Untitled

DEPRESSION AND ANXIETY 26 : 711–717 (2009) CHILDHOOD NEGLECT AND ABUSE AS PREDICTORS OF ANTIDEPRESSANT RESPONSE IN ADULT DEPRESSION Jeanette M. Johnstone, M.A.,1Ã Suzanne E. Luty, M.B. Ch.B. Ph.D. F.R.A.N.Z.C.P.,1 Janet D. Carter, Ph.D. Dip. Clin. Psych. M.N.Z.C.C.P.,2 Roger T. Mulder, M.B. Ch.B. Ph.D. F.R.A.N.Z.C.P.,1 Christopher M.A. Frampton, Ph.D.,1 and Peter R. Joyce, M.D. Ph.D. D.Sc. F.R.S.N.Z. F.R.A.N.Z.C.P.1 Background: Childhood neglect and abuse are recognized as risk factors fordepression, but are not often studied as predictors of treatment response indepression. Methods: Clinically depressed outpatients (n 5 195) were askedabout childhood experiences before beginning a randomized antidepressant trialwith either fluoxetine or nortriptyline. Three treatment outcomes weremeasured: Adequate trial, six-week response and two months sustainedrecovery. Results: Patients reporting low paternal care (paternal neglect), asmeasured by the Parental Bonding Instrument (PBI), were less likely tocomplete an adequate six-week trial of medication. Patients who reported highmaternal protection (maternal overprotection) on the PBI had poorer treatmentresponse in the short-term at six weeks, and longer term, for two months ofsustained recovery. However, abuse, whether sexual, physical, or psychological innature, did not predict treatment response. Conclusions: The experience ofhaving a neglectful father or an overprotective mother was more predictiveof response to treatment for depression than abuse, suggesting that the quality ofongoing intra-familial relationships has a greater impact on treatmentoutcomes for depression than experiences of discrete abuse in childhood.
Depression and Anxiety 26:711–717, 2009.
Key words: depression; Parental Bonding Instrument (PBI); childhoodadversity; treatment predictors; abuse; neglect Childhood adversity, whether abuse or neglect, has 1Department of Psychological Medicine, University of Otago, been associated with the development of depression in adult life.[1–5] Abuse can take various forms including 2Department of Psychology, University of Canterbury, physical, verbal, psychological, or sexual. All forms of abuse have been strongly implicated in predisposingchild victims to adult psychopathology.[6,7] Even the Contract grant sponsors: Health Research Council of New oft-downplayed experience of being called names as a Zealand; Lottery Health; Eli Lilly; University of Otago; MentalHealth Division of Canterbury Health.
child, when perpetrated by one’s parents, has been found to contribute to higher levels of depressive Correspondence to: Jeanette M. Johnstone, Department of symptoms in adulthood.[8] In two retrospective studies Psychological Medicine, University of Otago, Christchurch, PO involving more than 25,000 adult members of a large Box 4345, Christchurch 8140, New Zealand.
health maintenance organization, adverse childhood experiences were found to increase risk for depressive Received for publication 2 March 2009; Revised 7 May 2009; One well-documented approach to measuring child- hood experiences is the Parental Bonding Instrument (PBI) developed by Parker et al.[11] Created more than 25 years ago, this self-report tool has been used around the world with community and clinical samples; in Childhood neglect and abuse were assessed using the PBI[11] and Australia,[12] Japan,[13,14] the Netherlands,[15] New structured interview questions. The PBI is a series of 25 self-report Zealand,[3,16] the United States,[5] and throughout questions, which ask the respondent to rate separately the level of Europe.[17] The PBI asks the respondent to reflect on care and protection provided by his or her mother and father or the level of perceived care and protection received from caregiver in the first 16 years of life. Responses are coded 0–3: very one’s mother and father before age 17. Parker’s early unlike, moderately unlike, moderately like and very like, with scores work, which has been supported by others,[15] sug- ranging from 0 to 36 on the care scale; 0–39 on the protection scale.
