Supplementary Online Content

Supplementary Online Content Luijten M, Schellekens AF, Kühn S, Machielse MWJ, Sescousse G. Disruption of reward processing in addiction: an image-based meta-analysis of functional magnetic resonance imaging studies. JAMA Psych. doi:10.1001/jamapsychiatry.2016.3084 emethods. efigure 1. Reward Anticipation Meta-Analytic Results for the Different Subgroups. efigure 2. Reward Outcome Meta-Analytic Results for the Different Subgroups. efigure 3. Binarised Jack-Knife Analyses For Reward Anticipation (A) and Reward Outcome (B). efigure 4. Stability of Whole-Brain Meta-Analytic Results After Excluding the 2 Datasets From van Hell et al. efigure 5. Whole-Brain Meta-Analytic Results in the Ventral Striatum For Reward Anticipation After Excluding the 2 Datasets From van Hell et al. (2010). efigure 6. Stability of Whole-Brain Meta-Analytic Results After Excluding Studies With Partial Brain Coverage (n = 5). efigure 7. Funnel Plots Assessing Publication Bias of fmri Studies For Reward Anticipation (A) And Reward Outcome (B) in the Striatum. etable 1. Task Paradigms and Included Contrasts For Reward Anticipation and Reward Outcome Studies. etable 2. Number of Studies and Participants Involved in Specific Analyses etable 3. Reward Anticipation Meta-Analytic Results Addicted > Controls etable 4. Reward Anticipation Meta-Analytic Results Addicted < Controls etable 5. Reward Outcome Meta-Analytic Results Addicted > Controls etable 6. Reward Outcome Meta-Analytic Results Addicted < Controls ereferences This supplementary material has been provided by the authors to give readers additional information about their work.

Supplementary materials Psychiatric comorbidity In order to explore potential confounding effects of psychiatric co-morbidities on brain reward-processing, studies were categorized according to the presence of such co-morbidity among individuals included in the addicted groups. Several studies clearly excluded all cases with psychiatric co-morbidity, using well-validated structured clinical interviews. These were classified as not co-morbid. All other studies, including those that did not systematically assess psychiatric co-morbidity, were classified as co-morbid. There were no studies that specifically examined addicted individuals with psychiatric co-morbidity. Phase of addiction In order to explore potential confounding phase-dependent effects of addiction on brain reward-processing, studies were categorized according to the addiction phase of the individuals included in the addicted groups. Since reporting on the phase of addiction is often limited, studies could only be categorized into active substance use, initial or prolonged abstinence, or a mix of both. Most studies on substance dependence other than nicotine dependence did not report separately on smoking status of the participants, which was therefore not taken into account in categorizing these studies. Studies were thus categorized as follows: active addictive behavior (either gambling or substance use, until or during the day of scanning), initial abstinence (last use >24 hours before scanning, but not >4 weeks), prolonged abstinence (>3 months abstinence), or mixed sample (both active and abstinent users included). If no information on current addictive behavior was available, the sample under survey was considered mixed.

efigure 1. Reward anticipation meta-analytic results for the different sub-groups. Those Z-maps show the brain regions showing consistent responses across studies during reward anticipation for substance use disorder individuals and matched healthy controls (left) and gambling disorder individuals and matched healthy controls (right). Functional Z-maps are overlaid on the Colin 27 anatomical template and thresholded at p <.005 and k 10 (equivalent to a corrected p-value of.05, see methods).

efigure 2. Reward outcome meta-analytic results for the different sub-groups. Those Z-maps show the brain regions showing consistent responses across studies during reward outcome for substance use disorder individuals and matched healthy controls (left) and gambling disorder individuals and matched healthy controls (right). Functional Z-maps are overlaid on the Colin 27 anatomical template and thresholded at p <.005 and k 10 (equivalent to a corrected p-value of.05, see methods).

efigure 3. Binarised jack-knife analyses for reward anticipation (A) and reward outcome (B). Jack-knife analyses consist of systematic repetitions of the meta-analysis of interest after excluding one study at a time. The color bar represents the number of overlapping jackknife analyses at a threshold of p <.005 and k 10 (22 analyses in total for reward anticipation and 23 analyses in total for reward outcome). The results show that the striatal group differences observed during reward anticipation and outcome in the main analyses are observed in virtually all jack-knife analyses. This suggests that our results are highly replicable across many combinations of datasets, and that none of the studies is playing a disproportionate role and driving the results.

