Work ability of survivors of breast, prostate, and testicular cancer in Nordic countries: a NOCWO study
- First Online:
- Cite this article as:
- Lindbohm, M., Taskila, T., Kuosma, E. et al. J Cancer Surviv (2012) 6: 72. doi:10.1007/s11764-011-0200-z
- 319 Views
Cancer can cause adverse effects on survivors’ work ability. We compared the self-assessed work ability of breast, testicular, and prostate cancer survivors to that of people without cancer. We also investigated the association of disease-related and socio-demographic factors and job-related resources (organizational climate, social support, and avoidance behavior) with work ability and looked at whether these associations were different for the survivors and reference subjects.
Working aged cancer patients diagnosed between 1997 and 2002 were identified from hospital or cancer registries in Denmark, Finland, Iceland, and Norway (Nordic Study on Cancer and Work). A cancer-free reference group was selected from population registries. We collected information on work ability and other factors from 1,490 employed survivors and 2,796 reference subjects via a questionnaire.
The adjusted mean value of work ability was slightly lower among the breast and prostate cancer survivors compared to the cancer-free population. Co-morbidity, chemotherapy, low workplace support, and low organizational commitment were associated with reduced work ability. Avoidance behavior from supervisors or colleagues was only related to work ability among the cancer survivors.
Conclusions and implications
More attention should be paid to assisting cancer survivors in work life, particularly those who have chronic diseases or have undergone chemotherapy. Although most factors affecting the work ability of the survivors and reference subjects were the same, survivors’ work ability seemed to be particularly sensitive to avoidance behavior. The results suggest that there is a need to improve communication at the workplace and develop supportive leadership practices in order to avoid isolating behavior towards cancer survivors.
KeywordsCancerDisabilityJob resourcesSocial supportSurvivorshipWork ability
Some cancer survivors experience deterioration in their ability to work and are therefore at higher risk of leaving the workforce earlier than people without cancer . According to a review, the amount of survivors reporting impairment of work ability due to cancer varied from about 20–30% . Long-term breast cancer survivors have been observed as having increased physical impairment and symptoms, but not consistently poorer psychosocial or work functions. After testicular cancer, surveys have found slight, if any, differences between survivors and norms in most physical domains, although vitality and social function remained lower in long-term survivors, as they reported more pain . Of prostate patients who remained employed at 12 months after diagnosis, up to 30% indicated that they were not able to perform specific job tasks because they had been treated for cancer . The work ability of cancer survivors has, however, been observed to improve significantly over time during the 18 post-diagnosis months [5, 6].
The results of studies comparing work ability between long-term cancer survivors and the cancer-free population have been inconsistent. Taskila et al.  observed no difference in mean work ability between female or male cancer survivors and their cancer-free reference groups 2–6 years after diagnosis. Gudbergsson et al.  also noted no difference in male survivors compared with their controls, whereas female cancer survivors had lower mean work ability than female controls. Short et al.  found that cancer-free survivors diagnosed with various types of cancer (an average of 46 months post-diagnosis) suffered from a higher rate of work limitations than individuals with no chronic conditions. Hansen et al.  also found that 4 years post-diagnosis, breast cancer survivors reported higher levels of work limitations than a non-cancer comparison group. These inconsistent findings may be related to differences in study populations (e.g., by age, cancer stage, or cancer type) or country-specific differences in social security systems.
The effects of cancer on work ability may be related to the disease itself and its treatment. The work ability of an individual is, however, not only a health-related issue but also associated with work life and society. The multiple dimensions of work ability consist of the resources of the individual, factors related to work, and the environment outside of work . In the general population, work ability has been associated with health and functional capacity, socio-demographic factors (age, marital status, education, and occupation), physical and mental workload, values and attitudes, and psychosocial factors at work . It is, however, not well known how these factors affect cancer survivors’ work ability and whether their effect differs between survivors and the cancer-free population. Among cancer survivors, problems in work life may also arise because of attitudes and lack of knowledge among their supervisors and co-workers .
