The concept of addiction involving non-substance behaviors as repeated urges to engage in counter-productive activities was introduced in 1990. At the time, behavioral addictions were proposed to encompass obsessive-compulsive disorder (OCD), "compulsive spending" (including gambling), "overeating" (binge eating), "hypersexuality," and kleptomania. It was thought these behaviors were linked to addiction by poor impulse control and self-regulation, which led to repetitive engagement of the behavior despite negative consequences[1,2]. The concept of an obsessive-compulsive spectrum of disorders was proposed in 1993; disorders that featured an inability to control or delay repetitive behaviors were thought to fall on a spectrum from impulsivity to compulsivity [3]. Numerous psychiatric and neurologic disorders were included in this spectrum [4].

Further research suggested that impulsivity and compulsivity appeared in substance and behavioral addictions at different stages, and the "impulsive-compulsive disorder" model was proposed. In this model, impulsivity was dominant in earlier-stage addiction, when behavior is motivated and reinforced by reward, and compulsivity was dominant in later stages, when behavior is motivated and reinforced by avoidance of negative emotional states [5].

An influential 2006 study broadened the definition and core features of addiction as 1) a state of craving or urge that immediately precedes the behavior, 2) impaired ability to control the behavior, and 3) the behavior continues despite negative consequences [6]. Criterion 3 suggested addiction was no longer tethered to substance use. This was cemented in 2011, when the American Society of Addiction Medicine (ASAM) released their definition of addiction. The core features of this definition are [7]:

The reward deficit disorder model describes addictions as chronic relapsing disorders characterized by a compulsion to seek and take the drug or experience the behavior, loss of control over stimuli intake, and emergence of a negative emotional state (e.g., dysphoria, anxiety, irritability) when unable to access the stimulus. The negative emotional state is termed motivational withdrawal syndrome. A key component of addiction is negative reinforcement, which describes engagement in potentially harmful behaviors to alleviate a negative emotional state [45].

The cognitive-behavioral model addresses the role of motivational drives for reward-seeking and stress-reduction; behavioral control (related to executive inhibition); and decision-making (involving weighing the pros and cons of engaging in motivated behaviors) in addictive behaviors.

Predisposing personality factors include high impulsivity, shyness, neuroticism, low self-esteem, conscientiousness, self-directedness, and procrastination. Affective/cognitive responses to environmental stimuli influence the preference of specific Internet genres. For example, a perceived lack of social support, feelings of isolation, and loneliness are linked to the use (and potential misuse) of Internet genres with prominent communication features (e.g., social networking sites) [64,65]. Specific Internet-mediated addictions are associated with specific personality profiles, such as greater rates of ADHD and impulsivity.

The PFC region is involved in executive processes important in risk-reward decision-making, but is among the last brain regions to mature. Compared with mature adults, this aspect of brain development renders adolescents more vulnerable to addictions, other risky behaviors, and mental health disorders [44]. The brain mechanisms that mediate cognitive control, impulsivity, and sensitivity to reward mature during adolescence and young adulthood, and this is the period when substance use initiation and other problem behaviors most commonly emerge [67].

Immature connections between the PFC, the nucleus accumbens, and the amygdala are thought to largely influence goal-directed behaviors in adolescents [68,69]. Reduced prefrontal cognitive control may allow greater influence of affective systems over decision-making and behavior, which increases adolescent vulnerability to social and peer pressure [70]. Addiction can interrupt the normal neurodevelopmental trajectory from adolescence to early adulthood, and addiction in adolescence can be both a cause and effect of cognitive, affective, and behavioral dysfunction.

Almost all behavioral addictions were originally understood through the obsessive-compulsive spectrum model. By extension, and based on efficacy in OCD, selective serotonin reuptake inhibitors (SSRIs) became the most widely used drug class in the treatment of behavioral addictions, but they showed minimal benefit. More recently, neuroscience research identified similarities in the core features of behavioral and substance addictions, and the obsessive-compulsive spectrum model was discarded in favor of better-fitting addiction disorder models. Naltrexone, an opioid receptor antagonist with efficacy in substance use disorders, has suggested benefit in behavioral addictions, but there is evidence of possible adverse effect [22,85].

