Nicotine and Cannabis Co-Use and Sleep-Related Outcomes in Adults: A Narrative Review and Implications

Study and Population Characteristics

Overall, this review included six eligible studies examining nicotine and cannabis co-use behaviors and sleep-related outcomes among adults, with sample sizes ranging from 200 to 6,573 participants (Table 1). Populations included general adults from national surveys [24, 25], clinical populations such as first-episode psychosis patients (i.e., individuals experiencing their first psychotic episode, assessed for earliest symptoms and clinical profile) [26] and chronic rhinosinusitis patients [27], pregnant women [28], and local adolescent and young adult samples [29]. Study designs included three cross-sectional surveys [24, 25, 29], two prospective cohort studies [27, 28], and a retrospective cohort study [26]. Four studies were conducted in the U.S. [24, 25, 28, 29], one in Egypt [27], and one in Spain [26].

Nicotine and Cannabis Co-Use

Nicotine use was captured via cigarettes, e-cigarettes, or combined tobacco products (cigarette and e-cigarette co-use), while cannabis use included smoked cannabis, vaped cannabis, and general self-reported cannabis use without specifying the route of administration. Definitions of co-use varied widely. Three studies defined co-use as using both substances within the past 30 days [24, 25, 29]. Another study compared adults with current heavy daily cigarette use (≥20 cigarettes/day for ≥5 years) with those who also reported habitual cannabis use (≥1 joint/day for ≥1 year), classifying the latter group as co-users [27]. One study assessed tobacco and cannabis use history (lifetime, including past-30-day use) and defined co-use based on participants’ history of using both substances prior to the onset of first-episode psychosis [26]. One study examined tobacco (unspecified products) and cannabis co-use during pregnancy, defined as any reported or biologically verified use of both substances at any point during pregnancy [28]. Comparison groups for co-users included non-users (neither nicotine nor cannabis), nicotine-only users, and cannabis-only users, with slight variations across studies.

Across the different populations evaluated, sample rates of nicotine and cannabis co-use vary from 4% to 66%. In a sample of U.S. adults aged 18–64 (N=6,573), NHANES data showed that approximately 4% reported co-use of e-cigarettes and cannabis, and 9% reported co-use of traditional cigarettes and cannabis [24]. In another NHANES adult sample examining rest-activity rhythms (N=5,129), 6.7% of adults aged 18–59 reported current co-use of cigarettes and cannabis [25]. Among a mixed sample of adolescents and young adults (aged 16–22; N=221), co-use was notably higher, with about 43% classified as co-using cannabis and (unspecified) nicotine/tobacco products over the past 30 days [29]. In clinical populations, a study of 200 adult patients with chronic rhinosinusitis who smoked reported that 50% were co-users of tobacco and cannabis by design [27], and among another sample of 283 first-episode psychosis adult patients, approximately 51% reported lifetime co-use of cigarette and cannabis prior to symptom onset [26]. In a sample of 294 pregnant women, 66% of those who used cannabis during pregnancy also reported tobacco use, indicating substantial co-use during pregnancy [28].

Sleep Outcomes and Measures

Sleep outcomes were assessed mostly via subjective measures across the included studies with only one study reporting use of objective sleep measurements. Subjective measures included self-reported sleep duration (short <7 h, normal 7–9 h [NSF recommended], long >9 h over a 24-hour period) [24]; the Pittsburgh Sleep Quality Index (PSQI), a questionnaire assessing sleep quality and disturbances over the past month [28]; sleep items from the 20-Item Sino-Nasal Outcome Test [SNOT-20] assessing sleep difficulty, night awakenings, and fatigue [27]; single-item measures of sleep disturbance from the Beck Depression Inventory (BDI) [29]; and sleep disturbance as a general prodromal symptom of psychosis, as defined by the Symptom Onset in Schizophrenia inventory [26]. Objective measures were used in a single study, which employed actigraphy-derived rest-activity rhythms from the cross-sectional U.S. NHANES sample to assess interdaily stability, intradaily variability, and relative amplitude [25]. These actigraphy metrics are indirect measures of sleep, capturing rest-wake patterns rather than directly measuring sleep stages. Overall, most studies relied on subjective self-report, with only one study incorporating an objective but indirect measure of sleep patterns from a population-based cross-sectional study.

Associations between Co-use and Sleep

Across the six included studies, evidence linking nicotine-cannabis co-use to sleep outcomes was mixed. Three studies reported significant associations between nicotine-cannabis co-use and poorer sleep outcomes, including increased short or long sleep duration, disrupted rest-activity rhythms, and sleep difficulties and awakenings [24, 25, 27]. In contrast, three studies did not find significant effects of co-use on sleep [26, 28, 29]. These mixed findings may reflect differences in populations, sample sizes, co-use definitions, or heterogeneity in sleep assessment approaches.

