Trajectories and predictors of benefit finding among patients with colorectal cancer undergoing chemotherapy: A longitudinal observational study

Abstract

Objective

This study aimed to examine the longitudinal trajectories of benefit finding (BF) among patients with colorectal cancer undergoing chemotherapy and to identify predictors associated with these trajectories to inform tailored psychosocial interventions.

Methods

A total of 191 patients were recruited from three tertiary hospitals in Southwest China. Baseline data were collected before chemotherapy initiation (T0) using the General Information Questionnaire, the Brief Illness Perception Questionnaire, the Distress Disclosure Index, and the Benefit Finding Scale. BF levels were reassessed after the first (T1), third (T2), and sixth (T3) chemotherapy cycles. Growth mixture modeling was applied to identify distinct BF trajectories over time, followed by multivariate logistic regression to determine predictors of trajectory membership.

Results

Among 173 patients who completed all assessments, three BF trajectories were identified: moderate-to-increasing (11.6%), low-to-increasing (63.0%), and high-to-decreasing (25.4%). Marital status, place of residence, cancer stage, stoma history, baseline illness perception, and self-disclosure were significant predictors of trajectory classification (all P ​< ​0.05).

Conclusions

Patients with colorectal cancer exhibited heterogeneous patterns of benefit finding throughout chemotherapy. Marital status, residence, stoma history, illness perception, and self-disclosure emerged as key factors shaping these trajectories. Identifying distinct patient subgroups based on these characteristics may help guide personalized interventions to promote psychological adjustment during treatment.

Introduction

Colorectal cancer (CRC) is among the most prevalent and deadly malignancies of the digestive system worldwide.¹ Recent epidemiological data report approximately 1.93 million new CRC cases and 900,000 related deaths annually.¹ In China, an estimated 517,100 new cases and 240,000 deaths occur each year, placing CRC among the top five cancers in both incidence and mortality, and posing a significant public health challenge.² With ongoing advancements in medical technology and evolving therapeutic strategies, treatment options for CRC have become increasingly diverse. Chemotherapy, in particular, remains one of the most widely utilized and effective treatment modalities, playing a critical role in prolonging patient survival.³ Commonly employed regimens include capecitabine plus oxaliplatin (XELOX), 5-fluorouracil with oxaliplatin and folinic acid (FOLFOX), and 5-fluorouracil with irinotecan and folinic acid (FOLFIRI).³ While the duration of chemotherapy cycles varies among regimens, a standard cycle typically lasts 21 days, with a full course generally comprising 6–8 cycles over 4.5–6 months.⁴ Despite its efficacy, chemotherapy is often associated with adverse effects such as nausea, fatigue, and immunosuppression, which not only increase patients’ physical burden but also heighten the risk of anxiety and depression, thereby substantially compromising quality of life.^5,6

In recent years, with the rise of positive psychology, increasing attention has been directed not only to the physiological treatment of cancer patients but also to their psychological adaptation. Benefit Finding (BF), an indicator of positive psychological adjustment, refers to a cognitive coping strategy that individuals employ in response to threatening events, such as cancer, manifested through both cognitive and behavioral mechanisms.⁷ Importantly, BF is a dynamic process that may change continuously over the course of the disease and across different treatment stages.⁸ Conley et al. conducted a nearly two-year follow-up of breast cancer survivors and observed an upward trend in BF levels over time.⁹ In contrast, Bi et al. reported a slight decline in BF scores among breast cancer patients since diagnosis.¹⁰ Similarly, Zimmaro et al. observed dynamic changes in BF among CRC patients as disease duration increased.¹¹ Wang conducted five follow-up assessments over six months in breast cancer patients and identified four distinct trajectories: high-stable, low-stable, increasing, and declining,¹² highlighting both the dynamic nature of BF and individual differences. Compared with these studies, research on BF in CRC patients has largely relied on cross-sectional designs, providing limited longitudinal evidence and lacking systematic investigations of BF changes during chemotherapy. Therefore, longitudinal research is needed to delineate the developmental trajectories of BF in CRC patients throughout the chemotherapy process.

Several studies have examined the influence of sociodemographic and clinical factors on the level and trajectory of BF in cancer patients. Gender has been identified as a predictor of BF in CRC patients, with females exhibiting higher BF levels than males.¹¹ Additionally, age and educational attainment have been shown to predict changes in BF among prostate cancer survivors.¹³ Disease stage, number of chemotherapy cycles, and symptom burden have also been closely associated with both the level and evolution of BF.¹⁴ Accordingly, the present study aims to further investigate whether sociodemographic and clinical characteristics can predict the developmental trajectories of BF in CRC patients undergoing chemotherapy.

In addition to sociodemographic and clinical characteristics, cognitive and behavioral coping factors play a pivotal role in positive psychological adaptation. Illness perception reflects patients’ cognitive appraisal of their disease and their perceived control over it,¹⁵ and has been shown to be significantly associated with BF levels in cancer survivors.¹⁶ Notably, illness perception may influence not only BF at a single time point but also its longitudinal trajectory. Patients with positive illness perceptions tend to exhibit gradually increasing or high-stable BF trajectories over time, whereas negative perceptions may result in low or declining trajectories.¹³ Similarly, self-disclosure, a key mechanism for emotional expression and psychological coping, has been shown to positively predict BF. For instance, self-disclosure among thyroid cancer survivors was found to enhance BF levels.¹⁷ By expressing and sharing personal experiences, patients can alleviate negative emotions and facilitate cognitive restructuring, thereby improving their perception of positive meaning.¹⁸ Guided self-disclosure interventions have also been shown to promote BF in breast cancer patients.¹⁹ Consequently, it is hypothesized that patients with higher levels of self-disclosure are more likely to follow increasing or high-stable BF trajectories. Therefore, this study considers illness perception and self-disclosure as potential predictors of the developmental trajectories of BF in CRC patients undergoing chemotherapy.

