Behaviors National changes Social Distancing

Protective measures in the US

While quite a bit of research has reported which COVID19 policies work and don’t work on a country or state-level, VERY little peer-reviewed research has reported what individuals are doing to curb spread. In other words, are individuals conforming with the recommended protective health measures?

A study was just published in the International Journal of Environmental Research and Public Health.

Scientists used data from the COVID19 Impact Survey. This survey was designed carefully to pick a random set of households in the US to participate. They did this extra work so the results of the study could be generalized to the entire U.S. (instead of just one state/city or type of state/city). This survey was done three consecutive times (April, May, and June of 2020) and 25,269 people participated.

What did they find?

  • In the United States, 95% wash their hands, 90% kept 6 feet away, 86% wore a mask, and 82% avoided crowded places (Figure 1)
  • Of 19 protective measures, an average of 7 protective measures were taken
  • Who took protective measures more? Higher incomes, insurance, higher education levels, large household size, age 60+, females, minorities, those who have asthma, have hypertension, overweight or obese, and those who suffer from mental health issues during the pandemic
  • Who took protective measures less? Suburban and rural areas, and the Midwest and West
  • People who wore masks increased from April to June 2020 (Figure 2)
  • Participants who were positive for COVID19, knew an individual with COVID19, or knew someone who died from COVID19 had a stricter lifestyle. This was especially true regarding washing hands, avoiding public places, and canceling social events.
  • Some protective measures are linked to others (Figure 3)

Now, what people say they do may be different from what they actually do. However, this still makes my little epidemiologist heart happy. People know the social desirability of our community, which lines up with effective protective measures. Also, this data uncovers patterns that would otherwise have been obscured. Epidemiologists can leverage this knowledge to more quickly and effectively curb COVID19 spread.

I’m hoping that COVID19 Impact Survey’s hard work continues because I would love to see numbers for July-September too.

Love, YLE


COVID19 Impact Survey:

Behaviors National changes

BLM protests vs. Sturgis Motorcycle Rally

Why was the COVID19 spread so different for these two events? A valid question to follow-up the Sturgis post yesterday.

I (and other public health officials) hypothesized that BLM protests would exacerbate COVID19. Protesters reduced social distancing behavior, were shouting and yelling (i.e. spreading droplets), and congregated in large groups. This could have spurred a resurgence. However, as more and more data comes in, this does not seem to be the case.

One white paper, published by 5 scientists, looked closely at BLM protests in 315 of the largest US cities from May 15-June 9.

Their purpose? To see if the BLM urban protests impacted social distancing in communities, COVID19 cases, and COVID19 related deaths.

What did they find? • Stay-at-home behavior increased across these cities during the protests. The authors attribute this to fear of danger, avoiding travel outside the home due to additional traffic congestion or street closures, or due to lack of available activities from businesses closures near protest sites. • There is no evidence that protests sparked COVID19 case or death growth 5 weeks following the onset of protests

Why was our hypothesis wrong? • I didn’t think about non-participants’ behaviors when I made the initial hypothesis. More people, who were not protesters, stayed inside. This is consistent with the literature on crime, violence, and perceived safety amongst the general population.

So why was BLM different than Sturgis? • The sheer amount of people in one place. Both had about 500,000 people attend, but BLM was spread over 500 cities while Sturgis was in 1 city. The high density at 1 event makes it far more difficult to social distance. • Difference in spaces (Sturgis also took place in bars, tattoo parlors, hotel rooms, etc.) • Difference in testing behaviors. I can only speak for Texas, but we offered free COVID19 testing for protesters. We are hearing anecdotally, from contact tracers in the area, that bikers are either not interested in getting tested or refuse to admit they went to the rally if they are tested • Other hypotheses? Differences in face mask use, ages of attendees, and travel to and from the sites (protesters did travel to BLM events, however this was not 100% like Sturgis).

Why is this comparison important? It contributes to the current discussion around policies for controlling the spread of disease. Especially when it comes to large events. Because we ALL want to get make to normal one day.

Love, YLE


Behaviors Social Distancing

Spread on a bus

Back in January, there was an outside worship event at a Buddhist temple in China. 293 people attended, of which, 126 traveled to the temple in 2 buses.

Bus ride: The ride was 50 minutes each way. The buses had an air-conditioned system, windows (that were kept shut), and there was no bathroom on board. Everyone remained seated and no one wore masks (this is before public awareness of COVID19). Everyone sat in their original seats on the ride back.

The event: Once the buses arrived at the temple, worship took 150 minutes and the event included a luncheon (10 people at each round table). Bus riders were randomly mixed with others at the event. There was a slight breeze and everyone had close contact.

What happened?

• Bus #1 (59 passengers+1 driver) had no index case.

• Bus #2 (67 passengers+1 driver) had an index case. The index case was asymptomatic. He sat in the middle of the bus (3rd seat in the 8th row).

• Bus #2 had an attack rate of 35%. In other words, 24 other passengers were infected during the 50-minute drive there or back. Bus #1 had an attack rate of 0% (even though they mingled with everyone at the event).

• Cases were scattered on Bus #2. There was no apparent pattern, other than distance from index case

• Among the worshipers that attended the event (but were not in one of these two buses), 7 tested positive (4% attack rate).


• There is airborne transmission with recycled air.

• In closed environments, COVID19 is highly transmissible.

• The index case, who was asymptomatic, achieved enough viral shedding by just breathing to cause a high secondary transmission rate

Translation: Even if you don’t feel symptoms, it’s important to wear a mask. You could be spreading the disease without even knowing it. And, be sure you’re in a space with air circulation (that is not recycled); open your window on the bus.

Love, YLE

Data Source: Published 3 days ago in JAMA:

Behaviors Long-term effects

Domestic Violence and COVID19

For those of you that don’t know, I am a violence epidemiologist. My research lab focuses on how violence is contagious (just like infectious diseases) and predictable. Because if it’s predictable, then it’s preventable.

We, violence epidemiologists, have been paying very close attention on how stay-at-home policies and isolation have impacted all types of victimization, including domestic violence.

We have consistently hypothesized that pandemic stressors (like job loss, mental health, and lack of social support) have exacerbated the risk of violence at home. Also, domestic violence hotlines have reported 25-50% increase in calls and 150% increase in website traffic.

However, no research has asked victims, themselves, how violence has changed during the pandemic (compared to before the pandemic). Well, the first study was published yesterday…

What was found?

• The majority of victims reported that there was NO CHANGE in victimization during the pandemic compared to prior to the pandemic.

• Among victims that did say victimization changed, more victims said it got BETTER compared to getting WORSE.

• Victimization that got WORSE during the pandemic was due to an increase in hitting, slapping (physical) and rape (sexual).

• Victimization that got BETTER during the pandemic was due to a decrease in physical violence.

These results are surprising and certainly not what the authors hypothesized.

What could explain this discrepancy?

• Pandemic or not, the majority of domestic violence is through controlling behavior. With stay-at-home policies implemented, we can imagine that perpetrators may have more control over victims and more knowledge about whereabouts, thus running into less conflict.

• This study sample was recruited through social media, and then by default, the participants must have had access to a computer or smartphone and internet. It is possible that the most severe of victims do not have these freedoms and so were systematically missed.

Nonetheless, the majority of victimization did NOT change=victims continue to be isolated with their perpetrators. The realities of this crisis are immediate and much more research, engaging victims themselves, is needed.

If you are in need of help, here are a few national resources for you. Emergency domestic violence shelters are STILL open and STILL helping women AND men that are victims of abuse.

National Domestic Abuse Hotline: 1-800-799-7233

National Domestic Abuse Chat (if you can’t talk on your phone):

Dallas Area Rape Crisis Center: 972-641-7273

Love, YLE

Data Source: From the brilliant 🙂 Jetelina KK, Knell G, Molsberry RJ. Changes in intimate partner violence during the early stages of the COVID-19 pandemic in the USA. Injury Prevention.

Behaviors Long-term effects

Health Behavior Changes during COVID-19 Pandemic

Back in April, my colleague and I asked YOU to take a survey via YLE (and other social media platforms).

In total, we reached 47,796 social media users (i.e. people who read the post), and, ultimately, 2,440 of you took the survey. This is the first peer-reviewed, published analysis of YOUR data (that is if you participated).

The purpose of this paper: How did healthy and unhealthy behaviors (physical activity, substance use, and sleeping) change in the beginning of the pandemic (compared to before the pandemic):

• For the majority of people behaviors remained the same after immediate shut down policies

• Among those that reported changes…

o Tobacco use: More people increased use compared to decreased use after stay-at-home orders were put in place. This was especially true among females and those with depression

o Marijuana use increased…especially among those with depression

o Alcohol consumption increased…especially among those with kids, were college graduate, and had depression

o Physical activity decreased…especially among females, among those that spent more time at home, and among those with depression.

o Sleep quality worsened…especially among females, college graduates, those with more than 1 comorbid condition, and those with depression

• Why did we see changes?

o Boredom

o More worried

o Stress relief

o Less motivation

o More time available

So…what? These behaviors have a direct impact on our health and our overall quality of life RIGHT NOW. These changes will also have long-term effects. The full picture will take a few years to see, but we are starting to get a peek.

Other papers with YOUR data are in the pipeline. I’m particularly excited about my paper on intimate partner violence, which is in press now. We also have an alcohol consumption paper and a depression paper coming soon. Stay tuned.

Love, YLE

P.S. If you’re ever asked, please take part in research surveys and studies. PLEASE. (Census…cough cough). We really only know what’s going on when you are willing to share your experiences.

Data Source: