Comprehensive Mapping of Psychosocial Burden in Breast Cancer: A Multicenter Cross-Sectional Study in Türkiye

Breast cancer is the most common cancer in women globally, with approximately 2.3 million new cases annually, representing 24.5% of female cancers and 15% of cancer-related deaths. By 2050, incidence is projected to surpass 3.2 million, a 38% increase from 2022^{1, 2}. This rise presents growing challenges for diagnosis and treatment, especially in resource-limited settings³.

Beyond its physical toll, breast cancer imposes a significant psychological burden. Studies report that 29-47% of the patients experience psychiatric conditions such as depression, anxiety, and distress, with up to 8% reporting suicidal ideation and 2-6% attempting suicide, particularly in terminal stages^{4, 5}. One in five women continues to experience depressive symptoms two years after diagnosis⁶.

The severity and frequency of these symptoms often reflect underlying sociodemographic factors. Younger age, low education, limited social support, and lack of a stable relationship are consistently linked to greater psychological distress^7-9. In low- and middle-income settings, these effects are intensified by poor healthcare access, mental health stigma, and fatalistic cultural views of cancer^{10, 11}.

These psychological factors do not only reduce quality of life but also impair treatment adherence, exacerbate fatigue and pain, and may increase mortality^{12, 13}. By worsening disease progression and treatment outcomes, these psychological factors increase the overall burden of breast cancer, which remains one of the leading causes of lost healthy years measured in disability-adjusted life years worldwide¹⁴. These findings underscore the need for psychological assessment in comprehensive breast cancer care. Yet in many oncology settings, focus remains on physical symptoms, while depression and anxiety are often overlooked or poorly managed¹⁵.

This study aimed to quantify the prevalence of psychological distress in breast cancer patients, examine related sociodemographic and clinical factors, and emphasize the importance of integrating psychosocial support into routine oncology care. It is among the few studies in Türkiye to use validated tools the Patient Health Questionnaire-9 (PHQ-9) and the Generalized Anxiety Disorder-7 (GAD-7) widely used in psycho-oncology and known for strong psychometric reliability^{16, 17}. Their standardized use enhances data accuracy and generalizability, supporting evidence-based care and informing national health policy.

Study Design and Setting

This multicenter, cross-sectional descriptive study was conducted in six healthcare institutions across Türkiye. Participant recruitment began after ethical approval on January 6, 2025, and continued until mid-March. Data were collected and analyzed for sociodemographic, clinical, and psychological variables.

The study was conducted at three university-affiliated medical centers Trakya University Faculty of Medicine Hospital, Bursa Uludağ University Faculty of Medicine Hospital, and Çanakkale Onsekiz Mart University Faculty of Medicine Hospital and three state hospitals: Çanakkale Mehmet Akif Ersoy State Hospital, Kastamonu Training and Research Hospital, and Edirne Sultan 1^st Murat State Hospital.

Participants and Inclusion Criteria

Eligible participants were women diagnosed with non-metastatic breast cancer who voluntarily agreed to participate. Inclusion criteria were: (1) age 18 years or older, (2) no current or past diagnosis of psychiatric disorder or dementia, (3) no history of alcohol or substance dependence, and (4) sufficient Turkish language proficiency for effective communication. Pregnancy and the presence of another malignancy were exclusion criteria. The sample size was calculated using Cochran’s formula, assuming a 30% prevalence of depression/anxiety, with a 95% confidence level and 5% margin of error, resulting in a minimum requirement of 323 participants. This estimate was based on the lower end of previously reported prevalence rates (4-5) and reflects a conservative approach. However, to enhance the study’s statistical power and enable subgroup analyses, the final sample size was increased to 460.

