JONT

Introduction

Cancer survivorship often brings long-term psychological and emotional challenges, with one of the most persistent being the distress related to the possibility of cancer returning. For many survivors, particularly those with a history of breast cancer, concerns about recurrence can continue well after treatment completion and may influence psychological well-being, quality of life, and engagement with follow-up care [1, 2]. Although some level of worry is expected, sustained and intense concern has been associated with clinically significant symptoms of anxiety, depression, and impaired.

These worries are not limited to fear of death or physical decline. Survivors may experience distress related to body image, femininity, social roles, and personal identity, which are often deeply intertwined with cultural and societal expectations [3, 4]. This highlights the importance of using tools that assess both the emotional intensity and the multidimensional nature of recurrencerelated distress.

The Concerns About Recurrence Scale (CARS), developed by [5], offers a structured and validated way to capture such experiences. It includes both a general index of worry and subscales addressing specific domains: Health Worries, Womanhood Worries, Role Worries, and Death Worries. While the CARS has been successfully adapted into Dutch [6], Japanese [7]  and Spanish-speaking population [8], a Greek version has been lacking. Cultural factors, including communication styles, coping behaviors and attitudes toward illness, underscore the necessity of validating this tool within Greek-speaking populations.

The current study aimed to fill this gap by translating and culturally adapting the CARS for use among Greek breast cancer survivors. In addition to examining the internal consistency and factorial validity of the adapted scale, the study assessed its convergent validity with psychological distress indicators, specifically anxiety, depression, and perceived stress.

Methods

Participants

A total of 219 breast cancer survivors were recruited through convenience sampling from four oncology clinics located in Athens, Greece (Hippocration, Alexandra, Agios Savvas, and Elena Venizelou). Eligibility criteria included being 18 years or older, having completed primary cancer treatment (excluding ongoing hormonal therapy), and demonstrating sufficient fluency in Greek to complete self-report questionnaires. Patients with significant cognitive impairment or active psychiatric disorders were excluded.

Procedure

The study was approved by the ethics committee of the Medical School of the National and Kapodistrian University of Athens (Protocol no.2021032325). Data were collected between October 2024 and May 2025. After providing informed consent, participants completed a paper-based survey comprising demographic and clinical questions, the Greek-translated version of the Concerns About Recurrence Scale (CARS), the Hospital Anxiety and Depression Scale (HADS), and the Perceived Stress Scale (PSS14).

Translation and Cultural Adaptation

Translation of the CARS followed established cross-cultural adaptation guidelines. Two independent bilingual translators first translated the original English version into Greek. Discrepancies were reconciled through consensus, followed by a back-translation conducted by a native English speaker unfamiliar with the original scale. The back-translated version was reviewed for conceptual equivalence. To ensure clarity and cultural relevance, a pilot group of ten breast cancer survivors participated in cognitive debriefing interviews.

Measures

Concerns About Recurrence Scale (CARS)

The Concerns About Recurrence Scale (CARS) consists of 30 items [5]. The first four items assess the frequency, intensity, consistency, and functional impact of recurrence-related thoughts using a 6-point Likert scale. These items form a general index of recurrence concern severity. The remaining 26 items assess four domains—Health Worries, Womanhood Worries, Role Worries, and Death Worries—rated on a 5-point Likert scale from 0 (not at all) to 4 (extremely).

Hospital Anxiety and Depression Scale (HADS)

The Hospital Anxiety and Depression Scale (HADS) includes 14 items split into two subscales measuring anxiety and depression. Responses are scored on a 4-point scale (0–3), with subscale totals ranging from 0 to 21. The following cut-off points are commonly used to interpret symptom severity: 0–5 (normal), 6–8 (mild), 9–11 (moderate), and 12–21 (severe) [9]. The Greek version has demonstrated good psychometric properties in oncology populations [10].

Perceived Stress Scale (PSS-14)

The Perceived Stress Scale (PSS-14) measures perceived stress over the past month using 14 items rated on a 5-point scale from 0 (never) to 4 (very often). Although not formally validated in Greek oncology settings, it has shown good internal consistency in Greek samples [11].

Statistical Analysis

Descriptive statistics were used to summarize demographic, clinical, and psychometric data. Internal consistency for each scale and subscale was evaluated using Cronbach’s alpha and interpreted alongside item-level and structural evidence. Confirmatory factor analysis (CFA) was conducted to evaluate the fit of the original four-factor model using Jamovi version 2.6 [12]. Item-level functioning was examined through standardized CFA factor loadings and their statistical significance. Convergent validity was assessed using Pearson correlations between CARS subscales and HADS and PSS-14 scores. Group comparisons based on HADSdefined psychological distress levels were performed using oneway analysis of variance (ANOVA) with post hoc Tukey tests. Known-groups validity was examined by comparing CARS scores across subgroups defined by time since diagnosis, cancer stage, employment status, and income satisfaction. Due to the crosssectional design, temporal stability (test–retest reliability) could not be assessed.

Results

Participant Characteristics

The study included 219 breast cancer survivors, with a mean age of 52.60 years (SD ± 8.12; range: 39–73). Most participants lived in Athens (79.0%), were currently employed (76.3%), with a higher education degree (48.9%). The majority was married (56.6%) with children (76.7%). Clinically, 86.8% had early-stage breast cancer (Stage 0–II), and 57.5% were diagnosed within the past two years. The most common surgical treatment was lumpectomy (61.2%), followed by mastectomy (38.8%), with only 18.7% undergoing breast reconstruction. A total of 81.7% received radiation therapy, 58.9% received chemotherapy, and 74.9% were under hormonal treatment. Among those on hormonal therapy, 58.9% used aromatase inhibitors, and 39.2% were on SERMs. Regarding menopausal status, 65.3% were postmenopausal. A family history of breast cancer was reported by 30.1%, and 75.7% had at least one comorbid condition. Current smoking was reported by 33.3% and alcohol consumption by 48.4%. In more details see Table 1.

Table 1: Sociodemographic and Medical Characteristics of the Participants (N=219).

Descriptive Statistics and Internal Consistency

Table 2 presents descriptive statistics and reliability estimates for all measures. The CARS Total score (items 5–30) had a mean of 50.96 (SD ± 26.97) and demonstrated excellent internal consistency (Cronbach’s α = 0.97). Subscale internal consistency ranged from α = 0.85 (Death Worries) to α = 0.93 (Health Worries), with mean scores of 19.40 for Health Worries, 9.36 for Womanhood Worries, 11.15 for Role Worries, and 11.09 for Death Worries. The Overall Worries subscale (items 1–4) showed a mean of 11.54 (SD ± 5.28) and α = 0.93.

For the validation measures, HADS-Anxiety (Mean = 5.51, SD ± 3.98, α = 0.85), HADS-Depression (Mean = 16.12, SD ± 3.79, α = 0.81), and PSS-14 (Mean = 21.53, SD ± 9.31, α = 0.88) all demonstrated acceptable internal consistency. Item-level CFA results further supported adequate item functioning, with all items exhibiting statistically significant standardized loadings on their respective latent factors. Accordingly, high internal consistency estimates were interpreted in conjunction with the multidimensional factor structure and item-level evidence, rather than as standalone indicators of scale quality.

Table 2: Descriptive Statistics of Psychometric Measures.

Confirmatory Factor Analysis

Standardized factor loadings from the CFA supported the four-factor model, with all items loading significantly on their respective factors (range: 0.58–0.92; p < 0.001). Model fit indices indicated a moderate fit: χ² (269) = 1193.3, CFI = 0.810, TLI = 0.788, RMSEA = 0.125, SRMR = 0.087. Inter-factor correlations were high and significant (p < 0.001), with the strongest between Health and Death Worries (r = 0.97), followed by Role and Death Worries (r = 0.90). The confirmatory factor analysis model with standardized factor loadings and inter-factor correlations is presented in Figure 1.

Figure 1: Confirmatory factor analysis of the Greek version of the Concerns About Recurrence Scale (CARS-GR).

Latent variables are represented as ellipses: HW = Health Worries; WW = Womanhood Worries; RW = Role Worries; DW = Death Worries. Observed variables (CARS items) are shown in rectangles. Standardized factor loadings are displayed next to the singleheaded arrows, and standardized correlations between latent factors are displayed next to the double-headed arrows. All loadings and correlations were statistically significant at p < 0.001.

Convergent Validity

CARS Total scores showed moderate to strong correlations with psychological distress measures (all p < 0.001). The CARS Total correlated positively with HADS-Anxiety (r = 0.548) and PSS-14 (r = 0.347), and HADS-Depression (r = 0.440). Subscale correlations with HADS-A ranged from r = 0.304 (Overall Fear) to r = 0.533 (Womanhood Worries). Similar patterns were observed for PSS-14 and HADS-D. Details are presented in Table 3.

Table 3: Convergent Validity for CARS.

Correlations with Participant Characteristics

As shown in Table 4, younger age (r = -0.175**), shorter time since diagnosis (r = -0.185**), lower income satisfaction (r = -0.207**), and advanced cancer stage (r = -0.203**) were significantly associated with higher CARS scores. Radiation therapy (r = -0.167*), unemployment, and comorbidities (r = 0.173*) were also significant factors. Family status, residence, and surgical type showed smaller but significant associations. Menopausal status and educational level showed no consistent patterns.

Table 4: Correlations between CARS total score and subscales with characteristics of the study’s sample.

Known-Groups Validity

As shown in Table 5, CARS total scores significantly differed across both HADS-Anxiety and HADS-Depression subgroups.

For anxiety, participants in the Normal HADS-A group reported significantly lower CARS scores (Mean = 39.13, SD ± 24.93; a) compared to those in the Mild (Mean = 64.44, SD ± 21.06; b) and Moderate (Mean = 59.08, SD ± 21.26; b) groups. The Severe group had the highest scores (Mean = 75.78, SD ± 17.13; c), significantly different from all other groups.

For depression, individuals in the Mild (Mean = 79.22, SD ± 2.69; a) and Moderate (Mean = 64.90, SD ± 19.52; a) groups reported significantly higher CARS scores than those in the Severe group (Mean = 47.88, SD ± 26.77; b).

Table 5: CARS Total scores by HADS-A and HADS-D Groups.