Among treatment-seeking disordered gamblers, women initiate gambling at an older age than men (31.3 years versus 22.4 years) and have a significantly shorter time from gambling initiation to meeting DSM criteria for gambling disorder (8.33 years versus 11.97 years). As noted, this phenomenon is consistently found in women with substance use disorders as well [91].

Gambling disorder is highly comorbid with other mental health disorders, particularly substance use disorders, and shows a heritability rate of 50% to 60% [94]. Data published before widespread access to online gambling indicate that, compared with non-pathologic gamblers, pathologic gamblers show a six-fold greater risk for substance use, a four-fold greater risk for illicit substance use disorder, and a three-fold greater risk for mood disorders [95]. At-risk gamblers have higher rates of psychiatric comorbidity relative to the general population [96].

In Australia in 2010, gambling generated $19 billion in revenue; 55% came from electronic gaming machines in clubs and hotels, and 40% of electronic gaming machine revenue came from problem gamblers (only 0.6% of the population) [10]. Termed the "crack cocaine of gambling," electronic gaming machines are considered the most hazardous form of gambling, because they incorporate features that promote gambling-related cognitive distortions to perpetuate gambling, including near-win (or near-miss) outcomes and "loss disguised as a win." These features activate the sympathetic nervous system and reward-related neural circuitry, amplified further in disordered gamblers [99,100].

According to the DSM-5-TR, gambling disorder is categorized as a substance-related or addictive disorder. Diagnosis of gambling disorder requires four or more of the following [11]:

The exclusion of illegal activity from the DSM-5-TR criteria was not an endorsement to downplay its relevance. The presence of illegal behavior in patients with gambling disorder is directly linked to more severe pathologic outcomes and resistance to treatment; specific and more intensive treatment may be required for these patients [108].

Early sexual conditioning impacts sexual arousal templates. A critical developmental period of sexual behavior forms around initial experiences with sexual arousal and desire, masturbation, intercourse, and orgasm. Younger age of onset for regular Internet pornography use and greater preference for pornography over partnered sex is associated with persistence of this pattern. In 2014, nearly 50% of college-age men reported they were first exposed to Internet pornography before 13 years of age, compared with 14% in 2008 [176,177]. Young adults with addictive Internet pornography use show greater impairments in sexual arousal and erectile function in intimate relationships, but not with pornographic material.

Paraphilic hypersexuality is remarkable for the sizeable number reporting gynandromorphophilia, a rarely discussed erotic interest in persons with both male and female anatomy (usually full breasts and intact penis) typified by incompletely transitioned male-to-female transgender women. For many, this specific enduring erotic interest leads to confusion over sexual orientation or gender identity and hesitant self-reference as ''mostly heterosexual'' or as bisexual [138].

Massive, multiplayer, online role-playing games (MMORPGs) are unique in pathogenic potential among game genres. A 2015 report described a case of two brothers (22 and 19 years of age), high academic achievers from an intact, upper-middle class family who lived at home. Both began MMORPGs two years before hospital admission for Internet gaming disorder. In the initial months, time spent gaming progressed from 2 to 4 hours to 14 to 18 hours daily. Gaming interfered with their sleep and daily routine. Both deteriorated to the point that when engaged in MMORPGs, they urinated and defecated in their clothes, did not change clothes for days, did not bathe, skipped meals, and did not answer the phone. Fixated on gaming, they were indifferent to the presence of a burglar robbing their home. Both failed their classes, became abusive and violent when their gaming was disrupted, and were admitted to inpatient care by their parents [194].

The pooled prevalence estimates from more than 30 countries showed problematic Internet game use in 8% to 12% of young persons and addictive use in 2% to 5% of children, adolescents, and college students [17]. Nationally representative American samples show an Internet gaming disorder prevalence of 8.5% among those 8 to 18 years of age [203,204].

As discussed, maladaptive cognitions are a core feature of Internet gaming disorder and include self-regulation deficits, preference for a virtual life, cognitive bias, impaired cognitive control abilities, cognitive deficits, poor cognitive error processing, and decision-making deficits [204]. CBT addresses these cognitions and is the most widely used psychological treatment for Internet gaming disorder. The first stage of CBT treatment for Internet gaming disorder deals with the behavioral aspects of addicted gaming, and subsequent stages gradually shift the focus toward the development of positive cognitive assumptions. During therapy, addicted gamers identify false beliefs and learn how to modify them into more adaptive ones. CBT also trains patients to monitor their thoughts and to identify affective and situational triggers associated with their addictive gaming behavior [204].

