Exploring the influencing factors of spiritual well-being in patients with advanced cancer undergoing chemotherapy: a latent profile analysis

Abstract

Background

Spiritual well-being (SWB) is an integral component of palliative and holistic care for patients with advanced cancer undergoing chemotherapy. However, its latent characteristics have not been fully explored.

Objectives

This study aimed to identify latent profiles of spiritual well-being (SWB) among patients with advanced cancer undergoing chemotherapy and to examine the factors influencing profile membership.

Design

Cross-sectional study.

Methods

The study was conducted from February to June 2025 in the oncology departments of three Grade-A tertiary hospitals in Southwest China, involving 318 patients with advanced cancer undergoing chemotherapy. Data were collected using the Demographic Information Questionnaire, the Functional Assessment of Chronic Illness Therapy Spiritual Well-being Scale (FACIT-Sp), the Hospital Anxiety and Depression Scale (HADS), the Comprehensive Scores for Financial Toxicity Based on the Patient-reported Outcome Measures (COST-PROM), and the Family APGAR Index. Descriptive statistics, latent profile analysis, univariate analysis, and multivariate logistic regression were employed to identify latent profiles of SWB and their influencing factors.

Results

The SWB of patients with advanced cancer undergoing chemotherapy was categorized into three latent classes: “Low SWB Group” (28.9%), “Moderate SWB Group” (36.5%), and “High SWB Group” (34.6%). Marital status, residence, monthly household income per capita, payment method for medical expenses, anxiety and depression, FT, and family function were identified as significant factors influencing different levels of SWB (P < 0.05).

Conclusion

Patients with advanced cancer undergoing chemotherapy showed marked heterogeneity in SWB, which could be classified into three latent profiles. These findings suggest that healthcare providers may use a biopsychosocial (BPS) framework to guide stratified supportive care tailored to profile-specific factors, integrating symptom management with psychosocial, family, and spiritual support as well as resource linkage, to better address patients’ diverse needs and potentially improve overall well-being.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12904-026-02058-4.

Introduction

Cancer is one of the leading causes of death worldwide and imposes substantial economic and social burdens on public health systems [1]. In China, more than 2.5 million cancer deaths occurred in 2022, accounting for 26.4% of global cancer mortality [2]. Because early-stage cancer often presents with insidious onset and nonspecific symptoms, many patients are diagnosed at a middle-to-advanced stage [3], making advanced cancer a priority population in oncology care. Chemotherapy remains a commonly used and effective treatment for advanced cancer and can extend survival [4]. However, these patients face not only symptom burden but also high medical costs and family- and society-related stressors [5]. In addition, prolonged treatment and prognostic uncertainty can readily trigger anxiety and depression and exacerbate psychological distress. In this context, patients are more likely to seek spiritual care and support to make sense of life [6]. Therefore, spiritual well-being (SWB) is increasingly recognized as a key resource for alleviating suffering and improving quality of life [7].

SWB is a fundamental component of overall health, promoting well-being by helping individuals derive meaning and purpose in life [8], and it is reflected in perceived life meaning, positive self-identity, harmonious interpersonal relationships, and a sense of inner energy [9]. Prior studies indicate that higher SWB not only helps patients with cancer clarify life goals but also reduces negative emotional experiences and alleviates symptom burden [10, 11]. However, although SWB has been widely recognized in research, spiritual care for patients with advanced cancer receiving chemotherapy remains insufficient in clinical practice. Current assessments often rely primarily on total scale scores, which may not adequately capture individual heterogeneity in SWB. Previous studies have shown that patients with cancer commonly report high spiritual care needs, yet their actual SWB levels are often low [12, 13], highlighting the need for a more refined assessment approach in clinical care. Moreover, few studies have identified latent subgroups of SWB based on individual differences, limiting our ability to reveal within-population heterogeneity and, consequently, to support stratified assessment and targeted interventions. Latent profile analysis (LPA) is a person-centered analytic approach that can identify unobserved subgroups of individuals with similar response patterns [14]. Therefore, this study aims to use LPA to identify latent subgroups with different levels of SWB, thereby providing an evidence base for developing tailored clinical interventions to meet the needs of distinct patient subgroups.

As health care has gradually shifted from a purely biomedical model toward a more comprehensive biopsychosocial perspective [15], the biopsychosocial (BPS) model proposed by George Engel provides a strong theoretical foundation for the present study [16]. The BPS model posits that health and illness arise from the interplay of biological, psychological, and social factors. Although prior research has examined correlates of SWB across physiological, psychological, and social domains [17–19], evidence specific to patients with advanced cancer undergoing chemotherapy remains fragmented. Existing studies often focus on single domains or isolated variables, and few have examined how biological, psychological, and social factors jointly operate within an integrated theoretical framework. For this population, the disease itself and treatment-related adverse effects impose substantial physiological burden; patients commonly experience psychological distress such as anxiety and depression and may also face social stressors, including financial strain and inadequate family support. These interconnected factors may jointly shape SWB and overall health status. Accordingly, guided by the BPS model, this study integrates key biological, psychological, and social variables to systematically investigate factors associated with SWB among patients with advanced cancer receiving chemotherapy.

