A 2026 study across 22 clinics in Japan found that 83 of 407 adult psychiatric outpatients—20.4%—met criteria for current generalized anxiety disorder (GAD) in a structured diagnostic interview.
The result applies to people already receiving psychiatric care, not to the Japanese population as a whole. Participants had passed their first visit and had a recorded main diagnosis involving a mood disorder or an anxiety, stress-related or somatoform condition.
What the study examined
The adults in the study were between 18 and 79 years old. Their mean age was 43.9 years, and 60.9% were women. People with psychotic disorders, dementia, recent substance-use disorders, seizure disorders or imminent suicide risk were excluded.
Using the Structured Clinical Interview for DSM-5, researchers estimated that 28.3% of the sample had experienced GAD at some point in their lives. Major depressive disorder was present in 33.9% of all participants.
Among the 83 people who met criteria for current GAD, depressive disorders were reported in 65.1%, including major depression in 53.0%. Other reported co-occurring conditions included social anxiety disorder, specific phobia, agoraphobia, panic disorder, adult ADHD, obsessive-compulsive and related disorders, and bipolar spectrum disorders.
Possible gap between records and structured assessment
Only 22 patients had GAD documented in their medical records before the study, and 17 of those were confirmed to meet the study criteria. Among 175 patients previously diagnosed with a depressive episode, 37—21.1%—met criteria for GAD.
The researchers interpreted these findings as suggesting that GAD may not always be recognized in routine care. However, the study did not directly compare the structured interview with everyday clinical diagnosis, so it cannot establish the extent of underdiagnosis.
The study also described Japan as lacking an insurance-covered, GAD-specific treatment pathway. The researchers said care may therefore default to addressing depression or another recorded diagnosis. The findings do not show that any particular treatment is appropriate for an individual.
Participants with GAD reported broader difficulties
Compared with 225 outpatients who had another current psychiatric diagnosis but not GAD, those with GAD reported lower self-rated health and quality of life, along with more depressive symptoms, anxiety and psychological distress.
They also reported greater disability affecting social life and family or home responsibilities. Work or school impairment and days missed were not significantly different between the groups, although the GAD group reported slightly more underproductive days—3.3 compared with 2.6.
Patients with GAD were somewhat younger, more often women, more likely to be in households earning under 4 million yen annually and more likely to report adverse childhood experiences. After adjustment for other factors, age and sex were not significant. Because the study was cross-sectional, these associations do not demonstrate causes.
Screening tools did not identify every case
The study also assessed two questionnaires. At the standard GAD-7 cutoff of 10 or higher, 45 of the 83 patients with GAD were identified, equivalent to 54%. The measure cleared 80% of people without GAD. At a cutoff of 15, specificity rose to 94%, while sensitivity fell to 29%.
The K6 identified 72% of GAD cases at a cutoff of 9 or higher, but it cleared only 56% of people without GAD. The source reported that about seven in 10 positive K6 results were false alarms in this sample.
Agreement with the structured interview was fair for the GAD-7 and weaker for the K6. Across the cutoffs examined, 82% to 89% of people who screened negative truly did not have GAD. Still, 38 of the 83 patients with GAD scored below 10 on the GAD-7.
These results reinforce that questionnaires are screening tools rather than standalone diagnoses. The researchers described a structured diagnostic interview as more reliable for sorting through overlapping conditions in psychiatric outpatients.
Important limits on the findings
The study was cross-sectional and involved selected clinics and consenting participants, so the 20.4% figure should not be treated as a national prevalence estimate. People with psychotic disorders were excluded, and the participants were already in psychiatric care.
The PRADO study was funded and sponsored by Viatris Pharmaceuticals Japan. Three of the five authors were Viatris employees, while the other two reported fees or advisory roles with Viatris and other drugmakers. IQVIA handled statistical analysis and medical writing. These relationships are relevant when interpreting the findings.
A separate 2024 survey by the same research group estimated GAD prevalence in the general population at 7.6% using the GAD-7 and 6.0% using DSM-5-based questions. Those estimates used a different population and methods and should not be directly compared with the outpatient figure.
AI tools were used to assist with the preparation of this article.
Leave a Reply