There is increasing concern that consumption of food with high sugar content may be "addictive" and promote weight gain. Claims that sugar has addictive potential are based on animal studies, but direct human evidence for symptoms of sugar-related substance dependence is lacking [236]. A study evaluated 1,495 university students for potential "food addiction" (with DSM-IV substance dependence criteria applied to food) involving high-sugar foods. In this group, 12.6% met food addiction criteria; of these, 5% mainly consumed sugar-laden foods. Overweight/obesity was unrelated to sugary food preference. "Food addiction" was concluded to result from unique individual characteristics that determined reward response to food and promoted excessive eating [236,237].

Therapy variants in binge eating disorder include CBT combined with body image exposure or cognitive restructuring components, and ecologic momentary assessment to increase self-monitoring adherence [81,259,262]. A meta-analysis of studies that randomized patients to CBT or non-treatment (control) found improved binge frequency and abstinence outcomes with therapist-led, partially therapist-led, structured self-help, and guided self-help CBT. Therapist-led CBT outcomes were compelling, with a higher rate of abstinence (59% CBT vs. 11% control), reduction of binge episodes, reduced patient hunger and eating concerns, and improved sense of control over eating. Guided self-help CBT reduced global eating-related psychopathology. When all study results were pooled, CBT and control groups were similar in weight lost and depressive symptom reductions. Therapist-led CBT was much more effective in improving key behavioral and eating-specific psychological domains [81].

The prodrug lisdexamfetamine is the sole FDA-approved drug for the treatment of moderate-to-severe binge eating disorder. The FDA has withdrawn several investigational or approved binge eating disorder pharmacotherapies from the U.S. market over safety concerns and adverse effects (e.g., sibutramine, rimonabant, d-fenfluramine) [81].

Bariatric surgery is a treatment option for binge eating disorder with severe obesity (BMI ≥40 or BMI ≥35 with comorbid conditions). In the past, an eating disorder diagnosis was a contraindication to bariatric surgery, but this restriction has been relaxed. The disability and difficulty achieving stable weight loss with standard behavioral interventions has increasingly led to surgery candidates with binge eating disorder [245,274]. Bariatric surgery can be suitable for selected patients with binge eating disorder, but the extent of weight loss depends on post-surgery binge episode reduction, and 20% to 40% fail to lose sufficient weight or regain significant weight post-surgery [275]. Patients may experience emergent post-bariatric surgery psychiatric comorbidities. Considering the demonstrable and durable behavioral and neuroplastic changes associated with addictive eating behavior, obesity treatment is not realistic for some of these patients [276].

In one study, among 1,441 shopping mall visitors, the rate of compulsive buying disorder was 8.7% [283]. Compared with non-compulsive buyers, compulsive buyers were younger, less educated, more likely female, and more likely to have used licit and illicit substances. They also showed higher levels of impulsivity and obsessive-compulsive symptoms, lower levels of well-being and self-esteem, and greater psychological distress. Patients identified as compulsive buyers were five times more likely to meet diagnostic criteria for borderline personality disorder. These findings suggest that, among shopping mall visitors, compulsive buyers are prevalent and have high rates of important indicators for psychopathology [283].

Compulsive buying is a chronic, repetitive behavior that becomes a primary response to negative events and feelings [286]. Four distinct phases in compulsive buying events are described: anticipation, preparation, shopping, and spending. Negative emotions, such as anger, anxiety, boredom, and self-critical thought, were the most common antecedents to shopping binges, with euphoria or relief from negative emotions the most common immediate effects [281].

Compulsive buying disorder was classified in the DSM-III-R as an impulse control disorder not elsewhere specified, but it was omitted from the DSM-5 [11,300]. Since 2013, research demonstrates that compulsive buying disorder has several features that characterize addictions, including cue reactivity and cravings [289].

