Cognitive effects of methylphenidate in healthy volunteers: a review of single dose studies

Abstract

Methylphenidate (MPH), a stimulant drug with dopamine and noradrenaline reuptake inhibition properties, is mainly prescribed in attention deficit hyperactivity disorder, is increasingly used by the general population, intending to enhance their cognitive function. In this literature review, we aim to answer whether this is effective. We present a novel way to determine the extent to which MPH enhances cognitive performance in a certain domain. Namely, we quantify this by a percentage that reflects the number of studies showing performance enhancing effects of MPH. To evaluate whether the dose–response relationship follows an inverted-U-shaped curve, MPH effects on cognition are also quantified for low, medium and high doses, respectively. The studies reviewed here show that single doses of MPH improve cognitive performance in the healthy population in the domains of working memory (65% of included studies) and speed of processing (48%), and to a lesser extent may also improve verbal learning and memory (31%), attention and vigilance (29%) and reasoning and problem solving (18%), but does not have an effect on visual learning and memory. MPH effects are dose-dependent and the dose–response relationship differs between cognitive domains. MPH use is associated with side effects and other adverse consequences, such as potential abuse. Future studies should focus on MPH specifically to adequately asses its benefits in relation to the risks specific to this drug.

Cognition, Dopamine, Memory, Methylphenidate

Introduction

Methylphenidate (MPH; see Table 1 for a complete list of abbreviations) is a stimulant drug that is mainly prescribed in attention deficit hyperactivity disorder (ADHD, (Leonard et al., 2004)). In the past decades there has been a vast increase in the use of MPH (Diller, 1996). Not only by patients to whom MPH is prescribed, but also by the general population, who believe that MPH enhances their cognitive functions (Maher, 2008). This has raised ethical concern (Sahakian and Morein-Zamir, 2007; Larriviere et al., 2009; Stix, 2009). Discussions in popular scientific literature consider safety, potential abuse and side effects (Swanson and Volkow, 2008; Stix, 2009) of cognition-enhancing drugs. However, even if these issues were no reason for concern, a critical question would still be: ‘How effective are supposed cognition-enhancing drugs actually?’. Although there are some indications that MPH may improve certain aspects of cognitive function (e.g. working memory (Mehta et al., 2000b)), there are also reports of impaired performance on cognitive tasks after MPH (e.g. planning (Elliott et al., 1997)). While a reasonable amount of studies have examined cognitive effects of MPH in healthy volunteers, the evidence is undeniably mixed.

Cognitive effects of MPH in children with ADHD have been reviewed by Pietrzak et al. (2006). Their extensive review includes evidence on a broad range of cognitive functions and presents it classified into several cognitive domains. An equivalent review on cognitive effects of MPH in healthy volunteers is not available yet. Previous reviews on the cognition-enhancing effects of drugs have focused on attention, learning, memory and executive function (Repantis et al., 2010; Smith and Farah, 2011). In the current review the full spectrum of cognitive function is considered. The differences between studies on cognitive effects of MPH – such as differences in methodology, MPH dose, neuropsychological instruments employed – do not encourage conducting a meta-analysis (Pietrzak et al., 2006). As we did not want to limit our review to one particular measure or cognitive domain, we used a similar approach as Pietrzak et al. (2006). Because the review by Pietrzak et al. (2006) focused on studies involving children with ADHD, the reported tasks differ slightly from those in studies with healthy volunteers, since both clinical status (i.e. ADHD diagnosis) and age (children vs. adults) may call for a different selection of tasks. Inspection of the range of tasks used in studies with MPH in healthy volunteers, and the fact that the classification applied by Pietrzak et al. (2006) is not based on factor analysis or any other established way of classifying cognitive tasks (or this is not reported), we looked for a classification that was more appropriate for the available data (Pietrzak et al., 2006).

The categorization described by Nuechterlein et al. (2004) is based on factor analytic studies and fine-tuned by a committee of field experts. Although the identification considered cognitive function in schizophrenia patients, it is generally accepted that cognitive domains are broadly the same in healthy controls (Dickinson et al., 2006; Genderson et al., 2007). The cognitive domains (i.e. speed of processing, attention/vigilance, working memory, verbal learning and memory, visual learning memory and reasoning and problem solving) fit well with the literature on healthy volunteers (Riedel et al., 2006) and are employed in this review (Nuechterlein et al., 2004).

Considering the mixed nature of the literature, a descriptive statistic quantifying the evidence would facilitate its interpretation. Pietrzak et al. (2006) summarize the results within each domain as a percentage of studies reporting cognition enhancing effects of MPH. In order to also take into account factors that determine the statistical power of the studies (number of participants and number of tasks/measures), we established a weighed percentage reflecting the relative contribution of the studies reviewed to the MPH effects (Kleijnen et al., 1991).

Effects of dopamine on cognition are often described to follow an inverted-U-shaped curve in which intermediate levels of neurotransmitter activity lead to optimal cognitive performance, but lower and higher levels may lead to suboptimal performance (Husain and Mehta, 2011). Since MPH blocks the dopamine and noradrenaline transporters, thereby blocking reuptake of these neurotransmitters, MPH leads to increased levels of dopamine and noradrenaline availability (Volkow et al., 1998; Hannestad et al., 2010) and may demonstrate inverted-U properties. Moreover, dose–response relationships may vary between cognitive domains. Therefore, we report a weighed percentage reflecting MPH effects, equivalent to the one described above for low, medium and high doses, respectively.

The aim of this review is to answer the question whether MPH can enhance cognitive function in healthy individuals. This review specifically looks at the acute effects of MPH on healthy adults (i.e. excluding effects of other medications, effects on children, effects after sleep deprivation and long term effects). Effects on the elderly are discussed in a separate paragraph, as well as imaging data.