Methods

Study design and participants

This cross-sectional study employed a convenience sampling method. Patients with advanced cancer who were undergoing chemotherapy in the oncology departments of three tertiary hospitals in Southwest China between February and June 2025 were recruited as study participants.

Inclusion criteria were as follows: (1) age ≥ 18 years; (2) histopathological diagnosis of cancer with TNM stage III or IV; (3) adequate language and comprehension skills, confirmed by a brief screening interview conducted by the researchers to ensure the ability to understand study information and appropriately respond to simple questions; (4) awareness of their disease condition; and (5) provision of written informed consent and voluntary participation, in accordance with ethical standards. Exclusion criteria were as follows: (1) history of psychiatric disorders or hearing impairment; (2) severe cognitive impairment, as determined during the initial screening; (3) concurrent participation in other interventional studies; and (4) inability to independently complete self-administered questionnaires.

Sample size estimation

This study primarily involved LPA and multinomial logistic regression. According to Nylund-Gibson, the minimum sample size requirement for LPA is 300 participants [20]. For multinomial logistic regression, the sample size is generally estimated to be 5 to 10 times the number of independent variables [21]. Given that 25 independent variables were included in this study, the required sample size ranged from 125 to 250. Considering a 20% potential attrition rate, the final target sample size was determined to be 150 to 300. Ultimately, a total of 318 patients were successfully enrolled in the study.

Procedure and ethical considerations

In this study, the investigators and team members served as the principal researchers and data collectors. Eligible patients were initially identified by reviewing electronic medical records, and their basic demographic and clinical information was recorded. Prior to the formal survey, researchers explained the study objectives and procedures to patients. Upon obtaining informed consent, questionnaires were distributed on-site for data collection. Questionnaires were administered in quiet patient rooms or conference rooms. Each questionnaire took approximately 15 to 20 min to complete. Immediately after completion, researchers checked the responses for completeness in real time to minimize missing data. If missing or ambiguous responses were detected, participants were asked to provide clarification on the spot whenever possible; if this was not feasible, the missing values were retained. During the data analysis phase, questionnaires with substantial missing data were excluded using listwise deletion, while sporadic missing values were addressed through multiple imputation to minimize potential bias. To safeguard participants’ rights, the research team assured all participants that their personal information and questionnaire responses would be kept strictly confidential and used solely for research purposes. Participants were also informed that they could withdraw from the study at any time without any impact on their medical care.

A total of 340 patients with advanced cancer undergoing chemotherapy were initially approached. Of these, 12 declined participation, 5 withdrew due to physical discomfort, and 5 questionnaires were excluded because of excessively homogeneous responses. Ultimately, 318 valid questionnaires were obtained, yielding an effective response rate of 92.2%. The data collection process is illustrated in Fig. 1.

Fig. 1.

A flow chart about data collection

This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of the Chengdu Medical College. The ethical approval number for this study is [2024 NO.088]. Researchers explained the study to eligible participants. Each participant signed a written informed consent form prior to participation.

Measurements

Demographic information questionnaire

The participants’ demographic characteristics were collected, including gender, age group, marital status, educational level, religious belief, average monthly household income per capita (CNY), employment status, residence, primary caregiver, payment method for medical expenses. Disease-related information includes types of cancer, tumor metastasis, synchronize with other treatment methods, disease duration, chronic comorbidities, and round of chemotherapy.

Place of residence was self-reported as the participant’s usual living area and was categorized as city, town, or countryside according to the Chinese administrative/statistical context: city refers to urban districts of prefecture-level cities or above; town refers to county seats or designated town built-up areas outside city districts; and countryside refers to non-built-up rural areas, typically village-based communities administered at the township and village level [22].

The functional assessment of chronic illness therapy spiritual well-being scale (FACIT-Sp)

The validated Chinese version of the FACIT-Sp was used in this study [23]. This instrument is widely used in Chinese populations and has demonstrated good reliability and validity. It comprises 12 items across three domains (peace, meaning, and faith). Items are rated on a 5-point Likert scale from 0 (“not at all”) to 4 (“very much”), yielding a total score of 0 ~ 48; higher scores indicate better spiritual well-being.

Comprehensive scores for financial toxicity based on the patient-reported outcomes measures (COST-PROM)

Financial toxicity was assessed using the validated Chinese version of the COST-PROM [24], which has been applied in Chinese cancer populations and exhibits sound psychometric properties. The scale contains 11 items covering two domains (positive financial well-being and negative psychological response). Items are rated on a 5-point Likert scale from 0 (“not at all”) to 4 (“very much”), with total scores ranging from 0 to 44. Lower scores indicate greater financial toxicity.

The hospital anxiety and depression scale (HADS)

Anxiety and depression were measured using the validated Chinese version of the HADS [25], which has been validated in general hospital populations in China. The HADS consists of 14 items and includes two subscales: anxiety (HADS-A) and depression (HADS-D). Each item is rated on a 4-point Likert scale (0 ~ 3), with each subscale score ranging from 0 to 21. Higher scores indicate more severe anxiety or depression. A score of ≥ 8 on either subscale suggests the presence of clinically relevant anxiety or depression.

