Effect of prefrontal transcranial magnetic stimulation on spontaneous truth-telling

Abstract

Brain-process foundations of deceptive behaviour have become a subject of intensive study both in fundamental and applied neuroscience. Recently, utilization of transcranial magnetic stimulation has enhanced methodological rigour in this research because in addition to correlational studies causal effects of the distinct cortical systems involved can be studied. In these studies, dorsolateral prefrontal cortex has been implied as the brain area involved in deceptive behaviour. However, combined brain imaging and stimulation research has been concerned mostly with deceptive behaviour in the contexts of mock thefts and/or denial of recognition of critical objects. Spontaneous, “criminally decontextuated” propensity to lying and its dependence on the activity of selected brain structures has remained unexplored. The purpose of this work is to test whether spontaneous propensity to lying can be changed by brain stimulation. Here, we show that when subjects can name the colour of presented objects correctly or incorrectly at their free will, the tendency to stick to truthful answers can be manipulated by stimulation targeted at dorsolateral prefrontal cortex. Right hemisphere stimulation decreases lying, left hemisphere stimulation increases lying. Spontaneous choice to lie more or less can be influenced by brain stimulation.

Introduction

The distinction between brain processes involved in deceptive and honest behaviour have been the central issue in many studies (e.g., [1], [2], [3], [4], [5]). Various brain imaging methods, most commonly fMRI, have been used to study brain-process correlates of deceptive versus truthful behaviour [6], [7], [8], [9], [10], [11]. A “gold standard” in these investigations presumes comparing brain processes when individuals respond falsely or truthfully to experimental stimuli such as words, pictures, statements, etc. Typically a mock crime scenario or guilty knowledge situation is used and/or some stimulus is manipulated to be critical in terms of significance or its emotion-inducing capacity while subjects have to try to conceal or deny stimulus criticality. Also, factually correct personal information has been a matter of honesty or deceit in experimental settings. In one or another way criminality, deservedness for punishment, or personal significance have been modelled by an experimental design. However, deceptive behaviour and chances to discover it depend not only on actual guilty knowledge or “incriminating” circumstances, but also perhaps on natural propensity of a subject to lie more or lie less. This also refers to the ease with which a person is capable of or prone to counter-factual statements. Research on this aspect of lying and its brain bases has been largely overlooked. One aim of the present paper is to begin filling this gap (see the particulars later on in the text).

Neuroimaging studies in which specific cortical areas are shown to become activated during lying cannot be taken literally or interpreted simplistically to indicate a cortical “locus” for lying. As has been pointed out by several researchers [12], [13], [14] such studies only provide evidence of a correlation between the activation of certain brain regions and the occurrence of a specific behaviour. However, the same cortical areas – together or separately – can be involved in a variety of cognitive tasks when no deception is taking place. For example, the prefrontal, parietal, and anterior cingulated regions commonly activated in deception studies are also generally activated when executive processing is studied with no deception involved. Therefore, the activation of a cortical area during lying as a correlational cue does not prove a causal relationship between it and the occurrence of lying [13]. However, if by manipulating the state or activity of a certain area by transcranial stimulation artificially changes propensity to lying, we would be a step closer to unraveling the mechanisms involved in lying. Thus, another aim of this work is to use a causally relevant approach to studying the mechanisms behind spontaneous lying.

Studies of brain area activation through neuroimaging techniques can be said to fall broadly into two categories [15]: studies that attempt to elicit evidence of deception through the presence of a specific pattern of cortical activation and studies that aim at the neurocognitive processes underlying deception. In the latter category, lying is viewed as a complex, high-level cognitive and social task [2], [4], [6], [9], [16], [17]. It is important to specify whether deceiving and truth telling processes are special in any way, or whether they rely on a set of general-purpose processes [2], [16]. Studying causal effects should be important in this context.

In several works deception is interpreted as inhibition of expressing the truth (e.g., [6], [18]). Thus, creation of a lie may be mediated by the control processes in the prefrontal cortex (PFC) with the neural facilities of working memory involved. Consequently, when we artificially inhibit PFC structures involved in lying and/or disturb the balance between truth-telling and lying readiness through these mechanisms, measurable likelihood of the corresponding behaviour should also change. Furthermore, if lying means inhibitory cognitive control over truth-telling propensity then it is reasonable to assume that whenever subject has some (external) extra means for this control, lying should be easier and its detection more problematic. (Indeed, some works have pointed out the problem of cue-dependency in deception studies – in the most cases subjects are being told when to lie [2], [6], [16], [19]). To overcome also this problem and bring experiments closer to real life it may be advisable to use study designs in which subjects are free to choose when to lie or tell the truth like herein. Our present study follows this strategy.