The Stress and Coping Theory, proposed by Lazarus, encompasses stressors, cognitive appraisal, coping strategies, and stress responses.²⁰ According to this theory, individuals evaluate stressors cognitively and adopt corresponding coping strategies, leading to either positive or negative psychological and behavioral outcomes. In longitudinal studies, the theory explains why individuals exposed to the same stressor may exhibit distinct BF trajectories due to differences in cognitive appraisal and coping behaviors. In the context of chemotherapy for CRC patients, the physical and psychological burdens imposed by cancer and its treatment constitute stressors. Illness perception reflects cognitive appraisal, representing patients’ subjective evaluation of disease threat and controllability, whereas self-disclosure serves as a behavioral coping strategy aimed at obtaining emotional support and informational feedback. Although these factors occur at different stages of the stress process, both can directly influence the formation and progression of BF. Previous studies have demonstrated that interventions targeting cognitive restructuring or self-disclosure can enhance BF levels, highlighting their clinical applicability.^18,21 Based on this, the theoretical framework of the present study is illustrated in Fig. 1, positing that illness perception and self-disclosure may function as potential predictors of BF trajectories.

Fig. 1.

The theoretical framework of the study.

Methods

Study design and participants

This study is a prospective longitudinal study. From July 2024 to March 2025, CRC patients who received chemotherapy in the oncology departments of three tertiary hospitals in southwest China were selected by convenience sampling. Inclusion criteria were as follows: (1) histopathologically confirmed diagnosis of CRC; (2) age ≥ 18 years; (3) first-time hospitalization for chemotherapy with a planned regimen of 6–8 cycles of chemotherapy; (4) awareness of their diagnosis, condition and treatment of diseases; (5) normal verbal expression and comprehension abilities; and (6) provision of informed consent and voluntary participation. Exclusion criteria included: (1) with severe cardiac, hepatic, or renal insufficiency or other major comorbidities; (2) history of mental illness that impairs cooperation. Shedding criteria were: (1) experienced a change in treatment regimen due to disease progression or physical intolerance; or (2) voluntarily discontinued participation or were lost to follow-up, including due to death.

Sample size calculation

While the trajectory analysis ideally requires a larger sample, the current sample was calculated based on a single-group repeated-measures design.²² An average correlation coefficient of r ​= ​0.5, an effect size of f ​= ​0.14 (indicating a small effect), and a significance level of α ​= ​0.05 were used as parameters. To achieve a statistical power of 1-β ​= ​0.8, the minimum required sample size was 142. After accounting for a potential 10% attrition rate, the final estimated minimum sample size was 157 participants.

Selection of study time points

The selection of measurement time points in this study was informed by clinical relevance and prior research experience. Before the first chemotherapy session, CRC patients may undergo significant psychological changes due to disease diagnosis or surgery, which can disrupt their lifestyle and self-image.²³ At the first chemotherapy session (early stage of chemotherapy), patients have a clearer understanding of their disease and a confirmed treatment plan, which may lead to notable psychological adjustments.²⁴ By the third chemotherapy session (mid-stage of chemotherapy), most patients have accepted their cancer diagnosis and completed more than half of their chemotherapy, likely resulting in further psychological adaptation.²⁴ At the sixth chemotherapy session (late stage of chemotherapy), after multiple treatments, patients gradually acclimate to the side effects and discomfort of chemotherapy, exhibiting greater acceptance and cooperation with treatment, which may correspond with changes in their BF levels.²⁵

Drawing on prior longitudinal studies of CRC patients’ psychological health24, 25, 26, 27 and considering the clinical characteristics of CRC treatment, this study ultimately selected the time points before the first chemotherapy session and after the first, third, and sixth sessions, thereby comprehensively capturing changes in BF throughout the chemotherapy process.

Research instruments

Sociodemographic and disease-related information

Sociodemographic data included age, gender, marital status, education level, working status, place of residence, caregivers, per capita monthly household income, relationship with family members, and medical payment method. Disease-related information included cancer stage, treatment methods, number of comorbidities, history of surgical, and history of stoma.

Brief illness perception questionnaire

Illness perception was assessed using the Brief Illness Perception Questionnaire (BIPQ), originally developed by Broadbent et al.²⁸ and cross-culturally adapted into Chinese by Mei et al.²⁹ The questionnaire comprises nine items across three dimensions: cognitive, emotional, and comprehension. The first eight items are rated on a scale from 0 to 10, with items 3, 4, and 7 reverse-scored. The ninth item is open-ended and not included in the total score. The total score ranges from 0 to 80, with higher scores indicating a more negative perception of illness. In this study, only the first eight items were used, and the Cronbach's α coefficient for the scale was 0.826.

Distress disclosure index

Distress disclosure was measured using the Distress Disclosure Index (DDI), originally developed by Kahn et al.³⁰ and subsequently validated in Chinese populations by Li.³¹ The scale consists of 12 items rated on a 5-point Likert scale, where 1 represents “strongly disagree” and 5 represents “strongly agree”. Items 2, 4, 5, 8, 9, and 10 are reverse-scored. The total score ranges from 12 to 60, with higher scores indicating greater levels of self-disclosure. In this study, the Cronbach's α coefficient for the scale was 0.898.

Benefit finding scale

In this study, the Chinese version of the Benefit Finding Scale (BFS) adapted by Weaver et al.³² and translated by Liu et al.³³ was used. The scale consists of 22 items across six dimensions: acceptance, family relationships, worldview, personal growth, social relationships, and health behaviors. Each item is rated on a five-point Likert scale, where 1 indicates “not at all” and 5 indicates “very much”. The total score ranges from 22 to 110, with higher scores reflecting greater levels of benefit finding. In this study, the Cronbach's α coefficient for the scale was 0.894.

