Murderous Minds (22 page)

The human brain is not static. It is always consuming glucose and oxygen at fairly high rates compared to other organs. Depending on the nature and design of their experiments, neuroscientists are sometimes trying to pin down a dynamic system.

As Bennett puts it: “Our brains are constantly changing, and are never the same from hour to hour, day to day, month to month, or year to year. If I give you a cognitive test at 2 p.m., you will perform differently than you would at 2 a.m. Your brain activity will be somewhat different too. You could hypothesize a situation where a psychopathy classifier is very accurate when the subject’s data was acquired at the same time of day as the training data, but would misclassify if the subject’s data were acquired in the wee hours of the morning, or if some other bit of context were different.”

Bennett uses as an example the effect on men’s testosterone levels associated with the success or failure of their favorite sports teams. “As a male, if your team loses then your testosterone level tends to fall lower. Conversely, if your team wins then your testosterone level goes up. If testosterone level was part of a measure of future tendency toward violence, then what you watched on TV a few hours before could make a difference in your sentencing.”

Kiehl and his fellow authors, of course, know that they have much more work to do before they have validated their approach to predicting the chances a freed criminal will break the law again and return to prison. They acknowledge in their paper that future attempts to replicate their findings “should examine the robustness to variations in task, sample characteristics, sample size, anatomical region of interest (ROI), and analytic procedures.”

After the paper was published, Kiehl told Time magazine that “There are more papers coming out that show how MRIs predict who reoffends. We
are examining treatments that increase activity in the anterior cingulate. The goal is to see if we can help identify the best therapies to reduce recidivism.”
²⁶

Flanking Maneuvers

The neuroprediction report from Kiehl’s lab makes another interesting and potentially useful observation about how researchers may be able to progress in their search for clues about criminal behavior, psychopathy, and potential treatments.

It refers to a promising way to attack the challenge of reducing criminal behavior to pages of graphs and illustrations in scientific journals. Finding the biological and genetic bases of mental and personality disorders has been a frustrating experience for researchers. For instance, after years of research and millions of dollars in costs, no clear-cut examples of individual genes solely responsible for complex diseases have been identified. This includes schizophrenia, depression, and bipolar disorder, even though these disorders clearly have strong genetic influences. Even less is known about the nature of psychopathy.

This is one reason in recent years scientists have begun to look for biological markers or traits associated with complex disorders versus searching for particular genes. They call these traits endophenotypes to distinguish them from genotypes, the genes a person is born with. Endophenotypes may be more easily identified, observed, and measured than the complete, complex genetic bases of psychiatric conditions with their complex behaviors.
²⁷
This makes endophenotypes potentially attractive targets for getting at some of the root causes of complex psychiatric disorders.

The same outflanking, divide-and-conquer approach may be a way to dissect complex behavior patterns like criminal behavior and psychopathy. If the ACC can be shown to be underactive in impulsive criminals, it might be looked upon as a neurocognitive endophenotype, a defect or abnormality linked to criminal impulsivity but more readily accessible to study than the criminal behavior as a whole.

Turning Up the Power

The 96 subjects in the neuroprediction study conducted by Eyal Aharoni, Kent Kiehl, and their co-workers in 2013 was large by past standards of
psychopathy research, but more subjects will be needed to convince the scientific community that this preliminary study is valid. Many neurobiological studies of the psychopathic brain, however, have involved far fewer subjects. This might have drawbacks some researchers hadn’t considered, or else hadn’t spent much time thinking about, before Katherine Button and her colleagues pointed them out in a paper they published in Nature Reviews Neuroscience called “Power failure: why small sample size undermines the reliability of neuroscience.”
²⁸
Small sample sizes, the authors remind readers, make it more likely that results that appear to be statistically significant actually are not, and they also increase the chances that true effects are missed.

It is impossible to slide every criminal with a Hare psychopathy score near 30 into an fMRI machine even if you have one ready to go on a semitrailer within a reasonable driving distance. The best researchers can do is sample the population of criminal psychopaths. Then they must use mathematical manipulations and statistics to determine whether or not the effect they see—for example, decreased activity in a part of the brain involved in impulsivity—reflects what they would observe if they could test every criminal psychopath. To do this accurately requires something called statistical power. If a study has enough statistical power, then it is capable of finding a result in a sample that accurately reflects what exists in the entire population (which, as noted, is too large to test and still get home at a decent hour).

The statistical power of a study depends on the sample size and on the magnitude of the real effect. Not surprisingly, studies involving many subjects can pick out smaller effects. Studies involving fewer subjects can’t reliably pick out smaller effects, but they may be able to spot larger effects. Large studies are more powerful than small studies, but are more difficult to organize and cost more.

Button, a psychologist at the University of Bristol in the UK, together with her statistician, geneticist, and neuroscientist co-authors, analyzed 49 research papers in the field of neuroscience. Each of these papers had analyzed and combined results from multiple other studies to determine if a particular finding was real. (Such studies are called meta-analyses.)

They found that “the median statistical power in neuroscience is 21 per cent.” That means half of the studies have a power below 21 per cent and
half above. Low-powered studies miss too much. If the power estimate of 21 per cent is correct, many neuroscience studies could be missing around 80 percent of the real effects. These are called false negatives. Low power also increases the chances of obtaining false positive results; that is, a “statistically significant” effect reported in a small study has a good chance of mistakenly looking real.

This finding doesn’t mean that you can point to any particular paper and say it is poorly designed and provides false results. It means instead that as a whole, the scientific literature in the field of neuroscience is underpowered, and alarmingly so.

