The Science of Mental Health
The Science of Mental Health
Edited BySteven E. Hyman
Edition 1st Edition
First Published 2002
eBook Published 15 October 2013
Pub. location New York
Pages 350 pages
eBook ISBN 9781315054322
Hyman, S. (Ed.). (2002). Depression. New York: Routledge, https://doi.org/10.4324/9781315054322
First Published in 2002. Routledge is an imprint of Taylor & Francis, an informa company.
TABLE OF CONTENTS
Subsyndromal Depressive Symptoms Major Depression Sample General Population Lewis M.D., Martin P. Paulus, M.D., Kenneth B. Wells, M.D., M.P.H.,
the relationship between "independent" events and depression (Williamson et al 1995), where a dependent event was defined as one that could plausibly be a consequence the respondent's own actions (e.g. being fired from a job) The fact that independent events were generally significant predictors of Brown & Harris 1978). However, another type of confounding could have led
respondent's (e.g. being fired from while independent event could (e.g. losing plant closing). independent events generally significant predictors depression those studies distinguished independent dependent events considered evidence events cause depression (e.g.
about completeness of controls and linearity-additivity of multivariate influ-ences (Holland 1986, Sobel 1990). It is important to emphasize in this regard that misspecification can invali-date the control variable approach even when confounding variables are cor-rectly measured. The most obvious example involves the case where the researcher assumes that confounding variables have additive effects. Given what we know about stress-modifying effects, it is almost certain that this is not the case in most applications. Common causes of event exposure and depression, such as a genetic liability or various aspects of personality or access to supportive social relationships, are likely to modify the impact of stressful life events on depression and lead to a biased estimate of the magni-tude of this impact in the absence of a correct specification of interactions between the events and the controls.
Because methodological problems enumerated above, studies cited above cannot interpreted providing unequivocal support aggregated stressful events cause depression. However, other studies provide support. studies exploit relationship between stressful event and a health outcome can interpreted based
from these studies that chronic role-related stresses are significantly associated chronically depressed mood & Ross We :dso know from the handful of life event studies that have included assess-ments of them that chronic stresses are often associated with an exacerbation flf the effects of life events on episodes !lIc life domains affected by the events are the same as the domains affected by the chronic stresses (Brown et al 1987, McGonagle & Kessler 19(0). lbe effects of life events on depression. These studies suggest that enduring of of the effects
(Mirowsky 19lN, Pearlin 19R9). major depression, especially domains affected events are domains affected chronic stresses (Brown 1987, McGonagle Kessler 19(0). Recent research begun focus mediators effects of events on depression. These studies suggest enduring stressful sequelae stressful events account effects researchers grapple seriously complex methodological problems involved in conduct- naturalistic studies of agg regate stress effects community samples. Focused studies of individual randomly occurring events are limited because events occur reasons random respect
in treatment effects rather than life event effects. As a result, the studies generally lack control groups and usually focus on people in treatment rather than on representative community samples of those who have been exposed to the event under investigation. Consequently, it is impossible to draw any inferences about stress effects (Burgess & Holstrom 1979). Despite such limitations, the results from these studies provide a rich source of information about the meanings of stressful events in the lives of the people who experience them, the coping challenges these events elicit , and the resources and vulnerabilities associated with successful and unsuccessful coping efforts. These results could be invalu-able to future researchers who attempt to apply the logic of nonexperimental causal analysis to new investigations of the effects of these events on depression. Such studies should include prospective designs and use carefully constructed control groups in an effort to estimate the magnitude of life event effects , the pathways through which these effects operate, and the modifying factors that lead these events to vary in their effects across respondents. The methodological issues are a good deal more complex in research on the relationship between chronic stress and chronic depression. The possibility that chronic role-related stress is an important determinant of chronic depression is certainly consistent with the observation that people who have chronic major depression or dysthymia often report ongoing problems in one or more of their central life roles. However, a major problem in interpreting this fact is that both the chronic role-related stresses and the chronic depression by definition have occurred for so long that deciding unambiguously which came first is difficult. No serious efforts address this problem of causal order. The researcher, however, may focus on stresses that can be assumed to have occurred randomly with respect to other risk factors of depression and to be inescapable, in which case matched comparison can be used to make causal inferences about long-term stress effects . A good example is the matched comparison of the parents of children having cancer, diabetes , or some other serious childhood physical disorder with the parents of healthy children. Disorders of this sort are quite common and occur, in most cases, for reasons that are unrelated to other risk factors for parental psychiatric disorder (Pless 1994). The small amount of research shows that these childhood physical disorders have significant psychiatric effects on the family (Jessop et al J988, Krosnick 1970). The more usual case, however, is one in which nonrandom exposure
