An estimated 5–8 million Americans use long-term opioid therapy (LTOT) for the treatment of chronic pain (Reuben et al., 2015). Though the majority of individuals on LTOT do not engage in nonmedical opioid use, patients can become physically dependent on opioids and have higher rates of nonmedical opioid use than the general population (Karmali et al., 2020). In fact, increased prescribing of opioid medications in the late 1990s and early 2000s for both chronic and acute pain has been cited as one of the contributing factors to the increased rates of opioid use and opioid mortality in subsequent years (Kolodny et al., 2015). A deeper understanding of individual differences and the role of momentary experiences in opioid response among patients with chronic pain on opioid therapy may help mitigate harm.

Craving is defined as a powerful urge to use a drug, and is both a symptom of opioid use disorder and a contributor to continued opioid use (Kakko et al., 2019). Previous work suggests that as many as 45 % of patients with chronic pain on LTOT experience some degree of craving for opioids, and that individuals who experience craving face elevated risk for nonmedical opioid use (Wasan et al., 2009). Among patients with chronic pain, craving frequently occurs in the context of greater emotional distress and pain catastrophizing (i.e., amplified emotional and cognitive responses to pain) (Martel et al., 2014a, Martel et al., 2014b, Parisi et al., 2022), but has only been weakly linked to elevations in pain intensity (Martel et al., 2016). Craving has also been shown to mediate the relationship between negative affect and nonmedical opioid use in patients on LTOT (Martel et al., 2014a, Martel et al., 2014b), suggesting that craving may represent an important mechanism linking negative affective states to non-medical opioid use.

Findings that craving may be a risk factor for nonmedical opioid use among patients on LTOT should be considered against the backdrop of widespread state-based legalization of cannabis for non-medical and medical use in the United States. The number of medical cannabis patients increased 4.5-fold between 2016 and 2020, with chronic pain being the most commonly listed qualifying condition (Boehnke et al., 2022). A subset of individuals who use cannabis for medical use also report craving. One study that assessed adapted DSM-5 criteria for cannabis use disorder and found that approximately 14.4 % of persons using cannabis for chronic pain reported a strong desire to use cannabis that made it difficult to think of anything else (Bialas et al., 2023). In a separate study of patients seeking a medical cannabis waiver for chronic pain management, a subset reported elevations in craving for cannabis during periods of non-use (Coughlin et al., 2021). These findings demonstrate that at least a portion of patients with chronic pain who use cannabis for pain management experience craving associated with cannabis use.

Patients increasingly report using cannabis as an adjunct to LTOT (Boehnke et al., 2016), with an estimated 16 % of patients reporting use of both opioids and cannabis for pain management (Degenhardt et al., 2015). Patients with chronic pain who use medical cannabis have reported that it changes the perception and experience of pain and provides a desirable alternative to other pain medications (Haroutounian et al., 2016, Piper et al., 2017). Nonetheless, increased pain intensity and interference (Degenhardt et al., 2015), anxiety and depression (Buckner et al., 2023), and nonmedical opioid use (Nugent et al., 2018) have also been reported among persons using cannabis and opioids relative to those who only use opioids. Limited attention has been given to correlates of craving among persons using both cannabis and opioids for pain management, and previous studies among persons who co-use cannabis and opioids have primarily examined outcomes such as daily intake (Mun et al., 2022) or substitution patterns (Boehnke et al., 2021, Boehnke et al., 2019, Lucas et al., 2021). As medical cannabis use continues to climb among patients with chronic pain, it is imperative to examine whether craving is more likely to occur during moments of heightened pain intensity, pain catastrophizing, and negative emotional states.

In the present study, craving and related symptoms were assessed multiple times per day using ecological momentary assessment (EMA) - an intensive longitudinal method for capturing time-varying states in real time (Shiffman, 2009, Shiffman et al., 2008). EMA provides a unique opportunity to capture the dynamic nature of craving in order to examine variations in craving and related symptomatology. The dense sampling strategy afforded by EMA yields a daily measure of average craving intensity and reduces retrospective recall bias of craving, which is important since craving is, by nature, a transient experience known to vary in intensity and duration. The present study tested the following hypotheses: 1) opioid and cannabis craving would be elevated on days of opioid and cannabis co-use, 2) momentary craving would be elevated in the context of greater levels of pain intensity, pain catastrophizing, stress, and low affect, and 3) elevations in pain intensity, pain catastrophizing, stress, and low affect would precede elevations in craving.