Post-Traumatic Growth in Esophageal Cancer Patients: A Latent Profile Analysis for Targeted Psychosocial Care

Abstract

Purpose

This study’s aim is to assess the latent profiles of post-traumatic growth in patients with esophageal cancer and to analyze the relationship with perceived social support and cognitive appraisal, thereby highlighting distinct PTG subgroups that may warrant consideration of personalized psychosocial support strategies.

Patients and Methods

Patients with esophageal cancer visited a tertiary Class A hospital from January to April 2026 in Tai’an were selected using convenience sampling. These patients were surveyed using a general data questionnaire, the Chinese Version of Post-Traumatic Growth Inventory, the Perceived Social Support Scale and the Cognitive Appraisal Scale. Latent profile analysis was used to determine different types of post-traumatic growth among patients, and influencing factors for the various profiles were identified through univariate analysis and multinomial logistic regression analysis.

Results

Post-traumatic growth in esophageal cancer patients was categorized into three latent profiles: the “low-growth group”, the “moderate-growth group”, and the “high-growth group”. Compared with the high-growth group, higher scores for positive appraisal and friend support were associated with a lower likelihood of belonging to the low-growth group (OR = 0.489 and 0.824, respectively, p< 0.001) and moderate-growth group (OR = 0.528 and 0.889, p< 0.008); patients with squamous cell carcinoma were more likely to fall into the low-growth group (OR =2.563, p = 0.048) or moderate-growth group (OR =3.692, p = 0.001); publicly funded patients were significantly associated with the moderate-growth group (OR =4.185, p = 0.034) and showed a marginally significant trend for the low-growth group (OR =4.142, p = 0.054).

Conclusion

The post-traumatic growth levels of patients with esophageal cancer exhibited significant heterogeneity. These distinct profiles may inform the development of targeted psychosocial interventions, with potential implications for patient-centered care research.

Introduction

Esophageal cancer remains a major global health challenge. Worldwide data from 2022 show approximately 511,000 new cases and 445,000 deaths, making it the 11th most common cancer and the 7th leading cause of cancer death.¹ Demographic trends suggest the burden will rise substantially in the coming decades: an estimated 957,000 new cases and 880,000 deaths are expected by 2040, driven mainly by population aging and changing patterns of lifestyle-related risk.² China bears the greatest share of this burden, accounting for roughly 44% of global cases and 42% of cancer deaths.³ National surveillance data show an crude incidence rate of 15.87 per 100,000 (around 224,000 new diagnoses annually, ranking 7th among all malignancies) and a mortality rate of 13.28 per 100,000 (about 187,000 deaths per year, 5th leading cause of cancer mortality).⁴ The high incidence rates in East and Central China have significantly increased the strain on China’s public health system.⁵

While early detection and treatment greatly improve survival,⁶ patients face considerable physical and psychological challenges after treatment.⁷ Traditionally, psychosocial research has focused on negative outcomes—anxiety, depression, and related distress—with less attention to positive psychological adaptations following cancer-related trauma. Tedeschi and Calhoun called this phenomenon post-traumatic growth (PTG).⁸ PTG refers to a positive psychological change that arises from confronting challenging life situations, beyond usual psychological growth experiences and involves meaningful changes in self-perception, relationships, and life priorities, including greater appreciation for life, stronger social bonds, and a better sense of personal resilience.⁸ For patients with esophageal cancer, PTG offers a valuable framework for understanding the positive psychological changes they experience during their physical recovery. Furthermore, studies have shown that PTG levels are positively correlated with quality of life.⁹ This association may reflect patients’ capacity for positive psychological adaptation in the face of life-threatening illness. In Tedeschi and Calhoun’s framework, such adaptation typically takes the form of closer interpersonal relationships, reordered priorities, and a deeper appreciation of life.⁸ However, these positive changes are not uniformly experienced: emerging evidence indicates that PTG is heterogeneous, with individuals varying widely in their patterns of positive change.^10–12 Yet few studies have systematically characterized PTG subgroups among patients with esophageal cancer—a gap that is especially pronounced in China, which bears the world’s highest burden of this disease.^3–5 Although a range of psychosocial factors associated with PTG has been identified, it remains unclear which factors differentiate PTG profiles in this population. Drawing on existing theory and research, we focused on two candidate factors: perceived social support and cognitive appraisal.