Method

Literature search

A literature search was performed in PubMed and Psychinfo using the search term Methylphenidate combined with the terms cognition, neuropsychology, executive, working memory, vigilance and inhibition in separate searches. Reference lists of extracted articles were screened for omissions in our search. Studies that were included fulfilled the following criteria: used immediate release formulation of MPH; assessed the effects of acute doses of MPH; assessed healthy human adults 18 yr and older; assessed the effects using cognitive tasks; were written in English; published in a peer-reviewed journal. Both studies employing a within- and between-subjects design were included. Studies assessing chronic or subchronic effects of MPH were not included. The earliest study included was published March 1978 and the last included study was published May 2013.

Categorization

The neuropsychological tests were categorized into six different cognitive domains: speed of processing, attention/vigilance, working memory, verbal learning and memory, visual learning and memory and reasoning and problem solving (Nuechterlein et al., 2004). Tasks employed in the included articles were categorized following the classification by Nuechterlein et al. (2004). If more than one cognitive domain was applicable, tasks were still classified as measuring one domain only (i.e. the most salient domain). Tasks that could not be categorized in any of these six cognitive domains were not included in the analysis (for instance ‘spatial bias’ in (Dodds et al., 2008a)).

Outcome measure

For each domain a weighed percentage was calculated reflecting to what extent MPH affects task performance in that specific domain. It was calculated as follows:  

in which number of participants was used as reported for within-subjects designs, while for between-subjects designs, the number of participants that received MPH was used in the calculation. Outcome significance testing was defined as: 1 = significantly improved task performance; −1 = significantly impaired task performance; 0 = no significant effect; 0.5 = trend towards improved task performance; −0.5 = trend towards impaired task performance, and the factor to correct for multiple comparisons (relative contribution) was defined as: 1/number of tasks or measures within the study that are reported in Table 2 (i.e. if three measures were reported in the article, but only two were listed in Table 2., the factor was 1/2 = 0.5). Only data of participants aged between 18 and 60 were included in the calculation. Data on elderly participants are discussed in a separate paragraph.

Table 1

Abbreviations

ADHD	Attention deficit hyperactivity disorder
A/V	Attention/vigilance
CNV	Contingent negative variation
CPT	Continuous performance test
DSST	Digit symbol substitution
ID/ED	Intra-/extradimensional shift task
MPH	Methylphenidate
PAL	Paired associates learning
PASAT	Paced auditory serial addition test
RPS	Reasoning and problem solving
RT	Response time
RVIP	Rapid visual information processing
SERS	Stimulus evaluation/response selection task
SMS	Sternberg memory scanning task
SoP	Speed of processing
TMT	Trail making test
TOVA	Test of variables of attention
VEM	Verbal learning and memory
VSM	Visual learning and memory
WLT	Word learning test
WM	Working memory

