In a post earlier this week on a study raising doubts about some high-profile studies in neuroscience, I was remiss in implying that the problem existed only in fMRI studies. As the paper’s lead author, Niko Kriegeskorte, reminds me, “this is not only about brain imaging (as your title suggests), but equally affects other fields of systems neuroscience,” including EEGs.

I also reported criticism that Kriegeskorte and his colleagues had not listed the studies they found to be problematic, which leaves scientists scratching their heads about what’s reliable and what's (maybe) not. The reason they did not make the list public, Kriegeskorte explains, is that their aim was “to educate, so that an alarming trend can be nipped in the bud before many incorrect claims accumulate in the literature” and not to accuse. “So we decided not to list papers. Every case is different and we could not have done justice to particular studies had we been more specific. We didn’t even want to give a percentage [of how many studies resort to the ‘double dipping’ they criticize], but one of the four reviewers and the editors were adamant about this. . . . We feel that starting a political fight with hundreds of authors is not helpful to our field—especially when most of the studies affected are likely to be correct in their overall conclusion. . . . A few bad apples can and should be listed. But literally thousands of overall good apples, each with a little brown spot, can and should not."

Since the original criticism of fMRI studies earlier this year by Ed Vul and his colleagues, neuroscientists seem to have gotten more aware of statistical pitfalls that can skew their results. But there is no question that passions are running high and that back-biting and defensiveness have set in. At least one attempt to get critics and criticized together in the same room blew up.

Official reports of swine flu cases always lag behind actual cases. It takes time for people to get to a doctor, get diagnosed, and have the results reported to public health surveillance networks. And that assumes people get checked out: illegal aliens and other uninsured people in the U.S. might try to treat themselves and stay under the radar. But you don’t need health insurance or a doctor’s appointment to use Google. Led by research showing that the popularity of certain search terms correlates with a rise in the incidence of flu, the company just published flu estimates for 16 states in Mexico and the country as a whole, as well as U.S. data, in an effort “to help track the spread of the swine flu outbreak,” Google said in a statement.

The effort began last week, when epidemiologists at the U.S. Centers for Disease Control and Prevention asked the researchers at Google Flu Trends if they could track the incidence of swine flu in Mexico. Google Flu Trends, which was launched last November, analyzes the popularity of various search terms to detect a sudden increase in flu cases: historical data have shown that when lots of people start Googling terms such as “flu,” “aches,” “fever” and “buy thermometer,” it correlates with a rise in flu cases. (The researchers described their methodology in a February paper in Nature.) By comparing historical search data with historical data on flu cases, the Google team has been able to filter out search terms—such as plain old “swine flu”—that indicate curiosity rather than an actual case of the flu.

In a blog post today, Google software engineers Jeremy Ginsberg and Matt Mohebbi say “the system has detected increases in flu-related searches in Mexico City (Distrito Federal) and a few other Mexican states in recent days, beginning early in the week of April 19-25,” which was a week before widespread media reports of a swine flu outbreak. The flu incidence is still quite low, with only pockets of "moderate" activity. That suggests that the few hundred officials cases are not the tip of a huge iceberg.

It’s not clear if the Mexican search data are reliable, however. Unlike for the U.S., Google does not have data correlating actual flu incidence in Mexico with the popularity of particular search terms among people in Mexico. The lack of such historical data means the Google researchers “cannot be fully confident that the data is correct,” they say. But “we are cautiously optimistic that the graphs reflect actual flu activity. . . . While we would prefer to validate this data and improve its accuracy, we decided to release an early version today so that it might help public health officials and concerned individuals get an up-to-date picture of the ongoing swine flu outbreak.” Google Flu Trends will update the Mexican data every day.

Interestingly, Google Flu Trends shows low flu activity in the U.S. While it’s too soon to breathe a sigh of relief, that just might be an indication that public health authorities are not missing a huge number of swine flu cases, which as of 11 a.m. today numbered 91 in 10 states with one fatality. Google Flu trends promises, however, that it will “be keeping an eye on the data to look for any spike in [U.S.] activity.”

It’s bad enough when a medication for asthma, hypertension or anything else doesn’t do what it’s supposed to. Even more exasperating is knowing that the way the system of drug discovery and testing is set up, it’s practically guaranteed to produce medications that will not help a lot of the people they’re aimed at.

The reason is that, in order to participate in a clinical trial that the manufacturer runs to test whether an experimental compound is effective, you need to be sick but not too sick. Often, if you have something other than the illness the drug is targeted at, you don’t qualify. Thanks to this and other problems, conclude scientists at the University of Pittsburgh Graduate School of Public Health, the results of clinical studies used to get federal approval for common antidepressants do not apply for most patients with depression.

For their paper in the May issue of the American Journal of Psychiatry, epidemiologist Stephen Wisniewski and his colleagues compared depressed patients who met the criteria for joining a phase III clinical trial (which compares the experimental compound to a placebo) of antidepressants to those who did not. Although the inclusion criteria vary from study to study and are not subject to federal guidelines (the manufacturer gets to decide who’s in and who’s out),  patients with milder forms of depression are typically excluded (they’re more likely to respond to the placebo, making the experimental drug look less effective by comparison). Excluded, too, are patients with chronic depression (they're harder to get a response from, also making the new drug look ineffective) or with additional psychiatric and medical illnesses (who might show dangerous side effects or not respond to the drug).

So, how many patients with depression are thereby excluded? Assessing 2,855 patients treated with citalopram (sold by Forest Labs as Celexa), the scientists find that only 22 percent of people with depression qualify for phase III clinical trials of antidepressants. That means that 78 percent of people with depression are getting the supposedly helpful drug, even though there are no data on whether it is safe and effective in people like them. “This raises major concerns about whether results from traditional phase III studies can be generalized to most people with depression, who also often suffer from anxiety, substance abuse and other medical and psychiatric problems,” said Wisniewski.