chronic stress cannot ruled studies of relationship between chronic marital difficulties depression. Frequently plication systematic selection out of exposure (e.g. differential likelihood seeking a stressful based individual differences depresso- genic effects of stress) Standard longitudinal collection methods
ffects chronic stress o epression rt-term ffects events on episod epression. Nevertheless, there opportunities making provision onrand prospect resea rch vestigate entry first few years marri study selec cesses stressful noted there nside rable variation timated ending on wheth measur hecklist approach usuall traced to Meyer chart" summa information provid
placed these with a longer set of more concretely defined events within each of these categories. 3. The SRRS was a fairly short checklist that omitted a great many stressful events (e.g. rape, criminal victimization, witnessing a traumatic event) . Subsequent checklists added these events and sometimes customized event lists to special populations (e.g. Kessler et al 1992). In addition, methodological studies of the SRRS and the various checklists based on it were carried out. The main results Negative events are much more powerful predictors that lifc changc is not the central dimension linking stressful life events to life event effects. 2. Within the set of fairly serious events typically included in life event checklists like the SRRS, the usc of differential weights does not markedly increase the association between negative life event scales and measures psychological distress (Zimmerman 1983). 3. However, distinctions along a number of other dimensions (e.g. amount loss, amount threat , degree controllability consequences the event) do lead to substantial increases in the association between negative life event scales and measures distress (Thoits 1983). A key feature of the checklist approach is that all life events a given type are treated as equivalent. Death of a spouse, for example, was assigned a 100 point LCU score on the SRRS irrespective of the suddenness the death, the quality the marriage, or any circumstances surrounding the death that might have made it more stressful (e.g. the death occurred in an automobile accident in which the respondent was driving) or less stressful (e.g. the death occurred peacefully after the spouse had spent many years suffering from a degenerative ilIncss). There is clear evidence from focused studies individual events that the strength the relationship between life events and depression increases substantially when these sorts of distinctions are made. Based on this observa-tion, a number researchers have attempted to modify the checklist approach in such a way as to consider these distinctions. Two strategies have been used to make these adjustments. One is to allow each respondent to assign a subjective weight to his or her own events (Sara-son et al 1978). The other is to use objective information about the person and his life situation to construct an independent judgment how stressful the event would be for a typical person in that same situation. The first of these strategies has been rejected as confounding measurement the event with emotional reaction to the event (Turner & Wheaton 1995, Zimmerman 1983) and has consequently been abandoned. The second strategy has evolved into what has been termed the "contextual" approach to rating life events.
these studies follows: Negative events are powerful predictors mental health comes positive events (Zautra Reich 1983), which conclusion changc central dimension linking stressful events psychological disorder weighting approach leads underestimation event effects. Within fairly
The intellectual origins of the contextual approach can be traced to the work of Blown & Harris (1978), who developed a method of using a rating panel to assign scores on a variety of dimensions to stressful life events. According to this rating scheme, the death of a neighbor would be rated as more severe than otherwise if the respondent who reported it was a rural elderly person who lived alone and was a housebound invalid whose main source of social contact was the weekly visit the neighbor for lunch. To obtain adequate contextual the events believed to be relevant to contextual ratings. Detailed the event in relation to the onset of the depr ession. Precise dating is also used to identify the aspects the event that arc thought to affect depression onset. Methodological studies suggest that such intensive interviewing is much more effective than the comparatively mechanical procedures used in the administration of checklists in avoiding misdating events (McQuaid et al 1992), communicating to the respondent the importance of accurate recall (Cannell et al 1981), and facilitating the use of memory aids to improve recall events and accurate dating event occurrence (e.g. calendars, visual repre-sentations, reminders personally salient events reported previously in the interview). Such aids have been shown to signifieanlly improve accuracy of life event reports (Sobell et al 1990). Furthermore, the use of contextual ratings has been shown to substantially reduce the response errors in checklist meas-ures due to individual differences, such as checking off the occurrence of a "serious physical illness" that, in fact, was only a cold, or otherwise giving reports that are inconsistent with the implied severity thresholds in the check-list events (Kessler & Wethington 1991, Raphael et al 1991). There arc two practical problems with the contextual rating approach to life events measurement. The first is that great care is needed to make sure infor-mation about the occurrence depression after the event docs not contaminate the ratings of context, possibly by highlighting to the interviewer an aspect of the respondent's life situation that would not have been known were it not for the fact that the respondent became depressed. Concerns have been raised that some users of the Brown & Harris method are not sufficiently attentive to this possibility. which can result in confounding of stress ratings with outcomes (Bebbington [986). The second practical problem is that the intensive inter-views used to make contextual life event ratings arc very labor intensive. Interviewers must be highly trained. Interviews can take as much as five hours to complete. A complicated rating scheme typically requiring several addi-tional hours to complete is needed to review interview audiotapes. Weekly panel meetings sometimes last an entire day to review these ratings (Brown