Group psychotherapy that uses a CBT approach or cognitive-behavioral methods within an eclectic approach appears beneficial, with durable improvements in reducing distress associated with compulsive buying disorder and maladaptive buying behavior. Research suggests that impulse control training should be a core component of compulsive buying disorder treatment [289]. Attrition rates show that group psychotherapy is not acceptable to all patients with compulsive buying disorder; patient choice and suitability are important considerations [289].

Evidence suggests that a low-intensity, guided self-help approach to treating compulsive buying disorder was comparable to high-intensity group therapy. If patients can be treated with effective, brief, and less intensive psychological interventions first, this may increase service throughput and efficiency [304]. Studies of Internet-based, therapist-assisted self-help programs also usefully mimic the shift of behavior in online shopping [305]. Research indicates that excitability regarding online shopping and compulsive buying disorder are mediated by Internet use expectancies [294]. Treatments clearly need to reflect the context within which compulsive buying occurs [289].

Reward and punishment sensitivity was studied for impact on treatment outcome (12 weekly CBT sessions) in female patients with compulsive buying disorder or gambling disorder [46]. In compulsive buying disorder, higher reward sensitivity was related to poorer treatment adherence but reduced risk of dropout. Patients were likely to have stronger intrusive urges to buy, interfering with ability to curb buying behavior and carry out practice homework for CBT. The lower dropout risk may reflect patient motivation by social factors, with patients more likely to form a therapeutic alliance and not abandon treatment. High punishment sensitivity correlated with harm avoidance; these patients had increased risk of treatment drop-out.

A review of compulsive buying disorder treatment found few published pharmacotherapy studies. SSRI antidepressants are the most-studied pharmacotherapy, based on the same premise as their use in other behavioral addictions [85]. Evaluations of citalopram, escitalopram, and fluvoxamine have reported mostly negative results, and clinicians should consider psychotherapeutic options before pharmacotherapy [289]. A systematic review of treatment studies found that group CBT is effective in reducing symptoms of compulsive buying disorder, whereas pharmacotherapy with SSRIs or topiramate did not indicate superiority over placebo [306]. One literature review found that a combination of antidepressants and CBT is effective for management of compulsive buying disorder. Serotoninergic antidepressants are effective as monotherapy [302].

Community prevalence studies suggest that trichotillomania occurs in 0.5% to 2.0% of the population [309,310]. Lifetime prevalence is estimated to be 1% to 3% [311]. In adults, trichotillomania predominately affects women (4:1 female to male), but childhood trichotillomania shows equal sex distribution [309,310]. As a behavior, hair pulling appears quite common and often presents along a continuum from mild to severe. When hair pulling meets the criteria for trichotillomania, interventions should be considered [312].

With hair pulling in trichotillomania, the scalp is the most common site (72.8% of patients) followed by the eyebrows (56.4%) and the pubic region (50.7%) [312]. Triggers to pull can be sensory (e.g., physical sensations on the scalp, hair thickness, length, and location), emotional (e.g., feeling anxious, bored, tense, angry), or cognitive (e.g., thoughts about hair and appearance, rigid thinking, cognitive errors). Many patients report not being fully aware of their pulling behaviors at least some of the time—a phenomenon termed "automatic" pulling. "Focused" pulling, in contrast, generally occurs when the patient sees or feels a hair that is "not right" or if the hair feels coarse, irregular, or "out of place" [312,314].

Patients with trichotillomania have high rates of comorbidity, including major depressive disorder (39% to 65%), anxiety disorders (27% to 32%), and substance use disorders (15% to 19%). Trichotillomania is often misdiagnosed as OCD. Rates of comorbid OCD are significantly higher in clinical (13% to 27%) than community (1% to 3%) populations [312,318].

The age of trichotillomania onset is generally earlier than its common comorbidities. In a large trichotillomania survey, patients sought to alleviate negative feelings associated with hair pulling through use of tobacco products (17.7%), alcohol (14.1%), or illicit drugs (6.0%). Also, 83% reported anxiety and 70% reported depression due to pulling, indicating that clinicians should screen for trichotillomania and the secondary manifestations of the behavior to better ensure successful treatment outcomes [318,319].

