Substance‐related intrusive memories in cocaine use disorder are different from but associated with craving

BACKGROUND AND AIMS: Cocaine use disorder (CUD) is a significant global health concern, characterized by persistent craving despite severe consequences. Recent theories highlight maladaptive memory processes - such as intrusive, vivid recollections of past substance use that arise spontaneously in daily life - as key contributors to craving and relapse; however, empirical studies examining such substance-related intrusions in naturalistic contexts remain scarce. This study therefore used ecological momentary assessment (EMA) to investigate the frequency, phenomenology and emotional and behavioural correlates of substance-related memory intrusions in individuals with CUD, and how these relate to craving, cocaine use severity (CUS) and psychotherapy experience. DESIGN: A 14-day EMA study captured event-based reports of intrusions, craving and related experiences in participants diagnosed with CUD. SETTING: The study was conducted in Switzerland in a naturalistic setting, with data collected via individuals' smartphones. PARTICIPANTS: N = 43 participants (recruited in Switzerland, 16% female; 18-59 years old, median compliance rate 82.8%) with a current DSM-5 diagnosis of CUD made a total of n = 360 event-based reports that were analysed. MEASUREMENTS: EMA reports included intrusion episode types (pure intrusions, intrusions with subsequent or simultaneous craving or pure craving), intrusion characteristics (vividness, intrusiveness, modalities), craving intensity, episode triggers and cognitive-behavioural, emotional and physiological responses. CUS was assessed based on use quantity, duration and obsessive-compulsive use symptoms. FINDINGS: Participants recorded an average of 8.4 episodes (standard deviation = 5.8) across 14 days. Intrusions frequently occurred independently of craving (42.4%) but were statistically significantly associated with greater craving intensity when more vivid (unstandardized regression coefficient b = 0.53, P = 0.002) and intrusive (b = 0.48, P < 0.001). Episodes involving craving were characterized by greater distress (b = 1.52-2.17, all P < 0.001) and greater loss of control (b = 2.41-3.22, all P < 0.001) and were associated with higher odds of reporting obtaining cocaine (odds ratio = 19.90, P < 0.001). Higher CUS predicted more frequent intrusion episodes (unstandardized regression coefficient β = 0.52, P < 0.001), while psychotherapy experience was associated with lower vividness (b = -1.45, P = 0.008), intrusiveness (b = -1.33, P = 0.004) and craving intensity (b = -1.56, P = 0.010). CONCLUSIONS: Substance-related memory intrusions in people with cocaine use disorder are distinct cognitive-affective events that often occur independently of craving but are closely linked to its intensity, particularly when experienced as vivid and emotionally charged. Targeting these features through behavioural or pharmacological interventions may help mitigate craving-driven distress and impulsive use-related behaviour.