Fear of cancer recurrence eight years after early-stage breast cancer – results from a national survey

Abstract

Background

Fear of cancer recurrence (FCR) in breast cancer survivors (BCSs) is common, associated with reduced quality of life and effective interventions exist. There are knowledge gaps concerning FCR among long-term, early-stage BCSs and its associations with other late effects. Within a national cohort, we explored these knowledge gaps, with the ultimate aim of improved care for BCSs experiencing long-term FCR.

Methods

In this cross-sectional study, all BCSs aged 20–65 years with early-stage breast cancer in 2011–2012 (n = 2803), were identified by the Cancer Registry of Norway in 2019 and mailed a survey including the Assessment of Survivor Concerns used to measure FCR. Factors associated with moderate/high FCR (defined as a sum score of ≥ 6 of a possible range 3–12, or a single score on one of the items of ≥ 3) were explored using a three-block regression analyses including relevant sociodemographic-, health- and cancer-related variables.

Results

In total, 1311 BCSs were included (47%). Median age at survey was 60 years. Fifty-six % reported moderate-to-high FCR, associated with younger age (OR 0.96, 95% CI 0.95–0.97) and receiving chemo- and endocrine therapy (OR 1.59, 95% CI 1.15–2.20). After adding late effects into the model, FCR remained significantly associated with these variables, in addition to sleep disturbances (OR 1.58, 95% CI 1.18–2.10). In the final block, adding mental distress, FCR remained significantly associated with younger age (OR 0.97, 95% CI 0.96–0.99), receiving chemo- and endocrine therapy (OR 1.14, 95% CI 1.00–1.97), sleep disturbances (OR 1.44, 95% CI 1.08–1.94) and anxiety (OR 2.67, 95% CI 1.38–5.19).

Conclusions

FCR was prevalent eight years after early-stage breast cancer. Being younger, receiving intensive treatment, experiencing sleep disturbances and/or anxiety were associated with moderate/high FCR. Addressing FCR should be part of standard follow-up care of long-term BCSs.

Keywords:

• Fear of cancer recurrence
• breast cancer
• long-term survivorship
• late effects after breast cancer

Introduction

In 2020, more than 2.3 million women were diagnosed with breast cancer (BC) world-wide [1]. BC has excellent prognosis with a 5-year relative survival rate around 90% in several western countries [2]. Consequently, the population of breast cancer survivors (BCSs) is rapidly increasing. Fear of cancer recurrence (FCR), defined as fear, worry or concern relating to the possibility that cancer will return or progress [3] is believed to be one of the most prevalent unmet supportive care needs during cancer survivorship [4], consistently associated with reduced quality of life and an overuse of health care services [3,5] and affecting close to 60% of cancer survivors, according to a recent review [6].

During the first years after diagnosis, BCSs have been shown to have a higher risk of FCR compared to survivors of other cancers, and young BCSs seem to be especially at risk [7]. Experiencing mental distress and fatigue has been associated with increased risk of high FCR during this time [8]. A review on long-term cancer survivors (≥ 5 years) by Koch et al. concluded that most survivors do not experience a significant reduction in FCR as time goes on [9]. A prospective study by Schapira and colleagues reported that although the majority of young survivors of early-stage BC experience symptom improvement over time, one third report stable or worse FCR five years post diagnosis compared to time of diagnosis [10]. No predictors of worsening FCR during those first five years post-diagnosis were identified. Among long term, early-stage BCSs there are knowledge gaps with regards both to occurrence and associated factors.

Although BC treatment is increasingly personalized according to the individual tumor’s clinical characteristics and genetic profiling, follow-up care can be characterized as ‘one size fits all’. For clinicians, being able to identify BCSs vulnerable to high and persistent FCR is important as effective interventions are available, such as contemporary cognitive behavioral therapies [11]. The aim of this study was therefore to explore FCR and associated factors, including a range of late effects, in a national cohort of early-stage BCSs eight years after diagnosis.

Methods

Study design

The nation-wide, cross sectional “Survivorship Work and Sexual Health” (SWEET)-study was conducted in 2019 with the primary aims to explore late effects, work life and sexual health among long-term BCSs [12]. The Cancer Registry of Norway (CRN) identified all women diagnosed with early-stage (I-III) BC in 2011 or 2012, who were between 20–65 years of age at time of diagnosis. To be included in the study they had to be free of BC relapse and other cancers prior to and after their BC diagnosis (except from non-melanoma skin cancer and ductal carcinoma in situ). In total, 2803 women received invitation to participate, and one reminder was sent to the non-responders (n = 1684).

