Abstract
Background and aims
Cognitive control and reward-related abnormalities are centrally implicated in addiction. However, findings from longitudinal studies addressing neurocognitive predictors of addictive behaviors are mixed. Further, little work has been conducted predicting non-substance-related addictive behaviors. Our study aimed to assess predictors of substance and non-substance addictive behaviors in a community sample, systematically evaluating each neurocognitive function's independent influence on addictive behavior.
Methods
Australians (N = 294; 51.7% female; M[SD] age = 24.8[4.7] years) completed online neurocognitive tasks and surveys at baseline and 3-month follow-up. Self-report scales assessed problematic alcohol use, addictive eating (AE), problematic pornography use (PPU), and problematic internet use (PUI) at 3- and 6-month follow-ups. Linear regressions with bootstrapping assessed neurocognitive predictors for each addictive behavior across a 6-month period.
Results
Neurocognition at baseline did not predict AE or PUI severity at 6-month follow-up. Less delay discounting at baseline predicted higher PPU at 6-month follow-up (β = −0.16, p = 0.005). Poorer performance monitoring at baseline predicted higher AE at 3-month follow-up (β = −0.16, p = 0.004), and more reward-related attentional capture at 3-months predicted higher AE at 6-month follow-up (β = 0.14, p = 0.033). Less reward-related attentional capture (β = −0.14, p = 0.003) and less risk-taking under ambiguity (β = −0.11, p = 0.029) at baseline predicted higher PUI at 3-month follow-up. All findings were of small effect size. None of the neurocognitive variables predicted problematic alcohol use.
Discussion and conclusions
We were unable to identify a core set of specific neurocognitive functions that reliably predict multiple addictive behavior types. However, our findings indicate both cognitive control and reward-related functions predict non-substance addictive behaviors in different ways. Findings suggest that there may be partially distinct neurocognitive mechanisms contributing to addiction depending on the specific addictive behavior.
Introduction
A profile of neurocognitive dysfunction has been proposed to be critical across addictive behaviors (Yücel et al., 2019, 2021); specifically, reward-related learning and higher-order executive functions such as cognitive control (Bechara, 2005; Smith et al., 2014; Verdejo-Garcia & Albein-Urios, 2021; Volkow & Boyle, 2018; Volkow & Morales, 2015; Yücel & Lubman, 2007). However, a recent systematic review revealed that individual differences in these cognitive control and reward-related processes do not consistently predict addiction outcomes in longitudinal samples (Christensen, Brydevall, et al., 2023). This finding was in part attributed to methodological disparities in the literature, such as the inconsistent selection of neurocognitive domains, choice of tasks, and choice of confounders (or lack thereof) to include in models. There has also been very little literature investigating the neurocognitive predictors of non-substance addictive disorders, making it difficult to determine whether there are trans-addiction neurocognitive mechanisms or if different addictive behaviors have unique neurocognitive predictors. These insights would allow for the development of more targeted and effective interventions that address distinct pathways involved in specific addictive behaviors.
Longitudinal research on the neurocognitive mechanisms of addiction has predominantly focussed on substance use disorders (SUDs) and problem gambling (PG). However, several non-substance-related addictive behaviors merit investigation. Addictive Eating (AE), Problematic Pornography Use (PPU), and Problematic Use of the Internet (PUI) are common behavioral problems, affecting 2.6%–15% of the population globally (Kumar et al., 2021; Mennig et al., 2020; Meule & Gearhardt, 2019; Pan et al., 2020). In keeping with SUDs and PG, AE, PPU and PUI have been linked to significantly poorer psychosocial outcomes (Burmeister, Hinman, Koball, Hoffmann, & Carels, 2013; Burrows, Kay-Lambkin, Pursey, Skinner, & Dayas, 2018; Camilleri, Perry, & Sammut, 2021; Fineberg et al., 2018; Floros & Ioannidis, 2021; Kuss et al., 2014; Raj et al., 2022; Rodrigue et al., 2018).
Similar to SUDs, neurocognitive changes have been observed in individuals with AE, PUI, and PPU, such as decreased response inhibition, poorer performance monitoring ability, inflexible task shifting, and risky decision-making (Franken et al., 2018; Ioannidis et al., 2019; Müller et al., 2023; Odlaug et al., 2011; Rodrigue et al., 2018; Smith et al., 2014; Zhou et al., 2013), alongside enhanced attentional bias towards addiction- or reward-related cues (Adams, Sedgmond, Maizey, Chambers, & Lawrence, 2019; Albertella, Pelley, et al., 2019; Jeromin et al., 2016; Mechelmans et al., 2014; Nikolaidou et al., 2019), and steeper temporal discounting (Amlung, Vedelago, Acker, Balodis, & MacKillop, 2017; Antons et al., 2019) and riskier decision-making (Ioannidis et al., 2019). However, little longitudinal research exists expressly testing predictive factors. Only a handful of studies have evaluated the neurocognitive predictors of non-substance addictions and, of these, none have looked at AE, PUI, or PPU specifically (Christensen, Brydevall, et al., 2023). This is especially important given work in SUDs suggests cross-sectional correlates of addiction do not necessarily imply predictive mechanisms, for example, delayed discounting and decreased response inhibition have been shown to be key correlates of substance addiction (Amlung et al., 2017; Smith et al., 2014), yet these relationships have not been consistently replicated longitudinally (Christensen, Brydevall, et al., 2023).
The current lack of evidence suggesting neurocognitive functions predict consumption (i.e. frequency/quantity), severity, or diagnosis of addiction (Christensen, Brydevall, et al., 2023), may be due to the heterogeneous nature of past longitudinal studies, making them difficult to compare. In a recent systematic review of the literature by Christensen and colleagues (2023), 44% of studies focused on a single, specific neurocognitive function (Audrain-McGovern et al., 2009; Chen et al., 2021; Cousijn et al., 2015; Fröhner et al., 2022; Jones et al., 2021; Peeters et al., 2014; van Hemel-Ruiter et al., 2015). Those that looked at multiple functions predominantly evaluated a single domain, for example reward valuation, assessed via delay discounting and probability discounting tasks (Bernhardt et al., 2017; Kräplin et al., 2020), or multiple tasks that evaluated aspects of cognitive control (Bø, Billieux, Gjerde, Eilertsen, & Landrø, 2017; Rubio et al., 2008), for example response inhibition and set shifting. Rarely were multiple neurocognitive tasks that independently tapped into both reward-related and cognitive control functions included in the same model (Fernández-Artamendi et al., 2018; Whelan et al., 2014). This is a critical area for investigation given both cognitive control and reward-related neurocognitive processes interact and in many cases overlap (Criaud & Boulinguez, 2013; Ridderinkhof et al., 2004). Interrogating the unique role of specific cognitive control and reward-related functions in the same model, will allow us to determine each function's relative contribution to addictive behaviors. This knowledge can be used to identify individuals at higher risk of developing addiction and inform novel treatment targets and more tailored therapeutic strategies.
A further issue with current studies is the preponderance of clinical samples. The trajectory from the onset of addiction problems to receiving clinical care can take a median of 18 years (Chapman, Slade, Hunt, & Teesson, 2015). Studies that assess individuals in clinical treatment settings only capture a small subset of those affected by addiction (Grant et al., 2015; Hasin et al., 2013). Investigating general community samples affords the opportunity to identify and study these individuals before they would typically be accessible in treatment settings. We can also adopt a dimensional approach that encompasses a spectrum of addictive behavior severities. This is especially important as even individuals engaging in addictive behaviors at less severe levels also experience adverse outcomes (Shankman et al., 2009). Understanding the neurocognitive mechanisms that predict addictive behaviors dimensionally will better identify early risk indicators for addiction and provide insights into key treatment targets at different stages of illness.
The aim of this study was to evaluate the extent to which neurocognitive measures predict subsequent substance (alcohol) and non-substance (AE, PPU and PUI) addictive behaviors in a general community sample. This study took a systematic and comprehensive approach to neurocognitive assessment (Lee et al., 2023) by evaluating the independent influence of multiple individual neurocognitive functions on addictive behaviors. Neurocognitive functions were selected per expert-endorsed domains which are thought to drive addiction (Yücel et al., 2019), capturing cognitive control, working memory, reward learning, and reward valuation functions as defined by the Research Domain Criteria (RDoC: Insel et al., 2010). Given literature looking at the neurocognitive functions associated with AE, PPU and PUI is still sparse, we first evaluated the cross-sectional correlates of each addictive behavior, before then modeling these relationships longitudinally over a 6-month period. To determine whether these relationships were replicable, we evaluated the relationship between neurocognitive functions assessed at baseline and addictive behavior at 3-month follow-up, and then the relationship between neurocognitive functions assessed at 3-months and addictive behavior at 6-month follow-up. All models accounted for both trait impulsivity and compulsivity as well as key covariates (i.e. psychological distress, sex and age).
Methods
Participants
This study was embedded within a larger cohort study (Christensen, Albertella, et al., 2023) which aimed to collect normative data in a demographically-stratified sample and to investigate the neurocognitive correlates of addictive behaviors. Nine-hundred-and-forty-four community members were recruited from Prolific and online advertisements via popular social media sites and enrolled in the wider study; 400 of which, aged between 18 and 35 years, were invited to take part in the longitudinal protocol and made up the study sample for the present paper. This age band was chosen given the median age of onset for substance use/addictive behavior disorders is 25 years of age (Solmi et al., 2022). Participants were Australian residents, who were not color blind and self-reported an absence of a neurological disorder (i.e. stroke, brain injury, and dementia) and an absence of a history of a psychotic disorder.
Measures
All data were collected remotely via participants' laptops or desktop computers. Participants provided basic demographic information (i.e. age, sex, average household income, and education status), completed neurocognitive tasks and self-report surveys at baseline and 3-months, and underwent a brief assessment of self-report scales, including addictive behavior scales at 6-months. A detailed description of the study protocol and a full list of measures can be found in the supplementary materials for this manuscript.
Neurocognitive tasks
Tasks were selected to assess neurocognitive functions linked to addiction (Yücel et al., 2019), see Table 1 for details of each task. Four of the eight tasks were delivered by the BrainPark Assessment of Cognition application (BrainPAC), a novel expert-endorsed digital assessment tool for addictive disorders (refer to Lee et al., 2023 for psychometric information). The remaining four were delivered via Inquisit 5 (2018).
