Establishing clinically important differences in adults with attention deficit/hyperactivity disorder

ABSTRACT

Background

The adult attention deficit hyperactivity disorder (ADHD) Self‐Report Scale (ASRS) is one of the most widely used patient‐reported outcome measures (PROMs) for evaluating ADHD symptom severity and treatment‐related changes in clinical trials of adults with ADHD; however, clinically important differences (CIDs) for the ASRS remain poorly defined.

Aims

This study aimed to review how CIDs for the ASRS have been reported in randomised controlled trials (RCTs) of adult ADHD and to establish ASRS‐specific CID values for this population.

Methods

We first conducted a cross‐sectional literature search to examine the reporting of CID values in RCTs of adult ADHD interventions. We then conducted a secondary analysis of an RCT involving 56 adults with ADHD who received either active or sham transcranial alternating current stimulation. The ASRS total score served as the PROM. An anchor‐based approach using the Clinical Global Impression‐Improvement scale was applied. General linear mixed‐effects modelling estimated mean differences across patient‐defined improvement levels to identify categorical CID thresholds.

Results

None of the 11 identified studies reported CID values for the ASRS as a PROM in adults with ADHD. The mixed‐effects model indicated that a reduction of approximately 13.39 points in ASRS total score represents a clinically meaningful improvement (CID_{within‐group}), and a reduction of 19.51 points represents substantial improvement, corresponding to reductions of 29.2% and 42.6% from the baseline, respectively. Additionally, a between‐group difference of 6.13 points in ASRS total score represents a minimal clinically meaningful difference (CID_{between‐group}), and a 12.25‐point difference represents a moderately meaningful between‐group difference.

Conclusions

This study provides adult‐ADHD‐specific CID estimates that offer practical benchmarks for interpreting treatment response and informing power calculations in future clinical trials.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• The adult attention deficit hyperactivity disorder (ADHD) Self‐Report Scale (ASRS) is widely used in adult ADHD trials, but clinically important differences (CIDs) for interpreting score changes have not been well established, limiting its usefulness in determining whether treatment effects are clinically meaningful.

WHAT THIS STUDY ADDS

• This study found that recent randomised controlled trials in adults with ADHD did not report CID values for the ASRS and provided adult‐ADHD‐specific CID estimates using an anchor‐based approach. It identified quantitative thresholds for clinically meaningful within‐group improvement and between‐group differences in the ASRS total score.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• These findings provide practical benchmarks for interpreting treatment responses and informing power calculations and may support more rigorous endpoint selection and trial design in adult ADHD research.

Abbreviations

ADHD

attention deficit hyperactivity disorder

ASRS

Adult ADHD Self‐Report Scale

CGI‐I

clinical global impression‐improvement

CID

clinically important difference

PROM

patient‐reported outcome measures

RCT

randomised controlled trial

tACS

transcranial alternating current stimulation

INTRODUCTION

Attention deficit hyperactivity disorder (ADHD) is one of the most prevalent childhood‐onset neurodevelopmental disorders and may persist into adulthood. The global prevalence among adults is approximately 2.5%, and affected individuals often experience functional impairments. ¹ Well‐powered clinical trials have significantly advanced treatment options, and meta‐analyses provide robust evidence for the efficacy of stimulants in reducing core ADHD symptoms in adults. ²

Given the lack of reliable objective measurements for ADHD symptoms, patient‐reported outcome measures (PROMs) play an important role in evaluating ADHD symptom severity and treatment‐related changes. Most clinical trials in adult ADHD have evaluated the efficacy of interventions based on the statistical significance of rating scales, which were determined by a mathematical calculation and probability based on null hypothesis significance testing. However, simply measuring changes in scores may not fully reflect the meaningfulness of the treatment effect. ³ Statistical significance does not always correspond to meaningful clinical effects and may ‘deflect attention from the actual size of an effect’. ⁴ Previous ADHD clinical trials defined response to intervention as the reduction rate in the ADHD Rating Scale but lacked a uniform, evidence‐based threshold. Some studies set the threshold at a 30% change rate, ⁵ whereas others used 25% as the threshold for improvement and 40% for robust improvement. ⁶ Therefore, quantifying clinically meaningful symptom change is important for individuals with ADHD. The minimum clinically important difference defines the minimum change considered clinically meaningful. ⁷ The term clinically important difference (CID) refers to an evidence‐based threshold for determining whether an observed change or difference is clinically meaningful, extending interpretation and decision‐making beyond statistical significance. ¹⁸ Changes exceeding this threshold are considered clinically relevant. The most commonly used method for determining the CID is the anchor‐based method, ⁸ also called the ‘external reference’ method. Specifically, an anchor is selected as an external indicator, and the relationship between scores on the target instrument and this independent measure is examined. ⁹

