The Russian ASRS v1.1 gets cut-offs of its own for adult ADHD
- A team from St Petersburg and Moscow tested the Russian version of the Adult ADHD Self-Report Scale v1.1 on 500 adults: 300 people from the general population without an ADHD diagnosis and 200 patients with confirmed ADHD (F90.0).
- The 18 items formed two factors, inattention and hyperactivity/impulsivity (CFI = 0.930, TLI = 0.921, RMSEA = 0.050); Cronbach's α for the full scale was 0.91, and a retest after eight weeks gave ICC = 0.79 (95% CI 0.72–0.85).
- Part A, the first six items, at a cut-off of 4 points or more had a sensitivity of 84% and a specificity of 92%; Part B at 6 or more had 85% and 81%. The area under the ROC curve was 84% for Part A and 85% for Part B.
- The total score correlated with HADS anxiety at r = 0.611 and with the Barratt Impulsiveness Scale at r = 0.62, and it fell with age (r = -0.282 in the patient group).
ASRS v1.1, developed by Kessler and colleagues together with the WHO, is the most widely used adult ADHD self-report scale. Until this paper, no study had checked how its Russian version behaves on Russian-speaking respondents.
Two groups and a Youden cut-off
The translation followed the WHO protocol: forward and back translation, focus groups, review by at least three mental health specialists. The comparison group, 300 adults surveyed by questionnaire, had no ADHD diagnosis and, by self-report, no psychiatric history; their mean age was 33.4 years. The 200 patients, mean age 27.8, had been diagnosed by a psychiatrist under ICD-10 at the Neuro-psy clinic between 2023 and 2025. Items are rated from 0 to 4.
Each cut-off was chosen by maximising Youden's index. For Part A, 4 points or more gave 84% sensitivity and 92% specificity, with a positive predictive value of 73% and a negative predictive value of 91%. Part B at 6 or more gave 85% and 81%, with a negative predictive value of 82%. Sixty patients repeated the scale two months later.
A rule for the first appointment
For a first appointment, Part A is the working piece: six questions, under a minute to complete, and now a threshold drawn from a Russian-speaking sample. With a negative predictive value of 91%, a score below 4 lets me set ADHD aside for the moment and look at other explanations first. A score of 4 or more is a reason for a full diagnostic interview, not a diagnosis.
I read a positive result with two figures in mind. In this sample two respondents in five had ADHD; at an ordinary intake the share is lower, and the predictive values move with it. And the total score rose with anxiety, r = 0.611, so I ask about anxiety before I interpret the number.
Six questions and a Russian-sample threshold of 4 points make Part A a first filter at intake: a score below the line lets the clinician look elsewhere, a score above it calls for a diagnostic interview.
The comparison group had no clinical assessment; the absence of ADHD rested on the lack of a diagnosis and on self-report, which the authors name as a limitation. The retest covered 60 patients. The predictive values come from a sample in which 200 of 500 people had ADHD.
- adult ADHD
- ASRS
- screening
- psychometrics
- Russian-language instruments