kendler. “If you give females testosterone, they get more aggressive.”

Addictions follow the pattern of male externalizing. Epidemiological studies have shown, for example, that in families of women with bulimia, the men often have alcoholism and other addictions. Studies have repeatedly shown that even within the alcoholic population, females are more often diagnosed with depression, whereas more males express antisocial behavior.

The sex-based tendency to act out versus internalize is evident in the distribution of the personality disorders, which involve maladaptive patterns of thinking and relating to the world. Some of these, such as dependent or histrionic personality disorders, are hotly debated, and critics argue that they only exemplify learned gender-typical behaviors. “Can it be that human beings manifest certain symptoms in ways that are politically and socially acceptable within certain historical times?” asks feminist therapist Arlene Istar Lev of Albany, New York. But psychiatrist Larry Siever of Mount Sinai School of Medicine in New York City says there's more to it. “It used to be a non sequitur to link biology and personality,” he says. “Now we see a very real substrate.”

Take borderline personality disorder (BPD), which features extreme emotional instability, impulsivity, and self-harming behavior and is most often seen in post-adolescent girls. Although people with BPD frequently have a history of childhood abuse, says Siever, brain scans of patients also show abnormalities. In an unusual study now under review at Biological Psychiatry, Siever and psychiatrist Antonia New also found sex differences. They compared the brains of 17 males and 9 females with BPD and a history of impulsive aggression to normal controls matched for sex. They report that the males with BPD showed less neural activity in prefrontal areas involved in inhibition. This “presumably suggests a brain mechanism” for this type of aggression, Siever says. The fact that males with BPD are more prone than women with BPD to impulsivity and aggression could partly explain why more women get the diagnosis, he adds; the men may be seen as having antisocial personality disorder (ASP).

Researchers are also looking into the biological dimensions of ASP, in which males outnumber females three to one. People with ASP (formerly known as psychopaths) don't form deep attachments and feel little guilt. Psychologist Adrian Raine of the University of Southern California in Los Angeles believes he has hit upon a possible biological marker. He reported in 1997 in the journal Child and Adolescent Psychiatry, from a longitudinal study of 1800 children in the Republic of Mauritius, a strong correlation between slow heart rate among 3-year-old boys—suspected to reflect reduced autonomic reactivity—and their subsequent antisocial behavior as adolescents. A high threshold for reacting to physical or social threats can make for “fearlessness,” which in turn inhibits the learning of normal social inhibitions, says Raine.

Susman of Penn State has similar findings. In a 1997 study she reported that adolescents who had low cortisol levels prior to an anticipated physical stressor exhibited raised antisocial behavior 1 year later. She believes these individuals are unable to “anticipate” stresses. “They're not good at planning or regulating,” she says, so “they don't anticipate fear.” It may be that males are more prone to hypoarousal of the stress response system and females to hyperarousal, speculates Susman.

## Thought disorder

Sex differences also extend to cognitive functions such as memory, attention, and perception. Men's brains are more lateralized, which means that higher cortical functions tend to be centered in the right or left side of the brain, says genetic epidemiologist Kathleen Merikangas of NIMH, whereas in females there's more “crosstalk” between the sides and therefore more redundancy. Evidence for this comes from the fact that women are more likely than men are to recover language from strokes in the left hemisphere, where language is centered.

This redundancy may also be protective in girls, who have much lower rates than do boys of childhood developmental and mental disorders, including attention deficit hyperactivity disorder (ADHD) and autism, says psychiatrist Raquel Gur of the University of Pennsylvania in Philadelphia. Estrogen, too, says Gur, appears to have neuroprotective effects according to results of research on brain injury, epilepsy, and cognitive decline in aging.

Gur says these differences also seem to work against men afflicted with schizophrenia, the most complex and devastating mental illness of all. More males than females have schizophrenia, and they have earlier, more severe symptoms, says psychiatrist Jill Goldstein of Harvard University. In brain scans comparing men and women with schizophrenia, Goldstein has found that men tend to have greater deficits than women do in attention, language, visual-spatial perception, and other areas ruled by the cortex, such as olfaction and motor skills. These are also areas that she and others have found to be sexually dimorphic in normal subjects. She believes these deficits all begin prenatally during the period of sexual differentiation of the brain.

