The Therapist in Your Smartphone

How data tracking may bring
mental health treatments to the masses

cbt advances

Photo illustration: Alexander Hadjidakis

By Keri Wiginton

Medically reviewed by Neha Pathak, MD

March 26, 2021

Keri Wiginton is a journalist in Madison, WI. She has lived with depression and generalized anxiety disorder for most of her life.

All I need to do is plant my feet on the floor. But here I am, trapped in what feels like hardening cement, willing my brain to heave myself out of bed. But my inert body is impervious to my own pep talk. Even breathing feels like effort. 

This is how I spent most mornings before I pulled myself out of a lifelong depression. And I got better not with repeated sessions with human therapists, but with a tool almost everyone owns: a smartphone. It took only a few months of practice with a meditation app to learn how to interrupt my mental loops of worry and to pare back on my antidepressants.

My recovery left me wondering. If I can rely on this device to teach myself to change my thoughts and behaviors in the depths of my gloom, what else can it do? 

I already use my smartphone to monitor a lot of body functions, like my heart rate and daily step count. Apps also track my menstrual cycle and sleep patterns. Could our phones also detect the early warning signs of mental illness and alert our doctors, before we sink into depression or spiral into a manic episode?

That is precisely where behavioral research, led by David Mohr, PhD, director of Northwestern University’s Center for Behavioral Intervention Technologies, is headed.

Most people who need mental health therapy don’t get it. Pre-pandemic statistics show that 1 in 5 adults in the U.S. will have a mental illness in a given year. But less than 43% will get any kind of treatment. There are many barriers that put care out of reach, with cost and access at the top of the list.

Integrating mobile tools into the health care system might fix that.

The need for digital mental health treatment is more pressing than ever now that 40% of U.S. adults are grappling with depression, anxiety, trauma, or substance use because of the coronavirus pandemic.

“With COVID, we’ve moved to telehealth, and that may overcome some of the geographic barriers to care,” Mohr says. “But it doesn’t solve the problem that we’ll never have enough clinicians -- in this country or any country -- to take care of the need. Digital mental health provides another tool for reaching people. 

“The question is,” he adds, “how do we make it available?”

People want easy access to these kinds of tools. Studies show more than half of those with a mobile device have downloaded some kind of health app. The most popular ones offer meditation, mindfulness, or relaxation skills. 

That might make smartphones the ideal therapist -- or at least the perfect therapist assistant. Always available. Discreetly checking on you without being intrusive. Supportive but not too pushy. And intelligent enough to know exactly how and when to help. All that from a device that fits into your pocket or purse.

Revelations From Data Tracking

Most U.S. adults own a smartphone, and these gadgets know a lot about us. Built-in sensors can track how we communicate, travel, move, and interact with our devices. 

Researchers like Mohr want to harness this “digital exhaust” for our psychological benefit. 

Using sensor data to target users’ behaviors and emotions is called personal sensing. The field is young, but research is promising, Mohr says. If mental health professionals can access this data and it proves to be reliable, it could improve all kinds of treatment.

“Let’s say a therapist sees that a person’s sleep rhythms are really off. They go to bed at 3 a.m. one morning and then 11 the next night,” Mohr says. “We know that creating a more stable sleep habit is helpful in managing mental health problems.”

So how does your device know when you’re asleep? For some people, phone usage alone is a good tipoff. Mohr says younger folks tend to use their phones until they conk out. Then they reach for them first thing when they wake up. Light and sound sensors can also help. If it’s dark and quiet for hours, people are more likely to be asleep, he says.

GPS data show that people who are depressed tend to leave home less often compared to others. When they do go out, they visit fewer places and leave sooner.

Mohr and his colleagues also look at GPS sensors. They use this data to track how movement relates to someone’s mental state. For instance, their research found that people who spent more time at home -- at least before the pandemic -- were more likely to be depressed. And if they did go out, they went fewer places and were less likely to spend much time at each location.

But perhaps more insightful, Mohr says, is the rhythm of a person’s behavioral patterns, which can be more random when their mental health is off. For example, when people are depressed, they might start their workdays a little late. Or they skip their regular Saturday morning shopping trips or daily workouts.

“We tend to be creatures of habit,” Mohr says. “But when we’re distressed, those habits get thrown off.”