The concern is not merely hypothetical. When the scientists assessed how well patients did on antidepressants, they found that those few who would have been eligible for the clinical trial had higher remission rates (34 percent vs. 25 percent for the ineligible patients), and fewer serious side effects and adverse events. The clinical trials that serve as the basis for getting a new drug approved—and that the press trumpets as indicating how effective a drug is—“suggest more optimistic outcomes than may exist for real-world patients” with depression,” said Wisniewski.

Even when you’re skewering an entire field of science, the better part of valor might be to use terms such as “circular analysis” rather than, say, “voodoo.”

The latter is how a team of scientists characterized some findings from brain imaging, as I described in a print column and a previous post about an upcoming paper in Perspectives on Psychological Sciences (but available ahead of print here) by Ed Vul of MIT, Hal Pashler of UC San Diego and colleagues. Its original title of “Voodoo Correlations” in fMRI studies has been replaced by the much-politer “Puzzlingly High Correlations” in fMRI studies, but the message is the same: conclusions from brain-imaging studies of social and emotional aspects of human behavior (jealousy, altruism, social pain and the like) might be wrong and cannot be trusted unless they are re-done with greater statistical rigor.

Now a new study, published last night online for the May issue of Nature Neuroscience, offers an equally devastating critique. Nikolaus Kriegeskorte, Chris Baker and colleagues, of the National Institute of Mental Health, analyzed all the fMRI studies published last year in five top journals (Nature, Science, Nature Neuroscience, Neuron and Journal of Neuroscience). Of the 134 fMRI papers, 42 percent (57 papers) committed a statistical sin at least once: they were guilty of what the NIMH scientists call “double dipping.”

In double dipping, scientists start with a hypothesis that some region of the brain is involved in, say, feeling jealous, and therefore responds to (say) a photo of your romantic rival by becoming extremely active. It’s double dipping, and problematic from a statistical sense, if the scientists then look for those more-active brain regions and analyze only these areas to test the hypothesis. The problem is that brain regions may become more active when they see that photo (compared to seeing, say, a landscape) purely by chance, and analyzing only to these regions would give a misleading result. For the statisticians among you, it’s called non-independent selective analysis.

And it can turn dross into gold. When the NIMH team analyzed “noise”—that is, random data that was known not to show any effect—they still obtained results that seemed to connect a stimulus with a brain response. In other words, double dipping can do wonders for your study. As Kriegeskorte and his colleagues write, the practice “beautifies results, rendering them more attractive to authors, reviewers and editors, and thus more competitive for publication. These implicit incentives may create a preference for circular practices so long as the community condones them.”

As with the Vul et al. critique, this doesn’t mean that all the fMRI results are wrong. The point is, you can’t tell. “To decide which neuroscientific claims hold,” the NIMH scientists write, “the community needs to carefully consider each particular case, guided by both neuroscientific and statistical expertise. Reanalyses and replications may also be required.”

The British Psychological Society has a nice write-up of the new paper here. And Pashler, one of the “voodoo” authors, calls my attention to what he calls a “funny thing about the Kriegeskorte paper” that I didn't notice: unlike his team, the NIMH scientists “didn’t publish the list of which of the 2008 papers were, and were not, afflicted with problems. . . . I can well understand why, since he is a full time fMRI researcher and needs to avoid ticking those people off—something none of us were terribly worried about. And we do know how thin-skinned they are. But there is some tension between the idea of a secret list of bad studies, on the one hand, and the whole notion of science as a public self-correcting enterprise, with a Literature that can be relied upon.”

Science does tend to move at glacial speed, but isn’t it time the fMRI community came to grips with the growing criticism of its methods?

It isn’t every day that one federal agency says the work of another has such “serious scientific flaws” that the work is “not a product [we] should endorse as it does not reach the level of scientific rigor.” Nor is it every day that federal agency #1 (as we’ll call it) says that while federal agency #2 may have tried to get its act together in response to earlier criticism, the work is “essentially unchanged, and . . . [still] scientifically flawed.”

Yet this is how the U.S. Environmental Protection Agency characterizes an analysis by the U.S. Food and Drug Administration. The comments from EPA are in response to an analysis by FDA, issued in the waning days of the Bush administration, that lays the groundwork for a change in federal policy which currently warns people to avoid high-mercury fish and eat lots of the low-mercury kind (a list of what’s high and what’s low can be found here). FDA's analysis would support telling people to eat all the high-mercury fish they want. FDA had received 248 responses by the time the comment period closed earlier this week. (My thanks to the Environmental Working Group, whose own comments on the FDA proposal pointed me to and quote from the EPA response. You can find EWG’s letter here; the docket is a nuisance to navigate around, but if you go to that page and click on the pdf icon, you’ll get EWG’s comment. EPA’s is here; click on the Word icon.)

Here's the backstory. In January, FDA published what’s called a draft assessment of the benefits and risks of eating fish. It asked for public comment. While it’s hard summarize the technically dense, 350-page report, FDA’s conclusions seem to be that mercury risks are very small, and that telling women to eat more fish has greater public health benefits than telling them to eat low-mercury fish. (The focus is on women because fetuses are at the highest risk from mercury. Mercury is a neurotoxin and the developing brains of fetuses and kids are most vulnerable to its poisonous effects.) Next step (though FDA doesn't say this): stop telling people to limit their consumption of high-mercury fish.

Bizarrely, FDA has structured the choice as “either/or:” Eat more fish, or eat low-mercury fish. For reasons I can’t imagine, FDA left out of its analysis the scenario that combines the two: eat more fish, but only low-mercury fish. In every scenario it ran, fish benefits and mercury damage largely offset each other. Obviously, the eat-more-low-mercury fish approach would have greater benefits than doing one but not the other—as researchers (and FDA) have been saying for years. Why would you want to wipe out the brain-healthy effects of fish by having a little neurotoxin on the side?

