Psychological Adaptation to Cancer Progression and End-of-Life: State Anxiety, Stress-Related Growth, Well-Being, and Spiritual Experiences

Simple Summary

People living with cancer can experience profound psychological changes as illness progresses, but these changes do not always follow a simple pattern of worsening distress. This study followed the same 103 oncology patients from diagnosis to a terminal-phase assessment three years later and compared them with 103 survivors of severe non-cancer life-threatening events. State anxiety was lower at the terminal-phase assessment, while self-acceptance, stress-related growth, psychological well-being, and purpose/meaning were higher. Survivors of severe life-threatening events also showed a more favorable psychological profile than newly diagnosed oncology patients. Reports consistent with near-death experiences, deathbed visions, or other end-of-life experiences were documented more often in terminal-phase oncology patients. Because these experiences were not assessed identically across groups and time points, they should be interpreted as documented reports, not unbiased prevalence estimates. Future studies should use standardized assessment and matched longitudinal designs to clarify these patterns.

Abstract

Objective: This study examined psychological adaptation across cancer progression by following the same oncology patients from diagnosis to a terminal-phase assessment and by comparing them with survivors of severe non-cancer life-threatening events. Outcomes were state anxiety, unconditional self-acceptance, stress-related growth, psychological well-being, purpose and meaning in life, and exploratory reports of near-death experiences (NDEs), deathbed visions (DBVs), and other end-of-life experiences (ELEs). Methods: The analytic sample comprised 206 unique participants: 103 oncology patients with paired assessments at diagnosis (Time 1) and after three years in the terminal phase (Time 2), and an independent comparison group of 103 survivors of severe life-threatening events. The oncology cohort was drawn from 649 patients initially assessed in the same clinical setting; longitudinal analyses included the 103 patients with complete paired data through the final assessment. Repeated-measures ANOVA tested within-person psychological change; independent-samples t tests compared newly diagnosed oncology patients with life-threatening-event survivors; Pearson chi-square compared the frequency of documented spiritual-experience reports between the independent terminal-phase oncology and life-threatening-event groups; and McNemar’s exact test examined within-person change in spiritual-experience reporting from diagnosis to the terminal phase. Results: In the paired oncology cohort, state anxiety decreased and unconditional self-acceptance, stress-related growth, psychological well-being, and purpose/meaning increased from diagnosis to the terminal-phase assessment (all p < 0.001). Life-threatening-event survivors showed lower state anxiety and higher scores on the positive psychological outcomes than newly diagnosed oncology patients (all p < 0.001). Reports consistent with NDEs, DBVs, and ELEs were documented more frequently in terminal-phase oncology patients (94.2%) than in life-threatening-event survivors (63.1%), χ²(1, N = 206) = 29.59, p < 0.001, Cramer’s V = 0.38. Within the paired oncology cohort, spiritual-experience reporting also increased from 14.6% at diagnosis to 94.2% in the terminal phase; 85 patients changed from no report to a report and only 3 changed in the opposite direction (McNemar exact p < 0.001). Conclusion: Psychological functioning differed markedly across phases and contexts of mortality-related threat. Reports consistent with NDEs, DBVs, and ELEs were documented more frequently in the terminal phase and increased within the paired oncology cohort, but the observational design does not establish that these experiences caused or predicted psychological improvement. The findings support further longitudinal, controlled, and qualitative research on adaptation, meaning, and spiritual experience in oncology and end-of-life care.

1. Introduction

Cancer diagnosis and progression confront patients with uncertainty, changes in control, and the possibility of death. Psychological adaptation in this context is not necessarily linear: acute distress can coexist with later acceptance, meaning reconstruction and positive psychological change. Contemporary models of death acceptance and end-of-life adjustment therefore emphasize both vulnerability and the possibility of adaptive reorganization rather than assuming inevitable psychological deterioration as illness advances [1,2].

The present study focuses on five complementary psychological outcomes. State anxiety captures transient distress in response to an immediate threat; unconditional self-acceptance reflects the capacity to maintain self-worth without global self-condemnation; stress-related growth captures perceived positive change following adversity; psychological well-being represents broader eudaimonic functioning; and purpose and meaning in life reflect a central component of well-being that may become particularly salient during serious illness [3,4,5]. Considered together, these constructs provide a clinically relevant profile of distress and adaptation across the cancer trajectory.

Spirituality is relevant to end-of-life care because questions of meaning, connection, transcendence, and existential peace often intensify in serious illness [2,6]. Reports of near-death experiences (NDEs), deathbed visions (DBVs), and other end-of-life experiences (ELEs) have been described in the literature, although their phenomenology, interpretation, and clinical significance vary and should not be reduced to a single explanatory model [7,8]. Evidence also indicates that spiritual well-being is related to quality of life in palliative populations [9], and that appropriately delivered spiritual interventions may improve selected psychological and quality-of-life outcomes in people with cancer [10]. A recent synthesis of research in oncology further emphasizes the breadth of spiritual concerns in cancer and the methodological importance of careful assessment [11]. These intervention and broader spirituality findings should, however, be distinguished from observational associations involving spontaneous spiritual experiences.

A comparison group of individuals who had survived a severe non-cancer life-threatening event was included to examine whether psychological patterns associated with confrontation with mortality were specific to ongoing cancer or were also evident after another form of acute threat. In the present sample, these events primarily involved severe road-traffic trauma accompanied by critical medical conditions such as coma, cardiorespiratory arrest, and/or medically induced coma used in the management of extensive life-threatening injuries. Participants were assessed after survival and clinical stabilization, when they were able to complete the psychological evaluation. Trauma research indicates that some survivors report positive changes in appreciation of life, relationships, personal strength, or meaning after severe events, although such growth is neither universal nor evidence of an absence of distress [12]. This comparison is conceptually useful because survivors had already passed through an acute critical event, whereas newly diagnosed oncology patients were at the beginning of an uncertain disease trajectory. At the same time, differences in age, event characteristics, and time since the event can influence such comparisons and are treated as limitations.

Death anxiety is an important theoretical background construct in psycho-oncology; however, the present study did not administer a specific death-anxiety scale. Accordingly, all empirical anxiety findings reported below refer to STAI state anxiety and should not be interpreted as direct measurements of death anxiety. This distinction is important when interpreting both the longitudinal findings and their clinical implications.

The study had three aims. First, it examined within-person psychological change in the same oncology patients from diagnosis (Time 1) to a terminal-phase assessment three years later (Time 2). Second, it compared newly diagnosed oncology patients with an independent group of survivors of severe life-threatening events. Third, it examined documented spiritual-experience reporting in two complementary ways: between the independent terminal-phase oncology and life-threatening-event groups, and longitudinally within the same oncology patients from diagnosis to the terminal phase.

Hypothesis 1 predicted lower state anxiety and higher unconditional self-acceptance, stress-related growth, psychological well-being, and purpose/meaning at the terminal-phase assessment than at diagnosis in the paired oncology cohort. Hypothesis 2 predicted that life-threatening-event survivors would show lower state anxiety and higher positive psychological outcomes than newly diagnosed oncology patients. Hypothesis 3a predicted a higher documented frequency of reported spiritual experiences among terminal-phase oncology patients than among life-threatening-event survivors. Hypothesis 3b predicted a significant within-person increase in the reporting of spiritual experiences from diagnosis to the terminal phase among the paired oncology cohort. The spiritual-experience indicators were treated as exploratory categorical outcomes rather than as causal mechanisms.

2. Materials and Methods

2.1. Study Design

This study used a mixed longitudinal and cross-sectional design. The longitudinal component involved the same 103 oncology patients assessed at cancer diagnosis (Time 1) and again three years later, when they were clinically classified as being in the terminal phase of disease (Time 2). The cross-sectional component used an independent comparison group of 103 adults who had survived a severe non-cancer life-threatening event. Thus, the study included 206 unique participants; the two oncology assessments represent repeated observations from the same 103 individuals and were analyzed as paired data.

2.2. Setting and Participant Flow

Participants were recruited through the same clinical setting, in collaboration with a local county hospital and its psycho-oncology service. A total of 649 oncology patients were initially assessed while receiving standard oncological care, including chemotherapy and/or radiotherapy when clinically indicated. Patients routinely returned to the clinical service for treatment, medical review, laboratory investigations, follow-up examinations, and psycho-oncology support. During the three-year follow-up, attrition occurred because some patients died before the final assessment, whereas others did not return for or did not complete all scheduled psychological assessments. The longitudinal analyses therefore included the 103 patients who repeatedly returned to the psycho-oncology service, had complete paired data, and were clinically classified as being in the terminal phase at the final assessment. The available analytic records did not support a reliable numerical breakdown of attrition by death versus non-return or incomplete assessment, and no such breakdown is inferred.

