Effectiveness of Gross’s emotion regulation training on death anxiety in leukemia patients undergoing chemotherapy: a randomized controlled trial

Abstract

Background

Death anxiety is common among leukemia patients undergoing chemotherapy and can adversely affect quality of life. This randomized controlled trial evaluated the short-term effect of Gross’s emotion regulation training on death anxiety in this population.

Methods

This single-blind (data analyst-blinded), parallel-group randomized controlled trial was conducted at Omid Hospital, Isfahan, Iran, in 2024. Sixty adult leukemia patients undergoing chemotherapy were randomized 1:1 to receive either six virtual sessions of Gross model-based emotion regulation training or standard care. Death anxiety was assessed using the Templer Death Anxiety Scale at baseline and immediately after the intervention. The primary analysis used a linear mixed-effects repeated-measures model including group, time, and group × time interaction as fixed effects and participant as a random effect. ANCOVA was considered but not retained because the homogeneity of regression slopes assumption was not supported. Mean differences, 95% confidence intervals, and Cohen’s effect sizes were reported. Analyses were performed using SPSS v26, with α = 0.05.

Results

Baseline characteristics and death anxiety scores were comparable between groups. Baseline TDAS scores were 11.03 ± 4.59 in the intervention group and 10.87 ± 4.54 in the control group (mean difference = 0.17, 95% CI: −2.19 to 2.53; p = 0.888). Post-intervention scores decreased to 5.97 ± 3.64 in the intervention group but remained relatively unchanged in the control group (11.07 ± 4.67). The mixed-effects model showed a significant group × time interaction (β = −5.27, 95% CI: −6.40 to − 4.13; p < 0.001). The post-intervention between-group difference was significant (mean difference = − 5.10, 95% CI: −7.26 to − 2.94; p < 0.001; Cohen’s d = − 1.22). No adverse events were reported, and intervention adherence was 100%.

Conclusions

Virtual Gross model-based emotion regulation training was associated with a significant short-term reduction in death anxiety among leukemia patients undergoing chemotherapy. These preliminary findings should be interpreted considering the short-term assessment, single-center design, and absence of blinded outcome assessment.

Trial registration

IRCT20141127020108N5. Registered on September 7, 2022. ( https://irct.behdasht.gov.ir/).

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s40359-026-05142-9.

Background

One of the most significant psychological effects experienced by cancer patients is death anxiety, defined as a profound fear or concern regarding death or the dying process [1, 2]. The ambiguous and unpredictable nature of death serves as a substantial source of anxiety for individuals facing life-threatening illnesses [3]. Death anxiety refers to the negative emotions experienced when a person reflects on the inevitability of their own death, which is common among patients with cancer [4].

Research indicates that death anxiety is prevalent among patients with advanced cancer and can have a considerable impact on their quality of life, psychological adjustment, and end-of-life care. A recent systematic review highlights that death anxiety stands out as one of the most significant psychological challenges faced by cancer patients, necessitating focused interventions [2, 5–6].

In patients with leukemia, one of the most common malignancies in adults [7], the psychological situation is particularly complex because of the aggressive nature of the disease, prolonged treatment regimens, and significant side effects associated with chemotherapy. Leukemia can cause profound emotional problems, such as anxiety, feelings of shock, fear, isolation, anger, irritability, and confusion. These factors contribute to an increased likelihood of psychological distress, death anxiety, and depression [8, 9]. Cancer-related death anxiety may serve as a predictor of poorer psychological functioning throughout the treatment process [8]. Additionally, evidence shows that death anxiety is notably prevalent among patients with cancer, particularly among older adults and those with advanced disease, and may adversely affect psychological well-being, quality of life, and adaptation to the illness [10, 11].

Despite the significance of death anxiety and its effects on mental health, it is often undiagnosed and undertreated due to limited health literacy and knowledge of psychological treatments. Recent studies emphasize the need for educational and psychological interventions to improve the mental health of cancer patients [4]. Fear and anxiety about death are pressing concerns, and various psychological interventions have been introduced to alleviate these issues. As these challenges are typically emotional in nature [12], approaches focusing on both emotional and cognitive aspects are essential, including cognitive emotion regulation techniques [13].

In recent years, emotion regulation has received increasing attention as a psychological process relevant to anxiety, depression, and distress among patients with chronic illnesses, including cancer. Emotion regulation encompasses a range of cognitive and behavioral processes that enable individuals to monitor, assess, and modulate their emotional responses. Gross’s process model represents one of the primary theoretical frameworks in this area and outlines five key strategies: situation selection, situation modification, attentional deployment, cognitive change, and response modulation [14–19]. In the context of death anxiety, this model is clinically relevant because patients’ fear of death may be intensified by threat-focused attention, catastrophic appraisal of illness-related cues, and maladaptive emotional responses. Training patients to redirect attention, reappraise death-related thoughts, and modulate emotional responses may therefore reduce the intensity and persistence of death-related anxiety [5, 14–19].

