After Police Kill Unarmed Black People, Sleep Worsens — but Only for Black People

Black people in the United States are more likely than white people to report that they do not sleep much, research shows. On average, they live in louder neighborhoods, work longer hours and pick up more late-night shifts — concerning to public health experts, since sleep deprivation is linked to chronic health issues and early death.

But a group of public-health researchers from multiple universities and the National Institutes of Health wondered whether unequal exposure to police violence could also be contributing to racial sleep disparities, since those events are known to increase hypervigilance, worry and post-traumatic stress. They designed a pair of complex studies to measure how police killings of unarmed Black people affected sleep among Black and white people over time. The results were published Monday in the journal JAMA Internal Medicine.

Black people were consistently more likely to report harmfully low levels of sleep after such a killing than they did before it occurred, the researchers found, regardless of whether the killing was a nearby event or a high-profile incident captured in media. The researchers did not find substantial effects on sleep among white people in either case.

Sign up for The Morning newsletter from the New York Times

Dr. Atheendar Venkataramani, an associate professor of medical ethics and health policy at the University of Pennsylvania’s Perelman School of Medicine, was a co-author of the studies. He said the findings reflected “the general human tendency to interpret events — and disparities in events — in ways that apply to you and your future and your family’s future.”

Venkataramani’s lab, the Opportunity for Health Lab, uses statistical data to investigate the relationship between economic opportunity and health outcomes. He said that standard health questionnaires and clinicians, including himself, tended to ask patients about behavioral risk factors but that “we don’t really collect data with these kinds of timely social exposures in mind.”

“We’re never really asking, ‘Hey, did you see something on the news that made you kind of rethink your position in society or how you feel about your future?’” he said.

“These things change people’s outlook about where they stand in society,” he said, “and they can get under the skin to affect health. Sleep is one of those things that can move very exquisitely in the face of these types of events.”

The new studies on sleep involved federal data from the Behavioral Risk Factor Surveillance System and the American Time Use Survey between 2013 and 2019. Researchers used those time-stamped surveys from about 190,000 Black people and about 1,846,000 white people who had been called at random by phone and asked about, among other topics, how much sleep they got.

Then, using statistical data from the Mapping Police Violence database, the researchers identified whether a police killing of an unarmed Black person had occurred in a survey respondent’s state within the previous three months. If they found one, they compared the respondent’s sleep duration with that of people who had been called before the killing. They also compared the answers with those of people surveyed at a similar time, but outside the region.

Survey responses were sorted by whether the respondents’ total sleep duration fell below seven hours, which is considered “short sleep,” or six hours, considered “very short sleep,” since that threshold has been even more closely associated with poor health outcomes.

After controlling for an array of factors, such as seasonal temperatures and unemployment rates, they found that Black people were 2.7% more likely to experience less than seven hours of sleep in the first three months after an officer had been involved in killing an unarmed Black person in their state compared with before the killing, and 6.5% more likely to report less than six hours of sleep compared with before.

To address potential bias, the researchers looked at associations between sleep and other events, such as police killings of armed Black people or unarmed white people, but they found no significant links. They also applied regression models to samples of white respondents and found that associations between sleep and police killings were not statistically significant.

In order to account for the fact that police killings were likely to affect people across state lines, they designed a second study, this one looking at the influence of high-profile killings on a national level. The study compared changes in sleep patterns among Black survey respondents before and after the killings with changes among white respondents — essentially subtracting the differences seen in white respondents from the ones seen in Black ones.

Here, the magnitude of the findings was even larger. In the national-level analysis, researchers found that Black people were 4.6% more likely to report less than seven hours of sleep and 11.4% more likely to report less than six hours of sleep in the months after a killing compared with white people surveyed during that time.

Karen Lincoln, a researcher at the University of California, Irvine, who studies social determinants of health disparities and was not involved in the study, said she would have liked to see researchers include a more robust set of sleep metrics beyond duration, since other factors (such as how often it takes to fall asleep or how often they wake up during the night) are often linked to stress.

Still, Lincoln called the study “interesting and provocative,” saying it was “a very important conversation to have around structural racism and how we measure it.”

Venkataramani said he hoped that researchers aiming to measure other timely social factors in health — for example, the effects of immigration policy or school shootings on communities — might find the rather intricate study design to be instructive.

“The goal of this, really, was to be a ‘methods’ paper,” he said. “You have to cut this many different ways, try many different things, not privilege any one approach over the other” when assigning various exposures, he said. Then, he added, researchers can “show that regardless how you do it, you’re still seeing that same signal.”

c.2024 The New York Times Company