Data collection

In this study, patients who met the inclusion criteria were investigated at four time points: before the first chemotherapy (T0), after the first chemotherapy (T1), after the third chemotherapy (T2), and after the sixth chemotherapy (T3). Chemotherapy intervals were relatively consistent across patients, and questionnaires were administered one day prior to chemotherapy and 2–3 days following the first, third, and sixth sessions to ensure comparability of measurement points. Questionnaires were completed in quiet wards or conference rooms to minimize distractions. Prior to data collection, all researchers received standardized training to ensure a scientific and consistent data collection process. Investigators then explained the study's purpose, procedures, and follow-up arrangements in detail to eligible patients. After providing informed consent, patients' contact information was recorded for follow-up and data collection.

At T0, clinical characteristics were extracted from electronic medical records, while sociodemographic information, illness perception, self-disclosure, and BF scores were obtained via face-to-face paper questionnaires, with an average completion time of 15–20 minutes. At T1-T3, only BF scores were collected using a combination of face-to-face questionnaires and telephone follow-up, with an average completion time of 10–15 minutes.

Data analysis

SPSS 26.0 and Mplus 8.3 software were used for data analysis. Initially, normality tests, descriptive statistics, and between-group comparisons were performed in SPSS 26.0. The overall temporal trend of BF was then assessed using repeated measures ANOVA or the Friedman test. Subsequently, a growth mixture model (GMM) was constructed in Mplus 8.3 to model BF across the four time points and to identify heterogeneous trajectories among individuals. Baseline BF (T0) was included as a covariate to control for baseline differences in trajectory analysis. Model fit was evaluated using several indicators: Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), adjusted BIC (aBIC), entropy, the Lo-Mendell–Rubin test (LMR), and the Bootstrap Likelihood Ratio Test (BLRT). Lower values of AIC, BIC, and aBIC indicate better model fit, while entropy values closer to 1 suggest higher classification accuracy.³⁴ If the P values for LMR and BLRT are less than 0.05, the model with k classes is considered superior to the model with k-1 classes.³⁵ The optimal trajectory model was selected based on a comprehensive evaluation of fit indices. Finally, univariate logistic regression was conducted in SPSS 26.0 to screen for significant variables, which were then entered into a multivariate logistic regression model to identify factors associated with different BF trajectory classes. The difference was statistically significant, with P ​< ​0.05.

Results

Patient characteristics

A total of 191 patients were enrolled in the study at baseline (T0). At the first follow-up (T1), 3 patients refused to participate. At the second follow-up (T2), 188 patients were contacted; among them, 5 refused to participate, 4 experienced changes in treatment regimen, and 2 were lost to follow-up due to loss of contact. At the final follow-up (T3), 177 patients were assessed, including 3 who refused the investigation and 1 who had died. In total, 18 patients were lost to follow-up, with a total loss rate of 9.42%. A total of 173 patients completed all follow-up assessments. The detailed enrollment and follow-up process is illustrated in Fig. 2.

Fig. 2.

The detailed process diagram.

Supplementary Table S1 presents the sociodemographic and clinical characteristics of the participants at T0. The mean age of patients ranged from 60.2 ​± ​10.6 years. 64.4% of the participants were male, almost all of the participants were married (91.1%), and only 10.5% of the participants had a university education level or above. In addition, nearly half of the participants (44.0%) had a stage IV cancer, more than half had a history of surgery (67.5%), and 27.2% had a stoma. There was no significant difference between the baseline data of the follow-up group and the loss-to-follow group (P ​> ​0.05), as shown in Table 1.

Table 1.

Comparison of baseline characteristics between follow-up and lost-to-follow-up groups among CRC patients undergoing chemotherapy (N ​= ​191).