Perceived social support refers to the assistance patients subjectively experience from family, friends, and broader social networks. This support provides emotional comfort and practical help, strengthens patients’ confidence in managing their illness, and is positively associated with PTG development.^13,14 Cognitive appraisal describes how individuals interpret, reflect on, and make sense of traumatic experiences—processes critical to adapting to cancer.¹⁵ An appraisal style oriented toward benefit finding and optimism has been associated with more flexible coping, better psychological adjustment, and higher levels of PTG.¹⁶

Traditional approaches using total scores or arbitrary cut-offs (eg, low/medium/high splits) obscure individual differences in item-level responses, treating potentially heterogeneous patients as homogeneous.¹⁷ Latent Profile Analysis (LPA) overcomes this by classifying patients into empirically distinct subgroups based on their full item-level response patterns, rather than relying on researcher-imposed cut-points. This data-driven approach maximizes between-group differences while ensuring within-group consistency, thereby capturing patient heterogeneity that aggregate methods miss.¹⁸ Such heterogeneity is clinically meaningful, as it reflects patients’ differential capacity for positive adaptation following their cancer experience—a capacity that has been linked to psychological adjustment and social functioning in cancer populations.¹⁰ However, classification alone is insufficient; profiles must also be characterized and their correlates understood. The present study therefore aimed to identify latent PTG profiles among patients with esophageal cancer and examine whether perceived social support and cognitive appraisal are associated with the likelihood of belonging to a specific PTG profile. Findings are expected to contribute to profile-tailored psychosocial care—for example, by helping clinicians identify patients with low-growth profiles earlier and informing the future allocation of interventions that strengthen social support or foster cognitive appraisal.

Materials and Methods

Subjects

Patients with esophageal cancer who visited a tertiary Class A hospital in Tai’an, Shandong Province, between January and April 2026, were selected using convenience sampling. The inclusion criteria were as follows: (1) a pathological diagnosis of esophageal cancer; (2) age ≥18 years; (3) stable clinical condition with the ability to complete the questionnaire independently; (4) willingness to participate in the study and provision of signed informed consent. The exclusion criteria were: (1) presence of severe organic diseases or complications; (2) history of psychiatric disorders or visual, auditory, and cognitive impairments; (3) experience of major stressful events within the past six months. Existing studies recommend that the total sample size for LPA should exceed 300, and that no fewer than 30 individuals per potential profile.¹⁹ Accordingly, a total of 430 questionnaires were distributed in this study. We collected 409 valid responses, for a valid response rate of 95.12%. The study was approved by the Ethics Committee of Shandong First Medical University (Approval No. R202512310537). This study was conducted in accordance with the declaration of Helsinki.

Survey Tools

General Data Questionnaire

The questionnaire, developed by the research team based on a comprehensive review of literature and group deliberations, encompassed sociodemographic data (such as gender, age, ethnicity, education level, occupation type, marital status, religious, and medical payment method) as well as disease-related information (including treatment status, disease duration, clinical stage, and pathological type).

The Chinese Version of Post-Traumatic Growth Inventory, C-PTGI

The Post-Traumatic Growth Inventory (PTGI) was originally developed by Tedeschi and Calhoun. The original scale consisted of 21 items across five dimensions: relating to others, new possibilities, personal strength, appreciation of life, and spiritual change.²⁰ In the present study, we utilized the Chinese version of the PTGI (C-PTGI), which was adapted and validated by Wang for use in Chinese populations who have experienced traumatic events. The C-PTGI consists of 20 items categorized into five dimensions (Life perspective, Personal Strength, Relating with Others, Self-transformation, and New Possibilities) and employs a 6-point Likert scale. Each of the 20 items is scored on a scale of 0 to 5, resulting in a total score ranging from 0 to 100. A higher score indicates a greater level of post-traumatic growth. The Cronbach’s α coefficient for the total C-PTGI scale is 0.9, while the Cronbach’s α coefficient for the subscales range from 0.611 to 0.874.²¹ The Cronbach’s α coefficient for this study is 0.912.