Predictably, the fishing industry and professional mercury risk-deniers have lauded the FDA analysis. Industry comments describe it as “a comprehensive analysis supporting consumption of seafood.” David Martosko, research director at the Center for Consumer Freedom, wrote in his comments that FDA’s analysis shows that the current FDA/EPA advisory about limiting how much high-mercury fish you eat “has been sending consumers the wrong message.” He urges that the advisory be withdrawn or revised: just tell women to eat more fish, and stop telling them to consider the mercury levels in different fish.

Research scientists and EPA, however, say FDA’s analysis is junk science. For one thing, FDA’s estimate of mercury risk is based on a 22-year-old study in Iraq that observed the age when kids first talk—but the children’s actual ages were unknown. FDA “adjusted” that risk estimate with data from the Seychelles Islands, where the harmful effects of mercury were obscured by the benefits of eating fish. A report from a committee of the National Academy of Sciences concluded in 2000 that the Iraq and Seychelles studies are not the best evidence, and urged the government to use data from a study in the Faeroe Islands. FDA didn’t.

Which is one reason why EPA was so scathing in its comments on its sister-agency’s analysis. It takes FDA to task for dozens of errors and faulty assumptions. One example: FDA relied on research on the neurodevelopmental risks mercury poses which “had been completely abandoned by the scientific community as a basis for risk assessment for more than a decade,” says EPA. EPA’s comments go on to note many other “questionable, faulty or unfounded choices with the effect of boosting benefits or reducing risks from seafood consumption above what is justified scientifically,” EWG says in its comments. EPA concludes that “a fish consumption advisory strategy based on the design of the FDA draft analysis would be highly inconsistent with what is generally considered to be proper public health practice.”

In its soon-to-be 100 days in office (April 29), the Obama administration has begun to clear out the detritus of the junk science and politicized science it inherited from the Bush years. It will be interesting to see whether the idea that it’s fine to eat fish containing high levels of a neurotoxic compound will survive the culling or be recognized for the bad science it is.

One final note: as the battle over what information the public should be getting about fish and mercury rages, the tuna industry might need to watch its flank. In a little-noticed decision this Monday, the U.S. Supreme Court declined to hear an appeal from Tri-Union Seafoods, which makes Chicken-of-the-Sea tuna. Tri-Union sought to block a lawsuit by a New Jersey woman who got methylmercury poisoning from eating its product. The company argued that since the FDA considers canned tuna safe enough to be sold without a warning label, individuals can’t sue over alleged injuries. The Court’s decision lets stand a ruling by an appeals court that the lawsuit can proceed. The plaintiff, Deborah Fellner, ate almost noting but canned tuna for five years, and got a classic case of mercury poisoning, according to her lawsuit. She argues that the company was negligent in failing to warn her of its tuna's high mercury content.

Early birds wake up at the crack of dawn and struggle to stay alert and productive (especially in the cognitive realm) in the evening. Night owls perform well in the evening but are worthless if you yank them out of bed too early in the morning. There are less-well-known differences, too: early birds experience what scientists call “a faster build-up of homeostatic sleep pressure” during the day compared to night owls, who, like a certain battery-powered bunny, just seem to keep going and going, resisting the pressure to sleep. (That must be why I practically turn into a pumpkin by 9 p.m.) And when they do sleep, early birds experience a faster dissipation of that sleep pressure, feeling restored more quickly than night owls. Now a new study, in the journal Science, reports some intriguing differences between the brain-activity patterns of the two types that underlie the behavioral differences.

Scientists led by Christina Schmidt and Philippe Peigneux of the University of Liege in Belgium had 15 extreme night owls and 16 extreme early birds spend two nights in a sleep lab. The two groups were separated by about four hours in their sleep patterns; if early birds were happy waking up at 7, night owls slept til 11, and early birds were ready to go to sleep at 11 while night owls had no trouble staying up til 3 in the morning. An hour and a half after waking up, and again 10.5 hours after waking up, the volunteers had their brain activity measured by fMRI while they took a simple reaction-time test of their ability to maintain focused attention. Both the early birds and the night owls were sleeping and waking whenever they pleased, rather than being kept on an artificial schedule.

There was no real difference between the early birds and the night owls in their performance on the morning test. But the evening test was a different story: night owls were less sleepy and had faster reaction times than early birds. (Just to emphasize, 'evening' was a relative term: it was a different actual time for each group, but the exact same 10.5-hours-after-waking for both early birds and night owls.) So even though both groups were sleeping and waking according to their preferred schedule, night owls generally outlasted early birds in how long they could stay awake and mentally alert before becoming mentally fatigued. The fMRI supported the behavioral results: 10.5 hours after waking up, the early birds had lower activity in brain regions linked to attention and the circadian master clock, compared to night owls.

It’s hard to resist an astronomy discovery when it’s called a blob, even if the precise name is the Lyman-Alpha blob. In a paper being published this afternoon in Astrophysical Journal, astronomers are announcing that they spied such an object—thought to be an enormous body of gas that may be the precursor to a galaxy—dating from when the universe was a mere 800 million years old. Stretching for 55,000 light years (approximately the radius of our Milky Way galaxy’s disk), this Lyman-Alpha blob has astronomers scratching their heads.

Named Himiko for a legendary Japanese shaman queen, the blob is not the largest such object ever discovered. That honor goes to a Lyman-Alpha blob reported in 2006 and thought to be the biggest object in the universe. Instead, this one is notable because it is so far away, and in cosmic terms far away = long ago. “The farther out we look into space, the farther we go back in time,” says astronomer Masami Ouchi of the Observatories of the Carnegie Institution , who led the international team that made the discovery: because light travels at a finite velocity, it takes time for light from objects in space to reach the eyes of Earthlings or their telescopes, which means we are seeing the blob as it was near the dawn of time, when the universe was barely 6 percent of its current age of 13.7 billion years. That means light from Himiko has been traveling toward us for 12.9 billion years, which is equivalent to saying we are seeing it was it was 12.9 billion years ago.

