NESDA ANALYSIS PLAN Please fax, send or e-mail completed form to Marissa Kok, NESDA study, A.J. Ernststraat 1187, 1081 HL Amsterdam. Fax: 020-788 5664. E-mail: ma.kok@ggzingeest.nl NESDA is supported by the ZonMW Geestkracht program, which is intended to promote collaboration between disciplines and institutes. 1. First author information: Name of first author: Rosa E. Boeschoten E mail address: r.boeschoten@ggzingeest.nl Telephone: 020 788 4659 Site: UL/LUMC UMCG/RUG X VU/VUMC Trimbos WOK NIVEL Other If proposer is not a NESDA senior investigator, which NESDA senior investigator will supervise?: P. van Oppen, B. Penninx Funding project: Stichting MS Research 2. Working title of plan: Comparison of depressive symptoms (patterns) and other clinical characteristics in depressed patients with multiple sclerosis versus depressed patients with no chronic disease. 3. Give a brief summary of your analysis plan that includes the following: a. Research question and/or hypothesis 1. Does the profile of depressive symptoms differ between depressed patients with multiple sclerosis, compared with depressed patients with no chronic disease? Hypotheses: a) The profile of symptoms is expected to differ in severity and expression of depressive symptoms between depressed patients with MS and with depressed patients with no chronic disease. b) Given the overlap of MS symptomatology and symptoms of depression, depressive symptoms of MS patients in the domain cognition and vegetative/somatic symptoms will be elevated in frequency and severity compared to depressed patients with no chronic disease. c) Since inflammation dysregulations are more advanced in the atypical depressive subtype compared with melancholic depressive subtype, MS patients will more often show atypical features like hypersomnia and fatigue, given the immunological and inflammatory character of MS.
2. Do depressed MS patients differ in specific characteristics as socio demographic variables (gender, age etc.), anxiety (diagnoses and severity), medication usage, compared with depressed patients with no chronic disease. Hypothesis: MS patients display no different characteristics compared with depressed patients with no chronic disease 3. Do depressive symptoms and profile differ for different disease related variables as onset, duration, subtype and/or severity of MS. Hypothesis: There is no clear association between the severity and profile of depression with MS related disease variables. b. Brief background and rationale for addressing the research question in NESDA Depression in Multiple Sclerosis (MS) is highly prevalent (~25%) and elevated compared with the general population. Until now it has been difficult to understand the relation of depression with MS which is ascribed to the diversity of the disease itself. There is no clear association between the presence and severity of depression and disease related variables. Some studies report a relationship with physical disability measured with the Expanded Disability Status Scale (EDSS), but others do not report a link. The same situation pertains to disease course and duration (Feinstein 2011). Determining the presence of depression in MS could also be a challenge since overlap of MS symptomatology and symptoms of depression (fatigue, cognitive dysfunction, psychomotor slowing, sleeping disorders, lost of interest) that may lead to an over diagnosis of depression in MS. Somatic complaints may for instance inflate the total score and may only be reflective of the medical condition and not depression. Neurovegetative symptoms of depression consist of symptoms such as fatigue, sleep disturbance, appetite/weight changes, restlessness, diminished concentration and sexual dysfunction. It is conceptualized that such symptoms of depression can have a direct neurological and/or physiological etiology and could therefore be more prevalent in MS patients. Some authors (Huber 1993, Nyenhuis 1995) have therefore suggested that when evaluating depression in MS patients, neurovegetative symptoms should be discounted and/or not considered given the high overlap between symptoms. They may artificially inflate overall depression scores while mood scales may provide more accurate indices of depression in MS patients. However, it is also suggested that certain neurovegetative symptoms are specifically related to MS patients depression and are not simply indicators of physical disability or fatigue (Randolph 2000). In a general medical sample, Clark et al. (1983), found that sense of failure, suicidal ideation, sense of punishment, loss of social interest, dissatisfaction and indecisiveness were the best indicators of depression in medically ill. Canavaugh (in Strober 2007) determined that the frequency of vegetative items increased minimally as depression worsened. However, the severity of the vegetative symptoms increased linearly as the severity of depression increased. Strober and Arnett (2010) investigated a MS sample with (n=17) and without depression (n=67) with healthy controls (N=22). (Depressed MS patients were found to meet two out of the three criteria for depression: a diagnosis of a major depressive episode on the SCID interview (n=11), a