gested that the combination of low care (neglect) and High care scores indicate the respondent experienced a caring and high protection (overprotection) was a determinant in affectionate relationship with that parent or caregiver; lower scores, arejecting or cold attitude. High protection scores indicate the adult depression.[18–20] More often though, neglect or experience of an overprotective relationship with that parent; lower abuse by either parent has been identified as a common scores indicate the respondent felt able to differentiate from that antecedent for adult depression.[5,12,17,21–23] parent. The PBI has a high degree of reliability and validity.[28] Despite the substantial evidence for links between The abuse measure came from a structured nurse-administered childhood abuse and neglect with adult depression, a interview, which assessed the presence, frequency, and type of abuse small amount of literature examines these adverse the individual may have experienced before age 16. Six question childhood experiences as predictors of treatment categories asked about a range of experiences including psychologi- response. In one notable study, Nemeroff et al.[24] cal, physical, or sexual abuse or threat. Responses were collapsed into treated 681 chronically depressed patients with an three frequencies: 0 (not occurring), 1 (occurring 1–3 times), or 2 antidepressant (nefazodone), Cognitive Behavioural (occurring four or more times). Total abuse scores ranged from 0 (noabuse) to 12 (repeated abuse in all six categories). For analytical Analysis System of Psychotherapy or a combination purposes, the abuse score was then classified into four categories of both. Results showed that those reporting childhood according to frequency reported: ‘‘none’’ for a 0 score, ‘‘mild’’ if r2, adversity responded better to psychotherapy by itself ‘‘moderate’’ if r4, and ‘‘severe’’ if 44. The CSA measure was derived than to an antidepressant alone, and the combination of from the same interview as the abuse measure but included only those the two treatments was only marginally better than questions relating to sexual abuse. The three CSA categories were psychotherapy alone. Further examination of these data defined as: noncontact sexual abuse, such as exposure or being forced revealed that the likelihood of achieving remission to watch sexual activity, contact sexual abuse such as unwanted sexual from depressive symptoms was two times higher with touching or attempted intercourse, and sexual intercourse or rape. In psychotherapy for those with an adverse childhood reporting the CSA variable, ‘‘none’’ was recorded if the patient history,[25] suggesting that psychotherapy may be an reported no sexual abuse. ‘‘Severe’’ was defined as any sexual abuseinvolving intercourse or more than three reports of contact sexual essential element of treatment for those patients abuse. ‘‘Some’’ was defined by three or fewer incidents of contact reporting adverse childhood experiences. Other studies sexual abuse or any report of noncontact sexual abuse.
support the notion that childhood adversity is asso-ciated with decreased response to pharmacotherapy intervention for adult depression[26] and dysthymia.[27] Given the established risk between experiencing Following the baseline assessment, patients were randomly adversity in childhood and developing depression in assigned to treatment with either fluoxetine or nortriptyline. Theywere seen at least weekly for 20–40 min, depending on clinical need.
adulthood, we hypothesized that neglect, particularly Sessions were focused on patient support and education and were from one’s mother; overprotection and childhood designed to optimize treatment response rather than offer formal sexual abuse (CSA) would result in poorer outcomes psychotherapy. Patients randomized to fluoxetine initially received 20 mg per day for three weeks, after which time the clinician couldadjust the dose up to a maximum of 80 mg. At six weeks, the meanfluoxetine dose of the sample was 28.1 mg per day, with a range of 10–80 mg. Patients taking nortriptyline initially received 25 mg forone night, 50 mg for one night, then 75 mg. Blood levels were taken after one week and dosing adjustments made based on clinicalresponse, side effects, and individual levels. At six weeks, the mean Depressed patients (111 females and 84 males) were recruited from nortriptyline dose was 93.5 mg per day, with a range of 50–175 mg.
various outpatient sources. After receiving study information, All patients who improved were encouraged to continue taking the patients gave written consent and were screened over the telephone medication for the six months. Further details regarding clinical by a research nurse, then seen for an initial assessment with a characteristics of the patients, dosing, monitoring, and follow-up care psychiatrist or senior psychiatric registrar. Eligible patients were free from any medication aside from oral contraceptives and occasionalhypnotics for a minimum of two weeks before study participation.