Family adaptation, partnership, growth, affection, and resolve (Family APGAR)

Family functioning was assessed using the validated Chinese version of the Family APGAR [26], which has been widely used in Chinese populations. The scale comprises five items assessing adaptation, partnership, growth, affection, and resolve. Each item is rated on a 3-point scale (0 ~ 2): “hardly ever” (0), “sometimes” (1), and “almost always” (2). Total scores range from 0 to 10, with 7 ~ 10 indicating good family function, 4 ~ 6 indicating moderate dysfunction, and 0 ~ 3 indicating severe dysfunction.

Selection of analyzed variables

Guided by the BPS model, we prespecified candidate variables prior to analysis to capture key domains related to SWB during chemotherapy and to support clinical risk stratification. Disease-related variables (Types of cancer, tumor metastasis, chronic comorbidities, disease duration, round of chemotherapy, and synchronize with other treatment methods) were included to characterize disease severity and treatment burden. Psychological status was operationalized using the HADS, given its established relevance to SWB in oncology. Social factors were assessed using the COST-PROM and the Family APGAR to reflect financial stress and family support during chemotherapy. Sociodemographic characteristics (Age, marital status, educational level, religious belief, average monthly household income per capita (CNY), employment status, residence, primary caregiver, and payment method for medical expenses) were collected to describe the sample and were treated as potential confounders in subgroup comparisons and regression analyses. All variables are routinely obtainable in clinical practice and have direct implications for tailoring nursing interventions.

Statistical analysis

Data were analyzed using IBM SPSS Statistics version 27.0 and Mplus version 8.3. LPA was performed to identify SWB subgroups. The LPA model used the 12 FACIT-Sp item scores as continuous indicators. Models were estimated sequentially, starting with a one-class solution and increasing the number of classes until the optimal fit was achieved. Model fit was evaluated using the Akaike information criterion (AIC), Bayesian information criterion (BIC), and sample-size adjusted BIC (aBIC), with lower values indicating better fit. Classification accuracy was assessed using entropy (range 0 ~ 1), with values > 0.80 generally considered indicative of good classification [27]. The Lo-Mendell-Rubin adjusted likelihood ratio test (LMR) and the bootstrapped likelihood ratio test (BLRT) were used to compare adjacent class solutions; a P value < 0.05 indicates that the k-class model fits better than the (k − 1) - class model [28]. In selecting the final model, we also considered class proportions and clinical interpretability.

Descriptive and inferential analyses were subsequently performed in SPSS. The Shapiro-Wilk test was applied to examine the normality of continuous variables. Variables with a normal distribution were presented as mean ± standard deviation (M ± SD), whereas those with a skewed distribution were expressed as median and interquartile range. Differences in sociodemographic and clinical characteristics across SWB subgroups were examined using one-way analysis of variance (ANOVA), chi-square (χ²) test, or Fisher’s exact test, as appropriate. Variables showing statistical significance in univariate analyses were further included in a multivariate logistic regression model to explore factors associated with SWB subgroup membership among patients with advanced cancer receiving chemotherapy. A two-tailed P value < 0.05 was considered statistically significant. To minimize the impact of multicollinearity, categorical predictors were dummy coded prior to model fitting, and variance inflation factors (VIFs) were calculated based on the final set of predictors included in the model; VIF < 5 was considered acceptable. Bonferroni correction was applied to control the family-wise type I error rate arising from multiple comparisons and to improve the robustness of the findings.

Results

Characteristics of participants

Among the 318 patients with advanced cancer receiving chemotherapy, 61.3% were male and 38.7% were female. Approximately 62.3% of patients were aged 18 to 64 years, while 37.7% were aged 65 years or older. The vast majority were married (85.5%). With respect to educational level, 36.5% had completed primary school or below, whereas only 15.9% had received education at the college level or above. Regarding religious belief, 20.8% reported having a religious belief. In terms of employment status, most patients were farmers or unemployed (64.1%).

A total of 39.9% of patients resided in urban areas. Only 13.2% reported a monthly per capita household income exceeding 5,000 (CNY). Most patients had undergone 1 to 4 rounds of chemotherapy (42.8%), and 70.5% had a disease duration longer than 6 months. Comorbid chronic diseases were present in 36.8% of patients. With respect to adjunctive treatment, radiotherapy was the most common (44.3%). (See Supplementary Material 1).

Latent profile analysis results

Establishment of the latent profile model for patients’ SWB

Using the 12 items of the SWB scale as manifest indicators, latent profile models ranging from a one-class baseline model to a five-class model were sequentially fitted (Table 1). The results showed that as the number of classes increased, the values of AIC, BIC, and aBIC decreased progressively, while the entropy value increased. The differences were statistically significant for the two-, three-, and four-class models (P < 0.05). However, when the number of classes reached five, the LMR test was not statistically significant (P > 0.05). Although the two-class model yielded the highest entropy value, its AIC, BIC, and aBIC values were higher than those of the three-class model. Notably, BIC is generally considered more robust in model selection, particularly in avoiding overfitting, and the BIC of the three-class model showed a clear improvement compared to the two-class model. In contrast, the three-class model demonstrated statistical significance in both the LMR and BLRT tests (P < 0.001), indicating a better fit than the two-class model. The four-class model also showed significance in the LMR and BLRT tests (P = 0.005), but its entropy value (0.930) was lower than that of the three-class model. Therefore, considering model fit indices, BIC robustness, and clinical interpretability, the three-class model was identified as the optimal solution.