Data Collection Procedure

Data were collected through face-to-face interviews in outpatient clinics using structured questionnaires, which included three sections: sociodemographic data, clinical information, and standardized psychological tools (PHQ-9 and GAD-7). Participants were either receiving active oncological treatment (chemotherapy, radiotherapy, or endocrine therapy) or were in routine follow-up after primary treatment. All interviews were conducted by the researchers themselves, all of whom were directly involved in the clinical care of participants. Written informed consent was obtained from all participants after they were informed about the study’s purpose and procedures. Interviews took place in private rooms to ensure confidentiality and lasted approximately 10-20 minutes, depending on responsiveness.

Variable Definition and Grouping Strategy

To ensure analytical clarity and minimize multicollinearity risk, variables were organized into two main domains: (1) individual and sociodemographic characteristics, and (2) clinical variables related to disease progression and treatment. The inclusion of a broad range of variables was based on sample size capacity and subgroup balance.

Multicategorical variables were dichotomized when preliminary analyses showed no significant differences or when distributions were highly unbalanced. Grouping decisions followed conceptually meaningful thresholds and clinically relevant cut-offs reported in the literature [e.g., age: <50 vs. ≥50 years; body mass index (BMI): <30 vs. ≥30 kg/m²; marital status: married vs. single/divorced].

“Low income” was defined based on the net national minimum wage in Türkiye as of January 2025 (TRY 22,104.67; ~USD 631 at 1 USD: 35 TRY). Monthly household income below twice this amount (<TRY 44,209) was categorized as low income. Participants with income equal to or exceeding this threshold (≥TRY 44,209) were classified as having “moderate–high” income. Menopausal status was based on menstrual history; women with ongoing menstruation or <12 months of amenorrhea were classified as premenopausal. Subjective variables—physical activity, healthcare access, and perceived social support—were categorized using predefined criteria. Detailed definitions are provided in table footnotes.

Psychological Assessment Scales

Depression and anxiety symptoms were evaluated using two validated psychometric instruments: the PHQ-9 and the GAD-7. The PHQ-9 screens for major depressive disorder, while the GAD-7 assesses symptoms of generalized anxiety. Both scales measure symptom frequency over the past two weeks using a four-point Likert scale ranging from 0 (not at all) to 3 (nearly every day).

Total scores range from 0-27 for the PHQ-9 and 0–21 for the GAD-7, with higher scores indicating greater symptom severity. A cutoff score of ≥10 is widely accepted for both instruments as the threshold for clinically significant depression or anxiety. The Turkish versions of both scales have demonstrated strong psychometric validity in previous studies^{18, 19}. In this sample, both scales demonstrated high internal consistency, with Cronbach’s alpha values of 0.81 for the PHQ-9 and 0.89 for the GAD-7, indicating strong reliability.

This study received ethical approval from the Non-Interventional Research Ethics Committee of Trakya University Faculty of Medicine (decision no: 01/14, date: 06.01.2025). Institutional permissions were obtained from all participating centers prior to study initiation. All procedures adhered to the principles of the Declaration of Helsinki, and informed written consent was obtained from all participants.

Statistical Analysis

All statistical analyses were performed using IBM SPSS Statistics version 26.0 (IBM Corp., Armonk, NY, USA). Categorical variables were summarized as frequencies (n) and percentages (%), while continuous variables were expressed as means and standard deviations. Depression and anxiety were dichotomized based on established cut-off scores (≥10) for the PHQ-9 and GAD-7.

Initial group comparisons (e.g., depression vs. non-depression; anxiety vs. non-anxiety) were conducted using Pearson’s chi-square test. Variables with p-values <0.10 were included in univariate logistic regression to calculate crude odds ratios with 95% confidence intervals (CIs).

Variables significant at p<0.05 in univariate analysis were entered into a multivariate logistic regression model to estimate adjusted odds ratios (AORs). Model fit was assessed using the Hosmer-Lemeshow test, explanatory power by Nagelkerke R², and classification accuracy was calculated. Statistical significance was set at p<0.05.