Exclusion criteria for the study included schizophrenia, a history ofmania, but not hypomania, severe alcohol and/or drug dependence, Depression severity was assessed with the clinician-rated or severe antisocial personality disorder, which may interfere with Montgomery–A˚sberg Depression Rating Scale (MADRS),[33] at study compliance. Patients with a current major medical illness were also excluded. After assessment, patients were randomly assigned to Three treatment outcomes were assessed. The first measure, either fluoxetine or nortriptyline. The study was approved by the adequate trial, recorded whether or not the patient took an adequate Canterbury Ethics Committee, New Zealand.
dose of the initially prescribed medication to which they had been Research Article: Abuse and Neglect as Treatment Predictors randomized for a majority of the first six weeks of treatment. Second TABLE 1. Sample characteristics, pre- and post- was the patient’s percentage improvement on the MADRS after six treatment of 195 depressed outpatients presenting for weeks, based on the change from baseline and represented a patient’s response in the short-term. Finally, looking longer term, two monthssustained recovery measured whether or not the patient sustained a ‘‘much improved’’ or ‘‘very much improved’’ score on the Clinical Global Impression scale[34] for a minimum of two months, based on the consensus assessments of both the treating psychiatrist and the All data analyses were performed using the Statistical Package for the Social Sciences (SPSS, V.15, 2006). For categorical outcome variables, logistic regression was used plus a two-way, between groups Analysis of Variance for known interaction effects. For the continuous outcome variable, univariate Pearson’s correlations were used, followed by multiple regressions for multivariate analyses.
The descriptive characteristics, pretreatment mea- sures, treatment, and three outcomes examined for the 195 depressed outpatients in the sample are presented in Table 1. The mean age was 32 years (711) with 37%o25 years old. Of the depressed patients, 57% were female; 62% had recurrent depression, while 64% had chronic depression, defined as lasting more than two years. Lifetime rates of comorbid diagnoses are Of the 195 patients, 10 did not rate their mother on the PBI, while 21 did not rate their father. As such, the maternal rating total was 185 and the paternal rating total was 174. The mean maternal care score was 21.5 (79.3); paternal care, 17.9 (79.6). The mean maternal protection score was 16.1 (78.4); paternal protection 14.7 (78.1). The mean PBI scores of this sample are comparable to other depressed clinical and community samples, which show a pattern of lower care and higher protection scores, in contrast to never-depressed community norms. For example, the mean maternal care score in this study is nearly one standard deviationless than local community norms.[3,35] When abuse is broadly defined to include physical, psychological, or sexual experiences, more than half of the the only one associated with failing to complete an sample (62%) reported abuse (58% of men and 63% of adequate six-week trial of medication.
women). When abuse was defined specifically as sexual, In previous analyses,[31] drug and gender influenced 18% reported CSA (8% of men and 26% of women).
completion of an adequate trial, with more patients More than three quarters of the sample (79%) randomized to nortriptyline failing to complete than completed an adequate trial of medication. The mean those assigned to fluoxetine. Furthermore, women percentage improvement on the MADRS after six were less likely to complete if prescribed nortriptyline; weeks was 60.3% (731.5%). Two months of sustained men less likely to complete if taking fluoxetine. Given recovery was achieved by 47% of the sample.
these previously known effects, drug and gender wereincluded as covariates along with the childhoodvariables, in a multivariate analysis. The multivariate analyses showed that low paternal care was still associated with failure to complete an adequate trial Table 2 shows that of the six childhood neglect and (OR 5 1.07, 95% CI: 1.00, 1.14). Once again, this was abuse variables, low paternal care (paternal neglect) is the only significant childhood predictor (P 5.036).
