Influencing factors and predictive model construction of anxiety and depression in patients with cervical cancer

Abstract

BACKGROUND

Anxiety and depression are highly prevalent among patients with cervical cancer (CC). However, few studies have systematically analyzed the psychological effects of tumor stage, treatment methods, and related factors on these patients, or developed predictive models for these outcomes.

AIM

To identify factors influencing anxiety and depression in patients with CC and construct predictive models.

METHODS

We retrospectively analyzed data from 119 patients with CC treated at the Gynecology Department of Suzhou Ninth People’s Hospital between January 2017 and May 2025. Clinical data, psychological hope levels at diagnosis, and Self-Rating Anxiety Scale and Self-Rating Depression Scale scores during treatment were collected. Influencing factors were identified, and predictive models were developed. The model performance was evaluated using receiver operating characteristic (ROC) curves and the Hosmer-Lemeshow goodness-of-fit test.

RESULTS

During treatment, 64.71% of the patients experienced anxiety and 52.10% experienced depression. Significant differences in family income, tumor stage, treatment modality, and hope level were observed between patients with and without anxiety/depression (P < 0.05). Multivariate analysis showed that a family monthly income < 5000 yuan, stage III-IV tumor, comprehensive treatment, and low hope level were independent risk factors (P < 0.05). The predictive formula for anxiety was as follows: Logit (P) = 0.795 × monthly income + 0.594 × tumor stage + 1.095 × treatment method + 1.184 × hope level − 9.176; for depression: Logit (P) = 0.432 × monthly income + 0.518 × tumor stage + 0.727 × treatment method + 1.095 × hope level − 8.541. The area under the ROC curves were 0.865 for anxiety and 0.837 for depression. Goodness-of-fit test confirmed no overfitting (P > 0.05).

CONCLUSION

Family income, tumor stage, treatment method, and hope level are key determinants of anxiety and depression in patients with CC. Predictive models incorporating these factors can effectively assess risk of anxiety and depression during treatment.

Key Words: Cervical cancer; Depression; Anxiety; Influencing factors; Prediction model

Core Tip: Cervical cancer (CC) causes long-term distress in women, who are prone to negative emotions such as anxiety and depression. This study analyzed the factors influencing anxiety and depression in patients with CC and developed a predictive model. The findings provide guidance for the clinical assessment of psychological risks and support the adoption of targeted measures to reduce these risks.

INTRODUCTION

Cervical cancer (CC) is a malignant tumor of the cervix, vagina, and cervical canal and remains the most common malignancy of the female reproductive system[1]. Common clinical symptoms include vaginal bleeding, discharge, pain, urinary frequency or urgency, and constipation[2]. Standard treatments include surgery, radiotherapy, chemotherapy, and targeted therapy[3]. However, patients often face the psychological burden of diagnosis, while surgery may involve hysterectomy and radiotherapy or chemotherapy can cause adverse effects such as vomiting, diarrhea, fatigue, and hair loss. This challenges contribute to varying degrees of anxiety and depression, which may significantly affect prognosis[4]. In addition, as CC is strongly associated with human papillomavirus infection[5], some patients may feel shame or stigma, believing the disease to be linked to sexual behavior, which further increases anxiety and depression.

Anxiety is characterized by excessive worry, fear, or tension about potential threats, often accompanied by hyperarousal and vigilance. Depression is defined by persistent low mood, loss of interest or pleasure, and decreased energy, often with fatigue and diminished vitality. Studies have shown that depression and anxiety are highly prevalent among adult cancer patients in China[6]. Cross-sectional surveys report prevalence rates of 52.2% and 65.6% among patients with CC, respectively[7], highlighting the considerable psychological burden. Despite this, little research has examined how clinical features, such as tumor stage and treatment method, influence psychological outcomes in CC. Therefore, we retrospectively collected clinical data from patients with CC to analyze the factors associated with anxiety and depression and construct predictive models. The aim was to provide clinicians with a framework for assessing psychological risk and developing appropriate interventions.

MATERIALS AND METHODS

Research object

We retrospectively collected data from 119 patients with CC who received treatment in the Gynecology Department of Suzhou Ninth People’s Hospital between January 2017 and May 2025.

Inclusion criteria: (1) Met the diagnostic criteria for CC as stipulated in the guidelines[8] and confirmed via pathological examination; (2) Age ≥ 18 years; (3) No major family stress events occurred during diagnosis or treatment; and (4) Completed a hope scale assessment at diagnosis and underwent anxiety and depression scale assessments during treatment.

Exclusion criteria: (1) Presence of tumors at sites other than CC; (2) Severe cardiac, hepatic, or renal dysfunction; (3) Language or communication disorders; (4) Pre-existing psychiatric disorder before CC diagnosis; and (5) Incomplete medical records.