Variables	Follow-up group (n ​= ​173)	Lost-to follow-up group (n ​= ​18)	χ2/t/z	P
Age (year), mean (SD)	60.10 (11.05)	60.89 (5.10)	−0.539b	0.593
Sex
Male	111 (64.2)	12 (66.7)	0.045a	0.833
Female	62 (35.8)	6 (33.3)
Marital status
Married	158 (91.3)	16 (88.9)	0.264a	0.607
Unmarried/Divorced/Widowed	15 (8.7)	2 (11.1)
Education level
Elementary school and below	74 (42.8)	10 (55.6)	3.882a	0.275
Junior high school	66 (38.2)	7 (38.9)
High school	14 (8.1)	0 (0.0)
University and above	19 (10.9)	1 (5.5)
Working status
Farmer	30 (17.3)	4 (22.2)	3.163a	0.367
Employed	27 (15.6)	1 (5.6)
Unemployed	46 (26.6)	3 (16.7)
Retired	70 (40.5)	10 (55.5)
Place of residence
City	114 (65.9)	10 (55.6)	0.326a	0.568
Towns or rural areas	59 (34.1)	8 (44.4)
Caregivers
None	16 (9.2)	2 (11.1)	4.855a	0.302
Spouse	94 (54.3)	6 (33.3)
Children	57 (32.9)	10 (55.6)
Parents	3 (1.7)	0 (0.0)
Siblings	3 (1.7)	0 (0.0)
Per capita monthly household income, USD
< ​415	27 (15.6)	3 (16.7)	0.723a	0.697
415-692	65 (37.6)	5 (27.8)
> ​692	81 (46.8)	10 (55.5)
Relationship with family members
Positive	102 (58.9)	9 (50.0)	1.855a	0.395
Moderate	66 (38.2)	9 (50.0)
Poor	5 (2.9)	0 (0.0)
Medical payment method
SPCI	3 (1.7)	0 (0.0)	5.570a	0.234
UEBMI	55 (31.8)	7 (38.9)
URBMI	14 (8.1)	0 (0.0)
NCMS	92 (53.2)	11 (61.1)
CI	9 (5.2)	0 (0.0)
Cancer stage
II Or below	25 (14.5)	5 (27.8)	1.835a	0.399
III	68 (39.3)	9 (50.0)
IV	80 (46.2)	4 (22.2)
Treatment methods
Chemotherapy	80 (46.2)	6 (33.3)	2.720a	0.437
Chemotherapy ​+ ​radiotherapy	39 (22.5)	7 (38.9)
Chemotherapy ​+ ​targeted therapy	30 (17.3)	2 (11.1)
Chemotherapy ​+ ​immunotherapy	24 (13.9)	3 (16.7)
No. of comorbidities
None	114 (65.9)	10 (55.6)	0.832a	0.660
One	47 (27.2)	6 (33.3)
Two or more	12 (6.9)	2 (11.1)
History of surgery
No	58 (33.5)	4 (22.2)	0.950a	0.330
Yes	115 (66.5)	14 (77.8)
History of stoma
No	123 (71.1)	16 (88.9)	0.039a	0.843
Yes	50 (28.9)	2 (11.1)
Illness perception [mean (SD)]	51.09 (10.61)	52.67 (4.31)	−1.218b	0.230
Self-disclosure [M (P25, P75)]	31.00 (20.00, 48.00)	33.00 (29.75, 39.25)	−0.139c	0.889
Benefit finding [M (P25, P75)]	36.00 (27.00, 73.00)	43.00 (40.00, 58.75)	−1.410c	0.159

a, the chi-square test; b, the t test of two independent samples; c, the Mann–Whitney U test; SPCI, Self-Payment or Commercial Insurance; NCMS, New Rural Cooperative Medical system; URBMI, Urban Residents' Basic Medical Insurance; UEBMI, Urban Employee Basic Medical Insurance; CI, Commercial Insurance; SD, Standard deviation; CRC, Colorectal cancer; M, median.

BF scores and their temporal changes in CRC patients undergoing chemotherapy

Normality tests indicated that total BF scores and dimension-specific scores at the four time points (T0-T3) were not normally distributed; therefore, the Friedman test was used. The analysis revealed significant differences in total and dimension-specific BF scores across all time points (P ​< ​0.001). Pairwise comparisons showed no significant change in BF scores between T0 and T1. However, as chemotherapy progressed, both total and dimension-specific BF scores exhibited a significant upward trend (P ​< ​0.05). Detailed results are presented in Table 2.

Table 2.

BF and dimension-specific scores of CRC patients undergoing chemotherapy from T0 to T3 (N ​= ​173).

Variables	T0	T1	T2	T3	χ2	P	Multiple comparisons
BF	36.00 (27.00, 73.00)	45.00 (33.00, 65.00)	54.00 (41.50, 64.50)	56.00 (49.00, 66.00)	102.306	< 0.001	b∗∗c∗∗d∗∗e∗∗f∗
Acceptance	6.00 (3.00, 10.50)	6.00 (5.00, 9.00)	8.00 (6.00, 9.00)	9.00 (7.00, 9.00)	74.969	< 0.001	b∗∗c∗∗d∗∗e∗∗f∗
Family relationships	3.00 (2.00, 6.00)	4.00 (2.00, 6.00)	5.00 (3.00, 6.00)	5.00 (4.00, 6.00)	24.614	< 0.001	b∗c∗∗e∗
Worldview	6.00 (5.00, 12.50)	8.00 (6.00, 12.00)	9.00 (8.00, 12.00)	10.00 (9.00, 12.00)	88.106	< 0.001	b∗∗c∗∗d∗∗e∗∗
Personal growth	11.00 (7.00, 23.00)	14.00 (10.00, 21.00)	17.00 (12.00, 21.00)	18.00 (15.00, 21.00)	76.586	< 0.001	b∗∗c∗∗d∗e∗∗
Social relationships	5.00 (4.00, 9.50)	6.00 (4.00, 9.00)	7.00 (5.00, 8.00)	8.00 (6.00, 9.00)	50.303	< 0.001	b∗c∗∗e∗∗f∗
Health behaviors	6.00 (4.00, 9.00)	7.00 (5.00, 9.00)	8.00 (7.00, 9.00)	8.00 (7.00, 10.00)	56.596	< 0.001	b∗∗c∗∗d∗e∗∗

b, significant difference between T0 and T2; c, significant difference between T0 and T3; d, significant difference between T1 and T2; e, significant difference between T1 and T3; f, significant difference between T2 and T3. ∗P ​< ​0.05; ∗∗P ​< ​0.01. CRC, colorectal cancer; BF, benefit finding.

Identification of BF trajectories in CRC patients undergoing chemotherapy

Fitting models for 1 to 5 trajectory category groups, see Table 3. The results showed that as the number of categories increased, the AIC, BIC, and aBIC values gradually decreased; when the number of categories reached 2 and 3, the P-values of the LMR and BLRT tests were statistically significant, but the Entropy value for category 2 (0.947) was less than that for category 3 (0.996), indicating that the classification accuracy was higher for 3 categories, so the 2-category group was discarded, and the 3-category group was established as the optimal model. After adjusting for baseline BF, the trajectories of the three classes remained significant.

Table 3.

Growth mixture model fitting indices for BF trajectories in CRC patients undergoing chemotherapy (N ​= ​173).

