Longitudinal comparison of anxiety and depression between idiopathic pulmonary brosis and other interstitial lung diseases: a prospective cohort study

Background objective: Few studies have investigated the differences in the incidence of anxiety and depression among the interstitial lung diseases (ILDs). We evaluated the differences in the incidence of anxiety and depression between idiopathic pulmonary brosis (IPF) and non-IPF ILD, and the changes after a 1-year follow-up period. Methods The study participants were patients included in a prospective ILD cohort of Seoul National University Hospital between March 2013 and August 2018. Clinical variables were recorded at baseline and at 1 year. The Hospital Anxiety and Depression Scale was used to assess patient anxiety and depression at baseline and at 1-year follow-up. St. George’s Respiratory Questionnaire total score at baseline was associated with depression at baseline and at 1-year follow-up. Hospital anxiety and depression scale; ILD: Interstitial lung disease; IPF: Idiopathic pulmonary RA-ILD: Rheumatoid arthritis-related ILD; respiratory


Introduction
Many conditions fall under the category of interstitial lung disease (ILD). Idiopathic pulmonary brosis (IPF), the most common form of idiopathic interstitial pneumonia [1] is a speci c form of chronic, progressive brosing ILD of unknown cause occurring in adults. The condition is associated with poor prognosis and a median survival of 2.5-3.5 years [2]. Treatment and prognosis of non-IPF ILD is different for each disease. Systemic steroid therapy is recommended to treat idiopathic nonspeci c interstitial pneumonia, and 5-year survival has been reported to be greater than 70% [3][4][5]. Cryptogenic organizing pneumonia also has good response to systemic steroid therapy [6,7]. Less than 5% of patients with cryptogenic organizing pneumonia progress to acute respiratory failure or death [8][9][10]. Connective tissue disease-associated ILD has various prognoses based on the causative connective tissue disease.
Interstitial lung disease has been found to contribute about 13% to the excess mortality of patients with rheumatoid arthritis-related ILD (RA-ILD) compared with the general population [11]. The 10-year survival rate after diagnosis of systemic sclerosis-related ILD has been reported in the range of 29%-69%. In 45%-55% of patients, deterioration of pulmonary function is con rmed within the rst 3 years, and about 16% of patients develop severe restrictive pulmonary disease [12]. The 5-year survival rate of Sjögren's syndrome-related ILD has been reported to be 84% [13].
Depression is more common in patients with chronic disease than in healthy people [14]. Anxiety and depression might be strengthened by the uncertain course of critical illness and a lack of clarity about the prognosis [15]. Therefore, anxiety and depression are common in patients with IPF or ILD. One recent study observed that about one-quarter of patients with IPF had depression and anxiety [16]. Dyspnea and comorbidity contribute to anxiety and depression in ILD [17]. Past studies have demonstrated that depression and anxiety are associated with health-related quality of life, health status, and reduced forced vital capacity (FVC); however, no signi cant differences have been found in survival rate and hospitalization [16,18,19]. One prospective study found that depressive symptoms in patients with ILD were persistent, and presence of depression at initial evaluation was the most powerful predictor of depression severity at follow-up [18]. However, anxiety had not been assessed in that study. Few studies have evaluated the differences in prevalences of anxiety and depression between patients with IPF and non-IPF ILD. In this study, we evaluated the prevalences of depression and anxiety in patients with IPF and non-IPF ILD and analyzed the changes in prevalence after 1 year, using a prospective cohort.

Methods
We recruited patients from the prospective ILD registry in Seoul National University Hospital in Korea between March 2014 and August 2018. All patients provided written informed consent. This study was approved by the Institutional Review Board and Ethics Committee of Seoul National University Hospital (number: NCT03238989). Among the patients of the ILD registry, those who completed the) at baseline and 1-year follow-up or at baseline were included in the analysis.
Idiopathic pulmonary brosis was diagnosed according to the 2011 consensus of the American Thoracic Society, European Respiratory Society, Japanese Respiratory Society, and Latin American Thoracic Association [2] and non-IPF ILD was de ned as all ILDs except IPF. The HADS, which was a primary outcome, was used to assess depression and anxiety. The questionnaire had been validated by several studies, and we used a validated Korean version [20][21][22]. The HADS consists of 14 questions; 7 questions on depression (HADS-D) and the other 7 on anxiety (HADS-A). Scores for each subscale range from 0 to 21.
Other study variables were as follows: age, sex, smoking history, comorbidity, St. George's Respiratory Questionnaire (SGRQ) score, modi ed Medical Research Council (mMRC) dyspnea scale score, six-minute walk test performance, FVC, diffusing capacity of the lungs for carbon monoxide (DLCO), and gender-agephysiology (GAP) index. Pulmonary function tests (FVC and DLCO) were carried out by trained technicians according to standardized American Thoracic Society guidelines [23]. The GAP index for each patient was calculated according to methods described by Ley, et al [24]. The SGRQ score, FVC, and DLCO were measured at the 1-year follow-up.
Continuous variables were compared using the independent Student's t-test, and categorical variables were compared using the Chi-square test. Logistic regression was used to identify the factors in uencing the presence of anxiety or depression on baseline status. The generalized estimating equation was conducted to identify the factors affecting the presence of anxiety or depression, at baseline and the 1year follow-up. Statistical analyses were performed using R version 3.4.4 (The R Foundation for Statistical Computing, Vienna, Austria, https://www.r-project.org). The level of signi cance was set at 0.05.