Neurocognitive tasks included in the study
| Function | Task | Brief description |
| Response inhibition | The BrainPAC Stop Signal Taks (SST) Lee et al. (2023) | A gamified visual cue stop signal paradigm. The primary outcome metric is stop signal reaction time (SSRT). Higher SSRT indicated poorer inhibitory control. |
| Reward learning (reward-related attentional bias) | The BrainPAC Value Modulated Attentional Capture (VMAC) Task Lee et al. (2023) | A gamified version of a standard VMAC task (Albertella, Pelley, et al., 2019; Le Pelley et al., 2015). The primary outcome metric is the VMAC score, the difference in reaction time between trials with a high- versus low-value distractor present. The VMAC score is averaged across the last two blocks of the task. Higher values indicate more reward-related attentional capture. |
| Reward learning (goal-directed vs habitual) | The BrainPAC Sequential Decision-Making Task (SDT) Lee et al. (2023) | A gamified two-stage choice task. The primary outcome metric is mixing weight (w). Higher scores indicate more goal-directed (model-based) decision-making. |
| Reward valuation (risky decision-making under uncertainty) | The BrainPAC Balloon Analogue Risk Task (BART) Lee et al. (2023) | A gamified version of the BART stretch variant. The primary outcome metric is the mean pre-committed pumps across all trials. Higher values indicate riskier choice in the face of uncertainty. |
| Flexible updating | N-back Task (Ragland et al., 2002; Inquisit 5, 2018) | A letter sequencing go/no-go task that progressively increases working memory load. The primary outcome metric is the parametric measure of sensitivity (d’), which in this study was calculated from 3-back trials. Higher d’ values indicate more flexible updating. |
| Goal selection; updating, representation and maintenance | Category Switch Task (CST) (Friedman et al., 2008; Inquisit 5, 2018) | A task switching paradigm. The primary outcome metric is the latency switch cost, calculated as the difference in reaction time on switch versus non-switch trials. Higher values indicate poorer task switching. |
| Performance monitoring | Error Awareness Task (EAT) (Hester et al., 2007; Inquisit 5, 2018) | A visual go/no-go paradigm in which participants indicate their error awareness following any commission error. The primary outcome metric is percentage awareness of commission errors. Higher values indicate better performance monitoring. |
| Temporal discounting | Monetary Choice Questionnaire (MCQ) Kirby et al. (1999) | A 27-item questionnaire asks the participant to choose between two hypothetical reward options, a smaller reward now, or a larger reward at some point in the future. The primary outcome measure is discounting rate (log k). Higher values indicate preference for sooner but smaller rewards. |
Addictive behaviors
Problematic alcohol use was assessed by the Alcohol Use Disorder Identification Test total score (AUDIT; Saunders et al., 1993) which has shown excellent internal consistency: Cronbach's α = 0.96 (Noorbakhsh et al., 2018). Addictive Eating (AE) was evaluated via the modified Yale Food Addiction Scale symptom count (mYFAS 2.0; Schulte & Gearhardt, 2017) which has shown good internal consistency: Cronbach's α = 0.88 (Imperatori et al., 2019). Problematic Pornography Use (PPU) was assessed via the short version of the Problematic Pornography Consumption Scale total score (PPCS-6: Bőthe, Tóth-Király, Demetrovics, & Orosz, 2021) which has shown good internal consistency: Cronbach's α = 0.84 (Bőthe et al., 2021). Problematic Use of the Internet (PUI) was evaluated by the abbreviated Young's Internet Addiction Test (IAT-10; Tiego et al., 2021) total score, which has shown good internal consistency: Cronbach's α = 0.80 (Tiego et al., 2021). Participants reported addictive behaviors using the corresponding scale reflecting on the previous three months, with zero assigned if no engagement. All addictive behaviors were captured dimensionally.
Covariates
The total score of the Depression Anxiety Stress Scale was used to assess psychological distress (DASS-21; Szabó, 2010) which has shown excellent internal consistency: Cronbach's α = 0.93 (Liu et al., 2021). The Cambridge-Chicago Compulsivity Trait Scale (CHI-T; Chamberlain & Grant, 2018; Tiego et al., 2023) assessed trait compulsivity, which has shown good internal consistency: Cronbach's α = 0.80 (Chamberlain & Grant, 2018). The Short UPPS Impulsive Behavior Scale (SUPPS-P; Cyders et al., 2014) assessed trait impulsivity, with three scores: urgency, lack of planning/persistence, and sensation seeking, as validated by factor analyses (Billieux et al., 2012, 2021). The SUPPS-P has shown acceptable internal consistency: Cronbach's α = 0.75 (Liu et al., 2021).
Procedure
Assessments were delivered online using Qualtrics software (Version 08.23, 2023). At baseline and 3-month follow-up, participants completed three one-hour assessment sessions across three consecutive days. During each session, the neurocognitive tasks were separated by self-report surveys (trait and behavior scales). The order of task presentation was counterbalanced. At 6-month follow-up participants completed a single 30-min questionnaire that evaluated addictive behaviour engagement.
Statistical analysis
All data cleaning and analyses were performed using R Statistical Software (v4.1.2; R Core Team 2021). The data underwent a cleaning procedure to ensure data quality. The cleaning procedures were specified in advance of data curation by the study team. Implausible responses and poor performance presumedly due to lack of effort were identified and removed via attention check questions, neurocognitive task performance at less than chance levels (as per Lee et al., 2023; Albertella, Watson, et al., 2019), as well as task-specific cleaning procedures (e.g. SST go trial accuracy, stop trial accuracy [Verbruggen et al., 2019], and Independent Race Model check [Band, Van Der Molen, & Logan, 2003]). Differences between individuals whose data were filtered out compared with included individuals were investigated using Welch's t-tests and when normality was violated Mann-Whitney U tests were used and reported in the supplementary materials. Similarly, Welch's t-tests and where appropriate Mann-Whitney U tests were used to compare individuals who returned for each longitudinal assessment and those lost to follow-up and are reported below.
Power calculations using G-Power 3.1 (Faul, Erdfelder, Lang, & Buchner, 2007) deemed n = 190 was the minimum sample size required for multiple regression analyses with 16 predictors to find a small effect (f2 = 0.10, power = 0.80). Statistical outliers on neurocognitive measures (≥3 standard deviations from the mean) were removed (Field, 2012). All analyses were conducted on complete data sets (i.e. participants who provided data for all variables of interest). Bivariate Spearman correlations (adjusted for multiple comparisons using the Holm method: Holm, 1979), investigated relationships among variables of interest at each time point (Table S1–4).
Four multiple regression models were generated for each outcome behavior of interest. The first model evaluated cross-sectional relationships between neurocognition, traits, covariates and addictive behavior at baseline. The second model included baseline variables as predictors of addictive behavior at 6-month follow-up, accounting for baseline addictive behavior severity, the third model included baseline variables as predictors of addictive behavior at 3-month follow-up, accounting for baseline addictive behavior severity, and the fourth model included variables measured at 3-months to predict addictive behavior at 6-month follow-up, accounting for 3-month addictive behavior severity (Tables 3–6). Visual inspection of residual plots and VIF values were evaluated independently for each regression model and revealed assumptions of linearity, independence, and no multicollinearity were met. VIF values less than 2.5 were taken to indicate no issue of multicollinearity (Johnston et al., 2018). Normality and homogeneity of variance were violated for each model, additionally, the distributions of all four outcome variables at each timepoint were positively skewed (Figure SII), constituting the choice of linear regression models with bootstrapping of residuals (5,000 samples; Hesterberg, 2015; Neal & Simons, 2007) using the boot.pval (Thulin, 2023) and car (Fox, 2019) packages. Age, sex, psychological distress, impulsivity, compulsivity, and the respective baseline addictive behavior scale (measured at 3-month assessment for regressions predicting 6-month outcome) score were included as covariates in each regression model (Eisenberg et al., 2019; Sjoberg & Cole, 2018; Starcke et al., 2016; Tennant et al., 2021). To test the impact of serial correlation error the Heteroscedasticity and Autocorrelation Consistent (HAC) covariance matrix estimation was run for each longitudinal model (Andrews & Monahan, 1992), these results can be found in the supplementary materials (Tables S5–S8). To counter potential bias of our longitudinal models due to drop-out over the course of the study, as secondary analyses we ran each longitudinal regression using Mice multiple imputation for missing data (bootImpute::bootMice; 10 imputations), these results are presented in the supplementary materials (Tables S9–S12).
Demographic and behavioral characteristics
| Variables | BL | 3-mths | 6-mths |
| N | 294 | 175 | 138 |
| Mean age (SD) | 24.8 (4.7) | 24.7 (4.5) | 25.0 (4.4) |
| Sex, N (%) | |||
| Female | 152 (51.7) | 90 (51.4) | 73 (52.9) |
| Gender, N (%) | |||
| Man | 141 (48.0) | 85 (48.6) | 65 (47.1) |
| Woman | 149 (50.7) | 88 (50.3) | 71 (51.4) |
| Non-binary | 3 (1.0) | 1 (0.6) | 1 (0.7) |
| Not listed/Prefer not to say | 1 (0.3) | 1 (0.6) | 1 (0.7) |
| Ethnicity, N | |||
| Aboriginal or Torres Strait Islander | 1 | 0 | 0 |
| African | 2 | 0 | 0 |
| Asian | 84 | 57 | 49 |
| Black or African American | 1 | 1 | 0 |
| Hispanic or Latino | 3 | 2 | 1 |
| Middle Eastern | 3 | 3 | 3 |
| South Asian | 23 | 13 | 9 |
| White or Caucasian | 166 | 91 | 69 |
| Other | 11 | 8 | 7 |
| Household income in AUD, N | |||
| < $10,000 | 13 | 9 | 6 |
| $10,000 – $20,000 | 12 | 10 | 9 |
| $20,000 – $40,000 | 25 | 18 | 12 |
| $40,000 – $60,000 | 55 | 32 | 26 |
| $60,000 – $80,000 | 38 | 25 | 21 |
| $80,000 – $100,000 | 45 | 24 | 19 |
| > $100,000 | 106 | 57 | 45 |
| AUDIT | |||
| Mean (SD) | 3.0 (3.7) | 2.7 (4.3) | 2.2 (3.8) |
| Range | 0–24 | 0–30 | 0–21 |
| % classified as problematic (hazardous use | dependence) | 10.2 | 1.4 | 8.6 | 1.7 | 2.9 | 3.6 |
| mYFAS | |||
| Mean (SD) | 0.7 (1.7) | 0.7 (1.6) | 0.7 (1.8) |
| Range | 0–10 | 0–10 | 0–11 |
| % classified as problematic (mild | moderate | severe) | 5.8 | 3.1 | 4.1 | 8.0 | 4.7 | 3.4 | 6.5 | 5.1 | 2.9 |
| % met diagnostic threshold * | 5.4 | 6.2 | 9.4 |
| PPCS | |||
| Mean (SD) | 10.1 (5.8) | 10.3 (5.6) | 9.9 (5.9) |
| Range | 6–40 | 6–30 | 6–37 |
| % classified as problematic | 7.1 | 9.1 | 6.5 |
| IAT | |||
| Mean (SD) | 17.4 (6.8) | 17.3 (6.7) | 16.8 (5.7) |
| Range | 10–48 | 10–42 | 10–37 |
| % classified as problematic | 41.2 | 41.1 | 39.9 |
Note: Sex was defined as biological sex. Gender was defined as the participant's gender identity at the time of the baseline assessment. AUDIT: Alcohol Use Identification Test, harmful/hazardous use (≥8), likely alcohol dependence (≥15); mYFAS: modified Yale Food Addiction Scale 2.0, mild (2–3), moderate (4–5) severe (≥6) symptoms. PPCS: Problematic Pornography Consumption Scale, problematic use (≥20). IAT: an abbreviated version of Young's Internet Addiction Test, problematic use (≥17). * The mYFAS also provides a diagnostic score adapted from the DSM-V criteria for substance use disorder. To further characterize the sample, the percentage of individuals at each time point who met this diagnostic threshold is reported, however, these values were not used for statistical analyses.