Class	AIC	BIC	aBIC	Entropy	LMR (P)	BLRT (P)	Profile prevalence (%)
1	5167.774	5196.154	5167.655	–	–	–	1
2	5051.832	5099.131	5051.633	0.947	< 0.001	< 0.001	0.382/0.618
3	5019.485	5085.704	5019.207	0.996	0.035	< 0.001	0.116/0.630/0.254
4	4998.472	5083.611	4998.114	0.856	0.382	0.128	0.133/0.098/0.520/0.249
5	4983.787	5087.846	4983.349	0.866	0.240	0.600	0.295/0.098/0.335/0.023/0.249

AIC, Akaike information criterion; BIC, Bayesian information criterion; aBIC, adjusted BIC; LMR, Lo–Mendell–Rubin likelihood ratio test; BLRT, bootstrap likelihood ratio test; CRC, colorectal cancer; BF, benefit finding.

Plot the change trajectory of the 3-category model, as shown in Fig. 3. According to the initial level and change characteristics of the patients, the groups were named: (1) Moderate-to-increasing BF group (C1), in which patients had moderate BF levels at T0 and exhibited a significant gradual increase during follow-up (slope ​= ​1.953, P ​< ​0.01), comprising 20 cases (11.6%); (2) Low-to-increasing BF group (C2), the level of BF was lower at T0 and continued to increase during the follow-up period (slope ​= ​7.324, P ​< ​0.001), with a total of 109 cases (63.0%); (3) High-to-decreasing BF group (C3), the level of BF was higher at T0, and the trend continued to decrease over time (slope ​= ​−7.648, P ​< ​0.001), and finally reached the moderate level, with a total of 44 patients (25.4%).

Fig. 3.

BF score at each time point based on latent class membership. BF, benefit finding.

Univariate analysis of BF trajectory subgroups in CRC patients undergoing chemotherapy

Univariate analysis revealed significant differences in marital status (χ² ​= ​29.290, P ​< ​0.001), place of residence (χ² ​= ​11.068, P ​< ​0.01), cancer stage (χ² ​= ​20.234, P ​< ​0.001), history of stoma (χ² ​= ​50.798, P ​< ​0.001), illness perception score (F ​= ​30.718, P ​< ​0.001), and self-disclosure score (F ​= ​9.242, P ​< ​0.001). See Table 4.

Table 4.

Univariate analysis of potential categories of BF trajectories in CRC patients undergoing chemotherapy (N ​= ​173).

Variables	Class 1 n (%)	Class 2 n (%)	Class 3 n (%)	Statistics	P
Sex
Male	14 (70.0)	72 (66.1)	25 (56.8)	1.498a	0.473
Female	6 (30.0)	37 (33.9)	19 (43.2)
Marital status
Married	19 (95.0)	108 (99.1)	31 (70.5)	29.290a	< 0.001∗∗
Unmarried/Divorced/Widowed	1 (5.0)	1 (0.9)	13 (29.5)
Education level
Elementary school and below	6 (30.0)	49 (45.0)	19 (43.2)	2.725a	0.843
Junior high school	9 (45.0)	41 (37.6)	16 (36.4)
High school	3 (15.0)	8 (7.3)	3 (6.8)
University and above	2 (10.0)	11 (10.1)	6 (13.6)
Working status
Farmer	6 (30.0)	20 (18.3)	4 (9.1)	7.370a	0.288
Employed	2 (10.0)	17 (15.6)	8 (18.2)
Unemployed	7 (35.0)	29 (26.6)	10 (22.7)
Retired	5 (25.0)	43 (39.4)	22 (50.0)
Place of residence
City	8 (40.0)	70 (64.2)	36 (81.8)	11.068a	< 0.01∗
Towns or rural areas	12 (60.0)	39 (35.8)	8 (18.2)
Caregivers
None	0 (0.0)	14 (12.8)	2 (4.5)	11.464a	0.177
Spouse	13 (65.0)	55 (50.5)	26 (59.1)
Children	6 (30.0)	36 (33.0)	15 (34.1)
Parents	0 (0.0)	3 (2.8)	0 (0.0)
Siblings	1 (5.0)	1 (0.9)	1 (2.3)
Per capita monthly household income (USD)
< 415	2 (10.0)	19 (17.4)	6 (13.7)	3.881a	0.422
415–692	6 (30.0)	45 (41.3)	14 (31.8)
> 692	12 (60.0)	45 (41.3)	24 (54.5)
Relationship with family members
Positive	13 (65.0)	61 (56.0)	28 (63.6)	2.266a	0.687
Moderate	7 (35.0)	44 (40.4)	15 (34.1)
Poor	0 (0.0)	4 (3.7)	1 (2.3)
Medical payment method
SPCI	0 (0.0)	3 (2.8)	0 (0.0)	14.746a	0.064
UEBMI	6 (30.0)	34 (31.2)	15 (34.1)
URBMI	2 (10.0)	12 (11.0)	0 (0.0)
NCMS	12 (60.0)	55 (50.4)	25 (56.8)
CI	0 (0.0)	5 (4.6)	4 (9.1)
Cancer stage
II or below	8 (40.0)	6 (5.5)	11 (25.0)	20.234a	< 0.001∗∗
Ⅲ	5 (25.0)	47 (43.1)	16 (36.4)
Ⅳ	7 (35.0)	56 (51.4)	17 (38.6)
Treatment methods
Chemotherapy	8 (40.0)	54 (49.5)	18 (40.9)	2.545a	0.863
Chemotherapy ​+ ​radiotherapy	5 (25.0)	22 (20.2)	12 (27.3)
Chemotherapy ​+ ​targeted therapy	3 (15.0)	20 (18.3)	7 (15.9)
Chemotherapy ​+ ​immunotherapy	4 (20.0)	13 (11.9)	7 (15.9)
No. of comorbidities
None	13 (65.0)	76 (69.7)	25 (56.8)	2.661a	0.616
One	5 (25.0)	27 (24.8)	15 (34.1)
Two or more	2 (10.0)	6 (5.5)	4 (9.1)
History of surgical
No	6 (30.0)	40 (36.7)	12 (27.3)	1.375a	0.503
Yes	14 (70.0)	69 (63.3)	32 (72.6)
History of stoma
No	15 (75.0)	95 (87.2)	13 (29.5)	50.798a	< 0.001∗∗
Yes	5 (25.0)	14 (12.8)	31 (70.5)
Age, mean (SD)	58.2 (12.0)	60.1 (11.2)	61.1 (10.4)	0.471b	0.625
Illness perception, mean (SD)	46.6 (8.8)	55.2 (9.3)	43.0 (9.0)	30.718b	< 0.001∗∗
Self-disclosure, mean (SD)	37.2 (15.0)	29.7 (14.8)	40.4 (13.7)	9.242b	< 0.001∗∗