The similarity between repetitive motor symptoms of hair pulling and repetitive compulsive rituals in OCD led to proposals that both disorders shared common neurobiologic pathways. However, evidence indicates that trichotillomania and OCD are distinct. In contrast to OCD, patients with trichotillomania are more commonly female, and body-focused repetitive behavior disorders (e.g., skin picking, compulsive nail biting) more frequently occur in these patients and their first-degree relatives. In OCD, compulsions are often driven by intrusive and obsessional thoughts, which are seldom found in patients with trichotillomania and are not listed in diagnostic criteria. The typical age of onset is early adolescence for trichotillomania and late adolescence for OCD. Treatment response also differs, with SSRIs effective in OCD but not trichotillomania [320,321].

Hair pulling is considered a means of escaping or avoiding aversive experiences, and the temporary relief from negative emotions can maintain the behavior through a cycle of negative reinforcement. Patients with trichotillomania have demonstrated greater difficulty regulating negative affective states than controls. In some individuals, boredom may trigger hair pulling. Some have hypothesized that in a subgroup of patients, hair pulling alleviates negative emotions resulting from perfectionism and an inability to relax, with pulling serving the function of releasing tension [322].

In the few trichotillomania neuroimaging studies conducted, patients with trichotillomania showed volume deficits and disorganization in neurocircuits that mediate affective regulation, motor habit generation, and suppression (compared with healthy controls) [323]. An analysis of MRI scans of 23 girls/women with trichotillomania (compared with 16 healthy controls) implicated somatosensory, sensorimotor, and frontal-striatal circuitry, and partially overlapped with structural connectivity findings in OCD [324]. One study sought to determine whether recently identified subtypes of trichotillomania mapped to any unique neurobiological underpinnings [325]. In this study, 193 adults with trichotillomania and 58 healthy controls were recruited for a between-group comparison using structural neuroimaging. Differences in whole brain structure were compared across the subtypes, while controlling for age, sex, scanning site, and intracranial volume. Patients with trichotillomania with low awareness demonstrated increased cortical volume in the lateral occipital lobes compared to controls. Additionally, impulsive/perfectionist patients showed relative decreased volume near the lingual gyrus of the interior occipital-parietal lobe compared with controls [325]. Another study of subjects with trichotillomania failed to identify abnormalities in implicit learning or striatal/hippocampal activation, characteristics of OCD [326].

The evidence base for psychotherapy in trichotillomania is small, and habit reversal therapy has the greatest empirical support. Habit reversal therapy is a behavioral therapy initially introduced for the treatment of nervous habits and tics. The core components of habit reversal therapy include self-monitoring (e.g., asking the patient to track hair-pulling or skin-picking events), awareness training, competing response training, and stimulus control procedures (e.g., removing hair-pulling or skin-picking cues from the patient's environment) [322].

Habit reversal therapy is generally delivered in weekly 60-minute sessions for 4 to 22 weeks, with more frequent sessions for patients with greater symptom severity. It can be delivered individually, in a group format, or online using a self-help manual [328]. The clinical benefits of trichotillomania symptom reduction have been augmented by adding components of acceptance and commitment therapy or dialectical behavior therapy [329,330]. Patient gains from habit reversal therapy are usually maintained for three to six months. Many clinicians combine habit reversal therapy and CBT, but published data support habit reversal therapy use as a stand-alone, first-line psychotherapy approach in trichotillomania [307]. One small study reported success combining metacognitive therapy with habit-reversal techniques for treatment of trichotillomania [331].

The commonly quoted kleptomania prevalence of 0.6% in the general population was extrapolated from study data of patients with eating disorders [341]. A 2010 college student survey found a lower kleptomania prevalence rate of 0.38% [342]. This latter figure is likely more accurate because it was not derived from a specific subpopulation [341]. Among those arrested for shoplifting, kleptomania rates have ranged from 0% to 8%. This may be falsely low due to incomplete psychiatric evaluation, lack of strict diagnostic criteria for kleptomania, and/or selection bias [340].