Research indicates that high levels of death anxiety are linked to decreased resilience, heightened distress, and impaired social and emotional functioning. In this context, emotion regulation strategies are crucial for alleviating psychological distress, enhancing resilience, and improving the quality of life for cancer patients [20–22]. There is a significant inverse relationship between emotion regulation and death anxiety, suggesting that effective emotion management can lead to a decrease in death anxiety among the elderly [23]. A study conducted in 2021 found that enhancing emotional skills in older adults contributes to reduced death anxiety [1]. Furthermore, recent findings indicate that emotion regulation strategies—particularly cognitive restructuring and emotional suppression—play a mediating role in lowering death anxiety in cancer patients [5]. In a 2023 study by Orang et al. in Iran, an emotion regulation-based program significantly diminished death anxiety among elderly women with cancer [10]. Additionally, research by Chukwuorji et al. demonstrated that emotion regulation strategies alleviate death anxiety, particularly concerning death acceptance and intrusive death-related thoughts [24].

Although findings indicate positive effects of emotion regulation interventions on patients’ death anxiety, the results remain contradictory regarding their overall impact. Khamoushi et al. (2019) conducted a descriptive correlational study involving 150 cancer patients in Kermanshah, Iran, to explore the relationship between emotional self-regulation and death anxiety. The study found no significant association between emotional self-regulation and death anxiety in this population [25]. Additionally, Maqbool et al. (2025), in a study of 300 patients who had experienced myocardial infarction in Iran, reported no significant association between death anxiety and resilience, a construct related to emotional adaptation [21].

Although some efforts have been made to evaluate emotion regulation interventions in patients with breast cancer and older individuals with cancer, no study was identified that specifically evaluated emotion regulation interventions based on the Gross model in relation to death anxiety among leukemia patients. Given the significant prevalence of death anxiety in patients with leukemia, along with its detrimental impact on various aspects of health, the treatment process, and overall quality of life, there is a pressing need for more clinical trials in this area. Therefore, this randomized controlled trial aimed to evaluate the effect of Gross model-based emotion regulation training on death anxiety in leukemia patients undergoing chemotherapy.

Methods

Study design and setting

This study was a parallel-group, single-blind (data analyst-blinded) randomized controlled trial with a 1:1 allocation ratio. The trial was prospectively designed within a superiority framework to test whether Gross’s emotion regulation training was more effective than standard care in reducing death anxiety among leukemia patients undergoing chemotherapy. The study was conducted at Omid Hospital, affiliated with Isfahan University of Medical Sciences (IUMS), Isfahan, Iran. Omid Hospital is a specialized oncology center providing chemotherapy, radiotherapy, and stem cell transplantation services. This trial was reported in accordance with the CONSORT 2010 statement for randomized controlled trials, and the completed CONSORT 2010 checklist is provided as supplementary material.

Study participants

The study population comprised adult patients with leukemia undergoing chemotherapy at Omid Hospital. Inclusion criteria were as follows: age 18–65 years, literacy, absence of intellectual disabilities or sensory impairments (blindness or deafness), ability to operate a smartphone independently or with assistance from a family member, and access to the WhatsApp application. Exclusion criteria included concurrent participation in similar research studies, presence of acute psychological disorders at baseline, development of any medical or psychological condition during the study that precluded continued participation, withdrawal of informed consent at any point, and death during the study period.

Sampling and randomization

Convenience sampling was used by approaching eligible patients in the chemotherapy ward and hospital units. After the study objectives were explained and written informed consent was obtained, eligible participants completed the baseline assessment and were then randomly assigned to either the intervention group or the control group.

The target sample size was set at 30 participants per group, considering a two-sided type I error rate of 0.05 (α = 0.05), a statistical power of 80% (β = 0.20), and an anticipated attrition rate of 20%. Because no universally accepted minimal clinically important difference (MCID) has been established for the Templer Death Anxiety Scale in leukemia patients, the sample size was justified based on the ability to detect a large standardized between-group effect rather than a predefined MCID. With 30 participants per group, the study had approximately 80% power to detect a standardized effect size of about Cohen’s d = 0.74 at a two-sided α level of 0.05. Accordingly, a total of 60 participants were allocated to the two study groups.