neighbor obtain adequate contextual information for rating events, Brown Harris method requires intensive personal interviews qualitative probes specify precisely charac- teristics events believed relevant contextual ratings. Detailed probing establish event relation onset ession. Precise dating identify aspects
). Several attempts have been made to evaluate whether a short-cut can be devised to approximate the Brown & Harris method (Cooke 1985, Dohren-wend al \993, Kessler & Wethington 199\, Miller & Salter 1984) . The most promising that appears to generate information similar to the original in a shorter inter-view period and with fewer demands for complex postprocessing (Wethington et al 1995) . ing factors: aspects of the personal and situational environments of people exposed to stressful events that are associated with variation in the impact these events on their probability of becoming depressed. The contextual rating method subverts this investigation by absorbing information about stress modifiers into the ratings of life event severity. Indeed, the information used by the Brown & Harris ratings panel can be seen as hypotheses concerning stress modifiers that never have a chance to be investigated because they are assumed in rating event severity. A clearly preferable approach would be to subject these hypotheses to empirical test. The difficulty doing so until now, though, has been that studies the sort using the Brown & Harr is method have all had quite small sample sizes due to the labor-intensive nature of the method. This has meant that the number subjects has been too small to test specific hypotheses about the modifying effects particular aspects con-text.
these version method, still being pilot tested, appears generate information similar original inter- period fewer demands complex postprocessing (Wethington 1995) There problem contextual rating method. described below, events researchers greatly interested in stress-modify- confounding of measures event severity stress modifiers opposite well, suggesting individual differences emotional reactivity when, event people experience illustration is found in research relationship between widowhood depression, which
sions that lead to the effects of the event on depression and then measure these dimensions longitudinally in a sample of people who were exposed to the event and in an appropriate comparison sample of people not exposed to it. Standard multivariate procedures have then been used to study the mediating effects the stress dimensions on the overall relationship between the event and depression. Umberson et al (1992). for example. examined how much the hood on stress dimensions versus differences in the impact effect of widowhood on men than women is. at least in part, because the death a spouse leads to a number secondary stresses for men that do not exist for women. For example. death a spouse leads to a significant decrease in contact and emotional closeness between the surviving parent and adult chil-dren among widowed men but not widowed women. A consistent result in these studies has been that most of the association between some stressful events and depression is due to a mediating effect on role-related stresses. Caution is needed in interpreting this result. however. in light the fact that only one the three studies (Aseltine & Kessler 1993) was truly prospective (i.e. assessed both depression and the mediators before the onset of the stressful event) and none them controlled for other possible confounding effects. Future work on unpacking life event effects needs to be based on prospective designs that use carefully matched control groups and use intensive interview methods and contextual ratings to define the interven-ing chronic stress dimensions.
stress dimensions overall relationship between event depression. Umberson (1992). example. examined observed gender difference impact widowhood depression national sample male-female differences impact widow- dimensions versus differences impact these dimen- sions on depression. analysis documented greater depressogenic results discussed short-term effects of recent stressful events on episodes of depression. Another stress research concerned long-term effects of previous stresses (usually either
same investigators in two other large-scale community surveys yielded similar results (Kessler et al 1996a). This investigation has the potential to yield important information about life course variations in the effects of traumatic events. By stud ying differential effects as a function of time since occurrence the event, it might be possible to document variation in both initial impact and in the length the risk period the event and the age look for fairly rapid onset of the disorder after the traumatic event, while an investigation the association between childhood sexual abuse and lifetime depression might be interested in the possibility more long-term delayed reactions. By investigating the possibility of variation in the long-term effects of traumas as a function of the respondent's age at occurrence, in comparison, the researcher can investigate the hypothesis that some traumas have their most powerful health-damaging effects during certain critical developmental phases the life course.
event, might possible document variation initial impact length period associated event. would expect these effects depending event and exposure. investigation association between childhood parental depression, example, would probably rapid onset disorder after event, while Related research considers long-term effects traumatic stress on current adult depression. number carried out psychiatric patient samples and general population samples. These studies based retrospective persons
ByStress Current Depression
either first onset or episode recurrence. Their studies examined relationships between childhood adversities and speed of episode recovery. Replications these studies are needed focused samples. understanding history depression could important in evaluating short-term effects stressful events on episode onset. literature differences depression provides interesting illustration situ- ation. Point prevalence depression higher among women
current stress, which failure control history could serious estimation History depression, example, thought to influence interpersonal style provoke other people toward depressed people nonsupportive could increased exposure interpersonal loss events (Coyne 1976, Monroe Steiner 1(86). possibility implies number examined ationship between stressful ences occurred before onset of depressive episodes and speed of recovery these episodes. carried out patient samples (Brugha 1990, Huxley 1979, 1993, Keller 1986, 1981, Monroe 1992, Parker 1988, ssman
that reverse or resolve an earlier stress are associated with more rapid recovery (e.g. reemployment after a job loss). An especially interesting result is that otherwise negative events can sometimes lead to episode recovery either be-cause they resolve an ongoing difficulty (e.g. separation from an abusive spouse) or because they put the event that triggered the episode into perspec-tive (e.g. a ncar-death experience in an automobile accident leading to the realization that a recent job loss was not really life-shattering).
until now investigations document existence aggregate effects stress on depression. However, contemporary research stress and depression typically accepts association as a given focuses consistent finding majority people exposed
introduce similar controls seriously compromises our ability to draw any infer-ences from the larger literature on stress-modifying factors in depression. is important to reemphasize two earlier points in making this criticism of the literature on stress-modifying First, recurrent major depression can profoundly affect most the individual-level and environmental factors that have been studied in the modifier literature. Consistent with this observation, a lifetime history of of
effects. individual-level factors studied modifier literature. Consistent observation, Kessler Magee (l994b) found strong relationships between retrospective reports lifetime history depression variables normally considered stress modifiers their nationally representative survey. Second, history depression modifier. addition distinction between people viously recent research shown number subsarnple people history is related stress reactivity. (1986) first report phenomenon noted though stressful events often precipitate early episodes of depres-
ByPrior Episodes Depression Modifier
vulnerability to depression. The results showed that genetic liability is associ-ated with a threefold increase in the stress-depression relationship. lt is also important to recognize that genetic influences can account for what appears to be an effect of a substantive stress-modifying variable. Kessler et al (1992) found in a study of the stress-buffering effect of social support on the relationship between life events and depression that it was not support itself but the genetically determined component of social support that became more strongly associated with depression in the presence of stressful life events . Future research on stress modifiers should use genetically informative designs , such as twin or adoption studies, to investigate related possibilities in more detail.
evidence reviewed above clearly shows inventories predict subsequent depression. smaller number studies of people exposed single major events provide strong evidence least association events causing depression. other studies relationship reciprocal
Iw'-,-d to the events, which made it impossible to distinguish stress-modifying cllc crs from the effects way amon g people who were not exposed to the event s used to define the It is unclear why there has not been more collaboration between interven-nary differences in orientation are involved. Prevent ive interventions are typi-cally carried out by clinical or community psychologists, while naturali stic stress studies arc typically conducted by personality/social psychologists, epi-demiologists. and sociologists. Future advances in our understanding the relationship between life event s and depre ssion , in my view , will requ ire a collaboration between both perspect ives. The above review has repeatedly implied why such a collaboration would be useful to naturalistic researchers but has also emphasized that these re-searchers must work harder at approximating the conditions exp eriments to clarify the meanings of their results. One unique way doing this is to work with interventionists to construct control group s of people who were not exposed to the events under investigation and to use the manipulation of intervention exposure with parallel measurement in the control group to facili-tate analy sis of stress-modifying effects. It is also important to recognize, though , that there is anoth er side of the exchange: Nonexperim ental studies are equally important for inte rvention spec ialists. Such studies are needed to select intervention targets and to interpret the pathways involved in the effects of successful interventions. Once reliable data on pathways are obtained, nonex -perimental studies should search for factors that might effectively block these pathways. Thi s kind of iterative cross-ferti lization is our best hope for advanc-ing research on stressful events and depression in the future . Literature Cited