efigure S4. Stability of whole-brain meta-analytic results after excluding the two datasets from van Hell et al. (2010). Note that excluding the two datasets from van Hell et al. (2010) does not qualitatively affect the main results of our meta-analysis (see comparison with Figure 2 in main text). Note also that the results for gambling addicted individuals remain unchanged after excluding the two datasets from van Hell et al. (2010), since these datasets are substance addiction-related. Functional Z-maps are thresholded at p <.005 and k 10 (equivalent to a corrected p-value of.05).

efigure 5. Whole-brain meta-analytic results in the ventral striatum for reward anticipation after excluding the two datasets from van Hell et al. (2010). Excluding the two datasets from van Hell et al. (2010) revealed an increased response in the ventral striatum during reward anticipation in individuals with addictive behaviours compared with healthy controls. This seems to be due to these two datasets presenting relatively large effect sizes in the opposite direction in the ventral striatum, as can be seen from in the forest plot which reports mean variance of effect sizes for group differences estimated from individual studies (note that in contrast to the whole-brain results, effect sizes for the two datasets from van Hell et al. (2010) were incorporated in the forest plot in order to illustrate their impact on the ventral striatal results). Note however that this result is not observed separately in individuals with substance addiction, and observed only weakly and unilaterally in individuals with gambling addiction. In addition, a close examination of the functional clusters shows that the peaks are not located in the striatum itself, and that the activations observed in the ventral part of the striatum are in fact the tips of more ventral/posterior clusters spanning the dorsal amygdala and posterior orbitofrontal cortex. We thus think this is a potentially interesting but fragile result that will need further investigation in future meta-analyses, in order to disentangle whether it is a false positive or a real effect that only surfaces weakly in the present analysis because of e.g. methodological issues known to affect this region in individual studies (low sensitivity due to susceptibility artefacts, less accurate normalization ). Functional Z- maps are thresholded at p <.005 and k 10 (equivalent to a corrected p-value of.05).

efigure 6. Stability of whole-brain meta-analytic results after excluding studies with partial brain coverage (n = 5). Note that excluding these studies does not qualitatively affect the main results our metaanalysis (see comparison with Figure 2 in main text). Note also that the results for gambling addicted individuals remain unchanged after excluding partial brain coverage studies, since all the excluded studies are substance addiction studies. Functional Z-maps are thresholded at p <.005 and k 10 (equivalent to a corrected p-value of.05).

efigure S7. Funnel plots assessing publication bias of fmri studies for reward anticipation (A) and reward outcome (B) in the striatum. These funnel plots represent variance as a function of mean effect sizes of group differences extracted for all studies in the striatum (data are the same as in Figures 2 and 3). Funnel plots show no evidence of publication bias for reward anticipation (normal distribution over the funnel), and an indication of weak publication bias for reward outcome (slight left orientation of the distribution over the funnel). Compared with the original effect size observed in our meta-analysis, the effect size estimated after correction for publication bias based on Duval and Tweedie s trim and fill analyses showed no differences for the reward anticipation phase (estimated corrected effect size:.201, 95CI:.067 -.336). For reward outcome a slightly smaller estimated effect size was found after correction for publication bias (adding the red dots representing missing studies) compared with the original effect size observed in our meta-analysis (estimated corrected effect size: -.149, 95CI: -.265 - -.033). Nonetheless, the corrected effect size for reward outcome was still significant, indicating that if missing studies had been published, our results for reward outcome would still have been the same, except for a slightly lower effect size.

etable 1. Task paradigms and included contrasts for monetary reward anticipation and monetary reward outcome studies Study Task Reward anticipation Reward outcome Balodis et al. (2012) (A1 phase) win Beck et al. (2009) win Bjork et al. (2008) (single trials and first response of doubleresponse trials) win (single response trials) Bjork et al. (2012) (high and low reward magnitude combined) Higher order contrast between reward notification and hit notification Bustamante et al. (2014) (high and low reward magnitude combined) Choi et al. (2012) Cousijn et al. (2012) Iowa gambling task Outcome win > outcome loss Fauth- Bühler et al. (2014) Instrumental motivation task Parametric modulation reward anticipation Parametric modulation outcome Filbey et al. (2013) (all reward magnitudes) *

Goldstein et al. (2007) Hägele et al. (2015) Jansma et al. (2013) Forced choice monetary reward task (high reward magnitude) ** win (high and low reward magnitude combined) win (high reward magnitude) ** Jia et al. (2011) win Martin et al. (2014) Reward prediction task win > anticipation monetary loss Outcome expected win > outcome expected loss Miedl et al. (2010) Blackjack game Outcome win > outcome loss Nestor et al. (2010) win Patel et al. (2013) win > implicit baseline (high and low reward magnitude combined Outcome win > implicit baseline (high and low reward magnitude combined) Romanczuk- Seiferth et al. (2015) win > anticipation no win * win * Rose et al. (2014) win > anticipation no win ** win ** De Ruiter et al. (2009) Probabilistic reversal learning task Outcome win > implicit baseline Sescousse et al. (2013) Incentive win

