PropertyValue
is nif:broaderContext of
nif:broaderContext
is schema:hasPart of
schema:isPartOf
nif:isString
  • RRMS patients conforming to the McDonald diagnostic criteria [33], [34] were contacted for the study. Inclusion criteria were: (1) mild to moderate neurological disability, but with unimpaired ambulation (Expanded Disability Status Scale [EDSS] ranging from 1.5 to 3.5) [35]; (2) no clinical relapse and no corticosteroid therapy for at least six weeks before inclusion in the study; (3) no diagnosis of major depression, alcohol or drug abuse or other psychiatric disorders according to the DSM-IV criteria. Seventy-two patients, aged between 18 and 48 years old, were analyzed. Thirty-nine patients were treated with disease-modifying therapies (interferon β-1a or 1b in most cases) for a mean period of 3.3 years (ranging from two months to nine years), ten were taking antidepressants for minor mood symptoms which were not severe enough to fulfill the diagnostic criteria for depression as assessed in a psychiatric interview, and five had had symptomatic treatment for fatigue (amantadine or modafinil). The control group consisted of 38 healthy volunteers matched for age, gender and education, with no history of alcohol or drug abuse, major psychiatric disorders (major depression, psychosis, untreated bipolar disorders), head trauma, other neurological disorders, or systemic illness. Standard Protocol Approvals, Registrations and Patient Consents: The study was approved by the local university’s Ethics Committee, and all subjects gave written informed consent for their participation in accordance with the Declaration of Helsinki. The Name of the Ethic Committee is the following: “Commission d’Ethique de la recherche clinique, Faculté de médecine et de biologie, Université de Lausanne, Switzerland”. The name of the accepted project (2007) was: “Dissecting the decisional process in patients with Multiple Sclerosis”. In order to maintain motivation, participants were informed that they would receive money as a function of their final gains in the DM tasks. For ethical reasons, in actual fact they all received 20 Swiss Francs at the end of the study. All participants underwent a neuropsychological examination, to test whether DM deficits were associated with specific pattern of cognitive deficits. The Brief Repeatable Battery of Neuropsychological Tests (BRB-N) [36], [37] was used to assess verbal memory (Selective Reminding Test [SRT]), spatial memory (10/36 Spatial Recall Test), sustained attention/information processing speed (3-second version of the Paced Auditory Serial Addition Test [PASAT], Symbol Digit Modalities Test [SDMT]), and verbal fluency in semantic cues (Word List Generation [WLG]). In addition to the BRB-N, we administered a task which assessed more complex executive skills (Stockings of Cambridge [SOC] from the Cambridge Neuropsychological Test Automated Battery [CANTAB]) [38]. At the end of the second testing session, subjects filled out questionnaires assessing mood (Hospital Anxiety and Depression scale, HAD) [39] and behavioral changes (Dysexecutive Questionnaire, DEX) [40]. The DEX was used for quantifying behavioral disturbances commonly associated with executive impairment and has already been used in MS studies [41]. This last questionnaire was aimed at investigating whether specific behavioural changes/symptoms in MS are associated with eventual DM changes. The 20 items of the DEX encompass broad areas of likely changes (impulsivity, apathy, desinhibition, intentionality, etc.). Each item is scored on a 5-point scale ranging from “never” to “very often” (0 to 4). The Wheels of Fortune (WOF) task [23] was used to measure the emotional strategies associated with DM under risk. This task had previously shown that advantageous decisions in healthy controls could be induced through prior experiencing disappointment and regret (Fig. 1A) [23]. Figure data removed from full text. Figure identifier and caption: 10.1371/journal.pone.0050718.g001 Description of the two gambling tasks.A) Wheel of Fortune (WOF): the subject (with SCR recorded) has to choose between two possible lotteries with different risk and earning possibilities. The subject was asked to choose one of the two wheels by pressing a left or a right button (choice period). A spinning arrow then appeared at the center of the wheel, spins for a variable duration (wait period), and stopped revealing the outcome(s) (feedback period). In the partial feedback condition (30 trials) the subject does not know the outcome of the other lottery; in the complete feedback condition (30 trials) both outcomes are presented. At the end of each trial, subjects had to indicate their affective state using a rating scale. B) Cambridge gambling task (CGT): the subject has to bet points on a choice associated with a given level of risk. Trials are run in blocks (two sets of four blocks), each containing nine trials. A row of ten boxes (red or blue, with a ratio varying across trials) is presented at the top of the screen. Participants are told that a yellow token was hidden in one of the boxes. They then have to guess whether it is in a red or blue box (color). Then they decide how many of their points they wanted to gamble on their choice (point choice) by pressing when they choose: available bets (5, 25, 75, 95 and total) are presented on the right of the screen in a ascending or descending sequence. Then feedback is given about gain or loss and total ongoing fortune (left). Two wheels were presented on a computer screen (Gamble 1 and Gamble 2). Each wheel was divided into two sectors (black and light blue) associated with different amounts of money. The size of each sector represented the probability of obtaining the proposed outcome. The possible outcomes for each individual gamble were visible on the screen and were formed by any pair of the following values: −50, +50, −200, +200 (units corresponding to cents in Swiss francs), and associated with different outcome probabilities (0.8, 0.2, and 0.5). The subject was asked to choose one of the two wheels by pressing a left or a right button (choice period). A spinning arrow then appeared at the center of the wheel, turned for a variable duration (wait period), and stopped in one of the two sectors, revealing the outcome which resulted in a financial gain or loss (feedback period). At the end of each trial, subjects had to indicate their affective state using a rating scale ranging from −50 (extremely sad) to +50 (extremely happy). The task included two conditions given in separate blocks. In the “partial feedback” condition (PF; 30 trials), the spinning arrow and the related outcome were apparent for the selected wheel only. In this condition, the unfavorable comparison of the obtained outcome with a more favorable counterfactual (i.e., unobtained) outcome may have induced disappointment related to the financial consequence of a decision [23]. By contrast, in the “complete feedback” condition (CF; 30 trials), spinning arrows appeared, rotated and stopped in both the selected and the non-selected wheels, revealing both outcomes to the participants. CF trials induced not only disappointment but also regret, by showing the outcome that would have been obtained if participants had selected the other gamble (counterfactual outcome). Skin conductance responses (SCRs) were recorded during the WOF task using the PowerLab/GSR amplifier system (AD Instruments GmbH, Spechbach, Germany). This was done in order to obtain the physiological correlate of regret, which consisted of an increase of SCRs in the CF condition due to counterfactual processing, as compared to the PF condition of the WOF. The SCRs represent an indicator of sympathetic nervous system activation, which is believed to contribute to the process of making advantageous choices for the organism. SCR data were acquired continuously using flat-surface electrodes placed on the non-dominant hand and stored for off-line analysis using a second computer running Chart v4.2 software. The SCRs of interest were those generated during the five-second interval following the viewing of the obtained outcome (feedback period, post-choice SCRs). Sixty five-second time points were recorded for each subject. Artefactual signals (e.g., the subject’s movements) were cleared manually. Mean amplitudes (microSiemens, µS) recorded during each five-second time window were analyzed (i.e., the mean of the data points obtained in the five-second selection). Baseline SCR activity was assessed using three measurements per subject recorded: (1) at rest; (2) in response to a loud noise; and (3) after a deep breath. This method had been applied in a previous study [11] and had been shown to be reliable. The Cambridge Gamble Task (CGT) [21] was used to measure DM (Fig. 1B). Participants sat in front of a computer touch screen. Trials were run in blocks (two sets of four blocks per subject), each containing nine trials. A block could finish prematurely if it ended in bankruptcy. At the beginning of each block, they were given 100 points. A row of ten boxes (red or blue, with a ratio varying across the trials) was presented at the top of the screen and participants were told that a yellow token was hidden in one of the boxes. They then had to guess whether it was in a red or blue box and to decide how many of their points they wanted to gamble on their choice (5%, 25%, 50%, 75% or 95%, given in a progressively ascending or descending order depending on the blocks). A winning choice was rewarded by the total of points gambled, whereas a losing choice was punished by subtracting that number of points. The probability of each choice being correct was indicated