Terminal-phase status was determined by the treating oncology physicians rather than by a research-defined prognostic cut-off. In this clinical context, terminal status referred to advanced oncological disease characterized by lack of meaningful response to ongoing treatment and/or rapidly progressive or recurrent disease, with a markedly limited prognosis. Patients in this clinical situation were routinely referred or re-referred to psycho-oncology as part of supportive care.

The independent life-threatening-event survivor group comprised 103 adults recruited through the same clinical setting after survival and clinical stabilization. Their histories included severe road-traffic accidents and associated critical medical states such as coma, cardiorespiratory arrest, and medically induced coma used in the management of extensive life-threatening trauma. These descriptors characterize the clinical histories represented in the group and do not imply that every participant experienced every listed condition. Event type was not used as a stratification variable in the present analyses. Data collection spanned January 2022 to January 2025.

2.3. Participants

The final analytic sample comprised 206 unique participants. The paired oncology cohort included 103 patients (68 men, 66.0%; 35 women, 34.0%; age range 31–76 years) who contributed data at both diagnosis and the terminal-phase assessment. The initial oncology pool included diagnoses such as breast, colon, lung, skin, and brain cancers; the complete-case analyses were not stratified by tumor type. Participant-level demographic summaries and detailed staging/treatment distributions for the complete initial pool of 649 patients were not retained in the analytic dataset available for the present revision and therefore were not reconstructed or inferred. The independent life-threatening-event survivor group comprised 103 participants (45 men, 43.7%; 58 women, 56.3%; age range 19–59 years). Across the 206 unique participants, there were 113 men (54.9%) and 93 women (45.1%). Inclusion in the analytic sample required complete data for the measures relevant to the corresponding comparison. The life-threatening-event group was younger in range than the oncology cohort; age was not matched or statistically adjusted in the present analyses and is therefore treated as a potential confounder.

2.4. Variables and Data Sources

Stress-related growth was assessed with a 15-item short form of the Stress-Related Growth Scale (SRGS), rated from 0 to 2, with higher scores indicating greater perceived positive change following adversity [4]. In the Romanian version used in this study, internal consistency was α = 0.85. The construct is referred to throughout the revised manuscript as stress-related growth to match the measure administered.

State anxiety was measured with the 20-item State subscale of the State-Trait Anxiety Inventory (STAI), rated on a 4-point response scale [13]. Higher scores indicate greater current anxiety. The STAI assesses transient state anxiety rather than death anxiety. Reported reliability for the State subscale is high, and the Romanian adaptation has been used in psychological assessment contexts [13,14].

Unconditional self-acceptance was evaluated using the 20-item Unconditional Self-Acceptance Questionnaire (USAQ), which assesses acceptance of the self without global evaluative judgment [3]. Responses range from statements that are almost always false to almost always true and are summed to obtain a global score, with higher scores indicating greater unconditional self-acceptance.

Psychological well-being was assessed with the 42-item Ryff Psychological Well-Being scale (PWB-R), covering six dimensions of eudaimonic functioning [5]. The total score and the purpose-in-life dimension were analyzed. Responses were recorded on the study’s 7-point response format, and the scale showed a reliability coefficient of 0.86 in the study documentation. Scores were treated dimensionally rather than interpreted against diagnostic cut-offs.

Spiritual and end-of-life experiences were documented exploratorily as reports consistent with near-death experiences (NDEs), deathbed visions (DBVs), and end-of-life experiences (ELEs), rather than through a standardized psychometric instrument. For study-specific coding, an NDE referred to an unusual subjective experience remembered in connection with a documented life-threatening episode, such as an altered bodily perspective, an intense perception of light, unusual peace, or perceptions of familiar or deceased persons; a DBV referred to vivid perceptions or visions reported during advanced illness, often involving deceased or familiar persons, luminous presences, or other meaningful scenes; and ELE was used as a broader category for other unusual experiences reported around severe or advanced illness that did not fit the first two categories. These descriptors served as clinical-research coding guides rather than validated diagnostic criteria. In the oncology cohort, research coding was based on experiences documented during psycho-oncology contacts corresponding to the two study assessment periods, at diagnosis (Time 1) and during the terminal phase (Time 2). These experiences were generally raised spontaneously by patients during psycho-oncology encounters. When such material emerged, brief neutral follow-up questions were used only to clarify the reported experience sufficiently for study-specific categorization. No fixed standardized item set or validated NDE/DBV/ELE instrument was administered uniformly to oncology patients at both time points. For survivors of severe life-threatening events, the ascertainment pathway differed. After the study had been explained, data-protection information had been provided, and voluntary informed consent had been obtained, participants were asked in neutral terms whether they recalled any unusual perceptions or experiences associated with the documented critical episode and, when such experiences were reported, were invited to describe them briefly. Reports were subsequently categorized as NDE, DBV, or ELE using the study-specific descriptors and coded dichotomously as present or absent. For quantitative analyses, the categories were collapsed into a binary indicator denoting whether at least one such experience had been documented. A validated NDE scale exists in the literature [15], but it was not administered in the present study; consequently, the NDE category reported here should not be interpreted as a Greyson Scale-defined NDE. The study-specific categories were not a validated composite psychometric scale, so internal-consistency coefficients, normative cut-offs, and clinical thresholds are not applicable. Detailed narratives were not collected systematically across all participants and were not subjected to qualitative analysis. Because the ascertainment pathway was not identical across groups and opportunities for spontaneous disclosure could also differ across the oncology trajectory, these variables should be interpreted as indicators of documented reporting rather than as standardized measures of incidence or prevalence.

2.5. Procedure and Bias Control

All participants provided written informed consent. The oncology cohort completed questionnaires during contacts with the psycho-oncology service at the two study assessment periods. For the oncology cohort, spiritual/end-of-life experience indicators reflected experiences documented during the corresponding psycho-oncology contacts; such experiences were generally raised spontaneously and, when necessary, clarified with brief neutral follow-up questions for study-specific categorization. No fixed standardized spiritual-experience item set was administered uniformly at both oncology time points. Survivors of severe life-threatening events were identified through the same clinical setting while hospitalized or returning for medical follow-up; the study was presented to them separately, data-protection/GDPR information was provided, and only those who voluntarily agreed attended the research assessment and completed the psychological measures. Supervision during questionnaire completion was procedural: it allowed clarification of instructions, practical assistance when needed, and checking for inadvertently omitted items. Participation or refusal had no consequence for clinical care, and clinical management was independent of questionnaire responses. Because assessment occurred in a professional clinical context, however, perceived pressure or social-desirability responding cannot be excluded and is acknowledged as a potential source of response bias.

2.6. Sample Size

No a priori power calculation determined the final analytic sample. The longitudinal sample size was determined by the number of oncology participants from the initial cohort who had complete paired assessments through the three-year follow-up and met the final clinical status required for the longitudinal comparison. The independent life-threatening-event group contained 103 complete cases.

2.7. Quantitative Variables

All continuous outcomes were analyzed as dimensional variables. The study questions concerned within-person change and between-group differences rather than diagnostic classification. Accordingly, no ad hoc clinical thresholds were introduced where a validated diagnostic cut-off was not part of the instrument’s intended use in this study. Descriptive statistics, distributional characteristics, and effect sizes are reported to support interpretation of the observed differences, and no score is interpreted as establishing a psychiatric diagnosis.

2.8. Statistical Methods

All statistical analyses were performed using IBM SPSS Statistics, version 26.0 (IBM Corp., Armonk, NY, USA). All tests were two-sided with α = 0.05. Descriptive statistics and skewness/kurtosis were examined before inferential analyses. H1 was tested with repeated-measures ANOVA in the paired oncology cohort (diagnosis versus terminal-phase assessment). With only two time points, sphericity is not an applicable constraint. H2 was tested with independent-samples t tests comparing newly diagnosed oncology patients with life-threatening-event survivors; Welch’s t test was used when Levene’s test indicated unequal variances, and the conventional Student t test was used when homogeneity of variance was supported. H3a was tested with a Pearson chi-square test comparing presence versus absence of at least one spiritual experience between two independent groups only: terminal-phase oncology patients and life-threatening-event survivors. Effect size was estimated with Cramer’s V. H3b examined within-person change in binary spiritual-experience reporting between diagnosis and the terminal phase in the same 103 oncology patients using McNemar’s exact test. Effect sizes are reported as partial eta squared (ηp²), absolute Cohen’s d, and Cramer’s V, as appropriate.

3. Results

Descriptive statistics for all continuous outcomes are presented in Table 1. The diagnosis and terminal-phase rows refer to the same 103 oncology patients measured at two time points; the life-threatening-event rows refer to an independent group of 103 participants.

Table 1. Descriptive statistics for psychological outcomes by assessment condition/group.