TABLE 2. Univariate logistic regression of baseline To illustrate this relationship, of those reporting high variables with an adequate trial of medicationa maternal protection (scores of 416), the meanimprovement was just 52%, compared with 67% improvement in those who did not report high TABLE 4. Univariate logistic regression of baseline variables with two months sustained recovery in those who completed an adequate trial of medicationa cPsychological, physical, or sexual abuse or threat dichotomized into ‘‘none,’’ ‘‘mild,’’ ‘‘moderate,’’ or ‘‘severe.’’ dChildhood Sexual Abuse dichotomized into ‘‘none,’’ ‘‘some,’’ TABLE 3. Univariate Pearson’s correlation of baselinevariables with percentage improvementa PREDICTORS OF TWO MONTHSSUSTAINED RECOVERY Table 4 shows that high maternal protection (maternal overprotection) was the only childhood variable associated with fewer participants achieving two months sustained recovery in those who completed Previous data have shown an interaction effect between age and drug, with patients under the age of25 less likely to achieve two months of sustained aBased on the patient’s change from baseline on Montgomery-A˚sberg recovery when taking nortriptyline compared with Depression Rating Scale after six weeks of treatment.
fluoxetine.[31] Even when including these covariatesalong with all the childhood variables in multivariate It should be noted that the scales for parental care analyses, maternal overprotection was the only child- (0–36) and protection (0–39) are quite different to the hood variable, which predicted fewer patients achieving abuse scale (0–3). Thus, odds ratios and their two months sustained recovery. There was a little confidence intervals cannot be compared across these difference between the univariate (OR 5 0.93; 95% CI: childhood variables. To illustrate the effect of low 0.89, 0.97; Po.001) and multivariate (OR 5 0.90, 95% paternal care on completion of an adequate trial, for those with paternal care scores r17, nearly 30% did To illustrate the relationship between maternal not complete an adequate trial of medication, while protection and two months sustained recovery, of those only 11% failed to complete an adequate trial with who reported maternal overprotection (scores416), only 39% achieved two months sustained recovery incomparison to 61% of those who did not reportmaternal overprotection.
Table 3 shows that maternal protection was inversely correlated with the percentage improvement in symp- The aim of this study was to examine whether toms of depression after six weeks of treatment.
adverse childhood experiences predicted response to As previous analyses have shown that patients under antidepressant medication in adult outpatients. Three the age of 25 randomized to nortriptyline had poorer key findings emerged. The first notable finding was outcomes at six weeks,[31] these interaction effects were that low paternal care (paternal neglect) was associated included as covariates, along with the childhood with not completing a six week trial of medication. The variables, in multiple regression analyses. Results second finding was that in both the short and longer showed that high maternal protection remained sig- term, high maternal protection (maternal overprotec- nificant in predicting a reduced improvement in tion) was associated with poorer response. The third symptoms of depression after six weeks of treatment.
finding was that the presence of abuse, whether defined Research Article: Abuse and Neglect as Treatment Predictors broadly or specifically as sexual abuse, did not predict adulthood. This speculation finds some support in response to any of the treatment outcomes. These Parker’s early examination of maternal overprotec- three findings, initially observed in univariate analyses, tion.[46] In his study, a mother who was rated by her remained significant in multivariate analyses, which child as being overprotective rated herself ‘‘as having a included all the childhood variables as well as potential more external locus of control,’’ (p. 307) or perceiving confounding variables such as age, drug, and gender.
that the environment and external factors have the Our first finding, that paternal neglect predicted greatest impact on what happens in her life. An external failure to complete an adequate trial of medication, has locus of control may operate in parallel with reduced not been reported. Researchers examining adequate expectations for treatment. Patients with lower trial or ‘‘dropout,’ as it is often defined, suggest patient expectations of improvement with treatment do less characteristics are the primary predictors of attrition.
well than those who have higher expectations.[44,47] General patient characteristics implicated include Patients who have limited experience thinking for comorbidities, age, race, and education level. Tedlow themselves may not readily form positive opinions et al.[36] comparing dropouts and completers in a about their treatment, thus reducing its efficacy.
fluoxetine-only trial, noted that dropouts had higher Our third finding, that abuse did not predict rates of histrionic or narcissistic personality disorder treatment outcomes in this pharmacotherapy trial, runs than completers, and higher anxiety levels. Higher counter to our initial hypothesis. Research suggests rates of anxiety in dropouts have also been found by strong evidence for the epidemiological role of child- Arnow.[37] Younger patients, subjects with less educa- hood abuse in the development of adult depression but tion and those belonging to an ethnic minority, are also a weaker connection between childhood abuse and more likely to drop out.[37,38] One study examined poorer response to treatment. A few studies have found childhood experiences in relation to dropout, but in the childhood trauma associated with treatment-resistant context of bulimic women. Those who dropped out depression[48] and dysthymia,[27] while another study rated their families as very poor at showing emotional found differential response to psychotherapy vs.
concern for each other.[39] Perhaps this finding is pharmacotherapy, with those patients reporting child- important in understanding why those reporting a hood adversity responding better to psychother- neglectful father dropped out of treatment. Psychody- apy.[24,25] Given that abuse by itself did not predict namic theory posits that supportive, caring parental any of the treatment responses, perhaps it is the quality attachment is a foundation for forming and maintain- of ongoing intra-familial relationships that has a ing social relationships in adulthood,[40] a concept greater impact on treatment than experiences of supported by research.[41–43] Carrying this notion discrete abuse in childhood, a hypothesis supported beyond social constructs, the professional relationship between patient and psychiatrist may be negatively In summary, a significant body of research links influenced by the patient’s experience of a neglectful various forms of childhood neglect and abuse to the father, to the extent that the patient is not sufficiently development of adult depression. This study indicates trusting the psychiatrist to take the prescribed medica- that those who reported having a neglectful father were less likely to take an adequate dose of the initially In both the short-term (percentage improvement at prescribed antidepressant medication for the six-week six weeks), and longer term (two months sustained trial. Additionally, those who reported an overprotec- recovery), maternal overprotection predicted a poorer tive mother had a poorer response to treatment in both response. While maternal overprotection has been short and longer term. Finally, abuse, whether physical, implicated in the development of adult depression,[1] sexual, or psychological in nature, did not predict no previous research was found regarding the specific treatment response to antidepressant medication.
role of maternal overprotection in predicting treatment Findings may inform clinicians’ decisions with respect outcomes. Three patient characteristics shown to to prescribing antidepressants to patients who report predict poorer outcomes in pharmacotherapy trials specific types of childhood adversity.
levels,[44,45] minority ethnicity, and concurrent anxietyissues.[45] Other predictors of poor outcome included Acknowledgments. The principal author thanks living alone[44] and poorer functioning and quality of Dr. Virginia McIntosh for her invaluable input; thanks also to Robyn Abbott and Isobel Stevens for their early Speculating on the role of maternal overprotection as work on this study, and to the local psychiatrists who a predictor of poor treatment outcomes, perhaps this finding suggests the importance of developing an Role of funding source: This research was funded by independent sense of self in childhood, as an over- a grant from the Health Research Council of protective mother is one who does not encourage her New Zealand, a grant from Lottery Health, and an child to think and act independently of her. Children unrestricted grant from Eli Lilly (New Zealand). These discouraged from independent thought and behavior sponsors had no further role in study design; in the are less likely to develop these skills for use in collection of analysis and interpretation of data; in the writing of the report; and in the decision to submit the 14. Sato T, Sakado K, Uehara T, et al. Dysfunctional parenting as a paper for publication. The Clinical Research Unit of risk factor to lifetime depression in a sample of employed the Department of Psychological Medicine, Christch- Japanese adults: evidence for the ‘‘affectionless control’’ hypoth- urch, is supported by the University of Otago and the esis. Psychol Med 1998;28:737–742.
Mental Health Division of Canterbury Health. Jeanette 15. Overbeek G, Ten Have M, Vollebergh W, de Graaf R. Parental Johnstone is supported by a scholarship from the lack of care and overprotection: longitudinal associations with DSM-III-R disorders. Soc Psychiatry Psychiatr Epidemiol2007;42:87–93.
Contributors: Johnstone is the principal author, Luty 16. Carter JD, Joyce PR, Mulder RT, Luty SE. The contribution of the study investigator, treating clinician, supervisor, temperament, childhood neglect, and abuse to the development Carter the supervisor, Mulder the study investigator of personality dysfunction: a comparison of three models.
and treating clinician, Frampton was the statistician, J Personal Disord 2001;15:123–135.
and Joyce was the principal investigator, treating 17. Heider D, Matschinger H, Bernert S, Alonso J, Angermeyer MC.