Cross-sectional and longitudinal multiple regression models of problematic alcohol use
| Variable | Baseline (N = 294) | Baseline predicting 6-months (N = 198) | Baseline predicting 3-months (N = 206) | 3-months predicting 6-months (N = 138) | ||||||||||||||||
| β | SE | 95% CI | p | β | SE | 95% CI | p | β | SE | 95% CI | p | β | SE | 95% CI | p | |||||
| LL | UL | LL | UL | LL | UL | LL | UL | |||||||||||||
| Demographics | ||||||||||||||||||||
| Age | 0.04 | 0.05 | −0.06 | 0.13 | 0.498 | 0.15 | 0.04 | 0.03 | 0.20 | 0.008** | 0.04 | 0.03 | −0.03 | 0.10 | 0.276 | 0.06 | 0.05 | −0.06 | 0.16 | 0.372 |
| Sex (F) | 0.14 | 0.44 | 0.19 | 1.87 | 0.020* | −0.05 | 0.37 | −1.10 | 0.38 | 0.334 | 0.02 | 0.29 | −0.46 | 0.71 | 0.676 | 0.01 | 0.46 | −0.81 | 1.01 | 0.821 |
| Neurocognition | ||||||||||||||||||||
| SST: SSRT | −0.03 | 3.17 | −7.66 | 4.73 | 0.645 | −0.10 | 2.75 | −10.43 | 0.57 | 0.079 | −0.03 | 2.25 | −5.19 | 2.51 | 0.494 | −0.07 | 3.41 | −10.46 | 3.13 | 0.257 |
| VMAC: VMAC score | −0.10 | 5.26 | −19.66 | 1.34 | 0.079 | −0.06 | 4.63 | −14.26 | 4.02 | 0.273 | −0.01 | 4.51 | −8.25 | 5.99 | 0.768 | 0.02 | 5.39 | −9.15 | 12.60 | 0.761 |
| BART: M pre-committed pumps | −0.05 | 0.01 | −0.04 | 0.02 | 0.417 | −0.08 | 0.01 | −0.04 | 0.01 | 0.154 | 0.01 | 0.01 | −0.02 | 0.02 | 0.824 | −0.04 | 0.01 | −0.03 | 0.02 | 0.478 |
| CST: Switch cost latency | −0.08 | 0.00 | −0.00 | 0.00 | 0.189 | −0.04 | 0.00 | −0.00 | 0.00 | 0.416 | 0.02 | 0.00 | −0.00 | 0.00 | 0.574 | 0.05 | 0.00 | −0.00 | 0.00 | 0.357 |
| EAT: Error awareness | 0.07 | 0.01 | −0.01 | 0.02 | 0.231 | 0.02 | 0.01 | −0.01 | 0.01 | 0.694 | 0.01 | 0.00 | −0.01 | 0.01 | 0.735 | 0.01 | 0.01 | −0.02 | 0.02 | 0.867 |
| SDT: w | −0.03 | 0.61 | −1.37 | 0.93 | 0.692 | −0.06 | 0.51 | −1.54 | 0.43 | 0.285 | −0.02 | 0.40 | −0.98 | 0.57 | 0.601 | −0.03 | 0.58 | −1.45 | 0.81 | 0.572 |
| N-Back: 3-back d’ | 0.11 | 0.19 | −0.02 | 0.72 | 0.060 | 0.01 | 0.16 | −0.30 | 0.35 | 0.900 | 0.02 | 0.11 | −0.18 | 0.30 | 0.630 | 0.02 | 0.17 | −0.25 | 0.42 | 0.647 |
| DDT: Log k | 0.02 | 0.13 | −0.21 | 0.31 | 0.701 | −0.03 | 0.11 | −0.28 | 0.17 | 0.612 | 0.01 | 0.07 | −0.15 | 0.20 | 0.722 | −0.02 | 0.13 | −0.31 | 0.21 | 0.719 |
| Covariates | ||||||||||||||||||||
| AUDIT | – | – | – | – | – | 0.68 | 0.06 | 0.53 | 0.74 | <0.001*** | 0.84 | 0.06 | 0.76 | 0.92 | <0.001*** | 0.78 | 0.06 | 0.70 | 0.96 | <0.001*** |
| DASS: Total score | 0.22 | 0.02 | 0.03 | 0.12 | <0.001*** | 0.07 | 0.02 | −0.02 | 0.06 | 0.248 | 0.04 | 0.01 | −0.02 | 0.04 | 0.458 | 0.07 | 0.02 | −0.02 | 0.07 | 0.245 |
| SUPPS-P: Lack of perseverance and premeditation | 0.07 | 0.14 | −0.13 | 0.42 | 0.273 | −0.13 | 0.12 | −0.47 | 0.00 | 0.051 | 0.00 | 0.12 | −0.19 | 0.19 | 0.995 | −0.04 | 0.14 | −0.38 | 0.21 | 0.508 |
| SUPPS-P: Urgency | 0.00 | 0.11 | −0.23 | 0.22 | 0.985 | 0.17 | 0.10 | 0.06 | 0.44 | 0.013* | 0.09 | 0.08 | −0.01 | 0.29 | 0.064 | −0.06 | 0.12 | −0.34 | 0.13 | 0.406 |
| SUPPS-P: Sensation seeking | 0.27 | 0.08 | 0.18 | 0.51 | <0.001*** | −0.00 | 0.10 | −0.14 | 0.14 | 0.978 | 0.02 | 0.05 | −0.08 | 0.13 | 0.660 | 0.04 | 0.08 | −0.11 | 0.21 | 0.567 |
| CHI-T: Trait compulsivity | −0.04 | 0.05 | −0.11 | 0.06 | 0.552 | −0.08 | 0.04 | −0.12 | 0.03 | 0.231 | −0.06 | 0.04 | −0.10 | 0.03 | 0.269 | 0.07 | 0.05 | −0.06 | 0.16 | 0.360 |
Note. β: Standardized coefficient; SE: Standard error; *p < 0.05, **p < 0.01, ***p < 0.001.
Cross-sectional and longitudinal multiple regression models of AE
| Variable | Baseline (N = 294) | Baseline predicting 6-months (N = 198) | Baseline predicting 3-months (N = 206) | 3-months predicting 6-months (N = 138) | ||||||||||||||||
| β | SE | 95% CI | p | β | SE | 95% CI | p | β | SE | 95% CI | p | β | SE | 95% CI | p | |||||
| LL | UL | LL | UL | LL | UL | LL | UL | |||||||||||||
| Demographics | ||||||||||||||||||||
| Age | 0.01 | 0.02 | −0.03 | 0.05 | 0.799 | 0.06 | 0.03 | −0.03 | 0.07 | 0.364 | 0.08 | 0.02 | −0.01 | 0.07 | 0.175 | −0.01 | 0.03 | −0.06 | 0.05 | 0.891 |
| Sex (F) | 0.18 | 0.19 | 0.22 | 0.96 | 0.002** | −0.13 | 0.23 | −0.95 | −0.05 | 0.027* | −0.01 | 0.19 | −0.42 | 0.34 | 0.849 | −0.06 | 0.24 | −0.69 | 0.28 | 0.398 |
| Neurocognition | ||||||||||||||||||||
| SST: SSRT | 0.03 | 1.35 | −1.89 | 3.52 | 0.579 | 0.03 | 1.67 | −2.67 | 4.03 | 0.689 | −0.05 | 1.24 | −3.51 | 1.35 | 0.355 | −0.01 | 1.83 | −3.89 | 3.49 | 0.862 |
| VMAC: VMAC score | 0.03 | 2.22 | −3.10 | 5.68 | 0.610 | 0.01 | 2.78 | −5.12 | 5.85 | 0.894 | −0.02 | 2.31 | −5.31 | 3.82 | 0.740 | 0.14 | 2.86 | 0.53 | 11.62 | 0.033* |
| BART: M pre-committed pumps | 0.01 | 0.01 | −0.01 | 0.01 | 0.896 | −0.06 | 0.01 | −0.02 | 0.01 | 0.321 | −0.00 | 0.01 | −0.01 | 0.01 | 0.953 | −0.11 | 0.01 | −0.02 | 0.00 | 0.088 |
| CST: Switch cost latency | 0.02 | 0.00 | −0.00 | 0.00 | 0.778 | 0.10 | 0.00 | −0.00 | 0.00 | 0.104 | −0.05 | 0.00 | −0.00 | 0.00 | 0.421 | 0.01 | 0.00 | −0.00 | 0.00 | 0.934 |
| EAT: Error awareness | 0.03 | 0.00 | −0.00 | 0.01 | 0.640 | −0.06 | 0.00 | −0.01 | 0.00 | 0.306 | −0.16 | 0.00 | −0.01 | −0.00 | 0.004** | −0.01 | 0.00 | −0.01 | 0.01 | 0.970 |
| SDT: w | −0.06 | 0.26 | −0.76 | 0.24 | 0.303 | 0.04 | 0.31 | −0.43 | 0.79 | 0.535 | −0.03 | 0.24 | −0.62 | 0.36 | 0.600 | −0.07 | 0.31 | −0.93 | 0.26 | 0.253 |
| N-Back: 3-back d’ | 0.01 | 0.08 | −0.14 | 0.17 | 8.29 | −0.01 | 0.10 | −0.21 | 0.17 | 0.863 | 0.04 | 0.08 | −0.09 | 0.20 | 0.457 | −0.07 | 0.09 | −0.26 | 0.08 | 0.275 |
| DDT: Log k | 0.15 | 0.06 | 0.05 | 0.26 | 0.007** | 0.10 | 0.07 | −0.03 | 0.24 | 0.139 | −0.01 | 0.06 | −0.12 | 0.10 | 0.829 | 0.09 | 0.07 | −0.04 | 0.23 | 0.148 |
| Covariates | ||||||||||||||||||||
| mYFAS | – | – | – | – | – | 0.41 | 0.08 | 0.30 | 0.60 | <0.001*** | 0.60 | 0.06 | 0.50 | 0.74 | <0.001*** | 0.67 | 0.07 | 0.60 | 0.89 | <0.001*** |
| DASS: Total score | 0.32 | 0.01 | 0.03 | 0.06 | <0.001*** | 0.17 | 0.01 | 0.00 | 0.05 | 0.023* | 0.08 | 0.01 | −0.01 | 0.03 | 0.219 | 0.12 | 0.01 | −0.00 | 0.04 | 0.094 |
| SUPPS-P: Lack of perseverance and premeditation | −0.05 | 0.06 | −0.16 | 0.07 | 0.454 | 0.09 | 0.07 | −0.05 | 0.23 | 0.235 | 0.05 | 0.06 | −0.07 | 0.17 | 0.419 | 0.07 | 0.08 | −0.08 | 0.22 | 0.378 |
| SUPPS-P: Urgency | 0.05 | 0.05 | −0.05 | 0.14 | 0.406 | 0.13 | 0.06 | −0.01 | 0.22 | 0.083 | 0.09 | 0.05 | −0.03 | 0.16 | 0.200 | −0.05 | 0.06 | −0.15 | 0.09 | 0.556 |
| SUPPS-P: Sensation seeking | −0.13 | 0.03 | −0.14 | −0.01 | 0.023* | −0.01 | 0.04 | −0.08 | 0.08 | 0.902 | 0.07 | 0.03 | −0.03 | 0.11 | 0.251 | 0.03 | 0.04 | −0.06 | 0.11 | 0.671 |
| CHI-T: Trait compulsivity | 0.13 | 0.02 | −0.00 | 0.07 | 0.052 | 0.08 | 0.02 | −0.02 | 0.07 | 0.296 | 0.08 | 0.02 | −0.02 | 0.06 | 0.267 | 0.13 | 0.03 | −0.02 | 0.09 | 0.140 |
Note. β: Standardized coefficient; SE: Standard error; *p < 0.05, **p < 0.01, ***p < 0.001.