In contrast to community prevalence rates, studies of patients with psychiatric disorders suggest prevalence rates of kleptomania sufficient to potentially represent a public health concern. Among 203 inpatients with diverse psychiatric disorders, 7.8% met current and 9.3% met lifetime criteria for a diagnosis for kleptomania. The similar current and lifetime prevalence suggests that untreated kleptomania is chronic. Other studies show increased rates of kleptomania in patients with depression (3.7%), alcohol use disorder (3.8%), and gambling disorder (2.1% to 5%) [340,343].

Most kleptomania studies report female predominance. However, women are more likely to seek professional help, and the legal system is more likely to send female shoplifters for psychiatric evaluation and male shoplifters to jail [340]. So, the true prevalences in men and women may be more similar than reported. Among 101 adults meeting DSM-IV kleptomania criteria (73.3% female), the mean age of shoplifting onset was 19.4 years, persisting an average 8.2 years before full diagnostic criteria were met [344].

In contrast to community prevalence rates, studies of patients with psychiatric disorders suggest prevalence rates of kleptomania sufficient to potentially represent a public health concern. Among 203 inpatients with diverse psychiatric disorders, 7.8% met current and 9.3% met lifetime criteria for a diagnosis for kleptomania. The similar current and lifetime prevalence suggests that untreated kleptomania is chronic. Other studies show increased rates of kleptomania in patients with depression (3.7%), alcohol use disorder (3.8%), and gambling disorder (2.1% to 5%) [340,343].

Most kleptomania studies report female predominance. However, women are more likely to seek professional help, and the legal system is more likely to send female shoplifters for psychiatric evaluation and male shoplifters to jail [340]. So, the true prevalences in men and women may be more similar than reported. Among 101 adults meeting DSM-IV kleptomania criteria (73.3% female), the mean age of shoplifting onset was 19.4 years, persisting an average 8.2 years before full diagnostic criteria were met [344].

The average age for stealing behavior onset is adolescence, with the first professional evaluation in the mid- to late-30s. Women tend to present for evaluation at a younger age than men. The extended length of time between onset and mental health evaluation reinforces the guilt, shame, and secrecy involved in this disorder [340].

The DSM-5-TR diagnostic criteria for kleptomania are [11]:

Preliminary data from treatment-seeking patients show gender differences in clinical features and comorbid disorders that may reflect biologic and sociocultural factors with implication for prevention and treatment. Women and men with kleptomania both show substantial symptom severity and functional impairment. Women are more likely than men to be married (47.1% vs. 25.9%), experience shoplifting onset at a later age (20.9 years vs. 14 years), steal household items, hoard stolen items, and have a comorbid eating disorder. Men are more likely to steal electronic goods and have another impulse-control disorder [351].

Another study of patients with Parkinson disease identified addictive behaviors in 14% of patients receiving any dopaminergic medication and 17% of patients receiving dopamine agonists. Behaviors included compulsive gambling (5%), compulsive sexual behavior (3.5%), compulsive shopping (6%), and binge-eating disorder (4%). Compulsive behaviors were more common with dopamine agonists than levodopa and other medications (17.1% vs. 6.9%). The prevalence was similar with the two most commonly used dopamine agonists, pramipexole (17.7%) and ropinirole (15.5%) [357]. Underlying susceptibility factors are suggested by the relatively small proportion of patients with Parkinson disease exposed to dopamine agonists who develop impulse control disorders. In these patients, risk factors for impulse control disorder during dopamine agonist therapy include levodopa treatment, age younger than 65 years, being unmarried, high caffeine use, family history of gambling problems, and current cigarette smoking. Other associated factors were functional impairment, depression, anxiety, obsessive-compulsive symptoms, impulsivity, and novelty-seeking. These risk factors show similarity to those reported in substance use disorders and gambling disorder and suggest common neurobiologic substrates [357].

Aripiprazole, an atypical antipsychotic drug, has been linked to treatment-emergent pathologic gambling in eight adult male patients with schizophrenia or bipolar disorder [359]. All had histories of substance use disorders and non-pathologic gambling before taking aripiprazole. Patient histories showed aripiprazole initiation coincided with intensified urges to gamble or loss of control over gambling; discontinuation of aripiprazole coincided with cessation of severe gambling urges or loss of control [359]. While acknowledging this association, the authors of one analysis caution that since Parkinson disease is itself associated with pathologic gambling, further studies are warranted [360].