Randomization was performed using block randomization with a fixed block size of 4 and a 1:1 allocation ratio. No stratification was used. The randomization sequence was prepared by an independent statistician who had no direct contact with participants and was not involved in enrollment, intervention delivery, outcome assessment, or data analysis. Allocation concealment was maintained using sequentially numbered, opaque, sealed envelopes (SNOSE) prepared according to the randomization sequence. The envelopes were kept sealed until after participant enrollment and completion of the baseline assessment and were opened sequentially by the study coordinator in numerical order. Therefore, group assignment was not known before enrollment and baseline data collection.

Blinding

Participants were enrolled and completed baseline assessments before randomization. After allocation, participant blinding and interventionist blinding were not feasible due to the behavioral and educational nature of the intervention and the use of standard care as the control condition. The researcher who delivered the intervention also administered the outcome assessments; therefore, the outcome assessor was not blinded to group allocation. No separate blinded outcome assessor was available for post-intervention assessment; therefore, blinding of outcome assessment could not be implemented in this trial. However, the primary outcome was assessed using a standardized self-administered questionnaire with predefined scoring procedures. To reduce analytical bias, all data were coded and analyzed by an independent statistician who remained blinded to group allocation throughout the analysis. Accordingly, this trial is best characterized as a single-blind randomized controlled trial with blinded data analysis.

In summary, enrollment was performed by clinical staff; the randomization sequence was prepared by an independent statistician; allocation was concealed using SNOSE; intervention delivery and outcome assessment were conducted by the study researcher; and statistical analysis was performed by a blinded data analyst. The CONSORT flow diagram for participant selection, allocation, follow-up, and analysis is presented in Fig. 1.

Fig. 1.

CONSORT flow diagram of participant enrollment, allocation, follow-up, and analysis

Data collection instruments

Data were collected using self-administered questionnaires, specifically the Demographic and Clinical Information Form. This form collected information on age, marital status, employment status, income, duration of illness, education level, time elapsed since diagnosis, type of leukemia, and the number of chemotherapy sessions.

The Templer Death Anxiety Scale (TDAS), developed by Templer in 1970, was used to assess death anxiety. The TDAS is a validated 15-item yes/no questionnaire designed to evaluate death anxiety, with scores ranging from 0 to 15. Each “yes” response is scored as 1, and higher total scores indicate greater death anxiety. A score above 8 suggests a higher level of death anxiety. The TDAS has demonstrated acceptable reliability, with Cronbach’s alpha values ranging from 0.49 to 0.73 and test–retest reliability of 0.83. Additionally, the Persian version of the TDAS was validated by Rajabi and Bahrani through convergent validity, showing a correlation of 0.34 with the State–Trait Anxiety Inventory (STAI) [1, 5, 26–28]. In the current study sample, the internal consistency of the Persian TDAS was assessed using Cronbach’s alpha coefficient. The TDAS demonstrated good internal consistency (Cronbach’s α = 0.82), supporting the reliability of the questionnaire for assessing death anxiety among leukemia patients undergoing chemotherapy.

The primary outcome was death anxiety measured by the Templer Death Anxiety Scale (TDAS) at baseline and post-intervention; no secondary outcomes were assessed. Adverse events were monitored throughout the study by clinical staff and participant self-report. No harms or unintended effects were observed or reported during the trial.

Intervention & program overview

The intervention consisted of a program based on the Gross model of emotion regulation, comprising six sessions of two hours each. The program focused on developing adaptive strategies through education, self-assessment, and practice. This six-session program, adapted from interventions based on the Gross model, was designed to enhance emotion regulation skills in clinical and educational settings, such as nursing or psychology. Each session lasted approximately two hours and included a check-in (10 min), core teaching/exercises (90 min), homework review/assignment (15 min), and feedback (5 min) [5, 29–32].

The intervention group participated in six virtual training sessions on emotion regulation based on Gross’s model, conducted via WhatsApp, with two sessions per week. Each session included multimedia educational content designed for easy access on smartphones. In addition to the online sessions, participants were provided with multimedia resources via WhatsApp for offline use throughout the week. The training covered various topics, including emotional awareness, vulnerability assessment, problem-solving skills, interpersonal skills, attention shifting, cognitive restructuring, emotional discharge, relaxation techniques, and behavioral activation. The sessions were led by a master’s-level psychiatric nursing expert with prior experience in emotion regulation training.