Cross-sectional and longitudinal multiple regression models of PPU
| Variable | Baseline (N = 294) | Baseline predicting 6-months (N = 198) | Baseline predicting 3-months (N = 206) | 3-months predicting 6-months (N = 138) | ||||||||||||||||
| β | SE | 95% CI | p | β | SE | 95% CI | p | β | SE | 95% CI | p | β | SE | 95% CI | p | |||||
| LL | UL | LL | UL | LL | UL | LL | UL | |||||||||||||
| Demographics | ||||||||||||||||||||
| Age | −0.05 | 0.07 | −0.20 | 0.07 | 0.348 | −0.06 | 0.06 | −0.19 | 0.05 | 0.267 | 0.02 | 0.06 | −0.09 | 0.14 | 0.647 | −0.07 | 0.09 | −0.26 | 0.09 | 0.306 |
| Sex (F) | −0.51 | 0.62 | −7.03 | −4.64 | <0.001*** | −0.17 | 0.63 | −2.99 | −0.52 | 0.003** | −0.16 | 0.62 | −3.01 | −0.51 | 0.004* | −0.14 | 0.89 | −3.38 | 0.13 | 0.074 |
| Neurocognition | ||||||||||||||||||||
| SST: SSRT | −0.02 | 4.59 | −10.56 | 7.17 | 0.630 | 0.01 | 4.17 | −7.08 | 9.09 | 0.851 | 0.01 | 3.64 | −6.04 | 8.18 | 0.797 | −0.04 | 5.77 | −14.24 | 7.78 | 0.524 |
| VMAC: VMAC score | 0.00 | 7.57 | −14.79 | 14.30 | 0.993 | 0.01 | 6.62 | −11.90 | 14.54 | 0.840 | 0.06 | 6.77 | −4.64 | 22.63 | 0.186 | −0.02 | 9.05 | −20.57 | 16.08 | 0.782 |
| BART: M pre-committed pumps | 0.02 | 0.02 | −0.03 | 0.05 | 0.753 | −0.03 | 0.02 | −0.05 | 0.03 | 0.625 | −0.00 | 0.02 | −0.04 | 0.03 | 0.966 | 0.00 | 0.02 | −0.04 | 0.04 | 0.986 |
| CST: Switch cost latency | 0.04 | 0.00 | −0.00 | 0.00 | 0.476 | 0.07 | 0.00 | −0.00 | 0.00 | 0.197 | 0.07 | 0.00 | −0.00 | 0.00 | 0.176 | 0.04 | 0.00 | −0.00 | 0.01 | 0.448 |
| EAT: Error awareness | −0.03 | 0.01 | −0.02 | 0.01 | 0.647 | −0.06 | 0.01 | −0.03 | 0.01 | 0.298 | 0.06 | 0.01 | −0.01 | 0.02 | 0.268 | 0.03 | 0.01 | −0.02 | 0.03 | 0.632 |
| SDT: w | 0.06 | 0.84 | −0.72 | 2.59 | 0.300 | 0.02 | 0.79 | −1.25 | 1.74 | 0.734 | 0.07 | 0.74 | −0.40 | 2.46 | 0.136 | 0.05 | 0.96 | −1.25 | 2.58 | 0.417 |
| N-Back: 3-back d’ | −0.08 | 0.26 | −0.95 | 0.08 | 0.104 | 0.06 | 0.24 | −0.21 | 0.72 | 0.267 | 0.02 | 0.22 | −0.35 | 0.53 | 0.681 | 0.03 | 0.29 | −0.40 | 0.69 | 0.664 |
| DDT: Log k | −0.02 | 0.18 | −0.44 | 0.27 | 0.638 | −0.16 | 0.17 | −0.82 | −0.16 | 0.005** | −0.02 | 0.16 | −0.38 | 0.26 | 0.729 | −0.16 | 0.22 | −0.97 | −0.12 | 0.014* |
| Covariates | ||||||||||||||||||||
| PPCS | – | – | – | – | – | 0.63 | 0.05 | 0.49 | 0.70 | <0.001*** | 0.67 | 0.05 | 0.56 | 0.77 | <0.001*** | 0.68 | 0.08 | 0.61 | 0.93 | <0.001*** |
| DASS: Total score | 0.13 | 0.03 | 0.01 | 0.13 | 0.034* | −0.00 | 0.03 | −0.06 | 0.06 | 0.974 | −0.01 | 0.03 | −0.06 | 0.05 | 0.919 | 0.11 | 0.03 | −0.01 | 0.12 | 0.120 |
| SUPPS-P: Lack of perseverance and premeditation | 0.02 | 0.20 | −0.31 | 0.47 | 0.704 | 0.04 | 0.17 | −0.23 | 0.47 | 0.480 | −0.00 | 0.17 | −0.35 | 0.33 | 0.944 | −0.02 | 0.24 | −0.54 | 0.41 | 0.809 |
| SUPPS-P: Urgency | 0.12 | 0.16 | −0.03 | 0.60 | 0.075 | 0.05 | 0.14 | −0.17 | 0.40 | 0.459 | 0.06 | 0.14 | −0.13 | 0.42 | 0.281 | −0.02 | 0.19 | −0.41 | 0.35 | 0.871 |
| SUPPS-P: Sensation seeking | −0.04 | 0.11 | −0.30 | 0.14 | 0.451 | −0.03 | 0.10 | −0.26 | 0.15 | 0.552 | −0.05 | 0.10 | −0.28 | 0.10 | 0.345 | 0.05 | 0.14 | −0.16 | 0.38 | 0.447 |
| CHI-T: Trait compulsivity | 0.12 | 0.06 | −0.01 | 0.25 | 0.071 | 0.01 | 0.06 | −0.11 | 0.12 | 0.915 | 0.03 | 0.06 | −0.09 | 0.15 | 0.603 | −0.02 | 0.09 | −0.20 | 0.14 | 0.744 |
Note. β: Standardized coefficient; SE: Standard error; *p < 0.05, **p < 0.01, ***p < 0.001.
Cross-sectional and longitudinal multiple regression models of PUI
| Variable | Baseline (N = 294) | Baseline predicting 6-months (N = 198) | Baseline predicting 3-months (N = 206) | 3-months predicting 6-months (N = 138) | ||||||||||||||||
| β | SE | 95% CI | p | β | SE | 95% CI | p | β | SE | 95% CI | p | β | SE | 95% CI | p | |||||
| LL | UL | LL | UL | LL | UL | LL | UL | |||||||||||||
| Demographics | ||||||||||||||||||||
| Age | −0.12 | 0.08 | −0.31 | −0.02 | 0.027* | 0.01 | 0.07 | −0.12 | 0.16 | 0.796 | −0.00 | 0.08 | −0.15 | 0.15 | 0.997 | 0.09 | 0.08 | −0.04 | 0.27 | 0.144 |
| Sex (F) | −0.20 | 0.72 | −4.05 | −1.27 | <0.001*** | −0.07 | 0.65 | −2.14 | 0.40 | 0.193 | −0.06 | 0.70 | −2.21 | 0.50 | 0.237 | −0.05 | 0.71 | −1.87 | 0.84 | 0.456 |
| Neurocognition | ||||||||||||||||||||
| SST: SSRT | 0.07 | 5.03 | −3.34 | 16.87 | 0.192 | 0.04 | 4.73 | −5.37 | 13.52 | 0.392 | −0.01 | 4.68 | −10.20 | 7.82 | 0.803 | −0.01 | 5.37 | −11.35 | 8.83 | 0.804 |
| VMAC: VMAC score | −0.04 | 8.29 | −23.04 | 10.08 | 0.437 | −0.05 | 7.53 | −22.28 | 7.79 | 0.345 | −0.14 | 8.75 | −43.89 | −9.16 | 0.003** | −0.07 | 8.16 | −26.60 | 5.95 | 0.223 |
| BART: M pre-committed pumps | 0.07 | 0.02 | −0.01 | 0.07 | 0.180 | −0.09 | 0.02 | −0.08 | 0.00 | 0.075 | −0.11 | 0.02 | −0.10 | −0.01 | 0.029* | 0.07 | 0.02 | −0.01 | 0.06 | 0.236 |
| CST: Switch cost latency | 0.05 | 0.00 | −0.00 | 0.01 | 0.365 | 0.01 | 0.00 | −0.00 | 0.00 | 0.810 | −0.00 | 0.00 | −0.00 | 0.00 | 0.913 | −0.01 | 0.00 | −0.00 | 0.00 | 0.845 |
| EAT: Error awareness | −0.10 | 0.10 | −0.04 | 0.00 | 0.052 | −0.04 | 0.01 | −0.03 | 0.01 | 0.432 | −0.09 | 0.01 | −0.04 | 0.00 | 0.068 | 0.03 | 0.01 | −0.02 | 0.03 | 0.643 |
| SDT: w | −0.08 | 0.96 | −3.31 | 0.48 | 0.143 | −0.04 | 0.86 | −2.42 | 0.99 | 0.433 | −0.04 | 0.92 | −2.51 | 1.13 | 0.465 | −0.07 | 0.85 | −2.73 | 0.69 | 0.238 |
| N-Back: 3-back d’ | −0.04 | 0.29 | −0.83 | 0.34 | 0.390 | 0.03 | 0.27 | −0.38 | 0.65 | 0.603 | −0.00 | 0.29 | −0.59 | 0.55 | 0.909 | 0.04 | 0.25 | −0.31 | 0.68 | 0.484 |
| DDT: Log k | −0.07 | 0.21 | −0.72 | 0.11 | 0.158 | −0.03 | 0.19 | −0.48 | 0.28 | 0.555 | −0.04 | 0.21 | −0.60 | 0.21 | 0.365 | 0.00 | 0.20 | −0.38 | 0.40 | 0.936 |
| Covariates | ||||||||||||||||||||
| IAT | – | – | – | – | – | 0.65 | 0.05 | 0.46 | 0.67 | <0.001*** | 0.72 | 0.06 | 0.65 | 0.89 | <0.001*** | 0.77 | 0.06 | 0.54 | 0.75 | <0.001*** |
| DASS: Total score | 0.31 | 0.04 | 0.11 | 0.25 | <0.001*** | 0.02 | 0.03 | −0.06 | 0.08 | 0.791 | −0.05 | 0.04 | −0.10 | 0.05 | 0.477 | 0.04 | 0.03 | −0.04 | 0.08 | 0.520 |
| SUPPS-P: Lack of perseverance and premeditation | 0.09 | 0.22 | −0.07 | 0.80 | 0.107 | 0.01 | 0.20 | −0.36 | 0.42 | 0.837 | 0.10 | 0.22 | −0.02 | 0.84 | 0.063 | −0.05 | 0.22 | −0.57 | 0.27 | 0.492 |
| SUPPS-P: Urgency | 0.11 | 0.18 | −0.03 | 0.68 | 0.076 | 0.11 | 0.16 | −0.05 | 0.59 | 0.100 | −0.01 | 0.18 | −0.38 | 0.32 | 0.850 | 0.06 | 0.17 | −0.21 | 0.48 | 0.439 |
| SUPPS-P: Sensation seeking | −0.14 | 0.13 | −0.57 | −0.08 | 0.015* | −0.02 | 0.12 | −0.27 | 0.19 | 0.739 | 0.01 | 0.13 | −0.22 | 0.28 | 0.784 | −0.06 | 0.12 | −0.35 | 0.11 | 0.305 |
| CHI-T: Trait compulsivity | 0.22 | 0.07 | 0.12 | 0.40 | <0.001*** | 0.07 | 0.07 | −0.06 | 0.20 | 0.287 | 0.09 | 0.08 | −0.04 | 0.27 | 0.143 | 0.05 | 0.08 | −0.10 | 0.21 | 0.460 |
Note. β: Standardized coefficient; SE: Standard error; *p < 0.05, **p < 0.01, ***p < 0.001.
Ethics
The study was approved by the Monash University Human Research Ethics Committee [26088]. All subjects were informed about the study and all provided informed consent.
Results
Four hundred participants completed the baseline assessment, 283 completed the 3-month follow-up and 262 completed the 6-month follow-up. After data cleaning (Fig. 1) the final sample sizes included in analyses were as follows: n = 198 baseline predicting 6-month outcomes, n = 206 baseline predicting 3-month outcomes, and n = 138 for 3-months predicting 6-month outcomes. Participant demographics are displayed in Table 2. When comparing individuals lost to 6-month follow-up and those who were retained on characteristic at baseline, participants lost to follow-up were younger (mean age = 23.7 vs 25.4 years; rpb = −0.22, p = 0.002), had higher PPU scores (mean score = 11.2 vs 9.5; rpb = 0.18, p = 0.010), had higher delay discounting (mean log k = −4.2 vs −4.7; Hedges' g = 0.28, p = 0.025), and more males did not return for the 6-month follow-up (n = 110) than females (n = 42). When comparing individuals lost to follow-up and those retained at 3-month follow-up, participants lost to follow-up were younger (mean age = 23.3 vs 25.4 years; rpb = −0.26, p < 0.001) and showed better response inhibition (mean SSRT = 0.32 vs 0.34s; Hedges' g = −0.32, p = 0.005). Further, more females (n = 46) were lost to follow-up than males (n = 38). Individuals lost to follow-up at 6-months had higher urgency scores at 3-months than those retained (mean score = 9.8 vs 8.6; Hedge's g = 0.47, p < 0.001), and more males did not come back for 6-month follow-up (n = 20) than females (n = 17).
Flow diagrams mapping data collection, cleaning, and reasons for exclusion. Note: A is the flow diagram relevant for analyses involving the baseline sample i.e. cross-sectional model, baseline predicting 3- and 6-month longitudinal models. B is the flow diagram relevant for the longitudinal model involving predictors assessed at 3-month follow-up predicting outcomes at 6-month follow-up
Citation: Journal of Behavioral Addictions 2024; 10.1556/2006.2024.00041
Cross-sectional model
Steeper delay discounting (preference for sooner smaller rewards; β = 0.15, p = 0.007) was associated with greater AE. No further relationships were found between neurocognition and addictive behaviors.