a, the chi-square test; b, the one-way ANOVA; SPCI, Self-Payment or Commercial Insurance; NCMS, New Rural Cooperative Medical system; URBMI, Urban Residents' Basic Medical Insurance; UEBMI, Urban Employee Basic Medical Insurance; CI, Commercial Insurance; CRC, Colorectal cancer; BF, benefit finding; Class 1: Moderate-to-increasing BF group; Class 2: Low-to-increasing BF group; Class 3: High-to-decreasing BF group; ∗P ​< ​0.01; ∗∗P ​< ​0.001.

Multivariate analysis of factors influencing BF trajectories in CRC patients undergoing chemotherapy

Collinearity diagnostics were performed on variables that were statistically significant in the univariate analysis, and the results indicated no multicollinearity (VIF < 5). A multivariate logistic regression analysis was subsequently conducted, using the significant variables from the univariate analysis as independent variables and BF trajectory classification as the dependent variable. The original scores for baseline illness perception and self-disclosure were entered directly, while the remaining variables were coded as described in Supplementary Table S2.

The results indicated that, compared with the “Low-to-increasing BF group,” married patients (OR ​= ​0.042, P ​= ​0.015), urban residents (OR ​= ​0.270, P ​= ​0.033), patients without a stoma (OR ​= ​0.178, P ​= ​0.031), and those with higher illness perception scores (OR ​= ​0.888, P ​= ​0.001) were less likely to belong to other trajectory groups, suggesting a stronger tendency to remain in the “Low-to-increasing BF group.” Patients with cancer stage II or lower were more likely to be classified into the “Moderate-to-increasing BF group” (OR ​= ​11.101, P ​= ​0.003) or the “High-to-decreasing BF group” (OR ​= ​6.226, P ​= ​0.048). In addition, patients with higher self-disclosure scores had a greater likelihood of being classified into the “High-to-decreasing BF group” (OR ​= ​1.053, P ​= ​0.017). Detailed results are presented in Table 5.

Table 5.

Multivariate analysis of potential categories of BF trajectories in CRC patients undergoing chemotherapy (N ​= ​173).

Variables	Class 1 VS Class 2				Class 3 VS Class 2
	β	P	OR	95% CI	β	P	OR	95% CI
Marital status (Reference: Unmarried/Divorced/Widowed)
Married	−1.760	0.275	0.172	0.007, 4.051	−3.164	0.015∗	0.042	0.003, 0.539
Place of residence (Reference: Towns or rural areas)
City	−1.310	0.033∗	0.270	0.081, 0.899	0.205	0.780	1.227	0.292, 5.160
Cancer stage (Reference: IV)
II or below	2.407	0.003∗∗	11.101	2.259, 54.539	1.829	0.048∗	6.226	1.020, 37.996
Ⅲ	−0.321	0.646	0.726	0.185, 2.851	−0.443	0.502	0.642	0.176, 2.340
History of stoma (Reference: Yes)
No	−1.727	0.031∗	0.178	0.037, 0.855	−4.054	< 0.001∗∗∗	0.017	0.004, 0.079
Illness perception score	−0.119	0.001∗∗	0.888	0.827, 0.954	−0.188	< 0.001∗∗∗	0.829	0.767, 0.896
Self-disclosure score	0.028	0.180	1.028	0.987, 1.070	0.052	0.017∗	1.053	1.009, 1.099

CRC, colorectal cancer; BF, benefit finding; OR, odds ratio; CI, confidence interval. Class 1: Moderate-to-increasing BF group; Class 2: Low-to-increasing BF group; Class 3: High-to-decreasing BF group; ∗P ​< ​0.05; ∗∗P ​< ​0.01; ∗∗∗P ​< ​0.001.

Discussion

Main findings

The results of this study indicated that patients' BF levels increased gradually over time, which contrasts with the findings of Conley et al.⁹ This inconsistency may stem from differences in study populations, follow-up durations, and measurement tools, further supporting the notion of individual variability in BF development. GMM identified three distinct BF trajectories among CRC patients receiving chemotherapy: the “Low-to-increasing BF group” (63.0%), the “High-to-decreasing BF group” (25.4%), and the “Moderate-to-increasing BF group” (11.6%). Previous research has similarly shown heterogeneous developmental patterns of BF in cancer patients.¹² For instance, Zhu et al. categorized BF trajectories into high level-stable, extremely low level-slow increasing, low level-slow increasing, low level-rapidly increasing, and moderate level-stable groups.³⁶ Collectively, these findings highlight the marked heterogeneity in patients’ psychological adaptation and provide empirical evidence for understanding the mechanisms underlying diverse BF trajectories.

