Last July, researchers at UCLA reported that the brains of people diagnosed as "hypersexual" do not display distinctive responses to sexual images. The study's authors interpreted their findings as evidence that hypersexuality, defined as a harmful and distressing preoccupation with sex, is not a true addiction or mental disorder.

A week after the UCLA study was published, Ariel Castro pleaded guilty to 937 criminal charges stemming from his kidnapping of three women whom he imprisoned and repeatedly raped over the course of a decade. "I'm not a monster," he insisted at his sentencing hearing. "I'm just sick. I have an addiction, just like an alcoholic has an addiction."

Around the same time, disgraced former congressman Anthony Weiner, who resigned in 2011 after admitting he had "sexted" photos of his genitals to various women, was attempting an improbable political comeback by seeking the Democratic nomination for mayor of New York City. His campaign suffered a major setback when it emerged that Weiner, a married man, had continued conducting virtual liaisons with Internet acquaintances even after quitting Congress.

All of these stories raise the question of how we should view people who don't control their sexual urges. Are they addicted to sex in the same way that an alcoholic is bonded to booze or a junkie is hooked on heroin? If so, does that mean they are no longer responsible for their behavior? And to what extent does neuroscientific evidence such as the electroencephalograms (EEGs) used in the UCLA study illuminate these issues? The debate about the nature of hypersexuality exemplifies a broader cultural confusion about addiction, a concept that has been medicalized to the point that people think it means both less and more than it really does.

If I Only Had a Brain Scan

The UCLA study, reported in the journal Socio­affective Neuroscience and Psychology, involved 52 volunteers whose scores on tests of sexual compulsivity were similar to those of "people seeking help for hypersexual problems." The researchers, led by the neuroscientist Nicole Prause, hooked the subjects up to electroencephalographs and showed them pictures that ranged from romantic to sexually explicit. Prause and her colleagues found that the brain responses registered by the machines were related to the subjects' self-reported levels of sexual arousal but not to their hypersexuality scores. "Hypersexuality does not appear to explain brain responses to sexual images any more than just having a high libido," Prause said in a university press release.

The researchers presented their results as evidence that hypersexuality is not really an addiction. "A diagnosis of hypersexuality or sexual addiction is typically associated with people who have sexual urges that feel out of control, who engage frequently in sexual behavior, who have suffered consequences such as divorce or economic ruin as a result of their behaviors, and who have a poor ability to reduce those behaviors," the press release noted. But according to Prause and her colleagues, "such symptoms are not necessarily representative of an addiction-in fact, nonpathological, high sexual desire could also explain this cluster of problems."

For someone unblinded by neuroscience, it is hard to make sense of this explanation. Sex addiction is defined by a pattern of behavior, a pattern exhibited by the subjects in this study, who reported that their sexual preoccupations and actions were so unrestrained that these had gotten them into trouble at home or work. Are these problems less real because their EEG tests were not abnormal? If the essence of addiction is a life-disrupting attachment that continues despite undesirable consequences, why does it matter whether Prause's subjects displayed a special "P300 response"?

Prause's reliance on EEG data to cast doubt on the reality of the problems reported by her subjects is an example of what I call "neuroreductionism": the tendency to perceive human experiences as valid or genuine only when they can be linked to measurable brain activity. In their recent book Brainwashed: The Seductive Appeal of Mindless Neuroscience, Oasis Clinic psychiatrist Sally Satel and Emory University psychologist Scott Lilienfeld detail the faulty science and logic underlying this mindset.

Neuroreductionism is the latest twist on the medicalization of addiction. Medicine has come up with wave after wave of biological explanations for addiction, ranging from the belief that narcotics such as heroin have special chemical effects that cause addiction to wild claims implicating the endorphins as addictive culprits. All these theories posit that addiction is caused by the things to which people become addicted.

My view, by contrast, is that people can become addicted to anything, whether drugs, alcohol, food, shopping, gambling, love, or sex, if it is the focus of an encapsulating experience that alleviates bad feelings and buttresses their self-esteem. Contrary to the common view of addiction as a choice-nullifying disease, this approach holds people accountable for their actions. Addicts are actively involved in building their attachments and can modify their behavior when they have an incentive to do so. Alcoholics drink moderately at home with their parents, for instance, and addicted smokers wait all morning during work until they can smoke outdoors. They might prefer to indulge their addictive impulses instantly, but those impulses can be resisted and ultimately eliminated.