Table 2

Summary of studies on cognitive effects of MPH in healthy volunteers

Study	No. of pp M, F (age) design	Dose	Task/measure	Domain	Effect
(Agay et al., 2010) [1]	8M, 8F (32.9) placebo controlled between groups	15 mg1	CPT (omission errors)	A/V	ns
			CPT (commission errors)	A/V	ns
			Digit span	WM	sig
(Aman et al., 1984) [2]	5M, 7F(28.3) placebo controlled crossover	0.3 mg/kg	CPT (omission errors)	A/V	ns
			CPT (commission errors)	A/V	sig
			CPT (RT)	A/V	ns
(Anderer et al., 2002) [3]	10M, 10F (23–34) placebo controlled crossover	20 mg	Oddball (error rate)	A/V	ns
(Ben-Itzhak et al., 2008) [4]	9M, 17F (73.8) Placebo-controlled crossover	20 mg	Go-NoGo	A/V	sig2
			nonverbal memory	VSM	ns
(Bernard et al., 2011) [5]	9M, 9F (26.7) placebo controlled crossover	40 mg	Choice reaction time (recognition reaction time)	SoP	sig
			Choice reaction time (total reaction time)	SoP	ns
			Choice reaction time (motor reaction time)	SoP	ns
(Bishop et al., 1997) [6]	6M, 3F (21–35) placebo controlled crossover	10 mg (2x/day)	Divided attention	A/V	ns
			Auditory vigilance	A/V	ns
(Brignell and Curran, 2006) [7]	6M, 10F (18–35) placebo controlled between groups	40 mg	Fear conditioning	VSM	ns
(Brignell et al., 2007) [8]	16 HV (18–35) placebo controlled between groups	40 mg	Story task	VEM	Trend3
(Brumaghim and Klorman, 1998) [9]	12M, 20F (20.9) placebo controlled crossover	0.3 mg/kg	PAL (CVC pairs)	VEM	ns
(Brumaghim et al., 1987) study 1 [10]	19M (19.4) placebo controlled crossover	0.3 mg/kg	SMS	WM	sig
(Brumaghim et al., 1987) study 2	6M, 8F (20.0) placebo controlled crossover	0.3 mg/kg	SMS	WM	sig
(Callaway, 1984) [11]	8F (30–40)	5/10/20 mg	SERS	SoP	sig
	8F (60–75)	5/10/20 mg	SERS	SoP	ns
	placebo controlled crossover
(Camp-Bruno and Herting, 1994) [12]	7M, 8F(19–33) placebo controlled between groups	20 mg	CPT (RT)	A/V	sig
			CPT (sensitivity)	A/V	ns
			CPT (ln beta)	A/V	ns
			Word learning	VEM	ns
			Buschke selective reminding	VEM	sig
(Campbell-Meiklejohn et al., 2012). [13]	19F (23) placebo controlled between groups	20 mg	N-back	WM	ns
(Clark et al., 1986a) [14]	12M (18–30) placebo controlled crossover	0.65 mg/kg	Dichotic monitoring task (target detection)	A/V	ns
			Dichotic monitoring task (error rate)	A/V	ns
			Dichotic monitoring task (RT)	A/V	ns
			Dichotic monitoring task (signal detection)	A/V	ns
(Clark et al., 1986b) [15]	10M (18–30) placebo controlled crossover	0.65 mg/kg	Dichotic monitoring task (target detection)	A/V	ns
			Dichotic monitoring task (error rate)	A/V	sig
			Dichotic monitoring task (RT)	A/V	ns
			Dichotic monitoring task (signal detection)	A/V	ns
(Clatworthy et al., 2009) [16]	10M (22–32) placebo controlled crossover	60 mg	Reversal learning	VSM	ns
			Spatial WM	WM	ns
(Coons et al., 1981) study 1 [17]	13M (23.84) placebo controlled crossover	20 mg	CPT X (omission errors)	A/V	ns
			CPT X (commission errors)	A/V	ns
			CPT BX (omission errors)	A/V	ns
			CPT BX (commission errors)	A/V	ns
(Coons et al., 1981) study 2	23M (19.7) placebo controlled crossover	20 mg	CPT X (omission errors)	A/V	ns
			CPT X (commission errors)	A/V	ns
			CPT BX (omission errors)	A/V	sig
			CPT BX (commission errors)	A/V	ns
			CPT Double (omission errors)	A/V	sig
			CPT Double (commission errors)	A/V	ns
			Oddball (omission errors)	A/V	ns
			Oddball (commission errors)	A/V	ns
			Choice RT (omission errors)	SoP	sig
			Choice RT (commission errors)	SoP	ns
			Choice RT (response time)	SoP	ns
(Cooper et al., 2005) [18]	32M (22.3) placebo controlled crossover	5/15/45 mg	CPT (omission errors)	A/V	sig
			CPT (commission errors)	A/V	ns
			CPT (RT)/N back	A/V	sig
(Costa et al., 2013) [19]	54M(23.7) placebo controlled crossover	40 mg	Go/No-go Task	A/V	ns
			Stop signal task	A/V	ns
(Dodds et al., 2008b) [20]	14M, 6F (22.2) placebo controlled crossover	60 mg	Probabilistic reversal learning	VSM	ns
(Drijgers et al., 2012) 21	21M, 2F (65.4) Placebo crossover	10 mg	Letter digit substitution	SoP	ns
			Simple reaction time task	SoP	ns
			Choice reaction time task	SoP	ns
(Elliott et al., 1997) [22]	28M (21.3) placebo controlled crossover	20/40 mg	Spatial WM	WM	sig4
			Tower of London (old)	RPS	sig5
			Tower of London (new)	RPS	sig6
			Verbal fluency test	SoP	ns
			Spatial span	WM	sig4
			ID/ED shift task	A/V	ns
			Sequence generation	RPS	sig
			RVIP (response latency)	A/V	sig
			RVIP (performance)	A/V	ns
(Finke et al., 2010) [23]	9M, 9F (20–35) placebo controlled crossover	40 mg	Visual perceptual processing speed	SoP	sig
			Visual short-term memory storage capacity	VSM	ns
(Fitzpatrick et al., 1988) [24]	20M (19.7) placebo controlled crossover	0.3 mg/kg	Memory scanning task	WM	sig
(Halliday et al., 1986) exp 1 [25]	8F (30–40) placebo controlled crossover	5/10/20 mg	SERS	SoP	sig
			SMS	WM	ns
(Halliday et al., 1986) exp 2	12M (26) placebo controlled crossover	10 mg	SERS	SoP	sig
			CPT (commission errors)	A/V	sig
			motor task	SoP	sig
(Hermens et al., 2007) [26]	32M (22.3) placebo controlled crossover	5/15/45 mg	Oddball	A/V	sig
			CPT (RT)	A/V	sig
			CPT (omission errors)	A/V	sig
			CPT (commission errors)	A/V	ns
			CPT (total errors)	A/V	sig
			Maze	RPS	ns
			Mackworth clock (RT variability)	A/V	sig
			Mackworth clock (false negatives)	A/V	sig
			Mackworth clock (false positives)	A/V	ns
			Mackworth clock (total errors)	A/V	sig
			Verbal memory recall	VEM	ns
			Choice reaction time	SoP	ns
			Switching of attention (TMT A)	SoP	ns
			Switching of attention (TMT B)	SoP	ns
			PASAT (RT)	A/V	sig
			PAL (word pairs)	VEM	ns
(Hester et al., 2012) [27]	27M (22) placebo controlled crossover	30 mg	Go/No-go task (accuracy)	A/V	sig
			Go/No-go task (Go RT)	A/V	ns