Relationship between parental bonding and mood disorder in sixEuropean countries. Psychiatry Res 2006;143:89–98.
18. Parker G. Parental characteristics in relation to depressive disorders. Br J Psychiatry 1979;134:138–147.
19. Parker G, Kiloh L, Hayward L. Parental representations of 1. Parker G. Parental Overprotection: A Risk Factor in Psychoso- cial Development. New York: Grune & Stratton; 1983:325.
2. Carter JD, Joyce PR, Mulder RT, Luty SE, Sullivan PF. Early 20. Parker G. Parental ‘‘affectionless control’’ as an antecedent to deficient parenting in depressed outpatients is associated with adult depression. A risk factor delineated. Arch Gen Psychiatry personality dysfunction and not with depression subtypes.
21. Hall LA, Peden AR, Rayens MK, Beebe LH. Parental bonding: 3. Oakley-Browne MA, Joyce PR, Wells JE, Bushnell JA, a key factor for mental health of college women. Issues Ment Hornblow AR. Adverse parenting and other childhood experi- ence as risk factors for depression in women aged 18–44 years. J 22. Rey JM. Perceptions of poor maternal care are associated with adolescent depression. J Affect Disord 1995;34:95–100.
4. Bifulco A, Moran PM, Baines R, Bunn A, Stanford K. Exploring 23. Mackinnon A, Henderson AS, Andrews G. Parental ‘‘affection- psychological abuse in childhood: II. Association with other less control’’ as an antecedent to adult depression: a risk factor abuse and adult clinical depression. Bull Menninger Clin refined. Psychol Med 1993;23:135–141.
24. Nemeroff CB, Heim CM, Thase ME, et al. Differential 5. Enns MW, Cox BJ, Clara I. Parental bonding and adult responses to psychotherapy versus pharmacotherapy in patients psychopathology: results from the US National Comorbidity with chronic forms of major depression and childhood trauma.
Survey. Psychol Med 2002;32:997–1008.
Proc Natl Acad Sci USA 2003;100:14293–14296.
6. Tonmyr L, Jamieson E, Mery LS, MacMillan HL. The relation 25. Heim C, Newport DJ, Mletzko T, Miller AH, Nemeroff CB.
between childhood adverse experiences and disability due to The link between childhood trauma and depression: insights mental health problems in a community sample of women. Can J from HPA axis studies in humans. Psychoneuroendocrinology 7. Collishaw S, Pickles A, Messer J, Rutter M, Shearer C, 26. Kaplan MJ, Klinetob NA. Childhood emotional trauma and chronic Maughan B. Resilience to adult psychopathology following posttraumatic stress disorder in adult outpatients with treatment- childhood maltreatment: evidence from a community sample.
resistant depression. J Nerv Ment Dis 2000;188:596–601.
27. Hayden EP, Klein DN. Outcome of dysthymic disorder at 5-year 8. Sachs-Ericsson N, Verona E, Joiner T, Preacher KJ. Parental follow-up: the effect of familial psychopathology, early adversity, verbal abuse and the mediating role of self-criticism in adult personality, comorbidity, and chronic stress. Am J Psychiatry internalizing disorders. J Affect Disord 2006;93:71–78.
9. Chapman DP, Whitfield CL, Felitti VJ, Dube SR, Edwards VJ, 28. Parker G, Roy K, Wilhelm K, Mitchell P, Austin MP, Anda RF. Adverse childhood experiences and the risk of Hadzi-Pavlovic D. An exploration of links between early parenting experiences and personality disorder type and disordered personality functioning. J Personal Disord 1999;13:361–374.
10. Dube SR, Felitti VJ, Dong M, Giles WH, Anda RF. The impact 29. Mulder RT, Joyce PR, Frampton CM, Luty SE, Sullivan PF. Six of adverse childhood experiences on health problems: evidence months of treatment for depression: outcome and predictors of from four birth cohorts dating back to 1900. Prev Med the course of illness. Am J Psychiatry 2006;163:95–100.
