Prevalence, Risk Factors and Assessment of Depressive Symptoms in Patients With Systemic Sclerosis

Abstract

Objectives: This study aims to evaluate the prevalence of depressive symptoms among systemic sclerosis (SSc) patients using the Major Depression Inventory (MDI), identify possible risk factors, and analyze the current standard of care to raise awareness and improve clinical care for SSc patients.
Patients and methods: The study included 94 SSc patients (12 males, 82 females; mean age 58.3±13.6 years; range, 28 to 83 years) who completed the MDI, Short Form 36 Health Survey, Scleroderma Health Assessment Questionnaire, Brief Fatigue Inventory and Physical Activity Questionnaire. Clinical parameters were assessed according to standardized procedures. Discharge letters were analyzed for evaluation of depressive symptoms.
Results: The prevalence of depressive symptoms was 22.3%. It correlated with female sex (p=0.047), underweight (p=0.002), fatigue (p<0.001), decreased quality of life (p<0.001) and less physical activity (p=0.048). The latter three were confirmed as independent risk factors in a multivariable regression analysis. The analysis of the current standard of care revealed no assessment of depressive symptoms in the majority of patients (89.4%), including 19 with depressive symptoms according to the MDI score.
Conclusion: This study confirms the high prevalence of depressive symptoms in SSc patients. There is an unmet need of regular assessment of mental health during SSc consultations. Fatigue, decreased quality of life and reduced physical activity were ascertained as independent risk factors, while special attention should also be paid to weight loss and underweight.

Introduction

Systemic sclerosis (SSc) is a connective tissue disease, which is characterized by a combination of vasculopathy, autoimmune-mediated inflammation and organ fibrosis.(1) The etiology of the disease is still unknown and no therapy exists to reverse or halt disease progression. Previous studies have shown that patients with SSc have a higher risk to develop depressive symptoms than the average population and various disease-specific factors such as the disfiguring and disabling course of disease have been proposed. The frequency ranges from 20 to 68% in different studies.(2-11) This compares to a frequency of 8.1% among the average German population according to a survey by the Robert Koch-Institute.(12) Lack of emotional support,(4) low ability to work, pain,(3) disability,(13,14) disease severity,(5,11) fatigue, fear of progression,(15) being lower educated and unmarried(5,16) have been identified as potential risk factors in previous studies.

Many studies on depression in SSc patients used the Beck Depression Inventory.(3,4,6,7,11,17) In those studies, the prevalence of depressive symptoms was high with 46 to 65% with the exception of 20% in a study by Baubet et al.(17) Almost all of these studies had a sample size ≤100(3,4,8-10,17) or ≤150.(6,7,11) A larger Canadian study analyzed 376 SSc patients using the Center for Epidemiological Studies Depression Scale (CES-D) to detect depressive symptoms and found a prevalence of 35%.(5) The same level (36%(9) and 38%(10)) were detected with the CES-D9 and the Hospital Anxiety and Depression Scale 10 by two other studies with a sample size ≤100.(9,10)

Our study examined the prevalence of depressive symptoms in SSc patients at our German center by using the Major Depression Inventory (MDI). We investigated the impact of sex, weight, fatigue, quality of life, disability and physical activity on the development of depressive symptoms and assessed the current standard of care, which has not been addressed by previous studies. Additionally, we evaluated the correlation with other comorbidities such as fatigue, using the Brief Fatigue Inventory (BFI), quality of life, using the Short Form 36 Health Survey (SF-36®) and physical activity, using the Physical Activity Questionnaire (IPAQ). Considering the high prevalence of depressive symptoms among SSc patients(2) and a general underdiagnosis of depression in primary care,(18) there is a potentially increased risk for secondary comorbidities such as coronary heart disease(19) and an altered pain perception.(20) Hence, it is of great importance for the comprehensive treatment of SSc not to miss any signs of depression in SSc patients and to know how depressive symptoms in these patients are assessed and treated best in a clinical routine setting. Therefore, in this study, we aimed to evaluate the prevalence of depressive symptoms among SSc patients using the MDI, identify possible risk factors, and analyze the current standard of care to raise awareness and improve clinical care for SSc patients.