			Go/No-go task (No-go error RT)	A/V	ns
(Hink et al., 1978) [28]	12M (19–28) placebo controlled crossover	10 mg	Target detection task	A/V	ns
(Izquierdo et al., 2008) [29]	7M, 5F (40–74) placebo controlled crossover	10 mg	Incidental memory task	VEM	sig
(Izquierdo et al., 2008)	11M, 9F (35–74) placebo controlled crossover	10 mg	Formal memory task	VEM	Sig/ns7
(Kollins et al., 1998) [30]	5M, 5F (30.7) placebo controlled crossover	20/40 mg	DSST	SoP	ns
			circular lights task	SoP	sig
(Kratz et al., 2009) [31]	8M, 6F (20–40) placebo controlled crossover	20 mg	Go/No-go task (hit rate)	A/V	ns
			Go/No-go task (RT)	A/V	sig
			Impulsivity errors	A/V	ns
(Kupietz et al., 1980) [32]	5M, 4F (28.7) placebo controlled crossover	5/10 mg	Learning beginning reading vocabulary task
			(simultaneous method)	VEM	sig
			(progressive method)	VEM	ns
(Kuypers and Ramaekers, 2005) [33]	9M, 9F (26.2) placebo controlled crossover	20 mg	WLT	VEM	ns
			Syntactic resasoning task	WM	ns
			DSST	SoP	ns
(Kuypers and Ramaekers, 2007) [34]	9M, 9F (26.2) placebo controlled crossover	20 mg	Spatial memory task	VSM	ns
			Change blindness task	VSM	ns
(Linssen et al., 2011) [35]	19M (23.4) placebo controlled crossover	10/20/40 mg	CNV Lines	SoP	sig
			CNV Stoplight	SoP	sig
(Linssen et al., 2012) [36]	19M (23.4) placebo controlled crossover	10/20/40 mg	WLT	VEM	sig
			Spatial WM	WM	ns
			Set shifting	A/V	sig
			Stop signal task	A/V	sig
			Tower of London	RPS	ns
(Marquand et al., 2011) [37]	15M (20–39) placebo controlled crossover	30 mg	Spatial WM	WM	ns
(Mehta et al., 2000a) [38]	10M (34.8) placebo controlled crossover	40 mg	Spatial WM	WM	sig
(Moeller et al., 2012) [39]	14M, 1F (38.9)	20 mg	Stroop	SoP	ns
(Muller et al., 2005) [40]	4M, 8F (69.8) placebo controlled crossover	20 mg	4 choice motor reaction task	A/V	Sig8
(Nandam et al., 2011) [41]	24M (23) placebo controlled crossover	30 mg	Stop signal task	A/V	sig
(Naylor et al., 1985) [42]	8F (30–39) placebo controlled crossover	5/10/20 mg	SERS	SoP	sig
(Oken et al., 1995) [43]	11M, 12F (25) placebo controlled Crossover	0.2 mg/kg	Covert orienting of spatial attention task (RT)	A/V	sig
			Covert orienting of spatial attention task (Errors)	A/V	ns
			Parallel visual search task (RT)	A/V	ns
			Parallel visual search task (Errors)	A/V	ns
			Serial visual search task (RT)	A/V	ns
			Serial visual search task (Errors)	A/V	ns
			Digit span	WM	ns
(Pauls et al., 2012) [44]	16M (23.6) placebo controlled crossover	40 mg	Stop signal task (original version)	A/V	ns
			Stop signal task (adapted version)	A/V	sig
(Ramasubbu et al., 2012) [45]	5M, 8F (28) placebo controlled crossover	20 mg	2-back task (correct responses)	WM	sig
			2-back task (incorrect responses)	WM	ns
			2-back task (missed responses)	WM	sig
			2-back task (reaction time)	WM	ns
			0-back task (correct responses)	A/V	ns
			0-back task (incorrect responses)	A/V	ns
			0-back task (missed responses)	A/V	ns
			0-back task (reaction time)	A/V	sig
(Roehrs et al., 1999) [46]	2M, 4F (21–30) placebo controlled crossover	10 mg	Divided-attention task (central RT)	A/V	sig
			Divided-attention task (peripheral RT)	A/V	ns
			Divided-attention task (tracking deviations)	A/V	ns
			auditory vigilance task (mean RT)	A/V	ns
			auditory vigilance task (Errors)	A/V	ns
(Rogers et al., 1999) [47]	16M (20.4) placebo controlled between groups	40 mg	ID/ED shift task	A/V	sig (decreased performance)
(Rush et al., 2001) [48]	4M, 4F (28) placebo controlled crossover	20/40 mg	DSST	SoP	ns
(Rush et al., 1998) [49]	2M, 3F (36) placebo controlled crossover	5/10/20/40 mg	DSST	SoP	ns
(Schroeder et al., 1987) [50]	10M (18–40) No placebo between groups	0.15/0.30 mg/kg	Concurrent probability matching
			Concurrent probability matching (hit rate)	RPS	ns
			Concurrent probability matching (changeover)	RPS	sig (decreased performance)
			Concurrent probability matching (strategy)	RPS	sig (decreased performance)
(Stoops et al., 2005) [51]	2M, 5F (24) placebo controlled crossover	10/20/40 mg	Arithmetic problems	RPS	sig
(Strauss et al., 1984) [52]	22M (19.2) placebo controlled crossover crossover	20 mg	CPT Double (omission errors)	A/V	sig
			CPT Double (commission errors)	A/V	trend
			CPT Double (sensitivity)	A/V	sig
			CPT Double (RT)	A/V	sig
			PAL (CVC pairs)	VEM	ns
(Studer et al., 2010) [53]	5M, 6F (29.7) placebo controlled crossover	20 mg	Serial visual WM task	WM	ns
(Theunissen et al., 2009) [54]	5M, 11F (21.8) placebo controlled crossover	20 mg	Critical tracking task	SoP	ns
			Divided attention task	A/V	sig
			Mackworth Clock task	A/V	ns
			Stop signal task	A/V	ns
(Tomasi et al., 2011) [55]	16M (33) placebo controlled between groups	20 mg	N-back task (RT)	WM	sig
			N-back task (accuracy)	WM	ns
			visual attention task	A/V	ns
(Turner et al., 2003) [56]	60M (61.4) placebo controlled between groups	20–40 mg	Digit span	WM	ns
			PAL (nonverbal)	VSM	ns
			Spatial WM	WM	ns
			Spatial span task	WM	ns
			Tower of London	RPS	ns
			RVIP	A/V	ns
			ID/ED shift task	A/V	Sig9
			Stop signal task	A/V	ns
(Unrug et al., 1997) [57]	6M, 6F placebo controlled crossover	20 mg	WLT	VEM	ns
(Volkow et al., 2008) [58]	12M, 11F (32) placebo controlled crossover	20 mg	Numerical problems	RPS	ns
(Wetzel et al., 1981) exp 1 [59]	6M, 6F (27.5) placebo controlled crossover	0.5 mg/kg	PAL (word pairs)	VEM	sig (decreased performance)
			Picture recognition	VSM	ns
			Story recall	VEM	sig (decreased performance)
(Wetzel et al., 1981) exp 2	6M, 6F (26.6) placebo controlled crossover	0.1/0.25 mg/kg	PAL (word pairs)	VEM	ns
			Picture recognition	VSM	ns
			Story recall	VEM	ns
(Zhu et al., 2013) [60]	18M (19–24) placebo controlled crossover	20 mg	Go/No-go	A/V	ns