Van Hell et al. (2010) Van Holst et al. (2012) Guessing task Anticipation 70% chance monetary win > anticipation 30% chance monetary win (high and low reward magnitude combined) win Outcome win > outcome loss Van Holst et al. (2014) Guessing task Anticipation 70% chance monetary win > anticipation 30% chance monetary win (high and low reward magnitude combined) Outcome win > outcome loss Yip et al. (2014) * The data in the paper was analyzed including several covariates (e.g., age). We received data without the covariates. ** The original study included a pharmacological manipulation. We received data from the placebo condition

etable 2. Number of studies and participants involved in specific analyses Analyses Reward Anticipation Reward Outcome N Studies N Participant s N Studies HC versus Addicted individuals 20 1-20 HC: 505 AI: 526 N Participant s 20 1,3,4,6,8-17,19,21-25 HC: 492 AI: 506 HC versus substance users 15 2-4,7-13,15,16,18-20 HC: 351 AI: 366 15 3,4,8-13,15,16,19,21-23,25 HC: 296 AI: 330 HC versus Gamblers 6 1,5,6,13,14,17 HC: 171 AI: 160 7 1,6,13,14,17,24,25 HC: 187 AI: 176 HC: Healthy controls AI: Addicted individuals

etable 3. Reward anticipation meta-analytic results Addicted > Controls Brain region Hemisphere MNI coordinates SDM-Z X Y Z Brainstem Right 12-10 -14 2.855 Putamen / Pallidum Right 30-6 -12 2.646 Anterior medial Right 4-18 temporal lobe / 20 2.065 Amygdala Amygdala Left -16 0-18 2.663 Lateral orbital Left -36 44-14 2.607 / inferior frontal Superior frontal Left -24 60 12 1.837 Superior frontal Right/Left 0 48 46 1.729 Inferiolateral Left -38-64 46 1.807 parietal lobe Superior parietal Left -2-34 76 1.893 Middle temporal Left -64-20 -12 1.866 Posterior temporal Left -48-44 2 2.229 lobe Posterior temporal Left -48-56 -22 1.952 lobe Cuneus Right 4-86 42 1.923 Lingual Right 12-92 -14 1.846 Lateral occipital Right 14-100 0 1.783 lobe Lateral occipital Left -20-100 8 1.857 lobe Cerebellum Left -38-60 -34 1.755 Cerebellum Right 46-58 -36 1.935 Cerebellum Right 24-84 -22 1.975 Cerebellum Left -20-86 -20 2.474 Cerebellum Left -18-66 -34 2.048 Cerebellum Left -12-54 -16 1.79 Threshold: voxel-wise p <.005 and k 10. Anatomical localization of functional clusters was performed based on a probabilistic atlas (Hammers et al., 2003).

etable 4. Reward anticipation meta-analytic results Addicted < Controls Brain region Hemisphere MNI coordinates SDM-Z X Y Z Striatum (anterior middle caudate) Left -8 18 4-2.905 Right 4 18 2-3.964 Striatum (posterior dorsal Left -18 0 20-2.808 caudate) Right 16 2 22-2.472 Thalamus Left -12-10 4-3.152 Thalamus Right 9-14 10-2.543 Anterior cingulate (mpfc) Right 12 42 6-2.439 Anterior cingulate Right 8 32 18-2.763 Anterior cingulate Right 2 8 40-2.349 Posterior cingulate Right 8-22 38-2.958 Posterior cingulate Left -8-48 28-2.533 Inferior frontal Left -58 12 2-2.789 Middle frontal Right 28 8 50-2.645 Precentral Right 60-4 20-2.564 Precentral Left -58-4 26-2.452 Precentral Right/Left 0-24 60-2.434 Postcentral Left -12-42 54-3.035 Postcentral Right 6-32 54-2.559 Inferiolateral parietal lobe Right 50-24 28-2.538 Inferiolateral parietal lobe Right 44-24 34-2.444 Inferiolateral parietal lobe Left -48-28 26-2.651 Superior parietal Right 10-54 28-3.201 Posterior parietal lobe Right 32-42 -12-2.962 Posterior temporal lobe Left -30-50 2-3.099 Posterior temporal lobe Left -48-36 16-2.891 Cuneus Right/Left 0-74 30-2.594 Lingual / cuneus Left -6-72 8-2.465 Lateral occipital lobe Right 30-66 -4-2.801 White matter Left -24-16 -10-2.927 Corpus callosum / ventrical Left -4-38 6-2.790 Threshold: voxel-wise p <.005 and k 10. Anatomical localization of functional clusters was performed based on a probabilistic atlas (Hammers et al., 2003).

