ABSTRACT
Objectives We aimed to determine whether chronic nicotine use, alcohol consumption, and gambling alters brain glucose metabolism.
Methods We retrospectively analyzed data from 473 healthy men who participated in health checkups at Samsung Changwon Hospital Health Promotion Center during 2013 (baseline) and 2018 (follow-up). The health checks included a brain 18F-fluorodeoxyglucose positron emission tomography (PET), a questionnaire of tobacco use, the Alcohol Use Disorders Identification Test (AUDIT; Korean version), and the Problem Gambling Severity Index (PGSI). From brain PET scans, the mean uptake in regions-of-interest was scaled to the mean global cortical uptake by each individual, defining the standardized uptake value ratio. We established a model for tobacco use, AUDIT, and PGSI with regional SUVR as a dependent variable and tobacco use, AUDIT, and PGSI as predictors adjusted for age using Bayesian hierarchical modelling. Bayesian models were estimated using four Markov chains, each of which had 4,000 iterations including 1,000 warm-ups, thus totaling 12,000 post-warmup samples. The sampling parameters were slightly modified to facilitate convergence (max tree depth = 20). All data were analyzed using R (The R Foundation for Statistical Computing, Vienna, Austria).
Results This study included 131 healthy males (mean age at baseline and follow-up: 43.0 ± 3.4, 48.1 ± 3.3 years, respectively). Tobacco use was negatively associated with glucose metabolism in the caudate, thalamus, cingulate, and frontal lobe, and positively associated with the cerebellum, whereas AUDIT or PGSI were not associated.
Conclusion Tobacco use was associated with altered brain glucose metabolism in the caudate, thalamus, cingulate, frontal lobe, and the cerebellum. However, neither hazardous alcohol consumption, nor problem gambling showed any association with brain glucose metabolism. Our findings might provide new insights into the neural mechanisms of chronic nicotine use.
INTRODUCTION
The global number of smokers worldwide during 2019 was 1.14 billion1. The agestandardized prevalence of current tobacco smokers was 32.7% among men aged over 15 years and 6.62% among women1. Although the prevalence of smoking had decreased significantly since 1990, the total number of smokers has increased1. Tobacco smoking is responsible for the premature deaths of approximately 6 million people annually worldwide 2. Most smoking-related deaths arise from cancers, respiratory, and cardiovascular diseases2. Smoking is a risk factor for stroke, blindness, deafness, back pain, osteoporosis, and peripheral vascular disease2.
Cigarette addiction is derived from nicotine in tobacco2 that enters the bloodstream through the lungs and reaches the brain across the blood-brain barrier3. This results in elevated blood pressure, glucose release, increased respiration, heart rates, and alertness, as well as constricted arteries3.
Acetylcholine plays a role in mental and physical arousal, learning, memory, and emotion and binds to n-acetylcholine receptors (nAChRs) that are classified as nicotine and muscarine types3. Nicotine also binds to nAChRs that are localized throughout the brain and peripheral nervous system including the midbrain, striatum, nucleus accumbens (NAc), ventral tegmentum, muscle, adrenal glands, and heart4. The gratifying action of nicotine is caused by the activation of dopaminergic neurons in the ventral tegmental area that projects to the NAc and prefrontal cortex4. Nicotine administration increases dopaminergic neuron firing in striatal and extrastriatal areas resulting in elevated dopamine levels4. The effects of smoking on the brain have been investigated in detail. The volumes of grey and white matter are decreased in magnetic resonance imaging (MRI) scans of smokers compared with non-smokers5-7. According to a meta-analysis, activation of the striatum and anterior cingulate gyrus increases in response to smoking-related vs. neutral cues in smokers8. Also, connectivity is lower in the orbitofrontal cortex, superior frontal gyrus, temporal lobe and insula in smokers compared with non-smokers9. However, the chronic effects of nicotine on brain glucose metabolism have not been evaluated. Thus, we aimed to identify the effects of chronic nicotine inhalation measured as pack-years on brain glucose metabolism. We also evaluated correlations between brain glucose metabolism and alcohol consumption as well as gambling, since these activities decrease grey matter volume10,11.
