We examined whether self-rated physical health, mental health, and overall QoL were associated with cannabis use following a cancer diagnosis and whether these associations varied by cancer stage at diagnosis, treatment status, and pre-diagnosis cannabis use. Consistent with our hypothesis, survivors reporting average or below average physical health, mental health, and QoL were more likely to use cannabis after their diagnosis; however, only the association between mental health and post-diagnosis cannabis use was statistically significant. Additionally, while none of the interactions were statistically significant, average or below average mental health was significantly associated with higher odds of post-diagnosis cannabis use among survivors diagnosed with non-advanced cancer, post-cancer treatment survivors, and those who had never used cannabis prior to diagnosis. Of note, these findings should be interpreted with caution as estimates were imprecise.

Although limited, most previous studies have reported that cancer survivors who use cannabis tend to have lower QoL compared to those who do not use cannabis [20–23]. These studies, conducted primarily among colorectal cancer survivors and oncology patients, assessed QoL using validated multi-item instruments including the 30-item European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core 30 [20, 22], the Functional Assessment of Cancer Therapy-Colon [21], the Edmonton Symptom Assessment Score (ESAS) [23]. Across these studies, cannabis use was generally associated with poorer QoL or symptom outcomes, although Nathan et al. (2023) found that ESAS scores remained stable following cannabis initiation among older adults receiving medical cannabis for cancer-related symptoms [23]. These previous studies used comprehensive methodological approaches to evaluate cancer-related QoL among cannabis users and their findings were largely consistent with ours. Despite our reliance on single-item measures of health status and QoL, this consistency suggests that brief self-rated assessments may capture meaningful differences in health status associated with post-diagnosis cannabis use.

While previous studies have associated cannabis use with lower overall QoL, they have generally treated QoL as a single composite construct and have not examined physical health and mental health as distinct health domains. One exception is Xu et al. [24], who used Behavioral Risk Factor Surveillance System data to assess physically and mentally unhealthy days among cancer survivors and found that frequent cannabis use was associated with poorer outcomes across both domains. Do et al. [25] evaluated physical and mental health status as predictors of cannabis use using nationally representative longitudinal data but found that mental health was a significant predictor only among individuals without a history of cancer. Our study builds on this work by examining self-rated physical health, mental health, and overall QoL as separate health domains in relation to post-diagnosis cannabis use specifically among cancer survivors. This approach showed that average or below average mental health was the only health domain significantly associated with post-diagnosis cannabis. Our examination of effect measure modification by cancer stage at diagnosis, treatment status, and pre-diagnosis cannabis use represents a novel contribution, as previous studies have not systematically evaluated whether these associations differ by these characteristics.

Cancer stage at diagnosis is a key clinical marker associated with increased disease severity, treatment intensity, and prognosis. We hypothesized that cancer survivors diagnosed at more advanced stages would face more intensive treatment, ongoing symptoms, and greater physical and psychological burden that may motivate cannabis use [26]. In our prior work, we found that survivors diagnosed at stages III/IV had 63% higher odds of considering cannabis use compared to those diagnosed at stages I/II [17]. However, in this study, we found little evidence of heterogeneity in the associations between health status and QoL and post-diagnosis cannabis use, although stratified analyses suggested potentially higher odds of post-diagnosis cannabis use among survivors with average or below average mental health who were post-cancer treatment or who had never used cannabis prior to diagnosis. We hypothesize that survivors who are in post-treatment phases may experience fear of recurrence, persistent symptoms, financial stress, and reduced clinical support and may therefore turn to cannabis for emotional regulation, sleep, or coping during this period. Our findings by pre-cancer diagnosis use of cannabis suggest that cancer diagnosis may serve as an initiation point for cannabis use among some previously cannabis-naïve survivors, potentially for symptom management; however, this result may also reflect broad measurement of lifetime pre-cancer diagnosis use, post-diagnosis cessation among former users, reporting error, or residual confounding. Altogether, these findings suggest that both the decision to use cannabis and its perceived benefits may be shaped by survivors’ symptom burden, current health status, and previous experience with cannabis.