The aim of our study was therefore to compare the self-assessed work ability of breast, testicular, and prostate cancer survivors to that of people without cancer in four Nordic countries. We also investigated the association of disease-related factors, socio-demographic factors, and country and job-related resources (social climate, support from co-workers and supervisors, avoidance behavior of co-workers and supervisors, and commitment to the work organization) with work ability and examined whether these associations were different for the survivors and the reference subjects.
The data of the current study were collected in connection with a larger research project: the Nordic study on Cancer and Work (NOCWO) [14, 15]. The aim of the project was to clarify the problems that people with cancer face in work life, to examine the effect of cancer on return to work, and to identify the factors that either facilitate or hamper the process of returning to work. The participating countries were Denmark, Finland, Iceland, and Norway. The study was approved by either the Scientific Ethics Committees or Data Protection Agencies of the countries.
Materials and methods
Patients suffering from breast cancer, testicular cancer, prostate cancer, or lymphoma during the period 1997–2002, but with a good prognosis, were identified from the files of one large hospital (Finland and Norway) or the cancer registry (Denmark and Iceland). These cancer types were selected because they are common among working aged people and usually have a high survival rate. Because the study focuses on work ability, eligibility was limited to the age of 25–57 at the time of diagnosis. The selected patients had to have a good prognosis (no advanced stage of the disease, metastases, or recurrence) and not be currently undergoing chemotherapy or radiotherapy. In Iceland, patients with recurrent disease were excluded, but information on the stage of disease or metastasis was not available.
A reference group was selected for the study to separate the cancer-specific effects from those resulting from other factors. Reference subjects were selected from the national population registries. They represented the general population living in the same district and of the same age and gender distribution as the cancer survivors.
About 1,000 patients were selected from each country. All the available patients diagnosed during the study period and fulfilling the criteria were included. The total number of the identified cancer survivors was 3,599 and that of reference subjects was 7,379. We excluded lymphoma patients from the present analyses because information on these was only collected in Finland and Iceland. This meant that the number of survivors in the data was 3,345.
Data on the study subjects’ employment, work ability, and socio-demographic, health-, and work-related factors were collected using a questionnaire. The original Finnish questionnaire was translated into the other Nordic languages in accordance with the EORTC Quality of Life Group Translation Procedure . Translations followed a forward–backward procedure, independently carried out by two native speakers of the target language. The translated questionnaires underwent pilot testing.
The questionnaires were mailed to the study subjects in 2003–2005, depending on the country. The length of the follow-up period from diagnosis to receiving the questionnaire varied from 1 to 8 years after cancer diagnosis. In Finland, two reminders were sent by post; in Denmark, two reminders were sent and a phone call made to the non-responders; and in Iceland and Norway, only one reminder was sent.
Work ability score
We assessed work ability by asking the participants to estimate their current work ability compared with their lifetime best, using the following question: “Assume that your work ability at its best has a value of 10 points. How many points would you give to your current work ability? (0 means that you cannot currently work at all).” This question (the work ability score) is an item included in the Work Ability Index, which is a validated tool for measuring self-assessed work ability . The score has been proven a good predictor of mortality and retirement due to work disability . It has also been found to predict future sick leave and health-related quality of life among women on long-term sick leave .
Disease-related and socio-demographic factors
Information on diagnosis, date of diagnosis, and chemotherapy treatment was obtained from the hospital or registry files. Treatment was classified into two categories, chemotherapy or other treatment, because no information was obtained on other types of treatment. Data on chemotherapy were available on all subjects except Danish testicular cancer patients. In the questionnaire, the participants were asked to mark on a list their current chronic diseases or injuries diagnosed by a physician. The conditions included injury or accident, musculoskeletal disease, cardiovascular disease, respiratory disease, mental disorder, neurological or sensory diseases, digestive disease, genitourinary disease, skin disease, tumor (elicited only from reference subjects), endocrine and metabolic diseases, blood diseases, or other disorders or diseases. Anyone who marked “other” was asked to specify their condition . Information was also elicited concerning marital status (single, married, cohabitating, or other), education and occupation. The participants were classified into four educational categories as follows: comprehensive school (approximate length 1–9 years), secondary school/vocational school (10–12 years), college degree (13–16 years), and university degree (over 16 years). The occupations were coded according to the International Standard Classification of Occupations ISCO-88.
