Validation of the Japanese version of the death and dying distress scale Authors: Seraki Miyamoto, Tadahiro Yamazaki, Toshio Matsubara, Hiroshi Yamagata, Saki Harashima, Hiroshi Kobo, Hiroaki Abe, Masahiko Sumitani, Ken Shimizu, Gary Rodin, Kazuhiro Yoshiuchi Published: 2026 ABSTRACT AbstractBackgroundThe Death and Dying Distress Scale (DADDS) is a self-report instrument used to assess distress related to dying and death in patients with advanced cancer. We aimed to test the reliability and validity of the Japanese-translated version of the DADDS (DADDS-J) and evaluate its psychological properties in patients with advanced cancer in Japan.MethodsJapanese participants with advanced cancer (n = 200) completed the DADDS and additional tools measuring depression (Patient Health Questionnaire-9), anxiety (Generalized Anxiety Disorder-7), quality of life (Functional Assessment of Cancer Therapy-General; Japanese Version of the Quality of Life at the End of Life-Cancer Scale), symptom burden (Japanese Version of the Quality of Life at the End of Life-Cancer Scale), and spirituality (Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being). Exploratory factor analysis (EFA), parallel analysis, and confirmatory factor analyses (CFA) were performed on 15 DADDS items. Internal consistency was assessed using Cronbach’s alpha and McDonald’s ω, and construct validity was examined by calculating correlations with relevant scales.ResultsEFA and parallel analysis identified a three-factor structure in the 14-item DADDS-J after item 3 was excluded, with Cronbach’s alpha and McDonald’s ω ranging from 0.85 to 0.92. CFA showed an adequate fit for the unique structure. Each subscale was significantly correlated with age, depression, anxiety, symptom burden, and lower quality of life and spirituality, but not with social and family well-being or time since cancer diagnosis.ConclusionsThe DADDS-J has a three-factor structure and demonstrates good reliability and validity. Differences between the original and Japanese versions of the scale may be due to cultural factors. This is the first tool that can measure spiritual pain associated with death in patients with advanced cancer in Japan and is suitable for large-scale studies and clinic use. This scale can enhance research and clinical care for patients with advanced cancer in Japan.Supplementary informationThe online version contains supplementary material available at https://doi.org/10.1186/s13030-026-00362-2.Keywords: Death and dying distress scale, Japan, Reliability, Validity, Advanced cancer Introduction Patients with advanced and terminal diseases experience “total pain” [1], which includes physical symptoms, financial burden, changes in interpersonal relationships, uncertainty about future treatments, anxiety, fear of recurrence, and death [2]. Death-related distress is common in patients with advanced cancer, with up to 80% reporting distressing thoughts about death [3]. Death-related distress in patients with advanced cancer is associated with pain, depression, anxiety, hopelessness, attachment insecurity, and hindrance to advance care planning and end-of-life preparation [4, 5]. However, death-related distress or death anxiety among patients with advanced cancer is typically not adequately assessed or addressed by medical staff [6, 7]. This may be due to the lack of established instruments and interventions to address these states. A 15-item Death and Dying Distress Scale (DADDS) was developed and validated by Canadian researchers to evaluate death and dying-related distress in patients with advanced cancer [8, 9], and it holds promise for enhancing the assessment and care of terminally ill patients. The DADDS comprises two subscales: finitude and dying. This measure has been translated into several languages, validated [10–13], and used to assess the effectiveness of psychotherapeutic interventions in patients with advanced cancer [14]. However, no validated instruments currently exist for measuring death and dying-related distress in Japanese individuals with advanced cancer. Although truth-telling has become more common in clinical oncology in Japan [5], compared to Western countries where knowing what to expect about one’s physical condition’ to achieve a good death is highly valued [15, 16], dying without knowing that one is dying is still often considered a virtue [15], and the importance of preparing for death remains inadequately recognized [17]. The topic of dying and death is taboo in Japan and is often avoided, even in the terminal stages of cancer [17], which makes it difficult to determine how spiritual care should be provided [5, 17]. The nature and extent of death-related distress in patients with advanced cancer in Japan could be clarified through the collection of data using self-report questionnaires. We report the results of the first study to validate a Japanese version of the DADDS (DADDS-J), evaluate its reliability and validity, and examine the psychological properties of death and dying-related distress in patients with advanced cancer in Japan. Based on the Japanese cultural values regarding dying without being aware of one’s own mortality and the tendency to avoid discussing death, we hypothesized that death-related distress in patients with advanced cancer in Japan may have a unique factor structure, distinct from that of questionnaires designed for Western populations. Materials and methods Procedure This multicenter cross-sectional study was approved by the Ethics Committee of the Graduate School of Medicine, University of Tokyo; the Ethics Committee of the Cancer Institute Hospital, Japanese Foundation for Cancer Research; and the Institutional Review Board of the Center for Clinical Research at Yamaguchi University Hospital. This study was conducted in accordance with the principles of the Declaration of Helsinki. Participants were fully informed of the study procedures, and written informed consent was obtained from all participants prior to their enrollment in the study. Given the vulnerability of the study population, the following considerations were made with attention to the potential for participant