Researchers are still cautious about their conclusions. Despite evidence of a “huge number of … differences between men and women's brains,” says Cornell's Altemus, “it's hard to know which are functionally relevant.” Nonetheless, times have changed, observes Goldstein: “For many years we were not even allowed to say there were sex differences in the brain.”

19. # Poor Countries, Added Perils for Women

1. Greg Miller

When the Indian government disbursed the first round of financial aid to families in Tamil Nadu state, hard hit by the 26 December 2004 tsunami, they doled it out to men, the traditional household heads. That didn't work too well, says K. Sekar, a psychiatrist with India's National Institute of Mental Health and Neurosciences, who has coordinated mental health support for the tsunami survivors.

Many men have coped with the disaster by drinking, Sekar explains, and much of the money intended for families flowed straight into state-run liquor stores. The second round of aid, delivered to women, seems to be doing more good, he says. Even though women—especially those who lost children in the tsunami—appear to have suffered most in psychological terms, they've handled it differently. Drinking is socially unacceptable for women, Sekar says, and they have largely internalized their distress, showing signs of anxiety and depression.

The psychological aftermath of the tsunami in Tamil Nadu reflects two of the most robust trends in psychiatric epidemiology: Across the globe, anxiety disorders and depression are more common in women, and substance abuse is more common in men (see main text). The situation there also hints at how social factors and women's roles as childbearers influence the mental health of women in developing countries, often for the worse.

Although there are no reliable figures on the prevalence of mental disorders for many parts of the world, there are signs that women in poor countries are more vulnerable than women in richer parts of the world, says Ricardo Araya, a Chilean-trained psychiatrist at the University of Bristol, U.K. A recent study by Araya and colleagues is the first to attempt a direct comparison of the gender gap in the prevalence of depression and anxiety disorders between developing and developed countries. The researchers interviewed more than 10,000 men and women in urban areas of Chile and the United Kingdom about their mental health and reported in the April issue of Social Science and Medicine that the gender gap is greater in Chile.

For women in sub-Saharan Africa, especially, HIV is a major risk factor for depression, says Sylvia Kaaya, a psychiatrist at Muhimbili University College of Health Sciences in Dar es Salaam, Tanzania. Although little research has been done to examine HIV's toll on women's mental health in particular, Kaaya suspects that women experience extra stress that ups their risk of depression. Women have little say in negotiating condom use and other protective measures and are generally expected to care for infected relatives, Kaaya says.

Much of the research on women's mental health overall in developing countries has investigated its links to reproductive health. Women in poorer countries are more likely to have miscarriages or lose young children, and such events, especially when they occur more than once, take a heavy toll on a woman's psychological well-being, says Veena Das, an anthropologist at Johns Hopkins University in Baltimore, Maryland. Das has just completed a study in poor communities in Delhi, India, that documents sharply elevated rates of depression in women who have lost multiple pregnancies.

Women who give birth to healthy babies aren't immune either. Rates of postnatal depression run high in some developing countries. In India and Pakistan, for example, a handful of studies in the past few years have found that 20% to 30% of women suffer postnatal depression, about twice the prevalence in wealthy countries. That's not just bad for moms. A study published last September in the Archives of General Psychiatry found that Pakistani infants born to depressed moms were four times more likely to be underweight 6 months after birth.

Socioeconomic factors work against women in many societies, says Jill Astbury, a psychologist at Victoria University in Melbourne, Australia. Even in developed countries, she says, the most disadvantaged women—for example, single moms with low incomes, insecure work, inadequate housing, and lack of child care—have rates of depression two to three times higher than those of women in more favorable circumstances.

Unemployment and low income, aside from being bad for mental health in their own right, have been linked to high rates of another risk factor: domestic violence. “Factors in many developing countries such as low levels of education for women … [and a] lack of legal redress and property rights in divorce make it more likely that women living with violent partners will be forced to stay with them to survive economically,” Astbury says.

Such factors might help explain why some of the highest suicide rates in the world are found among women in developing countries. More than half of all female suicides worldwide take place in China, one of the few countries where more women than men die by suicide. A paper published last year in The Lancet reported a startling suicide rate of 148 per 100,000 among young women in Vellore, an inland city in Tamil Nadu, India. In the United States, roughly 4 women per 100,000 commit suicide each year.

Rising economies don't necessarily relieve the risk factors for poor mental health, however. A 1999 paper published in Social Science and Medicine suggested that growing income disparities created by rapid economic development in India, Chile, Brazil, and Zimbabwe may have increased the risk of anxiety and depression for women there. Development can add to the stress of everyday life for women, says Araya, one of the study's authors. As jobs become available, women are often expected to work outside the home in addition to their household duties. “They go and work in a sweatshop, and then they have to go home and cook,” he says.

A more comprehensive view of women's mental health around the world should come from a massive international survey now under way in 28 countries. Ronald Kessler, an epidemiologist at Harvard Medical School in Boston who is directing the project for the World Health Organization, says it will examine a variety of potential influences on women's mental health, such as access to birth control, property rights, education, and reproductive history, including age of puberty.

Some researchers have proposed that sex hormones are responsible for the higher incidence of depression and anxiety in women, but the main evidence for that hypothesis is that these disorders appear in midpuberty in the United States, Kessler says. In developing countries, puberty is often delayed by several years as a result of malnutrition, even though girls marry and are thrust into adult roles earlier. Says Kessler: “This creates a natural experiment to tease out the relative effects of biology and social roles on female mental illness.”

20. # Let's Talk About Sex--and Drugs

1. Martin Enserink

Seven years after Viagra was launched, many drugs are on the horizon to treat women's sexual problems. But several questions remain: Are they safe? And are they needed?

Laura was miserable. She had once enjoyed a healthy sex life, but any desire for sexual activity had vanished after her hysterectomy. Her 9-year marriage was on the rocks, and her husband was becoming emotionally abusive.

Her real-life story, told by marriage and sex therapist Jean Koehler to a panel of experts gathered by the U.S. Food and Drug Administration (FDA) in December, was meant to help persuade the committee to approve the first drug developed to treat women's sexual problems. Koehler had traveled from Louisville, Kentucky, to suburban Maryland to deliver 3 minutes of testimony on behalf of a testosterone patch called Intrinsa. Once her client started taking testosterone, Koehler said, her life changed for the better: She enjoyed sex again, her relationship improved, “and two little children were spared the trauma of impending divorce.”

Intrinsa, developed by Procter and Gamble (P&G), is part of a wave of new drugs stirring controversy before a single one has hit the market. Advocates for these drugs, including some prominent researchers of women's sexuality, say they have the potential to help tens of millions of suffering women; not just those who have had their ovaries removed, which often happens as part of a hysterectomy, but “naturally menopausal” women as well. Even if the drugs don't promise to save a marriage, market analysts see them as potential blockbusters.

But the FDA panel was not convinced. At the end of its daylong session, the group unanimously rejected the drug. Panelists decided that there weren't enough data to show that long-term hormone treatment—patients have to wear the patch continuously—does not cause serious side effects. The decision, decried as overly cautious by proponents of the patch, was a setback not just to P&G but to several other companies whose products contain testosterone.

But safety isn't the only issue. Some researchers also worry that the new pills, patches, gels, and nasal sprays will lead women to take drugs for what are really social or psychological problems that can be treated more effectively with education or psychological intervention. “Most of women's sexual complaints have to do with self-respect, self-image, and the quality of the relationship,” says clinical psychologist Leonore Tiefer of the New York University (NYU) School of Medicine, a leading critic of what she calls the “medicalization” of female sexuality. “They're things a pill can't treat.”

## “Our turn”

It's no surprise that the drug industry has turned its attention to women. Sildenafil, or Viagra—discovered when a candidate angina drug had surprising side effects—has grossed billions of dollars for Pfizer since its launch in 1998; copycats tadalafil (Cialis) and vardenafil (Levitra)—which also block the PD5 receptor, resulting in increased blood flow to the penis—have become successes in their own right. Meanwhile, interest in women's sexuality is growing; to wit, the adventures of four Manhattan women in the HBO smash hit series Sex and the City. “Women started asking: What's there for us?” says Harvard reproductive endocrinologist Jan Shifren, who directs the Vincent Menopause Program at Massachusetts General Hospital in Boston.