Likewise, many highly sexual people do not act on their urges in ways that create problems. Even among those who qualify for the "hypersexual" label, the vast majority do not kidnap women and rape them. Since sex addicts clearly can conform to social expectations, legal limits, and moral principles, it is right to punish those who, like Ariel Castro, egregiously fail to do so.

Anthony Weiner's transgressions were trivial in comparison with Castro's. But his evident lack of self-control still prompted debate about exactly what his problem was.

"The candidate's behavior meets a fundamental criterion for addiction: his exhibitionist acts continued despite negative consequences," Time's health writer Maia Szalavitz claimed. "It's hard to imagine a better example of compulsive repetition: although he lost his job and put his marriage, family and entire political future at risk, the former Congressman nonetheless engaged again in the exact type of online behavior that brought him to public humiliation."

Yet Szalavitz was not ready to call Weiner an addict. "It's still not clear whether sexual compulsions qualify as an addiction," she said. "The [UCLA study] argued that they don't, because hypersexual people process sexual cues just like normal people do-and differently from the way addicts respond to drug cues. But the question is far from resolved. And whatever you want to call the issue, Weiner still has a problem, since compulsions can be just as disruptive as addictions, and equally difficult to overcome."

Szalavitz is at odds with herself. She says Weiner-who displayed the defining characteristic of addiction, persistence in destructive behavior-may not have been truly addicted because scientists found that sex addicts don't have special brain responses to sexual stimuli. Instead, Weiner may suffer from a destructive sexual compulsion. But that is a distinction without a difference. Szalavitz, who is as well-informed on addiction and drugs as anyone who has ever written for a mainstream news outlet, still gets tripped up by neuroreductionism. According to this way of thinking, if a person behaves like an addict but doesn't produce the right brain scan, he has to be called something else.

Before Cocaine Was Addictive

Although addiction historically referred to just about any strong attachment or hard-to-break habit, physicians began using the term to describe a medical syndrome at the turn of the 20th century. They had in mind addiction to drugs. Specifically, they had in mind addiction to opiates, since they erroneously believed that only opiates caused tolerance (a need for larger doses to achieve the same effect) and "physical dependence" (withdrawal symptoms). Alcohol and barbiturates were added to the list later, but physicians and pharmacologists continued to describe other drugs as merely "habituating." The U.S. surgeon general's landmark 1964 report on the health hazards of smoking, for instance, deemed nicotine habituating but not addictive. The same distinction was applied to cocaine, amphetamines, marijuana, and various other drugs.

By the late 1980s, the federal government had changed its mind: The 1988 surgeon general's report on smoking was titled Nicotine Addiction. The American Psychiatric Association likewise broadened its definition of addiction, noting in the fourth edition of its Diagnostic and Statistical Manual of Mental Disorders (DSM-IV, published in 1994) that "neither tolerance nor withdrawal is necessary or sufficient for a diagnosis of Substance Dependence." The DSM-IV defined substance dependence as "a maladaptive pattern of substance use" resulting in "clinically significant impairment or distress" meeting at least three out of seven criteria. It referred to less severe "substance use disorders" as "substance abuse."

The fifth edition of the manual-the DSM-5, published last May-eliminates the distinction between dependence and abuse. Instead it classifies substance use disorders as mild, moderate, or severe. Thus the DSM-5 does not explicitly recognize such a thing as drug addiction or dependence. But under "Substance Use and Addictive Disorders" the manual includes a category called "Behavioral Addictions" that so far consists of a lone entry: "Gambling Disorder." This term identifies what used to be called "pathological gambling," compulsive gambling that causes severe life problems.

It is unlikely that organized psychiatry will stop at labeling a single nondrug activity as addictive. Already the DSM-5 lists "Internet gaming disorder" as another possible behavioral addiction in Section III, which describes disorders under consideration for future inclusion in the manual. They're holding off for now because they want to see whether "persistent and recurrent use of Internet games, and a preoccupation with them, can result in clinically significant impairment or distress." But we already know that such preoccupations can disrupt people's lives, as commentator after commentator tells us they do, just as preoccupations with gambling, sex, and many other activities can. We are headed one way with addiction, which is to expand the concept.

