Personality changes related to presence and treatment of substance use (disorders): a systematic review

Heavy substance use (SU) and substance use disorders (SUD) have complex etiologies and often severe consequences. Certain personality traits have been associated with an increased risk for SU(D), but far less is known about personality changes related to SU(D). This review aims to synthesize the existing literature on this research question. A systematic literature search was conducted from November 2022 to February 2023 in PubMed, EbscoHost, and Web of Science. Peer-reviewed original papers on SU(D)-related personality changes were included. Of 55 included studies, 38 were observational population-based studies and 17 were intervention studies. Overall, personality and SU measures, samples, study designs, and statistical approaches were highly heterogenous. In observational studies, higher SU was most consistently related to increases in impulsivity-related traits and (less so) neuroticism, while interventions in the context of SU(D) were mostly associated with increases in conscientiousness and self-efficacy and lasting decreases in neuroticism. Findings for traits related to extraversion, openness, conscientiousness, and agreeableness were mixed and depended on SU measure and age. Studies on bidirectional associations suggest that personality and SU(D) both influence each other over time. Due to their strong association with SU(D), impulsivity-related traits may be important target points for interventions. Future work may investigate the mechanisms underlying personality changes related to SU(D), distinguishing substance-specific effects from general SU(D)-related processes like withdrawal, craving, and loss of control. Furthermore, more research is needed to examine whether SU(D)-related personality changes vary by developmental stage and clinical features (e.g. initial use, onset, remission, and relapse).

Personality traits are relatively stable and enduring patterns of thoughts, feelings, and behaviors that differ between individuals (Kandler, Zimmermann, & McAdams, 2014).A large proportion of these differences can be described by the Big Five traits extraversion, agreeableness, conscientiousness, neuroticism, and openness to experience (McCrae & Costa, 2008).Extraversion refers to one's sociability, assertiveness, and preference for social interactions.Neuroticism reflects negative affect, emotional instability, and insecurity.Openness includes intellect and willingness to explore new experiences.Conscientiousness refers to orderliness, responsibility, and self-discipline.Agreeableness refers to cooperativeness, empathy for others, and kindness.The Big Five are linked to life outcomes in various domains including health (Strickhouser, Zell, & Krizan, 2017).For example, individuals high in neuroticism are more susceptible to stress and detrimental coping strategies like substance use (Connor-Smith & Flachsbart, 2007), and individuals low in agreeableness may be less likely to comply with legal regulations, increasing the likelihood of illicit drug use (Dash, Martin, & Slutske, 2023).More broadly, personality also subsumes other traits, such as self-esteem or impulsivity.The terms impulsivity and impulsiveness are often used interchangeably in the literature yet their precise definition may vary by context.Here, we use the more common term impulsivity, which describes the tendency to act on impulse and without considering the consequences.
Although personality traits are relatively stable in the short term, they canand dochange over time.Previous studies have shown that personality develops across the entire lifespan and in relation to major life experiences (Bleidorn, Hopwood, & Lucas, 2018;Denissen, Luhmann, Chung, & Bleidorn, 2019).Theoretical framework (e.g.TESSERA) assume that personality may change due to repeated sequences of triggering situations, expectancies, state expressions, reactions, and associative/reflective processes (Wrzus & Roberts, 2017).These components may be affected by SU(D).For example, individuals with SUD often invest increasing time in obtaining/using drugs and neglect other areas of life (e.g.social relationships and work), which might lower conscientiousness.Moreover, substance-specific physiological effects may lead to changes in personality traits.Taken together, personality might affect whether individuals engage in SU (selection effects).At the same time, personality might change due to SU (socialization effects).
Consistently, different models on the relationship between SU and personality have been proposed (Klimstra, Luyckx, Hale, & Goossens, 2014;Samek et al., 2018).The vulnerability model suggests that certain personality traits (e.g.high neuroticism) predispose to SUD.Conversely, the scar model posits that SUD leads to personality changes (e.g.increasing neuroticism).The common cause model posits that SUD and specific personality trait levels (e.g.high neuroticism) stem from shared etiological factors but do not directly influence each other.The transactional model suggests that SUD and personality bidirectionally affect each other over time (Samek et al., 2018).The spectrum model suggests that personality traits and (problematic) behaviors lie on the same continuum (Klimstra et al., 2014;Krueger et al., 2021;Krueger & Tackett, 2003).For example, SU could be a manifestation of neuroticism, so that increasing neuroticism should be correlated with increasing SU.
The predictive role of personality traits in SUD has been widely studied (for a meta-analysis see Kotov et al., 2010).In line with the vulnerability model, numerous studies have shown that certain trait levels, especially high neuroticism, impulsivity, and sensation seeking, as well as low self-directedness and harm avoidance, relate to an increased risk for SUD (Dash et al., 2019;Kotov et al., 2010;Nevid et al., 2020;Sher, Bartholow, & Wood, 2000;Whiteside & Lynam, 2009).In contrast, much less is known about personality changes associated with SU(D) (Blonigen et al., 2015;Kroencke, Kuper, Bleidorn, & Denissen, 2021).Different substances might also be associated with different (changes in) personality traits due to substance-specific physiological mechanisms, social norms, and legal regulations (e.g.alcohol v. illicit drugs) (Robinson & Adinoff, 2016).
Research on SU(D)-related personality changes is important for several reasons.First, it may identify important mechanisms underlying personality development.For example, how does SU affect personality, and how might these effects explain agegraded personality changes (e.g. during adolescence)?Second, it may improve knowledge concerning the impact of personality changes on SU(D).To illustrate, does increasing neuroticism lead to heavier drinking?Examining such questions may provide predictive markers of symptom progression and recovery, which are needed for early detection and targeted intervention (Choi et al., 2023;Conrod, 2019;Debenham et al., 2021;Kroencke et al., 2021).Third, findings on personality changes due to SU(D) interventions may elucidate indicators of treatment success (Blonigen & Macia, 2021;Hershberger, Um, & Cyders, 2017).
Despite its importance, research on SU(D)-related personality changes is fragmented and lacks a comprehensive synthesis (Kroencke et al., 2021;Nevid et al., 2020).Thus, this systematic review aims to summarize existing evidence in this field by addressing the following research questions: (1) How does personality change before, during, and after the onset or remission of SU(D) (within-person comparisons)?(2) How are personality changes related to changes in diagnostic features (e.g.symptom severity) of SU(D) (within-person comparisons)?(3) Do personality changes differ between individuals with v. without SU(D) (between-person comparisons)?

Literature search
This review was preregistered in the PROSPERO Systematic Reviews Database (CRD42022370973) and follows the PRISMA and other best practice guidelines for systematic reviews (Page et al., 2021;Siddaway, Wood, & Hedges, 2019).
A systematic literature search was conducted on EBSCOhost, PubMed, and Web of Science (11/2022-01/2023).Additionally, literature cited in the identified papers was manually retrieved for further analysis.The review was restricted to peer-reviewed original papers published in academic journals in either English or German language, focusing exclusively on studies in humans.No restrictions were applied regarding the date of publication or other formal aspects of the search.Search terms were applied to the title, keywords, and abstract of potential studies.To ensure inclusion of the most recent studies, the search was repeated prior to the final submission.
Search terms (online Supplementary Table S1) included a combination of personality (e.g.Big Five domains, self-efficacy, impulsivity), change (e.g.change, development), and SU/addiction (e.g.addiction, substance use disorder, heavy drinking) keywords.The PRISMA flowchart (Fig. 1) depicts the procedure of the search, screening, and data extraction.

Screening and data extraction
To select the studies for inclusion, 2 reviewers applied the eligibility criteria.One reviewer screened the records and selected the studies based on the inclusion/exclusion criteria, while the other reviewer double-checked these decisions.Disagreements were resolved by consulting a third independent reviewer and finding a consensus decision.The same procedure was carried out for extracting and checking the retrieved data.The reviewers followed a standardized protocol with several steps, including examination of the titles and abstracts of references obtained from the databases and reading the full texts of potentially relevant articles.Of the studies meeting inclusion criteria, the aims, methods (e.g.country, sample, design, outcome variables), results, limitations, and potential biases were recorded and summarized.Given the heterogeneity of personality traits, SU measures, samples, study designs, and statistical approaches, a narrative synthesis rather than a meta-analysis was conducted.

Results
Of the 55 studies that met the inclusion criteria (publication dates: 1986-2022), 38 were observational studies on SU-related personality changes in population samples (Table 1) and 17 were (clinical) intervention studies in individuals with or at high risk for SU(D) (Table 2).Because the literature search yielded a plethora of personality constructs, we created eight clusters based on conceptual overlap among them: extraversionrelated traits, neuroticism-related traits, openness-related traits, conscientiousness-related traits, agreeableness-related traits, impulsivity-related traits, self-efficacy/self-esteem-related traits, and other.See online Supplementary Table S3 for more information.We did this to better structure the results and to identify similarities and differences in the findings.

Samples
The sample sizes of the population-based studies varied significantly from 99 to 56 786 participants.Thirteen studies focused on adolescents, 17 on young adults (mostly university students), one on middle-aged adults only, and 7 on the entire adult lifespan.

Study designs
In line with our search criteria, all studies were based on a longitudinal design, with number of waves ranging from 2 to 14 (median: 3 waves).The time span over which personality changes were assessed varied between 17 months and 30 years (median: 5 years).We present standardized effect sizes as exemplary but refrain from interpretations due to the large heterogeneity of studies.

Substances
In the observational studies, 31 studies investigated alcohol, 12 nicotine, 9 cannabis, and one ecstasy use.In addition, 4 studies  Higher ecstasy use at wave 1 was linked to higher sensation seeking at T2, adjusted for sensation seeking at wave 1.
There was no significant association between ecstasy use at wave 1 and impulsivity at wave 2, adjusted for ecstasy use at wave 1. i Rank-order correlations were moderate to high for all psychopathic traits and facets.
For both women and men, there were decreases over the study period in total psychopathic trait scores, as well as in the interpersonal, affective, and lifestyle facets.
There was no significant change in the antisocial facet.Over the study period, higher sensation seeking was linked to increased alcohol use at the following waves.In return, higher alcohol use was linked to increased negative urgency, sensation seeking, and lack of premeditation at the following waves.Thus, only sensation seeking and alcohol use showed bidirectional associations.Moreover, sensation seeking at wave 2 was a mediator between alcohol use at wave 1 and 3. Alcohol use/problems at wave 2 was a mediator between sensation seeking at wave 1 and 3, and between alcohol use at wave 1 and negative urgency at wave 3.Over the study period, there was positive correlated rank-order change between alcohol abuse and extraversion as well as two of its facets, and negative correlated change with conscientiousness and one of its facets.
There was positive correlated rank-order change between cannabis use and one facet of extraversion, and negative correlated change with conscientiousness.There were no significant associations between rank-order changes in SU and any other personality traits or facets.
Higher alcohol abuse was linked to higher extraversion and one of its facets and one facet of openness at following waves, and it was linked to lower agreeableness and conscientiousness and one of its facets.Cannabis use was linked to lower extraversion and one of its facets, and it was linked to higher openness and two of its facets at following waves.There were no significant associations for any other personality traits or facets predicting later SU.
Openness and one of its facets were linked to higher cannabis use at the following waves.Conscientiousness was linked to lower cannabis use at the following waves.There were no significant associations between any other personality trait or facet predicting later SU.Over the study period, three SU groups were identified: light, moderate, heavy.Affiliation, autonomy, exhibition, impulsivity, and play were linked to higher SU, whereas achievement, cognitive structure, and harm-avoidance were linked to lower SU.Across user groups, there were no significant differences in mean changes of personality traits over the study period.
There were no significant associations for self-esteem and nine other personality traits which were not reported.Over the study period, there was one significant association between SU and anxiety sensitivity, that is, higher nicotine use at wave 1 was linked to increased anxiety sensitivity at wave 2. Both alcohol use and binge drinking at wave 2 were linked to increased hopelessness at wave 3. Higher nicotine use at wave 1 was linked to increased hopelessness at wave 2. Conversely, hopelessness at wave 2 was linked to decreased binge drinking at wave 3.
Higher alcohol use at waves 1 and 3 was linked to increased sensation seeking at the following waves.Binge drinking at waves 2 and 3 was linked to increased sensation seeking at the following waves.Higher nicotine use at wave 2 was linked to increased sensation (Continued ) Over the study period, change in alcohol use was positively linked to change in impulsivity and sensation seeking.Higher initial levels of alcohol use were not significantly linked to changes in impulsivity and sensation seeking, but higher levels of sensation seeking were linked to increases in alcohol use.Over the study period, change in cannabis use was positively linked to change in impulsivity but not sensation seeking.Higher initial levels of cannabis use were not significantly linked to changes in impulsivity and sensation seeking, but higher levels of sensation seeking were linked to higher increases in cannabis use.Over the study period, change in nicotine use was positively linked to change in impulsivity but not sensation seeking.Higher initial levels of nicotine use were not significantly linked to changes in impulsivity and sensation seeking, but higher levels of impulsivity were linked to increases in nicotine use.Higher levels of heavy drinking at wave 1 were linked to increases in impulsivity and sensation seeking over the study period.Likewise, higher levels of impulsivity and sensation seeking at wave 1 were linked to increases in heavy drinking over the study period.
senior year of college) For autonomy, higher levels of autonomy were linked to lower increases in heavy drinking, but higher levels of heavy drinking showed no association with change in autonomy.Using biometric analysis, the authors also found that links between AUD and personality traits were largely explained by one source of shared genetic influence.
(Continued )  investigated multiple drug use including cocaine and sedatives, and 2 examined SU in general without differentiation between different substances.Indicators of both SU and personality are described in online Supplementary Table S2.)
For cannabis and other drug use, one study found that higher cannabis use was associated with increases in extraversion in university students (Klimstra et al., 2014), while 2 studies found no significant links in adult samples (Kroencke et al., 2021;Turiano et al., 2012).Taken together, the evidence on alcohol, nicotine, and other SU and changes in extraversion was inconsistent, with different directions of change, depending on age and personality measure.
For cannabis and other drugs, one study found that increasing sedative use was associated with subsequent increases in neuroticism (Kroencke et al., 2021), while 2 studies found no significant associations (Klimstra et al., 2014;Turiano et al., 2012).In sum, this indicates an association between higher alcohol use and increases in some neuroticism-related traits, while the evidence is less consistent for nicotine and other substances.
For nicotine use, one study found that openness decreased among persistent smokers (Stephan et al., 2019), while 4 studies found no significant association between nicotine use and change in openness (Allen et al., 2015;Jokela et al., 2018;Kroencke et al., 2021;Turiano et al., 2012).For cannabis and other drug use, one study found that higher cannabis was linked to increases in openness and its facet unconventionality in a college sample (Klimstra et al., 2014), while other research found no significant associations in adults (Kroencke et al., 2021).Overall, some evidence suggests that alcohol and cannabis use may be associated with increases in openness in adolescents and young adults.
Four studies found that nicotine use was associated with decreases in conscientiousness, mostly in young adults (Allen et al., 2015;Roberts & Bogg, 2004;Stephan et al., 2019;Welch & Poulton, 2009), while 4 studies found no significant associations in adults of all ages (Jokela et al., 2018;Kroencke et al., 2021;Stein et al., 1987;Turiano et al., 2012).Roberts and Bogg (2004) found that cannabis use was associated with decreases in social responsibility in young women, while Kroencke et al. (2021) found no significant associations in an older sample of both genders.There were no significant associations between other drug use and changes in conscientiousness-related traits (Kroencke et al., 2021;Stein et al., 1987;Turiano et al., 2012).In summary, there is some (albeit inconsistent) evidence that SU may relate to small decreases in conscientiousness.
For cannabis use, Quinn and Harden (2013) found increases in impulsivity but not sensation seeking from age 15 to 25, consistent with Wright andJackson (2022). De Win et al. (2006) found that higher ecstasy use was linked to increases in sensation seeking but not impulsivity.Wright and Jackson (2022) found that impulsivity and sensation seeking increased more strongly in adolescent cocaine users v. non-users.In summary, several studies suggest that SU relates to increases in impulsivity and sensation seeking, especially in adolescents and young adults.
Self-esteem and related traits (4 studies) Stein et al. (1987) found that alcohol use was associated with increases in one facet of self-esteem in women, while Kroencke et al. (2021) and Wright and Jackson (2022) found no such association.Wright and Jackson (2022) found that self-esteem increased less consistently during adolescence in smokers v. nonsmokers, whereas 2 studies found no significant association (Kroencke et al., 2021;Stein et al., 1987).Moreover, self-esteem increased less consistently in adolescent users (Wright & Jackson, 2022) and decreased in adult users (Kroencke et al., 2021) of sedatives/downers, whereas Stein et al. (1987) reported increases in one facet of self-esteem in young adults, although not differentiating between types of drugs.Taken together, previous research suggests no consistent association between SU and changes in self-esteem.

Directionality and within-person changes
Most observational studies investigating the directionality of change (online Supplementary Table S4) yielded bidirectional associations between SU and personality change.However, there was slightly more evidence for personality predicting subsequent changes in SU than for SU predicting subsequent personality changes.Studies that modeled between-person v. within-person effects evidenced within-person increases in impulsivity with higher/increasing SU (in college students) (Riley et al., 2018) but found little evidence for associations between SU and within-person changes in the Big Five (Jokela et al., 2018;Kroencke et al., 2021) (in adults of all ages).

Substances
Of the 17 intervention studies, 4 focused on alcohol use, one focused on nicotine use, one focused on heavy use of opioids, cocaine, and heroin, respectively, and nine were intervention studies that focused on treatment programs for heavy SU of various kinds and did not test for substance-specific effects.Indicators of SU(D) and personality that were considered in these studies are described in online Supplementary Table S2.
Extraversion-related traits (6 studies) Two studies found that extraversion increased from admission to 19-week follow-up in patients from an opioid treatment clinic using agonist medication and behavioral interventions (Carter et al., 2001) and from pre-treatment to 12-month follow-up in patients from a residential alcohol treatment program using social skills training and supportive therapy (Oei & Jackson, 1980).Three studies found that extraversion increased in SUD patients receiving cognitive behavioral therapy (CBT) and other multimodal interventions only from pre-to post-treatment, but not until 6-month (Stieger et al., 2022), 12-month (Blonigen & Macia, 2021), or 15-month follow-up (Piedmont, 2001).One study found no changes in extraversion in relation to SUD treatment (Borman et al., 2006).In summary, there is some evidence for increases in extraversion in SUD patients during treatment that, however, are not sustained in the long term.
Openness-related traits (4 studies) Two studies found increases in openness-related traits in SUD patients from pre-to post-treatment (i.e.CBT and comprehensive multimodal interventions) that, however, were not maintained until 6-month (Stieger et al., 2022) and 15-month (Piedmont, 2001) follow-up.Conversely, no changes in openness-related traits were found in relation to multimodal (Borman et al., 2006) and opioid (Carter et al., 2001) treatment.Thus, previous evidence suggests no lasting changes in openness-related traits following SUD treatment.
Conscientiousness-related traits (5 studies) Four studies found lasting increases (up to 15-month follow-up) in conscientiousness (Carter et al., 2001;Piedmont, 2001;Stieger et al., 2022) and constraint (Blonigen & Macia, 2021) in relation to (cognitive) behavioral approaches (e.g.d = 0.37 from pre-to post-treatment in Stieger et al., 2022).In contrast, Borman et al. (2006) found no significant changes in conscientiousness and persistence but increases in selfdirectedness from admission to 6-month follow-up in an intervention focusing on physical, emotional, social, and spiritual factors.Taken together, there is initial evidence that cognitive-behavioral SUD treatment relates to increases in conscientiousness-related traits.
Agreeableness-related traits (4 studies) One study found increases in agreeableness in SUD patients up to 6-month follow-up (Piedmont, 2001), another found increases from pre-to post-treatment, but this effect disappeared at 6-month follow-up (Stieger et al., 2022).Two studies found no significant changes in agreeableness-related traits related to SUD treatment (Borman et al., 2006;Carter et al., 2001).The evidence provides little support for lasting increases in agreeableness following SUD treatment.
Impulsivity-related traits (4 studies) Kazemi et al. (2014) found a decrease in impulsivity among university students who participated in an alcohol intervention to motivate change in drinking behavior (either mandated or voluntarily) from baseline to 6-month follow-up, including booster sessions.In inpatient multimodal SU treatments, Littlefield et al. (2015) found decreases in some facets of impulsivity, while Aklin et al. (2009) found no significant changes in impulsivity but decreases in risk taking, both after 4 weeks of intervention.Gonçalves et al. (2014) found no changes in impulsivity following a 4-week motivational interviewing and chess intervention for cocaine users in a psychiatric clinic.For sensation seeking, no significant changes in relation to SUD treatment were found (Kazemi et al., 2014;Littlefield et al., 2015).These results provide some, but inconsistent evidence for decreases in impulsivity-related traits in relation to SU(D) interventions.
Self-efficacy and related traits (7 studies) Only studies that focus on general (but not domain-specific) selfefficacy are considered, as the focus of this review is on major personality traits.RCTs in SUD patients found that clinically guided self-help plus standard treatment (i.e.methadone maintenance and counselling) (Nurco et al., 1995), mindfulness-based psychoeducation (Bayır & Aylaz, 2021), CBT with music intervention (Kayaoğlu & Şahin Altun, 2022), and advocacy training to reduce smoking in university students (Winkleby et al., 2004) led to increases in self-efficacy (or changes from external to internal locus of control, Nurco et al., 1995).Two studies without a randomized controlled design also found that self-efficacy increased in university students following an advocacy intervention to reduce smoking (Winkleby et al., 2001) and in alcohol use patients from pre-to post-treatment (i.e.CBT plus exercise incentives) but not to 4-or 6-month follow-up (Can Gür & Okanli, 2019).One study found no changes in self-efficacy after 6-week CBT (Chodkiewicz & Gruszczyńska, 2019).Taken together, these findings suggest that most SU(D) interventions increase self-efficacy.

Discussion
This systematic review synthesized existing evidence on personality changes associated with SU(D).In observational studies, higher or increasing SU was most consistently linked to increases in impulsivity, sensation seeking (to a lesser extent), and less consistently neuroticism and related traits.SU(D) interventions were linked to decreases in neuroticism, which is consistent with meta-analytic findings that psychological interventions were related to reductions in neuroticism (Roberts et al., 2017).For impulsivity and sensation seeking, there was only weak support for intervention-related changes.However, intervention studies provided support for initial increases in self-efficacy and sustained increases in conscientiousness with treatment/amelioration of SU(D), whereas observational studies provided no consistent evidence for changes in self-esteem and suggested that only severe SU may be linked to decreases in conscientiousness.For traits related to extraversion, agreeableness, and openness, findings were inconsistent and rather weak in both types of studies, as discussed below.
Differences in findings might be due to differences in personality and SU measures, samples, and study designs, often referring to different age groups (see online Supplementary Table S5).Some studies suggest that developmental factors play an important role, such that certain SU-related personality changes especially occur in younger individuals.For example, positive associations between alcohol/cannabis use and openness-related traits were restricted to younger individuals, and nicotine use was associated with increased impulsivity in adolescents but not in young adults, consistent with findings that differences in impulsivity between non-clinical adolescent users and non-users did not persist into adulthood (Wright & Jackson, 2022).Furthermore, the personality trait, the severity of SU(D), and the context are important.For example, alcohol use has been associated with higher levels of extraversion particularly in the college context (Alexander, Howard, & Maggs, 2022), but it is possible that increases in extraversion are only found when considering the frequency of (binge) drinking (e.g. at student parties) (Klimstra et al., 2014) rather than negative consequences and SUD symptoms (Littlefield et al., 2009(Littlefield et al., , 2010b)).Conversely, the evidence for a negative link between SU and changes in conscientiousness-related traits is most consistent when studies examine symptoms of SUD (e.g.Hicks et al. 2012;Littlefield et al. 2010b;Samek et al. 2018) rather than just frequency of SU.We also identified a gap in research: observational studies in adolescents have almost exclusively investigated changes in impulsivity and sensation seeking, whereas studies in (middle-aged) adults have almost exclusively examined change in the Big Five.Thus, future research in the context of SU may additionally focus on changes in the Big Five during adolescence.
Regarding bidirectional associations between SU and personality, observational studies provided slightly more evidence for personality predicting changes in SU (vulnerability model) than for SU predicting changes in personality (scar model), but these effects were rarely directly compared.Consistent with the transactional model, almost all studies reported (at least some) associations in both directions, highlighting that SU and personality are closely intertwined: Certain personality trait levels (e.g.high impulsivity) predispose to increased SU (selection effects).At the same time, increasing SU accentuates these trait levels over time, leading to a vicious cycle of SU and associated (personality) problems (socialization effects).
Intervention studies found that neuroticism decreased during SU(D) treatment, while other personality changes varied by type of intervention.Extraversion increased with social skills training, while conscientiousness, agreeableness, and self-efficacy increased with cognitive-behavioral interventions that focused on stress coping and regaining control over SU.Consistent with this idea, Littlefield et al. (2010b) found that more functional coping mediated the association between decreases in SU and increases in conscientiousness.Personality changes were often not maintained until follow-up, highlighting the need for additional booster sessions after treatment.However, only 4 studies were based on randomized controlled designs, so there is limited causal evidence that SU(D) interventions induce personality changes.
Most observational studies focused on the frequency and intensity of SU without assessing diagnostic criteria for fullthreshold SUD (but see Hicks et al., 2012;Östlund et al., 2007;Samek et al., 2018).Thus, they refer only to personality changes related to more (or less) frequent SU regardless of clinical features.The intervention studies mostly focused on individuals with fullthreshold SUD but did not directly assess whether changes in SU(D) were related to personality changes.That is, there is little (direct) evidence on personality changes before, during, and after onset or remission of SU(D) or in relation to specific clinical features, as specified by our research questions.In general, the studies included in this review were highly heterogeneous in terms of measures, samples, study designs, and statistical approaches (see also online Supplementary Table S5).
Because individual substances (e.g.sedatives v. stimulants) have different physiological effects, substance-specific personality changes are plausible.At the same time, SU-related personality changes may be due to general processes associated with heavy use (e.g.craving and withdrawal) independent of a particular drug (Chen, 2022).Observational studies found that personality changes varied by substance, but rarely focused on full-threshold SUD.Intervention studies focused on SUD but rarely distinguished between substances, highlighting the need for future research in this field.
To the best of our knowledge, this review provides the first comprehensive aggregation and systematic synthesis of evidence on SU(D)-related personality changes.However, it is not without limitations: First, our focus was on clinically relevant SU(D).However, because most studies focused on SU frequency and intensity irrespective of SUD, it was virtually impossible to analyze associations with specific clinical features and diagnostic transitions (e.g.onset and remission of SUD).At the same time, this limitation could be considered a strength given the increasing importance of dimensional approaches in clinical psychology (Kotov et al., 2021;Krueger et al., 2021).Second, we only included peerreviewed studies published in English or German, potentially limiting generalizability.

Conclusions
In observational studies, SU(D) was most consistently related to increases in impulsivity-related traits.In intervention studies, decreases in SU(D) were linked to decreases in neuroticismrelated traits and increases in conscientiousness and selfefficacy, although the available literature was sparse.Associations between SU(D) and other personality traits varied by substance and developmental stage.Overall, studies were highly heterogeneous in measures, samples, study designs, and statistical approaches.Future meta-analyses may investigate whether and how changes in specific personality traits vary by substance and different indicators of SU(D) (e.g.frequency of use v. clinical features).Practitioners may particularly target traits related to impulsivity, neuroticism, conscientiousness, and self-efficacy to treat but also prevent SU(D).Tailored interventions based on personality information, such as the PreVenture program (Debenham et al., 2021;Newton et al., 2022), have been shown to be effective, and could be implemented on a larger scale.

Preregistration statement
This systematic review was preregistered in the PROSPERO Systematic Reviews Database (CRD42022370973).
Supplementary material.The supplementary material for this article can be found at https://doi.org/10.1017/S003329172400093X.

Table 1 .
Observational population-based studies frequent, moderate, increasing, occasional, low increasing, decreasing, and rare.From high school through the end of college, the occasional, decreasing, and rare group decreased in sensation seeking, whereas the increasing group increased in sensation seeking.The moderate group decreased in impulsivity, whereas the frequent group increased in impulsivity.From the end of college through two years later, the frequent group decreased in impulsivity, whereas the increasing group increased in impulsivity.There were no significant associations for any other group and changes in sensation seeking or impulsivity.
Higher positive urgency and lack of perseverance at wave 1 were linked to increased alcohol use at wave 2. In return, higher alcohol use at baseline was linked to increased negative urgency, positive urgency, lack of premeditation, and sensation seeking at wave 2.
There were no significant findings for changes in openness, agreeableness and neuroticism in relation to changes in SU. d

Table 1 .
(Continued.) There were no significant associations with extraversion, conscientiousness, and openness and no significant associations between other substances and changes in any personality traits.(Continued)Psychological Medicine

Table 1 .
(Continued.) at age 43 was linked to decreases in social responsibility from age 43 to 52.There were no such significant associations between alcohol and tobacco use and changes in social responsibility.Higher levels of nicotine and cannabis use, but not alcohol use, at age 43 were linked to decreases in social responsibility from ages 21 to 43.In return, higher social responsibility at 21 was linked to increased alcohol, tobacco, and cannabis use at age 43, but not from age 43 to 52.
d Higher cannabis use

Table 1 .
(Continued.) SU(D), substance use (disorder); AUD, alcohol use disorder.a Studies used data from the Household, Income, and Labour Dynamics Study (Australia).b Analyses were run separately for substances.c Studies used data from the Minnesota Twin Family Study (United States).d Bidirectionality between SU and personality trait (change) was explicitly modeled.e Studies used data from the Health and Retirement Study (HRS) (United States).f Studies used data from the Midlife in the US Study (MIDUS) (United States).g Studies used data from the Socio-economic Panel (GSOEP) (Germany).h Studies used data from the Wisconsin Longitudinal Study (WLS) (United States).i Study's primary aim was to investigate stability (not change) of personality traits.j Studies used data from the Midlife in Japan Study (MIDJA) (Japan).k The same sample and measure for impulsivity was used, time frames were partly overlapping.l Analyses were run combinedly for substances.m The same sample across a similar time frame was used but different measures for impulsivity and extraversion were employed.n Studies used data from the National Longitudinal Study of Youth (NLSY).

Table 2 .
Intervention studies

Table 2 .
(Continued.) AUD, alcohol use disorder; DSM, Diagnostic and Statistical Manual of Mental Disorders; SU(D), substance use (disorder).a Study's primary aim was to investigate stability (not change) of personality traits.
found that SUD patients exhibited increases in trait resilience up to 6 months after CBT.