30. Joyce PR, Mulder RT, Luty SE, McKenzie JM, Sullivan PF, 11. Parker G, Tupling H, Brown LB. A parental bonding instrument.
Cloninger RC. Borderline personality disorder in major depres- sion: symptomatology, temperament, character, differential drug 12. Parker G, Hadzi-Pavlovic D, Greenwald S, Weissman M. Low parental care as a risk factor to lifetime depression in a community sample. J Affect Disord 1995;33:173–180.
31. Joyce PR, Mulder RT, Luty SE, et al. Patterns and predictors of 13. Narita T, Sato T, Hirano S, Gota M, Sakado K, Uehara T.
remission, response and recovery in major depression treated with Parental child-rearing behavior as measured by the Parental fluoxetine or nortriptyline. Aust N Z J Psychiatry 2002;36:384–391.
Bonding Instrument in a Japanese population: factor structure 32. Mulder RT, Joyce PR, Frampton C. Relationships among and relationship to a lifetime history of depression. J Affect measures of treatment outcome in depressed patients. J Affect Research Article: Abuse and Neglect as Treatment Predictors 33. Montgomery SA, Asberg MA. A new depression scale designed 43. Parker GB, Barrett EA, Hickie IB. From nurture to network: to be sensitive to change. Br J Psychiatry 1979;134:382–389.
examining links between perceptions of parenting received in 34. Guy W, ed. Clinical Global Impression (CGI). ECDEU childhood and social bonds in adulthood. Am J Psychiatry Assessment Manual for Psychopharmacology. Rockville: US Department of Health, Education, and Welfare; 1976.
44. Hirschfeld RM, Russell JM, Delgado PL, et al. Predictors of 35. Joyce PR. Parental bonding in bipolar affective disorder. J Affect response to acute treatment of chronic and double depression with sertraline or imipramine. J Clin Psychiatry 1998;59:669–675.
36. Tedlow JR, Fava M, Uebelacker LA, Alpert JE, Nierenberg AA, 45. Trivedi MH, Rush AJ, Wisniewski SR, et al. Evaluation of Rosenbaum JF. Are study dropouts different from completers? outcomes with citalopram for depression using measurement- based care in STARÃD: implications for clinical practice. Am J 37. Arnow BA, Blasey C, Manber R, et al. Dropouts versus completers among chronically depressed outpatients. J Affect 46. Parker G, Lipscombe P. Influences of maternal overprotection.
38. Warden D, Trivedi MH, Wisniewski SR, et al. Predictors of 47. Sotsky SM, Glass DR, Shea MT, et al. Patient predictors of attrition during initial (citalopram) treatment for depression: response to psychotherapy and pharmacotherapy: findings in the a STARÃD report. Am J Psychiatry 2007;164:1189–1197.
NIMH Treatment of Depression Collaborative Research Pro- 39. Waller G. Drop-out and failure to engage in individual gram. Am J Psychiatry 1991;148:997–1008.
outpatient cognitive behavior therapy for bulimic disorders. Int 48. Lara ME, Klein DN, Kasch KL. Psychosocial predictors of the short-term course and outcome of major depression: a long- 40. Bowlby J. Attachment and Loss. 2nd ed. Basic Books; 1999.
itudinal study of a nonclinical sample with recent-onset episodes.
41. Kendler KS, Kessler RC, Neale MC, Heath AC, Eaves LJ. The J Abnorm Psychol 2000;109:644–650.
prediction of major depression in women: toward an integrated 49. Edwards VJ, Holden GW, Felitti VJ, Anda RF. Relationship etiologic model. Am J Psychiatry 1993;150:1139–1148.
between multiple forms of childhood maltreatment and adult 42. Parker G, Barnett B. Perceptions of parenting in childhood mental health in community respondents: results from the and social support in adulthood. Am J Psychiatry 1988;145:

Source: http://empower-daphne.psy.unipd.it/userfiles/file/pdf/Johnstone%20M_%20-%202009.pdf