etable 5. Reward outcome meta-analytic results Addicted > Controls Brain region Hemisphere MNI coordinates SDM-Z X Y Z Ventral Left -4 14-4 4.217 striatum (ventral caudate) Ventral Right 12 8-10 3.149 striatum (ventral putamen / nucleus accumbens) Ventral Left -30 8 0 3.431 striatum (ventral putamen) Striatum Left -24 10-18 3.135 (lateral putamen) Insula Right 44 0-8 3.028 Insula Right 46 12-12 3.046 Thalamus Left -18-22 4 3.734 Anterior Right 30 44-12 3.311 orbital Posterior Left -24 24-16 3.495 orbital Posterior Left -6-20 30 3.352 cingulate Precentral Left -24-22 64 3.181 Precentral Left -42-10 48 3.409 Postcentral Left -16-24 42 3.37 Postcentral Right 34-18 40 3.524 Superior Right/Left 0 34 46 3.468 frontal Inferior Right 42 24-12 3.126 frontal Middle frontal Left -44 24 36 3.343

Superior frontal Right 26 16 50 3.688 Left -18 14 56 3.252 Right 10-8 48 3.261 Superior frontal Inferiolateral Right 34-40 34 3.752 parietal lobe Inferiolateral Right 42-48 24 3.836 parietal lobe Inferiolateral Right 42-68 32 3.271 parietal lobe Inferiolateral Right 62-28 20 3.44 parietal lobe Posterior Right 50-58 -6 4.019 temtopral lobe / Lateral occipital lobe Posterior Left -50-68 2 3.974 temporal lobe Posterior Left -56-48 -18 3.504 temporal lobe Lingual Right 4-90 -6 3.345 Lateral Right 28-78 32 3.117 occipital lobe Threshold: voxel-wise p <.005 and k 10. Anatomical localization of functional clusters was performed based on a probabilistic atlas (Hammers et al., 2003).

etable 6. Reward outcome meta-analytic results Addicted < Controls Brain region Hemisphere MNI coordinates SDM-Z X Y Z Dorsal striatum (dorsal caudate) Left -16 12 20-2.12 Anterior cingulate Left -10 46 14-1.43 Hippocampus Left -26-18 -20-1.27 Precentral Left -38-2 30-2.08 Right 34 4 28-1.75 Precentral Left -52-2 18-1.35 Postcentral Left -34-24 20-1.22 Superior frontal Left -24 48 36-1.51 Inferor temporal Right 52-12 -20-1.79 Left -46-8 -30-1.57 Posterior temporal lobe Right 34-42 -2-1.36 Posterior temporal lobe Left -32-54 0-1.96 Posterior temporal lobe Right 12-38 0-1.42 Superior temporal Left -44 4-18 -1.64 Lateral occipital lobe Left -20-94 6-1.34 Lateral occipital lobe Left -36-92 6-1.32 Lateral occipital lobe Left -26-62 28-1.74 Lingual Left -16-76 -2-1.46 White matter / Lateral ventricle Right 14 26 10-1.48 Ventricle Left -4-6 -14-1.25 White matter Left -20-12 -10-1.91 Threshold: voxel-wise p <.005 and k 10. Anatomical localization of functional clusters was performed based on a probabilistic atlas (Hammers et al., 2003).