Primary outcome

FCR was examined using the self-report instrument Assessment of Survivor Concerns (ASC) [13]. The ASC has shown excellent validity and is recommended for use among long-term cancer survivors [13]. The ASC consists of five items where the first three items (sub-scale 1) are specific of FCR (recurrence, new cancer diagnosis, future diagnostic tests) and the last two items (sub-scale 2) measures general health worries (death and future health). Sub-scale 1 was used to measure FCR in this study. Each item is rated on a 4-point scale from 1 (not at all) to 4 (very much), yielding a sum score from 3–12, with higher values reflecting higher levels of FCR. For the purpose of this study, moderate/high FCR was defined as a sum score of ≥ 6, or a single score on one of the items on sub scale 1 of ≥ 3. These criteria were selected in collaboration with the developer of the ASC. Moderate/high FCR, as a binary variable, was the primary outcome for this study.

Explanatory variables

Sociodemographic information included age at survey, self-reported information on higher education (>12 years) and living arrangements (partnered or not) (Table 1). Self-reported somatic co-morbidity was based on a short-version of the Charlson comorbidity index [14], and categorized into no comorbid conditions, 1–2 or ≥ 3 comorbid conditions. Cancer related variables, including stage, tumor characteristics and surgical treatment were provided by the CRN. Information on other BC treatments (chemotherapy, radiation- and endocrine therapy) was based on self-report.

Table 1. Characteristics of the study sample of long-term breast cancer survivors, according to the presence of moderate/high fear of cancer recurrence or not* (n = 1311).

	All BCSs		High FCR		No high FCR
	n	%	n	%	n	%
	1311	100	737	56.2	574	43.8
Demographic and somatic health variables
Age at survey, years, mean (SD)	59.8	8.7	58.5	8.7	61.5	8.3
Living with partner	960	73.2	542	73.5	418	72.8
Education >12 y	676	51.6	378	51.3	298	51.9
Somatic co-morbidity
1-2 conditions	712	54.3	401	54.4	311	54.2
≥ 3 conditions	309	23.6	180	24.4	129	22.5
Clinical variables
Stage 1	589	44.9	298	40.4	291	50.7
Stage 2	472	36.0	281	38.1	191	33.3
Stage 3	102	7.8	68	9.2	34	5.9
Systemic treatment burden
No systemic treatment	234	17.8	101	13.7	133	23.2
Endocrine (E) treatment only	166	12.7	83	11.3	83	14.5
Chemotherapy (C) only	220	16.8	124	16.8	96	16.7
C + E therapy	679	51.8	421	57.1	258	44.9
Surgical treatment
Breast conserving	775	59.1	408	55.4	367	63.9
Mastectomy	536	40.9	329	44.6	207	36.1
Radiation therapy	1055	80.5	584	79.2	471	82.1
Late effects
Cognitive dysfunction^a	565	43.1	367	49.8	198	34.5
Fatigue^b	538	41.0	358	48.6	180	31.4
Pain^c	607	46.3	390	52.9	217	37.8
Sleep disturbances^d	444	33.9	301	40.8	143	24.9
Arm problems^e	414	31.6	278	37.7	136	23.7
Breast problems^e	305	23.3	206	28.0	99	17.2
Mental distres
Anxiety^f	94	7.2	81	11.0	13	2.3
Depression^g	240	18.3	184	25.0	56	9.8

BCSs: breast cancer survivors; FCR: fear of cancer recurrence; *moderate/high FCR: an ASC sum score of ≥ 6, or a single score on one of the items of ≥ 3; BCSs: breast cancer survivors; SD: standard deviation. ^adefined as EORTC QLQ-C30 score above below 75, ^bdefined as EORTC QLQ-C30 score above 39, ^cdefined as a EORTC QLQ-C30 score above 25, ^ddefined as a EORTC QLQ-C30 score above 50, ^edefined as 3 (quite a bit) or 4 (very much) for at least one of the items within each scale of BR-23, ^fdefined as a sum score ≥ 10 on the General Anxiety Disorder 7-item scale, ^gdefined as a sum score ≥ 10 on the Patient Health Questionnaire -9.