For adults with ADHD, the Adult ADHD Self‐Report Scale (ASRS), a Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition‐based self‐report rating scale, ¹⁰ was developed for screening and evaluating treatment efficacy. It demonstrates good validity and reliability in both its English and Chinese versions. ¹¹ This scale comprises 18 items scored from 0 to 4 across 2 dimensions—inattention and hyperactivity/impulsivity—with 9 items in each dimension. To date, the application of the CID for the ASRS in the context of adult ADHD remains unclear.

In this study, we aimed to review the current reporting of CIDs for the ASRS in randomised controlled trials (RCTs) through a cross‐sectional survey and to establish CID values for the ASRS in adults with ADHD. To this end, we reanalysed data from one of our previous RCTs, which used transcranial alternating current stimulation (tACS) in adults with ADHD. ¹²

METHODS

Cross‐sectional survey

The literature search was conducted on 1 April 2025. We included RCTs in adults with ADHD, with no restriction on intervention type, provided that the ASRS questionnaire was used as an outcome measure. Eleven articles were identified for review and analysis. Each article was reviewed for any mention or definition of CID. Data were extracted using Microsoft Excel 365 (details in Supporting Information S1: Supplementary Methods).

CID estimation

Study design and participants

The RCT was a parallel‐group, double‐blind, sham‐controlled clinical trial investigating the efficacy and safety of tACS versus sham stimulation in adults with ADHD, conducted from January 2023 to August 2024 in Peking University Sixth Hospital. A total of 56 participants were included in the study. After inclusion and baseline assessment, participants were randomly assigned to either the tACS group, receiving stimulation at 77.5 Hz and 15 mA ¹³ , ¹⁴ or the sham group, for a total of 20 sessions over 4 consecutive weeks. Each stimulation session lasted 40 min, with a 180‐s ramp‐up and a 12‐s ramp‐down. Self‐rated ASRS scores were assessed at baseline, Week 2 and Week 4; the Clinical Global Impression‐Improvement (CGI‐I) was rated by clinicians at Week 2 and Week 4 (details are provided in Supporting Information S1: Supplementary Methods 2.2.1). Reporting followed the Consolidated Standards of Reporting Trials 2025 expanded checklist. ¹⁵

PROMs and anchor

We used an anchor‐based approach to estimate CIDs for both within‐group and between‐group comparisons. The ASRS was used as the response PROM, and the CGI‐I served as the anchor to define responder status for estimating both CIDs (details are provided in Supporting Information S1: Supplementary Methods 2.2.2).

Statistical analysis

Descriptive statistics, including t‐tests and chi‐square tests, were used to compare demographic characteristics between groups. The three categories representing worsening on the CGI‐I were combined for analyses due to small sample size (n = 1). To assess the suitability of the CGI‐I as an anchor, we first used Spearman's rank correlation to evaluate the association between CGI‐I ratings and the absolute change in ASRS scores from baseline to the end of treatment (Week 4).

For the longitudinal data (baseline, middle of treatment [Week 2] and end of treatment [Week 4]), we constructed a general linear mixed model (GLMM) for repeated measures with a random intercept for each individual. The model estimated the mean change score in each CGI‐I category (minimally improved or much improved vs. no change), which was taken as the CID. The model included the score difference between two timepoints for individuals meeting a given CGI‐I level, the fixed‐effect parameter associated with the anchor level, the random intercept for the individual and the residual. A p value < 0.05 was used as the threshold for statistical significance in comparing the two likelihood values. If statistical significance was reached, this indicated that the baseline dependency of the CIDs should be considered. The CID was estimated by the linear combination of coefficients, with parametric 95% confidence intervals (CIs). All analyses were performed in SAS 9.4 (SAS Institute Inc). In this study, CID_{within‐group} refers to a clinically meaningful within‐arm pre–post change in the ASRS total score, which is not specific to any treatment modality. By contrast, CID_{between‐group} refers to the clinically meaningful between‐arm difference in the ASRS total score change and was used to determine whether one intervention provides a clinically meaningful advantage over the other in an RCT.