METHODS
Participants
We retrospectively analyzed data from 473 healthy men who participated in health check-ups at Samsung Changwon Hospital Health Promotion Center during 2013 (baseline) and 2018 (follow-up). We excluded individuals with neuropsychiatric disorders (n = 5), malignancies (n = 3), as well as missing tobacco use, Alcohol Use Disorders Identification Test (AUDIT; Korean version), and Problem Gambling Severity Index (PGSI) data (n = 302). We finally analyzed data from 131 healthy men. The health checks included a brain 18F-FDG positron emission tomography (PET), questionnaires about tobacco consumption, AUDIT, and PGSI. All 131 men had been included in a previous study of the effects of aging on brain glucose metabolism12. The Institutional Review Board of Changwon Samsung Hospital approved the study protocol and the need for informed consent to participate was waived due to the retrospective study design.
Brain 18F-FDG PET and Image analysis
The participants avoided strenuous exercise for 24 hours and fasted for at least 6 hours before starting the PET study. The PET/computed tomography (CT) images were acquired from 60 minutes after injecting the participants with18F-FDG (3.7 MBq/kg) using a Discovery 710 PET/CT scanner (GE Healthcare, Waukesha, WI, USA). Continuous spiral CT images were acquired under the parameters of tube voltage and current of 120 kVp and 30-180 mAs, respectively. The PET images were acquired in 3-dimensional mode with an FWHM of 5.6 mm and reconstructed using an ordered-subset expectation maximization (OSEM) algorithm. The images were spatially normalized to Montreal Neurological Institute (MNI) space using PET templates from Statistical Parametric Mapping 5 (SPM 5; (University College of London, UK) with the reference tool for PET kinetic modeling pmod v. 3.6 (PMOD Technologies LLC, Zurich, Switzerland). Regions of interest (ROIs) was defined using the Automated Anatomical Labeling 2 (AAL2) atlas13, which comprised the caudate, cerebellum, cingulate, frontal lobe, hippocampus, insula, occipital lobe, pallidum, parietal lobe, putamen, temporal lobe, and thalamus. The mean uptake of each ROI was scaled to the mean global cortical uptake of each individual, and defined as the standardized uptake value ratio (SUVR). We smoothed SUVR images with a Gaussian kernel of FWHM 8 mm using Statistical Parametric Mapping (SPM) 12 (Wellcome Centre for Human Neuroimaging, UCL, London, UK). The statistical threshold was set at the cluster level for full-volume analysis and corrected using a false discovery rate with p < 0.05 in a regression model (corrected for age).
Tobacco use, AUDIT, and PGSI
Tobacco use was analyzed in terms of pack-years by multiplying the number of packs of cigarettes smoked daily by the number of years. The participants were screened using AUDIT for hazardous consumption or any alcohol use disorder. It consists of 10 questions that are each scored from 0 to 4 with higher scores indicating increased risk of alcohol use disorder. Scores from 0□7, 8□14, and ≥ 15 respectively indicate low-risk, hazardous or harmful alcohol consumption, and a likelihood of alcohol dependence14. PGSI is a Korean version of Canadian Problem Gambling Index15 that screens for gambling problems. It consists of nine questions that are each scored from 0□3, with higher scores indicating increased risk of problem gambling. Scores of 0, 1□2, 3□7, and ≥8 indicate a gambler without a problem, and low-risk, moderate-risk and problem gamblers, respectively.