1

Participants received 15 mg MPH unless high or low in body weight in which case they received 20 or 10 mg respectively.

2

MPH increased accuracy but did not affect response time.

3

MPH lessens effect of emotionally arousing material on memory, higher performance on neutral material compared to placebo.

4

Enhanced performance when MPH was taken on the first session.

5

MPH causing impairment when taken on second session.

6

Enhanced performance when MPH was taken on the first session, but impaired when taken on second session.

7

MPH improved performance of young volunteers but not old volunteers.

8

MPH impaired performance of easy task, but enhanced performance of difficult task.

9

MPH slowed response but not improve accuracy.

Another frequently considered evaluation parameter is effect size, and it would be interesting to compare this to our study evaluation measure. However, due to the fact that most of the reported studies did not convey all essential information necessary to calculate it, effect sizes were not reported.

Dose effects

The outcome measure as detailed above was also calculated for low, medium and high doses separately. A low, medium or high dose was defined as follows: low: ⩽10 mg or ⩽0.15 mg/kg; medium: >10 mg, ⩽20 mg or >0.15 mg/kg, ⩽0.3 mg/kg; high: >20 mg or >0.3 mg/kg. If it was not clearly stated which dose(s) led to significant effects, it was assumed that the reported effect was applicable to every administered dose.

Results and discussion

In total 60 studies met the inclusion criteria. Of these, 56 included healthy volunteers between 18 and 60 yr old. All studies are listed in Table 2. The extent to which MPH enhances cognitive performance (quantified by a weighed percentage) is reported in Table 3. The results show that MPH most effectively enhances performance in the domain of working memory. Second most affected were tasks measuring speed of processing, followed by verbal learning and memory, attention/vigilance, reasoning and problem solving, and visual learning and memory. In the next few paragraphs, the results per domain are discussed in more detail.

Table 3

Percentages of studies showing cognition enhancing effects of methylphenidate in each of six cognitive domains (Total) and per dose level (Low, Medium, High). Study reference numbers refer to studies described in Table 2

	Total	Low	Medium	High	No. of measures (No. of studies)	Study references
Working memory	65	0	74	41	21 (15)	1,9,12,16,22,24,25,33,36,37,38,43,45,53,55
Speed of processing	48	79	46	44	25 (15)	5,11,17,22,23,25,26,30,33,35,39,42,48,49,54
Verbal learning and memory	31	64	75	12	18 (11)	8,10,13,26,29,32,33,36,52,57,59
Attention and vigilance	29	37	32	38	87 (27)	1,2,3,6,13,14,15,17,18,19,22,25,26,27,28,31,36,41,43,44,45,46,47,52,54,55,60
Reasoning and Problem solving	18	1	18	74	10 (6)	22,26,36,50,51,58
Visual learning and memory	0	0	0	0	8 (6)	7,16,20,23,34,59

Working memory

Working memory involves the temporary storage and manipulation of information (Baddeley, 1992). For the purpose of this review, no distinction was made between short term memory and working memory tasks. A common working memory task is the digit span, requiring subjects to repeat a sequence of digits either in the exact same or reversed order. More complex tasks include memory scanning tasks and N-back tasks. Memory scanning tasks present participants with probes that they have to compare to a memory set that may vary in size, and hence, memory load (Sternberg, 1966). N-back tasks require a response to stimuli that are the same as the stimulus that was presented N-trials before. Other tasks specifically assess spatial working memory, which is often associated with dopaminergic activity (Mehta and Riedel, 2006).

Although the number of MPH studies including a test of working memory was not very high, the proportion that shows enhancing effects in this cognitive domain was the highest of all reported domains, namely 65%.

The proportion of the effect of MPH was similar across different types of tasks. The highest proportion of significant effects was observed with the medium dose. This was also reflected by a subset of tests, namely the spatial working memory tests. It is possible that the MPH-effect follows an inverted-U-shaped curve as a function of dose, since a high dose (60 mg) exerted no effect on a spatial working memory task that was affected by lower doses in other studies ((Elliott et al., 1997; Mehta et al., 2000b; Clatworthy et al., 2009), but note that Elliott et al. (1997) found no main effects).

Speed of processing

Tasks that are classified as a measure of speed of processing are relatively simple, involving fundamental processes such as perception and motor action. This category includes, for example, digit symbol substitution tests (DSST) and trail making tests (TMT); version A and B

The speed of processing domain was, with 48%, the second most affected domain in terms of performance enhancing effects of MPH. This is consistent with the notion that MPH speeds up response time in healthy volunteers (Elliott et al., 1997), reflecting response-readiness enhancing effects (Linssen et al., 2011).

A striking observation is that the low dose exerted the highest proportion of effects on cognitive performance in this domain, followed by the medium dose, with the highest dose showing the lowest proportion of effects. This suggests that within this domain the optimal dose is rather low and the higher the dose, the less healthy volunteers benefit from it.

Verbal learning and memory

Whereas MPH affected the former two domains in 48% or more of the reported measurements, the domain of verbal learning and memory was somewhat less affected. A proportion of 31% of the studies showed enhanced performance in this domain after administration of MPH. The cognitive domain of verbal learning and memory typically refers to declarative memory tasks such as word learning tests, verbal paired associates learning (PAL) and story recall. Verbal PAL tasks involve learning pairs of stimuli (words or letters) and recalling the corresponding member of the pair upon presentation of probe stimuli.

Positive effects of MPH on declarative memory fit well with accounts of related stimulant drugs affecting word list learning (Soetens et al., 1993, 1995; Zeeuws and Soetens, 2007). The relationship between MPH dose and its effect on word list learning seems to be a positive and linear one. This contrasts with the striking finding that a high MPH dose can decrease performance on a PAL task and story recall. Across the domain of verbal learning and memory the low and medium doses exert more cognition enhancing effects than high doses. Hence, even within one domain, different dose–response relationships may be observed for different tasks.

Attention/vigilance

As William James stated in 1890, ‘everyone knows what attention is’ (James, 1890). It is, however, difficult to give a clear and inclusive definition. It involves filtering information, focusing on certain aspects of what is perceived, while disregarding others. Many different forms of attention are distinguished (e.g. sustained attention, focused attention, divided attention). In this review, however, a collective domain ‘attention and vigilance’ is employed.

