Table 3.

Outcome measures and main results of studies not on folic acid

	Ampakine compound CX516
Reference	Measures of interest: assessment method	Main Results
Berry-Kravis 2006 [14]	Primary outcome, memory domain: Visual Memory and Visual Sequential Memory Subtests of the Test of Visual--Perceptual Skills (TVPS), the Memory for Words Subtest of the Woodcock-Johnson Tests of Cognitive Ability--Revised and the Repeatable Battery for the Assessment of the Neuropsychological Status (RBANS). Secondary outcome measures. Attention/Executive Function Domain: SNAP IV. Language Domain: Peabody Picture Vocabulary Test-III, Forms A and B (PPVT-III) or the Clinical Evaluation of Language Fundamentals-3 (CELF-3). Behavioral Domain: Autism Diagnostic Observation Scale (ADOS), the Gilliam Autism Rating Scale (GARS), the Childhood Autism Rating Scale (CARS), a clinician-rated scale to evaluate modification of autistic features, the ABC-C, and the behaviour component of CGI - Improvement (CGI-I) and VAS. Clinical Cognitive Improvement Measures: VAS (caregiver rated) for cognition and the subject's chosen task (described above) and CGI-I (clinician rated) for cognition and the task. Adverse events: Aberrant Behavior Checklist - Community Edition (ABC-C).	Wilkoson rank-sum test performed. No significant improvement in memory, the primary outcome measure, or in secondary measures of language, attention/executive function, behaviour, and overall functioning in CX516-treated subjects compared to placebo. There were minimal side effects, no significant changes in safety parameters, and no serious adverse events. There was a 12.5% frequency of allergic rash in the CX516 group and 1 subject developed a substantial rash.
	Dextroamphetamine Methylphenidate
Reference	Measures of interest: assessment method	Main Results
Hagerman 1988 [5]	Conners' Abbreviated Parent-Teacher Questionnaire. ADDH: Comprehensive Teacher Rating Scale (ACTeRS). Behavioural observation. Measure of movement: Large scale integrated sensor actometer (LSI). Delay task. Vigilance task.	Paired t tests performed. Compare to placebo clinical response in two thirds of patients, but no statistically significant difference between amphetamine and placebo for any of the ADHD measures, except for the improvement seen on the ACTeRS scale completed by the teacher. Social skills factor and improvements in attention significantly better with methylphenidate (mostly in mildly retarded persons) but not with amphetamine. Safety: significantly more side effects while taking amphetamine, mainly mood lability and irritability.
	L-Acetylcarnitine
Reference	Measures of interest: assessment method	Main Results
Torrioli 1999 [32]	Wechsler Intelligence Scale for Children-Revised (WISC-R); the Bender Gestalt test; and the Conners' Abbreviated Parent-Teacher Questionnaire.	Non-parametric Wilcoson independent-sample test performed. No statistically significant difference between L-Acetylcarnitine and placebo in Wechsler Scale and Bender Gestalt tests and Conners' questionnaire completed by teachers. The Conners' Abbreviated Parent questionnaire showed a significant reduction (P = 0.0065) of hyperactive behaviour at one year in the LAC-treated subjects. Safety: no side effects noted in LAC group.
Torrioli 2008 [13]	Conners' Global Index-Parents (CGI-P) and Conners' Global Index-Teachers (CGI-T). Vineland Adaptive Behavior Scales-Survey Form (VABS) to evaluate adaptive behaviour, four domains: communication, daily living skills, socialization, and motor skills. Wechsler Intelligence Scale for Children-Revised (WISC-R). Side effects.	T tests and repeated measures multivariate analysis using the general linear model were performed. Statistically significant stronger reduction of hyperactivity and improvement of social behaviour in patients treated with LAC, compared with the placebo group, in CGI-P y VABS. Both groups improved their behaviour, showing that psychosocial intervention has a significant therapeutic effect. Safety: no side effects in LAC group.