Perceived Social Support Scale, PSSS

The Perceived Social Support Scale (PSSS) was originally developed by Zimet¹⁴ in 1990 and later translated and adapted into Chinese by Jiang Qianjin²² in 1999. It comprises three dimensions—family support, friend support, and other support—with a total of 12 items. Each item uses a 7-point Likert scale, scored from 1 (Strongly Disagree) to 7 (Strongly Agree). Total scores range from 12 to 84; a higher score indicates a higher level of perceived social support among patients. The Cronbach’s α coefficient for this scale is 0.84. The Cronbach’s α coefficient for this study is 0.881.

Cognitive Appraisal Scale, CAS

Feng Shu¹⁵ developed the scale’s items and dimensions through interviews with cancer patients and a review of relevant literature based on Lazarus’s transactional model of stress in 2007. The scale consists of 16 items and is divided into four dimensions: uncertainty in illness, perceived disease threat, sense of responsibility for the disease, and positive appraisal. The scale employs a 5-point Likert scale, with each item scored from 1 (strongly disagree) to 5 (strongly agree). Higher scores indicate a more positive cognitive evaluation of cancer by the patient. The Cronbach’s α coefficient for the CAS total scale is 0.85. The Cronbach’s α coefficient for this study is 0.719.

Data Collection Method

Data were collected by two postgraduate nursing students. The researchers explained the purpose of the study and its significance to the patients, and emphasized the importance of maintaining confidentiality. After written informed consent was obtained, patients were instructed to complete the questionnaire independently based on their actual circumstances. The researchers provided assistance and conducted interviews when patients encountered difficulties in completing the questionnaire. The researchers immediately reviewed the completed questionnaires and asked patients to provide any missing information. Raw data were independently verified by two researchers, entered into Excel, and then cross-checked for accuracy.

Statistical Method

Latent profile analysis was conducted using Mplus 8.3. Model fit was evaluated using the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and sample-size adjusted BIC (aBIC), where lower values indicate better fit. The Lo-Mendell-Rubin adjusted likelihood ratio test (LMR) and the bootstrap likelihood ratio test (BLRT) were used to compare k-class models against k-1-class models, with significant p-values favoring the k-class model. Classification accuracy was assessed via entropy, with values >0.80 considered acceptable.

All other statistical analyses were performed using SPSS 26.0. Descriptive statistics were used to summarize demographic characteristics and scale scores. Categorical variables are presented as frequencies and percentages. Continuous variables are expressed as mean ± SD for normally distributed data, or median (IQR) for skewed data. Harman’s single-factor test with unrotated principal component extraction was applied to all items from the C-PTGI, PSSS, and CAS to assess common-method bias. Multinomial logistic regression was employed to examine the associations of perceived social support and cognitive appraisal with post-traumatic growth. A two-tailed p-value < 0.05 was considered statistically significant.

Results

Demographic Characteristics and Levels of Post-Traumatic Growth in Patients with Esophageal Cancer

This study included a total of 409 patients with esophageal cancer, and their demographic characteristics are presented in Table 1. The mean total score of posttraumatic growth among these participants was 58.65 ± 15.253.

Table 1.