Table 1.

Fit indices for the five models using LPA (n = 318)

Profile	AIC	BIC	aBIC	Entropy	LMR ( P )	BLRT ( P )	Profile prevalence
1	10777.916	10868.205	10792.082	—	—	—	1.00
2	8470.774	8609.970	8492.614	0.958	< 0.001	< 0.001	0.437 / 0.563
3	7774.212	7962.315	7803.725	0.940	< 0.001	< 0.001	0.289 / 0.365 / 0.346
4	7507.881	7744.890	7545.067	0.930	0.0050	< 0.001	0.164 / 0.299 / 0.264 / 0.273
5	7438.797	7724.713	7483.657	0.900	0.4385	< 0.001	0.138 / 0.198 / 0.217 / 0.189 / 0.258

AIC Akaike information criterion, BIC Bayesian information criterion, aBIC Adjusted Bayesian information criterion, LMR Lo–Mendell–Rubin likelihood ratio test, BLRT Bootstrapped likelihood ratio test

Naming of each latent profile

Figure 2 illustrates the latent profile plot of the three subgroups across the 12 manifest indicators. In this study, the probabilities of occurrence for Profile 1, Profile 2, and Profile 3 were 28.9% (n = 92), 36.5% (n = 116), and 34.6% (n = 110), respectively. The mean total SWB scores for patients in Profile 1, Profile 2, and Profile 3 were 12.99 ± 6.24, 24.18 ± 5.05, and 33.98 ± 3.43, respectively. According to the definition of FACIT-Sp, Profile 1, Profile 2, and Profile 3 were classified as the “Low SWB Group”, “Moderate SWB Group”, and “High SWB Group” respectively. No curve intersections were observed across the three profiles, and the trajectories showed a clear graded pattern. The “Low SWB Group” consistently scored lowest across the meaning, peace, and faith dimensions, the “Moderate SWB Group” exhibited intermediate levels of SWB, and the “High SWB Group” achieved the highest scores on all items. This pattern suggests that the latent profiles primarily reflect differences in overall SWB, with relatively consistent item-level contrasts across groups, supporting the discriminability and interpretability of the three-class solution.

Fig. 2.

Latent profile characteristics of SWB in advanced cancer patients undergoing chemotherapy

Comparison of general characteristics among different SWB subgroups

Univariate analyses were conducted using the three SWB latent classes of patients with advanced cancer undergoing chemotherapy as the dependent variable, and demographic characteristics, anxiety and depression, FT, and family function as independent variables. The results showed statistically significant differences across SWB profiles in terms of marital status (χ² = 16.099, P < 0.001), educational level (χ² = 19.387, P = 0.004), religious belief (χ² = 20.991, P < 0.001), average monthly household income per capita (CNY) (χ² = 34.338, P < 0.001), employment status (χ² = 29.658, P < 0.001), residence (χ² = 28.142, P < 0.001), and payment method for medical expenses (χ² = 30.253, P < 0.001). Moreover, significant differences were also observed for anxiety and depression (H = 14.776, P < 0.001), FT (H = 16.432, P < 0.001), and family function (F = 58.899, P < 0.001). These findings indicate that patient characteristics differed significantly across the three SWB latent profiles (P < 0.05). Detailed results are presented in Table 2.

Table 2.

Univariate analysis of factors associated with SWB latent profiles (n = 318)

