In the present study the principal question was: Are there significant differences in intertemporal decision-making scores between smokers and nonsmokers? If so, how are these differences biologically expressed? We employed a four-method multimodal image analysis on high dimensional 7 tesla MRI images examining functional connectivity, white matter microstructure and gray matter thickness in a group of cigarette smokers and nonsmokers. This is the first time a study such as this has been undertaken.
First, Smokers’ performance on the delay discounting task was significantly lower than nonsmokers (NS) implying that smokers exhibit aberrant inter temporal choices and lower response inhibition than NS. This finding is in line with a meta-analysis by Amlung et al, (2017) who showed that DD is strongly correlated with continuous measures of addiction severity and quantity-frequency in several drugs of abuse. The delay discounting AUC task also correlated with the number of cigarettes smoked.
Seed-based connectivity analysis revealed significantly increased functional coupling within the nodes of the Right FPN: [R-PPC and R-RPFC and Right PPC and L-RPFC] and also between the SN and right fronto-parietal networks in smokers. This finding is similar to Clewett et al, (2014) study who showed increased functional coupling between the Left FPN and Salience networks to be a predictor of performance on the delay discounting test. Another study found enhanced mPFC-left fronto-parietal coupling in smokers suggestive of cue exposure may increase communication between cue-evaluative and action planning brain regions (Janes et al, 2012). While the fronto parietal network is anatomically similar in smokers and non-smokers, amplitude within this network is enhanced in smokers. Resting state amplitude is thought to reflect neuronal activity, whereby an increase in amplitude may indicate greater local neuronal activity (Yang et al., 2007). The relationship between resting state amplitude and neuronal activity suggests that, compared to non-smokers, smokers may have greater intrinsic neuronal activity in the fronto parietal and salience network. This enhancement is consistent with preclinical studies that show nicotine inducing a persistent increase in the baseline sensitivity of brain reward systems (Kenny and Markou, 2006; Kenny et al, 2008).
Functional Connectivity and Delay discounting: Pearson’s correlations of subjects scores on DD tasks area under curve (AUC) and SN-FPN coupling showed that non-smokers were significantly correlated with the connectivity strength between SN-FPN (Figs. 1 & 3). This indicates that the greater the coupling strength, the more robust the cognitive control and the less impulsivity in making intertemporal choices in nonsmokers. Thus, stronger coupling predicted scores on the DD in nonsmokers. This key association was non-significant in smokers (Fig. 2) and reversed (Fig. 4). – a key difference that could plausibly be the reason for sustained use and abuse.
This finding is in line with several studies that have shown that cognitive control plays an important role in inter temporal DM (Luo et al, 2009, Figner et al, 2010). Stronger between and within network coupling in the SN-FPN lead to enhanced cognitive control which in turn lead to greater impulse control.
Functional connectivity and Age at first cigarette: Chen et al (2018) showed that stronger SN-FPN coupling mediates the higher discounting which in turn leads to cigarette smoking. This is shown in the present study (see graphs 6, 7 and 8). Three separate networks showed higher functional coupling being inversely correlated with age at first cigarette initiation. The implication of this is that the earlier the exposure to cigarettes, the higher the functional coupling in these networks’ indicative perhaps of a greater difficulty in quitting.
Functional connectivity and years of cigarette smoking: Fig. 5 showed that the longer one smoked tobacco cigarettes, the stronger the functional coupling between the R -FP and the salience networks. The strengthening of this association over time implies that quitting smoking may be harder for those who have been smoking for many years.
Functional connectivity and difficulty quitting (DSM- V): In line with the above statement, the present study showed the stronger the functional coupling between R-FP and Salience, the harder it is to quit smoking as measured by the DSM scale on quitting (Fig. 9).
DD correlated with DSM criteria for nicotine tolerance and with the number of cigarettes smoked indicating that inter temporal choices play a crucial role in both these aspects of nicotine addiction.
Examining the white matter correlates of this DD deficit, DTI analysis showed that there were no significant differences between the groups in fractional anisotropy, radial, axial and mean diffusivities. This is indicative of fairly intact structural connectivity in the WM of smokers. A review of WM integrity in young tobacco users (Gogliettino et al, 2016) showed FA increases in corpus callosum (genu, body and splenium), internal capsule and superior longitudinal fasciculus. The reason we did not find this difference may be due to a limitation of the FSL DTI method of FWE correction that was used. Examining whole brain WM intracellular neurite orientation and dispersion (NODDI), no significant differences were observed between the groups. This indicates that even in the multi compartment model of WM integrity, the effects of tobacco may be less pronounced. This may again be a limitation of the multiple comparison correction methods that we used. Another possible explanation for the lack of a difference in WM integrity between the groups in both models may be the degree of chronicity. Several studies have shown that subjects with chronic cigarette smoking habits having WM alterations (Baeza-Loya et al, 2016; Hudkins et al, 2010). Subjects in the present study had an average 7 years of use and incipient WM changes may be fallen below the multiple comparisons threshold.
The cortical thickness differences between the groups in the hubs of the salience network and FPN were also not significant. This is in line with Morales et al (2014) who did not find any differences in the thickness of the insula. Thickness deficits were however found in rodents (Zhu et al, 2012) and only one study found CT deficits in the medial orbitofrontal cortex (MOFC) in humans (Kuhn et al, 2010). The MOFC was not part of the regions that we investigated. Again as with WM, the reason for this lack of a difference could be chronicity of cigarette use.
Finally, the total number of cigarettes smoked was strongly correlated with the FTND score (which is a scale denoting nicotine dependency) implying that that the level of nicotine addiction was proportional to the number of cigarettes smoked.
In sum, tobacco cigarettes appear to affect cognition (inter temporal choices) and dynamic networks mediating cognition as shown by the present study. The underlying white matter appears to be largely spared but this may be due to shorter exposure to the effects of tobacco smoke. These key associations imply that higher discounting and greater functional coupling in smokers may be alterations that arise as independent developments lead to cigarette initiation and smoking dependence.