Real-World Medication Regimen Change in Adults with ADHD: A Large Claims Database Study
Abstract
Pharmacotherapy persistence in adults with attention-deficit/hyperactivity disorder (ADHD) remains concerning; however, evidence describing early treatment changes and discontinuation is limited. This study aimed to describe initial ADHD pharmacotherapy patterns and quantify claims-based initial regimen instability in Japanese adults. Using a Japanese claims database, we assembled a cohort of adults who initiated ADHD medication with methylphenidate, atomoxetine, guanfacine, or polytherapy (combination therapy) at initiation. Initial regimen instability was defined as the first observed regimen modification, including switching, add-on/polytherapy, de-escalation or combination-pattern change among initial polytherapy users, or a temporary drug-free interruption followed by restart after 31-179 d. A sustained medication-free period of ≥180 d without restart was treated as a competing event. We estimated cumulative incidence functions for initial regimen instability, treating the sustained medication-free period as a competing event. Fine-Gray models adjusted for age, sex, and major psychiatric comorbidities were used to compare initial regimens with methylphenidate as a reference. Landmark analysis was used to assess robustness. Among 14814 initiators, 3085, 8517, 3076, and 136 started methylphenidate, atomoxetine, guanfacine, and initial polytherapy, respectively. In the adjusted Fine-Gray model, the subdistribution hazard ratio for initial regimen instability was 0.74 for atomoxetine and 0.72 for guanfacine, whereas it was higher (2.65) for initial polytherapy in the main analysis. However, this estimate was sensitive to the operational definition of regimen instability. Landmark analysis yielded consistent differences between regimens. Initial ADHD pharmacotherapy showed heterogeneity in Japanese adults and included initial polytherapy, with initial regimen instability differing substantially depending on the starting regimen.
Related
- Association of Admission Medication Regimen Complexity with Delirium in Critically Ill Adults: A Prospective Multicenter Cohort Study
- Rising active surveillance/observation rates and regimen intensification in young men: A retrospective cohort study of national commercial health insurance claims data
- Antihypertensive Medication Class and Incident Alzheimers Disease and Related Dementias: External Replication Across Two Independent Healthcare Data Sources-the Vanderbilt EHR Synthetic Derivative and US Medicare Claims
- Association between maternal dysglycaemia and adverse perinatal outcomes: A retrospective cohort study using a Japanese mother–infant-linked claims database
- No Field for Lactation: Inferring Lactation Status From US Commercial Claims to Describe Postpartum Medication Dispensing
BibTeX
@article{yonezawa2026realworld,
title = {{Real-World Medication Regimen Change in Adults with ADHD: A Large Claims Database Study}},
author = {Natsuri Yonezawa and Kenji Momo},
journal = {Biological and Pharmaceutical Bulletin},
year = {2026},
month = aug,
doi = {10.1248/bpb.b26-00127},
url = {https://www.jstage.jst.go.jp/article/bpb/49/8/49_b26-00127/_article},
note = {CC BY-NC},
}
RIS
TY - JOUR
TI - Real-World Medication Regimen Change in Adults with ADHD: A Large Claims Database Study
AU - Natsuri Yonezawa
AU - Kenji Momo
PY - 2026
DA - 2026/08/18
JO - Biological and Pharmaceutical Bulletin
DO - 10.1248/bpb.b26-00127
UR - https://www.jstage.jst.go.jp/article/bpb/49/8/49_b26-00127/_article
AB - Pharmacotherapy persistence in adults with attention-deficit/hyperactivity disorder (ADHD) remains concerning; however, evidence describing early treatment changes and discontinuation is limited. This study aimed to describe initial ADHD pharmacotherapy patterns and quantify claims-based initial regimen instability in Japanese adults. Using a Japanese claims database, we assembled a cohort of adults who initiated ADHD medication with methylphenidate, atomoxetine, guanfacine, or polytherapy (combination therapy) at initiation. Initial regimen instability was defined as the first observed regimen modification, including switching, add-on/polytherapy, de-escalation or combination-pattern change among initial polytherapy users, or a temporary drug-free interruption followed by restart after 31-179 d. A sustained medication-free period of ≥180 d without restart was treated as a competing event. We estimated cumulative incidence functions for initial regimen instability, treating the sustained medication-free period as a competing event. Fine-Gray models adjusted for age, sex, and major psychiatric comorbidities were used to compare initial regimens with methylphenidate as a reference. Landmark analysis was used to assess robustness. Among 14814 initiators, 3085, 8517, 3076, and 136 started methylphenidate, atomoxetine, guanfacine, and initial polytherapy, respectively. In the adjusted Fine-Gray model, the subdistribution hazard ratio for initial regimen instability was 0.74 for atomoxetine and 0.72 for guanfacine, whereas it was higher (2.65) for initial polytherapy in the main analysis. However, this estimate was sensitive to the operational definition of regimen instability. Landmark analysis yielded consistent differences between regimens. Initial ADHD pharmacotherapy showed heterogeneity in Japanese adults and included initial polytherapy, with initial regimen instability differing substantially depending on the starting regimen.
SN - 1347-5215
SN - 0918-6158
ER -