Variables	Categories	Profile 1	Profile 2	Profile 3	Statistics	P value
		n (%)	n(%)	n(%)
Gender	Male	70 (54.3)	57 (62.9)	94 (65.5)	16.557 a	< 0.001**
	Female	37 (34.6)	51 (47.2)	26 (21.7)
Age	18 ~ 64 years	62 (67.4)	72 (62.1)	64 (58.2)	1.811 a	0.404
	≥ 65 years	30 (32.6)	44 (37.9)	46 (41.8)
Marital status	Married	85 (79.4)	92 (85.2)	101 (84.2)	16.099 a	< 0.001**
	Unmarried / Divorce / Widowed	22 (20.6)	16 (14.8)	19 (15.8)
Educational level	Primary school or below	28 (30.4)	37 (31.9)	26 (23.6)	19.387 a	< 0.004*
	Junior high school	32 (34.8)	40 (34.5)	20 (18.2)
	High school or vocational school	21 (22.8)	20 (17.2)	39 (35.5)
	College or above	11 (12.0)	19 (16.4)	25 (22.7)
Religious belief	Yes	7 (7.6)	22 (19.0)	37 (33.6)	20.991 a	< 0.001**
	No	85 (92.4)	94 (81.0)	73 (66.4)
Average monthly household income per capita (CNY)	< 1000	15 (16.3)	11 (9.5)	1 (0.9)	34.338 a	< 0.001**
	1001 ~ 3000	44 (47.8)	50 (43.1)	39 (35.5)
	3001 ~ 5000	27 (29.4)	46 (39.6)	43 (39.1)
	> 5000	6 (6.5)	9 (7.8)	27 (24.5)
Employment status	Working	10 (10.9)	11 (9.5)	9 (8.2)	29.658 a	< 0.001**
	Not working	36 (39.1)	38 (32.8)	37 (33.6)
	Peasants	36 (39.1)	39 (33.6)	18 (16.4)
	Retired	10 (10.9)	28 (24.1)	46 (41.8)
Residence	City	22 (23.9)	42 (36.2)	63 (57.3)	28.142 a	< 0.001**
	Town	36 (39.1)	28 (34.1)	25 (22.7)
	Countryside	34 (37.0)	46 (39.7)	22 (20.0)
Primary caregiver	None	17 (18.5)	13 (11.2)	19 (17.3)	4.778 a	0.311
	Mate	49 (53.3)	68 (58.6)	68 (61.8)
	Child / Parent / Sibling	26 (28.2)	35 (30.2)	23 (20.9)
Payment method for medical expenses	NCMS	57 (62.0)	62 (53.4)	33 (30.0)	30.253 a	< 0.001**
	URBMI	21 (22.8)	24 (20.7)	25 (22.7)
	UEBMI	14 (15.2)	30 (25.9)	52 (47.3)
Types of cancer	Lung Cancer	24 (26.1)	40 (34.5)	35 (31.8)	6.615 b	0.581
	Gastrointestinal Cancer	51 (55.4)	49 (42.2)	55 (50.0)
	Gynecological Cancer	12 (13.0)	14 (12.1)	9 (8.2)
	Nasopharyngeal Cancer	2 (2.2)	4 (3.4)	4 (3.6)
	Other types of cancer	3 (3.3)	9 (7.8)	7 (6.4)
Tumor metastasis	Yes	72 (78.3)	85 (73.3)	80 (72.7)	0.959 a	0.619
	No	20 (21.7)	31 (26.7)	30 (27.3)
Synchronize with other treatment methods	No	9 (9.8)	14 (12.1)	11 (10.0)	4.710 b	0.796
	Targeted therapy	10 (10.9)	8 (6.9)	8 (7.3)
	Radiotherapy	37 (40.2)	54 (46.6)	52 (47.3)
	Immunotherapy	30 (32.6)	37 (31.9)	32 (29.1)
	Surgical treatment	6 (6.5)	3 (2.6)	7 (6.4)
Disease duration	Half a year or below	26 (28.3)	38 (32.8)	30 (27.3)	1.475 a	0.831
	Half to one year	34 (37.0)	42 (36.2)	46 (41.8)
	More than one year	32 (34.8)	36 (31.0)	34 (30.9)
Chronic comorbidities	Yes	30 (32.6)	45 (38.8)	42 (38.2)	0.983 a	0.612
	No	62 (67.4)	71 (61.2)	68 (61.8)
Round of chemotherapy	1 ~ 4	43 (46.7)	53 (45.7)	40 (36.4)	4.489 a	0.344
	5 ~ 8	30 (32.6)	31 (26.7)	40 (36.4)
	≥ 9	19 (20.7)	32 (27.6)	30 (27.2)
Anxiety and depression		20.63 ± 8.67	14.72± 8.63	11.75± 6.48	31.774 c	< 0.001**
Financial toxicity		19.75 ± 7.33	20.23 ± 9.37	25.68 ± 8.15	16.432 c	< 0.001**
Family function		7.00 (5.00, 8.00)	8.00 (8.00,10.00)	10.00 (9.00,10.00)	58.899 d	< 0.001**

Profile 1, Low SWB Group; Profile 2, Moderate SWB Group; Profile 3, High SWB Group

CNY Chinese yuan, NCMS New rural cooperative medical system, URBMI Urban residents’ basic medical insurance, UEBMI Urban employee basic medical insurance

^aChi-square test, ^bFisher exact probability, ^cThe one-way ANOVA, ^dKruskal-Wallis test 

*, P < 0.05; **, P < 0.001

Multinomial logistic regression analysis of factors associated with SWB subgroups

First, we assessed multicollinearity among the predictors planned for inclusion in the final model. No substantial multicollinearity was detected; all VIFs were < 5 (See Supplementary Material 2). We then entered variables that were statistically significant in the univariate analyses into an unordered multinomial logistic regression model, with SWB latent class membership as the dependent variable and the “Low SWB Group” as the reference category. To avoid misinterpretation of direction, odds ratios (ORs) were interpreted consistently as follows: OR > 1 indicates higher odds of being in the target group (relative to the “Low SWB group”), whereas OR < 1 indicates lower odds of being in the target group (that is, a greater tendency to be classified into the “Low SWB group”).