Adherence was monitored by the intervention facilitator using session attendance records and follow-up phone calls during the three-week intervention period. Participants were contacted to confirm receipt of the WhatsApp-based educational materials, resolve possible access or technical problems, and encourage continued participation. Session completion was defined as participation in the scheduled WhatsApp-based session or confirmed review of the session materials when asynchronous access was needed. Full intervention adherence was defined as completion of all six planned sessions. Retention for the primary outcome analysis was defined as completion of both baseline and post-intervention TDAS assessments.

Meanwhile, the control group received standard care throughout the study period and was offered the training after the completion of the study. No additional concomitant care was provided beyond the allocated intervention or standard care. The program is presented in Table 1, organized into distinct sessions [5, 29–32].

Table 1.

Training sessions for emotion regulation training based on Gross model

Session	Duration	Objectives	Key Activities	Homework
1	2 h	Build group rapport; introduce Gross model stages.	Icebreakers; overview of emotion process (situation, attention, appraisal, response); emotion identification exercise.	Track daily emotions and triggers in a journal.
2	2 h	Recognize emotions and situations.	Discuss emotion types, short/long-term effects; role-play arousing situations; self-assessment of vulnerability.	Identify 3 personal emotion-arousing situations.
3	2 h	Deploy attention effectively.	Teach attention diversion; exercises to end rumination/worry; mindfulness practices.	Practice attention shifts in real-life stressors (e.g., 10-min daily).
4	2 h	Apply cognitive reappraisal.	Identify maladaptive appraisals; reframe negative thoughts; group discussions on impacts.	Reframe one daily negative appraisal and note outcomes.
5	2 h	Modify situations and responses.	Problem-solving skills; situation selection; teach expression, relaxation, catharsis.	Apply problem-solving to one trigger; practice relaxation.
6	2 h	Integrate and consolidate skills.	Review progress; role-play real-world application; address barriers; set maintenance plan.	Ongoing skill use; optional self-monitoring as a self-practice exercise

Following ethical approval from the Ethics Committee of IUMS (code number: IR.MUI.NUREMA.REC.1401.069), and the necessary institutional permissions, eligible patients provided written informed consent and completed baseline questionnaires. After randomization, the intervention group engaged in the training sessions while the control group continued with standard care. Recruitment for the study was carried out from January 2024 to May 2024. Each participant group underwent a three-week intervention, comprising six sessions in total. Immediate post-intervention assessments were conducted upon completion of the intervention, with the final group concluding in June 2024. Upon concluding the intervention, all participants completed the same set of questionnaires.

Data analysis

Data were analyzed using SPSS version 26. Descriptive statistics were reported as means and standard deviations for continuous variables and as frequencies and percentages for categorical variables. Baseline demographic and clinical characteristics were compared between the intervention and control groups using independent t-tests for continuous variables, chi-square tests for categorical variables, and the Mann–Whitney U test for ordinal variables.

For the primary outcome, death anxiety measured by the Templer Death Anxiety Scale (TDAS), a linear mixed-effects repeated-measures model was used as the main analysis. Time (baseline and post-intervention), group (intervention and control), and the group × time interaction were entered as fixed effects, and participant was included as a random effect to account for the within-subject correlation of repeated measurements. The group × time interaction was considered the primary test of intervention effectiveness.

ANCOVA with post-intervention TDAS score as the dependent variable, group as the fixed factor, and baseline TDAS score as the covariate was initially considered. Before selecting ANCOVA, the homogeneity of regression slopes assumption was examined by testing the group × baseline TDAS interaction. Because this assumption was not supported in the present dataset, conventional ANCOVA was not retained as the primary analysis. Therefore, a linear mixed-effects repeated-measures model, one of the recommended approaches for repeated-measure randomized controlled trial data, was used instead.

Within-group pre-post changes were additionally examined using paired t-tests, and between-group comparisons at each time point were examined using independent t-tests. Mean differences were reported with 95% confidence intervals. Cohen’s d was calculated for between-group comparisons using the pooled standard deviation. For within-group pre-post comparisons, Cohen’s dz was calculated as the mean change divided by the standard deviation of the change scores. Values of 0.2, 0.5, and 0.8 were interpreted as small, moderate, and large effects, respectively.

Normality of TDAS scores and change scores was assessed before inferential analysis. Because normality was not consistently supported across all TDAS distributions and because the TDAS is a bounded 0–15 score, non-parametric sensitivity analyses were also conducted using the Wilcoxon signed-rank test for within-group comparisons and the Mann–Whitney U test for between-group comparisons. All statistical tests were two-tailed, and statistical significance was set at α = 0.05. Analyses were conducted on a complete-case basis because no outcome data were missing.

