An N-of-1 Trial As an Individualized Withdrawal Treatment Approach to Psychological Methylphenidate Dependence

symptoms during everyday life on an 11-point Likert scale [from ‘very dissatisfied’ (0) to ‘very satisfied’ (10)], i.e. aggressiveness, impulsiveness, concentration, hypersomnia, sensation of hunger, and agitation. Furthermore, the numbers of eating attacks and losses of control were documented, and physical activity was assessed by a pedometer counting daily steps. As part of the self-assessment, the patient was asked every evening which medication he believed to have received this day. Additionally, on 4 days a week, the patient performed a cancellation test (d2-R), which is validated for the assessment of attention and concentration and suited to quantify attention deficits in ADHD patients [4] . The test was performed under supervision 2 h after drug intake, when maximum improvement in performance was expected [3] . Because of possible learning effects upon repetitive testing, the time to perform the d2-R test was reduced according to the test manual [4] . The primary statistical evaluation was a time series analysis, as described earlier [5] , using SAS version 9.3 (SAS Institute Inc., Cary, N.C., USA). The level of significance was set at p ≤ 0.05. This N-of-1 randomized controlled trial found no evidence for an ability of methylphenidate to improve ADHD symptoms in this patient on long-term treatment. However, the subjectively assumed intake of methylphenidate was correlated with a better general condition (p = 0.005) and higher subjective power of concentration (p = 0.007). Aggressiveness (p = 0.029), impulsiveness (p = 0.020), the sensation of hunger (p = 0.041), and the number of eating attacks (p = 0.012) were all negatively correlated with the assumed but not with the actual methylphenidate intake. Indeed, there is evidence that the belief that one receives active treatment sometimes predicts subjective responses better than the actual treatment [6] . Interestingly, the subjectively perceived improved concentration was not substantiated in better d2-R performance, which is in agreement with findings in healthy prescription stimulant-naive college students, who reported enhanced subjective arousal when assuming methylphenidate intake, while cognitive performance was in fact unchanged [7] . Although biased by numerous factors, the post hoc assessment of subjective well-being was a likely critical element of this trial because it helped to reveal that methylphenidate and expected beneficial effects did not coincide. Interestingly, the patient assumed an intake of methylphenidate twice as often as of placebo ( fig. 2 ), and particularly at the beginning of the trial, he was convinced that he recognized methylphenidate effects immediately, and tended to explain his behavior accordingly (‘I had an eating attack, since I had received placebo’). Thereafter, the patient became more and more confused regarding drug exposure, which shifted his focus gradually from a medication-centered view towards a state of imWe cared for a 27-year-old patient with an 18-year history of attention deficit hyperactivity disorder (ADHD) on long-term treatment with methylphenidate whose dose had been gradually increased from 40 to 140 mg/day during the previous 9 years, with an additional 10-mg ‘on-demand’ rescue dose for eating attacks or aggressive feelings ( fig. 1 ). The comprehensive psychological assessment confirmed a methylphenidate overuse in combination with an extensive psychological dependence complicated by a family environment that powerfully reinforced his abuse. Considering that nonresponse manifests in 20–30% of patients on methylphenidate [1] , the appropriateness of methylphenidate therapy was doubted and a dose reduction and drug discontinuation recommended, which was rejected by the patient because he was afraid of symptom exacerbation. Consequently, an N-of-1 randomized controlled trial was suggested to test the ability of methylphenidate to improve the ADHD symptoms in this individual [2] . In contrast to several earlier outpatient and inpatient withdrawal attempts, which had all failed, the patient welcomed this individualized treatment approach as a cooperative venture with the physician and was motivated to participate. After consultation of the local ethics committee and obtaining informed consent from the patient, the methylphenidate dose was gradually reduced to the recommended maximum dose of 80 mg/day. Subsequently, methylphenidate and identically looking placebo were administered in a randomized double-blind multiple-crossover fashion for 32 days, each day using either the drug or placebo ( fig. 1 ). Taking into account the patient’s fluctuating symptom severity, randomization was designed in blocks, each comprising 1 day of methylphenidate and 1 day of placebo treatment. The short half-life of methylphenidate, the rapid onset of action, and the disappearance of the effect within 1 day enabled a daily switch between placebo and the active drug [3] . Every evening, the patient assessed his self-perceived health status by rating his general condition and the most troubling Received: February 24, 2014 Accepted after revision: June 4, 2014 Published online: October 16, 2014