And that makes astronomers a bit uneasy. Whether the blob is ionized gas powered by a supermassive black hole, a primordial galaxy, the collision of two young galaxies or a single giant galaxy with a mass of 40 billion Suns—all of which are on the table—it’s too big for its age. As Ouichi puts it, “I have never imagined that such a large object could exist at this early stage of the universe’s history. According to . . . Big Bang cosmology, small objects form first and then merge to produce larger systems. This blob had a size of typical present-day galaxies when the age of the universe was about 800 million years old.” In fact, other blobs had the decency to wait to show up, appearing when the universe was 2 to 3 billion years old. No extended blobs had been found from when the universe was younger, until Himiko, which means astronomers need to scurry back to the drawing boards to figure out how an object this massive managed to grow up so fast.

There’s all sorts of mumbo-jumbo about how sports drinks boost athletes’ performance, especially in endurance events such as yesterday’s Boston Marathon. But according to an intriguing new study, it isn’t the sports drinks’ calories (athletes benefit even if they spit out the drink rather than swallow it) or their sweet taste (drinks with artificial sweeteners do not boost performance). Instead, suggest Ed Chambers of the University of Birmingham and colleagues in a paper in The Journal of Physiology, carbohydrates in the drinks fit into receptors in the mouth that in turn activate the brain’s pleasure and reward centers, spurring athletes to push themselves harder without realizing how hard they're working.

For their study, the scientists prepared drinks containing either glucose (a sugar), maltodextrin (a tasteless carbohydrate) or plain water, mixed with artificial sweeteners so they tasted identical. Eight endurance cyclists rinsed their mouths for 10 seconds with one of the three drinks, and then got on a stationary bike. Results: athletes who swished with the glucose or maltodextrin drinks outperformed those on sweetened water by 2 to 3 percent, raising their pulse and sustaining a higher average power output—even though they said they didn’t feel they were working harder.

Chambers explains it this way: “Much of the benefit from carbohydrate in sports drinks is provided by signalling directly from mouth to brain rather than providing energy for the working muscles.”

That was born out by neuroimaging. Using fMRI to monitor brain activity after the athletes rinsed their mouths with one of the three drinks, the scientists found that glucose and maltodextrin increased activity in regions associated with reward or pleasure (the anterior cingulate cortex and striatum). The artificially sweetened water did not. They propose that the sugar or carbohydrate glommed onto receptors in the mouth, causing a signaling cascade that activated these brain regions, with the result that the athletes felt they were not working as hard as they actually were—contributing to endurance and power output. Once again, it seems as if the brain, not the muscles, ultimately govern how well we do even on what seems to be a purely physical task. As the scientists put it, “carbohydrate in the human mouth activates regions of the brain that can enhance exercise performance.”

For patients suffering from locked-in syndrome, in which they are completely paralyzed and able to do no more than blink their eyes, the greatest hope is not walking, not feeding themselves, not anything else having to do with moving: it is communicating. (An episode of House last month did a good job of depicting the horror of locked-in syndrome, which can be caused by amyotrophic lateral sclerosis, aka Lou Gehrig’s disease, brain-stem stroke or high spinal cord injury.) Hence the intense research effort to build brain-computer interfaces (BCI) for such patients. As a 2007 publication from the National Institutes of Health described a BCI system being developed there, “eight electrodes hitched to the computer . . . record the user’s electrical brain waves, which the computer analyzes and translates into specific commands, such as writing emails, selecting computer icons, or moving robotic devices. No surgery is required and users typically master the system within an hour or two.”

Writing emails is all well and good, but now brain-computer interfaces have made the big leagues: a BCI has been used to Tweet. Earlier this month Adam Wilson, a graduate student in biomedical engineering at the University of Wisconsin-Madison, sent “using EEG to send tweet.” He used what has become the standard methodology, in which EEGs pick up electrical signals from the brain and translate them into movements of a cursor, in this case on a screen with the 26 letters of the alphabet, as the scientists show in this video.

“The way this works is that all the letters come up, and each one of them flashes individually,” says Justin Williams, a UW-Madison assistant professor of biomedical engineering and Wilson’s adviser. “And what your brain does is, if you’re looking at the ‘R’ on the screen and all the other letters are flashing, nothing happens. But when the ‘R’ flashes, your brain says, ‘Hey, wait a minute. Something’s different about what I was just paying attention to.’ And you see a momentary change in brain activity.”

Although it’s a tedious process—Wilson likens it to texting, when you may have to press a key four times to get the desired character—people get better with practice. “I’ve seen people do up to eight characters per minute,” he says. Which just goes to show that Twitter is not the civilization-ending toy that so alarms some people: locked-in patients may be able to use it to give friends and families status updates merely by thinking.

You can always count on studies of daycare to scare the living daylights out of parents, especially when they find that the more hours kids spend in daycare the more likely they are to be aggressive (a conclusion that, critics said, reflected shortcomings in the study) and that poor-quality daycare can hinder kids’ cognitive development, as the original report of a long-running study and a more user-friendly write-up both note.

These findings and more have emerged from the longest-running and most comprehensive study of daycare, the Study of Early Child Care and Youth Development, which the National Institute of Child Health and Human Development began in 1991. But there has always been something funny about these findings, namely, the small size of the effect of daycare. It raises the possibility that daycare does have strong effects on kids, but that the effects differ depending on the child, with the result that when you take an average over thousands of kids in a study you wash out the individual effects.

Psychologist Jay Belsky suspects something like that is going on. Now at Birkbeck College in London (he also blogs for Psychology Today), he suggests that children have what he calls a “differential susceptibility” to their environment, including the environment called childcare. (I wrote about his work in a story on how children’s genetic differences affect how they’ll respond to various forms of parenting, such as learning from mistakes.) According to this model, the reason studies don’t find a greater effect of child care is that they mix apples and oranges: the apples are children who are susceptible to the effects of daycare and the oranges are children who are not. As Belsky and colleague Michael Pluess put it in a paper in the April issue of the Journal of Child Psychology and Psychiatry, “Inconsistencies regarding developmental effects of non-maternal childcare may be caused by neglecting the possibility that children are differentially susceptible towards such experiences.”