Patients and Methods

In this study, 94 SSc patients (12 males, 82 females; mean age 58.3±13.6 years; range, 28 to 83 years) from our center at the Department of Rheumatology, Charité University Hospital Berlin, Germany were recruited between April 2013 and October 2015. They met the American College of Rheumatology/European League Against Rheumatism 2013 classification. All patients completed the MDI, SF-36®, Scleroderma Health Assessment Questionnaire (SHAQ), BFI and IPAQ. The study protocol was approved by the Charité University Hospital Berlin Ethics Committee (EA1/013/13). A written informed consent was obtained from each patient. The study was conducted in accordance with the principles of the Declaration of Helsinki.

We used the MDI to assess the prevalence of depressive symptoms among SSc patients. The questionnaire is used as a diagnostic instrument for Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV) major depression,(21) but also as a depression severity scale. The scale provides information about the severity of depressive symptoms. The MDI contains 10 items with 50 being the highest possible summarized score, which can be transferred into the categories of low (score of 20-24), moderate (score of 25-29) and severe depressive symptoms (score of ≥30).(22)

We measured quality of life using the SF-36®. It results in weighted scores for eight domains and a total value from 0 to 100 with a low score being equivalent to low quality of life.(23) The test has been used in many studies with SSc patients before.(13,24)

The German version of the SHAQ was used to measure the severity of disability arising from SSc. It results in a weighted total score, which can be translated in no to mild disability (DI) (SHAQ-DI >1), moderate (SHAQ-DI 1 to <2) and severe impairment (SHAQ-DI 2 to 325) and it is commonly used and well-validated in SSc.(7) The score correlates positively with the degree of impairment.(25)

The level of fatigue was measured using the BFI. Mild fatigue accounts for values from 10 to 30, moderate from 40 to 60 and severe from 70 to 100 (in decimal steps).(26) Fatigue is a symptom frequently mentioned by SSc patients.(18)

The IPAQ was used to measure physical activity in the categories of work, transportation, work at home and leisure time. It results in a weighted continuous score and a categorical score reflecting levels of low, moderate and intense activity. A high score in the test shows a more distinct level of physical activity.(27) Additionally, most patients completed the ABILHAND Questionnaire, a tool to assess a patients’ ability to perform daily activities with their upper limbs.

Clinical parameters including sex, age, disease duration, cutaneous subsets, type and number of organ involvement, autoantibodies, weight and the resulting body mass index (BMI, kg/m(2)) were collected and maximum grip strength was measured with Jamar hand dynamometer® (Patterson Medical by Sammons Preston, Bolingbrook, Illionis, USA).

At the time when the patients were included in the study, discharge letters were analyzed with respect to the assessment of depressive symptoms during their consultation to evaluate current standard of care. We analyzed whether depressive symptoms and mental health were assessed; (i) at all, (ii) if depression was mentioned in the list of diagnoses or in the examination, and (iii) if there was any treatment prescribed or suggested, for example social or psychological intervention or treatment with antidepressants.

Statistical analysis

The data were analyzed using the IBM SPSS version 23.0 program (IBM Corp., Armonk, NY, USA. The data are presented as percentages for categorical variables and means and standard variations for continuous variables. For the analysis of risk factors, patients were divided into two subgroups, those with low to severe (≥20) and those without depressive symptoms (<20) according to the MDI. The subgroups were compared with regard to clinical parameters and results of the patient questionnaires using chi-square and Mann-Whitney U tests to identify potential risk factors with significant differences between the two subgroups. In order to detect any potential association between the MDI score and clinical characteristics or results of the questionnaires, we conducted a multivariable linear regression analysis to test for independent influence of previously identified potential risk factors. Correlation analysis of the MDI score with scores of the SF-36®, SHAQ, BFI and IPAQ was performed to assess related comorbidities. Results with a significance level of p≤0.05 were considered as statistically significant.

Results

We enrolled a cohort of 94 SSc patients with 82 females (87%) and 12 males (13%). Our cohort is representative as it reflects the skewed proportion between females and males, as well as the proportions of patients with limited or diffuse cutaneous SSc and age profile (Table 1). Mean age was 58.3±13.6 years and mean disease duration was 7.6±5.9 years. Fifty four patients (57.4%) had limited cutaneous SSc and 40 patients (42.6%) had diffuse cutaneous SSc. In 32 patients (34.0%), anti-centromere antibodies were detected, while anti-topoisomerase I antibodies were detected in 31 (33.0%). The mean number of organ involvements out of 12 added up to 4.2±1.4. Mean BMI was 24.5±5.4 kg/m(2) and 38 patients (40.4%) were overweight, whereas six patients (6.4%) were underweight.