I have advocated a broader understanding of addiction since 1975, when the Harvard psychiatry researcher Archie Brodsky and I published our book Love and Addiction. We argued not only that the distinction between "addictive" and "habituating" drugs should be abandoned but that nondrug activities, including love, sex, eating, and gambling, could also be addictive, or not, depending on the way people became involved in them and how destructive the involvement was. As we wrote in 1975, "If addiction is now known not to be primarily a matter of drug chemistry or body chemistry, and if we therefore have to broaden our conception of dependency-creating objects to include a wider range of drugs, then why stop with drugs? Why not look at the whole range of things, activities, and even people to which we can and do become addicted? We must, in fact, do this if addiction is to be made a viable concept once again."

While psychiatry is gradually moving in this direction, it is doing so under the pretense of locating addiction in the brain. In a February 2010 press release defending the decision to describe gambling as an "addictive disorder," the head of the DSM-5 working group on substance use disorders, Charles O'Brien, cited "substantive research" indicating that "pathological gambling and substance-use disorders are very similar in the way they affect the brain and neurological reward system." So according to O'Brien, gambling is addictive because it affects the brain as drugs do.

The physician and historian Howard Markel likewise defends the neuroscientific soundness of the DSM-5's approach. "The concept of addiction has been changing and expanding for centuries," he wrote in a 2012 New York Times essay. "Initially, it wasn't even a medical notion." (Imagine!) "'Addiction' entered the medical lexicon only in the late 19th century, as a result of the over-prescription of opium and morphine by physicians. Here, the concept of addiction came to include the notion of an exogenous substance taken into the body. Starting in the early 20th century, another key factor in diagnosing addiction was the occurrence of physical withdrawal symptoms upon quitting the substance in question."

Markel continues: "This definition of addiction was not always carefully applied (it took years for alcohol and nicotine to be classified as addictive, despite their fitting the bill), nor did it turn out to be accurate. Consider marijuana: in the 1980s, when I was training to become a doctor, marijuana was considered not to be addictive because the smoker rarely developed physical symptoms upon stopping. We now know that for some users marijuana can be terribly addictive, but because clearance of the drug from the body's fat cells takes weeks (instead of hours or days), physical withdrawal rarely occurs, though psychological withdrawal certainly can."

Instead of seeing addiction as a constantly shifting idea, Markel retrofits history to define addiction neurochemically. According to him, it was scientific progress when addiction was defined as use of a specific substance that might lead to physical withdrawal. Yet by his reckoning, it also was progress when those conditions were abandoned. Why? Brain scans!

Markel explains why he is now convinced that gambling, like marijuana, is addictive: "A team of scientists led by Nora Volkow at the National Institute on Drug Abuse have used positron emission tomography (PET) scans to show that even when cocaine addicts merely watch videos of people using cocaine, dopamine levels increase in the part of their brains associated with habit and learning. Dr. Volkow's group and other scientists have used PET scans and functional magnetic resonance imaging [fMRI] to demonstrate similar dopamine receptor derangements in the brains of drug addicts, compulsive gamblers and overeaters who are markedly obese." By this light, food is addictive and sex isn't.

Where Are the Sex Parts?

Section III of the DSM-5 does not include sex addiction, hypersexuality, or excessive pornography consumption, so evidently they are not being considered for later inclusion. That omission has prompted intense criticism from the UCLA psychologist Rory Reid. In a 2012 study of people treated for hypersexuality, Reid and his collaborators found that the proposed DSM-5 criteria for the condition were both reliable (leading to diagnostic agreement among clinicians) and valid (corresponding to real-world consequences). Conforming with an essential requirement for inclusion in the DSM, the criteria included "evidence of personal distress caused by the sexual behaviors that interfere with relationships, work or other important aspects of life."

The Reid study, which was reported in the Journal of Sexual Medicine, was produced at the same UCLA research center as the recent EEG study supposedly showing that hypersexuality is not a real addiction. "Science Supports Sex Addiction As a Legitimate Disorder," announced the UCLA press release heralding the 2012 study, while the headline on the press release about the 2013 study asked, "Is Sexual Addiction the Real Deal?"