Pain, sleep disturbances, fatigue and cognitive functioning were assessed using symptom- and function scores from the quality-of-life questionnaire developed by the European Organization for Research and Treatment of Cancer (EORTC QLQ-C30, v. 3). Items are rated from 1 (not at all) to 4 (very much) and then transformed to 0–100 scales according to the scoring manual [15]. In order to ease the interpretation of scores, we applied threshold values for each of these late effects according to previously described recommendations [16]. These threshold values were as follows; 25 for pain, 50 for sleep disturbances, 39 for fatigue and 75 for cognitive functioning [16]. Symptom values above and function values below these thresholds were labeled as clinically relevant problems. Cronbach’s alphas were 0.87, 0.89, 0.72 for pain, fatigue and cognitive function respectively. Arm and/or breast symptoms were assessed using the breast cancer specific module (QLQ-BR23) [17]. Survivors reporting 3 (quite a bit) or 4 (very much) to at least one of the items within each scale were labeled as having arm/breast problems. Cronbach’s alpha was 0.79 for arm- and 0.77 for breast problem. A major depressive episode (referred to as depression from now on) was defined as a sum score ≥ 10 for the Patient Health Questionnaire 9-item scale (PHQ-9) [18]. Cronbach’s alpha for PHQ-9 was 0.85. A generalized anxiety disorder (referred to as anxiety from now on) was defined as having a sum score ≥ 10 on the General Anxiety Disorder 7-item scale (GAD-7) [19]. Cronbach’s alpha for GAD-7 was 0.87.

Statistical analysis

Descriptive statistics are presented for the total sample and according to presence of moderate/high FCR or not (yes/no), with frequencies and proportions for categorical variables and means with standard deviations for continuous variables. Responses for all five items of the ASC are given.

Factors associated with moderate/high FCR were explored using univariable and hierarchical multivariable logistic regression analyses. Based on univariate associations and a-priori knowledge, we developed a directed acyclic graph (DAG) illustrating possible causal pathways between the different variables and FCR, and potentially overlapping pathways indicative of multi-collinearity. Stage was omitted from the block as it is highly related to treatment. Variables were included block-wise in the multivariable model. In the first block we included socio-demographic and clinical variables previously shown to influence FCR, i.e. age and systemic treatment [5]. In the second block we included physical late effects, and the third block included anxiety and depression. In this way, we investigated the contribution of each block of variables to the outcome of interest. Associations are presented as odds ratios (ORs) with 95% confidence intervals (CI). P-values < 0.05 were considered statistically significant. There were no evidence of multi-collinearity or interaction between included variables. For the EORTC-QLQ-C30, QLQ-BR23 and the GAD-7 missing values were imputed based on mean values of answered items when at least 50% of the items had been completed [20]. For the PHQ-9 mean imputation procedure was performed if no more than two items were missing [21]. In the multivariate analyses, BCSs with missing information on any of the included variables, were omitted (n = 96). All analyses were performed using IBM SPSS statistics version 26.0.

Results

Of 2803 invited participants, 1361 (49%) completed the survey. Three responders were excluded due to incomplete consent, three due to self-reported BC recurrence and 44 due to missing information on FCR. The final sample thus consisted of 1311 BCSs (47%). Mean age at survey was 60 years (range 30–74 years). The majority had higher education (52%) and lived with a partner (73%). Pain and cognitive dysfunction were the most frequently reported late effects (46 and 43% respectively) (Table 1). The most prevalent cancer-related concern was BC recurrence, affecting 81% of the BCSs of whom 51% reported worrying ‘a little bit’ and close to 10% reported very much worry. Worrying about experiencing another cancer illness was the second most prevalent cancer related concern, affecting 78% to some degree (Table 2). The mean FCR score based on sub scale 1 of the ASC was 6.1 (SD 2.3) (results not shown).

Table 2. Assessment of Survivor Concerns (ASC) including sub scales for cancer worry and health worry with response alternatives among long term breast cancer survivors (n = 1311)*.

ASC items	Sub scale	Not at all (1)	A little bit (2)	Somewhat (3)	Very much (4)
I worry about:		n (%)	n (%)	n (%)	n (%)
1. Future medical tests?	Cancer worry	629 (48.0)	515 (39.3)	119 (9.1)	48 (3.7)
2. Another cancer illness?	Cancer worry	287 (21.9)	673 (51.3)	225 (17.2)	126 (9.6)
3. Disease recurrence?	Cancer worry	244 (18.6)	662 (50.5)	264 (20.1)	141 (10.8)
4. Dying?*	Health worry	561 (42.8)	520 (39.7)	143 (10.9)	83 (6.3)
5. General health?*	Health worry	237 (18.1)	689 (52.6)	259 (19.8)	122 (9.3)

*missing = 4 for item 4 and 5.

Eight years after diagnosis, 737 (56%) reported moderate/high FCR. Among BCSs reporting moderate/high FCR, a larger proportion had BC stage 3 (9 vs 6%) and had received both chemo- and endocrine therapy (57 vs 44%), compared to BCSs without moderate/high FCR (Table 1). All late effects were more prevalent among BCSs with moderate/high FCR, most pronounced for anxiety which was five times more prevalent in this group, compared to BCSs without moderate/high FCR (11 versus 2%) (Table 1).

In univariable regression analysis, moderate/high FCR was associated with younger age; for one year increase in age the odds of moderate/high FCR decreased by 4% (OR 0.96, 95% CI 0.95–0.97). BCSs’ treated with chemotherapy (OR 1.70, 95% CI 1.17–2.47) or chemotherapy and endocrine therapy (OR 2.15, 95% CI 1.59–2.91) had increased odds of moderate/high FCR compared to those not receiving systemic treatment (Univariate analysis, Table 3). Moderate/high FCR was associated with all late effects with the highest associations estimated for fatigue (OR 2.08, 95% CI 1.66–2.61), sleep disturbances (OR 2.09, 95% CI 1.64–2.66), depression (OR 3.08, 95% CI 2.23–4.25) and anxiety (OR 5.30, 95% CI 2.92–9.63).

Table 3. Univariable and hierarchical multivariable logistic regression analyses assessing associations between age, cancer treatment, somatic late effects and mental distress and moderate/high fear of cancer recurrence as dependent variable (n = 1215).

	Univariable analysis			Multivariable analysis, block 1			Multivariable analysis, block 1-2			Multivariable analysis, block 1-3
	OR	95% CI	P	OR	95% CI	P	OR	95% CI	P	OR	95% CI	P
Block 1
Age (years)	0.96	0.95–0.97	<.001	0.96	0.95–0.98	<.001	0.97	0.95–0.98	<.001	0.97	0.96–0.99	<.001
Endocrine (E) therapy*	1.32	0.88–1.96	0.18	1.42	0.93–2.16	0.11	1.39	0.91–2.14	0.13	1.34	0.87–2.07	0.18
Chemotherapy (C)*	1.70	1.17–2.47	0.01	1.36	0.91–2.03	0.13	1.31	0.87–1.97	0.20	1.28	0.85–1.93	0.24
C + E therapy*	2.15	1.59–2.91	<.001	1.59	1.15–2.20	0.01	1.42	1.02–1.99	0.04	1.40	1.00–1.96	0.05
Block 2
Cognitive dysfunction^a	1.89	1.51–2.37	<.001				1.06	0.79–1.42	0.72	0.98	0.73–1.33	0.91
Fatigue^b	2.08	1.66–2.61	<.001				1.16	0.84–1.61	0.36	1.08	0.78–1.51	0.64
Pain^c	1.86	1.49–2.32	<.001				1.29	0.97–1.73	0.08	1.27	0.95–1.71	0.11
Sleep disturbances^d	2.09	1.64–2.66	<.001				1.58	1.18–2.10	<0.001	1.44	1.08–1.94	0.01
Arm problems^e	1.97	1.54–2.52	<.001				1.28	0.95–1.73	0.10	1.23	0.91–1.65	0.19
Breast problems^e	1.91	1.45–2.50	<.001				1.23	0.9–1.70	0.20	1.18	0.86–1.63	0.31
Block 3
Anxiety^f	5.30	2.92–9.63	<.001							2.67	1.38–5.19	<0.01
Depression^g	3.08	2.23–4.25	<.001							1.45	0.97–2.16	0.07

CI: confidence interval; OR: odds ratio; P: p-value, significant associations highlighted in bold; *reference: no systemic treatment. ^adefined as EORTC QLQ-C30 score above below 75, ^b defined as EORTC QLQ-C30 score above 39, ^cdefined as a EORTC QLQ-C30 score above 25, ^ddefined as a EORTC QLQ-C30 score above 50, ^edefined as 3 (quite a bit) or 4 (very much) for at least one of the items within each scale of BR-23 ^fdefined as a sum score ≥ 10 on the General Anxiety Disorder 7-item scale, ^gdefined as a sum score ≥ 10 on the Patient Health Questionnaire -9.

In multivariable regression analyses, moderate/high FCR was associated with younger age (OR 0.96, 95% CI 0.95–0.97) and receiving chemo- and endocrine therapy (OR 1.59, 95% CI 1.15–2.20) (Block 1, Table 3). After adding late effects into the model (Block 2, Table 3), moderate/high FCR remained significantly associated with age (OR 0.97, 95% CI 0.95–0.98), receiving chemotherapy and endocrine therapy (OR 1.42, 95% CI 1.02–1.99) and sleep disturbances (OR 1.58, 95% CI 1.18–2.10). In the final block, adding mental distress into the model (Block 3, Table 3), moderate/high FCR remained significantly associated with younger age (OR 0.97, 95% CI 0.96–0.99), receiving chemo- and endocrine therapy (OR 1.14. 95% CI 1.00–1.97), sleep disturbances (OR 1.44, 95% CI 1.08–1.94) and anxiety (OR 2.67, 95% CI 1.38–5.19).

Discussion

We found that FCR remains a prevalent concern almost a decade after diagnosis of early-stage BC with 56% of survivors reporting moderate-to- high FCR. Younger age, treatment intensity, sleep disturbances and anxiety were all associated with moderate/high FCR.

FCR may to some degree be considered a universal psychological reaction to surviving cancer [22]. ‘Clinically relevant FCR’, often including moderate and high levels of FCR [23], is characterized by persistent and high degrees of preoccupation and worry concerning disease recurrence and a hypervigilance to bodily symptoms [24]. Even though not all of these aspects were evaluated in the present study, it is one of few studies attempting to capture clinically relevant FCR among survivors of early-stage BC, beyond the first five years of BC survivorship. Koch et al. which focused on long-term survivors in their review from 2013, reported that most cancer survivors, regardless of cancer type, experience FCR many years after diagnosis [9]. In a more recent meta-analysis examining the prevalence of FCR across all phases of survivorship (from diagnosis to >10 years since diagnosis), 59% reported at least a moderate level of FCR and 19% experienced high FCR [6]. This figure is comparable to our finding that 56% of BCSs report moderate/high FCR, but direct comparison is hampered due to the use of different measurement tools and cut offs for FCR. Our results also align well with a Danish study including older BCSs (mean age 63 years at diagnosis) reporting a prevalence of 55% for high FCR during the first five years after the BC diagnosis [25]. In another survey of young adult survivors (19–39 years) of several cancer forms using the ASC to measure FCR 16 years after diagnosis, BCSs had the highest FCR scores (mean score 5.9) [26]. This score is only slightly lower than what we observed here among adult 8-year BCSs (mean score 6.1) confirming that high levels of FCR may persist for years beyond diagnosis.

Younger age was the only demographic variable associated with moderate/high FCR in the present study, which is in line with previous reports across cancer forms [4,6]. In a recent study on young women (18–45 years) with early-stage BC, 70% reported high levels of FCR [27]. It has been proposed that the underlying reasons for the high observed levels among younger BCSs, may be that they have poorer coping skills and less experience with health issues compared to older BCSs, and more responsibilities as caregivers, for instance for small children, adding to their burden of worries [4,26]. Younger age has also been associated with an overestimation of recurrence risk [28]. It is however important to acknowledge that younger age is associated with more aggressive disease and poorer prognosis [29]. Younger BC patients are likely to be made aware of these considerations at treatment initiation, with the potential for heightened cancer worry, and increased attention to bodily symptoms and signs of recurrence. Younger BCSs may be in ‘double trouble’; having perhaps the most limited set of coping skills and the highest objective risk of recurrence. Balancing their FCR at an appropriate level may therefore be challenging.

Although all late effects were more frequently reported among BCSs with moderate/high FCR, only sleep disturbances and anxiety were associated with FCR in the multivariate model. We did not find evidence of significant correlation or interaction between FCR and anxiety. Thus, although they may share clinical commonalities, our results suggest they may represent distinct entities. Among BCSs with anxiety in this cohort, 86% also reported moderate/high FCR (results not shown), while among the BCSs with moderate/high FCR, only 11% reported anxiety. Thus, FCR appears to be relatively independent of more generalized forms of anxiety. Sleep disturbances is a major concern among BCSs, associated with reduced quality of life [30]. Psychological distress, such as anxiety, is associated with sleep disruption among BCSs [31]. The association between FCR and sleep disorders is less studied, but there are reports that FCR increase the risk of poorer sleep quality [32]. Consequently, FCR, anxiety and sleep disturbances seem related, and have the potential to exacerbate each other. A prospective research design will be valuable for future research in order to identify the triggering factor.

Higher levels of FCR is a distressing concern, which may have both psychological and behavioral consequences, such as quality of life impairments, excessive self-checking- and health care seeking behaviors [33]. Cost-effective and noninvasive interventions, such as contemporary cognitive behavioral therapy, has shown positive effects on FCR among BCSs in randomized studies [34–36]. Such interventions may improve potential clusters of symptoms such as FCR, sleep disorders and anxiety in combination [37]. In order to succeed with this approach, clinicians need to know how to identify survivors at risk of these potentially debilitating and longstanding complaints.

Strengths and limitations

This study is based on high-quality data from the CRN, ensuring a high degree of completeness and accuracy as reporting new cancer cases to the CRN is mandatory. The large sample size of more than 1300 BCSs eight years after diagnosis offers valuable insight on long-term BC survivorship. Also, this large sample size allowed for evaluation of a range of sociodemographic, clinical and health related variables.

There are also possible limitations that need to be addressed. This is a cross-sectional survey, and causal inference is therefore not possible to assess. We cannot know the temporal relationship of the observed associations, for instance whether anxiety and sleep disturbances precede FCR or vice versa. The response rate of 49% is considered acceptable for population-based surveys [38], but we cannot exclude the possibility of non-response bias. A methodologically similar nation-wide survey exploring late effects among young cancer survivors, with a response rate of 39%, found no evidence of non-response bias [39]. Due to data restrictions at the CRN, information on non-responders in the SWEET-study (n = 1448) was limited to tumor characteristics, surgical treatment and age, and a complete attrition analysis was therefore not possible. Compared to responders, non-responders were on average 1.3 years older at diagnosis, had lower tumor proliferation markers (mean Ki67-value of 27 versus 31), and an additionally 4% were Her-2 negative (85% versus 81%) compared to the responders. The groups did not differ according to tumor size, nodal involvement, hormone receptor status, or type of surgery. Given the design of the study we know that non-responders are of same gender and have had similar diagnosis in 2011/2012 as responders.

We chose the ASC as it is a brief and validated measurement of FCR among long-term cancer survivors. Since 2019 other measurements of FCR, including the Fear of Cancer Recurrence Inventory (FCRI) and the FCRI-severity subscale have become more widely used. Further, the cutoff applied to define moderate/high FCR in the present study, has not been validated, and the ASC has not been directly compared to the FCRI scale. Consequently there is still a risk of misclassifications and results may not be directly comparable. Moving forward, efforts should be focused on reaching a consensus on how we can optimally assess clinically relevant FCR

Conclusions

FCR is prevalent among long-term BCSs despite excellent long-term prognosis. Younger BCSs and BCSs struggling with anxiety and/or sleep disturbances represents particularly vulnerable groups in risk of moderate/high FCR long after their BC diagnosis. Our findings underline the need for continuous attention to FCR during follow-up care and for improved patient communication concerning recurrence risk across the survivorship trajectory.

Ethical approval

This study was approved by the Regional Committee for Medical Research Ethics (2018/2170), the Norwegian Cancer Registry, and the Data Protection Officer at Oslo University Hospital. All procedures were performed in accordance with the ethical standards of the national and institutional research committee and with the 1964 Declaration of Helsinki and its later amendments.

Informed consent

Informed consent was obtained from all participants included in the study.

Author contributions

Study concept and design: KVR and KV; acquisition of data: KVR, KV; statistical analysis: SS, SB, KV, RSF, HCL, KVR; interpretation of data: KV, KVR, SB, HCL, RSF, SS, CK. drafting of the paper: KV; critical revision of the paper for important intellectual content: KV, KVR, SB, HCL, RSF, SS, CK.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available on request from the corresponding author, KV. All data are available at the National Advisory Unit for Late Effects after Cancer Treatment, Department of Oncology, Oslo University Hospital, the Radium Hospital, Oslo, Norway. The data are not publicly available due to restrictions e.g. their containing information that could compromise the privacy of research participants.

Additional information

Funding

This study was funded by the Pink Ribbon Movement and Norwegian Breast Cancer Society through Grant No. 197255.