Resumes & CVs

A multifactorial developmental model for the etiology of Major Depression in a population-based sample

A multifactorial developmental model for the etiology of Major Depression in a population-based sample
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Research report  A multifactorial developmental model for the etiology of Major Depression in a population-based sample Louise Sjöholm  a, ⁎ , Catharina Lavebratt   a  , Yvonne Forsell  b a   Department of Molecular Medicine and Surgery, Karolinska Institutet, Sweden  b  Department of Public Health Sciences, Karolinska Institutet, Sweden Received 6 March 2008; received in revised form 24 April 2008; accepted 25 April 2008Available online 16 June 2008 Abstract  Background:  Kendler et al. proposed gender specific developmental models for Major Depression (MD) using a population-basedsample of north-American adult twins. The aim of this study was to test whether any of the models could predict both MD andother depressive disorders among unrelated Swedes of both sexes. To test depression specific prediction the sample wasoverrepresented for other psychiatric diagnoses.  Methods:  Persons with and without psychiatric symptoms were randomly selected from a population-based questionnaire study andinterviewed by psychiatrists. Diagnoses were made according to DSM-IV. The study included 81 persons with MD, 132 personswith other depressive disorders, 136 persons with other psychiatric diagnoses and 744 persons without diagnosis. Path andcorrelation analyses were performed using 16 risk factors.  Results:  The path analysis revealed similar prediction values of MD for the female and male models, i.e. two-thirds of the variancein liability to MD and depression in general when controls consisted of persons without psychiatric diagnosis. Prediction of depression was only slightly weakened when distinguishing cases from those with other psychiatric diagnoses. The risk factorswith the strongest unique influence on depression were from early adolescence: neuroticism, low self-esteem, anxiety; from lateadolescence: trauma; and history of depression.  Limitations:  Possibly some limited recall bias. Conclusions:  The model was successfully replicated in both genders of unrelated Swedes. In addition, the model highly predictedother subtypes of depression, despite that the sample was overrepresented for other psychiatric diagnoses. The results support similar etiology of MD and other depressive disorders.© 2008 Elsevier B.V. All rights reserved.  Keywords:  Depressive disorder; Risk; Epidemiology; Etiology 1. Objectives of the study The etiology for Major Depression (MD) is multi-factorial and the risk factors are interrelated and shouldtherefore be integrated in a comprehensive model.However, only a few studies have reported such models. Journal of Affective Disorders 113 (2009) 66 – ⁎  Corresponding author. Department of Molecular Medicine andSurgery, Division of Neurogenetics, CMM L8:00, Karolinska Hospital,S-17176Stockholm,Sweden.Tel.:+46851775541;fax:+46851773909.  E-mail address: (L. Sjöholm).0165-0327/$ - see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.jad.2008.04.028  The first was Akiskal & McKinney who proposed a biological common pathway for depression. The modelsuggested that a combination of genetic, developmental, pharmacological and interpersonal factors leads to areversible impairment of the neurophysiological sub-strates of reinforcement, which manifests as depression(Akiskal and McKinney, 1973; Akiskal and McKinney,1975). Later, Kendler et al. reported a comprehensivedevelopmental model for MD where known vulnerabilityfactors were structured according to when they occur inlife (Kendler et al., 1993; Kendler et al., 2000; Kendler etal.,2002).Modelsweredevelopedseparatelyinfemale(Kendler et al., 2002) and male twins (Kendler et al., 2006). The models had five tiers: 1) childhood: familyhistory of MD, disturbed family environment/low parental warmth, childhood parental loss and childhoodsexual abuse, 2) early adolescence: neuroticism, self-esteem, conduct disorder and early onset anxiety, 3) lateadolescence: educational attainment, lifetime traumas,socialsupportandsubstancemisuse,4)adulthood:historyofdivorce and history ofMD, and 5)the lastyear: marital problems, difficulties as well as dependent and indepen-dent life events. The model for MD developed from a population of female twins explained 52.1% of thevarianceinliability toepisodesofMD inthe lastyearandthe similar model calculated from male twins explained48.7%.The aim of the present study was to test theapplicability of these two models developed by Kendler et al. for the prediction of MD and other depressiondiagnoses in a population of unrelated female and maleSwedes. In order to determine the power of the model todetect depression specifically, and not mental illness ingeneral, the selected persons were over sampled for symptoms of mental illness. 