The items measuring support from supervisors and co-workers, social climate at work, and commitment to the work organization were adopted from the general Nordic Questionnaire. This is a validated tool for measuring psychological and social factors at work, including job organization characteristics, as well as individual work-related attitudes . Social climate at work, commitment to the work organization, and support from supervisors were measured by three items, and support from co-workers was determined from two items. The scale for all the questions ranged from 1 (very seldom/little or not at all/disagree totally) to 5 (very often or always/very much/agree totally). The individual values of the items measuring the same resource were summed up and divided by the number of items to get the means of the resource.
Three lacking (or “negative”) job resources were developed for the study: avoidance behavior by colleagues, avoidance behavior by supervisor, and avoidance of informing. The measurement of this kind of discrimination was partly based on the Social Impact Scale, which was used in the study of Fife and Wright . The specific questions were developed on the basis of a qualitative interview study carried out in Finland among 26 cancer survivors of working age (unpublished data). Avoidance behavior by supervisor (e.g., “My supervisor avoids me”) was assessed using three items, avoidance behavior by colleagues (e.g., “My colleagues avoid me”), and avoidance of informing (e.g., “My supervisor/co-workers do not inform me of work-related issues”) was assessed using two items. The items were rated on a five-point scale ranging from 1 (not at all) to 5 (very much). The individual values of the items measuring the same lack of resources were summed up, and the sum was divided by the number of items to get the mean values. For the analyses, the means of supervisor’s and co-workers’ avoidance behavior were classified into three categories based on their distribution: none (1), low (>1.1–3.4), and high (≥ 3.5) avoidance behavior.
We applied analyses of covariance to current work ability scores stratified by gender. Explanatory variables in the models were socio-demographic factors, disease-related factors, country, and job-related resources. In order to find out whether the effect of the explanatory variables on current work ability scores was different among the survivors and their reference subjects, we included interactions between disease status (survivor or reference subject) and explanatory variables in the models. We also calculated adjusted means of work ability scores along with their standard errors. All statistical analyses were performed using the SAS statistical program package, version 9.1 .
The response rate was 72% among the cancer survivors and 58% among the reference subjects. Women participated more actively in the study (70%) than men (59%), and the survivors’ response rate varied from 60% (Iceland and Norway) to 83% (Finland) and that of the reference subjects from 44% (Iceland) to 69% (Denmark). For the analyses presented here, we included only those who were in paid work at the time of the questionnaire, whose cancer had not recurred, and who answered the question on work ability. We also excluded the reference subjects who reported having had cancer. Thus, the final analyzed population comprised 1,449 cancer survivors (957 breast cancer, 112 prostate cancer, and 380 testicular cancer survivors) and 950 male and 1,759 female reference subjects.
Characteristics of cancer survivors and reference group
Survivors, N (%)
Reference group, N (%)
Survivors, N (%)
Reference group, N (%)
Married or cohabiting
Number of other chronic diseases or injuries
Three or more
Work ability score
Mean scores of job resources among cancer survivors and reference group
Survivors, mean (SD)
Reference group, mean (SD)
Survivors, mean (SD)
Reference group, mean (SD)
Support from co-workers and supervisors (scale 1–5)
Social climate at work (scale 1–5)
Commitment to work organization (scale 1–5)
Avoidance behavior of supervisors (scale 1–5)
Avoidance behavior of co-workers (scale 1–5)
Avoidance of informing (scale 1–5)
Using multivariate analyses, we first investigated the association of the disease status (survivor or reference subject) and other explanatory factors with work ability. We tested all the variables that were related to work ability in univariate analyses for inclusion in the multivariate model. The final models included the following variables: age, marital status (men only), country, number of chronic diseases, education (men only), occupation (women only), support at work, commitment to organization, supervisor’s avoidance behavior and colleagues’ avoidance behavior (women only), and avoidance of informing (men only).
Adjusted means of work ability score according to disease status, age, country, chronic diseases, occupation, and avoidance behavior at the workplace among women (N = 2,716), estimated by covariance analysis
Breast cancer survivors
Diseases or injuries (other than cancer)
Three or more
Legislators, professionals, senior officials, and managers (reference)
Technicians and associate professionals
Service and care workers, sales personnel
Craft workers, plant and machine operators, assemblers and elementary occupations
Supervisors’ avoidance behavior
Colleagues’ avoidance behavior
Adjusted means of work ability score according to disease status, age, marital status, country, chronic diseases, education, and supervisor’s avoidance behavior among men (N = 1,442), estimated by covariance analysis
Testicular cancer survivors
Prostate cancer survivors
Married or cohabiting
Diseases or injuries (other than cancer)
Three or more
University degree (reference)
Supervisors’ avoidance behavior
Suffering from other chronic diseases was strongly associated with reduced work ability among both women and men (Tables 3 and 4). The more chronic diseases people had, the poorer was their work ability. Danish participants had better work ability than workers of other countries. Women employed in service, sales, or manufacturing work had lower work ability than women employed as professionals or managers (Table 3). Among the men, high age and low education was related to low work ability (Table 4). Low support from supervisor or colleagues, low organizational commitment, and a high level of avoidance behavior by one’s supervisor were all associated with low work ability among both men and women. In addition, among the women, colleagues’ high avoidance behavior was related to low work ability, whereas among the men, the associated factor was avoidance of informing.
We then included interaction terms between disease status (survivor or reference subject) and explanatory variables in our statistical models to find out whether the association of explanatory variables was different between the cancer survivors and those without cancer. Nearly all the associations of various factors (disease-related factors, demographic factors, country, and job resources) with work ability were similar both in the survivor and the reference group, i.e., these variables did not modify the association of disease status with work ability. For example, cancer survivors’ work ability was lower than that of the reference group both among the subjects who had no other diseases and those who had one or more diseases. It should be noted, however, that the breast cancer survivors with no other chronic disease reported better work ability (8.27) compared to the reference subjects who had some other chronic condition (7.89).
There was also a weak interaction among women between disease status and age (p = 0.058). Among breast cancer survivors, the youngest age group (30–44) had the lowest work ability score (adjusted mean, 7.27), whereas the middle (45–54) and older (55–64) age groups had higher scores (7.54 and 7.44). In the reference group, work ability decreased along with increasing age (from 7.80 to 7.61).
The analysis of the association of chemotherapy with work ability was restricted to the breast and testicular cancer survivors. A total of 42% of breast cancer and 46% of testicular cancer survivors had undergone chemotherapy. Breast cancer survivors who had had chemotherapy reported lower work ability than those who had not had chemotherapy (estimated mean adjusted for covariates 6.64 vs. 6.85, p = 0.04). Among testicular cancer survivors, there was no clear difference in work ability between those who had or had not had chemotherapy (adjusted estimated means 8.15 vs. 8.06, ns.).
On average, the mean of breast cancer survivors’ perceived work ability was only slightly lower than that of cancer-free women after controlling for the effect of other covariates (socio-demographic factors, country, other chronic diseases, and job-related resources). Among men, there seemed to be no clear difference in work ability between the survivors and the reference group. The findings confirm the results of earlier Nordic studies showing none or only a slight difference in the mean work ability score between the breast, prostate, and testicular cancer survivors and their cancer-free reference groups [7, 8]. When interpreting our results, it should be noted, however, that our study population consisted of employed breast, prostate, and testicular cancer survivors with a good prognosis. Thus, our conclusions do not necessarily reflect the experiences of cancer survivors in general.
Contrary to our results, breast cancer survivors were found to report greater work limitations than a non-cancer comparison group on average 4 years post-diagnosis in a preliminary web-based questionnaire study . The difference of the results may be related to the higher proportion of survivors treated with chemotherapy in this study (79%) than in ours (42%). In line with our results, quality of life studies have, however, shown that breast cancer survivors report scores comparable to or better than healthy controls, and most studies report good overall quality of life . The overall quality of life among testicular cancer survivors has also been found to be comparable to that of the general population on the group level . Similarly, no difference was seen in health status or days out of their usual role for prostate cancer patients compared with people without cancer .
Most factors related to work ability had a similar effect among both the cancer survivors and reference subjects. Low level of education and low occupational status, having other chronic diseases, and low level of job resources were related to low work ability in both groups. Earlier studies in the general population have also shown that well-educated people or white-collar workers perceive their work ability as better than those with less education or in blue-collar positions  perceive it. The only factors with different associations in the survivor and reference group were women’s age and supervisors’ or colleagues’ avoidance behavior in the workplace.
We found weak evidence of reduced work ability, especially in the youngest (30–44 years) age group of breast cancer survivors as compared to cancer free women. This finding is in accordance with several studies on the quality of life, which have shown that younger breast cancer survivors report lower health-related quality of life than equally aged women of the general population [3, 26–28]. Young survivors also seem to fare worse after breast cancer than older survivors in physical role function, bodily pain, social function, and mental health . Younger breast cancer patients are treated more aggressively because their disease relapses more often than that of older patients. Treatment-related menopause has been found to be particularly problematic for younger women and is associated with poorer emotional functioning .
All common chronic diseases have been shown to decrease the work ability of the general population . We also found a strong association between suffering from chronic diseases and reduced work ability among both the cancer survivors and reference subjects. For cancer survivors, work ability was lower than for the reference group both among those subjects who had no other diseases and those who had one or more diseases. However, the breast cancer survivors with no other chronic disease reported better work ability than the cancer-free reference subjects with some other chronic disease. Similar results were obtained in two other studies on work disability associated with cancer survivorship compared to disability associated with other chronic conditions. The disability rate for cancer survivors with no other chronic conditions was lower than that for adults with chronic conditions in one study . Another study indicated that the survivors’ work disability rate was lower than or similar to those for other chronic conditions. .
Breast cancer survivors who had undergone chemotherapy reported poorer work ability than those who had received other types of treatment, whereas a similar difference was not seen among the testicular cancer survivors. Our finding among breast cancer survivors is in line with most earlier studies, providing evidence of reduced work ability , lower quality of life scores, or higher levels of fatigue after chemotherapy treatment in disease-free breast cancer survivors [23, 33]. They are also in accord with the results of a review concluding that on a group level, treatment strategies do not really influence testicular cancer survivors’ quality of life [24, 32].
Danish subjects had better work ability than those in other Nordic countries. The effect of cancer on work ability did not vary by country, as indicated by the interaction analyses. In each country, the mean work ability score was slightly lower among the survivor than the reference group. A possible explanation for the higher work ability of Danish participants may be related to differences in legislation and social security systems between the countries. This explanation is supported by the higher proportion of retired survivors and reference subjects in Denmark than in other Nordic countries (unpublished data). In a study on sickness absence among public sector employees, Danish employees were also found to have higher work ability than Swedish employees . As a potential explanation, the authors suggest a more active labor policy in Sweden, leading to increased opportunities for people with chronic illness to remain in work in Sweden compared to Denmark.
Our findings on the importance of workplace support and commitment for the work ability of cancer survivors and the cancer-free population are in line with the results of other studies conducted in the general working population [35, 36] and cancer survivors . Supervisor and/or coworker support has also been found important for work ability or return-to-work process in some other medical conditions, for example, rheumatoid arthritis  and injury [38, 39]. The results of one study indicated that low social supervisor support predicted forthcoming work disability .
