The Missing Piece in the Mental Health Puzzle

Why revolutionizing the field requires quantifiable data supported by investors and healthcare leaders

• A significant barrier to accurate diagnosis and effective treatment of depression and anxiety is over-reliance on subjective behavioral data
• Improving outcomes in the mental health space requires companies to push beyond established approaches and focus on objective, quantifiable metrics
• Recent data-driven innovations in mental health measurement and monitoring, in particular physical and digital biomarkers, hold exceptional promise and offer a wealth of opportunities for patients, healthcare systems, and investors
• Biomarkers can be physical or digital, and tradeoffs include quality of data vs. barriers to collection
• New biomarkers will need patient, practitioner and payer buy-in to eventually change outcomes at scale

Last month, the mental health sector was set abuzz by the news that Apple is working with University of California, Los Angeles to develop iPhone and Apple Watch features that will help diagnose depression and anxiety. Why is the world’s most valuable company, best known for the screens we can’t live without, stepping into a field that’s normally the domain of medical experts? It’s one thing for the iPhone to track how many steps you take, for example, but quite another for it to monitor cortisol levels and mood swings. What’s the underlying reason — and the payoff — for Apple to use its hardware, algorithms, and one billion customers to collect and analyze data about mental health?

The answer is complicated, as are most things related to mental health. It’s connected to another key question we’ve been delving into at Able Partners for the last two years: Why is the brain the only organ, as it relates to mental health, that we diagnose and treat based primarily on subjective behavioral reports without objective data?

A cardiologist wouldn’t dare prescribe a treatment plan without the wealth of information an EKG provides, nor would an oncologist recommend chemotherapy without the benefit of diagnostic radiology or biopsy results. Sounds like a no brainer, right? But when it comes to mental health, practitioners rely on their patients’ own narrative accounts of their symptoms. It’s hard to imagine another medical condition we’d consider treating based on a similarly acute lack of impartial information. Though depression and anxiety disorders are the leading cause of disability worldwide, even the terms we use to describe them, including “depression” itself, really aren’t diagnoses so much as rough assessments made by practitioners — usually primary care doctors without mental health expertise. This hit-or-miss approach has become the norm, thanks to a lack of practical methods to measure mental health in any consistent and quantifiable way.

Investors have been paying attention. Having identified a large and growing market with limited solutions, investors are pouring funds into the behavioral health sector. In 2020, mental health venture funding was 5.5x the $275 million investors funded four years earlier, according to CB Insights. For example, Talkspace, which helps users find licensed therapists for virtual sessions, went public in June via a merger with a SPAC in a deal valued at $1.4 billion. LifeStance Health, which offers both online and in-person mental health care services, also IPO’d on NASDAQ in June with a valuation of $7.5 billion. Similarly, Lyra, an employee-focused provider of online mental health care, reported $200 million in new funding in June as well, doubling its valuation to $4.6 billion in less than 6 months.

To date, investments like these have focused mainly on improving the legacy approaches to mental health, such as updating access to traditional treatments including talk therapy and medication management. At Able Partners, however, we encourage investors and healthcare leaders to support companies that are pushing beyond these established approaches and into what we see as mental health’s next frontier: finding new, more efficient, and more accurate ways to not just treat depression and other mental illnesses, but to actually diagnose, measure and monitor them over time.

Improving measurement in mental health is critical for at least four reasons:

1. Current mental health diagnoses are largely based on subjective patient self-reports, rather than objective data, inevitably leading to inaccurate diagnoses and, in turn, ineffective treatments.

2. Those conditions that are accurately diagnosed tend to be imprecisely and broadly defined, making them difficult to correlate to specific effective treatments.

3. Imprecise diagnoses present real barriers to successful clinical trials for new and better treatments, and therefore, are also barriers to investments in new mental health interventions.

4. Identifying actual biological bases for mental illness, as opposed to relying on less precise anecdotal and observational factors, will help reduce the widespread social stigma of these conditions.

Let’s start with the imprecision and sheer inadequacy of mental illness diagnoses themselves. Terms like depression and anxiety have come to serve as blanket descriptions of conditions that are, in reality, far more nuanced and diverse than such monolithic misnomers reflect. Mental disorders are highly heterogeneous in nature and can change significantly over time. In fact, each diagnosis can encompass hundreds of distinct diseases, which are in turn rooted in myriad unique conditions or events — everything from neuroinflammation to endocrine disorders to childhood trauma to grief.

“Our history of medicine tells us that defining a disease by its symptoms is highly simplistic and inaccurate,” writes Joe Herbert, a Cambridge University neuroscientist, in a recent article in The Walrus. “Psychiatrists recognize two types of depression — or three, if you count bipolar — but that’s simply on the basis of symptoms,” he notes. Imagine the countless variations of depression, and yet the experts have only three buckets with which to classify patients! In other words, we need better measurement tools in order to segment diagnoses into more specific, and therefore more treatable, groupings.

Compounding the problem is that identifying the right treatment options once practitioners have made these diagnoses is almost as imprecise and haphazard as the diagnoses themselves. Because we lack enough objective insight into which medications will work for which individuals, patients are often forced to undergo an excruciating trial and error process with multiple drugs before finding relief. The (depressing) reality is that for all the glowing promises made by pharmaceutical companies about their antidepressant medications, just 30% to 45% of patients achieve remission on their first medication, and up to one third of patients fail to respond after trying at least two different classes of antidepressants.