General Data of Patients with Esophageal Cancer and Univariate Analysis of Latent Profiles of PTG [Case (Percentage, %)]

Variable	Group	C1(n=150)	C2(n=181)	C3(n=78)	χ2	P
Gender					36.573	<0.001
	Man	88(28.5)	154(49.8)	67(21.7)
	Woman	62(62.0)	27(27.0)	11(11.0)
Age					4.410	0.110
	18-65	86(33.7)	123(48.2)	46(18.0)
	≥65	64(41.6)	58(37.7)	32(20.8)
Ethnic					3.303	0.192
	Han	150(37.1)	178(44.1)	76(18.8)
	Hui	0(0.0)	3(60.0)	2(40.0)
Level of education					54.792	<0.001
	Junior High and below	81(54.0)	35(23.3)	34(22.7)
	High school and secondary specialized school	61(32.1)	97(51.1)	32(16.8)
	Junior college and above	8(11.6)	49(71.0)	12(17.4)
Occupation					35.899	<0.001
	Farmer	116(47.5)	84(34.4)	44(18.0)
	Official	12(15.0)	51(63.7)	17(21.3)
	Other	10(23.3)	23(53.5)	10(23.3)
	Worker	12(28.6)	23(54.8)	7(16.7)
Marital status					3.497	0.174
	Married	138(37.4)	165(44.7)	66(17.9)
	Unmarried	12(30.0)	16(40.0)	12(30.0)
Religious					3.303	0.192
	None	150(37.1)	178(44.1)	76(18.8)
	Islam	0(0.0)	3(60.0)	2(40.0)
Medical payment method					19.417	<0.001
	Publicly funded	55(27.4)	110(54.7)	36(17.9)
	Self-paid	95(45.7)	71(34.1)	42(20.2)
Treatment status					16.932	<0.001
	Chemoradiotherapy	111(43.9)	94(37.2)	48(19.0)
	Combined therapy	39(25.0)	87(55.8)	30(19.2)
Disease duration					3.882	0.693
	≤3	68(39.8)	69(40.4)	34(19.9)
	3.1-6	37(37.4)	46(46.5)	16(16.2)
	6.1-12	16(30.2)	28(52.8)	9(17.0)
	>12	29(33.7)	38(44.2)	19(22.1)
Clinical stage					4.469	0.346
	I	10(35.7)	12(42.9)	6(21.4)
	III	78(36.4)	88(41.1)	48(22.4)
	IV	62(37.1)	81(48.5)	24(14.4)
Pathological type					11.326	0.003
	Squamous cell carcinoma (SCC)	116(38.9)	137(46.0)	45(15.1)
	Non-SCC	34(30.6)	44(39.6)	33(29.7)

Abbreviations: SCC, squamous cell carcinoma; PTG, post-traumatic growth.

Latent Profile Analysis of Post-Traumatic Growth in Patients with Esophageal Cancer

We estimated and compared latent profile models with one to five classes (Table 2). Information criteria (AIC, BIC, aBIC) decreased monotonically with additional profiles, but Model 4 showed clear weaknesses: the LMR test failed to reach significance (p = 0.109), and one class comprised only 2.45% of the sample—too small for reliable interpretation or clinical utility. We therefore retained the three-class solution as the optimal balance of statistical fit and substantive meaning. Classification accuracy was strong, with correct assignment probabilities of 96.6%, 93.2%, and 90.8% for the respective classes. Figure 1 shows the three identified profiles, which have been named based on their overall growth levels. Profile 1 scored the lowest across all dimensions and is therefore designated as the low-growth group(C1). Profile 3 scored the highest across all dimensions and is designated as the high-growth group(C3). Profile 2 scored between these two profiles on all dimensions and is therefore designated as the moderate-growth group(C2).

Table 2.

Fitting Index for the Latent Profile Model of the PTG of Patients with Esophageal Cancer

Class Model	LL	AIC	BIC	aBIC	Entropy	P(BLRT)	P(LMR)	Class Proportion
1c	−2701.118	5422.237	5462.374	5430.642
2c	−2310.154	4652.308	4716.528	4665.757	0.859	<0.001	<0.001	50.12/49.88
3c	−2146.487	4336.974	4425.276	4355.466	0.891	<0.001	<0.001	36.68/44.25/19.07
4c	−2088.554	4233.108	4345.492	4256.643	0.914	<0.001	0.109	2.45/41.57/35.21/20.78
5c	−2030.417	4128.834	4265.301	4157.413	0.882	<0.001	0.197	2.42/24.87/18.78/16.87/37.07

Abbreviation: PTG, post-traumatic growth.