Assessment/Group	Outcome	N	Mean	SD	Skewness	Kurtosis
Oncology—diagnosis (T1)	State anxiety	103	58.76	6.85	0.502	−0.655
Oncology—diagnosis (T1)	Stress-related growth	103	10.68	2.06	−0.094	−0.661
Oncology—diagnosis (T1)	Unconditional self-acceptance	103	74.89	4.85	0.259	−1.207
Oncology—diagnosis (T1)	Psychological well-being	103	49.35	2.30	−0.636	−0.632
Oncology—diagnosis (T1)	Purpose/meaning in life	103	16.33	2.92	−0.317	−0.580
Oncology—terminal phase (T2)	State anxiety	103	28.26	4.67	0.811	0.274
Oncology—terminal phase (T2)	Stress-related growth	103	21.14	1.56	0.197	−1.104
Oncology—terminal phase (T2)	Unconditional self-acceptance	103	105.91	9.70	−0.162	−1.738
Oncology—terminal phase (T2)	Psychological well-being	103	51.61	1.87	1.052	0.300
Oncology—terminal phase (T2)	Purpose/meaning in life	103	27.12	2.05	−0.049	0.518
Life-threatening-event survivors	State anxiety	103	39.77	5.52	0.230	−0.036
Life-threatening-event survivors	Stress-related growth	103	22.50	2.11	−0.160	−0.859
Life-threatening-event survivors	Unconditional self-acceptance	103	97.93	8.35	0.596	−0.649
Life-threatening-event survivors	Psychological well-being	103	58.38	4.21	0.516	−0.943
Life-threatening-event survivors	Purpose/meaning in life	103	24.37	2.87	0.375	−1.364

3.1. Hypothesis 1

H1 predicted lower state anxiety and higher unconditional self-acceptance, stress-related growth, psychological well-being, and purpose/meaning at the terminal-phase assessment than at diagnosis in the paired oncology cohort.

Repeated-measures ANOVA supported H1 for all five outcomes. State anxiety decreased from diagnosis to the terminal-phase assessment, F(1, 102) = 1290.33, p < 0.001, ηp² = 0.927. Unconditional self-acceptance increased, F(1, 102) = 749.59, p < 0.001, ηp² = 0.880; stress-related growth increased, F(1, 102) = 1618.86, p < 0.001, ηp² = 0.941; psychological well-being increased, F(1, 102) = 86.72, p < 0.001, ηp² = 0.460; and purpose/meaning in life increased, F(1, 102) = 1092.05, p < 0.001, ηp² = 0.915. With only two repeated observations, no sphericity correction was required. Full results are summarized in Table 2.

Table 2. Repeated-measures ANOVA for the paired oncology cohort (diagnosis vs. terminal-phase assessment).

Outcome	Diagnosis T1 M (SD)	Terminal T2 M (SD)	F(1, 102)	p	ηp2
State anxiety	58.76 (6.85)	28.26 (4.67)	1290.33	<0.001	0.927
Unconditional self-acceptance	74.89 (4.85)	105.91 (9.70)	749.59	<0.001	0.880
Stress-related growth	10.68 (2.06)	21.14 (1.56)	1618.86	<0.001	0.941
Psychological well-being	49.35 (2.30)	51.61 (1.87)	86.72	<0.001	0.460
Purpose/meaning in life	16.33 (2.92)	27.12 (2.05)	1092.05	<0.001	0.915

Section 3 reports the observed statistical differences without attributing them to adaptation, meaning-making, disease progression, or spiritual mechanisms; possible explanations are considered in Section 4.

3.2. Hypothesis 2

H2 predicted lower state anxiety and higher unconditional self-acceptance, stress-related growth, psychological well-being, and purpose/meaning among life-threatening-event survivors than among oncology patients at diagnosis.

Independent-samples comparisons supported H2 for all five outcomes (all p < 0.001). Levene’s test indicated unequal variances for state anxiety, unconditional self-acceptance, and psychological well-being; Welch’s t tests were therefore used for those outcomes. Homogeneity of variance was supported for stress-related growth and purpose/meaning, for which conventional Student t tests were used. Life-threatening-event survivors showed lower state anxiety and higher scores on all four positive psychological outcomes than newly diagnosed oncology patients. Table 3 reports the test selected for each outcome, the corresponding degrees of freedom, effect-size magnitude, and the Levene test result used to select the appropriate t test.

Table 3. Independent-samples comparisons between oncology patients at diagnosis and life-threatening-event survivors.