stress. If the content of the questionnaire or participation in this study caused stress or other psychological burden, participants were free to withdraw at any time without any disadvantage. If the content of the questionnaire or participation in this study caused stress or other psychological burden, participants were free to withdraw at any time without any disadvantage. If the research participants experienced distress due to the content of the questionnaire, they could stop responding and withdraw from the study at any time. Consent could be withdrawn at any point, even after it had initially been given without any disadvantage. If the research participants experienced distress due to the content of the questionnaire, they could stop responding and withdraw from the study at any time. Consent could be withdrawn at any point, even after it had initially been given without any disadvantage. The research participants were encouraged to discuss their participation in the study with family members. The research participants were encouraged to discuss their participation in the study with family members. To avoid burdening the research participants, researchers could visit patients on their treatment or outpatient days and participants could have completed the questionnaires, or participants could send completed questionnaires by mail. To avoid burdening the research participants, researchers could visit patients on their treatment or outpatient days and participants could have completed the questionnaires, or participants could send completed questionnaires by mail. If the research participants’ mental health was adversely affected, psychological support was provided by a specialist in psychosomatic medicine or psychiatry through outpatient care and appropriate mental health support. If the research participants’ mental health was adversely affected, psychological support was provided by a specialist in psychosomatic medicine or psychiatry through outpatient care and appropriate mental health support. We followed the World Health Organization translation methodology [18] and collaborated with the original DADDS developers at the Princess Margaret Cancer Centre [5, 9]. The DADDS was translated into Japanese and refined through further discussion and consensus among the authors, who are all native Japanese speakers with expertise in psycho-oncology. This version was back-translated into English by a professional bilingual editor proficient in both English and Japanese to ensure conceptual equivalence between the two versions. The accuracy of the back-translation was largely confirmed by the original DADDS developers, thus only minor corrections were needed; specifically, they were limited to changing a Japanese particle in item 4, “Not knowing what happens near the end of life.” Participants Participants were recruited between April 2020 and December 2022 from the outpatient psycho-oncology clinic in the Department of Psychosomatic Medicine, the outpatient clinic and inpatient consultations in the Department of Pain and Palliative Medicine at The University of Tokyo Hospital, the outpatient psycho-oncology clinic at the Cancer Institute Hospital of the Japanese Foundation for Cancer Research, and the outpatient clinic at the Palliative Care Center of Yamaguchi University Hospital, Japan. The inclusion criteria were: 1) age ≥ 18 years; 2) fluency in Japanese; 3) ability to provide written informed consent; and 4) a diagnosis of an advanced solid tumor with or without distant metastases, including recurrent stage III or IV lung or ovarian cancer, pancreatic cancer of any stage, stage IV gastrointestinal tract cancer, stage IV gynecologic cancer, stage IV breast cancer, stage IV genitourinary cancer, stage IV sarcoma, stage IV melanoma, stage IV endocrine cancer, unresectable cholangiocarcinoma, unresectable liver cancer, and unresectable ampullary or periampullary cancer [19]. The exclusion criteria were: 1) difficulty communicating with healthcare providers, 2) cognitive impairment, and 3) severe physical or speech impairment, as identified by healthcare providers or noted in medical records. Participants were asked to complete the DADDS-J, the Patient Health Questionnaire-9 (PHQ-9), the Generalized Anxiety Disorder-7 (GAD-7), the Functional Assessment of Cancer Therapy-General (FACT-G, version 4), the Japanese version of the QUAL-EC (QUAL-EC-J), and the Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being (FACIT-Sp, 4th edition). All these questionnaires, except the DADDS, have been evaluated for reliability and validity in Japanese. Measures Using questionnaires and patient medical charts, we obtained data on age, sex, living condition, marital status, education level, primary tumor site, date of diagnosis, and time elapsed since cancer diagnosis. The DADDS The original DADDS is a self-administered questionnaire consisting of 15 items to assess death-related distress, validated in English [5, 9], German [10], Chinese [11], Spanish [12], and Italian [13]. Each item is rated on a six-point Likert scale: 0 = I did not experience this thought or concern, 1 = I experienced very little distress, 2 = I experienced mild distress, 3 = I experienced moderate distress, 4 = I experienced great distress, and 5 = I experienced extreme distress. The original DADDS obtained data on 386 patients with advanced cancer, and it has a two-factor structure: Factor 1, “Finitude” (items 1–10) and Factor 2, “Dying” (items 11–15) [9]. Both factors showed good internal consistency (Cronbach α = 0.861–0.949). Construct validity was demonstrated by a positive correlation between DADDS and demoralization, and a negative correlation with preparation for the end of life [9]. The Finitude factor primarily refers to distress about the shortness of time. It includes items rating distress about not having done all the things I wanted to do, saying all that I wanted to say to the people I care about, not having achieved my goals and ambitions and running out of time. The Dying factor refers to distress about the process of dying and includes items rating distress about the end of life happening suddenly