References 1. Balodis IM, Kober H, Worhunsky PD, Stevens MC, Pearlson GD, Potenza MN. Diminished frontostriatal activity during processing of monetary rewards and losses in pathological gambling. Biol Psychiatry. 2012;71(8):749-757. 2. Bustamante JC, Barros-Loscertales A, Costumero V, et al. Abstinence duration modulates striatal functioning during monetary reward processing in cocaine patients. Addict Biol. 2014;19(5):885-894. 3. Bjork JM, Smith AR, Chen G, Hommer DW. Mesolimbic recruitment by nondrug rewards in detoxified alcoholics: Effort anticipation, reward anticipation, and reward delivery. Human Brain Mapping. 2012;33(9):2174-2188. 4. Bjork JM, Smith AR, Hommer DW. Striatal sensitivity to reward deliveries and omissions in substance dependent patients. NeuroImage. 2008;42(4):1609-1621. 5. Choi JS, Shin YC, Jung WH, et al. Altered brain activity during reward anticipation in pathological gambling and obsessive-compulsive disorder. PLoS One. 2012;7(9):e45938. 6. Fauth-Bühler M, Zois E, Vollstädt-Klein S, Lemenager T, Beutel M, Mann K. Insula and striatum activity in effort-related monetary reward processing in gambling disorder: The role of depressive symptomatology. NeuroImage: Clinical. 2014;6:243-251. 7. Hagele C, Schlagenhauf F, Rapp M, et al. Dimensional psychiatry: reward dysfunction and depressive mood across psychiatric disorders. Psychopharmacology (Berl). 2015;232(2):331-341. 8. Jansma JM, van Hell HH, Vanderschuren LJ, et al. THC reduces the anticipatory nucleus accumbens response to reward in subjects with a nicotine addiction. Transl Psychiatry. 2013;3:e234. 9. Jia Z, Worhunsky PD, Carroll KM, et al. An Initial Study of Neural Responses to Incentives as Related to Treatment Outcome in Cocaine Dependence. Biological psychiatry. 2011. 10. Martin LE, Cox LS, Brooks WM, Savage CR. Winning and losing: differences in reward and punishment sensitivity between smokers and nonsmokers. Brain Behav. 2014;4(6):915-924. 11. Nestor L, Hester R, Garavan H. Increased ventral striatal BOLD activity during non-drug reward anticipation in cannabis users. NeuroImage. 2010;49(1):1133-1143. 12. Patel KT, Stevens MC, Meda SA, et al. Robust changes in reward circuitry during reward loss in current and former cocaine users during performance of a monetary. Biol Psychiatry. 2013;74(7):529-537. 13. Romanczuk-Seiferth N, Koehler S, Dreesen C, Wustenberg T, Heinz A. Pathological gambling and alcohol dependence: neural disturbances in reward and loss avoidance processing. Addict Biol. 2015;20(3):557-569. 14. Sescousse G, Barbalat G, Domenech P, Dreher JC. Imbalance in the sensitivity to different types of rewards in pathological gambling. Brain. 2013;136(Pt 8):2527-2538. 15. Van Hell HH, Vink M, Ossewaarde L, Jager G, Kahn RS, Ramsey NF. Chronic effects of cannabis use on the human reward system: an fmri study. European

neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology. 2010;20(3):153-163. 16. van Holst RJ, Clark L, Veltman DJ, van den Brink W, Goudriaan AE. Enhanced striatal responses during expectancy coding in alcohol dependence. Drug Alcohol Depend. 2014;142:204-208. 17. van Holst RJ, Veltman DJ, Buchel C, van den Brink W, Goudriaan AE. Distorted expectancy coding in problem gambling: is the addictive in the anticipation? Biol Psychiatry. 2012;71(8):741-748. 18. Yip SW, DeVito EE, Kober H, Worhunsky PD, Carroll KM, Potenza MN. Pretreatment measures of brain structure and reward-processing brain function in cannabis dependence: an exploratory study of relationships with abstinence during behavioral treatment. Drug Alcohol Depend. 2014;140:33-41. 19. Rose EJ, Ross TJ, Salmeron BJ, et al. Acute nicotine differentially impacts anticipatory valence- and magnitude-related striatal activity. Biological psychiatry. 2013;73(3):280-288. 20. Filbey FM, Dunlop J, Myers US. Neural Effects of Positive and Negative Incentives during Marijuana Withdrawal. PLoS ONE. 2013;8(5):e61470. 21. Beck A, Schlagenhauf F, Wustenberg T, et al. Ventral striatal activation during reward anticipation correlates with impulsivity in alcoholics. Biological psychiatry. 2009;66(8):734-742. 22. Cousijn J, Wiers RW, Ridderinkhof KR, et al. Individual differences in decision making and reward processing predict changes in cannabis use: a prospective functional magnetic resonance imaging study. Addict Biol. 2013;18(6):1013-1023. 23. Goldstein RZ, Alia-Klein N, Tomasi D, et al. Decreased prefrontal cortical sensitivity to monetary reward is associated with impaired motivation and selfcontrol in cocaine addiction. The American Journal of Psychiatry. 2007;164(1):43-51. 24. Miedl SF, Fehr T, Meyer G, Herrmann M. Neurobiological correlates of problem gambling in a quasi-realistic blackjack scenario as revealed by fmri. Psychiatry Res. 2010;181(3):165-173. 25. De Ruiter MB, Veltman DJ, Goudriaan AE, Oosterlaan J, Sjoerds Z, Van den Brink W. Response perseveration and ventral prefrontal sensitivity to reward and punishment in male problem gamblers and smokers. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology. 2009;34(4):1027-1038.