Table 1. Clinical characteristics of our cohort of 94 systemic sclerosis patients

The MDI score showed a prevalence of depressive symptoms of 22.3% (21 out of 94 patients) in the study cohort. The mean MDI score was 13.8±9.6 with a range from 0 to 47 points. There were 21 patients with a score ≥20 and 73 patients with a score <20; accordingly, patients were grouped into those with or without symptoms of depression. Graded by severity, three patients (3.2%) had mild, 10 (10.6%) moderate and eight (8.5%) severe depressive symptoms (Table 2). The mean SF-36® total score for quality of life was 54.6±17.2 with slightly higher scores in the mental domain (SF-36® Mental Health SUM) with 61.3±19.1. The SHAQ showed a mean of 0.98±0.65 points, which indicated only a mild degree of disability at first sight. The BFI added up to 39.6±21.0 points, which indicated a moderate level of fatigue in our patients. The analysis of the IPAQ revealed a mean of 2,812±2,310 metabolic equivalent of task (MET)/minute, which was equivalent to a moderate level of activity.

Table 2. Results of patient questionnaires (n=94)

There were 73 patients who did not show any signs of depressive symptoms (mean MDI score 9.5±5.2) and 21 with depressive symptoms (mean MDI score 28.7±6.0). There were no males in the group of patients with depressive symptoms (Table 3). Five out of 21 patients were underweight, whereas only one patient was underweight in the group without depressive symptoms. The higher prevalence of depressive symptoms in females (p=0.047) and underweight patients (p=0.002) turned out to be significant. In the group with depressive symptoms, the quality of life was significantly decreased (SF-36®, p<0.001), the degree of disability was increased (SHAQ, p=0.003), there was also a higher degree of fatigue (BFI, p<0.001) and less activity (IPAQ MET/minute, p=0.048). There were no significant differences regarding age, BMI, mean weight, obesity, duration of disease, cutaneous subsets, maximum grip strength, antibodies and type or number of organ involvement, including no differences regarding lung fibrosis and pulmonary hypertension between patients with or without depressive symptoms.

Table 3. Comparative analysis between two subgroups: no depressive symptoms versus depressive symptoms

There were significant correlations between the MDI score and the SF-36® total score (p<0.001), SF-36® Mental Health SUM (p<0.001), SHAQ (p<0.001), BFI (p<0.001), IPAQ (p=0.001) and ABILHAND (p=0.001) (Table 4).

Table 4. Correlation analysis between results of patient questionnaires and MDI scores

In order to assess factors that might coexist with depressive symptoms, the following variables, which were either significantly different between the two groups or correlated with the MDI, were tested in a linear regression model of the MDI: SF Mental Health SUM, BFI, IPAQ, ABILHAND and underweight. In the multivariable linear regression analysis, SF Mental Health SUM, BFI score and IPAQ were independently significant (r=0.850, pBFI<0.001, pSFMH<0.001, pIPAQ=0.043) in the model to predict severity of depressive symptoms.

The analysis of the discharge letters identified 5.3% of patients with a diagnosis of mood disorder. Of the patients, 1.1% reported symptoms of depression, while 9.6% denied any depressive symptoms. However, the vast majority of 89.4% was not assessed for symptoms of depression (Table 5). These included 19 out of 21 patients with depressive symptoms according to the MDI. Advice on social support or a referral to a psychiatrist/psychologist was given to seven (7.4%) patients. Thirteen (13.8%) were treated with antidepressants, either for depression, or possibly also for neuropathic pain, as no indication was mentioned with the drugs.