Statistical analysis
Normality was assessed using Shapiro–Wilk tests. After logarithmic transformation of the regional SUVR, we analyzed the effects of chronic tobacco use, alcohol consumption (AUDIT), and gambling (PGSI) on regional SUVRs using Bayesian hierarchical modelling with brms16-18 that applies Markov-Chain Monte Carlo sampling using the rstan package in R19. We established a model for tobacco use, AUDIT, and PGSI with regional SUVR as a dependent variable and tobacco use, AUDIT, and PGSI as predictors adjusted for age. These fixed effects were individually calculated and participants and ROIs were added as random intercepts to allow the SUVR to vary between them. Bayesian models were estimated using four Markov chains, each of which had 4,000 iterations including 1,000 warm-ups, thus totaling 12,000 post-warmup samples. The sampling parameters were slightly modified to facilitate convergence (max tree depth = 20). All data were analyzed using R (The R Foundation for Statistical Computing, Vienna, Austria).
RESULTS
Characteristics of participants
This study included 131 healthy men (mean age at baseline and follow-up: 43.0 ± 3.4, 48.1 ± 3.3 years, respectively). The mean and standard deviations of tobacco use (pack-years), AUDIT, and PGSI were 15.5 ± 11.4, 11.4 ± 6.1, 0.3 ± 0.9 at baseline, 15.1 ± 10.6, 10.4 ± 6.1, 0.2 ± 1.1 at follow-up.
Tobacco use, AUDIT, PGSI, and Brain glucose metabolism
Tobacco use was negatively associated with glucose metabolism in the caudate, thalamus, cingulate, and frontal lobe, and positively with that in the cerebellum at baseline and follow-up (Figure 1). Full-volume analysis consistently revealed a negative association between tobacco use and glucose metabolism in the caudate, thalamus, cingulate, and frontal lobe, and a positive association with the cerebellum (Figure 2). However, Bayesian models and full-volume analyses did not associate brain glucose metabolism with either AUDIT or PGSI.
DISCUSSION
Our main finding was that tobacco use was associated with altered brain glucose metabolism in the caudate, thalamus, cingulate, frontal lobe, and the cerebellum. However, neither hazardous alcohol consumption, nor problem gambling showed any association with brain glucose metabolism. These effects are also consistent over the studied period of 5 years in the middle adulthood.
The caudate nucleus plays a role in the establishment and expression of reward-related conditioned behavior20, and caudate reactivity to smoking cues is associated with a reward response 20. An increased caudate volume in young male smokers has been identified using MRI21, whereas others found that young male smokers had a smaller caudate volume22. The authors speculated that the difference might be due to nicotine dependence and the consequent accumulation of neurotoxic effects22. The thalamus expresses abundant nAChR23. Acute administration of nicotine activates the thalamus, and chronic nicotine use leads to decreased nAChR in the thalamus 24. The thalamus plays a role in drug addiction via a locomotor response to psychostimulants, cue-reward association, the expression of drug-induced place preference, withdrawal, and the reinstatement of drug-seeking25. Individuals who are dependent on alcohol, cocaine, nicotine, methamphetamine, opioids, cannabis, and synthetic cannabinoids have lower thalamic volumes26. Thalamic connectivity is also decreased in those who are dependent on alcohol, cocaine, and nicotine26. A meta-analysis of chronic cigarette smoking found a decreased volume of grey matter in the thalamus27. The cingulate cortex expresses nAChR and is associated with nicotine28,29. The cingulate gyrus of smokers becomes activated in response to smoking, rather than neutral cues30. The cingulate cortex volume is smaller among smokers compared with nonsmokers5,7,31,32. A decreased cingulate cortex volume is also associated with alcohol, methamphetamine, and cocaine dependance33. An increased severity of nicotine addiction is associated with weaker functional connectivity between the cingulate cortex and striatum28. Nicotine also affects the frontal cortex. Grey matter volume and frontal cortex density are decreased in smokers5,32. The volume of the frontal cortex negatively correlates with pack-years 6. Nicotine is associated with disturbed working memory and attention that correlate with the prefrontal cortex, the activity of which is reduced34,35. The cerebellum also functions in addiction36. The cerebellum correlates with functional resting state networks including the prefrontal cortex, cingulate, and insula that