Prevalence of Anxiety and Depression in Prostate Cancer Patients and Their Spouses: An Unaddressed Reality

Objectives To estimate the prevalence of unsuspected anxiety or depression in prostate cancer patients and their spouses, as well as factors involved in its onset. Materials and Methods. A prospective study of 184 patients and 137 spouses evaluated in our hospital during 2019 using the Memorial Anxiety Scale for Prostate Cancer (MAX-PC), Hospital Anxiety and Depression Scale (HADS) and Patient Health Questionnaire depression module (PHQ-9). This study provides an internal validity assessment of the scales and their correlation (alpha and rho coefficients; index r). The contributions of age, education level, months after diagnosis, pain, prostate-specific antigen (PSA) level, stage of the disease and treatment performed to the positivity of the questionnaires were studied using the Wilcoxon–Mann–Whitney and chi-square tests. Results The prevalence of anxiety was 10.9% (MAX-PC) and 28.3% (MAX-PC-PSA). The HADS-A questionnaire indicated pathology in 14.1% of the patients and 16.05% of the spouses. Depression was detected in 7% (HADS-D) and 9.2% (PHQ-9) of patients as well as in 8.8% (HADS-D) and 16.05% (PHQ-9) of their spouses. The greatest concordance between men and women was with the PHQ-9 (Spearman's rho: 0.78; p = 0.01). Education level is significantly related to the presence of anxiety and depression, regardless of the questionnaire applied. The probability of detecting pathology in the MAX-PC varied from 6% in patients with elementary education to 23.5% in university students (p = 0.04). The greatest differences were detected when applying the PHQ-9 to patients (4% pathological, elementary education vs. 35.3% pathological, university education). Our study confirms the lack of a relationship between rates of anxiety and depression and factors such as PSA level, age of the patient and number of comorbidities. Conclusion There is a high prevalence of unsuspected anxiety and depression in patients with prostate cancer and their wives. Education level correlates with such prevalence.


Introduction
Due to advances in the early diagnosis of prostate cancer (PC) and improvements in treatment in the different phases of the disease, we are witnessing a significant increase in the number of patients who "coexist" with their PC [1]. is increase in survival is not exempt from morbidity, and the functional sequelae of the treatments used have been widely described, including urinary incontinence, erectile dysfunction, intestinal disorders, hot flashes, weakness and fatigue.
Despite the high prevalence of patients and the potential symptomatology of adverse effects, there is a striking lack of research on the nature and prevalence of psychological disorders and psychiatric illness in this population [2][3][4]. is lack of research is multifactorial, but it may be due in large part to the prioritization of survival outcomes over the quality of life by health professionals involved in the follow-up of these patients. Likewise, there may be a lack of preparation among urologists and oncologists to recognize symptoms of depression among their patients with PC, even if the symptoms are severe. e problem is aggravated because o en the wives of patients, who tend to be caregivers, also suffer from anxiety or depression.
Our objectives are to estimate the prevalence of clinically relevant, not previously diagnosed or treated, depressive symptoms in PC patients with good control of their disease and in their spouses, how it is interrelated, and the possible clinical and oncological factors involved in the onset itself.

Patient Selection.
Outpatients diagnosed with PC who attend our Urological hospital (January-June 2019) for scheduled monitoring of their disease. Given that the main objective of the study was to detect unknown and/or untreated symptoms, all patients or spouses who were under psychological or pharmacological treatment for anxiety or depression as well as those patients without oncological control of their disease (rising prostate-specific antigen (PSA) level or progression of the disease) or visual analogue scale (VAS) >2 were excluded.

Study Variables.
Total scores for questionnaires and pathology percentages, age, months of follow-up since PC diagnosis, stage of the disease (localized, locally advanced, metastatic or castrate-resistance phase-CRPC), treatment performed (radical prostatectomy (RP), radiotherapy (RT), hormonal treatment (HT), active surveillance (AS) or multimodal treatment), comorbidities, and PSA levels.

Statistical Analysis.
Numerical variables are presented as the mean, median, maximum and minimum values, and standard deviation (sd). Categorical variables are presented as the number and percentage. e internal validity of the scales was assessed with Cronbach's alpha and total item correlation ( ). e correlation between the tests was assessed by determining Spearman's rho coefficient. e possible influence of the factors studied was performed using Wilcoxon-Mann-Whitney and Pearson's chisquare tests. e statistical package SPSS version 25 was used. A -value of 0.05 or less was considered statistically significant.