Underpowered research studies also make it more likely that any detected real effects will look greater than they actually are. These and other conclusions concerning the drawbacks of small, underpowered research lead Button to conclude: “The current reliance on small, low-powered studies is wasteful and inefficient, and it undermines the ability of neuroscience to gain genuine insight into brain function and behavior. It takes longer for studies to converge on the true effect, and litters the research literature with bogus or misleading results.”
²⁹

These problems, of course, would apply to any field of scientific research that publishes results of studies involving small numbers of subjects. They are not specific to neuroscience or to psychopathy research, although the authors conclude that “small, low-powered studies are endemic in neuroscience.”
³⁰

Between September 2012 and April 2013, nineteen Editorial, News and Analysis, Comment, or Perspectives and Review articles in research journals published by the Nature Publishing Group highlighted “failures in the reliability and reproducibility of published research” in different areas of biological research.
³¹

In the future, the Nature Publishing Group, the publisher of the prestigious journal Nature and other influential publications, will allow authors much more space to describe how they carried out their studies. At the same time, authors who want to publish in Nature journals will have to document and justify more fully how they designed their studies and how they analyzed their results. Statistics will be looked over with more critical
eyes and authors will be encouraged to publish their unprocessed or raw data. This will allow reviewers and other researchers to analyze the results for themselves.

Why did the publisher of Nature scientific journals need to change its publishing guidelines? Their many reasons reflect specific weaknesses in graduate-level education and scientific research in the U.S. One of the most obvious is that hiring, tenure, and grant-review committees have long valued a long resume or curriculum vitae (CV) over a short one. Long CVs mean more publications. It is easier to get a long list of publications by publishing small amounts of data. The researcher who carefully completes a long and powerful project that results in one publication likely will lose out to a researcher who breaks the work up into several papers. Quality often suffers for the sake of quantity in a research environment where grant money is scarce and competition is intense. The fault, of course, lies with the hiring, tenure, and grant-review committee members; that is, with the most influential scientists themselves.

Button’s University of Bristol colleague and co-author Marcus Munafò told an interviewer from the blog Neurobonkers that his “big concern is with the incentive structures that scientists have to work within. We are incentivized to crank the handle and run smaller studies that we can get published, rather than take longer to run fewer studies that might be more authoritative but aren’t going to make for as weighty a CV in the long run because, however much emphasis there is on quality, there is still an extent to which promotions and grant success are driven just by how heavy your CV is.”
³²

The length of a scientist’s list of publications may even be more important than the prestige of his or her alma mater. A close look at the careers of 182 academic biologists suggests that those who published early, before being awarded a Ph.D., and often, are more likely to be successful a decade after graduation than those who delay publishing.

“It doesn’t matter if you go to Harvard or a low-ranked university. If you begin publishing scientific articles when you’re still a graduate student, you are far more likely to succeed in the long run,” first author William Laurance, Ph.D., said in a press release issued by the James Cook University in Cairns, Australia.
³³
And thus the cycle of publishing more but shorter studies versus fewer but longer and more substantial ones is renewed.

A second reason is that researchers receive little encouragement or reward for publishing results that report successful or unsuccessful attempts to reproduce findings from other labs.
³⁴
Researchers also are not encouraged by editors and peer reviewers to tone down the significance of their results, to describe them as exploratory or speculative. Reading nearly any press release issued by a university or research institution describing recently published results shows the emphasis placed on stressing the excitement of the finding rather than the exploratory and less-than-definitive nature of the research. The combination of pressure to publish lots of papers and to hype them has resulted in what the editors at the Nature Publishing Group consider worrisome “failures in the reliability and reproducibility of published research.”
³⁵

The field of psychopathy research is just as reliable and suffers from the same concerns as other fields of neuroscience. Often in the past, and in many cases to this day, researchers out of necessity have had to make do with small numbers of subjects. They, like scientists in other fields, need to look more critically at how they design their experiments and investigations to suit that availability. One reason some reports of brain-imaging studies involving psychopaths have not been accepted by the highly competitive top-tier journals like Nature is because they could not include enough subjects to satisfy the reviewers and/or editors. Kent Kiehl and his colleagues at the University of New Mexico and the Mind Research Network are tackling the problem by using grants from the John D. and Catherine T. MacArthur Foundation, the National Institute of Mental Health, and the National Institute of Drug Abuse to create a database containing thousands of personal and medical histories with matching brain scans of individuals with psychopathy.

Despite the challenges of getting enough brains scanned during an expensive scanning session, psychopathy researchers over the past two decades have nevertheless managed to lay the groundwork for someday explaining the biological reasons one percent of our species differs so much from the other 99 percent in terms of morality. When working toward this goal, identifying specific brain regions that might be targeted for therapy someday is obviously important. So, too, is identifying the brain circuits that might underlie the processes that will explain how the psychopathic mind works.

Chapter Seven

Missing Fear and Empathy

T
HE HAUKELAND UNIVERSITY HOSPITAL
in Bergen, Norway covers a lot of land. The trees lining the streets of Bergen can’t hide its buildings, which cover over two million square feet of space. The single main building alone covers more than 1
¹
/
3
million square feet. The Hospital employs around 11,000 people who treat or assist in the treatment of nearly 600,000 patients per year. If you find yourself in Norway and incur severe burns, develop a brain tumor, or experience decompression sickness or the bends while scuba diving in a fjord, this is where you should ask to be taken for the best specialist care. The Hospital also has a Center for Research and Education in Forensic Psychiatry and a Regional Security Unit. If you are a criminal psychopath and volunteer for a brain scan, you might have it done at Haukeland.