are associated with cognitive control, stimulus salience, and interoception, and shares reciprocal connections with the dopaminergic system in basal ganglia36. Addictive cocaine, opioid, nicotine, and alcohol are associated with decreased grey matter volume36. Smokers have smaller cerebellar grey matter volumes, reduced grey matter density5,7,37,38, and functional connectivity38. The neurotoxic effect of nicotine use might be responsible for the decreased volume, density, and connectivity in grey matter 27,37. These effects might appear as negative correlations between pack-years and brain glucose metabolism measured by 18F-FDG PET in the caudate, thalamus, cingulate, and frontal lobe. However, the correlation between pack-years and brain glucose metabolism in cerebellum was positive in the present study. The cerebellum has nAChRs that are sensitive to nicotine, which increases both glutamatergic GABAergic transmission by binding to nAChRs, but the former is superior to the latter39. Chronic nicotine use increases glucose oxidation and neurotransmitter cycling by glutamatergic neurons in the cortex, subcortex and cerebellum40. However, GABAergic metabolism increases in the cortex and but not in the subcortical and cerebellum regions40. Thus, chronic nicotine use increases excitatory glutamatergic transmission in the cerebellum39. Furthermore, cerebellar glucose metabolism positively correlates with age in healthy persons41,42. Age-related degeneration of the cerebellum is assumed to start between the ages of 50 and 60 years, and its age-related volume loss is lower than that of the cerebrum43. The aging effect on the cerebellum and increased glucose oxidation due to chronic nicotine use might explain the positive correlation regardless of a decreased cerebellar volume. Further investigation is needed to validate this.
The effects of alcohol on the brain are established. Alcohol consumption negatively correlates with total cerebral brain and grey matter volumes10,44, even in middle aged individuals45. The findings of 18F-FDG PET imaging associated alcohol use disorder with decreased glucose metabolism in the frontal, temporal, and parietal lobes compared with age- and sex-matched healthy persons46. The present study did not identify a correlation between AUDIT scores and brain glucose metabolism. However, AUDIT scores do not include the duration of alcohol consumption, and high scores do not necessarily indicate alcohol use disorder. Thus, despite a higher AUDIT score, changes in the brain might not be eminent. Behavioral addictions, including pathological gambling, decrease grey matter volumes compared with healthy controls11. Relative glucose metabolic rates in the orbitofrontal and medial frontal cortex are significantly increased compared with healthy controls47.
The present study did not identify a correlation between the PGSI score and brain glucose metabolism.
The PGSI score of most participants in this study was 0, which might be responsible for the absence of a correlation between PGSI scores and brain glucose metabolism.
This retrospective study has several limitations. Because we included only men, our results might not be directly generalized to women. The study was based on a health check program that did not include brain MRI. Thus, MRI-based coregistration and partial volume correction of PET images were impossible and the results could not be compared with MRI-based indices of atrophy.
CONCLUSION
Tobacco use was associated with altered brain glucose metabolism in the caudate, thalamus, cingulate, frontal lobe, and the cerebellum. However, neither hazardous alcohol consumption, nor problem gambling showed any association with brain glucose metabolism. Our findings might provide new insights into the neural mechanisms of chronic nicotine use.
Data Availability
All data produced in the present study are available upon reasonable request to the authors
TABLES
Footnotes
Seunghyeon Shin (1st author), forladou{at}naver.com
Keunyoung Kim, nmpnuh{at}gmail.com
Jihyun Kim, march4900{at}hanmail.net
DATA AVAILABILITY The data that support the findings of this study are available from the corresponding author upon request.
FUNDING No
CONFLICT OF INTEREST No conflict of interest
Ethics approval The Institutional Review Board approved the study protocol.
Patient consent statement The IRB waived the need for informed consent to participate due to the retrospective study design.
Permission to reproduce material from other sources No materials reproduced from other sources
Clinical trial registration Not available (retrospective study)