Based on previous literature, associating attention with dopaminergic activity (Nieoullon, 2002; Nieoullon and Coquerel, 2003; Cools and Robbins, 2004), it was expected that the domain of attention/vigilance would be one of the most affected by MPH. However, with 29%, this domain was only the fourth most affected. A possibly confounding factor might be that tasks from the attention/vigilance domain are often included in MPH studies as control measures because they are thought to be sensitive to MPH effects. Indeed, this domain was assessed most frequently of all domains (87 measures across 27 studies). Furthermore, the tasks within this domain were often split into multiple measures, which may affect the final outcome.

Reasoning and problem solving

The next cognitive domain involves planning and decision making, aspects of cognition that are usually referred to as executive functioning. However, in order to allow a separate category on working memory, which is often included in executive functioning, this domain is named reasoning and problem solving. A typical task in this domain is the Tower of London, in which balls have to be arranged on pegs according to a predefined pattern in as few steps as possible.

Overall, 18% of the results reflected improved performance within this domain after MPH. The reported results suggested that improvement of performance occurs more frequently with a higher dose.

Visual learning and memory

The last cognitive domain to discuss is visual learning and memory. Tasks classified as assessing performance in this cognitive domain are, for example, picture recognition tests and other learning and memory tasks involving visual stimuli. None of the studies found an effect of MPH on visual learning and memory (Wetzel et al., 1981; Bullmore et al., 2003; Brignell and Curran, 2006; Kuypers and Ramaekers, 2007; Dodds et al., 2008b; Clatworthy et al., 2009; Finke et al., 2010). It is, however, important to note that the data reported within this domain only comprises eight measures across six studies. This gives the conclusion regarding the absence of an MPH effect on visual learning and memory somewhat less weight.

Imaging studies on cognitive performance under the influence of MPH

Of the studies included in this review, 26 studies obtained imaging data during cognitive testing. In 14 of these studies electroencephalography (EEG) measures were taken. Furthermore, there were seven studies with functional magnetic resonance imaging (fMRI) data, three with positron emission tomography (PET) data, one with transcranial magnetic stimulation (TMS) data and one with functional near-infrared spectroscopy (fNIRS) data.

The EEG data are summarized in Table 4. Components often assessed during attention, working memory or speed of processing tasks include N1, P2, N2 and P3 ERP components. While the former two components, N1 and P2, are thought to reflect perceptual processes such as orienting and directing of attention, the N2 and P3 are cognitive components reflecting allocation of attentional resources and stimulus evaluation (Anderer et al., 2002). The studies reviewed found no effects of MPH on the earlier components, while increased P3 source strength, increased P3b amplitude and shorter P3b latency were observed in several studies (Coons et al., 1981; Strauss et al., 1984; Anderer et al., 2002; Cooper et al., 2005). This is thought to reflect an MPH-induced enhancement of attentional processes or increased recruitment of attentional resources. Other studies did not report an effect of MPH on P3 (Hink et al., 1978; Callaway, 1984; Naylor et al., 1985; Fitzpatrick et al., 1988; Hermens et al., 2007; Studer et al., 2010). An explanation offered for these findings is that MPH affects response processing, as reflected by faster responses, but that MPH does not speed stimulus evaluation processing.

Table 4

Summary of EEG studies on cognitive effects of MPH in healthy volunteers

Study	Dependent variables	Task*	Domain	Main finding
(Anderer et al., 2002)	N1, P2, N2 and P3	Oddball	A/V	Increased P300 source strength
(Brumaghim et al., 1998)	P3b, P2, late slow waves	PAL (CVC pairs)	VEM	Increased parietal P3b amplitude
(Brumaghim et al., 1987)	P3b	SMS	WM	Shorter P3b latency
(Callaway, 1984)	P3b	SERS	SoP	No effect
(Coons et al., 1981)	LPC, CNV	CPT, Oddball, Choice RT	A/V, SoP	Increased LPC in CPT, increased CNV in choice RT
(Cooper et al., 2005)	N1, P2, N2 and P3	CPT	A/V	Shorter P3b latency, increased P3b amplitude
(Fitzpatrick et al., 1988)	P3b	Memory scanning task	WM	No effect
(Hermens et al., 2007)	N1, P2, N2 and P3	Oddball, CPT	A/V	No effect
(Hink et al., 1978)	N1, P2 and P3	Target detection task	A/V	No effect
Linssen et al. (2011)	CNV	CNV task	SoP	Increased CNV amplitude
(Naylor et al., 1985)	P3	SERS	SoP	No effect
(Oken et al., 1995)	Event-related desynchronization	Visual search tasks	A/V	No effect
(Strauss et al., 1984)	P2, P3a, P3b	CPT, PAL	A/V	Shorter P3b latency, increased P3b amplitude for CPT
(Studer et al., 2010)	P3, slow waves	Serial visual WM task	WM	No effect

*

Task during which the imaging data were acquired.

Another component assessed in the studies listed in Table 4 was the Contingent Negative Variation (CNV). MPH was observed to enhance CNV amplitude, reflecting increased response readiness (Coons et al., 1981; Linssen et al., 2011). Event-related desynchronization, measured during a visual search task, was not affected by MPH (Oken et al., 1995).

Of the seven studies in this review that acquired fMRI data, three studies assessed brain activity during response inhibition tasks, i.e. Go/No-go and stop signal (Hester et al., 2012; Pauls et al., 2012; Costa et al., 2013). Hester et al. (2012) observed that activation of the dorsal anterior cingulated cortex and the inferior parietal lobe differed between errors of which the participants were aware vs. unaware in a Go/No-go task. MPH increased these differences in activation of the dorsal anterior cingulate cortex and the inferior parietal lobe. This is suggestive of the underlying mechanism leading to improved response inhibition with MPH. Pauls et al. (2012) suggest that inhibition effects are mediated through MPH effects on the attentional mechanisms. Pauls et al. (2012) reported reduced activation of different regions that are associated with the ventral attention system, i.e. the right inferior frontal gyrus and insula, during various types of infrequent stimuli. Costa et al. (2013) showed that MPH increased activity in the putamen during error of inhibition but not during successful inhibition in the Go/No-go task, while no such effect was observed in the stop signal task. This effect may be due to saliency of errors in the Go/No-go task, suggesting that MPH interacts with saliency.