In recent years transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS) have been invoked to examine causal effects of the brain areas in mediating deceit or truth-telling. These non-invasive brain stimulation methods can be more or less comfortably used with normal subjects. Both have been used to test the validity of brain imaging findings [13] or to try establish a causal relevance between the state of a cortical region and deceptive behaviour through the transient inhibition of cortical excitability [14]. TMS can alter brain activity in a specific cortical region through the use of targeted magnetic fields temporarily disrupting neural processing in the focal area [13], thus allowing to study the functioning of a certain area of the brain in relation to an existing behaviour through a measurement of small but significant alterations in the behaviour [20], [21]. The specifics and the basics of transcranial stimulation and tDCS and some of their known effects can be found in numerous pertinent publications [12], [13], [17], [18], [22], [23], [24], [25], [26]. Let us emphasize only some of the items central to our present study.

TMS uses the principle of electromagnetic induction, in order to produce a temporary and rather narrowly localised bioelectrical noise in the brain and/or to temporarily switch off or inhibit the functions in a particular region. In case of repetitive TMS (rTMS) it could alter the excitability of the cortex, either increasing or decreasing it, depending on the parameters of stimulation [22], [23]. Inhibition typically emerges with stimulation at about 1 Hz and excitation with stimulation at about 5 Hz and higher [13], [22]. Anodal tDCS seems to increase excitability in the areas of interest, whereas cathodal stimulation seems to have an inhibitory effect [13].

For example Knoch et al. [24] found that stimulation over the left dorsolateral prefrontal cortex (DLPFC) showed a frequency-dependent rTMS effect: counting bias in numerical random sequence generation was significantly reduced after the 1 Hz stimulation compared with baseline, but significantly exaggerated after the 10 Hz stimulation compared with 1 Hz stimulation. Subsequently, Knoch et al. [25] inhibited the right DLPFC with low frequency rTMS and observed significant increases in risky decision making in their subjects, compared to subjects in which inhibitory stimulation was received on the left DLPFC or a sham stimulation was applied to either side. Similarly, Fecteau et al. [26] used tDCS to impart anodal stimulation to the right DLPFC while applying a cathodal tDCS to the left DLPFC and found that subjects made more often the safer choice, while taking less time to evaluate and choose between low risk and high risk possibilities, compared with sham stimulation. Several studies have found functional laterality of the hemispheres: reduction of responses within right hemisphere and increased excitability in the left hemisphere during deceptive responses when compared to honest responses [18]; and on the contrary, in the left DLPFC stronger activations for higher demands on goal hierarchy were apparent [17]. As lying is a complex and demanding cognitive activity, interaction of these bilateral systems may be considerably involved in deception.

But is it realistic at all to experimentally manipulate readiness for deception? Priori et al. [27] gave an affirmative answer. Stimulation of the right and left DLPFC through anodal and cathodal direct currents was applied separately to these areas. The frequency of deceptive responses did not change, but a demonstrable increase in the time it took to make deceitful responses was found. Mameli et al. [5] found that anodal tDCS stimulation of the DLPFC bilaterally speeded up reaction times in the production of general knowledge related deceptive responses. Karim et al. [14] showed that inhibition of the anterior prefrontal cortex (aPFC) by cathodal tDCS lead to significant within-subject increase of deceptive behaviour.

As we see, the data is quite unsystematic, incomplete and tDSC-biased, which makes it difficult to generate clear-cut hypotheses for a TMS-based study of spontaneous lying. Even data from the correlational fMRI research does not help much. For example, well-rehearsed lies that fit into a coherent story elicit more activation in the right anterior frontal cortices than do spontaneous lies that do not fit into a story [1]. However, both types of lies were accompanied by excitation of the anterior prefrontal cortices (bilaterally) along with excitation in some other cerebral locations when compared to truthfulness. We also can see from the few previous TMS studies [24], [27] that the bilaterality of the effects is not always clear. Further investigations in this field seem necessary, additionally motivating our present work.

We can expect that same mechanisms that have been found to be involved in lying/guilty-knowledge in correlational studies may be involved also in changes in the propensity to lie or tell the truth as a result of brain stimulation. Frontal cortical structures, including DLPFC have been found relevant [4], [28]. Furthermore, DLPFC is considered as an area important in modulating the competitive weights of alternative goal-relevant representations [29], which is an essential part of what people experience when engaged in a conflict between whether to lie or tell the truth. Therefore, we chose to stimulate DLPFC to see possible causal role of this structure or set of structures influenced by DLPFC stimulation in spontaneous truth-telling or lying in a situation with no “high stakes”, or “mock criminality”, or personally relevant concealed information implied or generated.

The aims of the present study were (i) to see whether the spontaneous propensity to telling truth or lie can be artificially changed (particularly by TMS applied on DLPFC) and if yes, (ii) is the effect hemispherically bilateral or unilateral. A simple hypothesis states that slow-paced rTMS applied on DLPFC changes the frequency of subsequent spontaneous truth-telling compared to rTMS applied on a control locus.