Longitudinal multivariate models
Predicting problematic alcohol use
None of the neurocognitive variables predicted problematic alcohol use for any of the longitudinal models. These results held after correcting for serial correlation error and were replicated in regression models using multiple imputation (Tables S5 and S9).
Predicting addictive eating (AE)
The 6-month longitudinal model did not find any neurocognitive variables significantly predicted AE. The 3-month longitudinal models showed poorer performance monitoring (β = −0.16, p = 0.004) significantly predicted higher AE at 3-month follow-up and greater reward-related attentional capture (β = 0.14, p = 0.033) at 3-month follow-up significantly predicted higher AE at 6-month follow-up. All of these results held after correcting for serial correlation error (Table S6) except the relationship between reward-related attentional capture and AE at 6-month follow-up reduced to trend levels (p = 0.059). Findings were also replicated in the regression models using multiple imputation (Table S10), although the performance monitoring and reward-related attentional capture effects reduced to trend levels (95% CI [−0.01, 0.00], p = 0.066; 95% CI [−0.86, 10.33], p = 0.097).
Predicting problematic pornography use (PPU)
The 6-month longitudinal model showed less delay discounting (β = −0.16, p = 0.005) predicted higher PPU at 6-month follow-up. The 3-month longitudinal models showed neurocognition at baseline did not predict PPU at 3-month follow-up, but less delay discounting (β = −0.16, p = 0.014) at 3-month follow-up predicted higher PPU at 6-month follow-up. After correcting for serial correlation error delay discounting remained a significant predictor for the 6-month model but was no longer a significant predictor for the 3-month model (Table S7), these findings were also reflected by the regression models using multiple imputation (Table S11).
Predicting problematic use of the internet (PUI)
The 6-month longitudinal model did not find neurocognition predicted PUI. The 3-month longitudinal models showed less reward-related attentional capture (β = −0.14, p = 0.003) and less risk-taking under uncertainty (β = −0.11, p = 0.029) significantly predicted higher PUI at 3-month follow-up, but neurocognition at 3-moth follow-up did not predict PUI at 6-months. These results held after correcting for serial correlation error (Table S8). Regressions using multiple imputation replicated these findings (Table S12), although the baseline risk-taking effect reduced to trend level (95% CI [−0.09, 0.00], p = 0.062).
Discussion
The aim of the present study was to empirically test proposed trans-addiction, longitudinal mechanisms of addictive behaviors, namely problematic alcohol use, Addictive Eating (AE), Problematic Pornography Use (PPU), and Problematic Use of the Internet (PUI). We did this by first investigating the cross-sectional correlates of each addictive behavior, then evaluated the predictors of each behaviour longitudinally. Overall, it is evident that in a general community sample, different neurocognitive functions predicted different addictive behaviors.
Cross-sectional correlates of addictive behaviors
Our cross-sectional analyses, accounting for key covariates, found steeper delay discounting was associated with more AE symptoms which is in line with previous literature (VanderBroek-Stice et al., 2017). However, neurocognition was not significantly associated with any of the other addictive behaviors: problematic alcohol use, PPU, and PUI. This contrasts with previous literature that has consistently shown problematic alcohol use and PUI are associated with neurocognitive deficits, namely poorer response inhibition and less flexible updating (Ioannidis et al., 2019; Stavro et al., 2013) and steeper discounting (Amlung et al., 2017; Cheng, Ko, Sun, & Yeh, 2021). Our findings may differ from other studies given we looked at a general community sample, compared to more severe and/or clinical samples investigated previously (Cheng et al., 2021; Ioannidis et al., 2022; Stavro et al., 2013). This is particularly of note for alcohol use, given the majority of the sample engaged in no or low risk use.
Longitudinal predictors of addictive behaviors
Only a minority of neurocognitive functions assessed in this study predicted addictive behaviors longitudinally, and effect sizes were generally of small magnitude. Further, neurocognition poorly predicted addictive behaviors over a 6-month time period. With the exception of delay discounting and PPU, none of the other addictive behaviors measured at 6-month follow-up were predicted by baseline neurocognition. However, when evaluating a shorter timeframe (i.e. 3-months) we found neurocognitive functions can predict addictive behaviors. This suggests that the temporal delay between the assessment of neurocognition and behavioral outcome may impact the likelihood of finding a relationship.
Poorer performance monitoring at baseline predicted greater AE at 3-month follow-up. While similar relationships have been found cross-sectionally (Franken et al., 2018), this is the first study to demonstrate this relationship longitudinally. Previous research on performance monitoring deficits associated with AE compared samples classified with food addiction to those who were not (Franken et al., 2018; Rodrigue et al., 2018; VanderBroek-Stice et al., 2017). In contrast, the present study identified these effects dimensionally in a community sample, underscoring the potential significance of performance monitoring as an antecedent for AE dimensionally.
We also found that greater reward-related attentional bias at 3-months predicted more AE symptoms at 6-month follow-up. The present study is the first to identify that reward-related attentional capture can explain (albeit with a small effect) significant variance in future AE behavior. The VMAC task evaluates how cues indicating high-value rewards detrimentally affect overall task performance. Greater reward-related attentional bias indicates the allure of a substantial reward is so distracting that it hinders one's ability to pursue task goals. Individuals prone to developing more severe AE are perhaps more susceptible to learning of highly rewarding food cues, even when it may work against their goals or satiety.
Reward-related attentional bias was also found to predict (with a small effect) PUI but in the opposite direction. Greater freedom from distraction by reward-related cues, thus more goal-directed performance was associated with higher PUI at 3-month follow-up. Addictive behaviors can in certain contexts be purposeful, primarily driven by relief-based motivations (Köpetz et al., 2013; Liu et al., 2021). More goal-directed VMAC performance may predict more goal-directed motives to engage in internet use (Albertella, Vd Hooven, Bovens, & Wiers, 2021). We also found less risky decision-making in the face of uncertainty (BART) was shown to predict more PUI at 3-month follow-up. While studies have shown problematic alcohol and other substance use is associated with more risky BART performance (Fernie et al., 2010; Hopko et al., 2006), a recent meta-analysis showed this was not the case for PUI (Müller et al., 2023). Less risk-taking under uncertainty seen here is akin to the patterns observed in individuals with obsessive-compulsive disorder (Pushkarskaya et al., 2015), and perhaps signifies a distinct difference between PUI and other addictive behaviors, suggesting PUI could be more characteristic of compulsive disorders. A further consideration is that PUI has been described as a quasi-trait in which meaningful variance of symptom severity occurs only at the more severe end of the spectrum (Tiego et al., 2019). Consequently, samples that contain individuals with both non-problem and problem use (such as that presented herein) may impact the psychometric stability of the IAT-10 (Tiego et al., 2019).
The only relationship we found between neurocognition and PPU severity was delay discounting. This finding was unexpected and contradicts those observed in a study by Antons et al. (2019) that showed individuals who problematically use internet pornography had steeper delay discounting compared to those engaging in recreational use (Antons et al., 2019). It is unclear why we found a negative relationship between delay discounting and PPU. While our research participants exhibited a prevalence of PPU similar to that found in other general population samples (7%; Mennig et al., 2020), the majority of the individuals in our current study demonstrated either no PPU or low severity. Such a low occurrence of more severe levels of PPU may have impacted our ability to observe meaningful associations with ‘problematic’ pornography use (cf. non-problematic use). Further, it is pertinent to note that individuals who failed to return for 6-month follow-up had significantly higher PPU scores at baseline and were steeper discounters than those participants retained in analyses. Our models were likely biased by participant attrition which may explain the counterintuitive discounting findings. As such, the relationship between PPU and neurocognition is still unclear, and additional research is warranted to replicate our findings, perhaps in higher severity samples, to better understand the neurocognitive predictors of PPU.
We did not find neurocognition predicted future problematic alcohol use. This aligns with findings from a systematic review by the authors (Christensen, Albertella, et al., 2023) that showed longitudinal studies conducted in community settings have failed to consistently demonstrate neurocognition predicting future problem use (Goudriaan, Grekin, & Sher, 2011; Jones et al., 2021; Whelan et al., 2014). Instead, neurocognition is seemingly more relevant at riskier levels of alcohol use. For example, poorer response inhibition has been shown to predict the development of alcohol dependence in a sample of already heavily drinking individuals (Rubio et al., 2008). It may be that neurocognitive impairments arising from exposure to alcohol use, or an interaction between the two, predict problematic future use. For example, Peeters and colleagues (2014) found that alcohol use at baseline predicted flexible updating performance, which, in turn, predicted further alcohol use outcomes. Rubio and colleague's sample were already at risk of transitioning to dependence (i.e. currently drinking heavily), whereas general population samples without any risk-indictors (i.e. family history, current regular use etc.) may never go on to have problems. Finally, neurocognition may be more relevant when individuals are attempting to change or control their drinking behavior (Albertella et al., 2021). In the current study, participants were not required to change their drinking. Without the motivation to change, some have argued that it is unlikely an individual's alcohol use would be predicted by cognitive disposition (Albertella et al., 2021).
A strength of this study is our broad, longitudinal evaluation of neurocognitive functions. Each neurocognitive domain was selected based on theoretical frameworks and expert-endorsement, attesting to their relevance in addiction. This approach allowed for both a comprehensive and targeted evaluation of neurocognitive functions and their role in predicting addictive behaviors. Another strength is that we investigated multiple addictive behavior types in the same sample. Given addictive behaviors often co-occur (Christensen, Albertella, et al., 2023; Ford & Håkansson, 2020), investigating influence of neurocognitive predictors on various addictive behaviors within the same group, helps us understand whether there are common factors that predict addictive behaviors across different types or if these predictors are specific to each behavior.
Limitations and future directions
While our choice to focus on a community sample contributes to the dimensional evaluation of addictive behavior, it is important to recognise that the selected sample exhibited relatively low levels of engagement in addictive behaviors, specifically in the areas of AE, alcohol use, and PPU. This limitation may have impeded our ability to identify more subtle associations. Importantly, our failure to identify trans-diagnostic neurocognitive predictors does not mean they do not exist. Rather, these functions may be more pertinent at later stages of addictive behavior and could play a crucial role in treatment success (Domínguez-Salas et al., 2016), although this line of reasoning remains speculative. Notably, no prior studies have conducted such an extensive neurocognitive evaluation in higher-severity samples or within the context of treatment success.
An additional limitation is the complexity of our regression models in relation to the sample size. This is particularly relevant to our 3-month predicting 6-month longitudinal model which, was underpowered to detect small effects due to participant drop-out across the course of the study. Another notable limitation is that participant drop-out resulted in distinct differences between those individuals who completed the baseline assessment and those who attended the 3-month follow-up. This made it difficult to compare the 3- and 6-month longitudinal analyses given the baseline sample was slightly different in each analysis, and likely explains our inability to replicate the longitudinal results. Participant drop-out also has the potential to bias longitudinal models, however, our multiple imputation analyses largely replicated the findings obtained from the raw data, statistically mitigating the significance of this concern to a degree. Further, the effect size differences comparing individuals who were retained versus lost to follow-up were only small.
It is also important to acknowledge that our inability to identify neurocognitive predictors may have been a result of single task-based measurement error, which is inherent in investigations that opt for individual paradigms to measure what are considered to be latent mechanisms (Enkavi et al., 2019; Hedge et al., 2018). Utilising multiple tasks that assess the same underlying construct (e.g. Go/No-Go task and SST for inhibitory control) and deriving latent variables that represent concurrence among measures (Goschke, 2014; Verdejo-Garcia & Albein-Urios, 2021) may provide a more robust and comprehensive way to measure and understand neurocognitive processes. Finally, once-off assessments of neurocognitive functions at each time-point are perhaps not the most appropriate method of capturing what are dynamic processes, sensitive to both intrinsic (physiological processes) and extrinsic (environmental) factors (Schmitter-Edgecombe, Sumida, & Cook, 2020). Ecological Momentary Assessment (EMA) paradigms have been suggested to account for this, providing multiple daily snapshots of neurocognition, able to track within-person fluctuations in different contexts and under different psychological states and environmental situations (Sliwinski et al., 2018).