Figure 1.

Item-response mean scores for the 3-profile PTG model. The X-axis represents the PTG dimensions.

Univariate Analysis of Post-Traumatic Growth Latent Profiles in Patients with Esophageal Cancer

The mean total PTG scores differed significantly across the three profiles: low-growth (42.64 ± 7.10), moderate-growth (62.50 ± 5.97), and high-growth (80.50 ± 5.84) (F = 961.59, P < 0.001).

Significant between-group differences were also observed in several demographic and clinical characteristics, including gender, education level, occupation, medical payment method, treatment status, and pathological type (P < 0.05). Additionally, the profiles differed in psychosocial predictors, namely perceived social support, friend support, other support, uncertainty in illness, sense of responsibility for the disease, and positive appraisal (P < 0.05). Further details are presented in Tables 1 and 3.

Table 3.

Differences in Cognitive Appraisal and Perceived Social Support Among Latent Profiles of PTG

Variable		Low-Growth Group	Moderate-Growth Group	High-Growth Group	F	p
Perceived social support	Total score	58.67±9.474	62.11±10.721	69.65±8.173	32.092	0.000
	Family support	24.79±3.132	24.36±3.955	24.03±3.381	1.274	0.281
	Friend support	12.74±6.480	16.10±6.126	21.94±4.231	61.283	0.000
	Other support	21.15±3.320	21.64±3.553	23.69±2.947	15.364	0.000
Cognitive appraisal	Uncertainty in Illness	24.53±3.849	23.69±3.773	19.97±6.891	26.885	0.000
	Perceived Disease Threat	11.73±2.93	11.64±2.514	11.50±3.920	0.149	0.861
	Sense of responsibility for the disease	6.47±1.924	8.05±1.208	8.55±1.792	57.265	0.000
	Positive Appraisal	13.70±2.371	15.57±2.283	18.10±1.542	104.492	0.000

Abbreviation: PTG, post-traumatic growth.

Multivariate Analysis of Post-Traumatic Growth Latent Profiles in Patients with Esophageal Cancer

Variables significant in univariate analysis were included in the multinomial logistic regression model, using the high-growth group as the reference (Table 4). Higher scores for positive appraisal and friend support were associated with a greater likelihood of belonging to the high-growth group (OR < 1 for both low- and moderate-growth groups; p < 0.05). Conversely, higher uncertainty in illness showed a marginally significant association with the low-growth group (OR =1.094, p = 0.064). Patients with squamous cell carcinoma were more likely to fall into the low- or moderate-growth groups. Publicly funded patients were significantly associated with the moderate-growth group and showed a marginally significant trend for the low-growth group. Male patients were marginally less likely than females to be assigned to the low-growth group, with no significant difference observed for the moderate-growth group.

Table 4.

Multivariate Analysis of Latent Profiles in Esophageal Cancer Patients

	β	SE	Wald χ2	P	OR	95% CI
C1vsC3a
Uncertainty in Illness	0.090	0.049	3.424	0.064	1.094	0.995–1.204
Positive Appraisal	−0.716	0.121	34.836	0.000	0.489	0.385–0.620
Friend support	−0.194	0.048	16.205	0.000	0.824	0.749–0.905
Gender(man)	−1.183	0.606	3.813	0.051	0.306	0.093–1.004
Medical payment (publicly funded)	1.421	0.739	3.702	0.054	4.142	0.974–17.618
Pathological type (SCC)	0.941	0.476	3.910	0.048	2.563	1.008–6.516
C2vsC3a
Positive Appraisal	−0.639	0.112	32.287	0.000	0.528	0.423–0.658
Friend support	−0.118	0.044	7.042	0.008	0.889	0.815–0.970
Medical payment (publicly funded)	1.431	0.676	4.489	0.034	4.185	1.113–15.730
Pathological type (SCC)	1.306	0.411	10.084	0.001	3.692	1.649–8.269

Note: ^aC3 is the reference category.

Abbreviation: SCC, squamous cell carcinoma.