or unexpectedly, being prolonged or drawn out, happening when I am alone, with a lot of pain or suffering, or happening very soon [9]. The PHQ-9 The PHQ-9 is a nine-item self-report questionnaire developed by Kroenke et al. [20] to assess depressive symptoms and is widely used in patients with advanced cancer [21, 22]. Each item is rated on a four-point Likert scale ranging from 0 to 3, with a total score ranging from 0 to 27; higher scores indicate stronger depressive symptoms [23, 24]. The Japanese version of the PHQ-9 has been validated and reported to have good internal consistency (Cronbach’s α = 0.91) [25]. The optimal cutoff score for diagnosing major depressive disorder in Japan is reported to be 10 or higher [24]. The PHQ-9 has also been used to validate the English, Chinese, and Italian versions of the DADDS [9, 11, 13]. The GAD-7 The GAD-7 is a seven-item self-report questionnaire developed by Spitzer et al. [26] to assess generalized anxiety disorder. Each item is rated on a four-point Likert scale ranging from 0 to 3, with total scores ranging from 0 to 21; higher scores indicate stronger symptoms of anxiety. The Japanese version of the GAD-7 has been validated and reported to have good internal consistency (Cronbach’s α = 0.92) [27]. The GAD-7 has also been used to validate the English, Chinese, and Italian versions of the DADDS [9, 11, 13]. The QUAL-EC-J The QUAL-EC is a 17-item self-report questionnaire developed by Lo et al. [28] to assess the quality of life of patients nearing the end of life. Each item is rated on a five-point Likert scale ranging from 1 to 5, with total scores ranging from 17 to 85; higher scores indicate a better quality of life. The QUAL-EC has also been used to validate the English and Chinese versions of the DADDS [9, 11]. The QUAL-EC-J has a unique three-factor structure: “Symptom Control,” “Acceptance of Disease and Life,” and “Preparation for End of Life”. The validated QUAL-EC-J has good internal consistency (Cronbach’s α = 0.88, 0.81, and 0.68) [29]. The FACT-G The FACT-G is a 27-item self-report questionnaire developed by Cella et al. [30] to assess the quality of life of patients undergoing cancer treatment. Each item is rated on a five-point Likert scale ranging from 0 to 4, with total scores ranging from 0 to 108; higher scores indicate better well-being. It contains four subscales: “Physical Well-Being,” “Social and Family Well-Being,” “Psychological Well-Being,” and “Functional Well-Being.” The reliability and validity of the Japanese version have been confirmed to be adequate [31], with Cronbach’s α reported as 0.81, 0.55, 0.74, and 0.79 for the respective subscales [31]. The FACT-G has also been used to validate the German versions of the DADDS [10]. The FACIT-Sp The FACIT-Sp is a self-report questionnaire developed by Peterman et al. [32] to assess the quality of life and spirituality of patients with cancer and other chronic illnesses. In this study, only 12 items related to spiritual well-being from FACIT-Sp fourth edition were used. The 12 items include two subscales, “Meaning/Peace” and “Faith.” Each item is rated on a five-point Likert scale ranging from 0 to 4, with higher scores indicating a stronger element for each subscale. The Japanese version of the FACIT-Sp has been validated and reported to have good internal consistency (Cronbach’s α = 0.89 and 0.82, respectively) [33]. The FACIT-Sp has also been used to validate the English, Chinese, and Italian versions of the DADDS [9, 11, 13]. Statistical analysis Statistical significance was set at p < 0.05 for a two-tailed test. Most data analyses were conducted using IBM SPSS Statistics version 29.0 (IBM Corp., Armonk, NY). The parallel analysis was performed using HAD (official name) [34]—a free psychological statistics analysis program created by Shimizu [35] for use with Microsoft Excel, which is widely used in Japan—and confirmatory factor analysis (CFA) was performed using Analysis of Moment Structures (AMOS) (IBM Corp., Armonk, NY). The following measurement properties (recommended in the COSMIN Risk of Bias tool [36]) were assessed: structural validity (exploratory factor analysis [EFA]), internal consistency (Cronbach’s α), and construct validity (by testing hypotheses on expected correlations with the PHQ-9, GAD-7, QUAL-EC-J, FACT-G, and FACIT-Sp). The sample size was set at seven times the number of items and at least 100, in accordance with the COSMIN recommendations [36]. Structural validity The participants were split into two subsamples to avoid using the same sample for EFA and CFA. The two subsamples were separated randomly using the “RAND” function in Microsoft Excel. One subsample was named the EFA subsample (for EFA and parallel analysis), while the other was named the CFA subsample (for CFA). First, the underlying factor structure of the 15-item DADDS-J was examined using EFA for the EFA subsample. The Kaiser–Meyer–Olkin (KMO) measure was used to assess sampling adequacy [37, 38]. The KMO test value ranges between 0 and 1, with a higher value reflecting a more suitable analysis. This value should be equal to or higher than 0.60 to continue with factor analysis [39]. More specifically, values of 0.90 and above are deemed excellent; 0.80 s, good; 0.70 s, middling; 0.60 s, mediocre; 0.50 s, poor but acceptable; and below 0.50, unacceptable [40]. The Bartlett Test for Sphericity (p < 0.05) was applied to confirm that the items have patterned relationships [41]. Diagonal squared multiple correlations (SMC) were used as prior communality estimates, and parallel analysis based on randomly simulated polychoric correlations was conducted on the EFA subsample to determine the number of factors [42]. The Diagonal SMC parallel analysis tends to overestimate but rarely underestimates the number of factors, making it useful as a criterion for determining the maximum number of factors [43]. In the present study, the original dataset was randomly permuted 50 times to generate a set of random eigenvalues. Subsequently, the 95th percentiles of these random eigenvalues were compared with the eigenvalues from the actual data. The eigenvalues of the EFA should be higher than those obtained from the parallel analysis. The number of factors was determined using parallel analysis, a scree plot, and cumulative contribution rate before rotation. Prior to the main EFA, the data matrix was evaluated. Individual variables in the EFA subsample were assessed using the Kolmogorov–Smirnov one-sample test (p < 0.001) and graphical methods (histograms and box plots), which together indicated that the data were not normally distributed. Therefore, EFA was performed using generalized least squares (to account for non-normality) and promax rotation, with a factor loading ≥ 0.50 as the cutoff [40, 44]. Factor retention was based on the Kaiser–Guttman criterion, which recommends retaining factors with eigenvalues > 1.0 [45]. Promax rotation is a computationally faster type of direct oblimin rotation and allows factors to be correlated [46–48]. We chose oblique rotation to represent relationships between factors within the DADDS-J [48]. Although the orthogonal rotation is more appropriate when analyzing completely independent factors, we assumed that the correlations between the factors in the DADDS-J were not negligible [11, 13], and a previous study has revealed good agreement between rotational strategies in the DADDS [13]. Next, CFA was performed to assess the model fit for the CFA subsample. The fit was evaluated using several indices: the confirmatory fit index (CFI) and Tucker–Lewis index (TLI) as incremental indices, the root mean square error of approximation (RMSEA) as a non-centrality parameter, and the standardized root mean square residual (SRMR) as a residual index. Factor models with a CFI ≥ 0.90, TLI ≥ 0.90, RMSEA ≤ 0.06, and SRMR ≤ 0.08 were considered to have good model fit [48]. For RMSEA, values ≤ 0.06 indicate adequacy of the model [49–52]; however, commonly used cutoff values are ≤0.08 indicating an acceptable fit; ≤0.10, a mediocre fit; and >0.10, a poor (unacceptable) fit [50, 53–56]. SRMR values up to 0.05 indicate a close fit, while values between 0.05 and 0.10 suggest an acceptable fit [57, 58]. Model re-specification was conducted by adding residual covariances between items within the same factor based on modification index values. Internal consistency The reliability of the newly developed subscales of the DADDS-J was measured using Cronbach’s α and McDonald’s ω, with values ≥ 0.70 generally considered to indicate acceptable internal consistency, value ≥ 0.80 are interpreted as good reliability, and ≥0.90 as excellent reliability [59, 60]. Construct validity and psychological properties The Kolmogorov-Smirnov one-sample test (p < 0.001) and graphical methods (histograms and box plots) demonstrated that the sample of this study followed a nonparametric distribution. There are no other questionnaires with constructs identical to those measured in the DADDS, and there is no gold standard [36]. Therefore, construct validity was assessed using Spearman’s rank-order correlations for subscale scores of the DADDS-J, total scores of the PHQ-9 and GAD-7, and subscale scores of the QUAL-EC-J, FACT-G, and FACIT-Sp [61]. Spearman’s rank-order correlations were interpreted based on Cohen’s criteria: 0.1 ≤ r < 0.3 as a weak correlation, 0.3 ≤ r < 0.5 as a moderate correlation, and r ≥ 0.5 as a strong correlation [62]. Correlations between the DADDS-J and patient age and the time since cancer diagnosis were examined to assess psychological properties. Results Two hundred patients who met the inclusion criteria completed the questionnaire (mean age: 63.5 years, standard deviation: 12.5). Gastrointestinal, pulmonary, and oropharyngeal tumors were the primary diagnoses for 79.5% of the participants. Most participants were married (69.0%), lived with their families (80.0%), and had a high school education or higher (94.0%) (Table 1). The cancer stage was verified through patients’ medical records. Missing responses were checked with the patients, ensuring no missing data. Table 1Socio-demographic and clinical characteristics (N = 200)Variablen (%)Mean (SD)Min-MaxSex Male103 (51.5) Female97 (48.5)Age63.5 (12.5)27–89Marital status Married138 (69.0) Unmarried31 (15.5) Divorced17 (8.5) Bereaved14 (7.0)Living conditions Living alone40 (20.0) Living with family160 (80.0)Education Junior high school6 (3.0) High school67 (33.5) Vocational school21 (10.5) College/University100 (50.0) Uncertain6 (3.0)Time since cancer diagnosis (months), SD31.7, 46.6Primary tumor Gastrointestinal68 (34.0) Lung64 (32.0) Oral and oropharyngeal27 (13.5) Breast12 (6.0) Gynecologic14 (7.0) Genitourinary10 (5.0) Sarcoma3 (1.5) Melanoma1 (0.5) Endocrine1 (0.5)DADDS-J20030.91 (16.79)0–69PHQ-92008.5 (5.34)0–23GAD-72006.1 (5.23)0–21QUAL-EC-J Symptom control2008.7 (3.24)3–15 Acceptance of disease and life20034.0 (6.96)18–53 Preparation for end-of-life20012.5 (3.66)4–20FACT-G Physical well-being20016.0 (6.37)2–28 Social and family well-being20015.6 (5.70)0–28 Emotional well-being20014.5 (5.27)0–24 Functional well-being20014.4 (5.94)0–28FACIT-Sp Meaning/peace20016.7 (6.98)2–32SD, standard deviation; DADDS-J, Japanese Version of the Death and Dying Distress Scale; PHQ-9, Patient Health Questionnaire-9; GAD-7, Generalized Anxiety Disorder-7; QUAL-EC-J, Japanese Version of the Quality of Life at the End of Life-Cancer Scale; FACT-G, Functional Assessment of Cancer Therapy-General Scale; FACIT-Sp, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Socio-demographic and clinical characteristics (N = 200) SD, standard deviation; DADDS-J, Japanese Version of the Death and Dying Distress Scale; PHQ-9, Patient Health Questionnaire-9; GAD-7, Generalized Anxiety Disorder-7; QUAL-EC-J, Japanese Version of the Quality of Life at the End of Life-Cancer Scale; FACT-G, Functional Assessment of Cancer Therapy-General Scale; FACIT-Sp, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Structural validity First EFA The KMO test confirmed excellent sampling adequacy (KMO = 0.91). Bartlett’s test for sphericity yielded statistically significant results (p < 0.001). Three factors, representing 73.17% of the total variance, were extracted based on eigenvalues ≥ 1 and the scree plot (Additional file 1). The 15 items of the DADDS-J were allocated to three factors based on the factor loadings. Item 3, which exhibited cross-loadings (Additional file 2), had a principal factor loading of 0.605 (≥0.5) and a secondary factor loading of 0.419 (≥0.3). Since the difference between the principal and secondary factor loadings was 0.186 < 0.20, it was excluded [63]. Second EFA After item 3 was excluded based on the first EFA, a second EFA was conducted on the remaining 14 items of the DADDS-J. The KMO test indicated good sampling adequacy (KMO = 0.88). Bartlett’s test for sphericity yielded significant results (p < 0.001), suggesting that the relationship between the DADDS-J items was sufficiently strong to justify conducting an EFA. In this model, three factors—accounting for 73.39% of the total variance—were extracted based on eigenvalues ≥ 1 and the scree plot (Fig. 1). Figure 2 presents the results of the parallel analysis, in which the simulated data intersected between the three-factor and four-factor models, supporting the appropriateness of a maximum of three factors for the EFA. The remaining 14 items of the DADDS-J were divided into three factors using generalized least squares and promax rotation with no items excluded (Table 2). The factor loading for item 5 was 1.006 > 1. However, this factor loading corresponds to a partial regression coefficient rather than a correlation coefficient because the EFA uses an oblique rotation. Therefore, factor loadings by oblique rotations could be more than 1. In addition, the commonality of item 5 is 0.896 < 1 and the unique variance is positive, which means that this does not fit the Heywood case [64]. The three-factor model was determined to be the most suitable for the EFA. “Factor 1,” consisting of items 4–10, represented distress related to concerns about the shortness of time (“Finitude”); “Factor 2,” consisting of items 11–15, represented distress related to dying (“Dying”); and “Factor 3,” consisting of items 1 and 2, represented distress related to incompleteness (“Incompleteness”). Scree plot of exploratory factor analysis on the 14-item DADDS-J. The scree plot of the exploratory factor analysis conducted on the DADDS-J shows the eigenvalues of the 14 components of the DADDS-J. DADDS-J: Japanese version of the death and dying distress scale Scree plot of the parallel analysis on the 14-item DADDS-J using the diagonal squared multiple correlation. This figure presents the scree plot comparing the observed eigenvalues with those generated from 50 simulated samples. The plot shows that there were three factors that exceeded their corresponding random eigenvalues, indicating that a maximum of three factors is recommended based on the parallel analysis Table 2Factor loadings and subscales of the 14-item DADDS-JDADDS itemFactor 1FinitudeFactor 2DyingFactor 3Incompleteness5. Not having a future (F)1.006−0.137−0.0167. Running out of time (F)0.876−0.0670.0344. Not knowing what happens near the end of life (F)0.7870.058−0.02410. My own death and dying (F)0.6960.243−0.0866. The missed opportunities in my life (F)0.6610.0110.2449. The impact of my death on my loved ones (F)0.5520.1210.0218. Being a burden to others (F)0.5420.2040.05912. Be prolonged or drawn out (D)−0.2030.8780.30613. Happen when I am alone (D)0.0280.834−0.00915. Happen very soon (D)0.0200.820−0.19811. Happen suddenly or unexpectedly (D)0.0540.817−0.03514. Happen with a lot of pain or suffering (D)0.2220.697−0.0352. Not having said all that I wanted to say to the people I care about (F)−0.015−0.0420.9311. Not having done all the things that I wanted to do (F)0.120−0.0410.775Cronbach’s α0.9170.9080.851McDonald’s ω0.9170.908—We conducted exploratory factor analysis using generalized least squares and promax rotation. This three-factor solution was considered appropriate after examining the magnitude and rate of change in eigenvalues. Original subscale allocations are shown in brackets. Bold values in this table indicate inclusion in the “Factor 1” and “Factor 2,” and “Factor 3” columns, respectivelyDADDS, Death and Dying Distress Scale; F, Finitude D, dying Factor loadings and subscales of the 14-item DADDS-J We conducted exploratory factor analysis using generalized least squares and promax rotation. This three-factor solution was considered appropriate after examining the magnitude and rate of change in eigenvalues. Original subscale allocations are shown in brackets. Bold values in this table indicate inclusion in the “Factor 1” and “Factor 2,” and “Factor 3” columns, respectively DADDS, Death and Dying Distress Scale; F, Finitude D, dying CFA To determine whether the EFA-supported three-factor model of the 14-item DADDS-J was valid, we conducted CFA on an independent sample of 100 participants (CFA subsample) (Fig. 3). The analysis of the three-factor structure revealed an acceptable fit (CFI = 0.959; TLI = 0.943; RMSEA = 0.090; SRMR = 0.038). Model respecification was conducted based on modification indices, and three residual covariances within factors were added: between items 5 and 14 (e = 0.27), between items 4 and 9 (e = −0.24), and between items 9 and 13 (e = −0.41). Confirmatory factor analysis diagram of 14-item DADDS-J. The figure shows the confirmatory factor analysis of the 14-item, 3-factor structure of the DADDS-J model Internal consistency The Cronbach’s α values were 0.92 for “Finitude,” 0.91 for “Dying,” and 0.85 for “Incompleteness” (Table 3). The McDonald’s ω values were 0.92 for “Finitude” and 0.91 for “Dying” (Table 3); a value for “Incompleteness” could not be estimated because it had fewer than three items. Table 3Spearman’s rank-order correlations between the 14-item DADDS-J subscales and related measures, age, and time after cancer diagnosisDADDS-JFinitudeDyingIncompletenessPHQ-90.58**0.46**0.38**GAD-70.61**0.52**0.44**QUAL-EC-JSymptom control−0.31**−0.34**−0.33**Acceptance of disease and life−0.25**−0.20**−0.18*Preparation for end of life−0.52**−0.34**−0.37**FACT-GPhysical well-being−0.40**−0.39**−0.37**Social and family well-being−0.07−0.05−0.10Emotional well-being−0.68**−0.53**−0.42**Functional well-being−0.39**−0.47**−0.38**FACIT-SpMeaning/peace−0.48**−0.38**−0.38**Faith−0.30**−0.17*−0.19**Age−0.21**−0.05−0.21**Elapsed time since cancer diagnosis−0.11−0.13−0.14* <0.05, ** <0.001DADDS-J, Japanese version of the Death and Dying Distress Scale; PHQ-9, Patient Health Questionnaire-9; GAD-7, Generalized Anxiety Disorder-7; QUAL-EC-J, Japanese version of the Quality of Life at the End of Life-Cancer Scale; FACT-G, Functional Assessment of Cancer Therapy-General Scale; FACIT-Sp, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Spearman’s rank-order correlations between the 14-item DADDS-J subscales and related measures, age, and time after cancer diagnosis * <0.05, ** <0.001 DADDS-J, Japanese version of the Death and Dying Distress Scale; PHQ-9, Patient Health Questionnaire-9; GAD-7, Generalized Anxiety Disorder-7; QUAL-EC-J, Japanese version of the Quality of Life at the End of Life-Cancer Scale; FACT-G, Functional Assessment of Cancer Therapy-General Scale; FACIT-Sp, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Construct validity and psychological properties Table 3 presents the correlations between each DADDS-J subscale and other measures, and the age and elapsed time since cancer diagnosis. “Finitude” had strong positive correlations with the PHQ-9 and GAD-7 and strong negative correlations with the “Preparation for End of Life” subscale of the QUAL-EC-J and “Emotional well-being” subscale of the FACT-G. “Finitude” showed negative moderate correlations with the “Symptom Control” subscale of the QUAL-EC-J, the “Physical Well-Being” and “Functional Well-Being” subscales of the FACT-G, and the “Meaning/Pace” and “Faith” subscales of the FACIT-Sp. “Finitude” had weak negative correlations with the “Acceptance of Disease and Life” subscales of the QUAL-EC-J. “Dying” had a strong positive correlation with the GAD-7 and a strong negative correlation with the “Emotional well-being” subscale of the FACT-G. Additionally, “Dying” showed a positive moderate correlation with the PHQ-9 and negative moderate correlations with the “Symptom Control” and “Preparation for End of Life” subscales of the QUAL-EC-J, the “Physical Well-Being” and “Functional Well-Being” subscales of the FACT-G, and the “Meaning/Pace” subscale of the FACIT-Sp. “Dying” also had weak negative correlations with the “Acceptance of Disease and Life” subscale of the QUAL-EC-J and “Faith” subscale of the FACIT-Sp. “Incompleteness” had moderate positive correlations with the PHQ-9 and GAD-7 and moderate negative correlations with the “Symptom Control” and “Preparation for End of Life” subscales of the QUAL-EC-J; “Physical Well-Being,” “Emotional Well-Being,” and “Functional Well-Being” subscales of the FACT-G; and “Meaning/Pace” subscale of the FACIT-Sp. “Incompleteness” showed weak negative correlations with the “Acceptance of Disease and Life” subscale of the QUAL-EC-J and the “Faith” subscale of the FACIT-Sp. Scores of both “Finitude” and “Incompleteness” showed a weak negative correlation with ages. Discussion We report findings from a validation study of the DADDS-J, a translated version of the DADDS into Japanese. We demonstrated that the DADDS-J has a factor structure distinct from that of the original version and adequate reliability and validity for use in patients with advanced cancer in Japan. The DADDS-J has a three-factor structure comprising “Finitude,” “Dying,” and “Incompleteness.” For the DADDS-J, the “Finitude” factor of the original DADDS was divided into two factors, and a new factor, “Incompleteness,” was extracted. Although there are some differences in the number of items that make up “Finitude” and “Dying,” “Finitude” and “Dying” are latent factors of the DADDS, which has been adopted in many countries, including the English [9], Chinese [11], Spanish [12], Italian [13], and Japanese versions. The German version of the DADDS has a unidimensional structure, and was evaluated using only the items that constitute “Finitude” in the English version [10]. In each version of the DADDS, “Finitude” and “Dying” are generally associated with depression [9, 10, 13], anxiety [9, 10, 13], fear of progression [10], demoralization [9, 13], worsening physical symptoms [9, 13], quality of life [10, 13], spiritual well-being [9, 13], and attachment insecurity [9], which are related to distress associated with death. Accordingly, it is thought that “Finitude” and “Dying” are elements that constitute the distress related to death in patients with advanced cancer, and is relevant to both Eastern and Western countries. “Incompleteness” comprises only two items, and the contents of these items are similar to that of the concept of “Regret” extracted from the Italian version of the DADDS. “Regret” is a subscale composed of four items (item 3, 2, 6, and 1 in the original DADDS), two of which are the same as the items that constitute “Incompleteness” in the DADDS-J. However, one of the remaining two items of “Regret”—“The missed opportunities in my life” (item 6)—was classified as “Finitude” in the DADDS-J, and “Not having achieved my life goals and ambitions” (item 3) was excluded because it showed cross-loadings in the DADDS-J. Differences in the factor structure of each version of the DADDS [11, 13] suggest that the characteristics of death-related distress among patients with cancer vary across cultures [29]. In a qualitative study conducted by Kanda et al. [65] among 30 Japanese individuals aged 20–80 years, eight important categories related to desirable death were extracted, including “Preparation for death” and “Satisfaction with life.” “Preparation for death” included “Leave thoughts in a will and testament,” while “Satisfaction with life” included “To feel that one has lived until fulfilling one’s life purpose” and “Having no regrets.” If these are not achieved, the experience may deviate from the desired death, and the pain associated with death may increase. This concept is common with “Incompleteness,” which is one of the elements that constitute the DADDS-J, which measures pain associated with death. For Japanese people, “Incompleteness” represents a deviation from the desired death and may serve as an element of pain associated with death. In fact, the distress due to uncompleted life tasks is reportedly one of the major existential concerns among terminally ill Japanese patients with cancer [66]. In contrast to the findings of Steinhauser et al. regarding the view of a “good death” in the United States [16], control over the future domain appeared in the relatively less important category in the Japanese survey [15]. Miyashita et al. [15] and Kanda et al. [65] identified unawareness of death as one of the major contributors to a good death in Japan, and unawareness of death was significantly positively correlated with the traditional Japanese style of death (dying as one sleeps, sudden death, and leaving the decision to a medical expert) [15]. In a longitudinal study in which Arai et al. first evaluated prognostic understanding in patients with advanced lung cancer in Japan from the time of diagnosis, approximately 20% of the patients had an inaccurate understanding of their prognosis. The authors also found that a consistently accurate prognostic understanding was associated with higher psychological distress [67]. In the DADDS-J, two items derived from the “Finitude” factor in the original version have been extracted as “Incompleteness.” For Japanese patients with advanced cancer, the distress associated with death may tend to focus more on past “Incompleteness” than on future-oriented “Finitude.” The tendency among Japanese patients to psychologically avoid confronting the finitude of time may have dispersed the emotional burden associated with the “Finitude” factor. Instead, it can be interpreted that past-oriented “Incompleteness”—which includes unfulfilled tasks and unsaid words—emerged as a more concrete and recognizable form of distress related to death. The constituent items for “Dying” in the DADDS-J are the same as those in the original DADDS, with the highest factor loading on Item 12, “Be prolonged or drawn out.” Although EFAs were also conducted for the Italian and Chinese versions, this item was prioritized most highly only in the DADDS-J. Among the ideal traditional Japanese styles of death, “dying as one sleeps” and “sudden death” [15] may have influenced the differences in the loading structure of the “Dying” factor. If living as usual without the feeling of facing impending death is considered a good death in Japan, supporting this may help alleviate the distress associated with death in patients with advanced cancer in this country. However, it is generally known that the psychological distress of patients with advanced cancer worsens over time [68, 69], and it may be difficult for patients to be completely unaware of their mortality. Higher levels of death-related distress in this study were associated with higher levels of depression and anxiety and a poorer quality of life, “emotional well-being,” “preparation for end-of-life,” and “spiritual well-being,” which confirmed the findings of previous research [9–11, 70, 71]. Similar to the findings of previous studies [5, 10], none of the subscales of the DADDS-J were significantly correlated with “social and family well-being” on the FACT-G, whereas a study of the Chinese version of the DADDS revealed that a better relationship with healthcare providers was an independent protective factor against death-related distress [11]. A moderate correlation has been found between the original DADDS and attachment insecurity [9], and in a recent test of terror management theory in advanced cancer [72], attachment security, self-esteem, and sense of meaning were shown to have a combined effect in protecting patients from death-related anxiety. Further studies are needed to evaluate the contribution of attachment security and each type of social support to death-related distress in patients with cancer. In the DADDS-J, scores of both “Finitude” and “Incompleteness” showed a weak negative correlation with ages. In this study, the elapsed time since cancer diagnosis was not correlated with the DADDS-J, while, in previous studies, death anxiety was negatively correlated with age [73] and time since diagnosis [74]. The diagnosis of cancer during adolescence and young adulthood causes higher levels of psychological distress [75]. Older patients with cancer have reported greater end-of-life preparation [76], and being aged ≥ 65 years was identified as a protective factor for quality of life [77]. Therefore, it is necessary to conduct further research using longitudinal data. The DADDS-J was correlated with existing scales that are considered to be related to the distress of death. The correlation coefficient for “Finitude” was generally higher than that for “Dying” or “Incompleteness,” although additional research is necessary to determine the difference in clinical significance between these three subscales. Although the DADDS-J reportedly has stronger correlations with the PHQ-9 (r = 0.60–0.63) and GAD-7 (r = 0.42–0.47), the discriminant validity of the DADDS-J in relation to hopelessness or demoralization, which are very likely to be related to the DADDS-J [2, 78, 79], has not been verified. In future research, the Beck Hopelessness Scale [80, 81], Demoralization Scale [82], and DADDS-J should be compared and conceptual differences should be determined. Additionally, the relationship with family functioning (Family Relationship Index) [83, 84] and ikigai (Ikigai-9 scale) [85] should be clarified, and the spirituality of the Japanese population investigated in more detail. On that basis, we plan to conduct in-depth interviews with patients to identify concerns specific to the culture that are not covered by the DADDS-J as well as to explore the cultural adaptation of this scale. It is very likely that the DADDS-J can be used to measure death-related distress in clinical settings. To date, this type of distress has often been overlooked in clinical practice in Japan, even at the end of life. This study reveals that death-related distress is closely related to depression, anxiety, quality of life, and spiritual well-being in Japan. Measuring and addressing this type of distress in the context of care from the time of diagnosis of advanced cancer to the end of life may be a groundbreaking approach to the practice of psycho-oncology in Japan. In fact, the DADDS is used in Managing Cancer and Living Meaningfully (CALM) to regularly assess distress related to death and dying [9, 14], and it has been proven useful in dignity therapy as a measure for assessing pain related to death [86]. CALM is a supportive–expressive therapy that relieves death anxiety in individuals with moderately severe levels of such distress at baseline [14]. CALM strategies include adjusting the patient’s relationship with healthcare professionals to help manage their symptoms; involving family members in sessions and creating a space where patients and their families can share their thoughts and feelings; addressing attachment insecurity triggered by cancer; and focusing on life’s meaning and personal beliefs. In CALM therapy, the therapist does not confront or inform the patient of only objective facts. Rather, interventions focus on the patient’s subjective emotional experience [87]. These psychological interventions have the potential to alleviate the distress associated with death in patients with advanced cancer in Japan. If “Incompleteness” is an important concept in the pain associated with death among patients with cancer in Japan, preventing “Not having done all the things that I wanted to do” and “Not having said all that I wanted to say to the people I care about,” which constitute “Incompleteness,” may be important for alleviating the distress associated with death. This study had some limitations. First, the participants needed to be physically and psychologically well enough to complete the DADDS-J. We did not include patients who refused to answer the DADDS-J, which may affect the generalizability of the findings. Selection bias may have also been introduced, as participants with severe illnesses who could not complete the questionnaires might have been excluded. Second, we were unable to obtain additional evidence (e.g., test–retest reliability, known-group validity) for this study owing to the psychological and physical burden on the research participants. Third, 79.5% of the primary tumors were of gastrointestinal, lung, or oropharyngeal origin. This limits external validity to those malignancies, which are not as common. The psychometric properties of variables are very likely to differ in such populations that are symptomatically and diagnostically distinct, for example, individuals with brain tumors and cognitive deficits. As distress may vary by cancer type, future longitudinal studies are needed to investigate changes in death-related distress among patients with advanced cancer in Japan, using larger samples that allow separate analysis by cancer type or disease stage. Fourth, the final CFA model required the specification of three residual covariances based on the modification index. Although this improved the model fit, these covariances suggest that certain item pairs share variance beyond the range explained by their respective latent factors. This may indicate that the DADDS-J does not fully capture all relevant dimensions of death-related distress, or that some items overlap in content or exhibit similar response trends among Japanese patients with advanced cancer. Future studies should examine whether these residual associations are replicated in independent samples and, in this case, whether additional items or improved subscales are needed. Conclusion The DADDS-J demonstrated good reliability, construct validity and a unique factor structure. Future researchers could use the responsiveness of scores on this outcome measure to assess the effectiveness of psychotherapeutic interventions in relieving death-related distress in patients with advanced cancer [88]. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Acknowledgements We thank Kenji Azuma, Reo Inoue, Rikuhei Tsuchida, Mitsuru Konishi, and our research staff for their invaluable contributions to this study. We sincerely thank all the participants and their family supporters, without whom this study would not have been possible, and the healthcare providers of the hospitals who assisted with data collection. Author contributions Conceptualization, Seraki Miyamoto, Tadahiro Yamazaki, Ken Shimizu, and Kazuhiro Yoshiuchi; methodology, Seraki Miyamoto, Tadahiro Yamazaki, Ken Shimizu, and Kazuhiro Yoshiuchi; investigation, Seraki Miyamoto, Tadahiro Yamazaki, Toshio Matsubara, Hiroshi Yamagata, Hiroshi Kobo, Hiroaki Abe, Seraki Miyamoto, and Ken Shimizu; writing (original draft preparation), Seraki Miyamoto, Tadahiro Yamazaki, Saki Harashima, and Kazuhiro Yoshiuchi; writing (review and editing), Seraki Miyamoto, Tadahiro Yamazaki, Toshio Matsubara, Saki Harashima, Gary Rodin, and Kazuhiro Yoshiuchi; All authors have read and agreed to the published version of the manuscript. Funding This research was supported in part by a Grant-in-Aid for Research Activity Start-up (JP22K20878). Data Availability The complete dataset supporting the findings of this study is available within the manuscript and supplementary materials. The dataset is available from the corresponding author upon reasonable request with approval from the institutional ethics committee. Declarations Ethics approval and consent to participate This multicenter cross-sectional study was approved by the Ethics Committee of the Graduate School of Medicine, University of Tokyo (2019346NI); the Ethics Committee of the Cancer Institute Hospital, Japanese Foundation for Cancer Research (2020-GA-1242); and the Institutional Review Board of the Center for Clinical Research at Yamaguchi University Hospital (H2020-054). The patients/participants provided their written informed consent to participate in this study. Consent for publication Consent to publish has been received from all participants in the study. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement The complete dataset supporting the findings of this study is available within the manuscript and supplementary materials. The dataset is available from the corresponding author upon reasonable request with approval from the institutional ethics committee.