Conclusion
In conclusion, we did not identify a core set of specific neurocognitive functions that reliably predicted addictive behaviors across multiple behavior types, which is in itself a significant finding from a theoretical standpoint. However, we showed that reward-related neurocognitive processes were implicated across each non-substance addictive behavior, but in different ways. Our findings suggest that there may be partially distinct neurocognitive mechanisms contributing to addiction, depending on the specific addictive behavior under consideration. However, more work should be done to interrogate these differences, particularly focusing on reward-related functions. Our findings would also benefit from being replicated in studies that specifically recruit individuals with more severe levels of addictive behavior. This could lead to a better understanding of the types/profiles of individuals who may be at risk of developing specific types of addiction and inform the development of early identification and intervention clinical and public health strategies for specific addictive behaviors.
Funding sources
Funding for this study was provided by the National Health and Medical Research Council (NHMRC) Project Grant [APP1162031] and NHMRC Medical Research Future Fund Investigator Grant [APP1193946]. This research was also supported by an Australian Government Research Training Program (RTP) Scholarship.
Authors' contribution
EC: study concept and design, data collection, statistical analysis and interpretation of the data, writing - original draft and editing. LA: study design, interpretation of study results, writing - review and editing. SRC: writing - review & editing. MB: software development, writing - review and editing. CS: software development. JEG: writing - review and editing. MY: obtained funding, writing - review and editing. RSCL: supervision, obtained funding, study concept and design, interpretation of data, writing - review and editing.
Conflict of interest
EC and MB were supported by an Australian Government Research Training Program (RTP) Scholarship. CS received no financial support for the research, authorship, and/or publication of this article. LA has received funding from the David Winston Turner Endowment Fund. RSCL was supported by an NHMRC Investigator Grant funded by the Medical Research Future Fund (APP1193946). MY receives funding from: government funding bodies such as the NHMRC, Australian Research Council (ARC), Australian Defence Science and Technology (DST), the Department of Industry, Innovation and Science (DIIS), the National Institutes of Health (NIH, USA); philanthropic donations from the David Winston Turner Endowment Fund, Wilson Foundation; sponsored Investigator-Initiated trials including Incannex Healthcare Ltd. MY has also received payments in relation to court-, expert witness-, and/or expert review-reports. These funding sources had no role in the data analysis, presentation, or interpretation and write-up of the data. MY also sits on the Advisory Boards of: Centre of The Urban Mental Health, University of Amsterdam; Monash Biomedical Imaging Centre; and Enosis Therapeutics. JEG has received research grant support from NIDA, and from Biohaven, Janssen, and Boehringer Ingelheim Pharmaceuticals. He receives yearly compensation from Springer Publishing for acting as Editor-in-Chief of the Journal of Gambling Studies and has received royalties from Oxford University Press, American Psychiatric Publishing, Inc., Norton Press, and McGraw Hill. SRC and LA receive honoraria for editorial work at Elsevier. The above funding sources have had no role in the present study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit this paper for publication.
Acknowledgement
We wish to thank all of the participants who generously devoted their time to this research study. We also wish to thank the following individuals for their assistance throughout recruitment and data collection: Lara Piccolli, Dr. Chang Liu, Daniel Koay, Emma Moon, Sashka Samarawickrama, and Hanbing Wu.
Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1556/2006.2024.00041.
References
Adams, R. C., Sedgmond, J., Maizey, L., Chambers, C. D., & Lawrence, N. S. (2019). Food addiction: Implications for the diagnosis and treatment of overeating. Nutrients, 11(9), 2086. https://doi.org/10.3390/nu11092086.
Albertella, L., Pelley, M. E. L., Chamberlain, S. R., Westbrook, F., Fontenelle, L. F., Segrave, R., … Yücel, M. (2019). Reward-related attentional capture is associated with severity of addictive and obsessive-compulsive behaviors. Psychology of Addictive Behaviors, 33(5), 495–502. https://doi.org/10.1037/adb0000484.
Albertella, L., Vd Hooven, J., Bovens, R., & Wiers, R. W. (2021). Reward-related attentional capture predicts non-abstinence during a one-month abstinence challenge. Addictive Behaviors, 114, 106745. https://doi.org/10.1016/j.addbeh.2020.106745.
Albertella, L., Watson, P., Yücel, M., & Le Pelley, M. E. (2019). Persistence of value-modulated attentional capture is associated with risky alcohol use. Addictive Behaviors Reports, 10, 100195. https://doi.org/10.1016/j.abrep.2019.100195.
Amlung, M., Vedelago, L., Acker, J., Balodis, I., & MacKillop, J. (2017). Steep delay discounting and addictive behavior: A meta-analysis of continuous associations. Addiction (Abingdon, England), 112(1), 51–62. https://doi.org/10.1111/add.13535.
Andrews, D. W. K., & Monahan, J. C. (1992). An improved heteroskedasticity and autocorrelation consistent covariance matrix estimator. Econometrica, 60(4), 953. https://doi.org/10.2307/2951574.
Antons, S., Mueller, S. M., Wegmann, E., Trotzke, P., Schulte, M. M., & Brand, M. (2019). Facets of impulsivity and related aspects differentiate among recreational and unregulated use of Internet pornography. Journal of Behavioral Addictions, 8(2), 223–233. https://doi.org/10.1556/2006.8.2019.22.
Audrain-McGovern, J., Rodriguez, D., Epstein, L. H., Cuevas, J., Rodgers, K., & Wileyto, E. P. (2009). Does delay discounting play an etiological role in smoking or is it a consequence of smoking? Drug and Alcohol Dependence, 103(3), 99–106. https://doi.org/10.1016/j.drugalcdep.2008.12.019.
Band, G. P. H., Van Der Molen, M. W., & Logan, G. D. (2003). Horse-race model simulations of the stop-signal procedure. Acta Psychologica, 112(2), 105–142. https://doi.org/10.1016/S0001-6918(02)00079-3.
Bechara, A. (2005). Decision making, impulse control and loss of willpower to resist drugs: A neurocognitive perspective. Nature Neuroscience, 8(11), 1458–1463. https://doi.org/10.1038/nn1584.
Bernhardt, N., Nebe, S., Pooseh, S., Sebold, M., Sommer, C., Birkenstock, J., … Smolka, M. N. (2017). Impulsive decision making in young adult social drinkers and detoxified alcohol-dependent patients: A cross-sectional and longitudinal study. Alcoholism: Clinical and Experimental Research, 41(10), 1794–1807. https://doi.org/10.1111/acer.13481.
Billieux, J., Heeren, A., Rochat, L., Maurage, P., Bayard, S., Bet, R., … … Baggio, S. (2021). Positive and negative urgency as a single coherent construct: Evidence from a large‐scale network analysis in clinical and non‐clinical samples. Journal of Personality, 89(6), 1252–1262. https://doi.org/10.1111/jopy.12655.
Billieux, J., Rochat, L., Ceschi, G., Carré, A., Offerlin-Meyer, I., Defeldre, A.-C., … Van Der Linden, M. (2012). Validation of a short French version of the UPPS-P impulsive behavior scale. Comprehensive Psychiatry, 53(5), 609–615. https://doi.org/10.1016/j.comppsych.2011.09.001.
Bø, R., Billieux, J., Gjerde, L. C., Eilertsen, E. M., & Landrø, N. I. (2017). Do executive functions predict binge-drinking patterns? Evidence from a longitudinal study in young adulthood. Frontiers in Psychology, 08. https://doi.org/10.3389/fpsyg.2017.00489.
Bőthe, B., Tóth-Király, I., Demetrovics, Z., & Orosz, G. (2021). The short version of the problematic pornography consumption scale (PPCS-6): A reliable and valid measure in general and treatment-seeking populations. Journal of Sex Research, 58(3), 342–352. https://doi.org/10.1080/00224499.2020.1716205.
Burmeister, J. M., Hinman, N., Koball, A., Hoffmann, D. A., & Carels, R. A. (2013). Food addiction in adults seeking weight loss treatment. Implications for psychosocial health and weight loss. Appetite, 60, 103–110. https://doi.org/10.1016/j.appet.2012.09.013.
Burrows, T., Kay-Lambkin, F., Pursey, K., Skinner, J., & Dayas, C. (2018). Food addiction and associations with mental health symptoms: A systematic review with meta-analysis. Journal of Human Nutrition and Dietetics, 31(4), 544–572. https://doi.org/10.1111/jhn.12532.
Camilleri, C., Perry, J. T., & Sammut, S. (2021). Compulsive internet pornography use and mental health: A cross-sectional study in a sample of university students in the United States. Frontiers in Psychology, 11, 613244. https://doi.org/10.3389/fpsyg.2020.613244.
Chamberlain, S. R., & Grant, J. E. (2018). Initial validation of a transdiagnostic compulsivity questionnaire: The cambridge-chicago compulsivity trait scale. CNS Spectrums, 23(5), 340–346. https://doi.org/10.1017/S1092852918000810.
Chapman, C., Slade, T., Hunt, C., & Teesson, M. (2015). Delay to first treatment contact for alcohol use disorder. Drug and Alcohol Dependence, 147, 116–121. https://doi.org/10.1016/j.drugalcdep.2014.11.029.
Chen, H., Mojtahedzadeh, N., Belanger, M. J., Nebe, S., Kuitunen-Paul, S., Sebold, M., … Smolka, M. N. (2021). Model-based and model-free control predicts alcohol consumption developmental trajectory in young adults: A 3-year prospective study. Biological Psychiatry, 89(10), 980–989. https://doi.org/10.1016/j.biopsych.2021.01.009.
Cheng, Y. S., Ko, H.-C., Sun, C.-K., & Yeh, P.-Y. (2021). The relationship between delay discounting and internet addiction: A systematic review and meta-analysis. Addictive Behaviors, 114, 106751. https://doi.org/10.1016/j.addbeh.2020.106751.
Christensen, E., Albertella, L., Chamberlain, S. R., Brydevall, M., Suo, C., Grant, J. E., … Lee, R. S. C. (2023). The neurocognitive correlates of non-substance addictive behaviors. Addictive Behaviors, 150, 107904. https://doi.org/10.1016/j.addbeh.2023.107904.
Christensen, E., Brydevall, M., Albertella, L., Samarawickrama, S. K., Yücel, M., & Lee, R. S. C. (2023). Neurocognitive predictors of addiction-related outcomes: A systematic review of longitudinal studies. Neuroscience & Biobehavioral Reviews, 152, 105295. https://doi.org/10.1016/j.neubiorev.2023.105295.
Cousijn, J., van Benthem, P., van der Schee, E., & Spijkerman, R. (2015). Motivational and control mechanisms underlaying adolescent cannabis use disorders: A prospective study. Developmental Cognitive Neuroscience, 16, 36–45. https://doi.org/10.1016/j.dcn.2015.04.001.
Criaud, M., & Boulinguez, P. (2013). Have we been asking the right questions when assessing response inhibition in go/no-go tasks with fMRI? A meta-analysis and critical review. Neuroscience & Biobehavioral Reviews, 37(1), 11–23. https://doi.org/10.1016/j.neubiorev.2012.11.003.
Cyders, M. A., Littlefield, A. K., Coffey, S., & Karyadi, K. A. (2014). Examination of a short English version of the UPPS-P impulsive behavior scale. Addictive Behaviors, 39(9), 1372–1376. https://doi.org/10.1016/j.addbeh.2014.02.013.
Domínguez-Salas, S., Díaz-Batanero, C., Lozano-Rojas, O. M., & Verdejo-García, A. (2016). Impact of general cognition and executive function deficits on addiction treatment outcomes: Systematic review and discussion of neurocognitive pathways. Neuroscience & Biobehavioral Reviews, 71, 772–801. https://doi.org/10.1016/j.neubiorev.2016.09.030.