The regression results showed that being married (OR = 4.092, P = 0.013) and having a religious belief (OR = 6.564, P = 0.024) were associated with higher odds of membership in the “High SWB Group”. In contrast, lower average monthly household income per capita (OR = 0.037, P = 0.017), residing in town areas (relative to countryside areas) (OR = 0.406, P = 0.048), and paying medical expenses through the New cooperative medical system (NCMS) (OR = 0.081, P = 0.009) were associated with lower odds of being in the “High SWB Group”, indicating a higher likelihood of classification into the “Low SWB Group”. In addition, higher anxiety and depression levels were associated with lower odds of being in the “High SWB Group” (OR = 0.906, P < 0.001), suggesting a greater tendency toward the “Low SWB Group”. Conversely, lower financial toxicity (OR = 1.059, P = 0.030) and better family function (OR = 1.368, P < 0.001) were associated with higher odds of membership in the “High SWB Group”. (Detailed results are presented in Table 3).

Table 3.

Multivariate logistic regression analysis (n = 318)

Variables	Profile 1 VS Profile 2				Profile 1 VS Profile 3
	β	P	OR	OR 95% CI	β	P	OR	OR 95% CI
Marital status (Reference: Unmarried / Divorce / Widowed)
Married	0.689	0.102	1.991	0.871 ~ 4.551	1.409	0.013*	4.092	1.348 ~ 12.424
Religious belief (Reference: No)
Yes	0.860	0.105	2.364	0.836 ~ 6.682	1.882	< 0.001***	6.564	2.153 ~ 20.010
Average monthly household income per capita (CNY) (Reference:> 5000 )
<1000	-0.329	0.713	0.720	0.125 ~ 4.143	-3.291	0.017*	0.037	0.002 ~ 0.554
1001 ~ 3000	0.500	0.506	1.648	0.378 ~ 7.178	-0.028	0.970	0.972	0.224 ~ 4.225
3001 ~ 5000	0.664	0.337	1.942	0.502 ~ 7.520	-0.042	0.951	0.959	0.255 ~ 3.612
Residence (Reference: Countryside )
City	0.331	0.643	1.393	0.343 ~ 5.654	0.715	0.398	2.044	0.389 ~ 10.725
Town	-0.900	0.048*	0.406	0.167 ~ 0.990	-0.698	0.211	0.498	0.167 ~ 1.484
Payment method for medical expenses (Reference: UEBMI)
NCMS	-1.653	0.055	0.191	0.035 ~ 1.036	-2.516	0.009**	0.081	0.012 ~ 0.536
URBMI	-0.556	0.424	0.573	0.146 ~ 2.224	-1.472	0.061	0.230	0.049 ~ 1.069
Anxiety and depression	-0.069	< 0.001***	0.933	0.898 ~ 0.970	-0.099	< 0.001***	0.906	0.862 ~ 0.951
Financial toxicity	0.001	0.964	1.001	0.957 ~ 1.047	0.057	0.030*	1.059	1.006 ~ 1.114
Family function	0.112	0.111	1.119	0.974 ~ 1.284	0.313	< 0.001***	1.368	1.149 ~ 1.628

Profile 1, Low SWB Group; Profile 2, Moderate SWB Group; Profile 3, High SWB Group

CNY Chinese yuan, NCMS New rural cooperative medical system, URBMI Urban residents’ basic medical insurance, UEBMI Urban employee basic medical insurance

*, P < 0.05; **, P < 0.01; ***, P < 0.001

Discussion

Using LPA, we identified three SWB profiles among patients with advanced cancer undergoing chemotherapy: the “Low SWB Group” (28.9%), “Moderate SWB Group” (36.5%), and “High SWB Group” (34.6%). Beyond supporting model fit, this person-centered classification operationalizes SWB heterogeneity into clinically meaningful nursing guidance for risk stratification, targeted assessment, and tiered intervention planning during chemotherapy. Clinically, the “Low SWB Group” may indicate a high-risk group, warranting prioritized supportive care assessment, screening for psychological distress, and a brief spiritual history focused on meaning and inner peace. This should be complemented by structured communication and, when indicated, referral to psychological/spiritual services and linkage to relevant resources to address their heightened risk of emotional distress and social stressors [29]. The “Moderate SWB Group” represents a “vulnerable but modifiable” subgroup; regular follow-up and dynamic monitoring across round of chemotherapy, combined with brief meaning-centered or mindfulness-based support and strengthened mobilization of family and social resources, may enhance coping capacity and help prevent deterioration in SWB [30]. In contrast, the “High SWB Group” appears to possess stronger spiritual resources; care should focus on maintenance and consolidation through periodic reassessment, facilitation of family communication, and reinforcement of positive sources of meaning, while leveraging protective factors observed in this group to inform supportive strategies for other patients [13]. Overall, SWB profiling provides a stratified entry point and prioritization framework for supportive care during chemotherapy, enabling more efficient allocation of nursing and social support resources.