Participant characteristics
Two-hundred fty-three patients were recruited from the ILD registry. Twenty-three patients were excluded from the study due to incomplete HADS questionnaires. Six were excluded due to uncertainty of diagnosis. Therefore, the data of 224 patients were eligible for analysis at baseline. Of these 224 patients, 154 were diagnosed with IPF, and 70 were diagnosed with non-IPF ILD. In the patients with non-IPF ILD, there were 26 (37.1%) with connective tissue disease-related ILD, 11 with nonspeci c interstitial pneumonia (15.7%), and 1 with cryptogenic organizing pneumonia (1.4%). One hundred eleven patients completed the HADS questionnaire 1 year after enrollment in the registry (at 1-year follow-up). Of these patients, 77 had IPF, and 34 had non-IPF ILD (Figure 1).
At baseline, there were 118 (76.6%) males in the IPF group and 31 (44.3%) males in the non-IPF ILD group. Average age in the IPF and non-IPF ILD groups were 71.4±7.8 and 61.9±14.1, respectively. Compared with the IPF group, there were more patients in the non-IPF ILD group who had never smoked (56.7% vs. 30.5%). Comorbidities such as diabetes mellitus, hypertension, liver disease, history of pulmonary tuberculosis, gastrointestinal disease, brain vascular disease, articular disease, heart disease, and cancer were similar between the two groups. Compared with the non-IPF ILD group, SGRQ symptom scores were higher in the IPF group (41.9±22.1 vs. 34.5±20. 5 Table 2).
The prevalence of anxiety increased after 1 year in both the IPF and non-IPF ILD groups, and the p-value for interaction was 0.878. The prevalence of depression in the non-IPF ILD group decreased after 1 year, and the difference in prevalence between the IPF group and non-IPF ILD group increased; however, the pvalue for interaction was 0.376 ( Figure 2).

Factors associated with anxiety and depression
At baseline, whether the patient had been diagnosed with IPF or non-IPF ILD was not associated with anxiety and depression (odds ratio [OR] 1.65; 95% CI 0.71-3.88 in anxiety; OR 1.27; 95% CI 0.64-2.53 in depression). This nding was the same even after sex, age, whether the patient had been diagnosed with IPF or non-IPF ILD, SGRQ total score at baseline, FVC percent predicted at baseline, DLCO percent predicted at baseline, mMRC dyspnea scale score, and six-minute walk test performance were adjusted (OR 1.72; 95% CI 0.64-4.63 in anxiety; OR 1.18; 95% CI 0.52-2.70 in depression). St. George's Respiratory Questionnaire total score at baseline was signi cantly associated with anxiety and depression at baseline after adjusting the above-mentioned factors (OR 1.72; 95% CI 1.01-1.05 in anxiety; OR 1.04; 95% CI 1.02-1.06 in depression) ( Table 3). Whether the patient had been diagnosed with IPF or non-IPF ILD was not associated with anxiety and depression after 1 year (OR 1.48; 95% CI 0.54-4.10 in anxiety; OR 1.69; 95% CI 0.64-4.50 in depression). Only SGRQ total score at baseline was associated with depression after 1 year, after adjusting for sex, age, whether the patient had been diagnosed with IPF or non-IPF ILD, FVC percent predicted at baseline, DLCO percent predicted at baseline, mMRC dyspnea scale score at baseline, and 6-minute walk test performance at baseline (OR 1.03; 95% CI 1.00-1.05) ( Table 4).

