Digital Therapeutics for Anxiety: What the App-Based Treatment Evidence Shows

There are thousands of mental health apps available. Many claim to help with anxiety. Some are backed by randomized trials; most are not. The distance between the best-studied digital treatments and the average app in the store is enormous, and understanding that distance matters for making an informed choice.

The spectrum of digital treatments

Digital therapeutics for anxiety range widely in format, intensity and evidence:

Guided internet CBT. Structured, multi-session programs delivered online with regular therapist support (typically asynchronous messaging or brief calls). Examples include programs from research groups in Sweden, Australia and the Netherlands that have been tested in dozens of trials.

Unguided internet CBT. The same structured programs, but without therapist contact. The patient works through the modules independently.

Smartphone apps. Standalone apps offering exercises, tracking, guided meditations, chatbot conversations or brief CBT-based techniques. These range from well-designed products with some evidence to products with no evidence and questionable design.

Virtual reality exposure. VR-based tools for specific phobias and social anxiety, typically used in a clinical setting rather than independently.

Chatbots. AI-driven conversational agents that simulate therapeutic interactions.

Where the evidence is strong: guided internet CBT

The strongest evidence in digital mental health is for guided internet CBT. A meta-analysis of 20 trials that directly compared internet CBT with face-to-face CBT for various conditions found equivalent outcomes [1]. The pooled difference was essentially zero, with the confidence interval spanning a very small range on either side.

An earlier meta-analysis of computer-delivered CBT for anxiety and depression found large effect sizes compared with control conditions, with guided programs consistently outperforming unguided ones [6].

These are not fringe findings. Guided internet CBT for anxiety is recommended in NICE guidelines and is part of the IAPT program in England. The programs are structured (typically 8–12 modules), based on established CBT protocols, and include regular therapist check-ins. They are, essentially, conventional CBT delivered through a different medium.

Where the evidence weakens: unguided programs and apps

Remove the therapist, and the picture changes. A systematic review and meta-analysis of standalone smartphone apps for mental health found a small overall effect on depression but no significant effect on anxiety compared with control conditions [3]. The number of anxiety-specific trials was small, and the quality was variable.

A meta-analysis of acceptance, mindfulness and compassion-based apps found small effects on anxiety symptoms, but noted high heterogeneity and the predominance of waitlist comparisons [2].

The consistent finding across the literature is that guidance matters. Programs with therapist support produce larger effects, have lower dropout rates and show more durable gains. The guidance need not be intensive — brief weekly messages or calls often suffice — but its presence is a reliable moderator.

The dropout problem

High dropout is the single biggest challenge in digital mental health. A meta-analysis of smartphone app trials for depression found an average dropout rate of about 47%, with some trials exceeding 70% [4]. Anxiety app trials show similar patterns.

This creates an interpretive problem. If only 30% of participants complete the program, the positive results apply to that 30%, not to everyone who downloaded the app. The people who drop out may be those for whom the app was not helpful, or those who were too impaired to engage, or simply those who found it boring. The per-protocol results (analyzing only completers) look better than the intention-to-treat results (analyzing everyone who started), and the gap between the two is often large.

What FDA clearance means

Several digital therapeutics for mental health have received FDA clearance in the US. This is sometimes presented as validation that they “work.” The reality is more complicated.

FDA clearance for digital therapeutics typically uses the 510(k) or De Novo regulatory pathway, which requires demonstrating that the product is safe and has some evidence of effectiveness. The evidentiary bar is lower than for drug approval. Some cleared products have supporting RCTs; others have been cleared based on predicate devices or limited evidence.

An assessment of commercially available digital therapeutics found that many lacked adequate published evidence, and that marketing claims often outpaced the supporting data [5]. “FDA-cleared” is not equivalent to “proven effective in a rigorous independent trial.”

How to evaluate an anxiety app

If you are considering a digital treatment for anxiety, the following questions can help separate evidence from marketing:

1. Is there a published RCT of this specific product? Not of CBT in general, not of apps in general, but of this app. Check PubMed or the company’s website.
2. What was the comparison? A trial against waitlist tells you the app is better than nothing. A trial against an active comparison is more informative.
3. What was the dropout rate? If more than half the participants stopped using it, the results represent a self-selected minority.
4. Was the trial independent? Trials conducted by the company that makes the product are subject to the same allegiance effects as any researcher-led trial.
5. Is guidance included? If the trial tested a guided version and the product you’re buying is unguided, the evidence may not apply.

What this means

The digital treatment landscape for anxiety contains genuinely effective interventions (guided internet CBT), promising but less proven ones (some unguided programs and apps) and a large volume of untested products. The most important distinction is not between “digital” and “traditional” but between “tested” and “untested.”

For how to evaluate treatment evidence more broadly, see what a Cochrane review tells you. For the related question of whether online therapy matches in-person therapy, see online versus in-person therapy.

When to look for more support. A well-supported digital program can be a good first step or a complement to other treatment. But if your anxiety is severe or worsening, an app is not a substitute for professional care. Talk to your GP or a psychologist. If you are in crisis, call or text 988 (US) or Samaritans on 116 123 (UK and Ireland).