Our findings indicate that marital status is an important correlate of SWB. Compared with patients who were unmarried, divorced, or widowed, married patients were more likely to be classified into the “High SWB Group”, which is consistent with the findings reported by Feng et al. [31]. A plausible explanation is that married patients are more likely to receive sustained daily care and emotional support from their spouses, which may help reduce psychological stress and strengthen adaptation to illness. Notably, within the Chinese context, cancer care is often characterized by a “family-centered” pattern, and spouses commonly assume major caregiving and accompaniment roles, providing continuous presence and practical assistance and playing a key part in treatment communication and decision-making. In addition, as one of the primary caregivers, a spouse can often provide deeper psychological support [32], which may alleviate negative emotions, enhance coping with advanced cancer, and thereby contribute to higher SWB. Therefore, health care professionals should pay particular attention to patients who are unmarried, divorced, or widowed by conducting early assessments of gaps in social support and caregiving resources, and by strengthening support networks through family meetings, caregiver support, and peer-support programs, with the aim of improving SWB [33].

We also found that patients with religious belief were more likely to be classified into the “High SWB Group”, suggesting that religion may contribute to SWB by providing meaning, spiritual comfort, and coping resources under disease-related stress, consistent with Miller et al. [34, 35]. This interpretation is also supported by relevant guidelines, which recognize religion as a potential source of comfort and peace for patients with advanced cancer [36]. Importantly, in the Chinese cultural context, spirituality is not fully equivalent to formal religious affiliation. Individuals’ belief systems are often shaped by Confucian and Taoist traditions that emphasize harmony between humans and nature, exploration of life meaning, and interpersonal harmony, which may provide support similar to that derived from religion [37]. Moreover, within the local cultural milieu, religious belief and spiritual experience often intersect, reflecting a culturally grounded orientation toward seeking meaning through inner strength, which may enhance self-coping capacity and buffer psychological distress [38]. Accordingly, we recommend strengthening culturally sensitive spiritual assessment and communication to better identify and respond to patients’ religious and spiritual needs. For patients with religious belief, clinicians should respect their faith and encourage them to draw positive attitudes and optimism from their beliefs, thereby supporting treatment persistence [39]. For those without religious belief, spiritual support may be provided through meaning-oriented communication, hope-maintenance strategies, and family involvement, helping patients sustain hope and a sense of value during treatment and ultimately improve SWB [40].

Our findings showed that patients residing in towns were more likely to be classified into the “Low SWB Group” than those living in rural areas, suggesting lower spiritual well-being among town residents. This differs from the results reported by Qi et al. [41], who found higher SWB among patients living in cities than those in rural areas, which they attributed to better economic conditions and access to medical care in urban settings. A plausible explanation for this discrepancy is that town represents a transitional form between urban cores and rural areas. Although towns may have advantages over rural areas in terms of healthcare access and economic resources, they remain relatively disadvantaged compared with urban districts. Such relative gaps in healthcare resources and socioeconomic conditions may intensify existential distress among patients with advanced cancer undergoing chemotherapy, thereby undermining SWB [42]. In contrast, rural patients may benefit from stronger community cohesion and interpersonal support networks, as well as greater reliance on traditional cultural or religious practices, which may offer partial protection for SWB [43]. These findings suggest that spiritual care interventions should take patients’ residential context and social environment into account. For patients living in towns, nursing care may need to place greater emphasis on psychosocial support and resource linkage; for those in rural areas, efforts should focus on meeting basic healthcare and financial needs while reinforcing community networks and culturally embedded resources to promote spiritual well-being and overall quality of life.

In addition, patients with lower household income, those whose medical expenses were covered by the NCMS, and those with more severe financial toxicity were more likely to belong to the “Low SWB Group”, consistent with previous studies [10, 19]. From a BPS model perspective, financial toxicity and socioeconomic disadvantage can be viewed as cross-domain stressors. At the social level, economic burden may restrict access to treatment and supportive resources and increase caregiving demands and role strain within the family. At the psychological level, sustained financial stress may heighten helplessness, self-blame, and worries about the future, thereby exacerbating anxiety and depression and ultimately eroding SWB. Conversely, higher SWB may buffer the subjective burden of financial stress, as patients may adopt more proactive help-seeking and coping approaches and more effectively connect with available resources, which can mitigate the psychological and practical impact of financial difficulties [44]. Notably, Kurt et al. [45] reported an inverse association between higher income and lower SWB, which contrasts with our findings. This inconsistency may reflect differences in study populations, cultural contexts, and healthcare payment systems. In addition, when financial pressure is lower, reliance on faith-based coping may decrease, potentially resulting in lower scores on the faith subscale without necessarily reducing overall well-being. Clinically, these findings support integrating financial risk identification into routine assessment. For patients with high financial toxicity or marked economic stress, nursing care should extend beyond standard health education to include coordinated cost and resource navigation with social work and multidisciplinary teams, facilitating access to supportive services and reducing the downstream transmission of economic stress into psychological distress and SWB decline [46].

