Pain, fatigue, anxiety and depression in older home‐dwelling people with cancer

Abstract Aim Globally, cancer incidence counts for more than 14 million cases and the number increases with age. The aim of this study was to investigate the occurrence of pain, fatigue, anxiety and depression in association with demographic and clinical factors. Design A cross‐sectional descriptive design was used. Methods We performed descriptive statistics to analyse the questionnaires completed from 174 older home‐dwelling people with cancer. Results The people with cancer reported low occurrence of pain, fatigue, anxiety and depression. We found strong correlation between anxiety and depression. Women reported significantly higher scores of anxiety and depression than men. A higher pain score was associated with higher scores of fatigue and anxiety. Conclusions Home care personnel meeting older home‐dwelling people with cancer should emphasize these people’ mental health problems and be aware that pain, fatigue and anxiety may occur at the same time.


SOLVIK et aL.
Cancer illness effects the diseased person both physically and mentally due to the treatment side effects and the stress connected to the diagnosis (Dawn, 2011;Farrell, Brearley, Pilling, & Molassiotis, 2013). Anxiety and depression in people with cancer can reduce the quality of life and may contribute to worse disease outcome as compared people not experiencing mental health problems (Utne, Miaskowski, Bjordal, Paul, & Rustoen, 2010). Mental illness can be associated with increased mortality in people with cancer (Chan, Wan Ahmad, Yusof, Ho, & Krupat, 2015). Anxiety and depression in people with cancer are more frequently seen among women than men and among those having impaired activity status (Hellstadius et al., 2016). Fatigue is hard to understand due to its expression where the feeling of tiredness might be present despite completed treatment (Peters et al., 2014;Shao et al., 2016;Wang et al., 2014;Zeng et al., 2012).
Recent health reforms in Norway emphasize that the inhabitants should live at home for as long as possible, even when treated for serious illness. The local health authorities have to take responsibility of delivering health services to older people with cancer living at home. However, after this healthcare reform we need to know more about symptoms and complaints among those receiving health services from the community. The aim of this study was therefore to examine the occurrence of pain, fatigue, anxiety and depression in association with demographic factors in a selection of older homedwelling people with cancer.

| ME THODS
A cross-sectional descriptive design was used. The sample consisted of 174 home-dwelling Norwegian people (mean age: 77.4 years) who had received their first-time cancer diagnosis. The questionnaire was completed during discharge from the hospital to the home, or during the week after arriving home. Inclusion criteria for the participants were as follows: (a) ≥65 years of age; (b) living at home; and (c) needing assistance from home care nursing. Clinical variables are as follows: cancer diagnosis (type of cancer); time since diagnosis (dichotomized into "<1 year" and "≥1 year"); ongoing treatment (receiving cancer treatment: "yes" or "no"); functional level (dichotomized into "normal"-normal activity and "limited"-limited activity/up more than half part of the day, needing assistance/only in bed, needing full assistance); BMI (dichotomized into "≥22" and "<22"); and comorbidity (other diseases: "yes" or "no"). Demographic variables are as follows: age (dichotomized from median age of the sample: "≤77 years" and ">77 years"); civil status (dichotomized into "paired"-have a cohabitant/married, having a partner and "non-paired"-widow/widower/ separated/divorced); education (dichotomized into "<13 years" and "≥13 years"); and network (someone to talk to -family or friends: "yes" or "no").
The questionnaire, the Norwegian version of the Edmonton Symptom Assessment System (ESASr), is a psychometric tested instrument on symptom burden including pain, tiredness/fatigue, anxiety and depression (Bergh, Aass, Haugen, Kaasa, & Hjermstad, 2012;Bruera, Kuehn, Miller, Selmser, & Macmillan, 1991). The instrument has a scoring range from 0 (no complaints)-10 (worst complaints) and the assessments were self-reported, as recommended by gold standard in research (Polit & Beck, 2017). Cross-culturally the concept "tiredness" in the ESAS-r is explained as "lack of energy" corresponding to the word "fatigue" (www.palli ative.org/NewPC/pdfs/ESAS-r.pdf). However, translation of instruments is challenging both linguistically and culturally (Utne et al., 2017), particularly when languages have no exact word for the phenomenon.
By means of the SPSS, version 24, we performed descriptive statistics such as frequencies, comparison between groups (t tests for continuous variables and Chi-square for categorical variables) and linear regression analysis to determine possible associations between pain, fatigue, anxiety, depression and the demographic and clinical variables. Correlations between examined symptom levels were measured by Spearman's rho. Statistically significant levels were set at p ˂ .05 and all tests were two-tailed.

| Ethics
Research Ethics Committee approval was obtained by the Regional

| RE SULTS
People with cancer (N = 174) with a mean age of 77.4 years participated. Table 1 shows the comparison of demographic and clinical characteristics of the total and when divided by sex. There was a significant difference between men and women in terms of civil status, ongoing treatment, anxiety and depression. There were more single women (55% vs. 35%; p = .01) and more women who underwent treatment (52% vs. 35%; p = .04) than men. Average anxiety (2.34 vs. 1.28; p = .00) and depression (2.55 vs. 1.55; TA B L E 1 Elderly people with cancer in Norway: sociodemographic and clinical characteristics of the total sample and by sex (N = 174) Note: p < .05 is in bold. p = .01) scores were significantly higher in women than in men.
Standard deviation was higher than average scores of symptoms pain, fatigue, anxiety and depression, which means there is a large spread in the study's sample. However, when comparing the groups, a non-parametric test (Mann-Whitney U test) showed the same pattern as the parametric test and we therefore kept the t test for this presentation. Table 2 shows several significant but weak relationships, and the strongest relationship was between anxiety and depression. We therefore included only the anxiety variable in further regression analysis due to the high correlation between anxiety and depression and to avoid multi-collinearity. Including both variables in regression analysis could disturb the results as a kind of tautology.

| D ISCUSS I ON
The prevalence for anxiety, depression and pain in this sample of older home-dwelling people with cancer showed low mean values (1.9, 2.1 and 2.4, respectively) which indicate a low symptom burden. This finding indicates that for this sample substantial treatment in addition to what is already offered is not needed since the cut-off value for pain is ≥4, indicating moderate and severe symptom burden and for anxiety and depression ≥2 (Vignaroli et al., 2006). The low symptom burden reported by the participants may be due to appropriate treatment. Of note, the variation among the participants in reporting symptoms highlights the importance to assess these symptoms to find out who needs substantial treatment. For the symptom fatigue, the mean score was above four, which indicates that relevant interventions might be needed. While we have used the item "tiredness" from ESAS-r as equivalent with "fatigue", more information is needed weather the people with cancer in this sample need interventions for fatigue or not.
Despite the large samples in latter mentioned studies, the comparison with our findings has some limitations. For example, Aass et al.
(1997) divided people with cancer into different age groups, just one of these is a relevant age group (≥65 years) for our study. In addition, the women in the study by Pashos et al. (2013) were significantly older than men, which could explain the higher fatigue scores reported by the women. In the study of Husson et al. (2015), where they used the According to Landi et al. (2005), more women experienced daily pain compared with men. Although not including people with cancer, we consider this study relevant for this discussion due to the sample of older home-dwelling people. However, sample differences can to a certain degree explain dissimilarity in the results. Cancer-related pain might be more prominent than non-cancer pain, independent of sex.
Other studies examining people with cancer found that women reported more pain compared with men. In a study by Kim et al. (2012),

| Symptoms in oldest and youngest participant groups
In this sample of older home-dwelling people with cancer, there was no significant difference in anxiety, depression, pain and  (Peters et al., 2014). The inconsistent findings of differences between the oldest and the youngest participants in these four symptoms highlight the importance of assessing demographics and settings in the different studies.

| Pain, fatigue, anxiety, depression and civil status
It is interesting to note that we did not found any differences in examined symptoms between those who were in paired-relationships and those who lived alone. A contrasting result was found in a study by Husson et al. (2015), where people living alone reported significantly higher fatigue scores than those with a partner. A possible explanation might be that those who live alone have no one at home to share everyday activities; they might become more tired compared with those with partners who most likely share the daily chores. Another plausible explanation for the divergent findings could be related to the sample, where only 32% of the participants were above the age of 70. A third explanation might be the assessment tool (Fatigue Assessment Scale) which does not correspond with the single item from ESAS-r used in our study.
In studies among people with breast cancer, support from one's surroundings turns out to have a positive effect on psychological well-being (Salonen et al., 2013;Yoo et al., 2014). We also assume that people, who have someone nearby most of the time, receive more support because of the partner's availability. Therefore, our results are surprising. A possible explanation might be presence of the home care nurses compensating support from others.

| Combination and connection between pain, fatigue, anxiety and depression
Relatively low levels of anxiety and depression among the participants in this study are consistent with the findings of another study examining older people with cancer (Harrison et al., 2011).
As in our study, a study by Agasi-Idenburg et al. (2017) revealed a strong significant correlation between depression and anxiety.
The correlation matrix that includes pain, fatigue, depression and anxiety showed the correlation between anxiety and depression to be the strongest.
Several studies confirm association between pain, anxiety, depression (Landi et al., 2005;Utne et al., 2010) and fatigue (Paiva et al., 2012). Some presented the relationship between pain and anxiety (Mystakidou et al., 2006) and some between pain and depression (Bair et al., 2004;Kim, Malone, & Barsevick, 2014;Laird, Boyd, Colvin, & Fallon, 2009;Landi et al., 2005). The study of Paiva et al. (2012) used an ESAS-scale for symptom measurement, the same as in our study. In the study by Landi et al. (2005), the pain was registered by using health personnel's observation evaluation (proxy rating) which may give an under-or overestimation of the person's symptoms (Blomqvist & Hallberg, 1999).
Even though several studies confirm different associations among these four symptoms, there is no clear evidence of a causal relationship between pain, fatigue, anxiety and depression. According to a population study, people who are suffering from continuous pain are in greater risk of experiencing poor health and disability than those who are in less severe pain (Nahin, 2015). Another study shows strong connection between pain and fatigue in people with cancer undergoing treatment. Those who had both pain and fatigue tended to experience higher degree of depression (Kim, Shaffer, Carver, & Cannady, 2014).
It is possible that well-adjusted pain treatment may reduce fatigue, anxiety and depression (Kim et al., 2013) and individual follow-up programmes may have a positive effect on symptoms like fatigue, anxiety (Paiva et al., 2012), pain and depression (Shi et al., 2015). These explanations are pointing out the connection between the person's physical and mental states. Other studies show the link between physical and mental health and argues that untreated pain can result in increased anxiety and depression levels in people with cancer (Mystakidou et al., 2006). For example, one study found a strong correlation between people with cancer's depression symptoms and impaired physical activity (Kim & Yi, 2015). This connection can be explained by the possible side effect of cancer treatment. Fatigue may result in impaired physical activity, leading to the loss of normal daily activities and to mental distress. This understanding may be linked to findings in two other studies. People with cancer with lower activity levels are at greater risk of developing depression (Grov, Fossa, & Dahl, 2010).
Older people's lack of physical activity and impaired social roles may predict anxiety and fatigue can lead to depression (Aass et al., 1997). A more advanced study methodology is needed to confirm the hypothesis that pain, fatigue, anxiety and depression are connected and to understand the possible causal relationship between these symptoms.
However, the theoretical assumption that mental and biological processes are connected (Kandel, 1998) seems to be supported in this study. It is therefore worth to shed light on the nursing perspective claiming the person's condition and surroundings to incorporate physical, physiological and mental aspects (Schaller, Larsson, Lindblad, & Liedberg, 2015).
Based on our findings and the results of previous studies, we assume that to be a woman may result in increased mental distress. Therefore, women facing cancer should be given attention by health care personnel in the community.

| Limitations of the study
The use of a single item to measure symptoms is a limitation of this study. However, in a review of the measurement of psychological distress in palliative care, the authors reported that assessment measures like Edmonton Symptom Assessment Scale (ESAS), play an important role in clinical research (Kelly, McClement, & Chochinov, 2006). In addition, Hotopf, Chidgey, Addington-Hall, and Ly (2002) noted that a single-item questionnaire has obvious advantages in palliative care populations.
We had no possibility of cross-checking the self-rated symptom burden with their healthrecords. More female participants reported living alone and undergoing treatment than male participants. This could explain why some symptoms were more prominent in women than in men. At recruitment, only 77 participants were receiving treatment; this could result in low average scores of pain, anxiety and depression.
A large variety of pain, fatigue, anxiety and depression assessment tools have been used to measure symptoms in previous studies. Further, differences in demographic and clinical variables of the samples (e.g. differing age groups), different study aims and methods create a challenge when comparing results.

| CON CLUS IONS
Physical and mental processes in people with cancer are linked, and this connection have been confirmed by several studies. The main results from our study present a strong association between anxiety and depression as well an association of pain with fatigue and anxiety. Home care personnel meeting older home-dwelling people with cancer should emphasize the person's mental health problems.
Focus on the complexity of the symptom burden and awareness of pain, fatigue and anxiety symptoms occurring at the same time is important.
More studies are needed to investigate causal relationship between these symptoms and sex differences. We recommend a larger sample of older home-dwelling people with cancer than used in this study with particular focus on cluster analyses.

ACK N OWLED G EM ENTS
Thank you to Kristin Vassbotn Guldhav (RN, MSc) collecting data from Western Norway to the sample. The authors thank Mary Scott Stumoen (RN, MNSc) for valuable comments and proofreading.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data file supporting the conclusions of this article will be available in the Norwegian Centre for Research Data (NSD) (https :// www.nsd.uib.no/nsd/engli sh/index.html), after completing the study. Data can be shared with readers on request to the corresponding author.