Using the NICHD data, they therefore looked for relationships between a child’s temperament (which is under at least partial genetic control), childcare and outcomes when the children were about four-and-a-half. Their conclusion: “children with difficult temperaments as infants exhibited both more behavior problems when faced with low quality care and fewer when experiencing high quality care than children with easy temperaments.” These difficult children—hard to soothe, cranky, sensitive—really feel the effects of daycare, for good or ill. That doesn’t mean you shouldn’t put such a child in daycare. Look carefully at the last part of the quote above. Although these “negatively-emotional infants” are apt to emerge from poor-quality daycare at age 5 with more social and behavioral problems that children not in daycare, they have fewer such problems than the stay-at-home kids when they receive high-quality daycare. Mellower children are like Teflon: the effects of daycare, of high or low quality, simply do not stick to them as much.

Message: know your child. Know your daycare.

There is no clearer evidence of how controversial geo-engineering (altering the atmosphere so as to reduce global warming, perhaps by lofting a haze of sulfate aerosols into the stratosphere to reduce incoming sunlight) is than the tempest stirred up when White House science adviser John Holdren told the Associated Press that the administration was discussing it. To lots of people concerned about global warming, merely mentioning geo-engineering detracts from the urgency to reduce emissions of greenhouse gases, conveying a message of, “oh, no problem; we can keep making this mess but deploy a techno-fix when we need to.” But let’s leave politics aside for a nanosecond. The risk of geo-engineering is, to put it bluntly, that we’re not smart enough to know what climate effects it will produce.

A 2007 study has already shown that using sulfate aerosols to reduce incoming sunlight, and thereby cool the planet, would also mess up precipitation patterns enough to trigger serious droughts. Now a new study pinpoints what can only be described as a truly ironic side-effect: lofting sulfates into the stratosphere would also reduce the effectiveness of solar power, the renewable, zero-carbon energy source that has the best chance of averting catastrophic climate change.

Writing in Environmental Science & Technology, Daniel Murphy of the National Oceanographic and Atmospheric Administration explains that while enhancing the stratospheric aerosol layer makes some of the incoming sunlight bounce back to space (producing cooling), it scatters even more of the sunlight. For every 1 watt of sunlight reflected away from the Earth, another 3 watts of direct sunlight are converted to diffuse sunlight. Unfortunately, it is direct, rather than scattered, sunlight that large concentrated-solar-energy facilities need in order to reach maximum efficiency; they can’t use diffuse light. (Concentrating systems focus sunlight onto photovoltaic cells, to produce electricity directly, or onto tubes to produce steam or a hot fluid.) Put another way, every 1 percent reduction in sunlight due to aerosols causes a 4 to 10 percent loss in output from concentrated-solar collectors.

Flat solar hot water and photovoltaic panels would experience less of a loss, because they can use diffuse sunlight. But “the performance loss will still exceed the reduction in total sunlight because a tilted panel does not capture diffuse sunlight as efficiently as direct sunlight,” notes Murphy. Bottom line: “a significant reduction in the efficiency of solar power generation systems.”

It’s long been a puzzle to me why so many of my fellow commuters carry umbrellas when the weather forecast has called for, say, a 20 percent chance of rain. Me, I figure the odds are 4-to-1 in my favor. I assumed that other people are risk-averse—that is, even a small chance of getting drenched is worth the cost of carrying an umbrella. But a new study suggests something else is afoot: many people have no idea what “20 percent chance of rain” actually means.

Many people think it means that rain will fall over 20 percent of the area covered by the forecast, in which case people commuting to Manhattan from the outer boroughs, New Jersey, Connecticut or Westchester probably figure that they’ll get wet either coming or going, according to a study led by cognitive psychologist Susan Joslyn of the University of Washington (in Seattle, where they know something about rain). Others think “20 percent chance” means it will rain for 20 percent of the time period covered by the forecast, she and her colleagues report in the Bulletin of the American Meteorological Society. Only about half the population knows that “20 percent chance of rain tomorrow” means “it will rain on 20 percent of the days with exactly the same atmospheric conditions,” Joslyn said.

For their study, the scientists had 450 college students answer questions about probabilistic forecasts of precipitation. Asked how much of the time it would rain on the day being forecast and over what area, only 43 percent of the participants correctly answered, “can’t tell from this forecast.” Only when the participants were shown a forecast explicitly giving the chance of rain (20 percent) and the chance of no rain (80 percent), or a pie chart icon showing the chance of rain (as in this “probcast” the researchers developed), did more than half (52 percent) of the students know that the forecast had nothing to do with how much of the time or over what area it would rain.

If you misinterpret a probabilistic forecast to mean that the rain, snow or storm will definitely occur (in some percent of the area or for some percent of the time), you’re more likely to take precautions than if you understand that the event has a less-than-certain chance of occurring. If the precaution is just toting an umbrella, there’s no huge cost. But if it’s evacuating before a hurricane, closing schools before a snowstorm or the like, then the consequences of statistical illiteracy are much greater.

Literary agent and Edge impresario John Brockman is reporting that John Maddox, long-time editor of the journal Nature, died Monday night at the age of 93. Maddox, who trained as a physicist, edited Nature for 23 years, making the publication arguably (well, I would argue) the best scientific journal in the world. Brockman has a nice Q&A with Maddox from 1997, in which Maddox is his usual irascible self.

I’ll most remember Maddox for two things. In his fierce commitment to empirical reasoning, in 1988 he arranged for one magician (James Randi) and one science-fraud investigator (Walter Stewart) to observe an experiment in which French scientist Jacques Benveniste claimed, essentially, that water could retain a “memory” of compounds it had been exposed to, which is the basis for homeopathy. The compound was an antibody, and the claim was that even when the antibody had been diluted billions of times, it still had an effect on white blood cells. (Randi describes it here.)

Maddox had allowed the original claim to be published in the June 30 issue of Nature, because it was receiving coverage in the popular press. But his, Stewart’s and Randi’s investigation debunked it. The headline of the Nature story describing the result of the investigation, in July 1988, says it all: “’High-dilution’ experiments a delusion.” Benveniste claimed he was the victim of a witch hunt. The New York Times archives’ has a nice write-up of the whole affair here.

But I will also always recall the many editorials that Maddox wrote or ran casting doubt on anthropogenic global warming. As he told Brockman in their 1997, “I accept that global warming, because of carbon dioxide, is going to be a reality at some stage in the future. I disagree with the way in which the forecasts have been made by the organization called the Intergovernmental Panel on Climate Change. . . the real problem is that all this is based on computer modeling, and . . . I think it’s dangerous to rely on computer modeling when you are trying to make predictions about the real world.” It’s not just that this is wrong from the standpoint of 2009; it was also wrong in 1997. I’ve always wondered to what extent scientific and political efforts to come to grips with climate change were thwarted by the skepticism of the world’s leading science journal. Nevertheless, science will miss him.

With two seders behind us, Good Friday upon us and Easter two days away, let us sing the praises of religion for its power to improve mental health (pace, Hitchens and Dawkins). Yes, “religion features in a lot of psychotic delusions,” as British medical writer Tom Rees notes in his blog Epiphenom. But as two new studies show, religion can also improve mental health--but only if you believe in the right god.

In a paper in the online edition of the Journal of Religion and Health, psychologist Kevin Flannelly of the Spears Research Institute, HealthCare Chaplaincy and colleagues analyzed the link between particular beliefs about god and psychiatric symptoms in 1,306 adults in the U.S. They used data from a 2004 survey asking people what they thought about god as well as asking about their mental health. For the first part, the survey asked whether the god they believed in was close and loving, approving and forgiving, or creating and judging. Eliminating people who professed no religion (3 percent), Flannelly and his team then compared the incidence of general anxiety, depression, obsessive-compulsion disorder, paranoid ideation and social anxiety in the three groups.

If your goal is mental health, they found, it’s a whole lot better to believe in a close and loving god. People with that belief had significantly lower rates of anxiety, depression, obsessive-compulsion and paranoid ideation than people who believed in an approving and forgiving god. Belief in that sort of god reduced incidence of those disorders slightly. Belief in a creating and judging god raised the risk of all those mental disorders, especially of paranoid ideation. Well, sure: if you think an omnipotent and omniscient god is watching your every move and will smite you for any infraction, that can indeed make for a bit of paranoia.

Before we jump to causal conclusions here, however, it’s important to keep in mind that pre-existing mental states, including a tendency toward forms of mental illness, might predispose people to believe in one kind of god rather than another. (Yes, the family and culture you’re raised in has a strong effect on your religious beliefs, but as people reach adulthood they tend to customize their faith.) To be blunt, it’s possible that psychotic people gravitate toward a stern, judgmental god of the Old Testament stereotype (Sodom, Gomorrah et al.) while people without psychiatric problems gravitate toward a kinder, loving god, and not that particular religious beliefs make you more or less psychotic.

But in a paper in the same journal, Flannelly and colleagues argue that the arrow of causation might indeed run from belief system to mental health. They hypothesize that the human brain contains what they call an Evolutionary Threat Assessment System that alerts you to threats in the environment. If this system malfunctions, identifying dangers that are not there, it may produce symptoms of paranoia, depression and OCD. In contrast, a mellower alert system should decrease these symptoms. Belief in a close and loving god who’s looking out for you is one way to decrease activation of the threat-assessment system, they argue. Result: less chance of becoming paranoid, depressed, anxious and the rest.

Well, who are we here at Lab Notes to judge? Whatever gets you through the night. And up the next morning.

When I was working on a 2007 column and then a longer story last year about why people believe in paranormal and supernatural phenomena, one of the most insightful scientists I spoke to was psychology researcher Bruce Hood of Britain’s University of Bristol. As he does in his blog, he explained that believing in ghosts, ESP, telepathy or even that you can tell when someone is looking at you from behind is not a matter of religion or culture, but instead reflects the normal workings of the brain.

Now Hood has a new book called Supersense that delves into “why we believe in the unbelievable,” as his subtitle puts it. “Humans are born with brains designed to make sense of the world and that sometimes leads to beliefs that go beyond any natural explanation," he argues. "We are inclined from the start to think that there are unseen patterns, forces and essences inhabiting the world. . . . This way of thinking is unavoidable, and it may be part of human nature to see ourselves connected to each other at this deeper level.”

The numbers alone support his argument that many people—a lot of them smart and highly educated—believe in supernatural phenomena. Polls show that 73 percent of U.S. adults believe in at least one supernatural phenomenon (41% in ESP, 37% in haunted houses, 32% in ghosts, 31% in telepathy, 26% in clairvoyance, 21% in communicating with the dead). And superstitions are rampant. Tony Blair wore the same pair of shoes to answer Parliamentarians’ questions. John McEnroe made sure not to step on the lines as he walked on or off a tennis court during a tournament. John McCain carries a lucky feather and lucky coins. Barack Obama, having played basketball on the morning of the day he won the Iowa caucuses, played every other primary and caucus day thereafter. These and other “secular rituals,” as Hood calls them, show that “everyone is susceptible to supernatural beliefs.”

These beliefs are supported by what we think are personal experiences of the supernatural—having a premonition that someone is about to call and then hearing their voice when we answer the phone, for instance. Indeed, people cite such first-hand experiences of the supernatural, not religious teachings, for their belief in it. And once we believe, we find ever-more evidence to support that belief, due to “confirmatory bias,” the well-established psychological phenomenon of noticing and remembering events that confirm a belief and forgetting those that challenge it. Don’t you remember the times your premonitions were right (besides that phone call, maybe that something bad would happen to someone) and forget the times you were wrong? I bet McEnroe would vividly remember any time he lost a match because he inadvertently stepped on a line.