Our findings indicated that patients with more severe anxiety and depression were more likely to be classified into the “Low SWB Group”, consistent with the study by Yang et al. [47]. This finding also suggests that the peace dimension of SWB may be most closely associated with anxiety and depression. A higher level of peace may reduce uncertainty and hopelessness, thereby alleviating emotional distress. Patients with higher SWB may also be better able to mobilize internal resources and adopt adaptive strategies, such as problem-focused coping and emotion regulation, to mitigate negative emotional responses. In patients with advanced cancer, anxiety and depression are highly prevalent and represent key determinants of quality of life and treatment adherence [48]. Therefore, alleviating anxiety and depression may contribute to improving spiritual well-being. Bovero et al. [49] further noted that emotion regulation strategies, such as self-distraction and behavioral disengagement, may help relieve depression and anxiety symptoms and, in turn, improve SWB. In future clinical practice, standardized screening tools for SWB and psychological distress (for example, the FACIT-Sp and HADS) could be incorporated into routine care pathways to facilitate early identification of patients at high risk for anxiety and depression, enable timely management of negative emotions, and create conditions conducive to improving SWB.

In addition, we found that patients with higher Family APGAR scores were more likely to belong to the “High SWB Group”, which is consistent with prior evidence [50] and suggests a close link between family support and SWB. Family represents an integral spiritual resource for patients with advanced cancer. From a BPS model perspective, family support is not only a social resource but may also enhance SWB by facilitating emotion regulation and strengthening treatment confidence and the maintenance of hope [51]. Encouragement and trust from family members may help patients buffer negative emotions during chemotherapy, improve illness adaptation, and reinforce a sense of life value [52]. Nixon et al. [53] reported that family support is among the most important spiritual needs of patients with cancer, further underscoring the role of the family in alleviating spiritual distress. Accordingly, integrating family caregiving into cancer care may be an effective approach to improving SWB and quality of life. When developing interventions, we recommend routinely assessing family functioning and support needs within nursing assessments and strengthening the effectiveness of family support systems through caregiver education, family-inclusive communication, and, when indicated, linkage to multidisciplinary resources, thereby providing a foundation for improving SWB and psychological outcomes [54].

Limitations

This study has several limitations. First, the cross-sectional design prevents inference of temporal ordering or causality among the observed associations. Longitudinal studies are needed to examine trajectories of SWB across round of chemotherapy and to test potential mediating pathways within the biopsychosocial framework. Second, participants were recruited by convenience sampling from three tertiary hospitals in Southwest China, which may introduce selection bias and limit external validity. Replication in multicenter samples from diverse regions and care settings is warranted. Third, all key constructs were assessed using self-reported questionnaires, which may be affected by recall and social desirability biases and may contribute to common method variance. Future work should incorporate multi-method assessments (e.g., qualitative interviews, clinician-rated measures, and clinical / symptom indicators). Finally, although we included multiple cancer types to reflect real-world chemotherapy practice, heterogeneity in cancer site and treatment regimens may limit the applicability of the findings to specific subgroups. Future studies focusing on more homogeneous cancer populations or stratifying by cancer type and treatment characteristics could provide more tailored evidence for clinical decision-making.

Conclusion

Guided by the BPS model, this study identified distinct SWB subgroups among patients with advanced cancer undergoing chemotherapy, demonstrating substantial heterogeneity in SWB within this population. Marital status, place of residence, per capita monthly household income, method of medical expense payment, anxiety and depression, FT, and family function were significantly associated with SWB levels. These findings provide a basis for individualized and subgroup-specific nursing interventions that integrate biological, psychological, social, and spiritual care.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

SWB

Spiritual well-being

BPS

Biopsychosocial

FT

Financial toxicity

CNY

Chinese yuan

NCMS

New rural cooperative medical system

URBMI

Urban residents’ basic medical insurance

UEBMI

Urban employee basic medical insurance

FACIT-Sp

Functional assessment of chronic illness therapy spiritual well-being scale

HADS

Hospital anxiety and depression scale

COST-PROM

Comprehensive scores for financial toxicity based on the patient-reported outcome measures

Authors' contributions

**First author Ao Tang: ** conceptualization, investigation, study design, data acquisition, data analysis, manuscript drafting. **Co-first author Haiyan Liao and Huiru Zeng: ** conceptualization, methodology, formal analysis, writing-original draft. **Ying Zheng: ** writing-review and editing, resources. **Qing Li: ** writing-review & editing, visualization. **Fangyi Li: ** writing-review and editing, visualization. **Tian Xiao: ** software, methodology, data curation. **Linyu Zhou: ** methodology, data curation. **Jinlian Wu: ** writing-review and editing, visualization. **Biao He: ** methodology. **Nianfei Tang: ** methodology. **Corresponding author Xiaoju Chen: ** conceptualization, methodology, formal analysis, writing-review and editing, resources, funding acquisition. All authors had full access to all the data in the study, and the corresponding author had final responsibility for the decision to submit for publication. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Funding

This work was supported by the Chengdu Medical College-Chengdu Pidu District People’s Hospital 2024 project [Grant numbers 24LHLNYX1-50] and the Sichuan Research Center of Applied Psychology (Grant numbers: [CSXL-25220]). The funder had no role in study design, data collection, analysis, interpretation, or manuscript preparation.

Data availability

The dataset used in this research is available from the corresponding author upon a reasonable request.

Declarations

Ethics approval and consent to participate

This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of the Chengdu Medical College. The ethical approval number for this study is [2024 NO.088]. Researchers explained the study to eligible participants. Each participant signed a written informed consent form prior to participation.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.