Why Long-Term Follow-Up Data Is Scarce in Anxiety Research

If you look up any anxiety treatment, the evidence you find will almost certainly come from trials lasting weeks or months. The question most people actually want answered — “will this still be helping me in five years?” — has much less evidence behind it. Understanding why helps you read what does exist more honestly.

How long most trials run

The typical anxiety treatment trial lasts 8–16 weeks. This is enough to measure whether symptoms improve during treatment, but it’s a snapshot. Some trials include follow-up assessments at 3, 6 or 12 months after the treatment phase ends. Very few track people beyond two years.

An umbrella review of CBT meta-analyses noted that the evidence for CBT’s durability is one of its most cited advantages, but observed that the follow-up data varies considerably in quality and length [1].

Why long-term data is hard to get

The scarcity is not accidental. Several structural factors work against long-term follow-up.

Funding cycles. Research grants typically cover a defined period. A three-year grant might fund a 12-week treatment phase, a few months of follow-up and the time needed to analyze and publish. Extending follow-up to five or ten years requires additional funding that is harder to obtain, because funders often prioritize new questions over following up old ones.

Participant dropout. People move, change phone numbers, lose interest, seek other treatment or simply stop responding to researchers’ calls. In anxiety trials, dropout during the treatment phase itself is already substantial. A review of dropout rates in anxiety treatment trials found that roughly 15–25% of randomized participants leave before the end of the acute phase [5]. Over years, attrition compounds.

The people who drop out may not be a random sample. If those who are doing badly are more likely to stop responding, the follow-up sample over-represents successes, making the treatment look better than it is.

Ethical constraints. In an acute trial, a waiting-list control group waits weeks. Asking them to wait years is ethically indefensible, so long-term follow-up studies usually compare treatment groups with each other rather than with an untreated control. This makes it hard to separate treatment durability from natural improvement.

Treatment contamination. Over years, participants seek additional treatment. Someone initially randomized to CBT may later take medication, try another therapy or join a self-help group. By the five-year mark, the “CBT group” and the “medication group” may no longer differ much in what they actually received.

What the limited data shows

One of the longest follow-up studies of CBT for anxiety disorders traced trial participants in Scotland 2–14 years after their original treatment. The results were mixed: while some people showed lasting improvement, a substantial proportion had sought further treatment, and outcomes were variable [3].

For medication, a systematic review of relapse prevention trials found that people who continued antidepressants after responding had significantly lower relapse rates than those switched to placebo. The pooled relapse rate was about 36% on placebo versus about 16% on continued medication across anxiety disorders, OCD and PTSD [4]. But these trials typically measured relapse over months, not years, and the comparison is between continuing versus stopping medication, not between medication and no medication from the start.

A meta-analytic framework from depression (where more long-term data exists) found that relapse rates after psychotherapy tend to be lower than after medication discontinuation, but rise steadily over time. The authors cautioned that the evidence base was limited and that definitions of relapse varied [2].

What “relapse” means in this context

Relapse rates depend heavily on definition. A return of symptoms to a specified threshold is different from seeking treatment again, which is different from meeting full diagnostic criteria again. Studies using stricter definitions report lower relapse rates; studies using broader ones report higher rates. This means relapse figures from different studies are often not comparable. For how measurement choices shape results, see what an anxiety study actually measures.

Questions to ask about follow-up data

1. How long after treatment did the follow-up occur?
2. What proportion of the original sample was followed up, and could the dropouts differ systematically?
3. Was additional treatment allowed or measured during the follow-up period?
4. Was there still a comparison group, or is the follow-up uncontrolled?
5. How was relapse or maintained improvement defined?

What remains uncertain

We know surprisingly little about what happens to anxiety disorders over five, ten or twenty years, with or without treatment. Observational studies suggest that many anxiety disorders follow a chronic, fluctuating course, with periods of remission and recurrence. Whether treatment changes this long-term trajectory, or mostly helps during the acute episode, is an open question. The most honest thing most reviews can say is that short-term treatment produces short-term benefit, and the rest is informed extrapolation.

Thinking about the long term. If you’ve had treatment for anxiety and are wondering whether to continue, taper or try something different, that decision is best made with your prescriber or therapist, who can consider your history, current symptoms and preferences. If anxiety is affecting your daily life, talk to a GP or mental health professional. If you are in crisis or thinking about harming yourself, contact emergency services or a crisis line now: in the US, call or text 988; in the UK and Ireland, Samaritans on 116 123.