Methylphenidate dose–response in children with ADHD: evidence from a double-blind, randomized placebo-controlled titration trial

Study design

Children were recruited from mental health clinics in The Netherlands between May 2017 and December 2019. Inclusion criteria were as follows: (a) a clinical diagnosis of ADHD according to DSM-5, (b) 5–13 years of age, (c) IQ > 70, (d) indication for MPH treatment, as determined by the treating physician, and (e) no pharmacological treatment for ADHD 4 weeks prior to study entry. Comorbid diagnoses were not an exclusion criterion. Diagnostic status was confirmed by the first author (K.V.) using the (1) Kiddie–Schedule for Affective Disorders and Schizophrenia for School-Age Children–Present and Lifetime Version (K-SADS), a semi-structured standardized, investigator-based parent interview [25] and (2) teacher rated Disruptive Behavior Disorder rating scale (DBDRS) assessing the presence and severity of symptoms of ADHD [26].

The PCT protocol was based on the titration protocol used in the MTA study [18], modified to improve clinical usability by weekly instead of daily dose changes [27]. All participants received the following treatment conditions: placebo and 5, 10, 15 and 20 mg of MPH twice daily (20 mg only for children > 25 kg [28] in a semi-randomized order. The randomization and blinding procedure is described in more detail in Supplement S1. We used dosing twice daily versus three times daily as this is the dominant Dutch clinical practice. The titration procedure started with a lead-in phase [29], consisting of 4 days in which all doses were administered in ascending order. If a dose was not tolerated it was excluded from the PCT. Duration of the PCT was 3 to 5 weeks, depending on the child’s weight and MPH doses tolerated. During PCT, treatment with a particular dose started on a Saturday and was administered for seven consecutive days, twice daily, at breakfast (around 8 a.m.) and at lunch time (around 12 a.m.).

At baseline and at the end of each week, parents and teachers completed the Strength and Weakness of ADHD symptoms and Normal Behavior Rating Scale (SWAN) [30] rating scale and an adapted version of the MTA Side Effect Rating Scale (see description below) [19]. The local ethics committee approved the study (METC VUMC, # 2016.594 & Netherlands trial register # NL8121).

Outcome variablesADHD symptoms

ADHD symptom severity was measured with the SWAN [30], adapted to measures symptom severity in the past week. This questionnaire contains the following two scales: the Inattention scale and the Hyperactivity/Impulsivity scale, each comprising nine items based on the DSM-IV symptoms of ADHD. Items are scored on a 7-point Likert scale ranging from − 3 to 3, with lower ratings reflecting worse symptoms. Items reflect both ends (strong and weak) of the behavior captured in each symptom. Therefore, it has the potential to reveal clearer and additional dose–response effects due to assessment across the full range of positive and negative manifestations of the behavior underling the symptoms of ADHD [31].

Side effects

Side effects were reported using an adapted version of the MTA Side Effect Rating Scale [19], using a scoring system according to Wigal [32]. Commonly reported side effects were rated on a 4-point scale (0 = not at all, 1 = just a little, 2 = pretty much and 3 = very much) measuring side effect severity in the past week. The total score was used in the analyses.

Predictors of individual MPH dose–response curves

To explore clinically useful predictors, a range of variables that have been associated with dose–response effects [12, 20, 21, 24] and can be routinely identified in standard clinical practice, were assessed at baseline. Candidate predictors included clinical characteristics, demographic variables and attitudes towards ADHD diagnoses and treatment and are described in the Supplement S2 and S3.

Statistical analyses

Analyses were performed using STATA (version 16.0). To examine the effects of MPH dose on ADHD symptoms and side effects, mixed model analyses were conducted given the hierarchical data-structure. To investigate the dose–response curves at group level, two analyses were performed. First, to determine the function best describing the dose–response relationship linear dose–response curves were fitted as function of absolute dose (mg) for each of the following six outcome measures: parent- and teacher-rated inattention symptoms, hyperactivity-impulsivity symptoms and side effects. Next, it was evaluated if a second-, or third order polynomial better described the relationship by adding a quadric and S-shaped component, respectively. Additionally, the analysis were repeated with relative dose (mg/kg). Second, we treated dose as a categorical variable, comparing the effects of 5, 10, 15, and 20 mg (children > 25 kg only) on each of the six outcomes with a mixed model analysis.

As linear dose–response curves provided the best description of the dose–response curves at the group level, linear dose–response curves were also fitted to the individual dose–response curves. The regression coefficient of these individual dose–response curves was used to quantify the dose–response relationship for each outcome measure. Positive β-values indicate that increased doses of MPH were related to steeper linear dose–response curves, with increased doses resulting in larger reductions in ADHD symptoms or a smaller increase in side effects. Non-positive β-values indicate that higher MPH doses were related to more shallow linear dose–response curves, with increased doses resulting in smaller reductions in ADHD symptoms or a smaller increase in side effects.

Finally, the possibility to predict the variation in the individual linear dose–response curves, using variables assessed at study entry was explored. Mixed model analyses, with univariate preselection of significant predictors and backward selection procedure, were used to construct prediction models. Complying with the convention in this type of analysis, the threshold for significance was set at p-values < 0.10 [33]. The amount of variance explained (R.2) was calculated, with 1%, 9% and 25% used as the thresholds for small, medium and large effects, respectively [34].

留言 (0)

沒有登入
gif