Discussion

This study employed latent profile analysis to identify three types of PTG among esophageal cancer patients: the low-growth group (36.68%), the moderate-growth group (44.25%), and the high-growth group (19.07%), confirmed the heterogeneity in the psychological adaptation process among esophageal cancer patients. The vast majority of esophageal cancer patients exhibited low-to-moderate levels of PTG, with only about one-fifth reaching high levels, suggesting significant room for improvement in clinical interventions. This distribution is consistent with the findings of previous studies using latent profile analysis in cancer populations, which also indicate that the majority of cancer patients exhibit low-to-moderate levels of PTG.^23,24 Post-traumatic growth in cancer patients carries multidimensional psychosocial significance: it serves as a psychological adaptation mechanism that helps patients reconstruct worldviews shattered by illness;⁸ as a positive psychological resource associated with better mental health and social functioning;¹⁰ and as a social process realized through deepened interpersonal connections and help-seeking. Understanding PTG enables a shift from disease-centered management toward whole-person care, facilitating meaningful life transformation amid illness. However, the majority of patients in this study did not achieve high levels of post-traumatic growth, suggesting that these psychosocial benefits remain unrealized for a substantial proportion of the population.

Different subgroups exhibit distinct psychological characteristics in this study. The low-growth group exhibited a relationship-isolation-type pattern characterized by a relatively prominent “relating with others” dimension but generally low scores across all other dimensions, suggesting limited development in each growth dimension. Taylor noted that seeking relationships is an innate human tendency to cope with stress;²⁵ however, without deep cognitive processing, social support is unlikely to be effectively transformed into internal psychological resources.^8,26 Consequently, although this group is able to perceive relational support, they have failed to integrate it into the construction of self-concept and life meaning. Consistent with this view, recent research also indicates that the mere presence of social support does not necessarily promote personal growth unless individuals engage in a reflective process of meaning-making;²⁷ The moderate-growth group exhibited a relationship-dependent moderate growth pattern characterized by the highest scores on the “relating with others” dimension. This group demonstrated moderate performance in “life perspective” and “personal strength” but scored relatively low on “new possibilities” and “self-transformation”, reflecting an adaptive strategy centered on relational support but marked by uneven development across growth dimensions. This pattern suggests that if individuals rely excessively on relational support while neglecting other pathways of growth, they may face adaptive vulnerability. This is consistent with the findings of Sun et al;²⁸ The high-growth group consistently scored high across all dimensions with minimal variation, indicating a balanced and integrated pattern of growth in which no single dimension predominated This pattern suggests that integration across growth dimensions—rather than reliance on a single pathway—is a hallmark of optimal post-traumatic growth outcomes.⁸

Positive appraisal was the strongest predictor distinguishing the high-growth group from the low- and moderate-growth groups, with higher scores significantly decreasing the likelihood of classification into either of the latter two groups. This finding is consistent with the social cognitive model of post-traumatic growth, which posits that an individual’s positive appraisal of a traumatic event helps reconstruct their system of meaning, thereby promoting a higher level of growth.⁸ Esophageal cancer patients often face prolonged treatment cycles and significant physical symptoms (such as dysphagia, pain, and weight loss), factors that may undermine patients’ sense of control over disease outcomes.^29,30 In this context, positive appraisal may play a dual role: on the one hand, it is directly associated with patients’ search for meaning in their illness;³¹ on the other hand, it is associated with reduced the psychological exhaustion caused by disease uncertainty,³² thereby helping patients cross the critical threshold from “moderate growth” to “high growth”.

Similarly, friend support also demonstrated an independent potentially protective association, suggesting that a strong social support network is a vital resource for resisting a state of low growth. Unlike family support, support from friends may provide a “non-obligatory” emotional resource—patients need not worry about being a burden to family members and can express vulnerability and fear more freely.³³ This support model is particularly important for esophageal cancer patients, as disease-related difficulties with eating and weight loss often lead to significant social embarrassment, and non-judgmental companionship from friends may effectively alleviate this pressure.