Outcome	Diagnosis M (SD)	LTE Survivors M (SD)	Test	df	p	|d|	Levene p
State anxiety	58.76 (6.85)	39.77 (5.52)	Welch t = 21.898	195.211	<0.001	3.05	0.012
Stress-related growth	10.68 (2.06)	22.50 (2.11)	Student t = −40.716	204	<0.001	5.67	0.730
Unconditional self-acceptance	74.89 (4.85)	97.93 (8.35)	Welch t = −24.222	163.852	<0.001	3.38	<0.001
Psychological well-being	49.35 (2.30)	58.38 (4.21)	Welch t = −19.109	157.934	<0.001	2.66	<0.001
Purpose/meaning in life	16.33 (2.92)	24.37 (2.87)	Student t = −19.921	204	<0.001	2.78	0.632

All inferential p values in Table 3 refer to the t test of the group difference. Levene’s p values are reported separately only to document the variance assumption and the choice between Student and Welch t tests.

3.3. Hypothesis 3

3.3.1. H3a: Independent-Group Comparison

H3a predicted that the frequency of documented reports of spiritual/end-of-life experiences (endorsement of at least one NDE, DBV, or ELE category) would be higher among terminal-phase oncology patients than among survivors of severe life-threatening events. Because these two groups are independent, a Pearson chi-square test was appropriate. Such experiences were documented in 97 of 103 terminal-phase oncology patients (94.2%) and in 65 of 103 life-threatening-event survivors (63.1%). The difference was statistically significant, χ²(1, N = 206) = 29.59, p < 0.001, Cramer’s V = 0.38. No expected cell count was below 5 (minimum expected count = 22.00), supporting the chi-square approximation. Because ascertainment procedures and opportunities for disclosure differed between groups, this comparison is interpreted as a difference in the frequency of documented reports rather than as an unbiased estimate of underlying population prevalence. It does not establish that these experiences caused the psychological differences observed in H1 or H2. The corresponding results are presented in Table 4.

Table 4. Presence of at least one spiritual experience in the two independent groups included in H3a.

Group	With Spiritual Experience n (%)	Without n (%)	Total	χ2(1)	p	Cramer’s V
Terminal-phase oncology	97 (94.2%)	6 (5.8%)	103	29.59	<0.001	0.38
Life-threatening-event survivors	65 (63.1%)	38 (36.9%)	103
Total	162 (78.6%)	44 (21.4%)	206

3.3.2. H3b: Longitudinal Change Within the Oncology Cohort

H3b examined whether spiritual-experience reporting changed within the same 103 oncology patients from diagnosis to the terminal-phase assessment. At diagnosis, 15 patients (14.6%) reported at least one NDE, DBV, or ELE indicator, whereas 97 patients (94.2%) reported at least one such experience in the terminal phase. Of the 88 patients who did not report a spiritual experience at diagnosis, 85 reported one at the terminal-phase assessment and 3 remained without a report. Of the 15 patients who reported an experience at diagnosis, 12 continued to report one at the terminal-phase assessment and 3 did not. The discordant-pair pattern was highly asymmetric (85 no-to-yes transitions versus 3 yes-to-no transitions), and McNemar’s exact test indicated a statistically significant change over time, p < 0.001. The corresponding results are presented in Table 5.

Table 5. Paired change in spiritual-experience reporting from diagnosis to the terminal phase among the 103 oncology patients.

Diagnosis (T1)	Terminal: With	Terminal: Without	Row Total
With spiritual experience	12	3	15 (14.6%)
Without spiritual experience	85	3	88 (85.4%)
Column total	97 (94.2%)	6 (5.8%)	103

The longitudinal H3b result documents a marked increase in the frequency of recorded spiritual/end-of-life experience reports within the paired oncology cohort. Because the indicators were binary, detailed timing and qualitative phenomenology were not collected systematically, and opportunities for disclosure may have varied across the disease trajectory, this finding should be interpreted as a change in documented reporting rather than as evidence of incidence, prevalence, or a specific psychological or spiritual mechanism.