Importantly, analytic comparison groups varied across studies. While some articles compared co-use, single-substance use, and non-use groups [24, 25, 29], others contrasted co-use only with non-use or tobacco-only groups [26–28]. The absence of cannabis-only comparators in these studies limits the ability to determine whether co-use confers additional sleep-related risk beyond cannabis use alone [26–28]. In addition, the wide variation in sleep assessment methods may also likely contribute to the different findings. Together, this heterogeneity complicates interpretation and makes it difficult to isolate whether observed sleep outcomes reflect a distinct effect of nicotine-cannabis co-use or is primarily driven by nicotine, cannabis, or correlated behavioral or other factors. To better contextualize these findings, results are summarized below by population.

Population-based Studies

The two U.S. population-based studies provide the strongest evidence for examining co-use patterns in relatively large and population-based samples, though they assessed sleep using different approaches. One study using NHANES 2015–2018 data (N=6,573), found that adults who co-used e-cigarettes and cannabis had higher odds of short or long sleep durations relative to non-users [24]. The study also included single-use comparisons and reported that e-cigarette and cannabis co-users had higher odds of short sleep than e-cigarette-only users, suggesting that co-use may increase vulnerability to insufficient sleep beyond nicotine vaping alone. In sub-analyses, cigarette and cannabis co-users had higher odds of long sleep duration compared with non-users, indicating that cigarette-cannabis co-use may be linked to broader sleep dysregulation (i.e., sleep outside the NSF recommended normal range), rather than only short sleep. The second study used data from NHANES actigraphy substudy, which included 5,129 of U.S. adults and compared no use, cigarette-only, cannabis-only, and co-use groups using rest-activity rhythm metrics (capturing 24-hour patterns of rest and activity) [25]. Co-use was associated with lower interdaily stability and lower relative amplitude, reflecting less stable and less consolidated daily rest-activity patterns, with effects largely driven by cigarette smoking rather than cannabis use (as their analyses showed cannabis use alone was not significantly associated with rest-activity rhythm disruptions). While these studies provide valuable population-level insights, sleep outcomes were either self-reported [24] or indirectly captured through actigraphy-derived rest-activity rhythm metrics rather than direct measures such as sleep duration and quality [25]. Importantly, the use of objective sleep measures such as actigraphy or polysomnography is challenging in large national surveys due to cost, logistics, and participant burden. Future research may benefit from hybrid designs that combine population-based sampling with more intensive sleep assessment in nested subsamples to better clarify how co-use relates to sleep duration, efficiency, fragmentation, and architecture.

Clinical Populations

Findings from clinical samples highlight that co-use may worsen sleep-related symptoms in some health contexts, though patterns differ by condition. Among 200 chronic rhinosinusitis patients, a study found that co-use of tobacco and cannabis was associated with more self-reported sleep difficulties, nocturnal awakenings, and daytime fatigue compared with tobacco-only use [27], suggesting that in populations already burdened by inflammatory or respiratory symptoms and using cigarettes, cannabis smoke exposure may exacerbate sleep disruption. However, another study observed that 283 first-episode psychosis patients with cigarette use had higher prevalence of sleep disturbances [26], but co-use did not appear to increase sleep disturbance beyond tobacco-only use, suggesting that tobacco exposure (versus cannabis) may have been the primary driver of sleep disturbance among those with psychosis. These differences likely reflect the specific health contexts, including variations in underlying conditions, patterns and intensity of use, and neurobiological sensitivity to nicotine versus cannabis. For example, sleep disruption in psychosis may be more strongly linked to nicotine-related arousal or medication effects, whereas in respiratory conditions, additional smoke exposure from cannabis may have more direct physiological consequences on sleep quality [30].

Other Populations

Other populations also showed mixed evidence. In 294 pregnant women, one study compared cannabis users versus non-users during pregnancy and reported no statistically significant differences in sleep quality between cannabis users and non-users [28]. Notably, many cannabis users (66%) and one-third of non-cannabis users (33%) reported tobacco use during pregnancy, which may have influenced sleep outcomes and limits conclusions about the independent effects of co-use [28]. The study also did not report sleep outcomes stratified by co-use versus cannabis-only or tobacco-only groups, limiting conclusions about whether co-use may have unique associations with sleep during pregnancy. In another study of 221 adolescents and young adults, non-use, nicotine/tobacco-only, cannabis-only, and co-use groups were compared and no significant associations were observed between either co-use or single-substance use and sleep disturbance, overall or by sex [29]. However, this study assessed a relatively younger population, which may differ from older adults in sleep patterns, substance use behaviors, and susceptibility to sleep disruption. Additionally, the sample was small, and sleep was measured using a single-item assessment.

Limitations

Several limitations should be considered when interpreting this literature, which consists of only six studies. Half of the studies were cross-sectional [24, 25, 29], all but one relied on self-reported sleep measures [25], and across studies there was limited information on timing, frequency, dose, or mode of nicotine and cannabis use. Only one study incorporated objective monitoring, and it relied on actigraphy-derived rest-activity rhythms, which capture rest-wake patterning but do not directly measure sleep stages or architecture [25]. Several studies assessed populations in which sleep outcomes were secondary to other research aims (e.g., psychosis prodromal symptoms, chronic rhinosinusitis) and underlying medical or psychiatric conditions may have confounded observed associations. Sample sizes varied widely and only two studies included large population-based samples, but both derived from NHANES across different years [24, 25], underscoring the need for larger, independent datasets and replication across settings.