Course and Outcome of Bipolar Disorder Philip B. Mitchell, Dusan Hadzi-Pavlovic, and Colleen K. Loo
Contents 1 2 3 4 5
Historical Legacies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 The Course of BPD Prior to the Introduction of Modern Psychotropic Agents . . . . . . . . . . . . 2 Methodological Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Childhood and Adolescent Antecedents of BPD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 The Onset of BPD: Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5.1 Specific Age-Related Presentations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 6 The Onset of BPD: Polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 7 Subtypes and Patterns of BPD Based on Course of Illness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.1 Bipolar I and II Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.2 Rapid-Cycling Illness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.3 Nature of the Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.4 Predominant Polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 8 Duration and Outcome of Individual Episodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9 Recurrence Rates and Clinical Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 9.1 Retrospective or Cross-Sectional Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 9.2 Prospective Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 10 Predictors/Correlates of Recurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 11 Other Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 11.1 Disability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 11.2 Death Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 12 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
P.B. Mitchell (*) and D. Hadzi-Pavlovic School of Psychiatry, University of New South Wales, Sydney, NSW, Australia Black Dog Institute, Prince of Wales Hospital, Sydney, NSW, Australia e-mail: [email protected] C.K. Loo School of Psychiatry, University of New South Wales, Sydney, NSW, Australia Black Dog Institute, Prince of Wales Hospital, Sydney, NSW, Australia Mental Health Unit, St George Hospital, Kogarah, NSW, Australia
H.K. Manji and C.A. Zarate (eds.), Behavioral Neurobiology of Bipolar Disorder and its Treatment, Current Topics in Behavioral Neurosciences 5, DOI 10.1007/7854_2010_66 # Springer‐Verlag Berlin Heidelberg 2011, published online 11 August 2010
1
2
P.B Mitchell et al.
Abstract Bipolar disorder (BPD) is capricious illness. For some, it is a condition of few episodes; for others, it is unremitting. For some, elevated moods predominate; for others, depression is the major key. For a minority, the condition is predictably cyclical; for most, it is unpredictably chaotic. This chapter examines those studies that have attempted to draw coherence from this enigmatic disorder. Where possible, we will focus on data derived from prospective longitudinal investigations, while using as necessary the more limited retrospective or cross-sectional reports. For the sake of parsimony, we will limit discussion to those studies that have used the conservative historical definitions of BPD (as used in DSM-IIIR, DSM-IV, and ICD-10), eschewing the recent controversial concepts of “pediatric” BPD and “soft BPD spectrum.” Keywords Bipolar disorder Cycling Longitudinal Outcome
1 Historical Legacies The current distinction between the two major psychoses [schizophrenia and bipolar disorder (BPD)] that underpins the DSM-IV and ICD-10 nosologies derives from the writings of Emil Kraepelin (1899). Drawing on the work of his nineteenth century German predecessor Kahlbaum (Angst and Gamma 2008; Healy et al. 2008; Hippius and Mu¨ller 2008), Kraepelin’s central differentiation of manicdepressive psychosis from dementia praecox was that the former was a disorder of “good prognosis.” He described a characteristic pattern of full remission between the manic or depressive episodes, as opposed to the deterioration or chronicity which he observed in dementia praecox. However, Kraepelin also described interepisodic “chronic mild weaknesses” in manic-depressive psychosis, i.e., low-grade manic and depressive symptoms, reflecting his belief (Angst and Gamma 2008) of a continuum of severity from health to illness.
