A splashy headline appeared on the websites of many U.K. newspapers this morning, claiming that men whose brothers or fathers have been convicted of a sex offense are “five times more likely to commit sex crimes than the average male” and that this increased risk of committing rape or molesting a child “may run in a family’s male genes.” The study, published online today in the International Journal of Epidemiology, analyzed data from 21,566 male sex offenders convicted in Sweden between 1973 and 2009 and concluded that genetics may account for roughly 40% of the likelihood of committing a sex crime. (Women, who commit less than 1% of Sweden’s sexual offenses, were omitted from the analysis.) The scientists have suggested that the new research could be used to help identify potential offenders and target high-risk families for early intervention efforts.
But independent experts—and even the researchers who led the work, to a certain degree—warn that the study has some serious limitations. Here are a few reasons to take its conclusions, and the headlines, with a generous dash of salt.
Alternate explanations: Most studies point to early life experiences, such as childhood abuse, as the most important risk factor for becoming a perpetrator of abuse in adulthood. The new study, however, did not include any detail about the convicted sex criminals’ early life exposure to abuse. Instead, by comparing fathers with sons, and full brothers and half-brothers reared together or apart, the scientists attempted to tease out the relative contributions of shared environment and shared genes to the risk of sexual offending. Based on their analyses, the researchers concluded that shared environment accounted for just 2% of the risk of sexual offense, while genetics accounted for roughly 40%. Although there is likely some genetic contribution to sexual offending—perhaps related to impulsivity or sex drive—the group “may be overestimating the role of genes” because their assumptions were inaccurate, says Fred Berlin, a psychiatrist and sexologist at Johns Hopkins University in Baltimore, Maryland.
Data on sexual crimes are tricky to obtain and parse: It’s extremely difficult to collect sufficient data about sexual offenders and their families to detect statistically robust patterns. Sweden is unusual because its nationwide Multi-Generation Register allows researchers to mine not only anonymized criminal records, but also to link them with offenders’ family records as well. Even with access to a nationwide database, Seena Fazel, of the University of Oxford in the United Kingdom, and colleagues had to include a very diverse range of offenses, from rape to possession of child pornography and indecent exposure, to maintain a large sample size.
The team did do some analysis by type of offense, separating rape from child molestation, for example. But some researchers worry that attributing a genetic basis to such a wide swath of behaviors is premature. There are also problems with relying on conviction records: Many more sexual crimes are committed than reported, and the proportion of those that go to trial is even smaller.
In addition, families with one member who has been convicted of a sexual offense are likely to be under much higher scrutiny by social services and law enforcement, leading to potential detection bias that artificially enhances the perception that sex crimes run in families, says Cathy Spatz Widom, a psychologist at the City University of New York who studies the intergenerational transmission of physical and sexual abuse. In a recent study, for example, Widom found that parents with a formal record of being abused as children were 2.5 times more likely to be reported to Child Protective Services for abusing their own children than parents in a control group who admitted to abusing their children, or whose kids said they had been mistreated.
The absolute risk of becoming a sex offender is very low: One of the study’s more dramatic-sounding findings is that brothers and fathers of sex offenders are four to five times as likely as men in the general population to commit sex crimes themselves. That statistic seems pretty striking until you look at the low prevalence of sex offense convictions in Sweden overall—a number comparable to that in the United Kingdom and many other countries. “If you look at the absolute rates, we’re looking at quite small numbers,” Fazel says. According to Fazel, convicted sex offenders make up about 0.5% of Sweden’s general population while about 2.5% of the brothers and fathers of convicted sex offenders are themselves convicted. Widom thinks even those estimates are likely too high. Because the risk of being a sex offender in both groups is so low, the difference between the two groups is hard to interpret, she says: “Who knows what the [relative difference in risk] even means when the prevalence is that low?”
In summary, there’s no doubt that some families are at a higher risk for abuse and criminal behaviors, including sexual offenses. But we’re a long way from pinning down genes that can explain why a person commits rape or any other sex crime.