What Exactly Creates a Sociopath?

Sociopathy is a lay term. It has never been a diagnosis in the DSM, it has no validated measurement instrument, and no one has shown that its clinical presentation can be distinguished from psychopathy.

The diagnosis the word maps to is antisocial personality disorder. The DSM-5-TR criteria require a pervasive pattern of disregard for and violation of the rights of others since age 15, shown by at least three of seven indicators: repeated unlawful behaviour, deceitfulness, impulsivity, irritability and aggressiveness, reckless disregard for safety, consistent irresponsibility, and lack of remorse. The person must be at least 18, and there must be evidence of conduct disorder before age 15. Six of the seven criteria describe behaviour. Only the last describes an internal state.

Psychopathy is a separate research construct, measured by the Hare Psychopathy Checklist-Revised rather than diagnosed from the DSM. It weights interpersonal and affective traits, including shallow affect, grandiosity, and absence of empathy, that the ASPD criteria barely capture. The overlap is one-directional. Most people who score above the psychopathy threshold also meet ASPD criteria, while most people with ASPD do not meet the psychopathy threshold. In a study of 496 prisoners in England and Wales, 44.9% met ASPD criteria, and of those, 31.8% scored at or above a PCL-R of 25. Psychopathy is covered separately from this article.

The popular split, that psychopaths are born and sociopaths are made, comes from a 1995 proposal by the psychologist David Lykken, who acknowledged he had constructed the taxonomy from the armchair rather than from data. It has not held up as a claim about causes. Both antisocial behaviour and psychopathic traits show substantial heritability, and childhood maltreatment is associated with both. There is no clean division where genes produce one condition and trauma produces the other.

Conduct disorder is the childhood diagnosis and a required precursor: an adult cannot be diagnosed with ASPD without evidence of it before age 15. The path forward is not guaranteed. Reviews of the follow-up literature put the conversion at roughly 25% of girls and 40% of boys with conduct disorder later meeting ASPD criteria, though national survey data using different methods produce much higher figures. The rate depends heavily on how it is measured.

ASPD affects about 4.3% of US adults over a lifetime by the largest national survey, 6.4% of men and 2.4% of women. Earlier surveys using different criteria produced estimates between 2.6% and 3.6%.

Twin and adoption studies consistently find a moderate genetic contribution. The largest meta-analysis of this literature, covering 51 twin and adoption studies, found additive genetic influences of .32 and nonadditive genetic influences of .09 on antisocial behaviour, for a total genetic contribution near 41%. Shared environment accounted for .16 and nonshared environment for .43.

Studies measuring ASPD criteria specifically report similar or somewhat higher figures. A Norwegian twin sample of 2,794 people interviewed with a structured diagnostic instrument found heritability of 51% for the ASPD criteria, with a confidence interval from 40% to 67%. An earlier study in the same population found 38% for dimensional ASPD traits.

Two things follow from this that the popular version usually drops. The first is that heritability is a statement about variance in a population, not about an individual. A heritability of 51% does not mean half of any person's condition was caused by genes. It means that in that studied population, about half the variation between people was statistically attributable to genetic differences. The second is that the same studies put nonshared environment at .43, larger than the additive genetic component. Experiences unique to the individual account for more of the variation than genes do.

When the search moves from twin designs to actual DNA, the numbers collapse. The largest genome-wide study of antisocial behaviour, pooling 85,359 people, found one genome-wide significant hit, in an intron of the FOXP2 gene, and common variants together explained between 3.4% and 7.7% of variance depending on the sample. The gap between twin heritability and measured-DNA heritability is expected and appears across most behavioural traits, but it means no one can currently point to the genes that produce the twin estimate.

MAOA is the gene most often named in coverage of this topic. It codes for an enzyme that breaks down serotonin, norepinephrine, and dopamine, and a common variant in its promoter region produces lower enzyme activity. The nickname "warrior gene" attached to the low-activity variant.

As a standalone cause, it does not hold up. Meta-analyses of the association between MAOA genotype and antisocial behaviour find no reliable main effect. Carrying the low-activity variant, on its own, does not predict antisocial behaviour. The low-activity variant is also common: 37% of the males in the original study carried it.

What the research supports is narrower and more conditional. In 2002, a study following 442 men from a New Zealand birth cohort found that childhood maltreatment predicted antisocial outcomes far more strongly among those with the low-activity variant. Men with both the low-activity variant and a history of maltreatment made up about 12% of the male cohort but accounted for 44% of its violent convictions. Among those severely maltreated and carrying the low-activity variant, 85% developed some form of antisocial behaviour. There was no effect of the gene by itself.

