Metacognition in addictive behaviors

In this blog we simply cite and use excerpts from an interesting meta-analytic study (1)  into the various types of thinking problems that people with addictive bahaviours seem to exhibit.

“The term ‘metacognition’, which is most often associated with the work of John Flavell (1979, 1987), can be broadly defined as knowledge and cognitive processes that are involved in the appraisal, control, or monitoring of thinking.

More recently, as a result of the work of Adrian Wells and his colleagues, metacognition has applied to conceptualizing and treating psychological distress. Wells and Matthews (1994, 1996) have proposed a multi-process model, the Self-Regulatory Executive Function (S-REF) model, to represent dysfunctional cognition in psychological distress. The novel features of this model are: (1) the identification of a common or transdiagnostic set of processes and structures; (2) the modeling of cognition within an explicit cognitive architecture; (3) emphasis on top-down or strategic influences on processing bias; and (4) an explicit role assigned to metacognitive beliefs in the underpinning of coping styles that lead to psychological distress.

The initiation and cessation of S-REF activity are influenced by first level automatic processing (e.g. an intrusion related to body symptoms) and by the third level in the model: metacognitive knowledge. Metacognitive knowledge is conceptualized as information and beliefs about cognition that are positive and negative in content (e.g. “Worrying will help me cope” and “Some thoughts are dangerous”) and generic plans for guiding cognition. Wells and Matthews (1994) argue that a particular thinking style is central to psychological disorder; the Cognitive Attentional Syndrome (CAS). The CAS consists of a variety of coping styles including extended thinking (e.g. desire thinking, rumination and worry), monitoring for threat, thought suppression and avoidance, that have paradoxical effects on self-regulation and discrepancy reduction. According to the SREF model, the CAS is problematic because it causes negative thoughts and emotions to persist, leading to failures to modify dysfunctional metacognitive beliefs and stably resolve self-discrepancies.

Applying the S-REF model to addictive behaviors

Spada and Wells (2009) and Spada, Caselli, and Wells (2013) have applied the S-REF model to addictive behaviors.

In their formulation the CAS and metacognitive beliefs are conceptualized across three temporal phases of the addictive behavior episode: pre-engagement, engagement, and post engagement. What follows is an exposition of these different phases in nicotine use.

In the pre-engagement phase triggers in the form of urges, images, memories or thoughts activate the S-REF and associated metacognitive beliefs to guide appraisal and coping style. Positive metacognitive beliefs such as “Thinking about having a cigarette will make me feel better” and negative metacognitive beliefs such as “I cannot control my thoughts of smoking” activate the perseverative processing of intrusions and attempts at their suppression (the CAS) leading to an escalation of negative affect and craving. As a consequence the smoker becomes more likely to use in order to regulate these feelings and escape the escalating discrepancy between current and desired states.

The activation of the pre-engagement phase requires awareness of a preference. It is important to note that if the addictive behavior becomes habitual then the pre-engagement phase may be bypassed. When habit has been established, the pre-engagement phase can occur in three circumstances: (1) in a relatively new contextual environment; (2) when the habitual addictive behavior is interrupted through external control; or (3) through conscious attempts to remain abstinent.

This helps to explain why severely addicted individuals may not report pre-engagement extended thinking (for example Caselli & Spada, 2011).

In the engagement phase positive metacognitive beliefs about engagement (“Smoking will help me control my thoughts/reduce my worrying”) are paralleled by changes in metacognitive monitoring (the ability to monitor internal states as a guide to knowing how close one is to resolving discrepancies and achieving the desired state). These two factors contribute to a reduced ability to regulate behavior.

Specifically, reductions in metacognitive monitoring are thought to result from: (1) behaviors that distract from self-awareness and from monitoring the flow of goal-progress information thereby limiting the opportunity to identify a stop signal for engagement; and/or (2) the chemical effects of the addictive behavior (e.g. alcohol/nicotine) which affects higher order functioning and therefore impacts negatively on metacognitive monitoring.

Over the course of time and as the addictive behavior escalates in severity, negative metacognitive beliefs about its uncontrollability emerge, contributing to its perseveration.

These include negative metacognitive beliefs about the power of thoughts about the addictive behavior causing uncontrollable engagement (“Thinking about smoking can make me do it”) and negative metacognitive beliefs about the uncontrollability of the addictive behavior once it is initiated (“Once I start smoking I find it difficult to stop”). In the post-engagement phase, an intrusion (e.g. a self-blaming thought or withdrawal symptoms) leads to accessing positive metacognitive beliefs about post-engagement rumination (e.g. “If I analyze why I am feeling this way I will understand why I smoke”) and the activation of the associated coping styles of rumination and thought suppression. The latter leads to a worsening of negative affect increasing the likelihood of returning to engagement as a means of achieving self-regulation.

A review of research evidence on the components of the triphasic metacognitive formulation of addictive behaviors

The triphasic metacognitive formulation of addictive behaviors proposes that aspects of the CAS such as attentional bias, extended thinking (e.g. desire thinking, rumination and worry), disruption in metacognitive monitoring and thought suppression should be associated with addictive behaviors and lead to maladaptive consequences including increased levels of craving and engagement.

Attentional bias

In the S-REF model attentional bias is a function of both automatic and strategic processes, however what is emphasized is the role of strategic processing. Consistent with this view the emotional Stroop effect appears to be more dependent on slow disengagement processes than on fast automatic bias (Phaf & Kan, 2007) and on the voluntary maintenance of attention upon target-related cues (Field, Mogg, Zetteler, & Bradley, 2004). To this extent attentional bias is likely to be sensitive to the individual’s coping styles which in turn are influenced by metacognitive beliefs and the motivation to continue or disengage from processing personally relevant stimuli.

Attentional bias should therefore play a role in developing and maintaining: (1) addictive behavior; (2) risk of relapse; and (3) a strong and perseverative experience of craving (for a review see Field & Cox, 2008).

For example, among users of different substances, substance-related attentional bias is directly proportional to the quantity and frequency of the substance used (Field & Cox, 2008). The association between attentional bias and substance use has been well-replicated for alcohol misuse (e.g. Field, Schoenmakers, & Wiers, 2008; Sharma, Albery, & Cook, 2001), cannabis use (Field, Eastwood, Bradley, & Mogg, 2006) and nicotine use (Mogg, Field, & Bradley, 2005).

Longitudinal designs have also demonstrated the association between attentional bias and risk of subsequent relapse in alcohol abusers (Cox, Hogan, Kristian, & Race, 2002), tobacco smokers (Waters, Shiffman, Bradley, & Mogg, 2003), heroin users (Marissen et al., 2006) and cocaine users (Carpenter, Schreiber, Church, & McDowell, 2006).

Extended thinking

Extended thinking refers to recurrent, dysfunctional and rigid thinking styles that perpetuate the accessibility of intrusions. Desire thinking, rumination and worry are the main types of extended thinking that have been identified in the literature. Desire thinking has been characterized as a voluntary process involving the elaboration of a desired target at a verbal level and an imaginal level (Caselli & Spada, 2010; Kavanagh, May, & Andrade, 2009). The target of desire thinking may be an activity, an object, or a state (Kavanagh, Andrade, & May, 2004, 2005).

Rumination and worry are characterized by heightened selffocused attention involving persistent, recyclic, and predominantly verbal internal questioning about the causes, meaning, and consequences of one’s internal experiences. Rumination is focused on depressive symptoms and their consequences (Nolen-Hoeksema & Morrow, 1991), while worry is characterized by an apprehensive expectation of possible negative outcomes in the future (Borkovec, 1994).

Research has shown that desire thinking occurs in nicotine dependence, problematic gambling and problem drinking (Caselli & Spada, 2010).

In addition desire thinking has been found to: (1) predict craving across a range of addictive behaviors (Caselli, Soliani, & Spada, 2013; Caselli & Spada, 2011); (2) rise across the continuum of drinking and smoking behaviors (Caselli, Ferla, Mezzaluna, Rovetto, & Spada, 2012; Caselli, Nikčević, Fiore, Mezzaluna, & Spada, 2012); and (3) be associated with levels of problematic gambling (Fernie et al., 2014).

On similar lines desire thinking has been found to discriminate between problematic and non-problematic Internet users (Spada, Caselli, Slaifer, Nikčević, & Sassaroli, 2013).

Research has also demonstrated that rumination is higher for problem drinkers compared to social drinkers (Caselli, Bortolai, Leoni, Rovetto, & Spada, 2008), that it prospectively predicts alcohol use in community and clinical samples (Caselli et al., 2010) and that it brings increases in craving in experimental conditions (Caselli et al., 2013).

Several studies have also supported the association between high levels of worry and the tendency to use alcohol in problem drinkers (Goldsmith, Tran, Smith, & Howe, 2009; Smith & Book, 2010).

Thought suppression

Thought suppression is a mental control strategy involving the attempt to keep certain thoughts out of awareness. The engagement in thought suppression can lead to an increase in the suppressed thought (Wenzlaff & Wegner, 2000). This paradoxical effect has been observed in individuals with addictive behaviors. For example, when alcohol dependent individuals try to suppress thoughts regarding alcohol, these thoughts may become hyper-accessible immediately afterward (Klein, 2007).

In addition, alcohol dependent individuals that have been exposed to a suppression condition have been shown to be faster to endorse alcohol outcome expectancies following exposure to alcohol cues than individuals in a control group (Palfai, Monti, Colby, & Rosenow, 1997). It also appears that both trait and state thought suppression contribute towards the depletion of neurocognitive resources needed to regulate urges (Garland, Carter, Ropes, & Howard, 2012).

Finally, it has also been found that a greater use of smokingrelated thought suppression in everyday life is significantly associated with a greater desire to smoke (Erskine et al., 2012), attempts to quit smoking, and number of cigarettes smoked (Erskine, Georgiou, & Kvavilashvili, 2010). A similar association has also been observed between thought suppression and problem gambling (Riley, 2014).

Metacognitive monitoring (the ability to monitor internal states as a guide to knowing how close one is to resolving discrepancies and achieving the desired state; Spada & Wells, 2006; Spada, Nikčević, Moneta, & Wells, 2007) is likely to be affected by addictive substances. For example, Steele and Josephs (1990) have demonstrated that alcohol’s pharmacological properties disrupt attentional processes (through the narrowing of perception to immediate cues and reduction of cognitive abstracting capacity), and Hull (1981) has shown that alcohol use reduces self-awareness (corresponding to the encoding of information in terms of self-relevance).

Evidence also suggests that alcohol intoxication impairs neurological systems that underlie meta-level processing (Nelson et al., 1998). Furthermore research evidence has shown that not attending internally to the change in cognition and emotion (poor metacognitive monitoring) that occurs during alcohol and nicotine use is associated with excessive use (Nikčević & Spada, 2010; Spada & Wells, 2006). Finally, poor metacognitive monitoring has been associated with perseveration of gambling activity in individuals with gambling disorder (Spada, Giustina, Rolandi, Fernie, & Caselli, 2014).

Metacognitive beliefs

Positive and negative metacognitive beliefs about engagement in addictive behaviors have been identified in nicotine dependence, gambling and problem drinking. Positive metacognitive beliefs relate to the effects of engaging in addictive behavior as a means of controlling and regulating cognition (e.g. “Smoking helps me to control my thoughts”) and affect (e.g. “Gambling will improve my mood”) (Nikčević & Spada, 2010; Spada et al., 2014; Spada & Wells, 2006, 2008; Toneatto, 1999).

Negative metacognitive beliefs concern the perception of lack of executive control over the engagement in the addictive behavior (e.g., “My smoking persists no matter how I try to control it”), uncontrollability of thoughts related to the addictive behavior (“The thought of gambling is stronger than my will”), thought–action fusion (“Thinking about using alcohol can make me drink”), and the negative impact of the engagement in the addictive behavior on cognitive functioning (“Drinking will damage my mind”) (Hoyer, Hacker, & Lindenmeyer, 2007; Nikčević & Spada, 2010; Spada et al., 2014; Spada & Wells, 2006, 2008; Toneatto, 1999).

Positive and negative metacognitive beliefs about alcohol use have been found to predict the severity of alcohol use in binge drinking university students (Clark et al., 2012), problem drinking in clinical and non-clinical samples (Spada & Wells, 2009, 2010) and drinking behavior independently of alcohol outcome expectancies in non-clinical samples (Spada, Moneta, & Wells, 2007).

Positive and negative metacognitive beliefs have also been found to play a role in desire thinking. Positive metacognitive beliefs about desire thinking (“Imagining something I desire gives me control over my choices”) relate to the use of desire thinking as a form of coping with cognitive–affective triggers present in addictive behavior (Caselli & Spada, 2010, 2013).

Negative metacognitive beliefs about desire thinking (“I cannot stop thinking about a desire activity”) concern the uncontrollability of desire thinking and its negative impact on executive control over behavior, self-image, and cognitive performance (Caselli & Spada, 2010, 2013).

Research has also shown that general negative metacognitive beliefs are elevated across addictive behaviors including alcohol (Spada, Caselli, & Wells, 2009; Spada & Wells, 2005; Spada, Zandvoort, & Wells, 2007), gambling (Lindberg, Fernie, & Spada, 2011), nicotine (Nikčević & Spada, 2008; Spada et al., 2007) and Internet use (Spada, Langston, Nikčević, & Moneta, 2008).

These studies have tended to show that negative metacognitive beliefs about the need to control thoughts and lack of cognitive confidence positively predict addictive behavior.”

Reference

1. Spada, M. M., Caselli, G., Nikčević, A. V., & Wells, A. (2015). Metacognition in addictive behaviors. Addictive behaviors, 44, 9-15.