In this study, the identified trajectories corresponded closely to the core processes of stress and coping theory. Patients in the “Low-to-increasing BF group” exhibited low initial BF levels at the onset of chemotherapy, likely due to heightened perceptions of illness threat, treatment-related anxiety, or limited psychological resources. Nevertheless, through cognitive reappraisal, positive emotional experiences, and social support, these patients gradually achieved psychological adaptation,³⁷ leading to an upward BF trajectory. This finding is consistent with stress and coping theory, which posits that external support and positive cognition buffer stress and facilitate psychological recovery.²⁰ Accordingly, the early stage of chemotherapy represents a critical intervention window. Psychological interventions such as meaning-centered therapy, cognitive behavioral therapy, or narrative therapy³⁸ may be employed to stimulate positive cognition, enhance resilience, and strengthen self-efficacy, thereby accelerating and consolidating the upward trajectory. In contrast, patients in the “High-to-decreasing BF group” displayed initially high BF levels, possibly due to sufficient early psychological resources or strong social support. However, as treatment-related side effects accumulated, psychological resources were depleted and external support diminished, resulting in a decline in positive psychological adaptation³⁹ and a downward trajectory. For these patients, the mid-to-late stages of chemotherapy constitute a key intervention window. Priority should be given to the management of symptoms such as nausea, vomiting, and fatigue, supplemented with sustained psychological support and aerobic exercise therapy⁴⁰ to maintain positive adaptation. Patients in the “Moderate-to-increasing BF group”, although the smallest in proportion, demonstrated relatively stable and gradual improvements in BF during chemotherapy. This pattern reflects a moderate pace of psychological adaptation, potentially due to balanced illness threat appraisal, adequate coping strategies, and the presence of a supportive social network, which enable gradual adjustment through secondary appraisal and coping. For such patients, intensive psychological interventions may not be necessary; rather, regular follow-up, low-intensity mindfulness training, or online psychological support may be more appropriate to consolidate their slow but steady trajectory.⁴¹ In summary, health care providers should conduct stage-specific psychological assessments and deliver tailored interventions according to distinct BF trajectories. Clarifying the timing and type of intervention can more effectively address patients’ diverse support needs throughout the treatment process.

This study found that married patients undergoing chemotherapy for CRC were more likely to be classified into the “Low-to-increasing BF group” compared to unmarried, divorced, or widowed patients. Marriage provides not only a stable emotional attachment, offering comfort during treatment-related stress, but also practical care and interaction that enhance patients' sense of security and hope,⁴² thereby fostering more positive cognitive appraisals. In contrast, individuals without spousal support are more susceptible to feelings of loneliness, helplessness, and loss of control when confronted with illness, which increases psychological vulnerability.⁴³ Furthermore, this study revealed that patients living in urban areas were more likely to remain in the “Low-to-increasing BF group”. This finding may be explained by the availability of abundant medical resources, psychological interventions, and health education opportunities in urban settings,⁴⁴ higher health literacy,⁴⁵ and more flexible family and social network structures,⁴⁶ all of which strengthen self-regulation, coping capacity, and emotional support. It is important to note that urban environments may also impose psychological stress, such as fast-paced lifestyles and heightened social expectations.⁴⁷ Nevertheless, the overall results of this study suggest that an urban background has a positive influence on BF. Therefore, nursing practice should fully consider patients’ marital status and residential background. For those without a spouse or residing in rural or resource-limited areas, particular attention should be given to psychological education, social support, and emotional counseling to improve long-term psychological adaptation.

In this study, patients without a stoma were more likely to be grouped into the “Low-to-increasing BF group”. This finding is consistent with Simpson et al.,⁴⁸ who reported that patients who did not undergo ostomy generally exhibited more positive mental health outcomes. In contrast, having an ostomy significantly affects body image, self-esteem, and social functioning. Patients often face challenges such as excretory control difficulties, odor concerns, and social restrictions, all of which can severely impact quality of life.⁴⁹ Evidence also suggests that individuals with an ostomy frequently experience heightened levels of shame, loneliness, and depression,⁵⁰ which may hinder their ability to recognize or construct positive meaning from the illness experience. Moreover, the complexity of adapting to ostomy care and managing postoperative demands may further reduce patients’ motivation to engage in recovery efforts.⁵¹ Furthermore, this study revealed that patients with cancer at stage II or below were more likely to be classified into the “Moderate-to-increasing BF group” or “High-to-decreasing BF group”. This finding indicates that, although early-stage patients generally face a lighter physical burden and have a more favorable prognosis, factors that may facilitate positive psychological experiences,⁵² their BF trajectories remain heterogeneous. Some patients demonstrate progressive improvement, whereas others experience declines due to treatment-related side effects and the depletion of psychological resources. Overall, clinical practice should incorporate staged and tailored psychological support and nursing interventions according to the distinct BF trajectories observed in stoma and early-stage cancer patients. Such interventions should include postoperative psychological support, stoma management guidance, and ongoing psychological assessment,⁵³ all of which aim to promote psychological adaptation.

