Behavioral and Brain Functions

Cortico-limbic brain activity associated with anger may be susceptible to nicotine and, thus, may contribute to smoking initiation and nicotine addiction. The purpose of the study was to identify the brain regions that are most reactive to nicotine and show the greatest association with anger task performance. Twenty adult nonsmokers (9 women, 11 men) participated in two laboratory sessions to assess brain metabolism with fluoro deoxy-glucose Positron Emission Topography (FDG-PET) in response to nicotine and placebo patches during an anger provocation task. Outcome variables for the anger provocation task were reaction time, intensity and length of retaliation. Reaction time was associated with nicotine-induced changes in the left thalamus. Length of retaliation was associated with a functionally linked set of cortical and subcortical structures such as right frontal lobe, right anterior cingulate (BA 24), right uncus, left parietal lobe, left BA 11, left cingulate, left BA 25, left amygdala, left BA 30, left BA 38 and BA 9. These findings reveal the underlying brain circuitry targeted by nicotine during anger provocation. Findings Cigarette smoking is a major public health problem, and angry dispositions have been identified as an important risk factor for smoking initiation and subsequent nicotine dependence [1-4]. In particular, behavioral studies revealed a smoking-anger linkage [5-8], but little is known about the underlying brain circuitry subserving such linkage. Dysfunctional cortico-limbic brain activity associated with anger may be normalized by nicotine [4]. Functional imaging studies on regional brain activity in response to nicotine and smoking have documented the involvement of cortico-limbic structures such as the prefrontal cortex (including the dorsolateral prefrontal cortex, inferior frontal, medial frontal, and orbitofrontal gyri) [9-15], cingulate [12-16], thalamus [11,13,15,17-19], and amygdala [12,13,15,19]. Similarly, anger has been associated with activation of the medial prefrontal cortex, orbitofrontal cortex, and anterior cingulate cortex [20-22]. However, little is known about which brain areas are most responsive to the effects of nicotine during anger provocation. The Published: 24 April 2009 Behavioral and Brain Functions 2009, 5:19 doi:10.1186/1744-9081-5-19 Received: 12 November 2008 Accepted: 24 April 2009 This article is available from: http://www.behavioralandbrainfunctions.com/content/5/1/19 © 2009 Gehricke et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


Findings
Cigarette smoking is a major public health problem, and angry dispositions have been identified as an important risk factor for smoking initiation and subsequent nicotine dependence [1][2][3][4]. In particular, behavioral studies revealed a smoking-anger linkage [5][6][7][8], but little is known about the underlying brain circuitry subserving such linkage.
The cortico-limbic circuitry is part of the "prefrontal system", which was first described by Rosvold and Schwatzbart [23] (see also review by Divac [24]) and may be particularly sensitive to nicotine modulation. The objective of the present study was to identify the brain regions within the prefrontal system that are most reactive to the effects of nicotine and show the greatest association with anger task performance in nonsmokers. Nonsmokers were studied to determine the effects of a fixed low-dose of nicotine via patch without the confounding effects of smoking and nicotine withdrawal. It was hypothesized that nicotine-induced anger task performance is associated with nicotine-induced changes in the medial prefrontal cortex, orbitofrontal cortex, and anterior cingulate cortex.
Twenty adult nonsmokers (9 women, 11 men), with a mean age of 24 (SD = 6), were recruited from the community via ads and flyers. All participants were physically and mentally healthy based on medical history obtained by a physician (JM). Similar to previous research [25], nonsmokers were defined by self-reported smoking abstinence of at least 2 years prior to participation in the study and not having smoked more than 100 cigarettes in their lifetime. All participants provided written informed consent approved by the Institutional Review Board of the University of California, Irvine.
Each participant completed two laboratory sessions in randomized order: one with a 3.5 mg low-dose transdermal nicotine patch (Nicoderm; GlaxoSmithKline) and the other with a look-alike placebo patch. A 3.5 mg nicotine patch (ie, 7 mg patch cut in half) was used to minimize potential side-effects in nonsmokers. The two laboratory sessions examined the acute effects of nicotine compared to placebo on brain metabolism using fluoro deoxy-glucose Positron Emission Topography (FDG-PET) during an anger provocation task [26]. The anger provocation task was a computerized Competitive Reaction Time task (CRT) designed to induce and measure retaliatory tendencies. Each participant competed with an ostensible "opponent" (of the same sex) to respond to a visual stimulus (ie, red square) as quickly as possible. We enhanced the saliency of the "opponent" by incorporating a video clip of this person. The participant was asked to set the level of white noise (0-105 dB) she/he wanted the opponent to receive if the opponent lost. After the participant set the noise level, a yellow square appeared as a "get ready" cue, followed by the red square. The objective was to press a mouse button as fast as possible. A scale on the left side of the monitor indicated the noise level the opponent set for the participant to receive. Provocation was manipulated by increasing the noise level presumably set by the opponent across time. There were 25 trials, with white noise presented during approximately 50% of the trials. The noise blasts were delivered through headphones for a variable length of time (set by the participant from trial to trial). Output variables were average reaction time, intensity, and length of retaliation.