Materials and Methods.
e clinical baseline information of each patient was collected from the existing data in the clinical history. e tumour stage used was the most recently available.
Patient comorbidities were grouped by system (none, cardiovascular, endocrine, trauma, respiratory, neurological, and other pathologies) and assigned a numerical value based on the number of affected systems.
Pain was assessed according to a VAS scale (0-10). e following questionnaires validated for the Spanish language were administered:

Ethical Requirements.
e research protocol was approved by the ethics committee of our hospital, and all participants gave written informed consent.

Results
A total of 255 patients were evaluated. Twenty-seven (10.6%) were excluded because they took medication for anxiety or depression, and 14 (5.5%) were excluded because they did not present good oncological control of the disease. Of the 214 remaining patients, a total of 184 (86%) provided informed consent and agreed to be part of the study.
A total of 179 spouses were evaluated, of which 26 (14.5%) were excluded for taking medication. Of the 153 remaining spouses, 137 (89.5%) provided informed consent and completed all the questionnaires. e population characteristics are provided in Table 1. e median ages of the patients and of their wives were 71 years (49-92) and 68 years (45-87), respectively. e majority had only completed elementary education (53.8%), and only 9.2% had a university degree. e percentage of wives with elementary education was 64.2%. In the majority of cases, patients had undergone RP (53.8%) as primary treatment, 26.6% had initiated HT, 10.9% were included in an AS programme, and 8.7% had been treated with some form of RT. A total of 22.3% had undergone multimodal treatment. e median time from diagnosis to our evaluation disease was 18 months for patients who underwent RP, 12 months for those treated with RT, 23 months for patients who underwent HT and 48 months for those who opted for AS as the first treatment modality. e mean VAS score was 0.42 (median 0), with 60.9% presenting no pain (VAS 0). None of the scores exceeded 2, which confirms that this population is fundamentally asymptomatic or has very little symptomatology. e median PSA level was 0.1 (range: 0-88), and all patients were in the response phase to the proposed treatment regardless of the stage of their disease or the treatment performed. e scores obtained for the questionnaires and their internal validity are presented in Table 2. e median score for the MAX-PC was 17 (11-41 range), and 10.9% of the respondents were considered pathological. e most frequently detected pathological aspect was anxiety caused by PSA (pathological in 28.3% of patients) and, to a lesser extent, general anxiety (10.32%) and fear of recurrence (9.8%).
In the HADS-A questionnaire, 14.1% of patients were considered pathological and 16.05% of the spouses.
Regarding depression in patients, that detected by the HADS-D questionnaire (7.06%) was similar to that detected by the PHQ-9 (9.2%); for spouses; the values were higher: 8.8% were considered depressed according to the HADS-D, and 16.05% according to the PHQ-9. e internal validity of the questionnaires was very high, with a Cronbach's alpha higher than 0.8 in all cases except for the HADS-D for patients and spouses. e item-total correlation was above the value of 0.35 in almost all cases except for the MAX-PC (0.3) and HADS-D (0.26).
Likewise, the correlation between the different questionnaires according to sex (MAX-PC, HADS-A, HADS-D and PHQ for men and HADS-A, HADS-D and PHQ-9 for the spouses) was excellent, with statistical significance in all possible comparisons (Table 3). e relationships among PSA variables, age, time since diagnosis, VAS scale and comorbidities are presented in Table 4. e level of statistical significance varied according to the questionnaire applied. PSA values only showed a statistically significant relationship with the global MAX-PC score and with fear of recurrence. Age was only significantly associated with fear of recurrence (MAX-PC). Time since diagnosis was statistically significant for the HADS in both patients and spouses. Level of pain was related to the HADS-D, MAX-PC-A, MAX-PC and PHQ scale for patients. Finally, comorbidities only showed a statistically significant association with the MAX-PC and HADS-D for spouses.
Education level (Table 5) was clearly and consistently related among all the questionnaires. is association is independent of the questionnaire used and was demonstrated equally in spouses. e probability of detecting pathology in the MAX-PC varied from 6% in patients with elementary education to 23.5% in university students (푝 = 0.04). e greatest differences were detected when applying the PHQ-9 to patients (4% pathological, elementary education vs. 35.3% pathological, university education).
No association was found between the primary treatment used or the stage of the disease and the possibility of presenting a pathological test (Table 6); however, patients who had multimodal treatment had higher levels of anxiety and depression in MAX-PC, HADS-A, and PHQ scales.

Discussion
e current treatment for PC includes not only an oncological approach to the disease but also a broader vision that includes the global nature of the aspects that influence the health and well-being of the patient and their environment. Anxiety and depression are the most frequent psychological findings in cancer patients [5], and it has been previously reported that patients with PC have higher levels of anxiety and depression than do the general population [8]. ese disorders, if not investigated and treated, have a direct negative impact on the overall survival of patients [9].
It is difficult to advise a single instrument for measuring anxiety or depression, as there is no scientific evidence of the superiority of some over others [10]. All the scales used in this study have demonstrated their usefulness, and the percentage of pathological patients indicated was very similar regardless of the method used. In terms of internal consistency, Cronbach's alpha indices were acceptable or good for all scales, always above 0.70 and in the vast majority of cases above 0.80. e high statistical concordance between the results obtained from the questionnaires used in our patients indicates the possibility of using any of the instruments, although the MAX-PC (PSA) is possibly the most robust when determining the anxiety produced by the periodic determination of this marker.
e fundamental findings of the study are the detection of high levels of anxiety and depression in patients with good control of their disease, long a er the initial diagnosis (median since diagnosis: 18 months), and who are mostly asymptomatic (median VAS: 0); notably, all those who were in treatment and/or taking medication indicated for anxiety or depression were excluded. To our knowledge, there is no study that has focused on assessing risk in this population, in which, "theoretically", the risk should be minimal. e median score for the MAX-PC was 17 ± 6.43, which was significantly higher than those previously reported in the literature by Johanes et al. [8] (10.47 ± 4.64), Dale [11] (7.57 ± 7.26) or Rodríguez Vega et al. [12] (15.7%). e most influential factor is the anxiety produced by the determination of the PSA level (28.3%).  Prostate Cancer 4 HADS-D; 9.2 vs. 16.05% PHQ-9), even though the difference did not reach statistical significance (푝 = 0.09). ese findings are consistent with those found in the literature [15,16]. e possible causes of the higher prevalence of anxiety and depression in wives have not been studied; therefore, any interpretation of the data is speculative. Possible reasons include (a) the discordance between communication between couples, assuming that the wife has to openly discuss the problems and feelings of the family unit and the husband wants to minimize the effects and (b) the change in leadership because as the disease progresses, the patient requires more care, which is taken on by his wife, who must become the true support of the family nucleus.
Depression was detected in 7.06% of patients according to the HADS and in 9.2% of patients according to the PHQ-9. A meta-analysis recently published by Watts [13] demonstrated a depression rate of 15-18% in 4000 patients at all stages of PC. Our study confirms that despite the favourable circumstances established in the selection of patients and the known tendency of decreased psychiatric symptoms as the time from initial diagnosis of the disease increases [14], the level of unknown depression is maintained at significantly high levels throughout the course of the disease.
Notably, the levels of anxiety and depression in spouses are slightly higher than those in patients (7.06 vs. 8.8%  anxiety and depression were also not related to the initial treatment used or the stage of the disease. ese data are consistent with those described by Blank and Bellizzi [18], who concluded that the long-term psychological effects produced by PC depend more on the personality of the patient than on the primary treatment performed or on the side effects that could have occurred. ere seems to be a relationship between perceived pain by the patient and the positivity of certain questionnaires. However, these results should be interpreted with caution because 60.9% of the participants did not present any pain (VAS 0). Chabowski et al. [19] evaluated in 2018 the levels of depression, anxiety and pain in patients with lung cancer at risk of malnutrition. It could be another interesting factor, but this item was not analyzed in our population.

Conclusions
Our results confirm the existence of high rates of anxiety and depression in PC patients and their wives, which are unaddressed by healthcare professionals. In our opinion, this fact justifies the knowledge and periodic use of instruments for their detection.

Limitations
Despite the consistency of the findings found, we must highlight that the studied population was mostly white, married, with a low level of education and with all health costs covered by the state. Extrapolation of the results to other groups can be difficult because other groups may express their anxiety to the same initial circumstances in a different way.
e role of education level in PC anxiety rates has been partially communicated in some studies, with frankly contradictory results. Nelson et al. [17] concluded that patients with elementary level education had higher levels of anxiety regarding PSA levels than did those with a university education. In our study, there was a clear association between education level and rates of anxiety or depression independent of the test used to quantify it. e results were statistically significant and were clearly higher in women. ere is no clear explanation for this result. Possibly, patients with higher education levels have greater access to all types of information, greater awareness of the possible negative evolution of the disease, and find it more difficult to reconcile their usual tasks with the inconveniences and deficits they experience living with PC.
Our study confirms the lack of a relationship between rates of anxiety and depression and factors such as PSA level, age of the patient and number of comorbidities. Rates of