But exactly what sex drugs should do in women is much less obvious than in men. When the FDA put together “draft guidelines” in 2000 for companies interested in producing drugs to treat female sexual dysfunction (FSD), it followed the 4th edition of the Diagnostic and Statistical Manual of Mental Disorders. DSM-4 says that FSD has four major components: decreased desire to have sex; decreased arousal (such as blood flow to the genitals and lubrication); pain during intercourse; and difficulty or failure to have orgasms. Companies must choose which component their drug affects and show efficacy in women with that problem.

Not everyone subscribes to that delineation, based on the classic model of the “human sexual response cycle” proposed by William Masters and Virginia Johnson in the 1960s. The distinction between desire and arousal, for instance, doesn't make much sense, says Ellen Laan, a sex researcher at the University of Amsterdam in the Netherlands. How to measure a drug's effect on people's sex lives is controversial, too. Because of those debates, the final version of FDA's guidelines has yet to appear.

In the meantime, about two dozen companies now have products in development for FSD (see table). Viagra, once considered a candidate to treat arousal problems, is no longer among them; Pfizer gave up last year after disappointing trials. But other companies have products that would do the same: Vivus in Mountain View, California, for instance, is in phase III trials with alprostadil, a vasodilating agent that women apply directly to their genitalia.

Most candidate drugs, however, focus on what clinicians say is by far the most common disorder: decreased interest in sex, also known as hypoactive sexual desire disorder (HSDD). Some of these compounds act on the central nervous system. One is flibanserin, a pill previously studied and rejected as an antidepressant, from German pharma giant Boehringer Ingelheim. Another, PT 141 from Palatin Technologies in Cranbury, New Jersey, is a nasal spray that stimulates melanocortin receptors in the brain.

Plain old testosterone is the basis for most of the desire-enhancing products. It usually comes in the form of skin gels, sprays, or patches, because the hormone is broken down quickly by the liver when taken orally. Women naturally produce testosterone, although at much lower levels than men, and production declines after menopause. Levels also drop on average by 50% after a woman's ovaries are removed, a condition called “surgical menopause.” Several small trials suggested that testosterone enhances sexual desire in women, and U.S. doctors widely prescribe the hormone “off-label”—without being specifically approved—to women with HSDD. (European women are generally more reluctant to take hormones, Laan says.)

Because no specific product has been approved for women, doctors prescribe male testosterone drugs at about one-tenth of the dose or order pharmacists to produce special formulations that contain smaller amounts. A product aimed at and approved for women would be more convenient and safer, says Shifren, as well as opening a huge new market.

Those hopes were dealt a blow by the FDA panel in December. P&G had asked for approval of Intrinsa in surgically menopausal women first. The panel concluded that the results of two trials in this group were “clinically significant” if not exactly mind-blowing. (Roughly, patients who took placebo went from 3 to 4 “satisfying sexual episodes” per month, whereas those who got testosterone went from 3 to 5-5.5.) Side effects, such as increased facial hair growth and acne, were limited.

But the panel balked at the long-term safety data. The two trials together had enrolled 1095 women for 24 weeks—not nearly enough time to detect subtle risks resulting from long-term use, says panel member Steven Nissen, a cardiologist at the Cleveland Clinic in Ohio. Fresh on the panel's mind, he says, were the disturbing results of the Women's Health Initiative (WHI), a huge study funded by the U.S. government, which discovered in 2002 that long-term use of estrogen, alone or in combination with progestin, can increase women's risk of cardiovascular disease (see p. 1570). “We have a bad history with manipulating hormones in women,” says Nissen, and the decision “wasn't even close.”

Some critics say the vote smacks of a double standard, because drugs like Viagra, or even testosterone treatments in men, were never subjected to the long-term safety trials that the panel wished to see. “Be as conservative for men as you are for women,” says Shifren. Clinical psychologist Sheryl Kingsberg of Case Western Reserve University in Cleveland, Ohio, calls the panel “very, very overconservative” and says it's “paternalistic” to deny women the choice to use testosterone.