With a user’s consent, a smartphone can silently monitor hiccups in someone’s routine.

Finding Meaning in the Math

For his research, Mohr and his team had volunteers install an app on their phones that would track their movements. From time to time, the people in the study answered questions about how depressed they felt at any given moment. They also reported when they went to bed and when they awoke.

Researchers then entered that information into a program that used raw sensor data -- GPS movement -- to develop an algorithm that permitted them to map a user’s location over each response. 

Eventually, the hope is to transmit this kind of data -- with smartphone users’ permission, of course -- directly from their digital devices to a program that can synthesize the data. In theory, machine learning could alert a doctor or an app that a user is displaying symptoms of depression, social anxiety, bipolar disorder, or just heightened stress. A health care professional or smartphone app could then suggest behavioral interventions without any action by the user.

Studies on personal sensing have produced impressive results, Mohr says, but the research hasn’t translated into many real-world therapeutic interventions or applications yet. Companies are working on it, but he thinks it will be several years before this kind of technology is available broadly. The challenge, he says, is taking the raw data and integrating it into digital mental health tools in a meaningful way for health care professionals and patients.

“If your phone can detect your depression is getting worse, certainly that’s useful information. But what do you do with that?” Mohr asks. “What [apps or health care pros] need to do is understand how to use that information so they can give people some sort of positive action they can do to help themselves get better. ... There’s still a lot of challenges before we get there.”

My Experience With Therapy

For now, Mohr and other researchers believe current digital mental and behavioral health tools are an important way to bridge the gap in care. I know it did for me. 

I first sought help for my mental health when I was in high school. I’d vomit from stress and was afraid to sleep. I had nightmares almost every night for months. I felt hopeless. My therapist diagnosed me with clinical depression at age 17.

"Dredging up my past every week during therapy sessions with psychologists left me feeling only worse."

Keri Wiginton

Keri Wiginton

I felt better in college. I wasn’t on medication, but mine was a high-functioning depression. I made good grades, socialized with friends, and stayed in bed during the weekends if I needed.  

But at age 25, I ended up at an urgent care clinic with a panic attack. I couldn’t breathe. My heart refused to stop racing. My normally low blood pressure was sky high. At the time, I was working my dream job as a photojournalist, and for a year I had woken up every morning with a knot in my stomach. I don’t know how long I’d had insomnia. My doctor added generalized anxiety disorder to my chart. Medication helped. But not enough. 

Even with insurance, I’ve never had a wealth of options for mental health care. I tried traditional talk therapy, but I never found a good fit, at least not with the psychologists in my insurance network. The co-pays were and still are too much. And dredging up my past every week left me feeling only worse. 

I felt as if I was drowning. I wanted to learn how to swim. So I turned to my phone. 

A friend told me the Headspace meditation app helped ease his anxiety. I decided to check it out. One day, I sat down on my bedroom floor, took some deep breaths, closed my eyes and began. “Relaxation, here I come,” I thought. 

My initial attempts at meditation were jarring. The first 30 seconds were especially agonizing. The app awakened me to a swirl of negative thoughts in my head: Am I doing this right? This isn’t going to work. I’m never going to feel better. What’s the point? It felt like I’d entered a room full of screaming people. Had this been a 45-minute session with a live therapist, I might have walked out, vowing never to return. But my phone was always there, which meant I could do a few minutes at a time. 

So I stuck with it. After all, it took just 10 minutes a day. And after about a month of daily practice, I started looking forward to meditating. I worked up to 20 minutes, twice a day.


The app taught me something in-person therapy never had: I could alter my thought patterns if I paid attention to them. And not just during my sessions. I could subdue my whirling anxiety during the day, too. After a few months, something amazing happened. I slept better and my mind started to feel settled.  

Five years later, I still practice app-based mindfulness and meditation. But after my first 6 months, and with my doctor’s help, I was able to taper off my antidepressant. Now 39, I had my first relapse in years when my cat died of a brain tumor during the pandemic. But this time, I knew what to do. I reached for my phone to get myself centered.

None of my former therapists or doctors told me about these apps and digital therapeutics, even though I repeatedly asked them about skills-based options. It took my own curiosity and a little bit of serendipity.

