Weighted analysis of the association between cannabis smoking and chronic pain intensity among hemodialysis patients

Abstract

Chronic pain is a prevalent discomfort among hemodialysis patients. Managing this pain is challenging, which requires researching alternative treatments. Therefore, this study aimed to explore the average difference in chronic pain severity associated with smoking cannabis in hemodialysis patients. A cross-sectional observational study was conducted among 186 males. The exposure was the use of cannabis smoking and chronic pain severity was the primary outcome. Confounding variables were collected through a structured questionnaire. The severity of chronic pain was measured using the Numerical Rating Scale. Statistical analysis was performed using the Inverse Probability of Treatment Weighting to reduce imbalance in measured covariates between control and smoker groups. After weighting, all confounder variables were well balanced. The average difference in pain severity associated with smoking cannabis was β =-1.541 (95% CI, -2.219 to -0.863). The sensitivity analysis, including the analgesic use, indicates that the association was modestly attenuated compared to the main model, β = -1.299 (95% CI, -1.868 to -0.730). Cannabis smoking is associated with a decrease in mean pain severity compared with non-smokers. The sex-specific sample, observational design and potential unmeasured confounding constitute the main limitations of this study. Further research should focus on prospective studies with more exposure details.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1038/s41598-026-57047-7.

Introduction

Regardless of its origin, kidney disease is a silent disease that progresses slowly. This condition is a public health concern worldwide, managed by renal replacement therapies. Hemodialysis is the most widely used treatment for patients with end-stage chronic kidney disease (ESKD). However, patients undergoing hemodialysis suffer from multiple physical symptoms, with chronic pain being a prevalent discomfort among this population¹.

Comorbidities, ESKD complications and primary renal diseases are the main causes². Chronic pain is continuous, varying in intensity from moderate to severe and can affect the shoulders, knees, head, and back¹. In addition, chronic pain has a negative impact on health-related quality of life. This symptom is associated with alteration of physical function³, psychological disorders: anxiety and depression⁴ and daily activities⁵. Pain can lead to a reduction in hemodialysis session duration, an increase in the possibility of hospitalization and emergency department visits⁶. Multiple biological, sociological and clinical factors are involved in chronic pain in hemodialysis patients. The sex, Vitamin D⁷ and the Parathormone (PTH) level⁸ are associated with pain in hemodialysis patients. Residence, marital status and level of education can contribute to chronic pain in patients treated with hemodialysis⁹. In addition, comorbidities, number of painful sites and hemodialysis vintage can be related to chronic pain in this group⁹.

Chronic pain management in hemodialysis patients is a significant challenge. The pharmacological approach is based on acetaminophen for mild pain, Non-Steroidal Anti-Inflammatory Drugs (NSAIDs) and opioids for moderate to severe pain; the use of these drugs is associated with different side effects. NSAIDs can lead to cardiovascular side-effects, such as hypertension, heart failure, congestive heart failure, myocardial infarction and atrial fibrillation, and can even induce death¹⁰. Opioid consumption in hemodialysis patients is associated with multiple increased risks, including hospitalization, worsened morbidity, dialysis discontinuation and death¹¹. Despite the prescription of these medications, chronic pain in this population is inadequately controlled¹². Therefore, other drugs need to be explored.

The medical use of cannabis is growing in interest, particularly in the setting of chronic pain. In fact, there is significant evidence supporting the use of this substance to manage chronic pain¹³. Cannabis (marijuana) is a flowering plant that belongs to the Cannabaceae family, originally from Central or South Asia. It contains 150 different cannabinoids; the most active components are Δ-9-tetrahydrocannabinol (THC) and cannabidiol (CBD). The two cannabinoids have comparable chemical compositions; however, their modes of action and impacts on brain activity are very different¹⁴. THC has a psychoactive effect, is used to treat seizures, Huntington’s disease, sleep disorders and multiple sclerosis¹⁴. CBD has a therapeutic opportunity in anxiety, sleep quality¹⁵, autism spectrum¹⁴ and opioid addiction¹⁶. Furthermore, both molecules possess analgesic properties¹⁴.

Cannabis is cultivated in Morocco, especially in the northern regions of Al-Hoceima, Chefchaouen and Ouazzane¹⁷. Dried marijuana (Kif and Hashish) is used for recreational purposes. Mashed leaves can be smoked with tobacco or alone using a Sebsi (Moroccan pipe) and the extracted resin is smoked combined with tobacco. Cannabis is the most commonly used psychoactive substance among male students in middle and high schools in northern Morocco, with a prevalence rate of 13.5%¹⁸. Cannabis has been studied in several populations; however, the evidence on its effectiveness remains unclear, particularly in patients treated by hemodialysis. In a descriptive cross-sectional study of Ho et al¹⁹ describing the characteristics and perspective of cannabis use among hemodialysis patients, the authors did not explore the relationship between pain and cannabis use. In addition, in Bassat’s study²⁰, the authors assess the safety of medical cannabis in hemodialysis patients; however, the recreational use of cannabis is not included. Studies are needed to determine the relationship between chronic pain and recreational cannabis use among patients treated with hemodialysis. Therefore, this study aimed to explore the average difference in chronic pain severity associated with smoking cannabis in patients undergoing hemodialysis.

Methods

Study design and data collection

This is a cross-sectional study conducted in the public hemodialysis centers of the northern region of Morocco, assessing the weighing comparison of smoking cannabis on chronic pain severity in patients treated with hemodialysis. The study included male patients receiving chronic hemodialysis who were at least 18 years old, had been undergoing chronic hemodialysis for a minimum of six months and provided their consent to participate. Excluded criteria were cancer diagnosis, mental or hearing disorders, or the use of other forms of cannabis (non-smoked). A combination of sociocultural and epidemiological factors, without any intentional selection bias, led to the exclusion of women from the study. In Morocco, from a sociocultural perspective, cannabis use is more socially acceptable among men than among women. This difference in social norms may make women less likely to report this type of behavior due to fear of judgment, stigma or potential repercussions for their families, resulting in non-disclosure or reluctance to participate in the study. From an epidemiological perspective, the available data consistently show a significantly higher prevalence of cannabis use among men in the general population.

A total of 190 patients met the eligibility criteria and were included in the study, and were selected through convenience sampling. The response rate for the questionnaire was 97.89%; four patients (2.11%) were excluded as they were unable to fill out the questionnaire for the following reasons: one due to high blood pressure, two due to cramps and one suffered from headaches. These cases were treated as missing data; consequently, they were removed from the analysis, and 186 were included in the study (each observation with missing data has been excluded). Data were collected from questionnaires and medical records during dialysis sessions, ensuring participant anonymity, between 26 September 2024 and 15 May 2025.

The exposure of this study was the recent use of cannabis smoking, classified as a binary variable: cannabis smokers were the exposure group and the non-smokers were the control group. A cannabis smoker was defined as any participant who reported recreational cannabis use (marijuana) in the previous three months, with a minimum frequency of at least once per month. The consumption was through traditional pipe smoking (sebsi), mixed with tobacco. The non-smoker was a person who had either never used cannabis or had quit using it at least three months before the study.

The primary outcome was the severity of chronic pain as a continuous variable. The chronic pain was described as interdialytic discomfort that is frequent or lasts longer than three months²¹. Chronic pain related to cancer has been excluded. The Numerical Rating Scale (NRS) was used to assess pain intensity. This pain screening tool allows patients to select a number from 0 to 10 that best describes their level of pain²². The additional covariates were assessed using a structured questionnaire completed face-to-face between the investigator and participants. The data included age, education status, residence and marital status. Patients’ records were used to collect the following data: diabetes, hypertension, Body Mass Index (BMI), dialysis vintage, duration of dialysis sessions and parathormone (PTH).

Ethical consideration

The study was conducted following the tenets of the Helsinki Declaration and approved by the Ethics Committee for Biomedical Research of the MOHAMMED V Faculty of Medicine and Pharmacy in RABAT (CERB 28 − 24). All methods were performed in accordance with the relevant guidelines and regulations. Participants were provided with a written consent form along with the questionnaire. This form outlined the purpose of the study and assured them of the privacy and confidentiality of their data.

Statistical analysis

The study was conducted using the Statistical Package for the Social Sciences (SPSS) version 26. The Kolmogorov-Smirnov test was used to assess the normality of the variables. Statistical analysis was based on the Inverse Probability of Treatment Weighting (IPTW) method²³. The Propensity Score (PS) was utilized to adjust the measured confounders by establishing balance between cannabis smokers and the control group to address bias. The score was based on the estimated probability (ranging from 0 to 1) of cannabis smoking for each participant in the two groups, considering their characteristics. The PS was calculated using logistic regression. The model included confounders as covariates: age, illiteracy, urban, married, diabetes, hypertension, BMI, dialysis vintage, dialysis duration session and PTH. The PS was used to calculate the IPTW for each individual. Every patient was weighted (W) by the inverse of the probability of using cannabis:

W _Exposure = 1/ PS and W _control = 1/ (1 - PS).

Figure 1 summarizes the different steps of statistical analysis using the Inverse Probability of Treatment Weighting method.

Fig. 1.

Different steps of statistical analysis using the Inverse Probability of Treatment Weighting method.

Weighting by this value generated a pseudo-population where the confounder distributions among the exposed and control groups were identical to the overall distribution of these confounding factors in the total population of origin. The IPTW balance was checked using the standardized differences (SDs) between the exposure and control groups for all characteristics, both before and after weighting. The SD was calculated for continuous variables by the difference in means and for categorical variables by the difference in prevalence. SDs < 0.1 were considered negligible and an acceptable imbalance between the two groups. However, SDs ≥ 0,1 after weighting were large, indicating a significant imbalance, necessitating a revision of the PS model. To estimate the population-averaged association between cannabis smoking and chronic pain severity, the IPTW-weighted generalized estimating equations (GEE) were used, specifying a Gaussian distribution (normal distribution), identity link function, independent working correlation structure and robust standard errors. The result was significant at p < 0.05. Results were expressed as beta coefficients (β), representing adjusted mean differences in NRS.

For the purpose of evaluating the robustness of the observed association between cannabis smoking and the severity of chronic pain, the principal IPTW-weighted GEE model was further adjusted for analgesic use as a sensitivity analysis. Considering analgesic use may reflect pain management strategies and could potentially be on the causal pathway between exposure history and pain outcomes, it was excluded from the primary propensity score model. Therefore, variation in the direction, magnitude and statistical significance of the association was examined to determine the stability of the primary results under this alternative model specification.

Results

Study population

A total of 190 hemodialysis patients were eligible for the study. Four patients were excluded due to missing data, resulting in 186 patients included in the analysis. The mean age of patients was 53.46 ± 15.47. Participants aged 50 or over represented 71.4% of the sample. The results revealed that 38.2% of patients resided in urban areas, 65.6% were illiterate, and 75.8% were married. Among participants, 25% were overweight (> 25.39) and 9.1% had a shorter dialysis duration (< 4 H). The mean duration of dialysis was 64.05 ± 56.67 months, with 50% of patients undergoing dialysis for 48.50 months. Table 1 displays the characteristics of participants included in the study.

Table 1.

Characteristics of participants included in the study (n = 186).

Characteristics	Overall N (%)
Age	53.46 ± 15.47
Illiterate	122 (65.6)
Urban	71 (38.2)
Married	141 (75.8)
Diabetes	33 (17.7)
Hypertension	123(66.1)
Dialysis vintage/ month	64.05 ± 56.67
Dialysis duration session <4 h	17 (9.1)
BMI	23.23 ± 3.99
PTH (pg/ml)	307.92 ± 216.92

BMI: Body mass index; PTH: Parathormone.

Confounders balance

Before balancing, the control group consisted of 152 patients undergoing hemodialysis, while the smoker group included 34 patients. Before weighting, the baseline comparison showed a significant imbalance across several assessed confounders, suggesting that the cannabis-exposed and non-exposed groups presented significantly different patient characteristics. The urban residency revealed the most significant imbalance (SMD = 0.48), indicating a significant baseline disparity and an elevated risk of confounding bias. Dialysis vintage (SMD = 0.21), dialysis session duration (SMD = 0.149), and several other covariates exceeding or approaching the standard threshold of 0.10 were found to have additional moderate imbalances.

After weighting, covariate balance significantly improved for all included variables. Importantly, most variables showed near-complete balance (urban residency; SMD = 0.001; married; SMD = 0.0079), with all post-weighting SMDs decreasing below the generally recognised threshold of 0.10. This significant decrease in imbalance indicates that the propensity score model effectively generated a pseudo-population in which measured confounders were well balanced between groups (See Table 2 in supplementary files). Figure 2 shows the love plot of standardized mean differences before and after weighting of patients’ characteristics included in the propensity score model.

Fig. 2.

Standardized mean differences before and after weighting according to the characteristics included in the propensity score model.

Outcome model and sensitivity analysis

In the GEE of weighting data, smoking cannabis (cannabis exposure) was independently associated with a statistically significant reduction in pain severity, β = -1.541 (95% CI, -2.219 to -0.863). This negative beta coefficient indicates that, after adjustment for measured confounders, patients with prior cannabis exposure had, on average, lower pain severity scores (-1.541 in mean) compared with non-exposed patients. The magnitude of this association suggests a clinically significant inverse relationship, which supports the hypothesis that prior exposure to cannabis may be associated with a reduction in pain burden in this population.

In the sensitivity analysis, which included the use of analgesics, the association between cannabis use and lower chronic pain intensity remained highly statistically significant, β = -1.299 (95% CI, -1.868 to -0.730). Although the magnitude of this association was modestly attenuated compared to the main model, the persistence of a strong negative coefficient supports the robustness of the main findings and indicates that differences in analgesic treatment do not entirely explain them.

The use of analgesics was independently associated with greater intensity of chronic pain (β = 4.620, p < 0.001), indicating that patients taking analgesics reported, on average, a higher intensity of pain. This finding is clinically plausible, as analgesic use may reflect a greater underlying symptom burden or more severe pain requiring pharmacological treatment. Therefore, accounting for analgesic use did not reverse or eliminate the association between cannabis and pain, which reinforces the robustness of the primary results of the IPTW analysis. Table 3 displays the IPTW-weighted mean difference in chronic pain severity associated with cannabis smoking among hemodialysis patients, including sensitivity analysis additionally adjusted for analgesic use.

Table 3.

Average difference in chronic pain severity between cannabis smoking vs. non-smoking among hemodialysis patients: primary IPTW-weighted analysis and sensitivity analysis.

Variable	IPTW primary Model β (95% CI)	P value	Sensitivity model* β (95% CI)	P value
Cannabis smoker group	-1.541 (-2.219 to -0.863)	< 0.001	-1.299 (-1.868 to -0.730)	< 0.001
Analgesic use			4.620 (4.042 to 5.199)	< 0.001

*Sensitivity model additionally adjusted for analgesic use.

Discussion

Recreational cannabis is occasionally used by patients undergoing hemodialysis. This study determines the average difference in chronic pain severity associated with smoking cannabis in patients undergoing hemodialysis using the IPTW. After weighting, all variables were well-balanced, and the SMDs were less than 0.1. According to the results, the average difference in chronic pain severity associated with smoking cannabis versus non-smoking in hemodialysis patients was β=-1.541 with p < 0.001. The sensitivity analysis, additional adjustment of the primary IPTW-weighted model for analgesic use, reveal that the association between cannabis smoking and chronic pain severity remained highly statistically significant. However, the magnitude of the effect was modestly attenuated compared with the primary weighted analysis. Specifically, the regression coefficient for cannabis smoking changed from (primary β=-1.541) in the main IPTW model to (sensitivity β= -1.299) after additional adjustment, suggesting that analgesic use may partially explain the observed association.

Meglin’s study demonstrated the efficacy of smoking cannabis in improving restless legs syndrome and sleep quality²⁴. Furthermore, Hill’s clinical review of 24 trials indicates positive findings, suggesting that marijuana may be effective for chronic pain, neuropathic pain and multiple sclerosis^25, and has a potential contribution in chronic pain relief²⁶. In this research, the result indicates that cannabis smoking is negatively associated with chronic pain severity.

Similarly, the National Academies declared that there is significant evidence for the beneficial use of cannabis and cannabinoids in the treatment of chronic pain²⁷. A double-blind, randomized study examined the effectiveness of cannabinoid extracts (THC: CBD) in managing cancer pain of at least moderate intensity. The results indicated a significant improvement in the mean NRS scores for patients receiving the cannabinoid extracts compared to those who received a placebo, with scores of 1.37 versus 0.69, respectively²⁸. Likewise, the finding of this study revealed that the mean of the NRS score for chronic pain was significantly 1.299 points less in the cannabis smoker group compared to the control group, after adjusting for analgesic use.

Another randomized experimental study evaluated various types of cannabinoids for managing chronic pain in patients with fibromyalgia. The results showed a 30% reduction in the Visual Analog Scale (VAS) score after patients inhaled vaporized cannabinoids, along with a significant increase in their pressure pain threshold²⁹. Additionally, a randomized controlled trial assessing the effect of smoking cannabis on chronic neuropathic pain, the finding showed a reduction in average daily pain level using THC (9.4% v. 0%), with a difference of 0.71, 95% [CI, 0.02–1.4] ³⁰.

This study demonstrates the association of chronic pain severity and smoking cannabis in patients undergoing hemodialysis. In terms of the pharmacodynamics of cannabis, smoked cannabis improves mood and facilitates sleep³⁰. However, smoking cannabis could harm lung health and significantly raise the risk of cancer and Chronic Obtructive Pulmonary Disease in the general population³¹. These pharmacological considerations are presented solely to contextualize existing literature and should not be interpreted as mechanisms directly demonstrated by our study. The recommendation of cannabinoid smoking to treat chronic pain in patients undergoing hemodialysis should be made with particular caution, especially with the challenges of drug pharmacokinetics in this population. The results indicate that cannabinoids may be effective in treating chronic pain in patients undergoing hemodialysis.

Chronic pain in patients on hemodialysis is increasingly being approached from a biopsychosocial perspective. Depressive symptoms are a major factor contributing to poor treatment adherence among patients with ESKD undergoing hemodialysis³²; Furthermore, significant links have also been identified between the burden of chronic pain and anxiety as well as depression⁴, highlighting the complex interdependence between psychological distress, treatment adherence and symptom severity in this population. Psychosocial or talking therapies, particularly cognitive behavioral therapy, have demonstrated beneficial effects in hemodialysis populations, especially through improvements in health-related quality of life, anxiety and depression, which may indirectly contribute to better pain coping and symptom perception^33–35. In addition, several reviews of non-pharmacological interventions suggest that talking therapies may represent valuable complementary strategies for symptom management in dialysis settings, even though pain-specific outcomes remain less consistently studied^35,36. Nevertheless, there is currently a lack of direct evidence evaluating integrated multimodal approaches combining cannabis use with structured psychological therapies in hemodialysis patients. Therefore, while psychosocial interventions provide important contextual relevance for comprehensive symptom management, the role of combined cannabis and talking-therapy strategies remains speculative and should be considered a future research priority rather than a conclusion supported by current evidence.

Beyond psychosocial and symptom-related factors, the broader context of renal replacement therapy may also influence the burden of chronic pain, as kidney transplant recipients generally report a lower pain burden³⁷ and a significantly better quality of life than patients on dialysis³⁸. This suggests that chronic pain in dialysis patients may partly reflect the physiological, procedural and psychosocial stresses inherent in dialysis itself, rather than isolated factors related to exposure alone.

Although the present study focused on hemodialysis patients, cannabis use in peritoneal dialysis (PD) populations remains insufficiently studied. Current nephrology literature on cannabis in kidney failure is largely centered on hemodialysis, with very limited evidence including PD patients³⁹. Notably, one survey-based study examining cannabis use for dialysis complications in patients treated with dialysis included both hemodialysis and peritoneal dialysis participants, but it was not specifically designed as a PD-focused investigation⁴⁰. Modality-specific research is therefore needed to better understand cannabis exposure patterns and clinical outcomes in the growing PD population.

Limitations

The current study has some limitations. First, the sex-specific design limits the generalizability of the results to female populations, while convenience sampling and recruitment from a single region may reduce broader external validity through potential selection and contextual biases. Second, the absence of detailed data on cannabis dose, potential and duration of use precludes assessment of exposure intensity and chronicity. Furthermore, frequent co-consumption with tobacco via a traditional sebsi pipe may result in residual confounding that cannot be fully addressed. Third, although the propensity score model was expanded to include other clinically relevant variables, and the IPTW significantly improved the balance of the measured confounding factors, this approach does not account for unmeasured or unavailable factors, such as psychiatric symptoms, sleep disorders or other behavioral characteristics. Consequently, a residual confounding effect remains possible. Finally, the IPTW constitutes a helpful analytical approach for minimizing imbalance in measured confounding variables. Nevertheless, despite its methodological strengths, the inherent limitations of an observational design persist, and the findings should therefore be interpreted with caution, particularly in light of potential residual confounding and unmeasured variables. Conversely, this research represents the first study evaluating the average of chronic pain severity and smoking cannabis in this population.

Future directions

Future research should focus on prospective and longitudinal studies to provide a clearer understanding of the temporal and clinical relationship between cannabis use and the severity of chronic pain in patients undergoing hemodialysis. A more detailed assessment of exposure is particularly needed, specifically regarding dose, duration, route and cannabinoid composition. Furthermore, additional research should incorporate broader clinical and psychosocial confounding factors.

This study provides evidence that smoking cannabis is associated with chronic pain. The findings show that cannabis smoking is negatively associated with the severity of chronic pain in patients on hemodialysis. Additionally, the consistency of this inverse association, both in the main model and in the sensitivity analysis, strengthens confidence in the robustness of the observed relationship between cannabis use and the severity of chronic pain. Although the observational nature of the study precludes definitive causal conclusions. Despite these results, recommending cannabis smoking for treating chronic pain should be approached with caution, especially in this population. More research is needed to confirm these results, investigate the pharmacokinetics of these compounds in hemodialysis patients and evaluate safety and possible side effects.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, investigation, visualization and writing – original draft: FZB, Data Curation: NAW and KB, Resources: MA, Formal analysis and Methodology: AEG, Supervision: HZ, Validation and Writing – review & editing: MC.

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.