A 2014 meta-analysis tested this across 20 male cohorts and confirmed it. Early adversity predicted antisocial outcomes more strongly for the low-activity genotype, and the effect was specific to maltreatment rather than adversity in general. Across 11 female cohorts, the interaction did not hold, and the weak signal that appeared ran in the opposite direction.

So the honest version of the MAOA story is not that a gene causes violence. It is that in males, a common gene variant appears to change how much damage childhood maltreatment does. Single studies have replicated this inconsistently, candidate-gene research as a field has a poor replication record, and the effect has not been established in women. It is a probabilistic moderator in one sex, not a cause, and it cannot be used to predict or explain any individual.

Maltreatment is the environmental factor with the most evidence behind it, and the effect is real but smaller than the popular account suggests. A meta-analysis of 14 prospective studies covering 20,946 people found maltreated children had roughly twice the odds of adult antisocial behaviour, an odds ratio of 1.96 with a confidence interval from 1.42 to 2.71. A separate meta-analysis of prospective studies of maltreatment and violence, covering 39,271 people, found an odds ratio of 1.8. These are meaningful and they are not destiny. Most maltreated children do not become antisocial adults.

Two complications sit underneath that number. The first is that studies asking adults to recall their own childhoods produce stronger associations than studies that recorded the maltreatment at the time, which means retrospective research overstates the link. The second is genetic confounding: parents with ASPD maltreat their children at substantially elevated rates, so a child in that home inherits both the genes and the environment. Part of what looks like an environmental effect is genetic transmission running through the same families.

An umbrella review of meta-analyses on environmental risk factors for personality disorders found that no risk factor for ASPD met its highest three tiers of evidence. That does not mean maltreatment is irrelevant. It means the meta-analytic evidence for ASPD specifically is thinner than the confidence with which it is usually asserted.

Prenatal exposures follow the same pattern in sharper form. Crude analyses find maternal smoking during pregnancy roughly doubles the odds of offspring conduct problems. But studies designed to separate inherited from prenatal effects, comparing siblings and using other genetically informed designs, find most or all of that association disappears. Mothers who smoke during pregnancy differ from mothers who do not in ways that are themselves heritable. The crude association is well replicated. The causal story is not.

Family instability, harsh or inconsistent parenting, poverty, and community violence all correlate with antisocial outcomes. They are also entangled with each other and with parental antisociality, and few designs can separate them. Notably, shared family environment shows up clearly in childhood conduct problems and shrinks toward zero in adult ASPD twin models, which suggests the family effects that matter most are the ones that differ between siblings rather than the ones they have in common.

Imaging studies find group-level differences. A meta-analysis of 43 structural and functional studies of antisocial, violent, and psychopathic individuals found reduced prefrontal structure and function with an effect size of -0.60, concentrated in the right orbitofrontal cortex, right anterior cingulate, and left dorsolateral prefrontal cortex. A 2023 meta-analysis of 23 studies comparing 629 antisocial individuals with 745 controls found smaller total grey matter volume and smaller amygdala volume, and also found greater variability in the antisocial groups, which is itself evidence that these are not one uniform condition.

None of this establishes a cause. Cross-sectional imaging cannot distinguish a brain difference that preceded the behaviour from one produced by years of substance use, head injury, or incarceration, or from one that shares a common origin with the behaviour. Substance use disorders are common in ASPD samples and are themselves associated with reduced prefrontal grey matter.

The construct problem is at its worst here. Most imaging samples are small, male, forensic, and mix ASPD with psychopathy and general violence. Amygdala underactivity is discussed far more consistently in psychopathy and callous-unemotional research than in ASPD, where heightened threat reactivity and reactive aggression also appear. Pooling these groups produces an average that describes no one precisely.

No single factor does, and the evidence does not support any version of the story that names one.

What the research supports is a liability spread across many genes, each contributing very little, interacting with childhood conditions, expressed first as conduct problems in childhood and meeting criteria for a personality disorder only in adulthood. Genetic influence is moderate and real. Maltreatment roughly doubles the odds and most maltreated children are unaffected. In males, one common gene variant appears to change how much harm maltreatment does. Brain differences track the antisocial spectrum at the group level without being established as causes.

Where popular coverage goes wrong is almost always in one of two ways. It swaps constructs, reporting a psychopathy study or a criminal conviction record as though it measured sociopathy. Or it converts a heritability figure or an odds ratio of two into an origin story, when neither can carry one.