The fate of Intrinsa and similar drugs is unclear. P&G withdrew its application after the panel meeting; a spokesperson says the company “is working with the FDA” to design new trials. “The key questions are going to be: How long does a trial have to be, and how many patients?” says Stephen Simes, president of BioSante Pharmaceuticals in Lincolnshire, Illinois, a company developing a testosterone gel. Companies will abandon their efforts if the agency requires studies like the WHI, which enrolled more than 16,000 women for 5 years in its main trial, Simes predicts. But Nissen says a trial of a few thousand women for 2 years, plus thorough postmarketing surveillance, might allay the worries.

## Defining what's normal

But proving the safety of Intrinsa and its slew of competitors won't solve women's sexual problems, says NYU's Tiefer, who also gave a 3-minute presentation at the December meeting. Tiefer worries that women will feel compelled to start taking drugs, even if they're comfortable with their decreasing sex drives, once they become available. “I'm pro-sex,” she says. “I'm pro-porn, I'm pro-vibrators. … But sex is a hobby. It's fine not to do it if you're not interested.” (And certainly, an abusive husband like Laura's isn't a reason to put a woman on drugs, she adds.) Tiefer has founded a group, FSD Alert, that takes a feminist view of female sexual problems and puts more emphasis on sociocultural, political, and psychological factors.

There are other foes of FSD as a medical problem. In a series of articles over the past few years in the British Medical Journal, Ray Moynihan, a freelance journalist based in Sydney, Australia, called it the “corporate-sponsored creation of a new disease.” He implicates the media for what he says are titillating but sloppy stories.

Those who favor the new drugs—even while admitting that they receive corporate support—dismiss this idea as absurd and slightly conspirational. Women had sexual problems long before drug companies started paying attention, says Shifren. And counseling or a getaway weekend with their partners, she notes, are some of many other options before medication. For some of her clients, lack of desire really is a source of misery, Shifren says.

Irwin Goldstein of Boston University adds that the critics are now telling women what men heard in the pre-Viagra era: that it's all in their heads. “They talk about medicalization. I call what they do psycholization,” Goldstein says.

But even experts who believe that some women might benefit from medical treatment don't like the idea of large numbers of healthy women, nudged by wall-to-wall advertising on U.S. television, on FSD drugs. Already, the extent of the problem is being blown far out of proportion, says John Bancroft, a former director of the Kinsey Institute at Indiana University, who is now retired in England.

For instance, a 1999 study by Edward Laumann and his colleagues at the University of Chicago found that a staggering 43% of women between 18 and 55 suffer from sexual dysfunction—a number often repeated in scientific literature and the press that Bancroft calls “extreme.” A recent British study suggests that many problems are transient, he adds: Although 40% of women reported having a problem with sexual function that lasted at least 1 month, the study found, only 10% had complaints that lasted longer than 6 months.

Laan, at the University of Amsterdam, also believes that there's nothing medically wrong with most of the women who have arousal or desire problems. Instead, she says they just need more sexual stimulation. A recent German study among college students, for instance, showed that a woman's desire dropped with the duration of the relationship. “It's a huge taboo to say so, but many women who have lost interest in their partner still feel like having sex with the guy next door,” Laan says. But desire can be stimulated, she adds, by anything from romantic dinners to fantasizing: “It's just something that takes some work.”

Ironically, the drug trials themselves suggest that some women may not need desire-boosting drugs. Most show a considerable placebo effect; in the Intrinsa studies, for instance, some 36% of patients on placebo wanted to continue after the study closed. Maybe P&G should just market the placebo, Nissen quipped during the panel meeting. Talking about a sexual problem and deciding to tackle it might have a therapeutic effect by itself, say researchers.

Even with all the questions about FSD drugs, Bancroft believes that the increased attention will benefit the field. “We're having a very healthy debate,” he says. “The good thing is that we'll come out of this with a much better understanding of women's sexuality.”