2. Materials and methods 2.1. Source of study group The study group derived from a longitudinal popula-tion-based study, PART (in Swedish:  P  sykisk Hälsa,  A r- bete  R ela T  ioner; in English: mental health, work andrelations), of mental health ongoing in Stockholm,Sweden (Hällström et al., 2003). The participants wererandomly selected from the Stockholm city councilregisters. The age range was 20 – 64 years and onlySwedishcitizenswereincludedtoavoidlingualproblems.Thefirstwaveconsistedofaquestionnaire,sentto19,742 persons, including risk and protective factors for mentalillness as well as psychiatric symptom scales (The WorldHealth Report, 2001; Bech and Wermuth, 2000; Kastrupet al., 1993; Marks and Mathews, 1979; Saunders et al.,1993;Sheehan,1983).Psychiatristsinterviewedarandomsample of those with suspected mental illness ( n =884)and persons with few orno symptoms( n =209). The two-hour interview was performed using Schedules for Clinical Assessment in Neuropsychiatry (SCAN) (Winget al., 1990) and took place within 4 weeks of thequestionnaire answering. Present state in SCAN coversthe last four weeks, with exception of the module onalcohol use that covers the last 12 months. Previous psychiatric symptoms and signs were assessed and dated.The interviewers had a one-week introductory course led by WHO designed trainers followed by continuoussupervision. They were unaware of the person's answersto the questionnaire. The procedures, and the validity andthe reliability of the instruments are fully described in thetechnical report of the PART study (Hällström et al.,2003). The ethical committee, Karolinska Institutet,approved the study and written consent was obtainedfromalltheparticipantsaftertheprocedurehadbeenfullyexplained. 2.2. Outcome variable The outcome variable was defined as fulfilling thecriteria for a Major Depressive Episode according toDSM-IV at the time of the interview (meaning MDwithin the last 4 weeks). Analyses were also madeincluding other depressive disorders: Bipolar Depres-sion, Dysthymia, Minor Depression and Mixed AnxietyDepression all defined according to DSM-IV. 2.3. Predictor variables Source of data is given in Table 2. 2.3.1. Tier 1/childhood  Family history of MD was considered present if the person reported that any of the biological parents or sisters/brothers had a history of a MD. The questions onfamily history were asked to all interviewees at the endof the interview, when the concept of depression wasclear to the interviewees. Only cases where there was nodoubt were accounted for, that is those where thediagnosis was known by the interviewees. Disturbedfamily environment included severe financial or other severe problems in the family during childhood, divorceof parents, or having lived with a single parent duringthe entire childhood. Childhood parental loss was present if one of the parents died before the personwas 18 years old. Childhood sexual abuse was not included in the model due to missing data. 67  L. Sjöholm et al. / Journal of Affective Disorders 113 (2009) 66   –  76   68  L. Sjöholm et al. / Journal of Affective Disorders 113 (2009) 66   –  76   2.3.2. Tier 2/early adolescence The psychiatrists assessed early onset anxiety andlow self-esteem using SCAN. Neuroticism was mea-suredusingthemodule “ Worrying,tensionetc. ” inSCANand only the data on previous history that by theinterviewees was dated back to the adolescence periodwas accounted for. Symptoms were considered present if  both of the following criteria were fulfilled: (i) symptomsor signs scored ≥ moderate according to the definition inSCAN and (ii) symptoms or signs had led to interferencein every day activities as stated by the person. Conduct disorder was not assessed due to lack of time. 2.3.3. Tier 3/late adolescence Low education was defined as  ≤ 9 years of educa-tion. The psychiatrist assessed lifetime trauma. Thesocial network was considered insufficient if the personhad no person to speak freely to or to trust. Thisinformation was obtained using an instrument devel-oped by Orth-Gomer and was present in the ques-tionnaire (Unden and Orth-Gomer, 1989). Substancemisuse (alcohol and illicit drugs) was assessed inquestionnaire and during the interview. 