presumed modifiers operating exactly people exposed event s define intervention subjects unclear there collaboration between interven- tionists nonexperimental stress researchers. conceivable discipli- differences orientation are involved. Prevent typi-
disorders, including major depressive disorder, bipolar affective disorder their minor variants, poten- tial1y disabling conditions encountered physicians areas of medical practice. morbidity caused these disorders greatly underappreciated survey conducted World Health organization (WHO) revealed
episodes recurrent depression bipolar disorder appears become autonomous, or stress, as lifetime course illness unfolds refers progressive change vulner- ability as kindling, proposes results impact repeated episodes illness brain functioning, perhaps concert normal changes accompany aging. process increases incidence least adult population suffer disorder point, lifetime prevalence rates cumulative risks specific major mood disorders major depressive disorder bipolar disorder Minor forms ofmajor depressive disorder include dysthymia
pears that earlier onsets also are associated with greater lifetim e incidence and higher rates of recurrence (16) . not diminish the importance the depressive and bipol ar syndromes that be-gin in late life. Majo r depressive disorder and dysthymia are more common among women role cannot be discounted, although the impact gender differen ces in economic power, social support, social role conflicts, threatening forms oflife stress, and characteristic ways ofcoping
course, an earlier modal diminish importance depressive bipol depressive disorder dysthymia common among women (morbid 2.0-fold greater observed reproductive hormones hormonal pattern
The term rapid cycl ing is used when a patient with bipolar disorder suffers four or more episo des per year. Women are overrepresented among the bipolar 2 and rapid cycling subtypes (15, 21). Within the bipolar I grouping, women also tend to suffer a greater proportion of ofdepression will develop a re-
depre ssive episodes (15). Major depressive disord several subforms of clinical import. noted above, people suffer develop current (also known unipolar) disorder. Early course illness, it is uncomm unipolar disorder switch bipolar, especially onset of depression before (22). After three depres- inci- dence). three principal clinical subtypes depressive episode are melan- cholic psychotic, Melancholia, which tends to common after age represe nts severe end spectrum and charac signs and symptoms such as anhedonia (includ diminished libido),
Some recurrent depressive episodes follow a seasonal pattern, especially if they tend to begin in the fall or winter (26). Such depressions are more com-monly nonmelancholic and often are characterized by one or more reverse neurovegetative symptom . Many, ifnot
most, eventually criteria polar disorder (more typically the form), there artificial manipulation photoperiod, either moving equatorial locale using bright white lights phototherapy, suppress inherent seasonality. Suicide dreaded complication disorders, lifetime death rates 10-15% typically reported. Stated another disorders account about deaths by suicide, which, turn, ranks between second and sixth common cause death across adult groups (27). Suicide common among elders, individuals
tions ; both are clearly specified in treatment manuals; both use a present-oriented and pragmatic approach; and both are highly focused on the patient's depression and on the problems believed relevant to the onset and maintenance the depressive episode. IPT examines the relationship between terpersonal deficits . The major goals ofIPT ofrecurrence
depressive episode. examines relationship between depression interpersonal problems causes or effects patient's current depressive episode. techniques ofIPT developed manage basic interpersonal problem areas: unresolved grief, transitions, interpersonal disputes (often marital disputes) weeks amitriptyline. Patients received weekly medication, medica- tion, placebo therapy. Relapse rates highest for patients receiv- placebo alone (36%). Relapse rates other three active treatment groups medication alone, 16.7% alone, and 12.5% combined medication.
A second study, the Pittsburgh study rent Depression (39), contrasted IPT-M with maintenance pharmacotherapy of of of ofma- ofcognition 50). 53) ofparity ofcognitive-
Maintenance Therapies Recur- Depression (39), contrasted IPT-M maintenance pharmacotherapy (imipramine), combination pharmacotherapy/psychotherapy, condition (placebo therapy) period three years depressed patients clear histories recurrent depression treated acutely imipramine. Active medication
acute treatment (55,56). question minds those involved in unipolar disorders relative efficacy of older tricyclic antidepressants (TCAs) versus newer selective serotonin reuptake inhibitor antidepressants (SSRIs) Hundreds studies document efficacy TCAs. In CPR's meta-analysis (45), intent-to-treat efficacy
demonstrating the importance of dose adequacy in the pharmacother apy of unipolar disorders. An earlier study (63) documented the negative effect s of underdosing in the acute treatment tive effects of stopping treatment too soon . A more recent paper focused on the outcomes, as did a half-dose strategy in a controll ed trial (65); the cumulative probabil ity of remaining well with the half-dose strategy over a three-year pe-riod was less than 30%. Although data on the TCAs clearly indicate that a maximum tolerated dose strategy (generall y equivalent to approx imately 200 mg imipramine) represents a highly effective acute and maintenance treat-ment, strategies for maximizing the efficacy of long-term maintenance treat-ment with SSRls have yet to be developed.