21. # Bone Quality Fills Holes in Fracture Risk

Osteoporosis isn't the only factor behind broken bones. A better understanding of bone quality, coming from biochemical markers and refined imaging techniques, will help predict who is most at risk of debilitating fractures

When a woman is tested for osteoporosis, technicians shoot low-dose x-rays through her hip to get a picture of the bone and a measure of its density. The less bone, the higher the overall risk of breaks, including debilitating hip fractures. But over the last decade, researchers have come to a greater awareness that it's not just quantity that matters: Bone quality counts for a lot.

The importance of bone quality—a term covering aspects such as the organization of the tiny struts that make up the inner tissue—became obvious during clinical trials of drugs for osteoporosis. These drugs prevent the loss of bone, but it turned out that, statistically, bone mineral density (BMD) couldn't explain all of the reduction in fracture risk. That fit with observations by clinicians: Some women with osteoporotic bones don't suffer breaks, whereas many women with apparently healthy bones still end up with fractures.

Identifying women at risk before they fracture is “the most challenging public health question” facing osteoporosis researchers, says Ego Seeman of Austin Hospital in Melbourne, Australia. And it's not just an issue for women. Osteoporosis is becoming more common in men, and more commonly diagnosed, especially as they live longer.

Researchers are trying to get a better handle on bone quality in several ways. They're searching for new and better biochemical markers of bone change, to add to the handful already used in the clinic to assess the effects of drugs. Higher resolution imaging with computed tomography (CT) and magnetic resonance imaging (MRI) is beginning to probe the inner architecture of bones without the need for direct sampling.

The hope is that these advances may one day better identify patients in need of treatment, as well as provide a way to chart their progress on drugs, but the newer imaging techniques are still being developed and won't be widely available for several years. In the meantime, some researchers are trying to integrate proven risk factors to predict a woman's chance of fracture.

## Strong bones

Osteoporosis is a factor in more than 1.5 million fractures each year in the United States alone. Costs have been estimated at more than \$17 billion a year, particularly from hip fractures, more than 75% of them in women. Part of the reason is that women who are not in nursing homes are twice as likely as men to fall, perhaps because they lose muscle strength faster with age. But another major factor is that their bones tend to become much weaker with age than men's do.

Strength comes from two features of bones. The outer shell of dense material, called cortical bone, is like the metal tubing of a bicycle that makes a strong, light frame. Inside this cortex is a porous network of tiny support struts and rods, called trabeculae. Trabecular bone makes up just 20% of bone mass but most of its surface area.

Sex differences appear relatively early in life. Growing girls tend to add more mass to the inner side of the bone cortex, beefing up the trabeculae to create a storehouse of calcium for pregnancy and lactation. Boys, in contrast, tend to add more material to the outside of the cortex. The greater the diameter, the stronger the bone. The effect, as seen in cross-sectional studies, is “absolutely huge,” says Heather McKay of the University of British Columbia in Vancouver. In addition, girls tend to be less active than boys, she says, so many don't get the bone-building benefits of exercise.

The big kicker comes at menopause. Estrogen is a key regulating signal for the cells that are constantly remodeling bone, thus repairing damage and allowing bones to bulk up to the loads placed on them. When estrogen levels decline during menopause, the bone-building cells known as osteoblasts slacken their activity. But the bone-resorbing osteoclasts continue to remove bone mineral and break down collagen. That means women typically lose 1% to 2% of their bone per year around menopause, more of it from trabecular bone.

Several risk factors influence the likelihood that a woman will lose more bone than normal and eventually suffer a fracture. A previous fracture ups the odds substantially, as does a family history of fracture, although genetic factors remain fairly murky. Race matters, too. The incidence of hip fractures is 25% lower in Asian than in white women, for example, even for women with similar bone densities. Behaviors—poor diet and lack of exercise, especially in youth—are also negative influences on bone health, as discussed in a massive report from the Surgeon General last year.*

These risk factors are fairly weak predictors of an individual's absolute risk, however. Up until the 1980s, clinicians basically waited until a fracture occurred before treating patients for osteoporosis. Diagnosis—and research—got a considerable boost in the 1990s with the advent of dual x-ray absorptiometry (DXA). “It just revolutionized the field,” says B. Lawrence Riggs of the Mayo Clinic in Rochester, Minnesota. DXA enabled clinicians and researchers to follow patients over a long time and assess their responses to medications, helping bring the current crop of drugs to market, Riggs says (Science, 3 September 2004, p. 1420). In the United States, the National Osteoporosis Foundation recommends that women over the age of 65, or younger women who have one or more risk factors, be tested with DXA for osteoporosis.

