The RCT with a parallel design assessing the effect of multiple-dose OT administration in children with ASD was performed at the Leuven University Hospital (Belgium). The double-blind phase (phase I) was followed by a 4-week single-blind phase (phase II) during which all participants received intranasal OT. In both phases, OT administration effects were assessed immediately after the 4-week administration period (i.e., post-measurement T1 and T3) and at a follow-up session 4 weeks after cessation of the daily administrations (i.e., follow-up measurement T2 and T4). Please see Fig. 1A, for a visualization of the trial design and Fig. 1B for the CONSORT Flow diagram visualizing the number of participants randomized and analyzed. Please also see Additional file 1 outlining in more detail the trial design and the impact of COVID-19-related health restrictions on the recruitment and flow of participants in the trial.

Trial design (panel A) and CONSORT flow diagram of participants in the trial (panel B). Participants first underwent a double-blind phase (phase I) during which they were allocated to administer either oxytocin or placebo (4 weeks of twice daily intranasal administration). In phase I, nasal spray administration effects were assessed immediately after the last administration of the 4-week administration period (post, T1) and at a follow-up session, four weeks after cessation of the daily administrations (follow-up, T2). Phase I was immediately followed by a single-blind phase (phase II, during which all participants received four weeks of intranasal oxytocin. Also in phase II, nasal spray administration effects were assessed immediately after the four-week administration period (post, T3) and at a follow-up session, four weeks after cessation of the daily administrations (follow-up, T4) (panel A). The CONSORT flow diagram (panel B) visualizes the number of participants throughout the trial, indicating completed assessments at each session, separately for parent informant- and child self-reports

Written informed consent from the parents and assent from the child were obtained prior to the study. Consent forms and study design were approved by the Ethics Committee for Biomedical Research at the University of Leuven, KU Leuven (S61358) in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki). The trial was registered at the European Clinical Trial Registry (EudraCT 2018-000769-35) and the Belgian Federal Agency for Medicines and Health products. As indicated in the EudraCT registration, behavioral data collections were part of a broader assessment, including (neuro)physiological and biological assessments (reports in preparation). The trial was monitored by the Clinical Trial Center at the University Hospital of Leuven, and all trial staff had Good Clinical Practice certification and was trained in the study protocol.

Children with a formal diagnosis of ASD were recruited through the Autism Expertise Centre at the Leuven University Hospital between July 2019 and January 2021. The diagnosis was established by a multidisciplinary neuropediatric team based on the strict criteria of the DSM-5 (Diagnostic and Statistical Manual of Mental Disorders) [1]. Prior to randomization, the Autism Diagnostic Observation Schedule (ADOS-2) [24] and estimates of intelligence (four subtests of the Wechsler Intelligence Scale for Children, Fifth Edition, Dutch version) [25] were acquired (Table 1). The performance intelligence quotient (IQ) was derived from the subtests Block Design and Figure Puzzles. The verbal IQ was derived from the subtests Similarities and Vocabulary.

Inclusion/exclusion criteria Principal inclusion criteria comprised a clinical diagnosis of ASD, age (8–12 years old), intelligence quotient (IQ) above 70, native Dutch speaker, a stable background treatment for at least 4 weeks prior to the screening and no anticipated changes during the trial. Only premenstrual girls were included. Principal criteria for exclusion comprised any neurological (e.g., stroke, epilepsy, concussion) or significant physical disorder (liver, renal, cardiac pathology) or prior use of OT nasal spray (see Additional file 1: Table S1).

Sample size A total of 80 participants (40 in each treatment arm) participated in the trial, allowing to detect a medium effect size (d = 0.60) with α = 0.05 and 80% power, corresponding to effect sizes previously reported in a 4-week oxytocin trial with school-aged children [16].