Therapy on Demand

Emily Lattie, PhD, an assistant professor who works with Mohr at the Center for Behavioral Intervention Technologies, sees the value in meeting people where they are. Like my experience, she thinks this kind of easier “on-the-go” access might lead to more engagement and better therapy outcomes for other users. “I think mobile apps, if used with commitment, can be really helpful for people who want an individual self-management plan moving forward.” 

That includes those who need help with substance use disorders. Many quit-smoking or quit-drinking apps utilize techniques based on cognitive behavioral therapy (CBT), a form of psychotherapy that teaches you how to identify and reframe unhelpful thoughts and behaviors. 

There’s plenty of evidence that consistent CBT practice helps people learn new skills to control their substance use. That’s why it’s already a vital part of in-person treatment for alcohol or drug addiction. But like other forms of mental health care, those who need help the most don’t always have access to it. 

That’s why Kathleen Carroll, PhD, who was a professor of psychiatry at Yale School of Medicine, worked with a team to create Computer-Based Training for Cognitive Behavioral Therapy, or CBT4CBT. More than eight clinical trials show that this self-guided program is cost-effective and works as well as, if not better, than face-to-face substance use treatment, said Carroll, who died in December 2020. WebMD spoke to Carroll in late October.

She explained a CBT4CBT session like this: Users log on from a digital device, usually a computer, for a 30-minute session. They choose from a selection of different lessons, such as dealing with cravings or standing up for yourself. 

They watch a video clip in which a character gets into a tough situation -- maybe an insistent friend wants someone with alcoholism to go drinking. CBT4CBT uses narration and animation to explain how to say no and mean it. Users then watch the same scenario, but this time the character uses a CBT skill to avoid temptation.  

Clinicians in 35 states and Canada are already using CBT4CBT with their clients, Carroll said. Future versions may take aim at teen vaping and chronic pain-induced opioid dependence. But there are still challenges to broadly disseminating the program, in part due to the difficulties of obtaining the right kind of billing code to collect payment from insurers for these novel digital therapies.

“We offer a whole course of treatment for $100. But even that, for substance use, is a tremendous barrier for people,” Carroll said. “Some insurance companies see that it’s effective and saves money. And there’s an awful lot of interest by clinicians.”

But mental health care professionals can’t charge insurance companies for CBT4CBT because there isn’t an approved reimbursement code to file it under. “It’s a huge frustration, but we’re going to keep pushing,” Carroll said.

Carroll’s program isn’t the only one that comes with a fee. Users pay anywhere from $13 a month to $200 a year for some mental health apps. Ones that offer access to trained therapists charge even more. And right now, the U.S. health care system isn’t prepared to cover the costs of most digital tools, even if they are recommended by a doctor. But recently, there’s been some movement on the mobile health front.

To help people during the COVID-19 pandemic, the FDA has temporarily eased some guidelines on how doctors use evidence-based wellness apps and digital CBT to treat psychological conditions. 

In March 2020, the FDA authorized Somryst, a prescription-based app that provides CBT for insomnia (CBT-I). And some not-for-profit providers, like Kaiser Permanente, offer their clients free behavioral health platforms, like MyStrength, or cover the $70 annual subscription cost for Calm, a meditation and relaxation app.

Are these looser rules around mobile health tools here to stay?

“I hope so,” Lattie says. “I’m seeing so many individuals on the more traditional patient side and the more traditional provider side really getting a lot of benefits from the rapid changes that we’ve had to make, and so I’m hopeful those benefits will be able to stick around.”

There are thousands of options to download. But right now, it’s hard to know which ones are reliable.

Some nonprofits, like Psyberguide, offer expert reviews based on credibility, user experience, and transparency. And in January, Mohr and other researchers, along with representatives from the health care industry, published a set of guidelines on how to push digital mental health treatment forward in the U.S. They’re calling for a set of evidence-based standards and reimbursement codes. That way, doctors will know which ones to suggest and how to cover treatment costs.

“Once that’s accomplished,” Carroll said, “millions of untreated and undertreated people will have affordable access to these clinically validated, effective digital therapies.”

I’m an introvert who’s never clicked with a traditional psychologist. So digital therapeutics work really well for me. They are only one way I manage my mental health, though. I also exercise, eat healthy, and check in with my doctor. I take medication when I need to. But mobile apps are effective because they’re available 24/7, and all I have to do is reach for my phone.