Two fMRI studies on working memory showed increased activation of the dorsal attention network and altered default mode network activation with MPH (Marquand et al., 2011; Tomasi et al., 2011), suggesting that MPH may exert its effects on cognition partly by modulating the default mode network.

During a probabilistic reversal learning task MPH decreased activation in the ventral striatum during response switching, while activation changed in the prefrontal cortex if there was no switching of response. MPH reduced dorsal anterior cingulate cortex activation during errors on a stroop task (Moeller et al., 2012).

Among the imaging studies were three PET studies. One PET study revealed that MPH reduced the amount of glucose used by the brain during cognitive task performance (Dodds et al., 2008b). This is thought to reflect focusing of attention. MPH was also shown to reduce regional cerebral blood flow in the dorsolateral prefrontal cortex and posterior parietal cortex during MPH-induced improved performance of a working memory task (Mehta et al., 2000b). Finally, MPH-induced enhancement of cognitive performance could be predicted by its effect on dopamine receptor availability, as measured with PET (Clatworthy et al., 2009).

Using TMS, MPH was shown to alter motor system excitability in a response inhibition task (Kratz et al., 2009), suggesting MPH can fine-tune the motor system. Finally, a study employing fNIRS reported decreased oxyhemoglobin concentration with MPH compared to placebo in the right frontal lobe, consistent with the PET finding by Mehta et al. (2000) described above.

In sum, the imaging studies on cognitive performance under the influence of MPH show that MPH enhances attentional processing and modulates activity in brain areas associated with inhibition and attention. However, more studies are needed to draw more exact conclusions, as existing studies vary widely with respect to task, method and peak/area of interest.

Cognitive effects of MPH in the elderly

A limited number of studies investigated the effects of MPH on cognition in the elderly (Callaway, 1984; Turner et al., 2003; Muller et al., 2005; Ben-Itzhak et al., 2008; Izquierdo et al., 2008; Drijgers et al., 2012). The effect of MPH on working memory performance was tested in an extensive study by Turner et al. (2003) by means of a digit span task, a spatial working memory task and a spatial span task. None of the tasks was significantly affected by MPH.

Speed of processing was also not affected as measured with a SERS task in the study of Callaway (1984) and letter digit substitution and simple and choice reaction time tasks in the study by Drijgers et al. (2012). The four-choice motor reaction task employed by Muller et al. (2005), showed a performance enhancing effect of MPH on a difficult version of the task. In this uncued version, participants were asked to move a joystick towards a stimulus that was presented in an uncued location. Performance on the cued version of the task was impaired by MPH.

Effects of MPH on verbal learning and memory was assessed by Izquierdo (2008) with tests of formal and incidental memory in which memory of, respectively, unintentionally remembered an intentionally remembered information was tested. MPH was shown to reverse age-related decline of persistence of incidental memory. No such effect was observed for formal memory.

Reasoning and problem solving, as measured with a Tower of London task, did not reveal any effects of MPH on older adults' performance.

Effects of MPH on older adults' attention and vigilance task performance was tested in four different tasks i.e. RVIP task; ID/ED shift task; stop signal task (Turner et al., 2003); Go/No-go task (Ben-Itzhak et al., 2008). Performance of the RVIP and stop signal tasks was not affected by MPH. In the ID/ED task a significant slowing of responses similar to that in young volunteers was observed. However, this was not accompanied by an improvement in accuracy, as is the case in young volunteers. In the Go/No-go task, on the other hand, accuracy improved under the influence of MPH, while response time was not affected.

Within the domain of visual learning and memory, two tests were administered, i.e. a non-verbal PAL test and a test of nonverbal memory in which memory of the spatial orientation of geometric objects is tested with a recognition test (Ben-Itzhak et al., 2008). Performance of neither of these tasks by older adults was affected by MPH.

In sum, MPH does not improve working memory, reasoning and problem solving and visual learning and memory in older adults. Some effects were observed within the domains of speed of processing, verbal learning and memory and attention and vigilance. These results only partly parallel those in younger volunteers, with the main difference being the lack of effects on working memory in older adults. However, the limited number of studies in older adults limits the conclusions that can be drawn from these data.

Adverse effects and ethical issues

While we cannot speak of a global cognition enhancing effect of MPH, the research reviewed in this paper shows that MPH can improve working memory and speed of processing and, to a lesser extent, verbal learning and memory, reasoning and problem solving and attention and vigilance in healthy volunteers. Reports of cognition enhancing effects of MPH might encourage such use. We would, therefore, like to put our findings in perspective and discuss adverse effects of MPH and related ethical issues.

The effectiveness of MPH as cognitive enhancer should not be overestimated (Ragan et al., 2013). All experimental studies reviewed here reflect effects in relatively homogeneous groups assessed in controlled environments using sensitive tests. In daily life, circumstances are much more diverse and complex, which makes it hard to verify the enhancing effects of MPH on performance in authentic situations. Any such effect is likely to be relatively small in comparison to the effects of adequate sleep, education and work–life balance, and will be subject to individual differences.

Even if MPH substantially improved real-life performances, the question remains whether the benefits outweigh the risks. Common adverse effects of chronic MPH use include insomnia, nervousness, irritability, anxiety, jitteriness, increased heart rate, dizziness, drowsiness, headache, stomach ache, anorexia and appetite suppression (Diller, 1996; Klein-Schwartz, 2002; Repantis et al., 2010). Large doses of MPH can lead to psychosis, seizures and cardiovascular events (Lakhan and Kirchgessner, 2012). Cardiovascular effects associated with ADHD stimulant medications include hypertension and tachycardia (Lakhan and Kirchgessner, 2012).