Sample Characteristics

A total of 460 women with non-metastatic breast cancer were included. Mean age was 54.8±12.1 years (range: 23-87), and 10.9% were under 40. Most participants were married (73.5%), postmenopausal (65.9%), and unemployed (77.2%). Regarding education, 51.5% had only primary education, while 19.1% held a university degree or higher. The average education duration was 8.0±4.8 years.

In terms of income, 28.3% reported a household income below the minimum wage. Urban residency was reported by 70.7%, and 14.6% described healthcare access as “very difficult”. Among the participants, 27.2% had no children, and 24.1% reported low perceived social support.

Clinically, 34.1% were premenopausal. Obesity (BMI ≥30 kg/m²) was seen in 37.6%, and 27.4% had low physical activity. Poor sleep quality was reported by 17.8%. Cancer staging showed 44.1% in Stage II and 34.6% in Stage III.

Surgery was performed in 90.7% of cases, with breast-conserving surgery being most common (45.9%). Chemotherapy had been administered to 88.7%, and 12.6% were receiving active chemotherapy during data collection. Radiotherapy was given to 68.3%. Lymphedema was present in 23.7%, with 2.8% reporting severe symptoms. Time since diagnosis was less than two years in 59.3% of cases (complete descriptive data are provided in Table 1 A,B).

Depression and Anxiety Scores

The mean PHQ-9 score was 6.61±4.42, and the mean GAD-7 score was 5.32±4.21. Using a cut-off of ≥10, 113 participants (24.6%) screened positive for clinically significant depression, and 125 (27.2%) for anxiety. Regarding depression severity, 19.8% reported mild symptoms, 3.3% moderate, and 1.5% severe. For anxiety, 24.8% had mild and 3.0% had moderate to severe symptoms. Among those below the clinical threshold, minimal symptoms were most common 37.0% for depression and 48.5% for anxiety (see Table 2 for full distribution details).

Factors Associated with Depression

Separate multivariate logistic regression models were conducted to assess associations between depression and sociodemographic (Table 3A) and clinical variables (Table 3B).

In the sociodemographic model, younger age (<50 years) (AOR: 4.68, 95% CI: 2.53-8.67), childlessness (AOR: 2.47, 95% CI: 1.40-4.37), low income (AOR: 3.35, 95% CI: 1.72-6.52), limited healthcare access (AOR: 3.34, 95% CI: 1.95-5.70), and low perceived social support (AOR: 6.38, 95% CI: 3.61-11.26) were significantly associated with higher odds of depression. Although marital status was significant in univariate analysis (crude odds ratio: 1.97, 95% CI: 1.25-3.12), it did not remain in the multivariate model. Model performance showed a Nagelkerke R² of 0.307, classification accuracy of 83.7%, and a Hosmer-Lemeshow p-value of 0.005, indicating limited model fit.

In the clinical model, significant predictors included premenopausal status (AOR: 2.86, 95% CI: 1.78-4.61), poor sleep quality (AOR: 2.18, 95% CI: 1.09-4.38), lymphedema (AOR: 2.55, 95% CI: 1.53-4.25), advanced cancer stage (AOR: 1.65, 95% CI: 1.03-2.64), and active chemotherapy (AOR: 2.61, 95% CI: 1.39-4.89). BMI and physical activity were significant only in univariate analysis. The clinical model showed good fit (Hosmer–Lemeshow p= 0.696), with a Nagelkerke R² of 0.174 and classification accuracy of 75.9%.

Factors Associated with Anxiety

Separate multivariate logistic regression models were conducted to examine associations between anxiety and both sociodemographic (Table 4A) and clinical variables (Table 4B).

In the sociodemographic model, participants under 50 years (AOR: 2.93, 95% CI: 1.81-4.77), those with higher education (AOR: 2.43, 95% CI: 1.49-3.96), limited access to healthcare (AOR: 3.84, 95% CI: 2.33-6.34), and low perceived social support (AOR: 4.34, 95% CI: 2.57-7.34) showed significantly higher odds of anxiety. Other factors such as marital status, employment, income, and parental status were not significant in the adjusted model. Model fit was acceptable, with a Nagelkerke R² of 0.220, classification accuracy of 77.4%, and Hosmer–Lemeshow p= 0.079.

In the clinical model, significant predictors included premenopausal status (AOR: 3.05, 95% CI: 1.92-4.87), poor sleep quality (AOR: 2.42, 95% CI: 1.38-4.25), advanced-stage cancer (AOR: 1.85, 95% CI: 1.16-2.96), active chemotherapy (AOR: 2.09, 95% CI: 1.05-4.15), and shorter time since diagnosis (≤24 months) (AOR: 2.84, 95% CI: 2.10-3.76). Physical activity, BMI, and hormone therapy were significant only in univariate analysis. The model showed acceptable fit (Hosmer-Lemeshow p= 0.076), with a Nagelkerke R² of 0.200 and classification accuracy of 77.2%.

A visual summary of the multivariate logistic regression models is presented in Figure 1 for depression and Figure 2 for anxiety, based on AORs and 95% CIs for both sociodemographic and clinical predictors.

Reciprocal Association Between Depression and Anxiety

A significant bidirectional association was found between depression and anxiety. Among those with depression, 57.5% (65/113) also reported anxiety, compared to 17.9% (60/347) without depression. Conversely, 52.0% (65/125) of participants with anxiety showed depressive symptoms, versus 14.3% (48/335) without anxiety (p<0.001 for both).

Bidirectional logistic regression confirmed that the presence of anxiety significantly increased the odds of depression, and vice versa (OR: 6.48; 95% CI: 4.07–10.32; p<0.001). Both models showed a classification accuracy of 76.5% and a Nagelkerke R² of 0.195, indicating moderate explanatory power.

Anxiety is a prolonged state of alertness to perceived threats, while depression manifests as low mood, apathy, and reduced motivation²⁰. In our study, 24.6% of the women with non-metastatic breast cancer showed depressive symptoms, and 27.2% experienced anxiety. These figures are consistent with European data (depression 20-35%, anxiety 25-40%), but rates are markedly higher in low- and middle-income countries up to 62.6% and 77.4% in Morocco, and 83% for depression in Pakistan^{21, 22}. This gap may result from limited healthcare access, cultural norms, lack of validated tools, and methodological variation. Stigma, fatalism, and weak social support further hinder help-seeking and exacerbate distress^21-25. Altogether, these factors underscore the complex, multifactorial roots of psychological burden in underserved regions.

There was a notable overlap between depressive and anxiety symptoms, suggesting a bidirectional relationship. Patients experiencing one often reported the other. This co-occurrence may partly reflect shared symptom domains in assessment tools, potentially inflating comorbidity estimates. Still, some studies propose a linear progression typically with anxiety preceding depression^{26, 27}. Our lower depression rates may align with the Learned Helplessness Model, which posits that prolonged anxiety can evolve into depression over time²⁸. Depression prevalence can reach 66.1% during the remission phase²⁹. These findings highlight the need for continuous monitoring of symptoms and psychological support throughout cancer care. Furthermore, identifying high-risk groups remains essential within this symptom interaction framework.

Depression and anxiety were significantly associated with sociodemographic factors, particularly younger age. Among the participants under 50 years of age, depression and anxiety rates were 34.3% and 40.4%, compared to 19.0% and 19.7% in older adults (p= 0.001; AOR: 4.68 and 2.93). This disparity may reflect greater psychological burden linked to fertility concerns, career disruption, and body image issues among younger women. Similar patterns have been reported in previous studies^30-32. However, some studies suggest psychological symptoms increase with age, with depression risk rising 0.57% per year of age and 2.25-fold after 55^{21, 33}. These differences may be linked to factors like loneliness, chronic illness, or social isolation in older adults. Thus, age seems to be a context-sensitive rather than universal predictor of psychological vulnerability.

Perceived social support was the strongest psychosocial predictor of depression (46.8% vs. 17.5%, AOR: 6.38) and anxiety (44.1% vs. 21.8%, AOR: 4.34). These findings align with previous research emphasizing its protective role in psychological well-being^{23, 34, 35}. Family structures often serve as key sources of support. In this context, childlessness emerged as a significant risk factor for depression (AOR: 2.47). Marital status was significant in univariate analysis (p= 0.005; COR: 1.97) but not multivariate, suggesting that relationship quality may matter more than marital status alone. Prior studies confirm this, showing that dysfunctional relationships can exacerbate distress and that parenthood does not always offer protection^{22, 36, 37}. This is especially true when caregiving burdens are high, or when children are young, dependent, or emotionally impacted by the illness. Thus, social support should be evaluated not by the presence of family members alone, but by the emotional quality of those relationships, informing more personalized psychosocial care.

Socioeconomic status is a fundamental structural determinant of psychological symptoms. In our study, low income was significantly associated with depression (27.7% vs. 18.5%, AOR: 3.35), aligning with previous findings^{8, 22, 23, 38}. However, income was not significantly associated with anxiety in multivariate analysis, possibly due to confounding by factors like social support. Limited healthcare access strongly predicted both depression (33.8% vs. 19.8%, AOR: 3.34) and anxiety (41.4% vs. 19.8%, AOR: 3.84), consistent with prior research³⁹. This suggests that reduced access heightens uncertainty and perceived loss of control, thereby intensifying psychological distress. Education level was unrelated to depression (p= 0.074) but unexpectedly associated with higher anxiety risk (35.6% vs. 22.0%, AOR: 2.43), contrary to studies suggesting a protective role⁴⁰. This may reflect differences in awareness, expectations, and coping styles among highly educated individuals.

Several sociodemographic variables commonly associated with psychological symptoms in previous studies were not significant in our multivariate analysis. Employment status, though often discussed, may have variable effects depending on job security, autonomy, and emotional demands all of which can influence psychological vulnerability^{35, 41}. Similarly, no significant link between residence and depression or anxiety was observed. While rural areas are commonly associated with higher psychological burden due to isolation and limited care access³⁹, this may not apply in settings where rural urban gaps are smaller. Family history of breast cancer also showed no significant association, although some studies have reported greater distress in such cases⁴². These findings emphasize that the psychological impact of sociodemographic factors is not universal but shaped by context, individual perception, and cultural norms. Emerging evidence also suggests that stressful life events may increase breast cancer risk, highlighting the broader role of psychosocial stressors in both emotional and biological processes⁴³.

However, sociodemographic factors alone do not fully explain psychological vulnerability. Clinical factors also significantly influence emotional outcomes. Premenopausal status was significantly associated with higher rates of depression (35.0% vs. 19.1%, p= 0.001; AOR: 2.86) and anxiety (41.4% vs. 19.8%, p= 0.001; AOR: 3.05). Hormonal fluctuations may contribute to mood vulnerability by affecting regulatory pathways, although psychosocial factors likely play a more prominent role^{38, 44}. This underscores the importance of early identification of younger patients experiencing greater emotional burden.

Sleep quality is closely associated with psychological symptoms. Poor sleep was significantly linked to both depression (35.8% vs. 18.2%, p= 0.002; AOR: 2.18) and anxiety (38.3% vs. 17.6%, p= 0.001; AOR: 2.42). Sleep disturbances may disrupt mood regulation via neurohormonal pathways. Literature describes a bidirectional relationship: disrupted sleep alters serotonin and cortisol, while depressive and anxious symptoms impair sleep initiation and maintenance^{38, 45}. In cancer care, fatigue, pain, and anticipatory anxiety may intensify this cycle⁴⁶. A prospective study reported that depression, fatigue, and sleep disturbances often co-occur as a symptom cluster in breast cancer patients⁴⁷. Routine screening of sleep quality may help identify at-risk patients before more severe symptoms develop.

Stage III patients showed higher rates of depression (31.4% vs. 20.9%, p: 0.017; AOR: 1.65) and anxiety (35.2% vs. 22.9%, p: 0.006; AOR: 1.85), even in non-metastatic cases. Another study found that Stage IV patients had nearly twice the depression risk of earlier stages (OR: 1.9, p: 0.003)⁴⁸. These findings suggest that disease stage acts not only biologically but also as a psychological stressor, driven by uncertainty, intensive treatment, and prognosis concerns.

Receiving chemotherapy was significantly associated with depression (41.4% vs. 22.1%, p: 0.003; AOR: 2.61) and anxiety (53.4% vs. 23.4%, p: 0.001; AOR: 2.09). Side effects like hair loss, nausea, fatigue, early menopause, and neuropathy may lower quality of life and trigger depressive symptoms. Additionally, the treatment’s cyclical nature and frequent hospital visits may reinforce the “patient role,” heightening feelings of lost control. Our findings align with previous research showing chemotherapy’s emotional burden extends beyond physical effects to include symbolic and psychological dimensions^{38, 49-51}.

Surgical type showed no significant association with psychological symptoms (p>0.05), implying that postoperative complications especially lymphedema may be more influential²². Lymphedema was significantly associated with depression (35.8% vs. 21.1%, p: 0.003; AOR: 2.55), likely due to chronic pain, mobility limitations, and body image concerns, as previous studies suggest^{52, 53}. Many women avoid form-fitting clothing, which may undermine self-image and social confidence⁵⁴.

Anxiety was more common within the first two years post-diagnosis (32.6% vs. 19.3%, p: 0.002; AOR: 2.84, 95% CI: 2.10-3.76). This period may reflect a psychologically vulnerable window due to diagnostic shock, treatment adjustment, and abrupt lifestyle changes. Previous studies support this; for instance, in a five-year follow-up, anxiety peaked before treatment (38.0%) and fell to 25.3% by the first year’s end⁵⁵. The absence of a significant link between time since diagnosis and depression suggests symptoms may develop gradually, underscoring the need for sustained psychosocial monitoring.

Clinical factors like inactivity, high BMI, hormone therapy, and poor performance status were initially significant but lost relevance in multivariate analysis. Other variables type of surgery, family history, and residential setting showed no significant association. Yet literature indicates they may still affect depression and anxiety. At least 2.5 hours of weekly physical activity is associated with lower depression risk²². High BMI correlates with fatigue, poorer quality of life, and depression⁵⁶. Aromatase inhibitors may increase depression risk by 27-41%⁵⁷. Mastectomy without reconstruction is also tied to higher rates of depressive symptoms⁵⁸. The complex, interrelated nature of psychological distress highlights the need to assess mental health factors within a broad multivariate framework.

Study Limitations

This study benefits from a large sample, multicenter design, and validated psychological measures. However, several limitations warrant consideration. Its cross-sectional design limits causal inference and captures only time-specific associations. Depression and anxiety were self-reported, introducing potential biases like social desirability or recall error. Key psychosocial variables e.g., perceived support and healthcare access were measured with non-validated, study-specific tools, limiting comparability. Important domains like body image, sexual function, death anxiety, and post-traumatic stress were not assessed. Lastly, because participation was voluntary, individuals with higher psychological burden may have been underrepresented, which may affect the generalizability of the results.

This study shows that depression and anxiety are common even among non-metastatic breast cancer patients. Psychosocial and clinical factors like younger age, weak support, financial strain, and ongoing treatment worsen emotional distress. Psychological vulnerability often begins at diagnosis and deepens with uncertainty, isolation, and treatment stress. These findings call for a rethink of oncology models that prioritize tumor control but overlook mental well-being. Better outcomes demand a holistic approach with routine psychosocial screening and support. Addressing mental and physical health together may improve adherence, coping, and reduce disparities. This multicenter study not only informs future research but also urges urgent integration of psychosocial care into oncology where mental health is essential, not optional.