unipol depression, nega- effects of stopping treatment too paper focused importance adequate maintenance treatment (64). analysis sugges patients maintained their acute episode, outcomes achieved domized controlled trial However, lower doses achieved relatively Among important recent advances unipolar disorders examination treatment efficacy specific subtypes, includ ing minor depression, dysthymia depression, atypical depression. study described above provided major impetus
ByVarious Subtypes ofUnipolar Depression
So-called atypica l depression is generally defined by the presence "re- the absence rate clinical tria ls conducted by his group and concl uded that imi- to placebo, the monoamine oxidase inhib itor (MAO!) phenelzine is signifi-cantly more effec tive than both imipr amine and placebo, and anxiety (in-cluded in some earlier definitions of atypica lity) was not associated with dif-ferential medicatio n response. Subsequently, Stewa rt et al (74) found that phenelzine, but not imipramine, had significant prophylactic effects for those with atyp ical depression. Our own researc h group has focused on the treatment the recurrent form unipolar disorder, once considered a particular subtype unipolar condi-tions . Recent epidemiological evidence (59, 60), however, suggests that this may be the predominant form unipolar disorder. We have demonstrated that maintaining antidepressant treatment at acute dose levels can prevent new episodes in recurrent unipolar patients, both in midlife (39) and in late life (40) . As implied above, in both the midlife and late life studies we also found a modest effec t for monthly sessions of IPT.
versed" vegetative symptoms (overeating, overs leeping), other symptoms include ction sensitivity and "leaden" paralysis, absence melancholia. Stewart reviewed eight sepa- clinical conducted group pramine treatment atypical depression, marginally superior controversy long-term treatment bipolar disorder impor- tance treatment reflected blood level) in tection bipolar episodes. Gelenberg random assigned subjects strategy and found patients maintained lithium level of0.8 mEq/L significa better protection against
not appreciably less effective than lithium in protection against new episodes, lithium is clearly superior in protecting bipolar patients against suicide . These results, again, point to the need for the clinician to balance the relatively poorer tolerability a compound against important prophylactic advantages when A final issue in the treatment of of ofbipolar ofour ofany
important prophylactic advantages tolerated. final issue in bipolar disorder relatively unrecog- nized difficulty achieving adequate remission bipolar depressive condi- tions group remission pression (defined weeks with Hamilton Depression Rating
Acknowledgments Lewinsohn, P.M., Rohde, P., Seeley, J.R., Klein, D.N., and Gotlib, LA. "Natural Course of Community of
Adolescent Major Depressive Disorder Sample: Predictors Recurrence Young Adults." (2000): 1584-1591. Copyright 2000, American Psychiatric Association. Reprinted permission. Mueller, T Leon, Keller, M.B., "Recurrence after Recovery Major Depressive Disorder during Years Observational Follow-up." (1999): 1000-1006. Copyright 1999, American Psychiatric Association. Reprinted permission. Schulz, S.R., Ives, D.G., Martire, L.M., Ariyo, A.A., "Association between Depression Mortality Older Adults." (2000): 1761-1768. Reprinted permission American Medical Association.
Acknowledgments 301 Duman, R.S., Heninger, G.R., and Nestler, E.]. "A Molecular and Cellular Theory of Arch GenPsychiatry of CB. of on of Bio!Psychiatry PET 156 .H. Neurosci Annu ReoMed50 Markowitz.j.C; .H., Eng!] Med ArchGenPsychiatry
Depression." (1997): 597-606. Reprinted permission American Medical Association. Heim, Nemeroff, Impact Early Adverse Experiences Brain Systems Involved Pathophysiology Anxiety Affective Disorders." (1999): 1509-1522. Reprinted permission Elsevier Science. Lancaster,].L., ecipro cal Limbic-Cortical Function Negative Mood: Converging Findings Depression Normal Sadness. (1999): 675-682. Copyright 1999, American Psychiatric Association. Reprinted permission. Sheline, Sanghavi, Gado, "Depression Duration