2 The Course of BPD Prior to the Introduction of Modern Psychotropic Agents A major issue when considering the course of BPD is the impact of medications – not only in terms of improving outcome but also the potential for destabilizing the illness, for example, manic episodes induced by antidepressants or lithium withdrawal (Suppes et al. 1991). This chapter will not review evidence-based therapies for BPD, which are dealt with elsewhere in this book. Nonetheless, it is pertinent,
Course and Outcome of Bipolar Disorder
3
where possible, to examine the course of this condition in its “native” form, prior to the introduction of modern psychotropic agents in the 1950s. An intriguing overview of the natural course of BPD in the pre-drug era (in descriptions roughly dating from Esquirol in the nineteenth century to Astrup in the mid-twentieth century) was recently published by Alvarez Ariza et al. (2009). Alvarez Ariza concluded that there were a number of commonalities between these historical reports: a high probability of recurrence; onset usually with a depressive episode; manic episodes of shorter duration and more likely to fully remit than depressive episodes; and, overall, good recovery between episodes. Perhaps, the most comprehensive data on the natural course of BPD before the introduction of modern therapies were those collected in the Iowa 500 study (Winokur and Tsuang 1996). Using files from the Iowa Psychopathic Hospital, Winokur and colleagues obtained clinical details on BPD patients admitted between 1934 and 1944. As that hospital had routinely collected follow-up information from patients, relatives, or clinicians every 6–12 months, it was possible for Winokur’s group to determine the progress of this condition in the era prior to the advent of lithium, neuroleptics, or antidepressants. Informative data were obtained on 87 patients with BPD who had been followed up between 6 months and 20 years (with an average of 2.2 years). Over that period of time, 54% recovered fully at some stage, while 21% were continuously symptomatic over the follow-up period. This pattern was not distinguishable from unipolar depression, but contrasted markedly with schizophrenia, in which 78% showed continuous illness and a further 13% deteriorated. Most BPD patients took 2–3 years to recover from their index episode (of mania or depression); however, for those able to be followed up over 10–20 years, all fully recovered at some stage. About 5% of those initially diagnosed with depression went on to develop mania over the follow-up period, most within 3 years of the index admission.
3 Methodological Issues Interest in both the natural course of BPD and the response of this disorder to treatment has increased, leading to increasing recognition of the limitations of the current nomenclature regarding the course and outcome of this condition. Recently, a working group of the International Society for Bipolar Disorders published a report on this issue (Tohen et al. 2009). In that paper, consensus opinion was reached regarding the definition of nine key terms (response, remission, recovery, relapse, recurrence, subsyndromal states, predominant polarity, switch, and functional outcome) commonly used to describe course and outcome in BPD. Pertinent to this chapter, Tohen and colleagues distinguished between “relapse” (which they
4
P.B Mitchell et al.
defined as a new episode occurring within 8 weeks of an index episode) and “recurrence” (a new episode occurring after eight weeks) – definitions similar to those recommended for major depressive disorder (MDD) by Frank et al. (1991). Because most of the BPD literature had hitherto not distinguished between these time intervals, and tended to use these terms interchangeably, we will arbitrarily use the term “recurrence” here to cover both concepts.
4 Childhood and Adolescent Antecedents of BPD It has long been recognized (based largely on retrospective reports) that BPD often presents for the first time during adolescence. In the first prospective study of this, Geller et al. (2008) demonstrated a continuity of childhood/adolescent onset DSMIV BPD-I with adult illness. Elucidation of replicable antecedent clinical or biological abnormalities prior to the onset of syndromal BPD would assist clinicians in early diagnosis of this condition and enable development of early identification and intervention programs. In recent years, there has been a growing recognition of the importance of identifying such antecedent features, with a concurrent increased number of crosssectional and longitudinal “high risk” studies of young offspring of parents with BPD (recently reviewed by Duffy 2009). Two cross-sectional studies (Lapalme et al. 1997; Birmaher et al. 2009) reported increased rates of syndromal BPD and MDD in these children, as well as increased rates of non-specific symptoms such as anxiety and subthreshold mood symptoms. There was no increase in the rates of attention-deficit hyperactivity disorder (ADHD) in those studies, an issue of pertinence in view of the high rates of reported comorbidity between that disorder and “pediatric” BPD (Biederman 2006). Similarly, the prospective longitudinal studies (Akiskal et al. 1985; Hammen et al. 1990; Egeland et al. 2003; Hillegers et al. 2005; Shaw et al. 2005; Duffy et al. 2009) have not found increased rates of childhood or adolescent ADHD in these “high risk” samples, nor have they demonstrated increased rates of pre-pubertal mania (which would be expected if “pediatric” BPD were a true antecedent of adult BPD-I). These prospective trials found higher rates of anxiety disorders, major mood disorders (mainly depression), and subthreshold depressive and “activation” symptoms. Other prevalent symptoms included sleep disturbance, distractibility, and a greater sensitivity to stress. In a retrospective and uncontrolled investigation of the occurrence of any more “immediate” prodromal symptoms, Mantere et al. (2008) found that prodromal symptoms were present in 45% of those with BPD-I and 50% of those with BPD-II. The mean duration of such symptoms was 31 days prior to syndromal onset and did not differ between BPD-I and BPD-II disorders. Most symptoms were moodcongruent, though some were non-specific features, particularly anxiety.
Course and Outcome of Bipolar Disorder
5
5 The Onset of BPD: Age Most patients with BPD report that the illness first presents in the mid-to-late teenage years or twenties. The mean age of onset reported in three large datasets (STEP-BD, Stanley Foundation Bipolar Network, and the Sydney Black Dog Institute; reviewed by Mitchell et al. 2009) ranged from 17.4 to 22.9 years. Onset occurs earlier in those with a positive family history of BPD (Taylor and Abrams 1981). Furthermore, a number of reports indicate that age of onset is genetically determined. Bellivier et al. (2003) reported three genetically determined age of onset subtypes: early (mean 17 years), intermediate (25 years), and late (40 years). In a similar study, Hamshere et al. (2009) reported on a mixture analysis that found three distinct age of onset groups for BPD-I: 22 years, 25–37 years, and 40 years. Hamshere and colleagues described differential clinical characteristics in these groups. The early-onset group had a stronger family history of affective disorder, more frequent suicide attempts, a greater proportion of rapid-cycling patients, more episodes of mania, and more episodes of depression. This suggestion of a greater severity of illness in the early-onset group has been supported by Perlis et al. (2009) who found that those who had a first episode before the age of 13, had (over their lifetime) fewer days euthymic, more impaired functioning, and earlier recurrence. Coryell et al. (2009) also reported that an earlier age at onset portends greater depressive symptom burden. Lin et al. (2005) reported linkage of a number of loci with age of onset of the first manic (but not depressive) episode.
5.1 5.1.1
Specific Age-Related Presentations Postpartum BPD
For many women, BPD presents for the first time in the postpartum period. For those with pre-existing BPD, the postpartum period is a period of high risk for relapse (Freeman et al. 2002). Munk-Olsen et al. (2009) reported that 27% of mothers with prior BPD were admitted within 1-year postpartum, and that a previous diagnosis of BPD was the strongest predictor of a postpartum psychiatric admission.
5.1.2
Late-Onset BPD
Late presentation of BPD (onset in the 60s or later) is less likely to be associated with genetic factors, rather related to neurological conditions such as cerebrovascular disease or traumatic brain injury, or more subtle organic processes such as small vessel disease in the brain; for instance, Tamashiro et al. (2008) reported more white matter hyperintensities in those with late-onset presentations. While
6
P.B Mitchell et al.
late-onset manic presentations have usually been associated with greater morbidity and mortality (Tohen et al. 1994), a recent report (Oostervink et al. 2009) from the prospective, longitudinal European EMBLEM study found a more favorable outcome in the late-onset (>60 years) group compared to an early-onset elderly BPD sample.
6 The Onset of BPD: Polarity While most retrospective studies have reported a depressive onset to be more common than mania, this has not been consistently demonstrated. In the three large datasets reviewed by Mitchell et al. (2009), 52–53% of patients reported their first episode to be depressive, with 16–26% reporting the initial presentation as hypo/manic. A number of retrospective studies have explored clinical correlates of the polarity of the onset episode. Perlis et al. (2005) reported on 704 BPD-I patients from the STEP-BD study. Depressive onset was associated with an earlier onset of illness, more lifetime depressive episodes, and more time depressed. Examining the UK Wellcome Trust genetic sample, Forty et al. (2009) reported similar findings. Using linear regression, initial onset with depression (compared to mania) was associated with earlier age of onset, predominant lifetime depressive polarity, more frequent and more severe depressive episodes, and less prominent lifetime psychotic episodes. Another means of assessing the phenomenon of the illness onset being depressive is to prospectively evaluate a sample of MDD patients, investigating for rates and correlates of “conversion” to BPD. First, in terms of rates of such “conversion,” Angst et al. (2005) reported on a prospective evaluation of hospitalized depressed patients. They found a conversion rate of 1% per year to BPD-I and 0.5% per year to BPD-II disorder. Males and those with an early onset of depressive illness were at greater risk of converting to BPD-I, whereas BPD-II conversion was associated with being female, later onset of depression, and a positive family history of mania. Examining a community sample of 14 to 24-year olds followed up over 10 years, Beesdo et al. (2009) reported that 3.6% of those with initial unipolar depression subsequently developed hypo/mania, with risk particularly high in those with adolescent onset (
Contents 1 2 3 4 5
Historical Legacies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 The Course of BPD Prior to the Introduction of Modern Psychotropic Agents . . . . . . . . . . . . 2 Methodological Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Childhood and Adolescent Antecedents of BPD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 The Onset of BPD: Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5.1 Specific Age-Related Presentations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 6 The Onset of BPD: Polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 7 Subtypes and Patterns of BPD Based on Course of Illness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.1 Bipolar I and II Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.2 Rapid-Cycling Illness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.3 Nature of the Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.4 Predominant Polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 8 Duration and Outcome of Individual Episodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9 Recurrence Rates and Clinical Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 9.1 Retrospective or Cross-Sectional Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 9.2 Prospective Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 10 Predictors/Correlates of Recurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 11 Other Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 11.1 Disability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 11.2 Death Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 12 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
P.B. Mitchell (*) and D. Hadzi-Pavlovic School of Psychiatry, University of New South Wales, Sydney, NSW, Australia Black Dog Institute, Prince of Wales Hospital, Sydney, NSW, Australia e-mail: [email protected] C.K. Loo School of Psychiatry, University of New South Wales, Sydney, NSW, Australia Black Dog Institute, Prince of Wales Hospital, Sydney, NSW, Australia Mental Health Unit, St George Hospital, Kogarah, NSW, Australia
H.K. Manji and C.A. Zarate (eds.), Behavioral Neurobiology of Bipolar Disorder and its Treatment, Current Topics in Behavioral Neurosciences 5, DOI 10.1007/7854_2010_66 # Springer‐Verlag Berlin Heidelberg 2011, published online 11 August 2010
1
2
P.B Mitchell et al.
Abstract Bipolar disorder (BPD) is capricious illness. For some, it is a condition of few episodes; for others, it is unremitting. For some, elevated moods predominate; for others, depression is the major key. For a minority, the condition is predictably cyclical; for most, it is unpredictably chaotic. This chapter examines those studies that have attempted to draw coherence from this enigmatic disorder. Where possible, we will focus on data derived from prospective longitudinal investigations, while using as necessary the more limited retrospective or cross-sectional reports. For the sake of parsimony, we will limit discussion to those studies that have used the conservative historical definitions of BPD (as used in DSM-IIIR, DSM-IV, and ICD-10), eschewing the recent controversial concepts of “pediatric” BPD and “soft BPD spectrum.” Keywords Bipolar disorder Cycling Longitudinal Outcome
1 Historical Legacies The current distinction between the two major psychoses [schizophrenia and bipolar disorder (BPD)] that underpins the DSM-IV and ICD-10 nosologies derives from the writings of Emil Kraepelin (1899). Drawing on the work of his nineteenth century German predecessor Kahlbaum (Angst and Gamma 2008; Healy et al. 2008; Hippius and Mu¨ller 2008), Kraepelin’s central differentiation of manicdepressive psychosis from dementia praecox was that the former was a disorder of “good prognosis.” He described a characteristic pattern of full remission between the manic or depressive episodes, as opposed to the deterioration or chronicity which he observed in dementia praecox. However, Kraepelin also described interepisodic “chronic mild weaknesses” in manic-depressive psychosis, i.e., low-grade manic and depressive symptoms, reflecting his belief (Angst and Gamma 2008) of a continuum of severity from health to illness.