The Ethics Committee of Suzhou Ninth Affiliated Hospital of Soochow University (Suzhou Ninth People’s Hospital) approved this study.

Research methods

Data collection: Patients data were obtained from the hospital information system, including electronic medical records, pathology reports, treatment records, and questionnaire assessments. The following data were collected: Age, residence, marital status, education level, monthly family income, comorbidities (e.g., hypertension and diabetes), tumor stage, tumor pathology type, and treatment method.

Hope level assessment: Hope levels were assessed at diagnosis using the Brief Hope Measurement Instrument developed by Herth[9]. The scale comprises 12 items across three dimensions: Reality orientation, positive action, and interpersonal closeness. Each item was scored on a 4-point scale (1-4), with a total score of 12-48. Scores of 12-23 indicated low hope, 24-35 indicated medium hope, and 36-48 indicated high hope.

Anxiety and depression assessment: Anxiety was assessed using the Self-Rating Anxiety Scale (SAS), which consists of 20 items rated on a 4-point Likert scale. Fifteen items were score positively and five negatively scored. The raw total score was multiplied by 1.25 to obtain the standard score. A standard score < 50 indicated no anxiety; 50-59 mild; 60-69 moderate; and ≥ severe anxiety[10]. Depression was assessed using the Self-Rating Depression Scale (SDS), also with 20 items (10 positive, 10 negative). The standard scores were derived in the same manner as for anxiety. A score < 53 indicated no depression; 53-62 mild; 63-72 moderate; and ≥ 73 severe depression[11].

Statistical analysis

Data analysis was performed using SPSS version 25.0. Categorical data are expressed as n (%), with groups compared using the χ² test. Factors influencing anxiety and depression in the patients were explored using multivariate logistic regression to construct predictive model formulas. Model performance was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC). Goodness-of-fit was assessed with the Hosmer-Lemeshow test to exclude overfitting. A two-sided P < 0.05 was considered statistically significant.

RESULTS

Incidence of anxiety and depression in patients with CC and univariate analysis

The SAS scale assessment showed that the incidence of anxiety during treatment among the 119 patients with CC was 64.71% (77/119). The SDS scale assessment indicated that the incidence of depression was 52.10% (62/119). Analysis revealed significant differences in family monthly income, tumor stage, treatment modality, and hope level between CC patients with and without anxiety or depression (P < 0.05; Table 1).

Table 1 Univariate analysis of anxiety and depression in cervical cancer patients (n = 119), n (%).

Characteristics	Total	Anxiety assessment				Depression assessment
		Has anxiety (n = 77)	No anxiety (n = 42)	χ2	P value	Has depression (n = 62)	No depression (n = 57)	χ2	P value
Age (years)				2.747	0.097			1.780	0.182
< 60	66	47 (61.04)	19 (45.24)			38 (61.29)	28 (59.12)
≥ 60	53	30 (38.96)	23 (54.76)			24 (38.71)	29 (50.88)
Place of residence				0.135	0.713			0.555	0.456
Town/city	48	32 (41.56)	16 (38.10)			27 (43.55)	21 (36.84)
Rural area	71	45 (58.44)	26 (61.90)			35 (56.45)	36 (63.16)
Marital status				1.116	0.291			1.603	0.206
Married	99	62 (80.52)	37 (88.10)			49 (79.03)	50 (87.72)
Unmarried or divorced	20	15 (19.48)	5 (11.90)			13 (20.97)	7 (12.28)
Educational level				0.093	0.760			0.018	0.892
High school and below	87	57 (74.03)	30 (71.43)			45 (72.58)	42 (73.68)
College degree or above	32	20 (25.97)	12 (28.57)			17 (27.42)	15 (26.32)
Monthly family income				7.661	0.022			6.849	0.033
< 5000 yuan (RMB)	38	30 (38.96)	8 (19.05)			26 (41.94)	12 (21.05)
5000-8000 yuan (RMB)	54	28 (36.36)	26 (61.90)			22 (35.48)	32 (56.14)
> 8000 yuan (RMB)	27	19 (24.68)	8 (19.05)			14 (22.58)	13 (22.81)
Underlying disease				2.753	0.097			1.303	0.254
Have	24	19 (24.68)	5 (11.90)			15 (24.19)	9 (15.79)
Not have	95	58 (75.32)	37 (88.10)			47 (75.81)	48 (84.21)
Tumor staging				7.581	0.006			4.437	0.035
Phase I-II	59	31 (40.26)	28 (66.67)			25 (40.32)	34 (59.65)
Phase III-IV	60	46 (59.74)	14 (33.33)			37 (59.68)	23 (40.35)
Pathological type				0.627	0.731			1.370	0.504
Adenocarcinoma	24	14 (18.18)	10 (23.81)			10 (16.13)	14 (24.56)
Squamous cell carcinoma	90	60 (77.92)	30 (71.43)			49 (79.03)	41 (71.93)
Adenosquamous carcinoma	5	3 (2.90)	2 (4.76)			3 (4.84)	2 (3.51)
Tumor treatment methods				10.245	0.001			5.147	0.023
Monotherapy1	40	18 (23.38)	22 (52.38)			15 (24.19)	25 (43.86)
Combined treatment2	79	59 (76.62)	20 (47.62)			47 (75.81)	32 (56.14)
Hope level				15.139	0.001			11.783	0.003
Low	50	38 (51.95)	12 (23.81)			33 (53.22)	17 (29.83)
Medium	44	31 (37.66)	13 (35.71)			23 (37.10)	21 (36.84)
High	25	8 (10.39)	17 (40.48)			6 (9.68)	19 (33.33)