The 2012 study involved no brain measurements, just DSM criteria. Yet no one is ever sent to rehab because of a brain scan. People are treated because of life problems and distress they have experienced. That's how the DSM-5 is written. Psychiatry's bible includes no defining criteria that refer to brain scans for the simple reason that there are no such defining criteria. Contrary to the claims of neuroscientific enthusiasts such as Charles O'Brien, Howard Markel, and Nora Volkow, you cannot see addiction on an fMRI.

The 2013 UCLA study rejected sex addiction because addicts did not show measurable brain reactions to sex like those cocaine addicts show in response to cocaine. But brain imaging studies of cocaine addiction do not find differences between experienced cocaine users who were addicted and those who were not. In other words, as in the case of sex addiction, a brain scan cannot indicate whether the subject is an addicted or controlled consumer, to what extent he might be addicted, or whether he is ready and able to cut back or quit.

A 2011 New York Times profile of Volkow, aptly headlined "A General in the Drug War," describes her scientific obsession this way: "She must say it a dozen times a day: Addiction is all about the dopamine." Dopamine is the neurochemical graphically imagined to mediate the pleasure centers of the brain, so that people become dependent on dopamine-producing experiences-which include, along with sex and eating, seeing a child smile, meditation, and being in love. But another obvious factor affects addictive behavior: self-restraint.

In a highly publicized 2012 study reported in the journal Science, Cambridge University researchers examined 47 addicted people along with 49 of their nonaddicted siblings. They found that the brains of both the addicted and nonaddicted siblings differed from those of 50 unrelated, nonaddicted people. Here is how Volkow described the results in an accompanying commentary: "The inferior frontal gyrus is really one of the main 'brakes' of our brain. [The drug users and their siblings] have less connections that are linking the rest of the brain with the inferior frontal gyrus that form a network that allows you to inhibit responses." But how does that explain why one sibling is addicted and the other isn't?

Covering the study in Time, Szalavitz made an observation that raises further questions: "Interestingly, the authors note, these connectivity problems are similar to those seen in the brains of teenagers, a group that is characterized by impulsive behavior. It is almost as if the brains of addicts are less mature. Perhaps that helps explain why some addiction wanes with age. Studies find that most people who struggle with alcohol and other drugs in their 20s [age out] of their problems by their 30s, typically without treatment." How is it that the overwhelming majority of people outgrow this brain disease, which the study suggests is inherited?

I'm an Addict, Not a Patient

There is a difference between saying a person with a life-disrupting attachment to something is addicted to it and saying he has a disease. In addition to arguing against unjustifiably restricted definitions of addiction, I have long resisted the excessive medicalization of human problems. This tendency can hurt the very people it is intended to help, as Szalavitz noted in a June 2013 Time essay.

Arguing that "defining obesity as a disease may do more harm than good," she drew an analogy to heavy drinking: "When alcoholism is seen as a disease, doctors and patients are also more likely to believe that overindulging can't be stopped without professional help or attending groups like AA-and that it must be treated with total abstinence. The disease concept wound up creating a ghettoized treatment system aimed only at severe cases, with few options for the vast majority of people with alcohol problems who don't require such extreme measures."

In addition to the question of whether it is helpful to call compulsive eating or drinking a disease, there is the question of whether it is accurate. Some critics of the DSM argue that the lack of biological tests for the disorders it lists undermines psychiatry's putative status as a branch of medicine. Thomas Insel, director of the National Institute of Mental Health, hopes further research will remedy the DSM's "lack of validity." As a May 2013 New York Times article explained, Insel wants to "reshape the direction of psychiatric research to focus on biology, genetics and neuroscience." But as the Times noted, "Decades of spending on neuroscience have taught scientists mostly what they do not know, undermining some of their most elemental assumptions."

When it comes to addiction, this is not a temporary inconvenience. It's the nature of the beast. Addiction can be defined only in terms of people's experiences and behaviors. There is nothing more scientific about how attached people become to an addictive object-whether heroin, alcohol, sex, eating, video games, or shopping-than to understand how subjectively captivating and powerful it is for them. That's not to say these things are usually or permanently addictive. Addiction is a specific involvement a person forms in a particular period in his life. Nothing more scientific can be said than that.