Results

Of 79 patients assessed for eligibility, 19 were excluded before randomization because they did not meet the inclusion criteria (n = 18) or declined to participate (n = 1). The remaining 60 participants were randomized equally to the intervention group (n = 30) or the control group (n = 30). All randomized participants received their allocated condition and completed both baseline and post-intervention TDAS assessments. No participants withdrew or were lost to follow-up, and no adverse events were reported. In the intervention group, all 30 participants completed all six scheduled emotion regulation training sessions according to the session-completion criteria defined in the Methods section. Therefore, no participant was excluded from the primary outcome analysis, and the complete-case analysis included all 60 randomized participants. Participant flow is presented in Fig. 1 (CONSORT).

Baseline characteristics

Independent t-tests showed no significant differences between groups in age (Intervention: 41.13 ± 13.90 years; Control: 40.18 ± 13.10 years; p = 0.822), time since diagnosis (Intervention: 5.90 ± 1.21 years; Control: 5.70 ± 1.10 years; p = 0.752), or number of chemotherapy sessions (Intervention: 6.66 ± 1.87; Control: 6.36 ± 1.81; p = 0.590). Chi-square analyses revealed no significant differences in gender (p = 0.895), marital status (p = 0.910), employment status (p = 0.840), or leukemia type (p = 0.590). Education level did not differ significantly between groups (Mann–Whitney U: p = 0.730). Detailed baseline data are presented in Table 2.

Table 2.

Baseline characteristics of participants in the intervention and control groups

Variable	Intervention group (n = 30)	Control group (n = 30)	p-value
Age, mean ± SD (years)	41.13 ± 13.90	40.18 ± 13.10	0.822 *
Time since diagnosis, mean ± SD (years)	5.90 ± 1.21	5.70 ± 1.10	0.752 *
Number of chemotherapy sessions, mean ± SD	6.66 ± 1.87	6.36 ± 1.81	0.590 *
Gender, n (%)	Male: 18 (60.0%) Female: 12 (40.0%)	Male: 19 (63.3%) Female: 11 (36.7%)	0.895 **
Marital status, n (%)	Married: 9 (30%) Single: 21 (70%)	Married: 8 (26.7%) Single: 22 (73.3%)	0.910 **
Employment status, n (%)	Employed: 17 (56.7%) Unemployed: 13 (43.3%)	Employed: 19 (63.3%) Unemployed: 11 (36.7%)	0.840 **
Type of leukemia, n (%)	Acute [AML/ ALL]: 10 (33.3%) Chronic [CML/ CLL]: 20 (66.67%)	Acute [AML/ ALL]: 8 (26.67%) Chronic [CML/ CLL]: 22 (73.3%)	0.590 **
Education level, n (%)	Elementary 4 (13.3%) Middle school 6 (20.0%) Diploma 12 (40.0%) Academic 8 (26.7%)	Elementary 6 (20.0%) Middle school 7 (23.3%) Diploma 12 (40.0%) Academic 5 (16.67%)	0.730 ***

*Independent t-test, **Chi-square test, ***Mann-Whitney U test

Primary outcome – death anxiety

At baseline, mean death anxiety scores were comparable between the intervention and control groups (11.03 ± 4.59 vs. 10.87 ± 4.54, respectively; mean difference = 0.17, 95% CI: −2.19 to 2.53; p = 0.888; Cohen’s d = 0.04).

Following the intervention, the intervention group showed a significant reduction in death anxiety, from 11.03 ± 4.59 at baseline to 5.97 ± 3.64 after the intervention. The mean change was − 5.07 points (95% CI: −6.26 to − 3.88; p < 0.001), corresponding to a large within-group effect size (Cohen’s dz = − 1.59). In contrast, the control group showed no statistically significant change, with scores changing from 10.87 ± 4.54 at baseline to 11.07 ± 4.67 after the intervention. The mean change was + 0.20 points (95% CI: −0.01 to 0.41; p = 0.056; Cohen’s dz = 0.36).

The post-intervention between-group difference was statistically significant. The intervention group had lower post-intervention death anxiety scores than the control group (5.97 ± 3.64 vs. 11.07 ± 4.67; mean difference = − 5.10, 95% CI: −7.26 to − 2.94; p < 0.001), with a large between-group effect size (Cohen’s d = − 1.22).

ANCOVA was initially considered with post-intervention TDAS score as the dependent variable, group as the fixed factor, and baseline TDAS score as the covariate. However, the homogeneity of regression slopes assumption was not supported, as the group × baseline TDAS interaction was statistically significant (F(1,56) = 17.61, p < 0.001). Therefore, conventional ANCOVA was not retained as the primary analysis.