¹Indicates single treatment such as surgery alone, radiotherapy alone, or chemotherapy alone.

²It refers to a comprehensive treatment approach that combines surgery and radiotherapy, or surgery and chemotherapy, or surgery and radiotherapy and chemotherapy.

Multivariate analysis of anxiety and depression in patients with CC

Taking the presence of anxiety or depression during treatment (0 = no; 1 = yes) as the dependent variable, the statistically significant factors from the univariate analysis (family monthly income, tumor stage, treatment method, and hope level) served as independent variables (Table 2). Multivariate logistic regression showed that family monthly income < 5000 yuan (OR = 2.214, P = 0.022), stage III-IV tumor (OR = 1.811, P = 0.013), comprehensive treatment (OR = 2.989, P = 0.006), and low hope level (OR = 3.267, P = 0.001) were risk factors for anxiety (P < 0.05; Table 3). For depression, risk factors included family monthly income < 5000 yuan (OR = 1.540, P = 0.027), stage III-IV tumor (OR = 1.679, P = 0.030), comprehensive treatment (OR = 2.069, P = 0.012), and low hope level (OR = 2.988, P = 0.004) (P < 0.05; Table 4).

Table 2 Explanation of variable assignment.

Variable	Description of valuation
Monthly household income	0 ≥ 8000 yuan; 1 = 5000-8000 yuan; 2 ≤ 5000 yuan
Tumor staging	0 = phase I-II; 1 = phase III-IV
Tumor treatment methods	0 = monotherapy; 1 = combined treatment
Hope level	0 = high; 1 = medium; 2 = low

Table 3 Results of multivariate logistic regression analysis of anxiety in cervical cancer patients.

Variable	β	SE	Wald χ2	P value	OR (95%CI)
Family monthly income < 5000 yuan	0.795	0.347	5.249	0.022	2.214 (1.122-4.371)
Tumor stage III-IV	0.594	0.239	6.177	0.013	1.811 (1.134-2.892)
Comprehensive tumor treatment	1.095	0.398	7.569	0.006	2.989 (1.370-6.521)
Low hope level	1.184	0.359	10.877	0.001	3.267 (1.616-6.606)
Constant term	-9.176	2.593	12.523	< 0.001	-

Table 4 Results of multivariate logistic regression analysis of depression in cervical cancer patients.

Variable	β	SE	Wald χ2	P value	OR (95%CI)
Family monthly income < 5000 yuan	0.432	0.195	4.908	0.027	1.540 (1.051-2.257)
Tumor stage III-IV	0.518	0.238	4.737	0.030	1.679 (1.053-2.675)
Comprehensive tumor treatment	0.727	0.289	6.328	0.012	2.069 (1.175-3.644)
Low hope level	1.095	0.380	8.303	0.004	2.988 (1.419-6.297)
Constant term	-8.541	2.572	11.027	< 0.001	-

Prediction model construction

Based on multivariate analysis, risk prediction equations for anxiety and depression were derived. For anxiety: Logit (P) = 0.795 × family monthly income (0 ≥ 8000 yuan; 1 = 5000-8000 yuan; 2 ≤ 5000 yuan) + 0.594 × tumor stage (0 = I-II stage; 1 = III-IV stage) + 1.095 × treatment method (0 = single; 1 = comprehensive) + 1.184 × hope level (0 = high; 1 = medium; 2 = low) - 9.176.

The goodness-of-fit deviation test revealed no overfitting (χ² = 1.875, P > 0.05).

For depression: Logit (P) = 0.432 × family monthly income (0 ≥ 8000 yuan; 1 = 5000-8000 yuan; 2 ≤ 5000 yuan) + 0.518 × tumor stage (0 = I-II stage; 1 = III-IV stage) + 0.727 × treatment method (0 = single; 1 = comprehensive) + 1.095 × hope level (0 = high; 1 = medium; 2 = low) − 8.541.

The goodness-of-fit deviation test also indicated no overfitting (χ² = 2.073, P > 0.05).