Medication use, co-occurring conditions and participation in ongoing therapies/trainings The presence of co-occurring psychiatric conditions (with the explicit mentioning of examples in the screening interview including e.g., attention deficit hyperactivity disorder, depression, dyscalculia, dyslexia) and concurrent psychoactive medication use (defined as use within 4 weeks before study enrollment) were screened through parent-report (see Additional file 1: Table S2 for detailed information). Parents were also asked to report participation in ongoing therapies/trainings and whether these were aimed at psychosocial stimulation. Upon free report, parents indicated participation of their child in the following psychosocial trainings/therapies: Theory of Mind training, emotion recognition training, social skills training, cognitive behavioral therapy, psychotherapy, self-esteem training, mood regulation, music therapy, hippotherapy and an autism coach. To perform moderator analyses assessing the possible impact of receiving concomitant psychosocial training, the group of children was subdivided in those receiving a higher intensity of psychosocial training (3 or more sessions per month) versus those receiving no or low intensity psychosocial training (less than 3 training sessions per month) (see Table 1). Note that adopting a more lenient threshold for defining the subgroups (i.e., 1 or more session(s) per month) yielded a qualitatively similar pattern of moderator analysis results (data not shown).

Study medication Participants were randomized to receive OT (Syntocinon®, Sigma-tau) or placebo nasal sprays, administered in identical blinded amber 10-ml glass bottles with metered pump. The placebo spray consisted of all the ingredients used in the active solution except the OT compound. Nasal spray preparation, packaging, blinding and randomization (permuted-block randomization, RITA software [26]) were performed by the pharmacy of Heidelberg University Hospital (Germany). Participants were randomly assigned in a 1:1 ratio, and balanced according to sex (male/female), IQ and age. During the initial double-blind phase (phase I), all research staff conducting the trial, participants and their parents were blinded to nasal spray allocation. During the subsequent single-blind phase (phase II), trial staff were aware that all participants received intranasal OT, but participants and parents were still fully blinded regarding nasal spray allocation. Particularly, children and their parents participating in the trial were informed that during at least one of the two treatment phases, they would administer the active OT nasal spray. Only after the last visit of the last participant, trial staff were unblinded regarding treatment allocation in phase I.

Dosing Children (assisted by their parents) were asked to self-administer a daily dose of 2 × 12 IU nasal spray or placebo equivalent (3 puffs of 2 IU in each nostril), 12 IU in the morning and 12 IU in the afternoon (similar to the conservative dosing scheme adopted in young children with ASD [15]). The nasal spray was administered during 28 consecutive days during the initial double-blind phase (phase I) and for another 28 days during the single-blind phase (phase II). The duration of 4 weeks was similar to prior trials in children [16] and adults [14] with ASD. Participants received clear instructions about the use of the nasal sprays through a demonstration together with the trial staff [27].

Compliance monitoring Compliance was assured using a daily medication diary that recorded date and time of administration (phase I percentage compliance; OT: 96.75 ± 5.26%; placebo: 96.11 ± 5.29%; t(74) = 0.52, p = 0.603; phase II percentage compliance; OT-first: 94.55 ± 11.69%; placebo-first: 92.98 ± 13.92%; t(74) = 0.53, p = 0.597). The total amount of administered fluid was also monitored (phase I: OT: 14.86 ± 2.37 ml; Placebo: 13.79 ± 2.35 ml; t(75) = 2.00, p = 0.050; phase II: OT-first: 13.72 ± 3.47 ml; placebo-first: 12.83 ± 3.52 ml; t(74) = 1.10, p = 0.275).

Side effects During the nasal spray administration period, participants were screened for potential adverse events (weekly parent report) or changes in affect and arousal (daily diary by child and parent). Overall, reports of side effects were minimal and not treatment-specific (see Additional file 1: Tables S3 and S4).

Parent-reported beliefs about allocated nasal spray At the end of each trial phase (I and II), parents reported beliefs about nasal spray allocation (see Results). In the double-blind phase (phase I), the proportion of parents that believed their child had received the OT nasal spray was similar in both treatment arms: 39.5% in the OT group and 35.9% in the placebo group (p = 0.75). In the OT group, 18.4% of parents indicated to ‘have no explicit belief’ about nasal spray allocation versus 10.3% in the placebo group. In the single blind phase (phase II), during which all participants received the actual OT nasal spray, the proportions of parents that believed their child had received the OT nasal spray were similar as well (p = 0.30): 57.9% in the oxytocin-first group, 46.2% in the placebo-first group. Furthermore, the proportions of parents that believed their child had received the OT nasal spray did not differ between treatment phases (oxytocin-first: p = 0.11; placebo-first: p = 0.36).