RESULTS

Cross‐sectional survey

Eleven publications were included (figure S1A; table S1). Of these publications, none mentioned CID. Only one article defined the response threshold at a 30% decrease in the ASRS total score from baseline.

Participant characteristics

Fifty‐six participants were included in this study, with 28 in each group (tACS and sham) (figure 1A; table 1). The baseline demographics and clinical characteristics of participants were well matched between the two groups (table 1). At baseline, the mean ASRS total score was 45.77 (8.98) (table 2). The correlation between the CGI‐I (anchor) and ASRS (PROM) change score at Week 4 was 0.49 (p < 0.001), indicating sufficient association between the anchor and the PROM. At the end of treatment, 1 (2%) participant reported ‘minimally worse’, 13 (23%) participants reported ‘no change’, 23 (41%) participants reported ‘minimally improved’ and 19 (34%) participants reported ‘much improved’ (table 2).

FIGURE 1.

(A) The Consolidated Standards of Reporting Trials diagram of RCT, (B) CID_{within‐group} and (C) CID_{between‐group} for the ASRS total score at the end of treatment. ADHD, attention deficit hyperactivity disorder; ASRS, Adult ADHD Self‐Report Scale; CID, clinically important difference; GAD, generalised anxiety disorder; MDD, major depressive disorder; RCT, randomised controlled trials; tACS, transcranial alternating current stimulation.

TABLE 1.

Baseline characteristics of participants

Characteristics	tACS group (n = 28)	Sham group (n = 28)	Statistics	p value
Age (years), mean (SD)	26.07 (4.41)	27.39 (5.49)	t = −0.99	0.325
Gender, n (%)
Male	9 (32.1)	8 (28.6)	χ 2 = 0.08	0.771
Female	19 (67.9)	20 (71.4)
Education (years), mean (SD)	16.46 (2.63)	16.39 (1.40)	t = 0.13	0.900
Marital status, n (%)
Unmarried	25 (89.3)	26 (92.9)	χ 2 = 0.22	0.639
Married	3 (10.7)	2 (7.1)
ADHD diagnosis, n (%)
ADHD‐I	22 (78.6)	20 (71.4)	χ 2 = 0.38	0.537
ADHD‐C	6 (21.4)	8 (28.6)
Baseline ASRS score, mean (SD)
Inattention	28.79 (3.83)	27.21 (3.36)	t = 1.63	0.109
Hyperactivity/impulsivity	17.54 (7.56)	18.00 (6.71)	t = −0.24	0.809
Total	46.32 (9.94)	45.21 (8.05)	t = 0.46	0.649

Note: All statistics were χ ² values for categorical variables and t values for continuous variables.

Abbreviations: ADHD, attention deficit hyperactivity disorder; ADHD‐C, ADHD‐combined presentation; ADHD‐I, ADHD‐predominantly inattentive presentation; ASRS, Adult ADHD Self‐Report Scale; SD, standard deviation; tACS, transcranial alternating current stimulation.

TABLE 2.

Anchor‐based CID estimation of ASRS total score

CGI‐I	n (%)	ASRS total score (0–72)		CIDwithin‐group		CIDbetween‐group
		Baseline	Week 4	Threshold	95% CI	Threshold	95% CI
Overall	56 (100)	45.77 (8.98)	31.52 (10.30)	–	–	–	–
No change	13 (23)	43.84 (7.27)	37.07 (10.32)	−7.26	−9.81 to −4.71	–	–
Minimally improved	23 (41)	46.73 (7.39)	31.13 (7.39)	−13.39	−15.33 to −11.45	6.13	4.23–8.02
Much improved	19 (34)	45.78 (11.83)	27.21 (10.80)	−19.51	−22.35 to −16.68	12.25	8.47–16.03

Note: One (2%) participant was rated as minimally worse at the end of treatment.

Abbreviations: ADHD, attention deficit hyperactivity disorder; ASRS, Adult ADHD Self‐Report Scale; CGI‐I, Clinical Global Impression–Improvement; CI, confidence interval; CID, clinically important difference.

Compared with the sham group, the tACS group showed significantly greater reductions in the ASRS total score from baseline to Week 2 (mean difference, −5.4 [95% CI −9.7 to −1.0], T _time×group = −2.4, p _time×group = 0.019; Cohen's d: 0.60) and from baseline to Week 4 (mean difference, −4.9 [95% CI −9.3 to −0.5]; T _time×group = −2.2, p _time×group = 0.031; Cohen's d: 0.54) (table S2).

Estimation of CID_{within‐group} and CID_{between‐group} for ASRS total score

The GLMMs estimated that the CID_{within‐group} mean change was −13.39 points (95% CI −15.33 to −11.45) for patients reporting minimal improvement and −19.51 points (95% CI −22.35 to −16.68) for patients reporting much improvement (figure 1B, C; table 2). Relative to the baseline ASRS total score, the percentage change from baseline to Week 4 was 29.2% for minimal improvement and 42.6% for much improvement. These changes were statistically significant compared with the no‐change group (p < 0.001 for all). For CID_{between‐group}, GLMM results showed that a 6.13 point (95% CI 4.23 to 8.02) difference in the ASRS total score represented the threshold for minimal improvement in the active group relative to the control group, and a 12.25‐point (95% CI 8.47 to 16.03) represented the threshold for much improvement.

DISCUSSION

Main findings

In this study, we conducted a cross‐sectional survey of CID reporting for the ASRS in adults diagnosed with ADHD and quantified meaningful improvement by estimating CIDs from a randomised sham‐controlled clinical trial. Our findings indicate CID_{within‐group} values of −13.39 points (minimal) and −19.51 points (substantial), corresponding to reductions of 29.2% and 42.6% from baseline, respectively. The CID_{between‐group} values were 6.13 points (minimal) and 12.25 points (substantial). These CID values represent evidence‐based endpoints for future clinical trials.

As shown in our analysis, no publication reported or defined a CID using the ASRS as a PROM in adult ADHD RCTs. Although many studies report statistically significant results, the absence of CID thresholds limits the clinical interpretability of these findings. Instead of reporting a CID, one study defined the response threshold as a change rate of more than 30% in the ASRS from baseline, ¹⁶ but without an evidence‐based rationale. Moreover, this threshold was originally applied to the ADHD Rating Scale, which has a different scoring method from the ASRS. Some studies have also reported the effect size (Cohen's d) as an indicator of clinical significance. However, it is important to note that responder analysis and effect size cannot replace CID. ¹⁷

Our study calculated both CID_{within‐group} and CID_{between‐group} for the ASRS total score. These findings provide a pragmatic benchmark for treatment‐response evaluation in clinical research. Specifically, a reduction of 13.39 points in the ASRS total score (regardless of the intervention) may be interpreted as indicating a clinically meaningful improvement in pre–post comparisons (minimal CID_{within‐group} threshold). A reduction of 19.51 points may be interpreted as indicating a clinically substantial improvement (moderate CID_{within‐group} threshold). ¹⁸ Notably, the term substantial here is distinct from statistical significance; it denotes a greater magnitude of clinical benefit beyond the threshold for clinical meaningfulness. When comparing the efficacy of two interventions in an RCT, a between‐group difference in the ASRS total‐score reduction of 6.13 points may be considered clinically meaningful, favouring the group with the larger reduction (minimal CID_{between‐group} threshold). A difference of 12.25 points may be considered clinically substantial, indicating a pronounced clinically meaningful difference between the two interventions (moderate CID_{between‐group} threshold).

An intriguing finding was that when the CID_{within‐group} was expressed as a percentage change in ASRS total score from baseline, minimal improvement corresponded to 29.2% and substantial improvement to 42.6%. These two change rates closely align with the predefined threshold for responses (change rate of 30%) ¹⁶ and for robust responses (change rate of 40%). ⁶ Thus, our study provides evidence‐based response thresholds that are consistent with previously established thresholds. Furthermore, determining between‐group differences is particularly critical for disorders or symptoms with substantial placebo effects. Our analysis revealed a 7.26‐point reduction in the ASRS total score at the end of treatment even when the CGI‐I indicated no change in the overall population and an 11‐point reduction in the sham group. These findings are consistent with meta‐analytic evidence demonstrating moderate pre–post placebo effect sizes in ADHD clinical trials. ¹⁹ Consequently, the between‐group CID values established in this study serve as a crucial benchmark for designing and interpreting future RCTs in adult ADHD populations. Notably, the CID values should not be rigidly interpreted as definitive thresholds analogous to statistical significance criteria (e.g., p = 0.05) but rather as approximate reference ranges in the absence of more specific information. They may also be integral to guideline development. Furthermore, CID estimates may be influenced by patient demographics and the type of assessor (patient‐reported or clinician‐reported), among other factors. ²⁰ Accordingly, future studies are needed to establish more generalisable CIDs for the ASRS to facilitate accurate assessment of treatment efficacy and improve patient‐centred outcomes in clinical settings.

Limitations

This study has several limitations. First, owing to the specific PROM selected and the inclusion criteria, only 11 publications were included in the cross‐sectional survey, which may introduce selection bias. Second, the limited sample size of the RCT may have influenced CID estimates; notably, no participants reported ‘very much improved’ at the end of treatment. These factors may limit the generalisability of the CID values reported here. Future research should employ additional PROMs and larger sample sizes to replicate our findings and establish more generalisable CIDs.

Implications

This study bridges a vital translational gap by transforming abstract PROM scores into interpretable, real‐world indicators of treatment success. The proposed CIDs enable future trials to define clinically meaningful endpoints a priori, enhance comparability across studies and align regulatory and patient‐centred outcomes in the field of adult ADHD. For adults with ADHD, a within‐group reduction of approximately 13.39 points, a change rate of 29.2% from baseline to end of treatment, and a between‐group difference of 6.13 points in ASRS total score are clinically meaningful endpoints for future intervention trials.

CONCLUSIONS

This study provides the first adult ADHD‐specific CID estimates for the ASRS, addressing a vital translational gap by transforming abstract PROM scores into interpretable, real‐world indicators of treatment success. These CID values offer practical benchmarks for interpreting treatment response and informing power calculations and represent evidence‐based endpoints for future clinical trials.

LEARNING POINTS

• CIDs play an important role in interpreting the clinical significance of RCT results for adult ADHD and provide a benchmark for future RCT design in this field.

• For adults with ADHD, a within‐group reduction of approximately 13.39 points (CID_{within‐group}), a change rate of 29.2% from baseline to end of treatment and a between‐group difference of 6.13 points (CID_{between‐group}) in the ASRS total score represent clinically meaningful endpoints for future intervention trials.

AUTHOR CONTRIBUTIONS

Zhao Fu: Investigation; formal analysis; methodology; writing—original draft. Jiahui Deng: Investigation; methodology. Simin Kang: Investigation; methodology. Junbin Tian: Investigation; methodology. Qingjiu Cao: Investigation; methodology. Yufeng Wang: Investigation; methodology. Zongshi Qin: Conceptualization; funding acquisition; supervision; writing—review and editing; guarantors. Li Yang: Conceptualization; funding acquisition; supervision; writing—review and editing; guarantors.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

This study was performed in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Peking University Sixth Hospital. The study was registered on the Clinical Trial Registry platform (ChiCTR Identifier: ChiCTR2400081121). All the participants provided written informed consent.

Supporting information

Supporting Information S1

Figure S1

Biography

Zhao Fu is a PhD candidate at Peking University Sixth Hospital and Peking University Institute of Mental Health in China, where he received his Master of Medicine degree in 2022. His main research interests include the neuroimaging mechanisms of attention‐deficit/hyperactivity disorder (ADHD) and the clinical and mechanistic investigation of noninvasive brain stimulation for ADHD.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author on reasonable request and with permission of the university administration.