Eisenberg, I. W., Bissett, P. G., Enkavi, A. Z., Li, J., MacKinnon, D. P., Marsch, L. A., & Poldrack, R. A. (2019). Uncovering the structure of self-regulation through data-driven ontology discovery. Nature Communications, 10(1). https://doi.org/10.1038/s41467-019-10301-1.
Enkavi, A. Z., Eisenberg, I. W., Bissett, P. G., Mazza, G. L., MacKinnon, D. P., Marsch, L. A., & Poldrack, R. A. (2019). Large-scale analysis of test–retest reliabilities of self-regulation measures. Proceedings of the National Academy of Sciences, 116(12), 5472–5477. https://doi.org/10.1073/pnas.1818430116.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175–191.
Fernández-Artamendi, S., Martínez-Loredo, V., Grande-Gosende, A., Simpson, I. C., & Fernández-Hermida, J. R. (2018). What predicts what? Self-reported and behavioral impulsivity and high-risk patterns of alcohol use in Spanish early adolescents: A 2-year longitudinal study. Alcoholism: Clinical and Experimental Research, 42(10), 2022–2032. https://doi.org/10.1111/acer.13852.
Fernie, G., Cole, J. C., Goudie, A. J., & Field, M. (2010). Risk-taking but not response inhibition or delay discounting predict alcohol consumption in social drinkers. Drug and Alcohol Dependence, 112(1–2), 54–61. https://doi.org/10.1016/j.drugalcdep.2010.05.011.
Field, A., Miles, J., & Field, Z. (2012). Discovering statistics using R. SAGE Publications.
Fineberg, N., Demetrovics, Z., Stein, D., Ioannidis, K., Potenza, M., Grünblatt, E., … Chamberlain, S. (2018). Manifesto for a European research network into problematic usage of the internet. European Neuropsychopharmacology, 28(11), 1232–1246. https://doi.org/10.1016/j.euroneuro.2018.08.004.
Floros, G. D., & Ioannidis, K. (2021). Editorial: The impact of online addiction on general health, well-being and associated societal costs. Frontiers in Public Health, 9, 676498. https://doi.org/10.3389/fpubh.2021.676498.
Ford, M., & Håkansson, A. (2020). Problem gambling, associations with comorbid health conditions, substance use, and behavioural addictions: Opportunities for pathways to treatment. Plos One, 15(1), e0227644. https://doi.org/10.1371/journal.pone.0227644.
Fox, J., & Weisberg, S. (2019). An R Companion to applied regression (3rd ed.). Thousand Oaks CA: Sage. https://socialsciences.mcmaster.ca/jfox/Books/Companion/.
Franken, I. H. A., Nijs, I. M. T., Toes, A., & van der Veen, F. M. (2018). Food addiction is associated with impaired performance monitoring. Biological Psychology, 131, 49–53. https://doi.org/10.1016/j.biopsycho.2016.07.005.
Friedman, N. P., Miyake, A., Young, S. E., DeFries, J. C., Corley, R. P., & Hewitt, J. K. (2008). Individual differences in executive functions are almost entirely genetic in origin. Journal of Experimental Psychology: General, 137(2), 201–225. https://doi.org/10.1037/0096-3445.137.2.201.
Fröhner, J. H., Ripke, S., Jurk, S., Li, S., Banaschewski, T., Bokde, A. L. W., … the IMAGEN Consortium (2022). Associations of delay discounting and drinking trajectories from ages 14 to 22. Alcoholism: Clinical and Experimental Research, 46(4), 667–681. https://doi.org/10.1111/acer.14799.
Goschke, T. (2014). Dysfunctions of decision‐making and cognitive control as transdiagnostic mechanisms of mental disorders: Advances, gaps, and needs in current research. International Journal of Methods in Psychiatric Research, 23(S1), 41–57. https://doi.org/10.1002/mpr.1410.
Goudriaan, A. E., Grekin, E. R., & Sher, K. J. (2011). Decision making and response inhibition as predictors of heavy alcohol use: A prospective study. Alcoholism: Clinical and Experimental Research, 35(6), 1050–1057. https://doi.org/10.1111/j.1530-0277.2011.01437.x.
Grant, B. F., Goldstein, R. B., Saha, T. D., Chou, S. P., Jung, J., Zhang, H., … Hasin, D. S. (2015). Epidemiology of DSM-5 alcohol use disorder: Results from the national epidemiologic survey on alcohol and related conditions III. JAMA Psychiatry, 72(8), 757. https://doi.org/10.1001/jamapsychiatry.2015.0584.
Hasin, D. S., O’Brien, C. P., Auriacombe, M., Borges, G., Bucholz, K., Budney, A., … Grant, B. F. (2013). DSM-5 criteria for substance use disorders: Recommendations and rationale. American Journal of Psychiatry, 170(8), 834–851. https://doi.org/10.1176/appi.ajp.2013.12060782.
Hedge, C., Powell, G., & Sumner, P. (2018). The reliability paradox: Why robust cognitive tasks do not produce reliable individual differences. Behavior Research Methods, 50(3), 1166–1186. https://doi.org/10.3758/s13428-017-0935-1.
Hester, R., Simões-Franklin, C., & Garavan, H. (2007). Post-error behavior in active cocaine users: Poor awareness of errors in the presence of intact performance adjustments. Neuropsychopharmacology : Official Publication of the American College of Neuropsychopharmacology, 32(9), 1974–1984. https://doi.org/10.1038/sj.npp.1301326.
Hesterberg, T. C. (2015). What teachers should know about the bootstrap: Resampling in the undergraduate statistics curriculum. The American Statistician, 69(4), 371–386. https://doi.org/10.1080/00031305.2015.1089789.
Holm, S. (1979). A simple sequentially rejective multiple test procedure. Scand J Statist, 65–70.
Hopko, D. R., Lejuez, C. W., Daughters, S. B., Aklin, W. M., Osborne, A., Simmons, B. L., & Strong, D. R. (2006). Construct validity of the balloon analogue risk task (BART): Relationship with MDMA use by inner-city drug users in residential treatment. Journal of Psychopathology and Behavioral Assessment, 28(2), 95–101. https://doi.org/10.1007/s10862-006-7487-5.
Imperatori, C., Fabbricatore, M., Lester, D., Manzoni, G. M., Castelnuovo, G., Raimondi, G., & Innamorati, M. (2019). Psychometric properties of the modified Yale food addiction scale version 2.0 in an Italian non-clinical sample. Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity, 24(1), 37–45. https://doi.org/10.1007/s40519-018-0607-x.
Inquisit 5 [Computer software] (2018). Retrieved from https://www.millisecond.com.
Insel, T., Cuthbert, B., Garvey, M., Heinssen, R., Pine, D. S., Quinn, K., … Wang, P. (2010). Research domain criteria (RDoC): Toward a new classification framework for research on mental disorders. American Journal of Psychiatry, 167(7), 748–751. https://doi.org/10.1176/appi.ajp.2010.09091379.
Ioannidis, K., Grant, J. E., & Chamberlain, S. R. (2022). Problematic usage of the internet and cognition. Current Opinion in Behavioral Sciences, 44, 101104. https://doi.org/10.1016/j.cobeha.2022.101104.
Ioannidis, K., Hook, R., Goudriaan, A. E., Vlies, S., Fineberg, N. A., Grant, J. E., & Chamberlain, S. R. (2019). Cognitive deficits in problematic internet use: Meta-analysis of 40 studies. British Journal of Psychiatry, 215(5), 639–646. https://doi.org/10.1192/bjp.2019.3.
Jeromin, F., Nyenhuis, N., & Barke, A. (2016). Attentional bias in excessive Internet gamers: Experimental investigations using an addiction Stroop and a visual probe. Journal of Behavioral Addictions, 5(1), 32–40. https://doi.org/10.1556/2006.5.2016.012.
Johnston, R., Jones, K., & Manley, D. (2018). Confounding and collinearity in regression analysis: A cautionary tale and an alternative procedure, illustrated by studies of British voting behaviour. Quality & Quantity, 52(4), 1957–1976. https://doi.org/10.1007/s11135-017-0584-6.
Jones, C. B., Meier, M. H., Corbin, W. E., & Chassin, L. (2021). Adolescent executive cognitive functioning and trait impulsivity as predictors of young-adult risky drinking and alcohol-related problems. Psychology of Addictive Behaviors, 35(2), 187–198. https://doi.org/10.1037/adb0000636.
Kirby, K. N., Petry, N. M., & Bickel, W. K. (1999). Heroin addicts have higher discount rates for delayed rewards than non-drug-using controls. Journal of Experimental Psychology: General, 128(1), 78–87. https://doi.org/10.1037/0096-3445.128.1.78.
Köpetz, C. E., Lejuez, C. W., Wiers, R. W., & Kruglanski, A. W. (2013). Motivation and self-regulation in addiction: A call for convergence. Perspectives on Psychological Science, 8(1), 3–24. https://doi.org/10.1177/1745691612457575.
Kräplin, A., Höfler, M., Pooseh, S., Wolff, M., Krönke, K.-M., Goschke, T., … Smolka, M. N. (2020). Impulsive decision-making predicts the course of substance-related and addictive disorders. Psychopharmacology, 237(9), 2709–2724. https://doi.org/10.1007/s00213-020-05567-z.
Kumar, P., Patel, V. K., Bhatt, R. B., Vasavada, D. A., Sangma, R. D., & Tiwari, D. S. (2021). Prevalence of problematic pornography use and attitude toward pornography among the undergraduate medical students. Journal of Psychosexual Health, 3(1), 29–36. https://doi.org/10.1177/2631831821989677.
Kuss, D., Griffiths, M., Karila, L., & Billieux, J. (2014). Internet addiction: A systematic review of epidemiological research for the last decade. Current Pharmaceutical Design, 20(25), 4026–4052. https://doi.org/10.2174/13816128113199990617.
Le Pelley, M. E., Pearson, D., Griffiths, O., & Beesley, T. (2015). When goals conflict with values: Counterproductive attentional and oculomotor capture by reward-related stimuli. Journal of Experimental Psychology: General, 144(1), 158–171. https://doi.org/10.1037/xge0000037.
Lee, R. S. C., Albertella, L., Christensen, E., Suo, C., Segrave, R. A., Brydevall, M., … Yücel, M. (2023). A novel, expert-endorsed, neurocognitive digital assessment tool for addictive disorders: Development and validation study. Journal of Medical Internet Research, 25, e44414. https://doi.org/10.2196/44414.
Liu, C., Rotaru, K., Lee, R. S. C., Tiego, J., Suo, C., Yücel, M., & Albertella, L. (2021). Distress-driven impulsivity interacts with cognitive inflexibility to determine addiction-like eating. Journal of Behavioral Addictions, 10(3), 534–539. https://doi.org/10.1556/2006.2021.00027.
Mechelmans, D. J., Irvine, M., Banca, P., Porter, L., Mitchell, S., Mole, T. B., … Voon, V. (2014). Enhanced attentional bias towards sexually explicit cues in individuals with and without compulsive sexual behaviours. Plos One, 9(8), e105476. https://doi.org/10.1371/journal.pone.0105476.
Mennig, M., Tennie, S., & Barke, A. (2020). A psychometric approach to assessments of problematic use of online pornography and social networking sites based on the conceptualizations of internet gaming disorder. BMC Psychiatry, 20(1), 318. https://doi.org/10.1186/s12888-020-02702-0.
Meule, A., & Gearhardt, A. N. (2019). Ten years of the Yale food addiction scale: A review of version 2.0. Current Addiction Reports, 6(3), 218–228. https://doi.org/10.1007/s40429-019-00261-3.
Müller, S. M., Antons, S., Wegmann, E., Ioannidis, K., King, D. L., Potenza, M. N., … Brand, M. (2023). A systematic review and meta-analysis of risky decision-making in specific domains of problematic use of the internet: Evidence across different decision-making tasks. Neuroscience & Biobehavioral Reviews, 152, 105271. https://doi.org/10.1016/j.neubiorev.2023.105271.
Neal, D. J., & Simons, J. S. (2007). Inference in regression models of heavily skewed alcohol use data: A comparison of ordinary least squares, generalized linear models, and bootstrap resampling. Psychology of Addictive Behaviors, 21(4), 441–452. https://doi.org/10.1037/0893-164X.21.4.441.