2.3.4. Tier 4/adulthood  In the questionnaire it was asked if the person hadever divorced, and this was further assessed during theinterview. The psychiatrist assessed history of MD in parallel to the questions of depressive symptoms andsigns in present state. 2.3.5. Tier 5/last year  Difficulties included financial and housing problems.Stressful life events last year were recorded using a list of events and the person was to state if they occurred or not during the last 12 months. Independent stressful lifeevents were e.g. death of a friend. Other eventsoccurring primarily to the informant were recorded asdependent stressful life event e.g. own serious illness/ injury. Marital problems last year were assessed in thequestionnaire as well as in the interview. 2.3.6. Study population The study population of interviewees from wave oneconsisted of 81 persons that fulfilled the DSM-IVcriteria for MD at the psychiatric interview. Of those, 24also had other types of psychiatric disorder, mostlyanxiety syndromes. In the persons with MD a previousepisode of MD was present in 51 persons, and 10 of those had experienced more than one MD episode. Theduration of the present MD episode varied; 19 personshad been depressed ≤ 1 month, 41 persons 1 to 6 monthsand 21 persons ≥ 6 months. Of the 81 with MD, 16 weretreated with antidepressants, 3 with psychotherapy, 2with both and the rest were untreated. The study population also included 132 persons that fulfilled thecriteria for other types of depressive disorders; 2 Bipolar Depression, 30 Dysthymia, 29 Mixed Anxiety Depres-sion and 71 Minor Depression according to DSM-IV. Of these 132, 30 also had other types of psychiatricdisorders, mostly an anxiety syndrome. An additional136 persons had other types of psychiatric disorders; 65had anxiety syndrome whereas the others had alcoholdependency/abuse or various types of sleeping dis-orders. The number of persons who did not fulfil thecriteria for any psychiatric diagnosis was 744 (controls). 2.4. Statistical analysis The statistical analyses were performed using Mplus,version 4.2 (Muthén and Muthén, 1998 – 2006). Thiswas the same statistical program as used by Kendler et al., since our intention was to apply their develop-mental models. We performed path analysis, whichsimilarly to structural equational modeling (SEM),allows a simultaneous modeling of several relatedregression relationships, although without latent vari-ables. All the risk factors used in Kendler et al.'s SEMmodel, except childhood sexual abuse and conduct disorder, were included to build the model (Fig. 1a). Theoutcome variable MD was reported positive if MD was present at the interview, whereas in Kendler et al.'sstudies MD was scored positive if it occurred during the Fig. 1. a. Path and correlation estimates of Kendler's best-fitting model for women with Major Depression at interview as outcome variable. Two-headed arrows represent correlation matrix estimates and one-headed arrows represent probit regression coefficients. Data from sample set 1. b. Twointeracting etiological pathways for Major Depression: an internalizing pathway (red) and an adversity pathway (blue). Data from sample set 1.Table 1Descriptive statistics of the sample groupsMD Other depressive disorders Other mental illnesses No psychiatric diagnosis Total Number 81 132 l36 744 1093Females (%) 71.6 53.8 59.6 57.5 58.4Age mean (range) 46.3 (20 – 64) 42.7 (20 – 63) 41.9 (20 – 64) 41.9 (19 – 65) 42.2 (19 – 65)69  L. Sjöholm et al. / Journal of Affective Disorders 113 (2009) 66   –  76   Table 2Correlation matrix for potential risk factors and Major DepressionCH1 CH2 CH3 EA1 EA2 EA3 LA1 LA2 LA3 LA4 AF1 AF2 LY1 LY2 LY3 LY4I Q I+Q I I I Q I+Q Q I+Q I+Q I I+Q Q Q I+Q Childhood  CH1 familyhistory of MDCH2 disturbedfamilyenvironment 0.14 ⁎⁎ (0.048 – 0.223)CH3 childhood parental loss0.05(ns)0.14 ⁎⁎ (0.052 – 0.226)  Early adolescence EA1 neuroticism 0.02(ns)0.14 ⁎⁎ (0.052 – 0.226)0.16 ⁎⁎ (0.072 – 0.246)EA2 lowself-esteem0.12 ⁎⁎ (0.031 – 0.207)0.11 ⁎⁎ (0.021 – 0.197)0.04(ns)0.50 ⁎⁎ (0.430.564)EA3 earlyonset anxiety0.07 ⁎⁎ 0.223)0.13 ⁎⁎ (0.041 – 0.217)0.02(ns)0.45 ⁎⁎ (0.376 – 0.518)0.24 ⁎⁎ (0.1540.322)  Late adolescence LA1 loweducation0.08 ⁎ (0.012 – 0.147)0.20 ⁎⁎ (0.113 – 0.284)0.00(ns)0.12 ⁎⁎ (0.031 – 0.207)0.14 ⁎⁎ (0.052 – 0.226)0.08 ⁎ (0.012 – 0.147)LA2lifetime trauma0.02(ns)0.07 ⁎ (0.002 – 0.137)0.15 ⁎⁎ (0.062 – 0.236)0.39 ⁎⁎ (0.312 – 0.463)0.22 ⁎⁎ (0.134 – 0.303)0.07 ⁎ (0.002 – 0.137)0.16 ⁎⁎ (0.072 – 0.246)  7   0   L  . S    j    ö  h   o l    m e t   a l    . /    J   o ur  n a l    o  f   A   f     f    e c t   i    v eD i    s  or  d   er  s 1  1   3    (   2   0   0   9    )    6   6   –  7  6  
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks