Among adults tested for the coronavirus at 11 U.S. healthcare facilities in July, those who were infected were about twice as likely to have dined at a restaurant in the previous 14 days, according to a US study.

Otherwise, activity levels were similar in people with or without COVID-19 in other respects. Those included shopping, social gatherings at home, going to an office, salon, or gym, using public transportation or attending religious gatherings.

“Masks cannot be effectively worn while eating and drinking, whereas shopping and numerous other indoor activities do not preclude mask use,” researchers said in the report on Friday in the U.S. Center for Disease Control and Prevention’s Morbidity and Mortality Weekly Report.

“Eating and drinking on-site at locations that offer such options might be important risk factors associated with SARS-CoV-2 infection,” they added.

Severe COVID-19 less common in patients with GI symptoms

People with gastrointestinal symptoms related to the new coronavirus, like diarrhea, nausea and vomiting, may be significantly less likely to develop severe COVID-19 and die, a new study found. New York City doctors looked at 635 COVID-19 patients, expecting to see worse disease when the GI tract was involved.

To their surprise, patients admitted with GI symptoms had 50% lower odds of severe COVID-19 and death, compared to patients without GI symptoms, even after accounting for age, race, and underlying medical conditions.

Also unexpectedly, patients with GI involvement had lower levels of inflammatory proteins in their blood. A subset who underwent closer inspection of their intestines had virus particles in gut tissues, but relatively little inflammation, and low activity of genes responsible for making inflammatory proteins, doctors found, according to a paper posted on medRxiv on Wednesday ahead of peer review.

When the New York doctors collaborated with Italian colleagues to study 287 COVID-19 patients in Milan, they saw the same link between GI involvement and less-severe disease, Dr. Saurabh Mehandru of the Icahn School of Medicine at Mount Sinai told Reuters. Mehandru’s team has also found that factoring GI symptoms into the initial patient assessment may help identify those at risk for more severe disease.

Antibody-binding might not ‘neutralize’ the virus

A so-called spike protein on the surface of the new coronavirus helps it invade cells, and some antibodies being tested as treatments work by attaching to the spike and disabling it. But researchers have discovered in test-tube experiments that merely binding to the spike protein is not necessarily enough to “neutralize” the ability of the virus to break into cells.

When they exposed coronavirus particles to antibody-rich plasma from 25 people recovering from COVID-19, all of the antibodies attached themselves to the spike protein. However, a few plasma samples failed to neutralize the virus and were no more effective than plasma from uninfected people.

The findings might help explain why convalescent plasma therapy does not always work, the researchers say. They did not use active virus particles for their experiments. Still, study leader Andrés Finzi of Université de Montréal told Reuters the findings stress the need to learn more about the different shapes the spike protein may assume as the virus breaks into cells, and how to block them.

“Efforts to better understand the link between antibody interaction with the spike protein and virus neutralization might assist ongoing vaccine efforts aimed at eliciting neutralizing antibodies,” the researchers conclude in a paper posted on Tuesday on bioRxiv ahead of peer review.

New system groups hospitalized COVID-19 patients by risk

A simple 21-point scoring system helps assign hospitalized COVID-19 patients to different risk groups, UK researchers reported on Wednesday in The BMJ.

“The score does not require an app or any other technology, beyond perhaps a pen or pencil if you can’t count up to 21 in your head,” Dr. Calum Semple of the University of Liverpool told Reuters. The score takes 8 factors into account including age, other illnesses, kidney health, and oxygen levels in the blood.

Based on the result, patients are assigned to one of four groups. The risk of dying from COVID-19 is 1% in the low-risk group, 10% in the intermediate-risk group, 31% in the high-risk group, and 62% in the very high-risk group. The ISARIC Coronavirus Clinical Characterization Consortium developed its “4C” scoring system using data from 35,463 patients and validated its accuracy in another 22,361 patients.

As pressures on health services increase, being able to identify patients most likely to need escalated care becomes particularly important, Semple said in a news release.