score above the median on the DPRS (17), and/or a score of one and a half standard deviations or more above the mean of the significant others report of participants on the CMDI mood subscale (N=10)). The authors conducted chi square analyses between MS (non depressed and depressed) and healthy controls to identify which symptoms (measured with the Beck Depression Inventory) were more prevalent in MS. They showed that the non depressed and depressed MS group endorsed (i.e., had a score of 1 or more) the following items more often than healthy controls: fatigue, work difficulty, indecision, irritability, loss of libido, loss of interest, crying, dissatisfaction, and self criticism. Next, they examined which symptoms were more indicative of depression in MS by examining the endorsement patterns of the non depressed and depressed MS groups. They concluded that symptoms common to MS that were related to identified depressive symptoms included dissatisfaction, self criticism, crying, irritability, and work difficulty. With the exception of work difficulty, these symptoms were also found to be more severe in depressed MS patients. The authors suggested a hierarchy in assessing depression in MS: Reports of sadness, guilt, disappointment, feelings of failure and pessimism accompanied with appetite or weight changes should be given top priority. The least amount of merit should be given to reports of fatigue, loss of libido and indecision when determining the presence of depression in MS. However, conclusions are drawn from a relatively small sample. Next to that, a depressed control sample was lacking. Comparisons between depressed MS patients and depressed controls would have enhanced the ability to substantiate the findings that certain symptoms are more representative of depression in MS. Aim By comparing depressive symptoms between a depressed sample with and without MS or chronic disease measured by the Inventory of Depressive Symptomatology (IDS, Rush et al. 1986, 1996), we aim to investigate whether MS patients with depressive symptoms are characterized by different symptom patterns than depressed patients without MS/chronic disease (NESDA population). We will a) examine difference for all depressive symptoms separately, and b) look into symptom profiles using three domains of the IDS (mood, cognition and vegetative/somatic symptoms) and c) investigate the depressive subtype melancholic and atypical depression that appear to be more homogeneous phenotypes of depression (Kahn et al. 2006, Novick et al. 2005) but might differ between depressed patients with or without a chronic disease. In addition, we further go into the relation of depressive symptoms and subtypes for different disease related variables as onset, duration, subtype and/or severity of MS, medication usage, anxiety and socio demographics. In this way we expect to gain more insight in what symptoms are most indicative of depression in MS which could enhance our conceptualisation and detection of depression in MS and will improves clinical care. c. Variables to be used in main analysis (the main predictor and outcome variables must be identified) Baseline data from NESDA and baseline data from a randomised controlled trial Depression in MS: the effectiveness of web based self help treatment will be used. Only currently depressed or anxious persons will be selected (<6 months/1 year prevalence).
Main predictor: MS dichotomous Clinical characteristics MS patients (EDSS score, subtype MS, medication) Outcome variables: Individual items from the Inventory of depressive Symptomatology (IDS) o All IDS symptoms separately o IDS symptoms divided in three categories (mood/cognition/vegetative by face validity) (Schaakxs et al., NESDA Analysis plan 1501) o IDS symptoms on two dimensions: a typical/melancholic depression (Kahn et al. 2006, Novick et al. 2005) Potential confounders: Depression severity IDS total score Codes of variables needed: NESDA N1_x051 (Area of sampling, cohort, and setting) N1_x100 (DOB, Age, gender, nationality and education of respondents) N1_x101 (Marital status, partner status, sexual preference, marriage) N1_x354 (Medication current use) N1_x250 (Chronic diseases / conditions) N1_x256 & N1_257 (CIDI depression) N1_x258 & N1_250 (CIDI anxiety) N1_x235 (IDS) N1_x236 (BAI) N1_x234 (Mastery scale) N1_x239 (Social support / Close person inventory) RCT Area of sampling, cohort, and setting DOB, Age, gender, nationality and education of respondents marital status, partner status, sexual preference, marriage Medication current use Disease related information: Onset MS, duration, MS subtype (Benigne, RR, SP, PP) EDSS score CIDI (anxiety / depression) IDS BAI Mastery scale Social support / Close person inventory
d. Outline of analyses Preparation 1. Merge NESDA & RCT variables 2. Dichotomize all IDS symptoms (IDS 0/1= Not present, 2/3= present) 3. Create 2 groups (MS patients, patients without chronic disease) Analyses 1. Table with characteristics (socio demographics, anxiety (BAI), medication) 2. Compare overall severity of depression across the separate groups (as this may affect the frequency of depressive symptoms) using ANOVA 3. Frequency tables displaying the frequency of all individual IDS symptoms per group 4. Logistic regression analyses a. All IDS symptoms separately as outcome (present/not present) b. Compare groups in IDS symptoms on three categories (mood, cognition, vegetative/somatic) c. Compare groups in IDS symptoms on two dimensions: a typical/melancholic d. Test associations with disease related variables (EDSS, MS type, duration, onset)/clinical outcome that is associated with higher IDS score e. Adjust all analyses for depression severity 4. Proposed authors: Boeschoten, R. E., Schaakxs, R., Uitdehaag, B.M.J., Dekker, J., J. H. Penninx, Beekman, A. T. F., Smit, J., B. W. van Oppen, P. Order to be determined 5. Timeline for completion and submission of manuscript July November 2015 I hereby state that I will use the data only for addressing the research question described in point 3, and not for other purposes, unless I submit a new analysis plan. Signed Date