Another risk related to MPH use is the potential abuse of the drug. When MPH is used appropriately for ADHD there is little evidence of addiction and/or abuse of MPH (Diller, 1996). However, MPH can have reinforcing effects through its effects on dopamine in the brain (Volkow et al., 1999). Reinforcing effects depend on the rate of uptake of the drug (Volkow et al., 2002; Leonard et al., 2004), which explains why people who use MPH to induce a high administer it intravenously or intranasally, while people who merely take it to stay awake, take it orally (Klein-Schwartz, 2002; Volkow and Swanson, 2003). Self-administration of MPH is likely limited by the low rate of clearance of MPH from the brain, in which it differs from cocaine (Volkow et al., 1995).

MPH use without prescription could be dangerous for various reasons. Individual risks of which users may not be aware, such as interaction with other medicinal drugs or abnormalities in the cardiovascular system, cannot be monitored by a physician if MPH is obtained otherwise (Sahakian and Morein-Zamir, 2007). If users purchase the pills online it might not actually be MPH, which may create additional unknown risks (Ragan et al., 2013). This may be a motive to regulate off-label stimulant use (Sahakian and Morein-Zamir, 2007). However, making MPH or other medications easily and legally available can lead to other problems with coercion to take such drugs (for example, in specific military or medical professions or in children) and fairness (if only rich people can afford to buy such drugs).

Attitudes towards cognitive enhancement vary between authors: while some are generally positive and optimistic and mainly call for guidelines (Sahakian and Morein-Zamir, 2007; Greely et al., 2008), others are more conservative (Ragan et al., 2013) or even consider the current debate to be a ‘bubble’ (Lucke et al., 2011) or ‘media hype’ (Partridge et al., 2011) and merely stress that the evidence for increased stimulant use is weak. However, everyone agrees that more research is needed, whether it is to obtain accurate prevalence rates or to increase our knowledge about the benefits and harm of cognition enhancing drug use in healthy individuals. We would like to add one recommendation to the existing abundance of recommendations: research and reviews of research on cognition-enhancing drugs should preferably address one single drug at a time. ‘Cognition-enhancing drugs’ are often described as if there were a group of virtually identical drugs with similar effectiveness and side effect profiles. It is very important to study specific, individual features of a drug, as it is essential to study a drug's effectiveness in relation to its side effects and other potential risks. Any description of a group of drugs is likely to be a generalization and may lead to over- or under-estimation of risks and benefits.

Conclusions

In this review the performance enhancing effect of MPH within various cognitive domains was quantified by a percentage (weighed on the basis of specified criteria such as number of subjects). This is a major improvement compared to earlier reviews in which only some domains of cognition were considered, the categorization in domains was less well specified (Repantis et al., 2010; Smith and Farah, 2011) and/or the quality was either not taken into account or based on criteria that hardly differentiate between studies (e.g. double blind design, randomization (Jadad et al., 1996)). Furthermore, this review compares 59 studies, which is considerably more than in previous reviews. The studies reviewed show that MPH improves cognitive performance in the healthy population in the domains of working memory and speed of processing, and to a lesser extent may also improve verbal learning and memory, attention and vigilance and reasoning and problem solving, but not visual learning and memory (see Table 3).

There were quite large differences between the domains with respect to the cognition enhancing effect of MPH, the lowest percentage being 0% and the highest 65%. These differences confirm that in studying cognitive performance it is important to distinguish between different domains of cognitive function. MPH may not globally improve cognitive performance, but it has potential to enhance certain aspects of cognitive performance at certain doses.

The dose–response relationship of the cognition modulating effect of MPH differs across different cognitive domains. In some domains, low doses are more effective than high doses. An explanation for this may be that, in healthy volunteers, dopamine availability is already close to the optimal level. Enhancing dopamine activity may push the level of dopamine beyond the optimum without improving performance, or even leading to suboptimal performance.

Factors limiting the conclusions in this review relate to the decisions that were made on how to categorize, list and present the data. Tasks not described by Nuechterlein et al. (2004) were categorized by the authors. Other decisions that were well thought through, but might still be somewhat arbitrary include: (1) some tasks were split into several measures whereas others were not; (2) a statistical trend was given a weight of 0.5; (3) if it was not clear which of the reported doses led to significant effects, it was assumed that the reported effect was applicable to every dose; (4) sometimes two low doses are reported, but they are only weighed once; (5) there was no correction for multiple comparisons for multiple dosing. Finally a major limitation of any review is that studies that did not show significant effects are generally underreported (i.e. publication bias). Therefore, any review is likely to present an overestimation of the reported effects.

In sum, MPH improves working memory and speed of processing and, to a lesser extent, verbal learning and memory, attention and vigilance and reasoning and problem solving in healthy young volunteers. MPH effects are dose-dependent and the dose–response relationship differs between the different cognitive domains. Imaging studies show enhanced attention processing and altered attention- and inhibition-related brain activation. MPH improves cognitive performance in the elderly on tasks of speed of processing, verbal learning and memory and attention and vigilance. MPH effects are, however, considered to be relatively small and MPH use is associated with side effects and other adverse consequences, such as potential abuse. Future studies should focus on MPH specifically to adequately asses the benefit–risk ratio of this specific drug.

Statement of Interest

Anke Linssen and Anke Sambeth have no financial interests to disclose. Eric FPM Vuurman is employed full time by Maastricht University. He was involved in conducting clinical trials for several pharmaceutical companies over the last three years: MSD, GSK and Transcept pharmaceuticals. Financial compensation for his work was only to Maastricht University and raises no conflict of interest. There were no other commercial or financial relationships that could be construed as a potential conflict of interest. Wim J. Riedel is honorary Professor at Maastricht University. He was an employee of F.Hoffmann-La Roche Ltd, Basel, Switzerland until December 2011 and since then has been an employee of Cambridge Cognition Ltd, Cambridge, United Kingdom.