This study further demonstrated that baseline illness perception and self-disclosure levels are critical predictors of BF trajectories in CRC patients undergoing chemotherapy. Patients with stronger perceived disease threats were more likely to fall into the “Low-to-increasing BF group,” whereas those with higher baseline self-disclosure were more likely to belong to the “High-to-decreasing BF group.” According to stress and coping theory, illness perception, as a primary cognitive appraisal, directly shapes individuals' emotional experiences and meaning-making.²⁰ At diagnosis, heightened perceived threat is often accompanied by anxiety and uncertainty, resulting in lower initial BF levels. However, as treatment progresses and symptoms are alleviated, patients may engage in cognitive reappraisal, transforming perceived threats into motivation for psychological growth.⁵⁴ This supports the notion that “moderate threat” can foster positive psychological change. In parallel, self-disclosure, as a behavioral coping strategy, is contingent on the availability of supportive environments and resources.²⁰ High levels of early self-disclosure may promote positive psychological experiences by facilitating emotional processing, fostering meaning-making, and enhancing perceived social support.⁵⁵ Conversely, when external support is lacking or problems remain unresolved, persistent disclosure may result in disappointment and emotional exhaustion,⁵⁶ diminishing initial positive experiences and ultimately leading to a decline in BF. Overall, these findings highlight that distinct cognitive appraisals and coping strategies exert differential influences on the dynamic development of BF. From a clinical perspective, assessing patients’ initial illness perception and self-disclosure is essential. Illness perception can be redirected into positive motivation through cognitive reappraisal and emotion-regulation interventions,⁵⁷ whereas self-disclosure should be supported with approaches such as problem-solving therapy⁵⁸ to foster a virtuous cycle between expression and coping, thereby facilitating the long-term maintenance and enhancement of BF.

Implications for nursing practice and research

This study identified three distinct trajectories of BF among CRC patients undergoing chemotherapy and highlighted key predictors, including marital status, place of residence, cancer stage, history of stoma, illness perception, and self-disclosure. These findings have important implications for nursing practice. First, tailored nursing interventions should be implemented according to the trajectory type. For example, patients in the “Low-to-increasing BF group” may benefit from continuous psychological education and emotional support to facilitate their psychological growth, while those in the “High-to-decreasing BF group” may require intensified late-stage interventions to prevent psychological decline. Second, special attention should be given to patients with a stoma, advanced-stage cancer, or limited family support. Psychological care, peer support, and communication training can help enhance their self-regulation and positive adaptation. Finally, incorporating BF assessments and dynamic monitoring into nursing workflows may enable more personalized, continuous, and proactive psychological care, ultimately improving patients’ treatment experiences and quality of life.

Limitations

This study has several limitations. First, participants were recruited from only three tertiary hospitals, resulting in a relatively concentrated sample that may not fully capture the diversity of CRC patients across different regions. Future studies should consider incorporating multi-center, cross-regional, or even cross-cultural samples to enhance the generalizability and applicability of the findings. Second, the follow-up period of this study covered four time points from before chemotherapy to after the sixth chemotherapy. While this design effectively captured the dynamic changes in BF during chemotherapy, it did not extend to the completion of treatment or the long-term management phase. Future research could extend the follow-up period to examine post-chemotherapy psychological changes and better map long-term BF trajectories. Third, although this study included a range of sociodemographic, clinical, and psychological variables as predictors, some potentially influential factors, such as religious beliefs and personality traits were not considered. Their exclusion may limit the interpretation of the mechanisms underlying changes in BF trajectories. Future studies should aim to incorporate a broader range of psychosocial variables to build a more comprehensive predictive model. Finally, although the sample size in this study was adequate for repeated-measures analysis, it was relatively small for trajectory modeling, which may have compromised the accuracy of trajectory classification and model stability. Additionally, the limited number of participants lost to follow-up reduces statistical power, and key predictors, such as stoma history, may influence trajectory outcomes. Therefore, the results should be interpreted with caution. Future studies with larger samples are needed to validate BF trajectories and examine the effects of various predictors, thereby improving the robustness of the findings.

Conclusions

This study identified three distinct BF trajectories among CRC patients undergoing chemotherapy: “Low-to-increasing BF group”, “High-to-decreasing BF group”, and “Moderate-to-increasing BF group”. These findings underscore the importance of dynamic psychological assessment throughout the chemotherapy process and the need for individualized interventions based on patients’ trajectory patterns to enhance long-term psychological resilience. In addition, this study identified marital status, place of residence, history of stoma, cancer stage, baseline illness perception, and self-disclosure score as key predictors of the trajectory of BF. Based on the above predictors, health care providers should implement targeted psychological support and interventions to promote higher levels of BF and facilitate the development of more adaptive psychological coping models in clinical practice.

CRediT authorship contribution statement

Fangyi Li: Conceptualization, Methodology, Formal Analysis, Writing – Original Draft. Linyu Zhou: Data Curation, Formal Analysis, Visualization. Biao He: Data Curation, Formal Analysis. Xiaoli Huang: Data Curation, Supervision, Visualization. Chunmei Liu: Data Curation, Resources, Supervision. Qing Li: Data Curation, Supervision. Qiumei Ma: Software, Validation. Tian Xiao: Software, Validation. Ruihan Xiao: Software, Validation. Ao Tang: Writing – Review & Editing. Xueqin Qiu: Writing – Review & Editing. Xiaoju Chen: Project Administration, Funding Acquisition, Writing – Review & Editing. All authors have read and approved the final manuscript.

Ethics statement

The study was approved by the Ethics Committee of the Chengdu Medical College (Approval No. 2024NO.088) and was conducted in accordance with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. All participants provided written informed consent.

Data availability statement

The data that support the findings of this study are available from the corresponding author, XC, upon reasonable request.

Declaration of generative AI and AI-assisted technologies in the writing process

No AI tools/services were used during the preparation of this work.

Funding

This study was supported by the Chengdu Medical College - Mike IVD Clinical Joint Research Center (Grant No. 23LHNBZYB27); and the Research Project of Zigong Key Research Base of Philosophy and Social Sciences - Patient Safety Research Center in 2023 (Grant No. HZAQ-2023-07). The funders had no role in considering the study design or in the collection, analysis, interpretation of data, writing of the report, or decision to submit the article for publication.

Declaration of competing interest

The authors declare no conflict of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article: