Autism research hits the road

Some scientists are thinking creatively about how to collect data in flexible environments and meet communities where they’re at.

Mobile research: Caitlin Hudac packed everything she needed for an EEG study into two large suitcases, making it possible to assess participants in their homes across the United States.
Courtesy of Caitlin Hudac

On a clear September morning in 2021, developmental psychologist Caitlin Hudac loaded about 120 pounds of luggage into her sedan for a day trip. Hudac, then assistant professor of psychology at the University of Alabama in Tuscaloosa, isn’t usually such a heavy packer, but she was embarking on a bold new mission.

Two years prior, she had received about $110,000 in grant money to collect brain activity data from people with rare mutations in two autism-linked genes. Her goal was to assess whether their responses to basic stimuli such as sound differed from those of non-autistic people or those with idiopathic autism. Then, in early 2020, COVID-19 hit, upending those plans.

As the pandemic started to subside midyear in 2021, the families she had enrolled told her they felt comfortable with in-person data collection. But Hudac’s lab in Alabama — an hour’s drive from the closest major airport — wasn’t easy to visit, she says. “So I proposed, what if I just drove around the country?”

Hudac requested a research sabbatical to relieve her of the next semester’s teaching obligations, and she mapped an initial route from Seattle, Washington, where she had been summering, to San Diego, California. Then she prepared two large bags to fit in the back of her car. One sturdy case, the size of a typical bag checked for a flight, contained most of her technical equipment, including an electroencephalography (EEG) machine. In a worn-out suitcase from her undergraduate days, she packed plush toys, fruit snacks, chocolates and other items to keep her study participants — mainly babies and children — “still and chill” while she collected the EEG data.

In the months to come, dozens of other families enrolled in Hudac’s study, so she rerouted and lugged these bags to homes in cities and rural areas across 33 states. Over 92 days from September to December, she traveled 15,000 miles and collected EEG data from 56 people aged 9 months to late 30s.

Broad reach: Home visits often demand more resources than traditional lab assessments do, but they can also make studies accessible for a more diverse group of participants.
Courtesy of Caitlin Hudac

Hudac, now associate professor of psychology at the University of South Carolina in Columbia, is one of a growing number of researchers who are taking their research on the road to improve diversity and inclusion in autism research — which is overwhelmingly skewed toward white and affluent autistic children who have low-support needs and can more readily travel to a lab. As part of that itinerary, these scientists are adapting experimental protocols and partnering with marginalized communities.

Some of this work is already paying off: One autism project increased its proportion of participants from underrepresented ethnicities by up to 3 percent after offering home visits to collect saliva samples, according to results presented late last year at a conference.

Such gains, though still small, are critically needed, says Brian Boyd, distinguished professor of education at the University of North Carolina at Chapel Hill. “Having research samples that are more representative and reflective of the broader autism population will enhance our understanding of autism.”

Home visits often demand more resources than do traditional lab assessments. Researchers must manage scientific goals while being flexible when collecting data from participants with complex issues and social-communication difficulties. But, Hudac says, home visits are here to stay. “That’s a new forever thing for us.”


or decades, the overwhelming majority of autism research has focused on people of European ancestry and those from relatively affluent communities or with mild traits, who typically live nearer to major research centers or can travel there more easily than those from lower socioeconomic communities or with severe traits.

“Our lab-based studies are 80 percent white, high-education and high-income [families],” says Lauren Schmitt, assistant professor of pediatrics at the University of Cincinnati in Ohio, who uses EEG to probe the neural basis of fragile X syndrome, the most common inherited form of autism; children with fragile X also have varying levels of intellectual disability. “It makes sense, because you have to take four or five hours off of work to come and participate, and we don’t pay a huge amount for [participating in] these studies,” she says.

Like Hudac, Schmitt has been piloting home visits to conduct tablet-based cognitive assessments and collect blood samples and EEG data from a broader range of participants than can visit her lab. Her goal is to understand how levels of the FMRP protein, which is lacking in people with fragile X, influence brain activity.

“We want to capture the full fragile X phenotype by reaching these families in their homes if their kids have irritability or high anxiety, but also if they just can’t take time out of their day to get [to the lab],” Schmitt says.

Home assessments are similarly helping researchers to collect data from children who speak few or no words. Minimally verbal children produce twice as many utterances, and more varied utterances, when prompted in a familiar setting — for example, by their parents during everyday activities at home — than they do with trained examiners in the lab, according to a 2020 study.

“That was way higher than we might have expected,” says Helen Tager-Flusberg, director of the Center for Autism Research Excellence at Boston University in Massachusetts, who led the work.

Home is a more comfortable environment for the child, says Karen Chenausky, director of the Speech in Autism and Neurodevelopmental Disorders Lab at the Massachusetts General Hospital Institute of Health Professions in Boston. Chenausky didn’t take part in the study but has worked with Tager-Flusberg on other studies of minimally verbal children with autism.

Chenausky’s team, for example, has used Zoom to film minimally verbal and low-verbal children as they watch a YouTube video of their choice at home. Measuring the children’s facial movements and comparing them with those of non-autistic children or autistic children with typical speech and language abilities could reveal whether their motor systems differ, she says.

Unlike taking a traditional language assessment, watching a video is a task that autistic children with language difficulties can perform successfully, Chenausky adds. “It’s our way of potentially including everyone in language studies, even kids who don’t talk.”


ome visits may have the potential to reach a broader participant pool, but they present a variety of challenges for collecting data. For instance, things such as barking dogs can render portions of video or EEG recordings unusable, and without a partner, EEG caps can be difficult to place accurately over curly or Afro-textured hair.

Data collection at home requires researchers to be flexible yet rigorous, says Carol Wilkinson, assistant professor of pediatrics at Harvard Medical School. Too strict, and the results may not be generalizable to many people. Too lax, and the science suffers. “We’re figuring out what’s the happy medium,” says Wilkinson, who is collecting EEG data in primary care clinics but hopes to one day collect data in participants’ homes.

Braiding technique: EEG caps can be challenging to place close enough to the scalp on people with curly or Afro-textured hair, which can lead to noisier readings.
Courtesy of Caitlin Hudac

During one home visit with a Black family in New Mexico, Hudac picked an EEG cap for their child that was slightly bigger than what she would have used for someone with the same head circumference and fine, straight hair. With the cap in place, Hudac gently pulled some hair through the cap’s holes to get the electrodes as close to the scalp as possible, increasing their ability to record signals without noise.

“The best way to work with Afro-textured hair is to have a partner, and I didn’t have one, but I happen to have very big hands, so I was able to easily put the net on around [the child’s] hair,” Hudac says. It’s a technique she has mastered over the years and shared in a study published last year with her collaborators. “[With] hair that is in locs or twists or braids, there’s definitely challenges to traditional EEG-style setups,” she says.

For this reason, Black Americans are often excluded from certain clinical trials or, in some cases, asked to cut their hair in order to participate, says Arnelle Etienne. “There’s a lot of strong feelings and trauma, and that may cause people to not want to [participate in studies],” she adds.

While an undergraduate research assistant in the lab of Pulkit Grover at Carnegie Mellon University in Pittsburgh, Pennsylvania, Etienne invented a special clip that holds EEG electrodes against the scalp of someone wearing cornrows. Solutions such as Etienne’s clips could help scientists collect better data from people of color, using other neuroimaging technologies — such as functional near-infrared spectroscopy (fNIRS) — that are suitable for home-based studies.

Midway through her trip, Hudac visited the home of a child with severe sensory sensitivities, a situation that forced her to adapt in other ways. “He doesn’t like things on his head and also is a pacer — he likes to move around,” she recalls.

Hudac managed to get the EEG cap on the child’s head, but having him sit on a couch or chair while he completed an auditory task proved difficult. “He was fighting it, so we took a minute to recalibrate.” Hudac eventually decided to let him stand in front of the speakers, and she pushed all the furniture out of the way. To compensate for such changes, during each home visit Hudac logged anything that happened that could affect the data — including background sounds — and excluded the corresponding portion of the EEG recording from the analysis.

Other researchers rely on computational pipelines that provide information about the quality of the data and remove noise. “Having a measure of artifacts in the data is valuable,” Wilkinson says, and helps ensure that researchers can focus on differences related to biology rather than to the environment or circumstances during data collection.


espite the challenges in a home, most data collected there can be used for research, scientists are discovering. For example, Chenausky, who developed a remote-assessment protocol that includes speech and language tasks, says that for 38 of the 42 participants in one of her studies, the speech data collected at home was good enough to do acoustic analyses. “[That] was a pleasant surprise to me, given the uncontrolled conditions we were working with,” she says.

To collect data that successfully, though, researchers have to prepare in advance. Before each home visit during her road trip, Hudac scheduled phone calls with the parents to learn more about the child and their needs — for instance, if the child had attention issues — and parents had the opportunity to get a sense of what the visit would entail by looking at videos that Hudac posted online. She also shared with families “social stories” — short descriptions of a home visit that included details about what to expect and why.

On the road: Hudac ultimately drove 15,000 miles over the course of three months and collected data from 56 people in 33 states.
Courtesy of Caitlin Hudac

Now back in South Carolina, Hudac has applied for grants to do more home visits around the state — this time together with a small research team. “I would encourage anyone who’s doing [home visits] in the future to think about bringing two people along,” she says. “It’s hard to be one person and manage both the child and parent expectations.”

Having more people on the research team increases the costs, making home visits more expensive than standard lab visits, Hudac notes. And visiting homes every week isn’t sustainable, “but maybe one can try once a month or six times a year.”

To help make that possible, Hudac and her collaborators are applying for a grant that would allow them to create a mobile lab — a van equipped with cameras, EEG and fNIRS machines, as well as other pieces of equipment that researchers may need to assess autistic people at home.

“We keep on encountering families that don’t have furniture, so we bring our own highchair and table,” says Jessica Bradshaw, associate professor of psychology at the University of South Carolina, who is working with Hudac on the grant. “Two experimenters plus equipment and toys — it gets pretty cluttered pretty quickly.”

The mobile lab would increase access to research for marginalized communities and foster collaboration across researchers who are interested in studying similar populations from different angles, Bradshaw says. “We have to think creatively about how to do rigorous science in flexible environments.”

Hudac says she feels a sense of responsibility, and one of her goals is to reduce whatever barriers to research participation she encounters. Those barriers, she adds, “are going to be ever changing, so we can’t ever get complacent.”