The primary repeated-measures analysis confirmed a significant group × time interaction. In the linear mixed-effects repeated-measures model, the intervention group showed a significantly greater reduction in death anxiety over time than the control group (β = −5.27, 95% CI: −6.40 to − 4.13; p < 0.001). The equivalent two-way mixed repeated-measures ANOVA also showed a significant group × time interaction (F(1,58) = 79.73, p < 0.001, partial η² = 0.58), indicating a large intervention effect.

As a sensitivity analysis, non-parametric tests were also conducted to examine whether the main findings were robust to the distributional assumptions of parametric tests. These analyses were not considered the primary analysis but were used to support the robustness of the results. The Wilcoxon signed-rank test showed a significant pre-post reduction in the intervention group (p < 0.001), whereas the change in the control group was not statistically significant (p = 0.058). Mann–Whitney U tests also showed no significant baseline difference between groups (p = 0.748), but significant between-group differences at post-intervention and in change scores (both p < 0.001). These results are summarized in Table 3.

Table 3.

Death anxiety scores and within- and between-group comparisons

Outcome / comparison	Intervention group (n = 30)	Control group (n = 30)	Effect estimate	95% CI	p-value / effect size
Baseline TDAS, Mean ± SD	11.03 ± 4.59	10.87 ± 4.54	MD = 0.17	-2.19 to 2.53	p = 0.888; Cohen’s d = 0.04
Post-intervention TDAS, Mean ± SD	5.97 ± 3.64	11.07 ± 4.67	MD = -5.10	-7.26 to -2.94	p < 0.001; Cohen’s d = − 1.22
Within-group change, Post - Baseline	-5.07	+ 0.20	Difference in change = -5.27	-6.45 to -4.09	p < 0.001; Cohen’s d = − 2.31
Within-group paired comparison	p < 0.001; Cohen’s dz = − 1.59	p = 0.056; Cohen’s dz = 0.36	—	—	—
Linear mixed-effects model, group × time interaction	—	—	β = -5.27	-6.40 to -4.13	p < 0.001
Repeated-measures ANOVA, group × time interaction	—	—	F(1,58) = 79.73	—	p < 0.001; Partial η² = 0.58

Negative values indicate lower scores in the intervention group or a greater reduction from baseline to post-intervention. Within-group p-values were obtained from paired t-tests. Between-group comparisons were obtained from independent t-tests. The group × time interaction was examined using a linear mixed-effects repeated-measures model and an equivalent two-way mixed repeated-measures ANOVA. ANCOVA was considered but not retained because the homogeneity of regression slopes assumption was not supported

TDAS Templer Death Anxiety Scale, MD mean difference

Discussion

The findings of the present study suggest that the emotion regulation training program based on Gross’s process model was associated with a significant short-term reduction in death anxiety among leukemia patients undergoing chemotherapy. The intervention group showed an approximately 5-point reduction in TDAS scores, whereas the control group showed no meaningful reduction over the same period. Because TDAS scores above 8 are generally interpreted as indicating higher death anxiety, the reduction from a baseline mean of 11.03 to a post-intervention mean of 5.97 suggests that the observed change may be practically relevant. However, the clinical significance of this change should be interpreted cautiously because no established minimal clinically important difference (MCID) for the TDAS in leukemia patients is currently available, and the present study did not directly assess daily functioning, quality of life, treatment engagement, or other patient-centered outcomes. Therefore, the findings should not be interpreted as proving clinical benefit in daily life, but rather as preliminary evidence of a statistically significant and potentially meaningful short-term reduction in death anxiety.

In addition, death anxiety was assessed only at baseline and immediately after completion of the intervention, and no follow-up assessment was conducted. Therefore, although the findings indicate a significant short-term reduction in TDAS scores following Gross’s emotion regulation training, the sustainability and durability of this treatment effect over time cannot be determined from the present trial. This limits the ability to conclude whether the observed improvement would persist after the end of the intervention. Future randomized controlled trials should incorporate longitudinal follow-up assessments, for example at 3, 6, and 12 months after the intervention, to evaluate the maintenance of treatment gains and the long-term effectiveness of emotion regulation training in leukemia patients undergoing chemotherapy.

Although the TDAS demonstrated good internal consistency in the current study sample (Cronbach’s α = 0.82), this finding should be interpreted alongside previous reports showing variable reliability estimates for the scale. Such variability may affect score precision and introduce some degree of measurement error in self-report assessment. Therefore, the observed intervention effect should be considered in light of both the acceptable reliability observed in the present sample and the broader psychometric variability reported for the TDAS.