Nikolaidou, M., Fraser, D. S., & Hinvest, N. (2019). Attentional bias in Internet users with problematic use of social networking sites. Journal of Behavioral Addictions, 8(4), 733–742. https://doi.org/10.1556/2006.8.2019.60.
Noorbakhsh, S., Shams, J., Faghihimohamadi, M., Zahiroddin, H., Hallgren, M., & Kallmen, H. (2018). Psychometric properties of the Alcohol Use Disorders Identification Test (AUDIT) and prevalence of alcohol use among Iranian psychiatric outpatients. Substance Abuse Treatment, Prevention, and Policy, 13, 1–8. https://link.springer.com/article/10.1186/s13011-018-0141-x.
Odlaug, B. L., Chamberlain, S. R., Kim, S. W., Schreiber, L. R. N., & Grant, J. E. (2011). A neurocognitive comparison of cognitive flexibility and response inhibition in gamblers with varying degrees of clinical severity. Psychological Medicine, 41(10), 2111–2119. https://doi.org/10.1017/S0033291711000316.
Pan, Y.-C., Chiu, Y.-C., & Lin, Y.-H. (2020). Systematic review and meta-analysis of epidemiology of internet addiction. Neuroscience & Biobehavioral Reviews, 118, 612–622. https://doi.org/10.1016/j.neubiorev.2020.08.013.
Peeters, M., Monshouwer, K., Janssen, T., Wiers, R. W., & Vollebergh, W. A. M. (2014). Working memory and alcohol use in at-risk adolescents: A 2-year follow-up. Alcoholism: Clinical and Experimental Research, 38(4), 1176–1183. https://doi.org/10.1111/acer.12339.
Pushkarskaya, H., Tolin, D., Ruderman, L., Kirshenbaum, A., Kelly, J. M., Pittenger, C., & Levy, I. (2015). Decision-making under uncertainty in obsessive–compulsive disorder. Journal of Psychiatric Research, 69, 166–173. https://doi.org/10.1016/j.jpsychires.2015.08.011.
Qualtrics software, Version 08.23 (2023). Copyright © 2023 Qualtrics. Qualtrics and all other Qualtrics product or service names are registered trademarks or trademarks of Qualtrics. Provo, UT, USA. https://www.qualtrics.com.
R Core Team (2021). R: A language and environment for statistical ## computing. Vienna, Austria: R Foundation for Statistical Computing. URL https://www.R-project.org/.
Ragland, J. D., Turetsky, B. I., Gur, R. C., Gunning-Dixon, F., Turner, T., Schroeder, L., … Gur, R. E. (2002). Working memory for complex figures: An fMRI comparison of letter and fractal n-back tasks. Neuropsychology, 16(3), 370–379. PMC4332798 and PMID: 12146684.
Raj, K., Segrave, R., Tiego, J., Verdéjo-Garcia, A., & Yücel, M. (2022). Problematic Use of the Internet among Australian university students: Prevalence and profile. Computers in Human Behavior Reports, 8, 100243. https://doi.org/10.1016/j.chbr.2022.100243.
Ridderinkhof, K. R., Van Den Wildenberg, W. P. M., Segalowitz, S. J., & Carter, C. S. (2004). Neurocognitive mechanisms of cognitive control: The role of prefrontal cortex in action selection, response inhibition, performance monitoring, and reward-based learning. Brain and Cognition, 56(2), 129–140. https://doi.org/10.1016/j.bandc.2004.09.016.
Rodrigue, C., Ouellette, A.-S., Lemieux, S., Tchernof, A., Biertho, L., & Bégin, C. (2018). Executive functioning and psychological symptoms in food addiction: A study among individuals with severe obesity. Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity, 23(4), 469–478. https://doi.org/10.1007/s40519-018-0530-1.
Rubio, G., Jiménez, M., Rodríguez-Jiménez, R., Martínez, I., Ávila, C., Ferre, F., … Palomo, T. (2008). The role of behavioral impulsivity in the development of alcohol dependence: A 4-year follow-up study. Alcoholism: Clinical and Experimental Research, 32(9), 1681–1687. https://doi.org/10.1111/j.1530-0277.2008.00746.x.
Saunders, J. B., Aasland, O. G., Babor, T. F., De La Fuente, J. R., & Grant, M. (1993). Development of the alcohol use disorders identification test (AUDIT): WHO collaborative Project on early detection of persons with harmful alcohol consumption-II. Addiction, 88(6), 791–804. https://doi.org/10.1111/j.1360-0443.1993.tb02093.x.
Schmitter-Edgecombe, M., Sumida, C., & Cook, D. J. (2020). Bridging the gap between performance-based assessment and self-reported everyday functioning: An ecological momentary assessment approach. The Clinical Neuropsychologist, 34(4), 678–699. https://doi.org/10.1080/13854046.2020.1733097.
Schulte, E. M., & Gearhardt, A. N. (2017). Development of the modified Yale food addiction scale version 2.0: Development of the mYFAS 2.0. European Eating Disorders Review, 25(4), 302–308. https://doi.org/10.1002/erv.2515.
Shankman, S. A., Lewinsohn, P. M., Klein, D. N., Small, J. W., Seeley, J. R., & Altman, S. E. (2009). Subthreshold conditions as precursors for full syndrome disorders: A 15-year longitudinal study of multiple diagnostic classes. Journal of Child Psychology and Psychiatry, 50(12), 1485–1494. https://doi.org/10.1111/j.1469-7610.2009.02117.x.
Sjoberg, E. A., & Cole, G. G. (2018). Sex differences on the go/no-go test of inhibition. Archives of Sexual Behavior, 47(2), 537–542. https://doi.org/10.1007/s10508-017-1010-9.
Sliwinski, M. J., Mogle, J. A., Hyun, J., Munoz, E., Smyth, J. M., & Lipton, R. B. (2018). Reliability and validity of ambulatory cognitive assessments. Assessment, 25(1), 14–30. https://doi.org/10.1177/1073191116643164.
Smith, J. L., Mattick, R. P., Jamadar, S. D., & Iredale, J. M. (2014). Deficits in behavioural inhibition in substance abuse and addiction: A meta-analysis. Drug and Alcohol Dependence, 145, 1–33. https://doi.org/10.1016/j.drugalcdep.2014.08.009.
Solmi, M., Radua, J., Olivola, M., Croce, E., Soardo, L., Salazar De Pablo, G., … Fusar-Poli, P. (2022). Age at onset of mental disorders worldwide: Large-scale meta-analysis of 192 epidemiological studies. Molecular Psychiatry, 27(1), 281–295. https://doi.org/10.1038/s41380-021-01161-7.
Starcke, K., Wiesen, C., Trotzke, P., & Brand, M. (2016). Effects of acute laboratory stress on executive functions. Frontiers in Psychology, 7. https://doi.org/10.3389/fpsyg.2016.00461.
Stavro, K., Pelletier, J., & Potvin, S. (2013). Widespread and sustained cognitive deficits in alcoholism: A meta-analysis: Alcoholism and cognition. Addiction Biology, 18(2), 203–213. https://doi.org/10.1111/j.1369-1600.2011.00418.x.
Szabó, M. (2010). The short version of the Depression Anxiety Stress Scales (DASS‐21): Factor structure in a young adolescent sample. Journal of Adolescence, 33(1), 1–8. https://doi.org/10.1016/j.adolescence.2009.05.014.
Tennant, P. W. G., Arnold, K. F., Ellison, G. T. H., & Gilthorpe, M. S. (2021). Analyses of ‘change scores’ do not estimate causal effects in observational data. International Journal of Epidemiology, 51(5), 1604–1615. https://doi.org/10.1093/ije/dyab050.
Thulin, M. (2023). _boot.pval: Bootstrap p-Values_. R package version 0.5. https://CRAN.R-project.org/package=boot.pval.
Tiego, J., Lochner, C., Ioannidis, K., Brand, M., Stein, D. J., Yücel, M., … Chamberlain, S. R. (2019). Problematic use of the Internet is a unidimensional quasi-trait with impulsive and compulsive subtypes. BMC Psychiatry, 19(1), 348. https://doi.org/10.1186/s12888-019-2352-8.
Tiego, J., Lochner, C., Ioannidis, K., Brand, M., Stein, D. J., Yücel, M., … Chamberlain, S. R. (2021). Measurement of the problematic usage of the Internet unidimensional quasitrait continuum with item response theory. Psychological Assessment, 33(7), 652–671. https://doi.org/10.1037/pas0000870.
Tiego, J., Trender, W., Hellyer, P. J., Grant, J. E., Hampshire, A., & Chamberlain, S. R. (2023). Measuring compulsivity as a self-reported multidimensional transdiagnostic construct: Large-scale ( N = 182,000) validation of the Cambridge–Chicago compulsivity trait scale. Assessment, 107319112211490. https://doi.org/10.1177/10731911221149083.
van Hemel-Ruiter, M. E., de Jong, P. J., Ostafin, B. D., & Oldehinkel, A. J. (2015). Reward-related attentional bias and adolescent substance use: A prognostic relationship? Plos One, 10(3), e0121058. https://doi.org/10.1371/journal.pone.0121058.
VanderBroek-Stice, L., Stojek, M. K., Beach, S. R. H., vanDellen, M. R., & MacKillop, J. (2017). Multidimensional assessment of impulsivity in relation to obesity and food addiction. Appetite, 112, 59–68. https://doi.org/10.1016/j.appet.2017.01.009.
Verbruggen, F., Aron, A. R., Band, G. P., Beste, C., Bissett, P. G., Brockett, A. T., … Boehler, C. N. (2019). A consensus guide to capturing the ability to inhibit actions and impulsive behaviors in the stop-signal task. eLife, 8, e46323. https://doi.org/10.7554/eLife.46323.
Verdejo-Garcia, A., & Albein-Urios, N. (2021). Impulsivity traits and neurocognitive mechanisms conferring vulnerability to substance use disorders. Neuropharmacology, 183, 108402. https://doi.org/10.1016/j.neuropharm.2020.108402.
Volkow, N. D., & Boyle, M. (2018). Neuroscience of addiction: Relevance to prevention and treatment. American Journal of Psychiatry, 175(8), 729–740. https://doi.org/10.1176/appi.ajp.2018.17101174.
Volkow, N. D., & Morales, M. (2015). The brain on drugs: From reward to addiction. Cell, 162(4), 712–725. https://doi.org/10.1016/j.cell.2015.07.046.
Whelan, R., Watts, R., Orr, C. A., Althoff, R. R., Artiges, E., Banaschewski, T., … Garavan, H. (2014). Neuropsychosocial profiles of current and future adolescent alcohol misusers. Nature, 512(7513), 185–189. https://doi.org/10.1038/nature13402.
Yücel, M., Lee, R. S. C., & Fontenelle, L. F. (2021). A new consensus framework for phenotyping and treatment selecting in addiction and obsessive-compulsive–related disorders. JAMA Psychiatry, 78(7), 699. https://doi.org/10.1001/jamapsychiatry.2021.0243.
Yücel, M., & Lubman, D. I. (2007). Neurocognitive and neuroimaging evidence of behavioural dysregulation in human drug addiction: Implications for diagnosis, treatment and prevention. Drug and Alcohol Review, 26(1), 33–39. https://doi.org/10.1080/09595230601036978.
Yücel, M., Oldenhof, E., Ahmed, S. H., Belin, D., Billieux, J., Bowden-Jones, H., … Verdejo-Garcia, A. (2019). A transdiagnostic dimensional approach towards a neuropsychological assessment for addiction: An international Delphi consensus study: Transdiagnostic neuropsychological approaches to addiction. Addiction, 114(6), 1095–1109. https://doi.org/10.1111/add.14424.
Zhou, Z., Li, C., & Zhu, H. (2013). An error-related negativity potential investigation of response monitoring function in individuals with internet addiction disorder. Frontiers in Behavioral Neuroscience, 7. https://doi.org/10.3389/fnbeh.2013.00131.