to the subjects by the ratio between red and blue boxes. This produced a variety of situations, ranging from those in which one outcome was the most likely (e.g., nine red boxes to one blue box) to those in which both outcomes were almost equally likely (six to four). We used the standard version provided by CANTAB (http://www.cantab.com/cantab-tests-cambridge-gambling-task.asp) with a 5∶5 ratio of red-blue boxes also included in the design to ensure that participants perceived the task as a random trial sequence. Therefore, it was hypothesized that the CGT allowed participants to apply cognitive strategies, such as probabilistic judgment, in order to decide advantageously. The ascending and descending sequence in which potential wins were proposed enabled us to differentiate patients with impulsive response tendencies from patients with real risk preference (i.e., risk-preferent patients had to wait if they wanted to place high bets in the ascending conditions). Measures of behavior choices were quality of DM (the proportion of trials where participants chose the more likely outcome) and deliberation time. Statistical analyses were conducted using a STATA software package (Version 10.0). Non-parametric tests were applied to all the demographic and behavioral data because (1) the CGT and WOF data could not be transformed successfully to reach normality, and (2) the distribution of the WOF emotional ratings is by definition non-normal as the scale is restricted (values available from −50 to +50) and thus cannot be considered as a continuous variable. Differences were examined using Chi2 tests for the comparison of categorical variables and Wilcoxon signed rank tests for the comparison of continuous variables. To measure the possible influences of executive, attentional and emotional variables on decisional performances, we then computed single Spearman correlation analyses with certain outcome measures in the following executive tasks: SOC, SDMT, and PASAT. For the WOF task, we first analyzed the emotional evaluations of decisional outcomes and then tested two models of choice computed by regression analyses, using a panel logit procedure with an individual random effect. The panel data analysis took each subject as the unit and the trial as the time. The random effects model was used as the default model, and the parameters were estimated by maximum likelihood. This statistical procedure is extensively described elsewhere [23], [26]. The first model integrated the effects of anticipating disappointment (d, negative emotion expressed in the PF condition when facing a negative outcome) and regret (r, negative emotion induced by the counterfactual comparison of a negative obtained outcome with a more advantageous outcome for the rejected alternative) in addition to the maximization of expected values (EV, choices of the most favorable odds, reflecting quality of DM and risk taking). In fact, as we have already mentioned, a subject obtaining an outcome lower than expected might experience disappointment or regret. The greater the difference between the expected and the obtained outcome, the more intense this negative feeling will be. Thus, to avoid future disappointment or regret, the subject might progressively adapt his or her choice behavior by selecting a gamble that minimizes the difference between the lowest and highest outcomes, weighted by the probability of the worst possible outcome. This is what we call anticipation of disappointment and regret. In the second model [42], we used the logit regression to estimate the probability of the participant choosing the first gamble, as a function of the difference in EV (dEV) and standard deviation (a measure of risk) between the first and second lottery: The variables dEV and risk are defined as follows:where x1, y1 and x2, y2 are the two possible outcomes of the first and the second lotteries respectively, with x1>y1, and x2>y2. The probability of x1 is p and the probability of y1 is (1–p). The probability of x2 is q and the probability of y2 is (1–q). Thus a positive (negative) and significant dEV or risk coefficient indicates that subjects consistently choose the lottery with the highest (lowest) expected value or level of risk respectively. For the CGT, choice behavior and betting behavior were analyzed separately [30] on the basis of the scores calculated through the CGT program. It should be noted that trials with a 5∶5 ratio of red-blue boxes, included in the design to mimic a random trial sequence, were excluded from the statistical analyses. The analysis of the participant’s betting behavior was also limited to the trials in which the subjects chose the most likely color, in order to maintain independence from choice behavior. The patients’ and controls’ scores were directly compared using Mann-Whitney non parametric tests.
rdf:type