We also observed that avoidance behavior at the work place is related to low work ability, but only among the cancer survivors; colleagues’ avoidance behavior was related to female cancer survivors’ work ability; whereas supervisors’ avoidance behavior was related to male cancer survivors’ work ability. We have earlier shown that supervisor and colleague avoidance behavior is negatively associated with work engagement, especially among breast cancer survivors with low levels of optimism . Other studies have reported that discrimination experienced because of cancer affects the likelihood to preserve employment and to return to work [42, 43]. Claims related to job loss and terms of employment were also more likely in cancer survivors than employees with other types of impairments . A high level of avoidance behavior was, however, rare in our data, although about a quarter of the study subjects experienced some level of avoidance at the workplace. A recent qualitative study indicated that line managers and colleagues may be ill-equipped to deal with the ongoing needs of cancer survivors, not recognizing or being aware of the long-term impact of cancer and its treatment on the individual . Even though workplace attitudes to cancer appeared to be sympathetic and well-meaning, they were largely uninformed and short-lived once the demands of day-to-day work resumed.
The strengths of our study include the use of validated tools and the assessment of several potential confounding factors in the analyses. The limitations include the cross-sectional design, which does not allow interpretation in terms of causal associations. Another limitation is related to the moderate response rate, which was higher among the survivors (72%) than the reference subjects (58%). Thus, selective participation is possible and may have played a role in the fact that only slight differences were found between the survivors and the reference group. Restricting the analyses to countries with the highest response rate (Denmark and Finland), however, still yielded similar results.
A third limitation of the study is that our data included only participants employed at the time of the survey. It is possible that the difference in work ability between the survivor and reference group is larger in the non-employed than employed population if more cancer survivors than cancer-free subjects are selected out of the work force because of limited work ability. Thus, we may have underestimated the impact of cancer on work ability. This limitation may also have played a role for detecting only small differences in work ability between the survivor and reference group. Another potential explanation for the small difference in perceived work ability may be related to cancer survivors’ positive adaptation to the situation. Some studies have suggested that cancer patients with an objectively declining quality of life report no decrease in quality of life, due to an altered definition of health or the development of a coping style involving denial or minimization of distress . The potential effect of this response bias is likely to be small for our findings because the reference group also included subjects with some other serious chronic disease, which probably affected their reports on well-being.
On average, the work ability of the breast, prostate, and testicular cancer survivors who returned to work life was only slightly lower than that of the cancer-free population, although a subgroup of the survivors suffered from reduced work ability attributable to cancer. Most factors affecting the work ability of the survivors and reference subjects were the same, and their association on work ability was also similar in both groups. More attention should be paid to assisting cancer survivors in work life, especially young breast cancer survivors, and those that have other chronic diseases or have undergone chemotherapy. Supervisor and co-worker support seemed to play an equally important role for both the survivor and reference group, and avoidance of isolating behavior in the workplace was particularly important to survivors’ work ability. These findings suggest that there is a need to improve the lines of communication in the workplace and focus on developing supportive leadership practices in order to avoid isolating behavior towards cancer survivors.
We thank the Nordic Cancer Union (NCU) that supported the Nordic Study of Cancer and Work (NOCWO) and the following sponsors in the respective Nordic countries: Denmark, the National Labor Market Authority, the Danish municipal VAT foundation, and the Danish Graduate School in Public Health Science, the University of Copenhagen; Iceland, The Commercial Workers’ Union of Reykjavik, The Icelandic Nurses Association Science Fund, the Memorial fund of Ingibjörg Gudjónsdóttir under the auspice of the Icelandic Cancer Society, and the Icelandic Social Workers’ Association Science Fund; Finland, The Finnish Work Environment Fund and the Finnish Cancer Organizations; Norway, The Norwegian Foundation for Health and Rehabilitation (grant no. HO-54010/002). The principal investigators for the Nordic Study of Cancer and Work (NOCWO) were as follows: Denmark, K. Carlsen and C. Johansen; Finland, M.-L. Lindbohm, T. Taskila, E. Kuosma, R. Martikainen, P. Hietanen and J. Hakanen; Iceland, H. K. Gunnarsdottir, G. L. Rafnsdottir, N. Sigurdardottir, T. Jonsdottir, and H. Sigurdsson; Norway, S. B. Gudbergsson, A. Dahl and S. D. Fosså.
Conflicts of interest
The authors declare they have no conflicts of interest.