Theoretically, death anxiety in leukemia patients undergoing chemotherapy may be maintained by repeated exposure to illness-related threats, attentional fixation on mortality cues, catastrophic appraisal of symptoms or prognosis, and maladaptive emotional responses. By targeting attentional shifting, cognitive reappraisal, situation modification, relaxation, and response modulation, the intervention may help patients reinterpret death-related thoughts more adaptively and reduce the intensity and persistence of anxiety responses.

Although the intervention was grounded in Gross’s process model and included components targeting attention shifting, cognitive restructuring, emotional awareness, relaxation, and response modulation, the present trial did not directly measure changes in specific emotion regulation strategies. Therefore, the observed reduction in death anxiety cannot be interpreted as evidence that improvements in cognitive reappraisal, attentional control, or other regulatory processes mediated the treatment effect. In addition, because secondary psychosocial outcomes such as depression, psychological distress, quality of life, and resilience were not assessed, the broader psychosocial impact of the intervention remains unclear. Future trials should include validated process measures of emotion regulation strategies and secondary psychosocial outcomes at multiple time points and should conduct formal mediation analyses to determine whether changes in specific emotion regulation mechanisms explain reductions in death anxiety.

These findings are broadly consistent with previous evidence suggesting that emotion regulation and related psychological interventions may reduce death anxiety and improve psychological adjustment in cancer and other vulnerable populations. Orang et al. reported that an emotion regulation-based intervention reduced death anxiety among elderly women with cancer [10], and Chukwuorji et al. showed that emotion regulation strategies were associated with lower death anxiety dimensions [24]. Baghdadi et al. similarly found that improving emotional skills reduced death anxiety in older adults [1]. Other studies have highlighted the role of emotion regulation in reducing psychological distress and enhancing resilience among patients with cancer [20, 22], while Ghazipoor et al. identified cognitive reappraisal as a potential mediator in the reduction of death anxiety [5]. Recent intervention studies and reviews further support the potential value of psychological and educational approaches for reducing death anxiety in chronic disease and advanced cancer populations [33, 34].

However, our findings contrast with Khamoushi et al. (2019), who reported no significant association between emotional self-regulation and death anxiety among 150 Iranian patients with cancer, and with Maqbool et al. (2025), who observed no significant relationship between death anxiety and resilience [21, 25]. These discrepancies may reflect several substantive differences. First, the present study focused specifically on leukemia patients undergoing chemotherapy, whereas previous studies included different or more heterogeneous clinical populations; variations in cancer type, treatment intensity, disease stage, and symptom burden may influence both death anxiety and the relevance of emotion regulation processes. Second, previous studies primarily examined naturally occurring or trait-like psychological constructs, such as emotional self-regulation or resilience, whereas the present trial evaluated a structured training program targeting modifiable emotion regulation skills. Third, the specific emotion regulation strategies emphasized or measured may differ across studies; for example, cognitive reappraisal, attentional control, suppression, resilience, and broader emotional self-regulation may not have equivalent relationships with death anxiety. Finally, cultural context, social support, spiritual coping, illness severity, and other unmeasured psychosocial factors may partly explain inconsistent findings. Therefore, the present results should be viewed as adding experimental evidence to an inconsistent literature rather than resolving the mechanism or generalizability of emotion regulation effects on death anxiety.

Conclusion

This randomized controlled trial suggests that Gross’s emotion regulation training, delivered virtually over six structured sessions, was associated with a significant short-term reduction in death anxiety among leukemia patients undergoing chemotherapy compared with standard care. The magnitude of TDAS reduction in the intervention group indicates that the intervention may have practical relevance; however, clinical significance should be interpreted cautiously because no established MCID for the TDAS in leukemia patients is available and broader patient-centered outcomes were not assessed. Therefore, these findings provide preliminary evidence supporting the potential usefulness of Gross model-based emotion regulation training as a psycho-oncology intervention. Further large-scale, multi-center randomized controlled trials with longer follow-up periods, blinded outcome assessment, secondary psychosocial outcomes, and formal mediation analyses are needed to confirm the effectiveness, durability, clinical relevance, and mechanisms of this approach across diverse patient populations.

Limitations and recommendations

This study has several limitations that should be considered when interpreting the findings. First, the sample size was relatively small (n = 60), which may restrict the statistical power of the findings and limit the precision of effect estimates. Second, because the trial was conducted in a single center using convenience sampling, the generalizability of the results to broader leukemia populations and other oncology care settings is limited. In addition, the sample was confined to adults within a specific cultural context, which may further limit the applicability of the findings to other populations and healthcare systems.