Discussion
In this study, prevalences of anxiety and depression at baseline were 20.8% and 29.9% in the IPF group, and 14.3% and 24.3% in the non-IPF ILD group, respectively. At 1-year follow-up, prevalences of anxiety and depression at baseline were 27.3% and 31.2% in the IPF group, and 23.5% and 17.6% in the non-IPF ILD group. St. George's Respiratory Questionnaire total score at baseline was associated with anxiety and depression at baseline and depression at 1-year follow-up. The proportion of patients with depression in the non-IPF ILD group decreased slightly from baseline to 1-year follow-up. However, the difference in the change in prevalence of depression between the IPF group and the non-IPF ILD group at 1-year follow-up was not statistically signi cant.
According to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, the diagnostic criteria for depression are mood swings, marked irritability or anger, decreased interest in usual activities, etc. Anxiety refers to the anticipation of future threat. This mood state is associated with muscle tension and vigilance in preparation for future danger and cautious or avoidant behaviors [25].
Psychological symptoms are associated with disease progress in severe COPD [26]. Depression negatively impacts patient quality of life by decreasing treatment compliance [27]. Idiopathic pulmonary brosis is associated with poor prognosis; therefore, anxiety and depression are common in patients with IPF [18,19]. The prevalences of anxiety and depression in patients with IPF in the present study were similar to those in other studies [28]. However, there were no signi cant differences in the prevalences of anxiety and depression between patients with IPF and patients with non-IPF ILDs.
Dyspnea and depression in uence one another. Depressive symptoms are caused by dyspnea, and the awareness of dyspnea is worsened by depressive symptoms [18]. Some depressive symptoms, such as dyspnea, fatigue, appetite, sleep, and concentration, can overlap with pulmonary symptoms [29]. In the present study, there were no signi cant differences in mMRC dyspnea scale score and six-minute walk test performance between the IPF group and non-IPF ILD group (Table 1). Moreover, mMRC dyspnea scale score and six-minute walk test performance were associated with anxiety and depression independently of whether the patient had been diagnosed with IPF or non-IPF ILD (data not shown). This nding indicates that anxiety and depression were related to dyspnea in all ILD patients.
Depression has been found to be signi cantly associated with health-related quality of life [19]. The present study demonstrated that anxiety is in uenced by health-related quality of life, as measured by the SGRQ, in both patients with IPF and patients with non-IPF ILD. Although the SGRQ was not speci cally developed for use in patients with IPF, its psychometric properties have been determined to be adequate, and it can be a useful measure of health-related quality of life in these patients [30]. The SGRQ total score at baseline was found to be related to anxiety and depression at baseline in both groups. Our study suggests that the SGRQ can also aid in the psychometric testing of patients with non-IPF ILD.
In one study, it was observed that the tendencies toward anxiety and depression were maintained over 1 year [31]. In another study on the impact of palliative care in patients with IPF, depression increased from baseline to the 3-and 6-month follow-ups [32]. In our study, the prevalence of anxiety increased in both groups, and the prevalence of depression increased in the IPF group but decreased in the non-IPF ILD group from baseline to 1-year follow-up. Whether the patient had been diagnosed with IPF or non-IPF ILD was not found to be associated with anxiety and depression after 1-year follow-up. The changes in prevalence of anxiety and depression after 1 year were not signi cantly different between the IPF and non-IPF ILD groups; the p-value for interaction was 0.878 and 0.376 in anxiety and depression, respectively (Fig. 2). Dyspnea was found to be associated with persistent anxiety and depression in both groups over 1 year in the univariate analysis but not in multivariate analysis. It was similar results with other study [31]. In the present study, SGRQ total score at baseline was related to depression at baseline and at 1-year follow-up, regardless of whether patients had IPF or non-IPF ILD. Interstitial lung disease has a long-term effect on health-related quality of life. Therefore, it is important to comprehensively evaluate patients with ILD.
One previous study demonstrated that depression and anxiety were not related to prognosis in patients with IPF [16]. The role of palliative care in patients with IPF is unclear [32,33]. The present study included not only patients with IPF, but also patients with non-IPF ILDs. In addition, the patients were followed up prospectively for 1 year. However, a 1-year period is a relatively short amount of time to investigate changes in anxiety and depression in patients with ILD. Further studies that evaluate the changes in and impact of anxiety and depression on clinical outcomes in patients with ILD are warranted. Availability of data and materials

Conclusion
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Continuous data are presented as mean ± SD, and categorical data are presented as number (%).