The core of the book is an examination of the brain processes that underlie that susceptibility, the “mind design” in which supernatural beliefs originate. For instance:

• As children (and often as adults), we believe that minds and bodies are separate entities, not that the mind is what the brain does. From there, it is a short step to belief in telepathy and spirits, minds that are not tethered to physical bodies.
• We have brain circuitry dedicated to perceiving faces, which stands us in good stead when we need to recognize mom but also causes us to see Jesus in a grilled cheese sandwich. That exaggeration of a useful brain function, argues Hood, is adaptive: “If you are in the woods and suddenly see what appears to be a face, it is better to assume it is one rather than ignore it. It could be another person out to get you.”
• Similarly, babies tend to treat moving objects as if they had purpose, rather than as following, say, a gravitational trajectory. Children “naturally assume that the living world is permeated by invisible life forces [and] energies,” notes Hood. That lays the groundwork for imputing aspects of mind and intentionality to inanimate objects. From there it is a small step to believing that the sweater of a killer, in one of Hood’s favorite examples, is imbued with the murderer’s spirit. “This way of thinking emerges early and may support a supersense that there are secret agents operating throughout the world,” he writes.

I disagree with Hood when he argues that a sentimental attachment to objects reflects superstitious beliefs, namely that the essence of the person or place resides in the article of clothing or other memento. For many of us, the object is a powerful reminder of someone or someplace, not something spooky. Nor am I convinced that collectors of memorabilia are motivated by a spooky sense of supernatural connectedness—the “tendency to see objects as possessing invisible properties that originate from significant individuals,” in Hood's words—rather than a simple desire to be reminded of something. (“Few things are more irrational that the human obsession for collecting,” he argues unpersuasively.) And Hood is too kind to supernatural beliefs when he credits them with enabling us to share the “sacred values” that bind together our societies. Given recent events, I think we could use fewer sacred values (jihads, promised lands . . . ) and more values rooted in justice, compassion and altruism.

Full confession: I’m a sucker for examples of how sensory input changes the brain. And if the changes alleviate a problem that can range from annoying to devastating, extra points.

A study of a treatment for tinnitus therefore caught my eye. According to results presented last week at the American Academy of Audiology, a device called Neuromonics and made by a company of the same name brought measurable improvement. On the standard evaluation of how incapacitating the ringing in the ear was, the device reduced patients’ psychological and emotional misery from 46 to 20 after six months of treatment, scientists led by the Cleveland Clinic’s Sharon Sandridge reported. It was a small study (just 45 patients), but intriguing nonetheless.

Tinnitus affects something like 50 million Americans. Lots of things can cause it, including repeated exposure to loud noise, but what’s interesting is that the problem is in the brain, not the ears. The Neuromonics device therefore targets the auditory cortex. It delivers a special sound (determined by the frequency of the ringing a patient reports) embedded in soothing music. Listened to daily, the sound seems to rewire the auditory cortex, training the brain to filter out the internal sounds so the tinnitus doesn’t intrude on consciousness.

There are as many explanations for why scratching relieves itchiness as there are causes of itching, with some of the favorites being that scratching releases painkilling endorphins or distributes itch-causing histamines so the high local concentration is diffused. (A New Yorker article last year explored the world of itching and scratching so thoroughly you’ll need calamine lotion after reading it). But a new paper in Nature Neuroscience makes a good case for a dark horse explanation: scratching decreases activity in some spinal cord neurons that transmit the itch sensation to the brain.

Although the physiological mechanisms for how scratching relieves itch are poorly understood, scientists have at least figured out that neurons in a specific part of the spinal cord—the spinothalamic tract, or STT—are more active when we itch, transmitting that information to the brain. To see if scratching acts on those neurons, neuroscientist Glenn J. Giesler of the University of Minnesota at Minneapolis and colleagues injected histamine into monkeys (they used crab-eating macaques, or Macaca fascicularis). They found that scratching the skin blocks the activity of spinothalamic neurons but, interestingly, only when the skin is itchy as a result of the histamine. That suggests, the write, that “scratching inhibits the transmission of itch in the spinal cord in a state-dependent manner." The “state-dependent” qualifier means that the activity of STT neurons was not reduced by scratching if the monkey did not itch.

Instead, the reduced activity in the neurons as a result of scratching occurred only when the skin was itchy and the neurons were carrying that information to the brain. Scratching blocks the transmission of that "I itch!" signal, bringing relief.

Although colleagues have from time to time wondered if I’m a witch (this started when I wrote a column saying the full moon is not associated with weird stuff, and then a few days later the 2004 Asian tsunami hit—during a full moon), I am definitely not psychic. So when I wrote the column in the current issue on screening for cancer, and the limitations of early cancer detection, I did not know that a study would appear yesterday evening with the enticing title, “What is the point: will screening mammography save my life?”

Unfortunately, it concludes that screening mammography (that is, giving mammograms to women with no symptoms of breast cancer and no elevated genetic risk for the disease) makes so little difference when it comes to cancer mortality—and really, what does anyone care about when it comes to cancer except not dying of it?—that the average benefit of a single screening mammogram is minuscule. For a woman of 40, the researchers calculate, the survival percentage is 99.52 percent if she does not undergo a screening mammogram and 99.62 percent if she does.

For their study in the open access journal BMC Medical Informatics and Decision Making, John D Keen of the John H. Stroger Jr. Hospital of Cook County and James E Keen of the University of Nebraska define “life saved” to mean that screening identified and helped cure one woman with breast cancer who would have died from the disease without screening. They then analyze U.S. data to estimate several things: the risk of dying from breast cancer if a woman does not undergo screening mammography, for instance, as well as the separate effects on breast-cancer mortality of screening mammography and improved therapy, and the survival percentages without and with screening. As you can tell, this is basically a mathematical analysis, not another clinical trial of mammography, and people will certainly argue with some of the assumptions the researchers used for their calculations (more on that below).