Table 5. Depressive symptoms mentioned in discharge letters in patients without versus with depressive symptoms

Discussion

In this study, we found a prevalence of depressive symptoms of 22.3% among SSc patients. This is lower than the prevalence of 65% reported by Roca(4) in 1996 and other studies in the years after with results around 50%(3,6-8) or 35%(5,9,10) At the same time, this rate is almost three times higher than the rate of depression in the average German population.(12) Importantly, we used the MDI in order to evaluate the prevalence of depressive symptoms. This score has not been used in studies on SSc before,(2) but has become very popular in recent years because it has only 10 questions to answer and has been shown to provide a sufficient instrument for the use as a depression severity scale.(28) It has the advantage of being easy to handle for the patient and is quick to assess. Also, it allows to analyze the prevalence of depressive symptoms and not only that of major depression disorder (MDD) according to the DSM-IV classification or International Classification of Diseases, 10th Revision. We did not use the MDI as a diagnostic instrument,(22) because we were interested in general mental health issues in our patients. Including patients that fulfilled the diagnosis of MDD only would have led to an exclusion of many patients with high MDI scores and obvious mental health issues.

Many previous studies on depression in SSc used the Beck Depression Inventory(3,4,6,7,11,17) with a cut-off of ≥10 and symptoms lasting at least one week. The questionnaire has been revized in 1996(29) in order to increase specificity with a new cut-off of ≥13 and symptoms over the last two weeks. Nevertheless, these studies still used the old version, thereby overestimating the prevalence of depression. Therefore, it might be difficult to compare their results with the prevalence found in our study. Other studies used the Montgomery- Asberg Depression Rating Scale(8) and CES-D(5,9,14) The studies using the latter also found a lower prevalence of depressive symptoms in SSc, namely around 35%. On a critical note, estimated prevalence might vary considerably according to the specific questionnaires. Also, these questionnaires do not exclude non-psychiatric conditions that lead to similar symptoms, as for example diagnostic interviews can.(5)

Looking at potential risk factors, we found no significant differences between patients with or without depressive symptoms regarding disease duration and any disease-specific organ manifestations like in most other studies.(3,4,7,8,11,14,30) Instead, there were significant correlations between depressive symptoms and poor quality of life (SF-36®), fatigue (BFI) and increased disability (SHAQ and ABILHAND). In our regression analysis; quality of life, fatigue and disability could also be identified as being associated with depressive symptoms. This finding has also been supported by previous studies that identify quality of life(13,14) fatigue(15,30) and disability(14) as risk factors. Similarly, physical activity (IPAQ), which has not been previously studied in relation to SSc, was reduced in patients with depressive symptoms. In our regression analysis, it also independently correlated with the MDI. However, the result of a correlation analysis does not allow us to establish causalities between physical activity and depressive symptoms. Also, there might be a conceptual overlap between these variables and depressive symptoms, as questions asking for symptoms of depression might be very similar to questions asking for fatigue, disability or quality of life.

Interestingly, we found that being female or underweight is related to the occurrence of depressive symptoms. Even though females are more likely to suffer from depression(31) and males are less frequently affected by SSc, we were still surprised that there were no males present in the subgroup of patients with depressive symptoms. Particularly knowing that male patients with SSc often suffer from erectile dysfunction,(32) which in turn can be responsible for a higher likelihood of developing depression,(33) this result came as a surprise. It is possible that the low case number of male patients (n=12) in our study might be responsible for this result and that higher case numbers might lead to a different result.

Next, we analyzed the current standard of care of mental health among SSc patients. We reviewed discharge letters with respect to the assessment and the treatment of depressive symptoms. Our findings revealed that most patients had not been assessed for depressive symptoms as far as medical reports can tell, including 19 of 21 patients with depressive symptoms according to the MDI. Reasons why physicians might avoid asking about mental health could be the fear of patients’ reluctance to answer personal questions or insecurity regarding how to address these issues. There might also be a lack of time to ask patients about depressive symptoms or to document such questions. A possible solution to improve awareness of mental health among doctors and patients alike might be to offer specific disease management programs that convey patients skills for coping with the disease and to handle activities of daily life.

This study has some limitations. Firstly, it has a relatively small sample size for an extensive statistical analysis as we have performed on this dataset. This is a problem that occurs frequently in the field of rare diseases. Multi-center studies are often helpful to achieve studies in this field with a good statistical power. Secondly, the nature of our study is exploratory, which might be seen as problematic for the following reasons: There is a potential overlap between SSc symptoms such as sleep difficulties and fatigue and symptoms of depression. Our study does not address this factor. Also, it is not possible from our analysis to assess clinically meaningful differences in depressive symptoms as this has not been defined for the MDI. Hence, consequences of this study for clinical routines still have to be evaluated with care.

In conclusion, our study found a high prevalence of depressive symptoms in a fifth of SSc patients. Fatigue, reduced quality of life and less physical activity were ascertained as independent risk factors. Particularly, when there is progressive worsening over time along any of these dimensions, special care should be directed to assessing patients’ mental health and coping strategies.

The authors declared no conflicts of interest with respect to the authorship and/or publication of this article.

The authors received no financial support for the research and/or authorship of this article.

References

1. Leavens A, Patten SB, Hudson M, Baron M, Thombs BD. Influence of somatic symptoms on Patient Health Questionnaire-9 depression scores among patients with systemic sclerosis compared to a healthy general population sample. Arthritis Care Res (Hoboken) 2012;64:1195-201.
2. Launay D, Baubet T, Cottencin O, Bérezné A, Zéphir H, Morell-Dubois S, et al. Neuropsychiatric manifestations in systemic sclerosis. Presse Med 2010;39:539-47.
3. Matsuura E, Ohta A, Kanegae F, Haruda Y, Ushiyama O, Koarada S, et al. Frequency and analysis of factors closely associated with the development of depressive symptoms in patients with scleroderma. J Rheumatol 2003;30:1782-7.
4. Roca RP, Wigley FM, White B. Depressive symptoms associated with scleroderma. Arthritis Rheum 1996;39:1035-40.
5. Thombs BD, Hudson M, Taillefer SS, Baron M. Prevalence and clinical correlates of symptoms of depression in patients with systemic sclerosis. Arthritis Rheum 2008;59:504-9.
6. Benrud-Larson LM, Haythornthwaite JA, Heinberg LJ, Boling C, Reed J, White B, et al. The impact of pain and symptoms of depression in scleroderma. Pain 2002;95:267-75.
7. Beretta L, Astori S, Ferrario E, Caronni M, Raimondi M, Scorza R. Determinants of depression in 111 Italian patients with systemic sclerosis. Reumatismo 2006;58:219-25.
8. Legendre C, Allanore Y, Ferrand I, Kahan A. Evaluation of depression and anxiety in patients with systemic sclerosis. Joint Bone Spine 2005;72:408-11.
9. Nietert PJ, Mitchell HC, Bolster MB, Curran MY, Tilley BC, Silver RM. Correlates of depression, including overall and gastrointestinal functional status, among patients with systemic sclerosis. J Rheumatol 2005;32:51-7.
10. Richards HL, Herrick AL, Griffin K, Gwilliam PD, Loukes J, Fortune DG. Systemic sclerosis: patients' perceptions of their condition. Arthritis Rheum 2003;49:689-96.
11. Faezi ST, Paragomi P, Shahali A, Akhlaghkhah M, Akbarian M, Akhlaghi M, et al. Prevalence and Severity of Depression and Anxiety in Patients With Systemic Sclerosis: An Epidemiologic Survey and Investigation of Clinical Correlates. J Clin Rheumatol 2017;23:80-6.
12. Busch MA, Maske UE, Ryl L, Schlack R, Hapke U. Prevalence of depressive symptoms and diagnosed depression among adults in Germany: results of the German Health Interview and Examination Survey for Adults (DEGS1). Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2013;56:733-9.
13. Danieli E, Airò P, Bettoni L, Cinquini M, Antonioli CM, Cavazzana I, et al. Health-related quality of life measured by the Short Form 36 (SF-36) in systemic sclerosis: correlations with indexes of disease activity and severity, disability, and depressive symptoms. Clin Rheumatol 2005;24:48-54.
14. Nguyen C, Ranque B, Baubet T, Bérezné A, Mestre- Stanislas C, Rannou F, et al. Clinical, functional and health-related quality of life correlates of clinically significant symptoms of anxiety and depression in patients with systemic sclerosis: a cross-sectional survey. PLoS One 2014;9:90484.
15. Kwakkenbos L, van Lankveld WG, Vonk MC, Becker ES, van den Hoogen FH, van den Ende CH. Disease-related and psychosocial factors associated with depressive symptoms in patients with systemic sclerosis, including fear of progression and appearance self-esteem. J Psychosom Res 2012;72:199-204.
16. Levis B, Rice DB, Kwakkenbos L, Steele RJ, Hagedoorn M, Hudson M, et al. Using Marital Status and Continuous Marital Satisfaction Ratings to Predict Depressive Symptoms in Married and Unmarried Women With Systemic Sclerosis: A Canadian Scleroderma Research Group Study. Arthritis Care Res (Hoboken) 2016;68:1143-9.
17. Baubet T, Ranque B, Taïeb O, Bérezné A, Bricou O, Mehallel S, et al. Mood and anxiety disorders in systemic sclerosis patients. Presse Med 2011;40:e111-9.
18. Mitchell AJ, Vaze A, Rao S. Clinical diagnosis of depression in primary care: a meta-analysis. Lancet 2009;374:609-19.
19. O’Neil A, Fisher AJ, Kibbey KJ, Jacka FN, Kotowicz MA, Williams LJ, et al. Depression is a risk factor for incident coronary heart disease in women: An 18-year longitudinal study. J Affect Disord 2016;196:117-24.
20. Zambito Marsala S, Pistacchi M, Tocco P, Gioulis M, Fabris F, Brigo F, et al. Pain perception in major depressive disorder: a neurophysiological case-control study. J Neurol Sci 2015 357:19-21.
21. Bech P, Rasmussen NA, Olsen LR, Noerholm V, Abildgaard W. The sensitivity and specificity of the Major Depression Inventory, using the Present State Examination as the index of diagnostic validity. J Affect Disord 2001;66:159-64.
22. Olsen LR, Jensen DV, Noerholm V, Martiny K, Bech P. The internal and external validity of the Major Depression Inventory in measuring severity of depressive states. Psychol Med 2003;33:351-6.
23. Ellert U, Kurth BM. Methodological views on the SF-36 summary scores based on the adult German population. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2004;47:1027-32.
24. Wafki F, Amine B, Ibn Yacoub Y, Laatiriss A, Znat F, Abouqal R, et al. Depression among the Moroccan systemic sclerosis. Clin Rheumatol 2012;31:73-7.
25. Becker C. Validation of German version of “Scleroderma Health Assessment Questionnaire” to evaluate functional health in patients with systemic sclerosis. Medizinische Fakultät Charité - Universitätsmedizin Berlin, 2013. Available from: https://refubium.fu-berlin.de/handle/fub188/5941
26. Mendoza TR, Wang XS, Cleeland CS, Morrissey M, Johnson BA, Wendt JK, et al. The rapid assessment of fatigue severity in cancer patients: use of the Brief Fatigue Inventory. Cancer. 1999;85:1186-96.
27. Craig CL, Marshall AL, Sjöström M, Bauman AE, Booth ML, Ainsworth BE, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 2003;35:1381-95.
28. Bech P, Timmerby N, Martiny K, Lunde M, Soendergaard S. Psychometric evaluation of the Major Depression Inventory (MDI) as depression severity scale using the LEAD (Longitudinal Expert Assessment of All Data) as index of validity. BMC Psychiatry 2015;15:190.
29. Beck AT, Steer RA, Ball R, Ranieri W. Comparison of Beck Depression Inventories -IA and -II in psychiatric outpatients. J Pers Assess 1996;67:588-97.
30. Ferreira JF, Eugénio G, Daniel A, Rovisco J, Costa C, Rodrigues M, et al. Prevalence of anxiety and depession in systemic sclerosis. 4th Systemic Sclerosis World Congress. February 18-20, 2016, Lisbon, Portugal: World Scleroderma Foundation; 2016. p. 185.
31. Bromet E, Andrade LH, Hwang I, Sampson NA, Alonso J, de Girolamo G, et al. Cross-national epidemiology of DSM-IV major depressive episode. BMC Med 2011;9:90.
32. Foocharoen C, Tyndall A, Hachulla E, Rosato E, Allanore Y, Farge-Bancel D, et al. Erectile dysfunction is frequent in systemic sclerosis and associated with severe disease: a study of the EULAR Scleroderma Trial and Research group. Arthritis Res Ther 2012;14:R37.
33. Chou PS, Chou WP, Chen MC, Lai CL, Wen YC, Yeh KC, et al. Newly diagnosed erectile dysfunction and risk of depression: a population-based 5-year follow-up study in Taiwan. J Sex Med 2015;12:804-12.