The ROC curve analysis showed an AUC of 0.865 (95%CI: 0.792-0.938) for the anxiety model and 0.837 (95%CI: 0.763-0.911) for the depression model, indicating good discrimination ability (Figure 1).

Figure 1  Receiver operating characteristic evaluation of anxiety and depression prediction model for cervical cancer patients.

DISCUSSION

Cancer treatment is a long process, and patients are often troubled by the disease, making them prone to negative emotions such as anxiety and depression[12]. Thangarajah et al[13] reported that approximately 68.8% of patients with CC experience anxiety, 51.4% worry, and 26.3% panic, indicating that CC has become a significant psychological stressor that contributes to adverse psychological outcomes. In this study of 119 patients with CC, the incidence of anxiety during treatment was 64.71% and that of depression was 52.10%, findings consistent with those of Yang et al[7]. These symptoms may arise because patients with CC fear treatment failure, disease progression, and threats to life. Additionally, disease-related exhaustion and reduced social engagement may lead to feelings of helplessness, alienation from family or friends, and diminished emotional support, thereby contributing to depression and loneliness[14].

Our study identified low family monthly income (< 5000 yuan), stage III-IV tumor, comprehensive treatment, and low hope level as influencing factors for anxiety and depression in patients with CC. The reasons for this may include the following: (1) For low-income families, economic hardship combined with the stress of illness often heightens anxiety and depression, negatively affecting quality of life. Okyere Asante et al[15] similarly found that CC patients with lower income experience greater psychological pressure, supporting our results. Because CC treatment is both prolonged and costly, patients face not only direct medical expenses but also indirect financial losses due to reduced productivity, which further exacerbates psychological distress; and (2) Advanced tumor stage increases the extent of cancer spread, treatment complexity, and uncertainty of prognosis, leading patients to greater anxiety and depression. Tosic Golubovic et al[16] also observed a significant correlation between tumor stage and psychological distress, supporting our findings. Treatment methods (surgery, radiotherapy, and chemotherapy) vary across CC stages, and each can affect patient’s physical and psychological health differently. For example, a total hysterectomy may cause great stress and pain, heightening concerns about postoperative recovery and increasing anxiety and depression[17]. Furthermore, if surgical margins are positive or lymph node metastasis is present, radiotherapy is required to eliminate the residual carcinomas and reduce recurrence risk. However, close-field radiotherapy to the uterine cavity can cause toxic side effects, including radiation cystitis, radiation proctitis, and urinary organ changes, which may further increase patient’s pain and anxiety[18]. For patients unable to undergo surgical intervention, chemotherapy can inhibit the spread of cancer cells, but the drugs often cause nausea, vomiting, and fatigue. These physical discomforts directly affect patients’ mental state, increasing anxiety and depression[19]. This study also found that comprehensive tumor treatment was a factor influencing anxiety and depression in patients with CC. The likely reason is that combining multiple treatments produces cumulative side effects, which significantly affect patients during therapy. Consequently, CC patients receiving comprehensive treatment are more likely to experience anxiety and depression compared with those undergoing a single treatment. Chen and Bai[20] similarly reported that CC patients receiving combined treatment were more prone to psychological distress, indirectly supporting our results. Hope, as a belief, enables patients to maintain a positive outlook on the future and adjust their mindset to accept illness[21]. Patients with higher hope levels tend to believe more in the possibility of a cure, actively adapt their goals to their circumstances, and reduce the occurrence of negative emotions. Li et al[22] found that hope level correlated negatively with the negative emotional expression in CC patients-that is, higher hope corresponded with fewer negative emotions. Conversely, lower hope often results in pessimism, which promotes anxiety and depression. Therefore, in clinical practice, measures should be taken to enhance patients’ hope index, enabling them to better adapt to the disease and reduce negative emotions.

After CC diagnosis, many factors influence patients’ anxiety and depression during treatment, making the construction of an effective prediction model essential. In this study, multivariate logistic regression analysis was used to identify influencing factors, and predictive models for anxiety and depression in patients with CC were established from the regression coefficients and constant terms of each variable. ROC curve analysis and goodness-of-fit tests demonstrated high diagnostic efficacy. This may be because the study applied independent sample verification, which filtered out irrelevant or weakly related indicators, thereby reducing noise, while integrating multiple related indicators to provide complementary information and improve diagnostic performance.

The limitations of this study are as follows: (1) It retrospectively collected the data of patients with CC from a single center with a relatively small sample size, limiting representativeness; and (2) The prediction model has not undergone external validation, which may affect the generalizability of the findings. Future prospective and large-scale studies with external model validation are therefore required.

CONCLUSION

Low family income, advanced tumor stage, comprehensive treatment, and low hope level correlated positively with anxiety and depression in patients with CC. Constructing a predictive model based on these factors can effectively assess the risk of anxiety and depression and has guiding significance for clinical interventions aimed at reducing their incidence.