In terms of histological type, patients with squamous cell carcinoma were significantly less likely to belong to the high-growth group compared to patients with other types. In this study, squamous cell carcinoma accounted for 72.9% of cases, consistent with the epidemiological characteristics of esophageal cancer in China, where squamous cell carcinoma predominates. Patients with esophageal squamous cell carcinoma typically undergo long treatment cycles and experience a higher symptom burden, including dysphagia, pain, and malnutrition.³⁴ Persistent physical discomfort and treatment-related distress often trigger chronic psychological stress, which in turn may weaken their post-traumatic growth.³⁵ Compared with patients with adenocarcinoma or other less common subtypes, those with squamous cell carcinoma tend to have poorer functional status, necessitating greater psychological resource allocation to symptom management and thereby leaving less cognitive capacity for meaning-making and positive reconstruction.³⁶ It should be noted that these conclusions primarily apply to esophageal cancer patients with squamous cell carcinoma; caution is warranted when extrapolating these findings to patients with non-squamous cell carcinoma. Regarding medical payment, publicly funded patients were more likely to be classified into the moderate-growth group than the high-growth group (OR = 4.185, p = 0.034), with a similar trend for the low-growth group (OR = 4.142, p = 0.054). This unexpected pattern suggests that self-paid status may be associated with higher PTG classification, though wide confidence intervals—reflecting the small high-growth reference group (n = 78)—limit precise estimation and warrant caution. Tentatively, this resonates with the view that PTG arises from struggle with challenges rather than distress per se,⁸ as the financial burden of esophageal cancer may prompt life reevaluation and resource mobilization, aligning with challenge and threat theory.³⁷ Additionally, both gender and illness uncertainty showed marginal significance (p = 0.05–0.10) in low-growth group comparisons. Males (75.6% of the sample, consistent with the male-predominant epidemiology of esophageal cancer) appeared less likely than females to be classified into the low-growth group, though no significant differences emerged between the moderate- and high-growth groups (p > 0.05). Given the skewed gender distribution, this effect requires validation in more balanced samples. Higher illness uncertainty was associated with an increased likelihood of low-growth group membership. The uncertainties spanning diagnosis, treatment, and recovery—regarding whether patients can resume a normal diet, whether the tumor will recur, and whether additional radiotherapy or chemotherapy will be needed—may, from a subgroup perspective, leave patients “stuck” in a low-growth state, hindering active meaning-making.³⁸ As neither effect reached conventional significance, these findings remain exploratory and require replication in larger, more balanced cohorts. These findings further suggest that psychosocial and cognitive factors may help distinguish patients with different levels of post-traumatic growth, which could have implications for subsequent psychosocial care.

In addition, the identified PTG profiles may have potential implications for patient preference, treatment adherence, and satisfaction. Patients in the low-growth group might prefer more structured and repeated disease-related education and may require additional emotional support to sustain adherence to treatment regimens, as their heightened illness uncertainty and negative appraisal may undermine confidence in treatment decisions. In contrast, patients in the high-growth group might favor less intensive psychological follow-up and exhibit greater autonomy in care decision-making, which may align with their stronger internal coping resources and sense of personal meaning, potentially contributing to higher satisfaction with care. For the moderate-growth group, whose needs fall between these extremes, a flexible approach that balances informational support with autonomy may be most appropriate for optimizing both adherence and satisfaction.