The 0% attrition rate and 100% intervention adherence should also be interpreted cautiously. Although these findings may partly reflect the short duration of the intervention, the virtual delivery format, availability of offline materials, and regular follow-up contacts, they may also indicate that participants who agreed to enroll were more motivated, clinically stable, or able to engage with smartphone-based training than the broader population of leukemia patients undergoing chemotherapy. Therefore, the retention and adherence rates observed in this trial may not be fully generalizable to routine oncology settings.

Third, although the trial used blinded data analysis, participant blinding and interventionist blinding were not feasible because of the behavioral and educational nature of the intervention and the use of a standard-care control condition. In addition, the same researcher who delivered the intervention also administered the outcome assessments; therefore, the outcome assessor was not blinded to group allocation. This may have introduced a potential risk of detection bias. However, the primary outcome was assessed using a standardized self-administered questionnaire with predefined scoring procedures, and the data analyst remained blinded to group allocation throughout the analysis.

Fourth, death anxiety was assessed at baseline and immediately after the intervention, and no follow-up assessment was conducted. Therefore, the durability and sustainability of the observed effects over time cannot be determined. Fifth, the study relied exclusively on a self-report instrument, the Templer Death Anxiety Scale (TDAS), to assess death anxiety. Although the TDAS showed good internal consistency in the current study sample (Cronbach’s α = 0.82), previous studies have reported variable reliability estimates for this scale. Therefore, the results may still be influenced by the inherent limitations of self-report measurement, including response bias, expectancy effects, social desirability, and common-method variance.

Sixth, although the intervention was based on Gross’s process model, the trial did not directly measure changes in specific emotion regulation strategies, such as cognitive reappraisal, attentional control, attentional shifting, or response modulation. Therefore, it cannot be determined whether improvements in these regulatory processes mediated the observed reduction in death anxiety. In addition, no secondary psychosocial outcomes, such as depression, psychological distress, quality of life, or resilience, were assessed. This limits interpretation of the broader psychosocial impact of the intervention beyond death anxiety.

Future research should aim to replicate these findings in larger, multi-center randomized controlled trials with more diverse samples. Future pragmatic trials should also anticipate possible attrition in routine clinical settings and consider appropriate retention strategies, oversampling, and intention-to-treat analyses. Future trials should also use separate blinded outcome assessors, longer follow-up periods, and complementary clinician-rated or behavioral measures in addition to self-report questionnaires. Moreover, future studies should assess secondary psychological outcomes and directly measure emotion regulation strategies in order to conduct formal mediation analyses and examine potential mediators and moderators of treatment effects. Where feasible, attention-control or sham educational control conditions may also help reduce performance and detection bias and strengthen causal interpretation.

Abbreviations

DA

Death anxiety

RCT

Randomized controlled trial

TDAS

Templer Death Anxiety Scale

WhatsApp

Messaging application used for virtual delivery

Gross model

Gross process model of emotion regulation

CONSORT

Consolidated Standards of Reporting Trials

STAI

State-Trait Anxiety Inventory

α 

Cronbach’s alpha reliability coefficient

SPSS

Statistical Package for the Social Sciences

Authors’ contributions

RR: Study Design and Conceptualization, Literature search, Preparation of the Research Proposal, Statistical Analysis and Results Interpretation, Supervision, Drafting the Initial Manuscript. AM: Study Design and Conceptualization, Literature search, Preparation of the Research Proposal, Preparation and development of an emotion regulation program, Project Implementation, Conducting the Research and Data Collection, Statistical Analysis and Results Interpretation, Drafting the Initial Manuscript, project administration, Correspondence. ZG: Preparation and development of an emotion regulation program, Preparing program content, leader of Project Implementation, Conducting the Research and Data Collection, Manuscript review & editing. TM: Study Design and Conceptualization, Literature search, Preparing program content, Data acquisition, Manuscript review & editing. MN: Study Design and Conceptualization, Literature search, preparing program content, Data acquisition, Manuscript review & editing. All authors have seen and approved the manuscript.

Funding

No external funding was received for this study. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. In addition to de-identified participant data, the statistical code used for analyses (SPSS syntax) and study materials (including intervention manuals and questionnaires) are also available upon reasonable request, subject to approval of the ethics committee.

Declarations

Ethics approval and consent to participate

This article is the result of a research project approved by the Ethics Committee of IUMS (code number: IR.MUI.NUREMA.REC.1401.069). Written informed consent was obtained from all participants, who were informed that they could withdraw from the study at any time. During the study, the utmost accuracy will be prioritized in the use of textbooks and literature, ensuring compliance with literary rights. Additionally, trustworthiness will be maintained by accurately referencing materials translated into Persian. The study adhered to the Declaration of Helsinki principles.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.