4. Discussion

This study examined psychological adaptation across two contexts of mortality-related threat. In the paired oncology cohort, state anxiety was substantially lower and unconditional self-acceptance, stress-related growth, psychological well-being, and purpose/meaning were higher at the terminal-phase assessment than at diagnosis. In the independent comparison, life-threatening-event survivors also showed a more favorable psychological profile than newly diagnosed oncology patients. Reports consistent with NDEs, DBVs, and ELEs were documented more frequently among terminal-phase oncology patients than among life-threatening-event survivors, and the paired oncology analysis showed a marked increase in spiritual-experience reporting from diagnosis to the terminal phase. These findings should be interpreted as patterns of change and association, not as evidence that spiritual experiences caused later psychological improvement.

The longitudinal oncology findings challenge a simple assumption that psychological functioning must worsen uniformly as disease advances. Acute diagnosis can involve uncertainty and threat, whereas later phases may also involve accommodation, reorganization of priorities, and greater acceptance. Models of death acceptance and meaning-oriented adaptation describe such processes as possible responses to mortality awareness [1]. The present data are compatible with that interpretation, but they cannot establish the mechanism responsible for change. In particular, the complete-case cohort represents patients who survived and remained evaluable across three years, so selection and attrition may have contributed to the observed trajectory.

The decrease in state anxiety requires especially careful interpretation. The study measured current state anxiety with the STAI, not death anxiety. Therefore, the finding supports a reduction in general situational anxiety at the later assessment, but it does not demonstrate a reduction in fear of death specifically. Likewise, higher well-being and purpose scores at Time 2 should not be described as evidence that clinical deterioration itself produces psychological benefit. They indicate that positive psychological functioning can coexist with advanced illness in this selected longitudinal cohort.

The comparison with life-threatening-event survivors offers a second perspective on adaptation. Survivors showed lower state anxiety and higher self-acceptance, stress-related growth, well-being, and purpose/meaning than oncology patients assessed at diagnosis. These differences are compatible with processes described in the trauma-growth literature [12], but they do not demonstrate that the life-threatening event itself produced better adjustment. One plausible interpretation is that time after a survived threat permits cognitive and emotional reappraisal, but alternative explanations remain important. The comparison group was younger in range and heterogeneous with respect to the precipitating event, and it was not matched to the oncology cohort on age or clinical characteristics. In this sample, severe traffic trauma and some cardiorespiratory-arrest histories were represented among younger adults; this is a sample characteristic rather than evidence of a general age distribution. Age, event type, elapsed time since the event, and other unmeasured factors may therefore confound the between-group differences.

The spiritual/end-of-life experience findings require particularly careful interpretation. Reports consistent with NDEs, DBVs, and ELEs were more frequently documented in the terminal-phase oncology group than in the life-threatening-event survivor group, and documented reporting increased substantially within the oncology cohort over time. However, the ascertainment pathway was not identical across groups: oncology patients generally raised these experiences spontaneously during psycho-oncology encounters, whereas survivors of severe life-threatening events were directly asked, after informed consent, whether they remembered unusual perceptions or experiences during the critical episode. Opportunity for disclosure also likely varied across the oncology trajectory because patients in advanced disease often had more intensive supportive-care and psycho-oncology contact. These differences may have influenced what was disclosed and recorded. In addition, detailed narratives were not collected systematically for every participant; the study therefore cannot characterize phenomenology, emotional valence, personal meaning, cultural or religious interpretation, or the precise timing of each experience. Accordingly, the results should be understood as differences and changes in the frequency of documented reports, not as unbiased estimates of underlying prevalence or evidence of how these experiences affect psychological adaptation.

This distinction is clinically important. The present results do not show that spiritual experiences measured at an earlier time predict later well-being, nor do they justify using NDEs, DBVs, or ELEs as prognostic indicators. A defensible implication is that clinicians in oncology and palliative care should be prepared to respond nonjudgmentally when patients raise spiritual or existential concerns. Spiritual well-being has been associated with quality of life in palliative populations [9], and intervention evidence supports the potential value of appropriately delivered spiritual care for selected outcomes in cancer populations [10]. Recent psycho-oncology literature likewise underscores spirituality as an important but methodologically complex dimension of cancer care [11]. Those findings, however, should not be conflated with the observational associations reported here.

The study has several strengths. The three-year follow-up of the same oncology patients provides uncommon longitudinal information across a clinically difficult trajectory, and the inclusion of multiple psychological domains allows distress, self-acceptance, growth, well-being, and purpose to be considered together. The additional comparison with survivors of life-threatening events broadens the conceptual frame beyond cancer alone. The revised H3 analyses also distinguish correctly between independent-group comparison and within-person longitudinal change, thereby respecting the dependence structure of the data. At the same time, the strongest contribution of the present data remains descriptive and comparative rather than causal.

4.1. Clinical and Research Implications

The findings support routine attention to psychological and existential needs throughout the cancer trajectory rather than assuming that distress follows a single worsening course. Assessment should distinguish general state anxiety from death-specific anxiety, and unusual or spiritual/end-of-life experiences should be approached nonjudgmentally and only when clinically relevant or raised by the patient. Future research should use larger multisite cohorts, explicitly record cancer stage and treatment variables, match or statistically adjust comparison groups on age and other confounders, and employ a standardized ascertainment protocol with comparable assessment opportunities for NDEs, DBVs, and ELEs across groups and time points. Where NDEs are a primary outcome, validated instruments such as the Greyson NDE Scale should be considered [15]. Qualitative or mixed-method designs are particularly needed to capture the content, timing, subjective meaning, and emotional consequences of these experiences before translating differences in documented frequency into clinical recommendations.

4.2. Limitations

Several limitations constrain interpretation. First, recruitment occurred through a single clinical setting, limiting generalizability. Second, the longitudinal analysis used complete cases from an initial pool of 649 oncology patients; patients who died before the final assessment or did not complete follow-up were not represented in the paired analysis, creating potential attrition and survivor-selection bias. Exact attrition counts by reason, participant-level demographic summaries for the entire initial pool, and complete staging/treatment distributions were not available in the analytic dataset used for this revision; these data were therefore not reconstructed or inferred. Third, terminal-phase status was based on clinical classification by treating oncology physicians rather than on a research-defined prognostic algorithm, which limits reproducibility of that designation across settings. Fourth, the life-threatening-event group was younger in range, was not age-matched to the oncology cohort, and was heterogeneous with respect to the precipitating event; age and event characteristics may therefore confound the cross-sectional comparisons. Fifth, ascertainment of NDEs, DBVs, and ELEs was exploratory and not fully standardized across groups or time points. In the oncology cohort, the research indicators at diagnosis and at the terminal-phase assessment reflected experiences documented during the corresponding psycho-oncology contacts, where such experiences were generally disclosed spontaneously and clarified when necessary. By contrast, life-threatening-event survivors were directly asked, after informed consent, whether they recalled unusual perceptions or experiences associated with the documented critical episode. Opportunities for disclosure may also have varied across the oncology trajectory because the intensity and context of psycho-oncology contact were not necessarily identical at diagnosis and in the terminal phase. Accordingly, the resulting variables represent documented reports rather than unbiased estimates of incidence or prevalence. They were binary, were not derived from a validated composite scale, and detailed narratives were not collected systematically. Sixth, questionnaire completion occurred under professional supervision; although this facilitated procedural clarification and reduced missing data, it may also have increased perceived pressure or social-desirability responding. Seventh, the study did not measure death anxiety directly. Finally, the observational design and timing of the spiritual-experience measures do not establish causal or predictive effects on later psychological outcomes. Data collection in this population was also constrained by substantial clinical instability and vulnerability: physical condition, symptom burden, treatment status, and capacity to complete assessments could change rapidly, particularly among patients with progressive cancer and, to a lesser extent, among survivors recovering from severe life-threatening events. These circumstances limited the depth and uniformity with which all variables could be assessed at every contact. They also imposed considerable practical and emotional demands on research staff working repeatedly in the presence of severe illness, deterioration, and end-of-life concerns; this contextual burden is relevant to the feasibility of longitudinal research in such settings, although it does not alter the statistical procedures themselves.

5. Conclusions

Among oncology patients who completed both assessments, state anxiety was lower and unconditional self-acceptance, stress-related growth, psychological well-being, and purpose/meaning were higher at the terminal-phase assessment than at diagnosis. Life-threatening-event survivors also showed a more favorable psychological profile than newly diagnosed oncology patients in unadjusted comparisons. Reports consistent with NDEs, DBVs, and ELEs were documented more frequently among terminal-phase oncology patients than among life-threatening-event survivors and increased markedly within the paired oncology cohort from diagnosis to the terminal phase. Because ascertainment procedures and opportunities for disclosure were not identical across groups or necessarily constant over time, these findings should be interpreted as differences in documented reporting rather than as unbiased estimates of incidence or prevalence. They highlight heterogeneity in psychological adaptation to mortality-related threats and support further prospective, controlled, standardized, and qualitative research. They should not be interpreted as evidence that spiritual experiences cause psychological improvement or as support for prognostic use of such experiences.