SUVs, including crossovers are more likely to kill pedestrians than cars. This is according to a preliminary study conducted by the U.S.-based Insurance Institute for Highway Safety.

Thanks to advances in safety, the number of people killed in motor vehicle crashes in the U.S. has fallen from more than 50,000 in 1980 to 36,560 in 2018. Over the past decade, however, the number of pedestrians killed on American roads has ticked steadily upward.

Analyzing a sample of 79 crashes from three urban areas in Michigan, the researchers found greater risk to pedestrians from SUVs. Because the sample size is small and limited to one geographic region, more research will be required to see whether all of the findings hold up in a larger study.

In the Michigan crashes, SUVs caused more serious injuries than cars when impacts occurred at greater than 30 km/h. At speeds of 30 to 62 km/h, 3 out of 10 crashes with SUVs (30 percent) resulted in a pedestrian fatality, compared with 5 out of 22 for cars (23 percent). At 64 km/h and higher, all three crashes with SUVs killed the pedestrian (100 percent), compared with 7 out of 13 crashes involving cars (54 percent). Below 30 km/h though, there was little difference between the outcomes, with pedestrians struck by either vehicle type tending to sustain minor injuries.

The number of pedestrians killed by vehicles rose 53 percent from 2009 to 2018, the latest year for which statistics are available. Pedestrians now account for nearly a fifth of all traffic fatalities, a proportion not seen since the early 1980s.

Earlier research had shown that SUVs, pickup trucks and passenger vans were 2 to 3 times more likely than cars to kill a pedestrian in the event of a crash. However, most earlier studies were based on crash data collected in the 1970s, 80s and 90s. Since then, SUV manufacturers have made substantial design changes.

To provide an updated comparison, IIHS researchers analyzed detailed crash data compiled by the International Center for Automotive Medicine Pedestrian Consortium. Each crash involved one SUV or car and one pedestrian over the age of 13. The median model year for the vehicles involved was 2009, and three-quarters of them were built between 2004 and 2013.

The Michigan crashes are not necessarily representative of those that occur nationwide. However, the injury patterns were consistent with earlier, national studies in showing that SUVs were more likely than cars to throw pedestrians forward and nearly twice as likely to cause severe hip and thigh injuries. These injuries were mainly caused by impacts with the bumper, grille or headlights. That’s likely because the high point of the front profile or “leading edge” of most new SUVs is still considerably higher than that of the average car.

In recent years various SUV manufacturers have adopted more car-like designs. Part of those changes were intended to address the risk that SUVs posed to car occupants. Bumpers and other force-absorbing structures were lowered so that they aligned better with those of cars. As a result, SUVs no longer pose a greater threat to the occupants of other vehicles than cars of comparable weight.

There hasn’t been a similar widespread effort to address the danger that SUVs pose to pedestrians, and the changes made to improve compatibility with cars wouldn’t be expected to improve outcomes for pedestrians. In pedestrian crashes, the location of the force-absorbing structures is less important than the overall shape of the front end.

In a crash with a traditional, block-front SUV, the grille strikes the pedestrian’s pelvis or chest split seconds after the bumper hits the lower extremities, transferring more energy to the pedestrian’s body. It’s possible that a more sloping profile could do less damage.

IIHS plans to use the Michigan crash data to look into what kind of SUV profile poses the least risk to struck pedestrians.