Despite the pressing and growing need for better and more mental health solutions, pharma giants have actually cut research on psychiatric medicine by a stunning 70% over the past decade. Their retreat lies in large part because of those same, overly broad diagnostic categories, which make it very hard — and expensive — to design, execute and get positive results in a clinical trial for mental illness. In other words, the companies running trials enroll participants with numerous subtypes of the condition being studied, many of which are not yet clinically identified, and then take the risk that despite enormous variables, a statistically significant number of people will respond to the treatment being studied.

Lastly, our inability to test for the biological bases of a range of mental illnesses inevitably compounds the stigma already surrounding these conditions. If we could pinpoint, for example, the physiological factors that predispose someone to experience a crippling anxiety disorder, which can lead to problems in school or at work, we could begin to correct the damaging cultural narrative that has long surrounded mental illness. It’s far too common for someone suffering from a mental health issue to feel judgement or even discrimination. Hard data would flip a “blame the victim” perspective on its head by showing what may, in fact, have been present on a cellular level from the moment of conception.

It is time to boldly address the need for better mental health metrics with novel and even revolutionary approaches. The Apple / UCLA collaboration is just one example of how a slew of researchers and founders are — finally — doing just that. Just as the physical health sector has witnessed a spike in direct-to-consumer wellness measurement tools evolving from the humble pedometer to sleep and activity trackers (Oura) to in-home labs (Vessel Health), the mental health space is seeing a number of visionary approaches to measurement. Many of these initiatives center around mental health biomarkers: biological, and more recently, digital indicators of disease that can be measured and monitored far more objectively than is possible through traditional methods like practitioner observation and patient self-reporting.

The hope driving these ventures is that biomarkers will lead to more impartially measured and accurately defined subtypes. For example, a new diagnosis such as Inflammatory Depression Type 1 would correlate with specific treatment protocols. This approach would encourage companies to invest and enroll clinical trials with a higher degree of confidence in their outcomes, leading to new and better treatments. It would also allow practitioners to more accurately match treatments, pharmacological or otherwise, to those patients most likely to benefit from them, improving lives and reducing costs in the process.

The ultimate goal of many of these tools is not just to identify mental illness, track treatment results and monitor large patient populations over time, but to eventually predict and intervene to avoid deterioration in a patient’s condition. For example, Ksana Health’s app sends a notification to users who may have been sedentary on a particular day to remind them that movement improves their mental state, or might nudge a sleep-deprived patient with messaging about how a shut-eye deficit can worsen their symptoms.

That’s not to say this process isn’t somewhat complicated. Biomarkers themselves fall into many categories, including those used in other illnesses such as blood, saliva, genetics, and neuroimaging (EEG or fMRI), as well as newer measurements, such as voice or digital biomarkers based on smartphone usage. Alto Neuroscience, for instance, is looking at EEG (electroencephalogram) biomarkers, while MindX is developing blood tests that can assess patients for mental health conditions including PTSD, depression and bipolarity, anxiety, schizophrenia and more. Ceretype Neuromedicine is harnessing fMRI (functional magnetic resonance imaging) to connect neuromolecular activity to specific psychiatric outcomes, in order to help practitioners fine-tune diagnoses and treatment plans, as well as aid in the development of new drugs. Genomind is exploring the genetic basis for mental health conditions, while Ellipsis is developing “voice as a biomarker,” analyzing acoustics and word choices in 90 seconds of natural speech for signs of mental health deterioration over time.

Digital biomarking, also called phenotyping, represents another exciting new frontier for less invasive measurement. Ksana and HealthRhythms are working on digital phenotyping technology that could tap into a patient’s location, physical activity levels, vocabulary, and keystroke patterns via smartphone, offering practitioners a broader and more data-driven view of their patients’ behavioral patterns and mental states.

To be sure, there remain hurdles to getting these nascent but promising technologies out of labs and trials and into the hands of patients who stand to benefit from them most. The current insurance reimbursement model pays practitioners for administering the standard GAD-7 and PHQ-9 surveys, and insurers and other payers will need to see evidence of the effectiveness of these new approaches before covering them. Similarly, clinicians will need to learn and trust these new tools and adapt to new modes of treatment. There is also the challenge of striking the right balance between user-friendliness in these applications and methodologies, and the quality of data they are able to provide patients and practitioners. While smartphone apps like Ksana are both cheap and easy for patients to use, the data they provide will inevitably be less detailed than that offered by, say, Ceretype’s fMRI, which requires the use of an MRI machine operated by a specially trained therapist, and therefore a visit to a lab or hospital.

Still, taken as a whole, this sector holds exceptional promise. The advances being made in this space point to a future where conditions like depression will no longer be seen as monolithic illnesses to be navigated by subjective diagnostic techniques and addressed with limited (and often ineffective) treatment options, but to be diagnosed and treated in a far more nuanced, accurate, effective and data-driven way. Of course these innovations also offer investors a wealth of opportunities. We believe that mental health measurement technologies are one of the few interventions that have the potential to meaningfully improve mental health, at scale, within our lifetime. After all, it’s hard to improve what we can’t measure — a conclusion with which the leaders at Apple clearly agree.