The numbers that popped out of the calculations were less than encouraging. In addition to the findings above, they report that “repeated screening starting at age 50 saves about 1.8 lives over 15 years for every 1000 women screened. . . . The average benefit of a single screening mammogram is 0.034%, or 2970 women must be screened once to save one life.” Overall, “Less than 5% of women with screen-detectable cancers have their lives saved” by screening mammography.

The mortality reduction with screening mammography is so low, in part, because for women under 65 the risk of dying of breast cancer is low to begin with: 1 percent over 15 years starting at age 55. With numbers that low, mammography would have to be highly selective and specific, and cure rates much higher with cancers detected early rather than late, to make a substantial difference in mortality. None of these is the case.

Last year, another study by the two scientists concluded that screening mammography has much greater life-saving benefits among older women. In the current study, that finding is replicated: the life-saving benefit of mammography increases with age. Repeated screening starting at age 50 saves about 1.8 lives over 15 years for every 1000 women screened.

The journal calls the study “highly controversial,” which is surely an understatement. For one thing, more than 90 percent of women believe that “early detection saves lives,” notes the journal. As I said in my column, people cling desperately to the notion that they can do something—namely, undergo cancer screening—to keep cancer from taking their life.

Because the study is likely to incite debate, the journal solicited two commentaries on it. In one, Stephen Duffy of Cancer Research UK, a British cancer-awareness group, argues that empirical data might be more trustworthy than the estimates the Keens derived by combining data from disparate sources. (A 2003 study that Duffy led found that mammograms did reduce breast cancer mortality.) He questions the accuracy of their numbers, noting that clinical trials have found that screening 1,000 women can prevent 2.1 (in a Swedish study) to 3 deaths from breast cancer (not the 1.8 the Keens calculate). And while the Keens calculate that fewer than 5 percent of women who undergo screening mammography have their lives saves, says Duffy, a Swedish study found that 141 breast cancer death were averted in a mammogram study that detected 928 cancers, for a mortality benefit of 15 percent.

In the opposing corner is Michael Retsky of Harvard Medical School, whose research has raised questions about the value of screening mammography, especially for women under 50. He writes in his commentary that the Keens’ study is on the right track in that women under 50 who undergo screening mammography face very real risks (of false positives and the resulting biopsy) for little benefit.

For another view on screening, however, I highly recommend the blog post of the American Cancer Society’s “Dr. Len”—J. Leonard Lichtenfeld, whose post on the PSA test I mentioned yesterday. Writing on the benefits of cancer screening in general, he argues that “comments that screening for cancer isn’t the answer to reducing deaths from every cancer are not new. Also untrue is the implication that those of us who advocate screening for certain cancers aren’t aware of the risks of screening. . . . Yes, we do believe that the scientific evidence shows that mammograms save lives. . . I remember what it was like to treat women with ‘early’ breast cancer in the days before effective mammograms were available, when a lump was palpated by the woman or her doctor and we told her it was ‘early’ cancer. Many of those women went on to die from their disease.”

As the debate continues, what we need is less fighting about whether screening mammograms save 5 percent or 15 percent of women who have them, and more progress on finding a screening method—paired with effective treatment—to radically boost those numbers.

If you prefer to keep a little magic in your life—by which I mean believing in the possibility of UFOs—then read no further. For I am going to tell you about the latest UFO hoax.

You may remember the sightings of a UFO over Morristown, N.J., in January, which was blogged about and even captured on video that has been posted to YouTube as clips from TV broadcasts and an amateur astronomer.

It was all a hoax, as the perpetrators reveal in this month’s issue of eSkeptic.

Last November, Joe Rudy and Chris Russo, two 20-somethings, were sitting around discussing pseudoscience and the many people who believe one or another form of it. Rudy describes himself as “an avid reader of Skeptic magazine” who teaches science and gives private music lessons. Russo works in sales and says he “intends to continue his quest to spread reason and truth, one pseudoscience at a time. “We had always had a strong interest in why people were so easily fooled by such irrational superstitions as psychic ability, spiritual mediums, alien abductions, and the like,” they write. So they “set out on a mission to help people think rationally and question the credibility of so-called UFO ‘professionals’.”

They cooked up a spaceship hoax “to show everyone how unreliable eyewitness accounts are, along with investigators of UFOs.” They used five feet of fishing line to tie flares to each of five three-foot helium balloons and launched them from a field on Jan. 5, 2009. “Once all five balloons were ready for takeoff (with our fingers on the verge of frost bite),” they write, “we struck the 15-minute flares and released them into the sky in increments of fifteen seconds,” filming the UFOs as they floated away.

Media coverage was extensive. A lot of it featured Paul Hurley, a pilot, and his family, who appeared on several news broadcasts describing the strange lights they saw in the sky. (For some reason, reporters find pilots’ UFO sightings especially believable.) Rudy and Russo repeated the performance four more time, gaining media coverage for each. Conspiracy Web sites and radio shows covered the sightings, but “the icing on the cake came when the popular History Channel show UFO Hunters featured the Morristown UFO one week,” the duo recall. “Bill Birnes, the lead investigator of the show and the publisher of UFO Magazine, declared definitively that the Morristown UFO could not have been flares or Chinese lanterns.”

This was the pair’s main quarry, to expose the foolishness of UFO “investigators.” They write: “Are UFO investigators simply charlatans looking to make a quick buck off human gullibility? ... If a respected UFO investigator can be easily manipulated and dead wrong on one UFO case, is it possible he’s wrong on most (or all)\of them? Do the networks buy into this nonsense, or are they in it for the ratings?”

You can see their handiwork here and here. Nicely done, guys.