2 The Course of BPD Prior to the Introduction of Modern Psychotropic Agents A major issue when considering the course of BPD is the impact of medications – not only in terms of improving outcome but also the potential for destabilizing the illness, for example, manic episodes induced by antidepressants or lithium withdrawal (Suppes et al. 1991). This chapter will not review evidence-based therapies for BPD, which are dealt with elsewhere in this book. Nonetheless, it is pertinent,
Course and Outcome of Bipolar Disorder
3
where possible, to examine the course of this condition in its “native” form, prior to the introduction of modern psychotropic agents in the 1950s. An intriguing overview of the natural course of BPD in the pre-drug era (in descriptions roughly dating from Esquirol in the nineteenth century to Astrup in the mid-twentieth century) was recently published by Alvarez Ariza et al. (2009). Alvarez Ariza concluded that there were a number of commonalities between these historical reports: a high probability of recurrence; onset usually with a depressive episode; manic episodes of shorter duration and more likely to fully remit than depressive episodes; and, overall, good recovery between episodes. Perhaps, the most comprehensive data on the natural course of BPD before the introduction of modern therapies were those collected in the Iowa 500 study (Winokur and Tsuang 1996). Using files from the Iowa Psychopathic Hospital, Winokur and colleagues obtained clinical details on BPD patients admitted between 1934 and 1944. As that hospital had routinely collected follow-up information from patients, relatives, or clinicians every 6–12 months, it was possible for Winokur’s group to determine the progress of this condition in the era prior to the advent of lithium, neuroleptics, or antidepressants. Informative data were obtained on 87 patients with BPD who had been followed up between 6 months and 20 years (with an average of 2.2 years). Over that period of time, 54% recovered fully at some stage, while 21% were continuously symptomatic over the follow-up period. This pattern was not distinguishable from unipolar depression, but contrasted markedly with schizophrenia, in which 78% showed continuous illness and a further 13% deteriorated. Most BPD patients took 2–3 years to recover from their index episode (of mania or depression); however, for those able to be followed up over 10–20 years, all fully recovered at some stage. About 5% of those initially diagnosed with depression went on to develop mania over the follow-up period, most within 3 years of the index admission.
3 Methodological Issues Interest in both the natural course of BPD and the response of this disorder to treatment has increased, leading to increasing recognition of the limitations of the current nomenclature regarding the course and outcome of this condition. Recently, a working group of the International Society for Bipolar Disorders published a report on this issue (Tohen et al. 2009). In that paper, consensus opinion was reached regarding the definition of nine key terms (response, remission, recovery, relapse, recurrence, subsyndromal states, predominant polarity, switch, and functional outcome) commonly used to describe course and outcome in BPD. Pertinent to this chapter, Tohen and colleagues distinguished between “relapse” (which they
4
P.B Mitchell et al.
defined as a new episode occurring within 8 weeks of an index episode) and “recurrence” (a new episode occurring after eight weeks) – definitions similar to those recommended for major depressive disorder (MDD) by Frank et al. (1991). Because most of the BPD literature had hitherto not distinguished between these time intervals, and tended to use these terms interchangeably, we will arbitrarily use the term “recurrence” here to cover both concepts.
4 Childhood and Adolescent Antecedents of BPD It has long been recognized (based largely on retrospective reports) that BPD often presents for the first time during adolescence. In the first prospective study of this, Geller et al. (2008) demonstrated a continuity of childhood/adolescent onset DSMIV BPD-I with adult illness. Elucidation of replicable antecedent clinical or biological abnormalities prior to the onset of syndromal BPD would assist clinicians in early diagnosis of this condition and enable development of early identification and intervention programs. In recent years, there has been a growing recognition of the importance of identifying such antecedent features, with a concurrent increased number of crosssectional and longitudinal “high risk” studies of young offspring of parents with BPD (recently reviewed by Duffy 2009). Two cross-sectional studies (Lapalme et al. 1997; Birmaher et al. 2009) reported increased rates of syndromal BPD and MDD in these children, as well as increased rates of non-specific symptoms such as anxiety and subthreshold mood symptoms. There was no increase in the rates of attention-deficit hyperactivity disorder (ADHD) in those studies, an issue of pertinence in view of the high rates of reported comorbidity between that disorder and “pediatric” BPD (Biederman 2006). Similarly, the prospective longitudinal studies (Akiskal et al. 1985; Hammen et al. 1990; Egeland et al. 2003; Hillegers et al. 2005; Shaw et al. 2005; Duffy et al. 2009) have not found increased rates of childhood or adolescent ADHD in these “high risk” samples, nor have they demonstrated increased rates of pre-pubertal mania (which would be expected if “pediatric” BPD were a true antecedent of adult BPD-I). These prospective trials found higher rates of anxiety disorders, major mood disorders (mainly depression), and subthreshold depressive and “activation” symptoms. Other prevalent symptoms included sleep disturbance, distractibility, and a greater sensitivity to stress. In a retrospective and uncontrolled investigation of the occurrence of any more “immediate” prodromal symptoms, Mantere et al. (2008) found that prodromal symptoms were present in 45% of those with BPD-I and 50% of those with BPD-II. The mean duration of such symptoms was 31 days prior to syndromal onset and did not differ between BPD-I and BPD-II disorders. Most symptoms were moodcongruent, though some were non-specific features, particularly anxiety.