Cannabis use after a cancer diagnosis may reflect a symptom-driven, reactive approach to managing ongoing physical and psychological challenges, particularly among survivors who are no longer in active treatment but continue to experience long-term effects of cancer or its treatment. In our previous study, we found that survivors who perceived any of 15 listed benefits of cannabis use had more than a five-fold increase in the odds of reporting post-diagnosis cannabis use, whereas perceiving any risks was associated with a 59% decrease in the odds of post-diagnosis cannabis use [12]. Rather than representing a proactive pursuit of wellness, this pattern of use appears more closely tied to efforts to relieve distress or discomfort. This aligns with frameworks such as Symptom Management Theory, which highlights the central role of symptom experience and relief-seeking in patient behavior [27, 28]. In this context, cancer survivors may turn to cannabis to manage cancer-related symptoms while also considering potential side effects and broader social perceptions of use [29]. The Self-Medication Hypothesis further supports this interpretation, proposing that individuals may use substances like cannabis to cope with persistent physical or psychological distress [30–32], which is supported by our findings which consistently highlight mental health as a correlate of cannabis use among cancer survivors and thus a potential target of intervention given that clinicians may not be aware of cannabis use behaviors. Continued research is needed to better understand the complex and multifaceted drivers of cannabis use among cancer survivors.

This study has several strengths that enhance the robustness and interpretability of the findings. This study is based on a large, weighted sample drawn from an NCI-designated Comprehensive Cancer Center, allowing for greater generalizability to more diverse cancer survivor populations within academic clinical settings. The single-institution design may minimize variability in cancer care delivery, potentially reducing confounding related to treatment access or quality of care. Cannabis accessibility in California, including San Diego, is shaped by state-level legalization of medical and recreational cannabis during the study period, which may influence patterns of use and limit generalizability to regions with different legal frameworks. However, several limitations should be noted. The cross-sectional study design precludes assessment of temporality and does not allow descriptions of the causal direction of observed associations between cannabis use and self-rated health status and QoL. Because cannabis use and health outcomes are assessed concurrently in cross-sectional studies, the directionality of these associations remains unclear and may reflect bidirectional relationships. All single-item measures of health status and cannabis use were subject to recall or social desirability bias. While multi-item QoL scales are valuable for clinical and psychometric applications, the use of single-item QoL measures is well supported in epidemiological research and has several advantages including being more time- and cost-effective, while being equally valid and reliable [33–38]. Cannabis use was not verified through medical records or biological measures, potentially resulting in misclassification. Additionally, THC and CBD were assessed together and are notably different compounds with distinct pharmacologic effects; combining them precludes examination of differential associations with health outcomes and may obscure potential variation in benefits or risks related to specific cannabinoid use. Other aspects of the patient’s cancer journey such as information on treatment modality, time since diagnosis, and recurrence history were not collected, but may be important to consider as these factors may provide insight into the differences in self-rated health status across survivors. Although survey weights were applied to improve representativeness, the response rate was modest, and selection bias remains a possibility. Finally, residual confounding from unmeasured variables like provider recommendations or concurrent use of other complementary therapies cannot be ruled out as these factors may influence the survivor’s health status.

Future research should aim to better understand the diversity of the survivorship experience and how cannabis use fits into this landscape. As survivorship experiences vary, important differences in health status may emerge based on a multifactorial combination of clinical characteristics, treatment history, and psychosocial or behavioral factors. Future work therefore must investigate how these factors influence cannabis use and whether survivors with different clinical histories turn to cannabis for distinct reasons. Doing so would help ensure more reliable comparisons across studies and support a clearer understanding of how cannabis use relates to survivors’ perceived well-being. Longitudinal studies in particular will help clarify the timing and persistence of cannabis use relative to changes in health status. Qualitative approaches will also be important for adding depth by capturing how survivors perceive cannabis as part of their broader coping or wellness strategies. Incorporating detailed measures of cannabis exposure such as product type, frequency, and route of administration, will be important for guiding clinical discussions and informing survivorship care planning.