Regions of interest were identified and verified with the Talairach and Tournoux atlas [27] and included areas that are associated primarily with the prefrontal system [24]. These areas cut across prefrontal, temporal and limbic regions (see Table 1).
Each laboratory session lasted approximately 4.5 hours. First, the patch was worn for 3 hours. Then an intravenous injection of 5 mCi of FDG was given during the 30-minute CRT. Subsequently, the subject lay in the scanner for 1 hour (see Fallon et al. [25] for more details). Patch administration was verified by serum cotinine levels (see Table  2).
For the PET data, normalized metabolic averages per anatomical region per subject were used. The metabolic averages were rescaled in order to reduce inter-individual variability in brain metabolism, and the resulting absolute metabolic averages per anatomical region per subject were used. Rescaling was achieved by taking the mean of all the brain regions within a subject (ie, global mean) and scaling it to the canonical normal global cerebral blood flow (50 ml/min/dl). This is a common technique used when analyzing PET data and provides some protection against false positives from global cerebral blood flow changes under a treatment condition [25,28]. Preliminary analyses using t-tests revealed that BA 11 and right rectal gyrus metabolism was significantly reduced under nicotine compared to placebo patches, whereas BA 29 showed the opposite pattern (see Table 2). No other significant differences were found [see Additional file 1]. For the anger provocation task, the CRT outcome measures of reaction time, intensity and length of retaliation were averaged across the duration of the task. No significant mean differences were found in anger task performance between nicotine and placebo conditions, perhaps due to the low-dose nicotine patch (see Table 2). Group means, however, do not reflect individual differences in brain metabolism in response to nicotine. Therefore, stepwise linear regression models were used to examine delta scores for each outcome variable in order to contrast nicotine with placebo conditions (nicotine -placebo patch).
Stepwise linear regression models (SPSS 13.0) examined which areas of the prefrontal system under nicotine compared to placebo predicted CRT outcome variables in response to nicotine versus placebo. Dependent variables were the total scores on CRT outcome variables. Predictor variables were the regions of interest. The inclusion criterion used in each regression model was a probability value of 0.05. The exclusion criterion was a probability value of 0.06. To adjust for the number of brain regions, alpha was Bonferroni-corrected at 0.0008.  The results from the regression models revealed that participants who showed nicotine-induced changes in anger task performance also showed changes in brain metabolism. More specifically, nicotine-induced improvements in reaction time were associated with decreased brain metabolism in response to nicotine in the left thalamus. This area of the brain explained 56% of variance for reaction time (R = 0.75, R 2 = 0.56, F change (1, 16) = 20.59, p < 0.0001).
Nicotine-induced reductions in length of retaliation were associated with increased brain metabolism in response to nicotine in the left cingulate, left amygdala, right frontal lobe, left BA 30, and left BA 38. In particular, the amygdala has been shown to modulate prefrontal regional activity [29]. However, our subjects included both women and men so that some underlying amygdaloid contribution may have been diluted by mixed lateralization components in women and men [30,31]. Positive associations were found between nicotine-induced reductions in length of retaliation and nicotine-induced reduction in brain metabolism in the orbital-frontal areas (ie, right uncus, left BA 9, right anterior cingulate (BA 24), left BA 25, right BA 9, and left BA 11) and left parietal lobe. These predictors explained 100% of the variability in length of retaliation (R = 1.00, R 2 = 1.00, F change (1, 5) = 66.69, p < 0.0001). No significant associations were found for intensity of retaliation. See Table 3 for more details.
The present findings on nicotine-induced brain metabolism in nonsmokers emphasize the role of negative affect, particularly anger, in susceptibility to nicotine. Moreover, Mid-sagittal drawing of cortical-limbic brain circuitry associated with anger provocation Figure 1 Mid-sagittal drawing of cortical-limbic brain circuitry associated with anger provocation. Blue areas indicate decreased nicotine-induced brain metabolism associated with increased nicotine-induced task performance. Red areas indicate increased nicotine-induced brain metabolism associated with increased nicotine-induced task performance. Lines represent implied circuitry based on well-documented anatomical connectivity. F = frontal lobe; P = parietal lobe Length Retaliation the findings suggest that nicotine may critically regulate brain areas that are involved in the inhibition of negative emotions such as anger [43]. Deficits in anger management may be a risk factor for smoking initiation and nicotine addiction, which may highlight the importance of improving affect regulation as a potential smoking prevention and cessation strategy [4]. Novel neurotherapeutic and behavioral treatments (e.g., anger management training) that affect the cortical and limbic brain areas may aid smoking cessation efforts in anger provoking situations that increase withdrawal and tobacco cravings.