In summary, the identification of distinct subgroups through latent profile analysis may inform personalized interventions. Evidence suggests that individuals on different post-traumatic growth trajectories benefit from differentiated intervention strategies.³⁹ Based on our research, our findings suggest that efforts might focus on patients with squamous cell carcinoma who receive little social support and have low positive appraisal. While the cross-sectional design precludes definitive conclusions, the three identified profiles may represent different stages or trajectories of psychological adaptation during the disease course. Whether patients shift between these profiles over time remains unknown and requires longitudinal follow-up. If confirmed longitudinally, these distinct PTG configurations could guide stage-matched interventions: For the low-growth group, interventions should focus on deepening cognitive processing of existing relational support to promote its internalization;^40,41 for the moderate-growth group, it may be beneficial not to rely solely on social support, but also to cultivate additional pathways of growth, such as personal meaning and new possibilities;^8,42 and for the high-growth group, maintenance and consolidation strategies may suffice. Beyond individual patient care, these findings carry broader psychosocial implications: for clinicians, the stratified profiles may serve as an actionable reference for early risk screening and individualized care coordination; for healthcare systems, adopting tiered psychosocial support strategies could optimize resource utilization and potentially reduce the broader societal burden associated with cancer-related psychological distress. Therefore, implementing profile-guided, tiered interventions can enhance the precision and effectiveness of psychosocial care for esophageal cancer patients, and provide a practical approach to improving patients’ health-related quality of life.

Limitations and Future Research

First, the cross-sectional design limits causal inference and, given PTG’s inherently temporal nature,^8,12 fails to capture its nonlinear evolution. Our single-time-point data cannot reveal how profiles shift across critical milestones—immediately post-surgery, completion of adjuvant therapy, or transition to survivorship. Notably, disease duration was not significantly associated with profile membership in our sample, which may indicate relative stability of PTG configurations across these stages, or simply that our broad categories, though clinically practical, missed the finer transitions where PTG actually shifts. Future longitudinal research should track these subgroups across more granular illness milestones to better model PTG trajectories and strengthen causal inferences.

Second, the sample was drawn from a single medical institution, which may limit generalizability. Future studies should recruit from multiple centers with larger and more diverse samples to enhance external validity.

Third, the data relied on patient self-reports, which are subject to recall bias and social desirability bias. We conducted Harman’s single-factor test to assess common-method bias; the first unrotated factor explained 26.91% of the total variance, below the 50% threshold, suggesting that common-method bias was not a severe concern in this study. Nevertheless, these other forms of reporting bias may still affect accuracy. Future research could combine self-report measures with clinician-rated assessments or qualitative interviews to cross-validate findings.

Fourth, we cannot rule out residual confounding from unmeasured variables. Although we assessed several key covariates, other theoretically relevant factors—such as personality traits and pre-diagnosis psychological functioning—were not included and may have influenced the observed associations. Future research should incorporate these variables to better isolate the unique contribution of disease-related factors to PTG.

Fifth, the wide 95% CI for medical payment method (OR = 4.142, 95% CI: 0.974–17.618) reflects limited precision due to the small high-growth reference group (C3, n = 78). We addressed this in a sensitivity analysis (Table S1) by re-specifying the reference category; the consistency of effect directions across specifications suggests that the core qualitative conclusions are robust, though these findings should still be interpreted as preliminary trends requiring validation in larger, more balanced samples.

Conclusion

This study reveals the heterogeneity of post-traumatic growth among esophageal cancer patients. Through latent profile analysis, patients with esophageal cancer and post-traumatic growth were classified into three groups: low-growth, moderate-growth, and high-growth. The moderate-growth group accounted for the largest proportion, while the high-growth group accounted for the smallest. These profiles were further distinguished by key predictors: higher positive appraisal and friend support scores increased the likelihood of high-growth group membership, offering concrete markers for patient stratification. This profile-based stratification may offer a preliminary reference for clinical practice. For clinicians, it could guide early identification of characteristic patterns and inform the future development of stratified intervention strategies. Particularly for patients with low post-traumatic growth, future intervention efforts might focus on transforming external relational support into internal security, pending direct evaluation. For the healthcare system and society at large, such stratification may facilitate more appropriate allocation of supportive care resources. These findings provide an empirical basis for future research to develop and test targeted interventions in clinical trials.

Disclosure

Ms Zixuan Cao reports support for the manuscript from Shandong Provincial Natural Science Foundation, during the conduct of the study. Ms Panpan Jia reports support for the manuscript from Shandong Provincial Natural Science Foundation, during the conduct of the study. Professor Aihua Zhang reports support for the manuscript from Shandong Provincial Natural Science Foundation, during the conduct of the study.

The authors report no other conflicts of interest in this work.