Course and Outcome of Bipolar Disorder
5
5 The Onset of BPD: Age Most patients with BPD report that the illness first presents in the mid-to-late teenage years or twenties. The mean age of onset reported in three large datasets (STEP-BD, Stanley Foundation Bipolar Network, and the Sydney Black Dog Institute; reviewed by Mitchell et al. 2009) ranged from 17.4 to 22.9 years. Onset occurs earlier in those with a positive family history of BPD (Taylor and Abrams 1981). Furthermore, a number of reports indicate that age of onset is genetically determined. Bellivier et al. (2003) reported three genetically determined age of onset subtypes: early (mean 17 years), intermediate (25 years), and late (40 years). In a similar study, Hamshere et al. (2009) reported on a mixture analysis that found three distinct age of onset groups for BPD-I: 22 years, 25–37 years, and 40 years. Hamshere and colleagues described differential clinical characteristics in these groups. The early-onset group had a stronger family history of affective disorder, more frequent suicide attempts, a greater proportion of rapid-cycling patients, more episodes of mania, and more episodes of depression. This suggestion of a greater severity of illness in the early-onset group has been supported by Perlis et al. (2009) who found that those who had a first episode before the age of 13, had (over their lifetime) fewer days euthymic, more impaired functioning, and earlier recurrence. Coryell et al. (2009) also reported that an earlier age at onset portends greater depressive symptom burden. Lin et al. (2005) reported linkage of a number of loci with age of onset of the first manic (but not depressive) episode.
5.1 5.1.1
Specific Age-Related Presentations Postpartum BPD
For many women, BPD presents for the first time in the postpartum period. For those with pre-existing BPD, the postpartum period is a period of high risk for relapse (Freeman et al. 2002). Munk-Olsen et al. (2009) reported that 27% of mothers with prior BPD were admitted within 1-year postpartum, and that a previous diagnosis of BPD was the strongest predictor of a postpartum psychiatric admission.
5.1.2
Late-Onset BPD
Late presentation of BPD (onset in the 60s or later) is less likely to be associated with genetic factors, rather related to neurological conditions such as cerebrovascular disease or traumatic brain injury, or more subtle organic processes such as small vessel disease in the brain; for instance, Tamashiro et al. (2008) reported more white matter hyperintensities in those with late-onset presentations. While
6
P.B Mitchell et al.
late-onset manic presentations have usually been associated with greater morbidity and mortality (Tohen et al. 1994), a recent report (Oostervink et al. 2009) from the prospective, longitudinal European EMBLEM study found a more favorable outcome in the late-onset (>60 years) group compared to an early-onset elderly BPD sample.
6 The Onset of BPD: Polarity While most retrospective studies have reported a depressive onset to be more common than mania, this has not been consistently demonstrated. In the three large datasets reviewed by Mitchell et al. (2009), 52–53% of patients reported their first episode to be depressive, with 16–26% reporting the initial presentation as hypo/manic. A number of retrospective studies have explored clinical correlates of the polarity of the onset episode. Perlis et al. (2005) reported on 704 BPD-I patients from the STEP-BD study. Depressive onset was associated with an earlier onset of illness, more lifetime depressive episodes, and more time depressed. Examining the UK Wellcome Trust genetic sample, Forty et al. (2009) reported similar findings. Using linear regression, initial onset with depression (compared to mania) was associated with earlier age of onset, predominant lifetime depressive polarity, more frequent and more severe depressive episodes, and less prominent lifetime psychotic episodes. Another means of assessing the phenomenon of the illness onset being depressive is to prospectively evaluate a sample of MDD patients, investigating for rates and correlates of “conversion” to BPD. First, in terms of rates of such “conversion,” Angst et al. (2005) reported on a prospective evaluation of hospitalized depressed patients. They found a conversion rate of 1% per year to BPD-I and 0.5% per year to BPD-II disorder. Males and those with an early onset of depressive illness were at greater risk of converting to BPD-I, whereas BPD-II conversion was associated with being female, later onset of depression, and a positive family history of mania. Examining a community sample of 14 to 24-year olds followed up over 10 years, Beesdo et al. (2009) reported that 3.6% of those with initial unipolar depression subsequently developed hypo/mania, with risk particularly high in those with adolescent onset (