The primary outcome measure was change from baseline in parent-rated social responsiveness on the Social Responsiveness Scale-Children, second edition (SRS-2 total raw scores) [28, 29], which comprises five subscales examining social cognition, social communication, social awareness, social motivation, and rigidity/repetitiveness, using a four-point Likert-scale (65 items). Lower scores indicate higher social responsiveness.

Secondary outcome measures included changes from baseline in parent-rated repetitive behaviors (Repetitive Behavior Scale-Revised; RBS-R) [30], self- and parent-rated presence of anxiety symptoms (Screen for Child Anxiety Related Emotional Disorders; SCARED-NL) [31], and changes from baseline in constructs of self-rated attachment toward their mother (Attachment Questionnaire child-report) [32] and peers (Attachment Style Classification Questionnaire child-report) [33] (see Table 2 and Additional file 1: Table S5).

All outcomes were assessed five times: (i) at baseline (T0), (ii) immediately after the 4-week double-blind nasal spray administration period (phase I—post, T1); (iii) at a follow-up session, 4 weeks after cessation of the double-blind nasal spray administration period (phase I—follow-up, T2); (iv) immediately after the 4-week single-blind nasal spray administration period (phase II—post, T3); and (v) at a follow-up session 4 weeks after cessation of the single-blind nasal spray administration period (phase II—follow-up, T4). Post-sessions were scheduled approximately 24 h after the last administration, follow-up sessions within 28 ± 7 days.

Analyses were performed using a modified intention-to-treat approach that included all randomized participants who completed the baseline session and at least one post or follow-up session (Fig. 1B, CONSORT diagram). All statistics were executed with Statistica 14 (Tibco Software Inc.).

First, possible baseline differences on the questionnaires were assessed between randomized nasal spray groups, indicating no statistically significant differences (Table 1). Next, between-group differences in treatment responses of phase I (double-blind) on the primary and secondary outcome measures were assessed, by subjecting change from baseline scores of the post (T1) and follow-up (T2) sessions to independent sample t-tests. Cohen’s d effect sizes (change from baselineOT—change from baselinePLACEBO)/pooled SD) are also reported, where 0.2 is indicative of a small effect, 0.5 a medium effect and 0.8 a large effect. Additionally, single-sample t-tests were adopted to assess within-group changes (compared to baseline) in the OT and placebo group separately (Table 2 and Additional file 1: Table S6). Similar independent and single-sample t-tests were adopted to assess treatment responses of phase II (single-blind), although note that here changes in outcome measures were calculated relative to assessment session T2 (last session of phase I), i.e., allowing to examine treatment-induced changes, over and above changes induced in phase I (Additional file 1: Table S6 and S7).

Further, to assess whether the overall magnitude of treatment-induced changes at the last session of the trial (T4) (calculated as change from baseline T0, i.e., reflecting the total change over phases I and II) were reliable for individual participants (more than can be expected by measurement error), the Reliable Change Index (RCI) [34] was calculated, based on the test–retest reliability of the adopted Dutch parent-reported SRS scale (Cronbach’s alpha = 0.94) and corresponding standard error of measurement (SEM = baseline SD × SQRT(1—Cronbach’s alpha) = 5.19) using the formula: RCI = 1.96 × SQRT (2 × SEM × SEM) = 14.8. Change scores higher than the RCI-value (14.8) were considered reliable.

Finally, exploratory analyses were performed to investigate the potential influence of moderator variables on phase I treatment outcome. To do so, change from baseline scores were subjected to a mixed-effect model with ‘subject’ as random factor and ‘nasal spray’ (OT, placebo), ‘assessment session’ (T1 post, T2 follow-up) and the moderator variable included as fixed factors. Separate models were constructed to assess the modulating effect of concomitant psychosocial training (3 or more sessions per month, less than 3 sessions per month); medication use (present, not present; as listed in Table 1); biological sex (male, female); and parent-reported beliefs (OT, placebo).