But DXA's usefulness for making predictions is limited. “The number one clinical goal is to be able to sit down with a patient and give a numerical indicator of fracture risk,” says Lawrence Raisz of the University of Connecticut Health Center in Farmington. DXA doesn't do that well, although many researchers point out that it's a better predictor than is cholesterol level for heart disease. By factoring in bone quality as well, researchers and doctors eventually hope to do better.

## Sharpening the picture

One of the main approaches to gleaning details about the quality of bones is to measure the activity of osteoclasts and osteoblasts, the cells that remodel bone and thus influence its structural properties. The first cell activity marker approved by the U.S. Food and Drug Administration, in 1995, measures the products of bone breakdown and can pick out women with extremely high rates of bone loss. In general, however, markers are not currently useful for diagnosis of osteoporosis, because levels overlap between those who have and don't have the disorder. Researchers are trying to explain the variability and investigating new markers that might be more specific.

The main clinical use of markers at the moment is to help chart how patients respond to drugs. That kind of information may also encourage patients to keep taking their medicine, as Pierre Delmas of Claude Bernard University in Lyon, France, explained last month at a meeting on bone quality run by the National Institutes of Health and the American Society of Bone and Mineral Research in Bethesda, Maryland. His unpublished data showed that providing patients with progress reports from biomarkers could increase the numbers who stay on their medications by 20%. Biochemical markers may also help refine the assessment of fracture risk, but the results of large studies so far have been inconsistent.

Another way of getting new information about bone quality is by looking at bone architecture directly. A time-tested research method is to study actual bone from biopsies, cadavers, or hip replacement operations. CT and electron microscopy can resolve individual rods and struts, the crucial support elements inside trabecular bone. But direct sampling is too invasive and expensive to be used to track individual patients' health.

Researchers have been trying to get similar and more clinically useful information using imaging tools. One benchmark in the field is a 2001 paper in the Journal of Bone and Mineral Research (JBMR) by Felix Wehrli's group at the University of Pennsylvania, Philadelphia. The researchers showed in a study of 79 women with various bone densities and vertebral deformities that a souped-up MRI machine can reveal microscopic bone structure noninvasively. In April, a group led by Charles Chesnut of the University of Washington, Seattle, published online in JBMR the first such longitudinal study of bone microarchitecture with MRI.

The other main imaging techniques use quantitative CT, mainly to study peripheral bones, such as the forearm. Aspects of bone quality are then extrapolated to hip and spine. Given the small size of studies so far, CT and MRI haven't been used to assess fracture risk. Researchers say those results should come in the next few years: Larger trials are incorporating CT and MRI in subsets of patients.

One attempt to get at fracture risk is already under way. Tony Keaveny, a biomechanical engineer at the University of California, Berkeley, is using a technique called finite element analysis. Keaveny and colleagues take CT images of human vertebrae, including information about the trabecular architecture, and model how they respond to stress. In a paper published in Bone in 2003, he and his former student R. Paul Crawford showed that their analysis of CT images of cadaver bones predicted 85% of the variation in bone strength in experiments with actual loadings of the bones—“better than BMD did,” he says. Ultimately, Keaveny says, the method should be able to provide a personalized fracture risk assessment for patients, adjusted for their height and weight and other factors. Clinicians say the approach is exciting but might be prohibitively expensive for screening patients.

In the meantime, clinicians and researchers say much can be done to get more women checked for osteoporosis and give patients a better idea of their fracture risk. In one high-profile effort, a center at the University of Sheffield, U.K., sponsored by the World Health Organization has been designing a method to express a person's absolute risk of fracture during the next 10 years. “This will allow us to have a standard of care,” comments Ethel Siris of Columbia University College of Physicians and Surgeons. “It will give us a better threshold for determining treatment.”