Low pulse = higher crime rate? Cause-and-effect confusion in heart rate/crime stories

precogAn LA Times story crossed our radar recently with what seemed like an exceptionally provocative premise:

“What if a test could identify young men who were nearly 50% more likely than their peers to become violent criminals?”

To me, this sounded like something out of the 2002 Tom Cruise sci-fi flick Minority Report — which (coincidentally?) is slated for a reprise as a prime-time drama this fall.

But the Times story wasn’t talking about anything as exotic as the film’s freaky crime-predicting “precogs.” In fact, the test envisioned by the Times is so simple that “you can take it with two fingers and a clock,” the story says.

It’s talking about heart rate, of course, and a Swedish study apparently found that that among men, those with the lowest resting heart rates at age 18 were more likely to commit violent crimes as they got older.

What could your pulse rate possibly have to do with your propensity to commit a crime? The story mentions some “guesses” that scientists have come up with:

One theory is that men with slower hearts have lower levels of physiological arousal, which they try to boost by doing things that are dangerous or illegal. The other leading theory is that those with slower hearts have a muted response to risky or stressful situations, so they don’t feel the fear that others would.

Such speculation is fascinating, but I do wish the story had framed this discussion within a larger context of evidence quality. One thing the story never specifically mentions is that this was an observational study — one incapable of proving any definitive link between heart rate and criminal behavior.

The story works hard to tell us all the things that the researchers ruled out as explanations for the findings — things like physical, cardiovascular, psychiatric, cognitive, or socioeconomic factors. But it doesn’t include a single word about elements that cut the other way  —  e.g. the inherent limitations of the study design.

And the Times was not alone in its omission of this crucial caveat. Newsweek also had no such cautionary language in its story, which suggested that low heart rate “may be a strong predictor of which young men are more likely to commit violent crimes later in life.”

Reuters Health, while lacking a specific comment on the limitations of the study design, did responsibly quote one of the study authors who said: “It is obvious that low resting heart rate by itself cannot be used to determine future violent or antisocial behavior.”

But perhaps that fact wasn’t so obvious to news media outlets like the New York Daily News, which reported that  “A low heart rate could make you more prone to a life of violence, crime and injuries.”

HealthDay seems to have done the best job of acknowledging that the study “only found an association rather than a cause-and-effect link.” And that cautionary language also got picked up by CBS in its secondary coverage of the study, which referenced the HealthDay story.

Another troubling hole in many of these stories is the lack of any critical perspective on the findings. The Times story quotes from an editorial that accompanied the study in JAMA Psychiatry, but that editorial, titled “Low Resting Heart Rate as an Unequivocal Risk Factor for Both the Perpetration of and Exposure to Violence,” seems mainly concerned with telling us how important the findings are and what broad implications they have for society.

“Can the criminal justice system continue to turn a blind eye to the anatomy of violence?” editorialist Adrian Raine asks.

I suppose the journal JAMA Psychiatry bears some of the blame for that. I’ve never understood why journals sometimes commission editorials from authors who simply confirm whatever study’s being commented on and offer little in the way of critical analysis. Why publish something that amounts to a nodding of one’s head in agreement, when there are others who would vigorously challenge and explore the validity of the findings?

As the Philadelphia Inquirer pointed out, Raine (the editorialist) wrote extensively about the need to pay more attention to biological links to crime, including heart rate, in his 2013 book, The Anatomy of Crime — and so his likely reaction to the study could have been anticipated. According to the New York Times, Raine even goes so far in his book as to propose (apparently more than semi-seriously) a “Minority Report-style” government program where “All men 18 and over will undergo a brain scan and a DNA test, and those whose results indicate future criminality will spend the rest of their lives in a pleasant enough form of indefinite detention.”

In the editorial, Raine says the large size of the study puts the empirical basis of [the study’s] findings “beyond further dispute.” But it didn’t take me very long to find an expert who was willing to continue disputing. In response to a quick email I sent her, Susan Molchan, MD, a psychiatrist and HealthNewsReview.org contributor with experience in clinical research at NIH, said, “I had seen the headlines on this study and was rolling my eyes at them.”

What problems did she see?

For one thing, she noted that the statistical significance of the results comes from the comparison of men with lowest resting heart rates to those with the highest resting heart rates. She said the latter group had heart rates “which are quite high for young men. Perhaps there is some issue with this group.”

In other words, maybe it’s not that men with low heart rates are more likely to commit crimes, but that those with very high heart rates have some issue that makes them less likely to commit crimes.

Another problem: “The fact the rape doesn’t correlate along with the other crimes erodes the whole theory of men with low resting heart rate seeking more stimulation . . .  And we don’t know what absolute kinds of numbers we’re looking at—how many crimes were committed by the 1st quintile compared to the 5th?”

Many questions.

Many places where some caution and critical perspective could have been inserted into this coverage.

Many lost opportunities to inform the public about the true state of this science.

You might also like

Comments (3)

Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.

Veronika V

September 16, 2015 at 2:04 am

Quick search in PubMed does not support the association as extensively researched or even recognized as plausible by the scientific community.
Keywords bradycardia + violence return merely 11 records, most of which relate to the consequences of shaken baby syndrome and brain injury as the underlying cause of bradycardia later in life. http://www.ncbi.nlm.nih.gov/pubmed/?term=bradycardia+violence
The researchers accounted for cardiovascular fitness which did not reduce the correlation. They did not discuss medication and illicit drug use history as an underlying cause, effectively grouping a symptom with widely varied etiology and creating an apparent cause-effect association. As with previous studies in the search for “born criminals” (Lombardo, Dugdale’s “The Jukes”) or “hereditary geniuses” (Galton), we shall not forget that correlation does not equal cause.


September 16, 2015 at 6:44 pm

The Norwegian study didn’t control for race or ethnicity, at least not according to the press reports. This 1983 study by the Chicago Department of Health (http://aje.oxfordjournals.org/content/122/3/366), this 1991 study by the University of Texas (http://www.ncbi.nlm.nih.gov/pubmed/1842529), and this 2011 study by the CDC (http://www.cdc.gov/nchs/data/nhsr/nhsr041.pdf) all show that blacks and non-white Hispanics have lower resting heart rates than whites do. About 6.5% of Norway’s population is non-Western, which coincidentally matches the rate of criminality (6%) measured by the Norwegian study.

“Can the criminal justice system continue to turn a blind eye to the anatomy of violence…” indeed.


September 16, 2015 at 6:53 pm

Another article seemed to claim that this was a Norwegian study. It’s a Swedish study, my apologies. Sweden doesn’t collect demographics data, but I suspect the Norwegian data serves as an adequate approximation.