Research on early-onset bipolar disorder sometimes lumps childhood-onset in with adolescent onset. Researcher Terence Ketter et al. explored differences in illness among 502 patients at the Stanford Bipolar Disorder Clinic. The 107 patients with childhood onsets (before age 13) had a more difficult course of illness in almost all domains compared to the 238 patients with adolescent onsets (age 13–18) or the 157 patients with illness onset in adulthood (after age 18).
Considered separately, both patients with childhood-onset illness and patients with adolescent-onset illness had more comorbid anxiety disorders, alcohol use disorders, eating disorders, prior suicide attempts, rapid cycling in the prior year, and at least five mood episodes over the course of their lifetimes than those patients whose illness began in adulthood. Patients whose illness began in childhood had higher rates of each of these unfavorable illness characteristics and were more likely to have a first-degree relative with a mood disorder.
These data mirror those from the Bipolar Collaborative Network in which this editor (Robert Post) is an investigator, and the larger STEP-BD network led by Perlis et al. All three suggest that in the US, two-thirds of the bipolar disorder seen in adults begins in childhood and adolescence, with about a quarter beginning before age 13.
Ketter suggests that research should not combine childhood and adolescent onset illnesses, which come with different rates of anxiety, alcohol, and eating disorder comorbidity, rapid cycling, and prior episodes. The statistical relevance of some findings can be diluted when the two groups are combined.
Editor’s Note: However, the primary message is that childhood onset bipolar disorder is a more severe version of the illness that deserves greater attention, treatment, and research so that its course can be made more benign. It is a problem that there are no Federal Drug Administration–approved treatments for children under 10 years of age with bipolar disorder.
At the 2014 meeting of the International Society for Bipolar Disorder, researcher Kiki Chang discussed Pediatric Acute Onset Neuropsychiatric Syndromes (PANS), a newly identified phenomenon in which children suddenly develop obsessive compulsive disorder (OCD) and/or a restrictive eating disorder following an infection or other process that stimulates an immune/inflammatory reaction in the brain. A similar phenomenon, Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections (PANDAS), was initially identified by Susan Swedo of the National Institute of Mental Health (NIMH) and refers to children (usually under 6–10 years old) who develop OCD and/or tics following a case of strep throat or scarlet fever.
PANS may have an autoimmune component. In addition to acute onset of OCD and eating restriction, other symptoms include mood episodes (depression, mania), high aggression/irritability, anxiety (particularly separation anxiety), cognitive problems (ADHD, handwriting regression), regressive behaviors, and somatic signs such as sleep difficulties and urinary urgency. Biological abnormalities may include: abnormalities in red blood cell sedimentation rate, elevated C-reactive protein (CRP), high Anti DNase B and/or Antistreptolysin O (ASO) titers (anti-Streptococcus antibodies), mycoplasma IgG or IgM antibodies (signs of some types of pneumonia), ANA (antinuclear antibodies, sign of an autoimmune disease), ferritin (a protein that stores iron), copper, and a panel of tests (the Cunningham Panel) by the company Moleculera Labs that measures antibodies for four neural antibodies (dopamine D1 receptors, dopamine D2 receptors, lysoganglioside (LysoGM-1), and tubulin) and calcium/calmodulin-dependent protein kinase activity (CaMKII).
PANS is three times more likely to affect males than females, and in the Stanford PANS Clinic sample of 50 youth, PANS was associated with strep infections (65%), mycoplasma bacteria (13%), viral or urinary tract infection (58%), and ear and other infections in 16%.
Symptoms included OCD (86%), anxiety (92%), mood disturbance (88%), and aggression (82%).
Treatments include steroids, the immunosuppressant mycophenolate, intravenous immunoglobulin (IVIG), plasma exchange, the tumor necrosis factor blocker infliximab, and sometimes the antibiotic amoxicillin.
Chang also described a case in which a 15-year-old developed minocycline-induced OCD and acute onset of severe mania that included urinary incontinence and was unresponsive to medication. The patient had elevated ANA, anti-thyroid antibodies, and reduced complement C4 proteins, along with elevated antibodies to dopamine D1 and D2 receptors, LysoGM-1, and tubulin.
While the reasons why one person develops bipolar disorder and another does not remain mysterious, the current thinking is that genes contribute some risk while immunological abnormalities contribute other risks. Researchers have identified certain antibodies whose levels spike during an episode of mania, as if the patient is having an immune reaction. These are referred to as biomarkers or inflammatory markers.
While various biomarkers for mania have been identified, until recently their effects had only been examined independently. A 2013 article by Dickerson et al. published in the journal PLOS ONE examined four biomarkers in combination. Each was a type of antibody: to the NR peptide of the NMDA receptor, to gliadin (a protein derived from gluten), to Toxoplasma gondii (a parasitic protozoan), and to Mason-Pfizer Monkey Virus. Measures of these four types of antibodies made up a combined inflammation score for participants in the study.
The study compared 57 patients presenting with a manic episode with 207 non-psychiatric controls and 330 patients who had had recent onset of psychosis, schizophrenia, or bipolar depression. The combined inflammation score of the mania group was significantly higher than the other groups at the time of hospital admission and at the time of evaluation several days later. It had returned to normal (i.e. not different from the other groups) at followup six months later, although those with the highest combined inflammation scores were at risk for re-hospitalization during that period.
The findings of this study suggest that hospitalization for mania is associated with immune activation, and the level of this activation predicts subsequent re-hospitalization. Treatments for mania that target this inflammatory response should be investigated.
Most children recover from an episode of bipolar disorder after a considerable period of time, but the majority eventually relapse. At the 2013 meeting of the American Academy of Child and Adolescent Psychiatry (AACAP), Boris Birmaher of the University of Pittsburgh presented new data on the long-term prospective course of bipolar disorder in 255 children with bipolar I, 30 children with bipolar II, and 153 children with bipolar NOS (not otherwise specified), who together had an average age of onset of 9.3 +/- 3.9 years. The children participated in the study for an average of 8 years. Most of the children (81.5%) recovered from their episode, but only after an average of 2.5 years of follow up treatment. Yet 62.5% of those who recovered experience a recurrence after an average of 1.5 years.
Editor’s Note: It takes a long, long time to achieve recovery, and longer for bipolar NOS (more than 2 years on average) than for either Bipolar I or II (about 1.8 years). However, the high rate of relapse within 1 to 2 years is equally disturbing. These data are similar to those in many other prospective follow up studies of children, and suggest that it is important for parents to be aware that this illness is difficult to treat, and good results within weeks are not likely to be the norm. At the same time, 43% of the children with a bipolar diagnosis eventually achieved euthymia (wellness) in the longer term, so there is cause for some optimism.
Four Trajectories in Children with Bipolar Illness
Birmaher described four different long-term,trajectories observed over an average of 8 years of follow up with 438 children with bipolar disorder.
- Predominately euthymic (24%)
- Ill early then much improved (19%)
- Mild to moderately ill—euthymic only 47% of the time (34.6%)
- Predominantly ill—euthymic 11.5% of the time (20.3%)
The predominantly well group (1) was associated in a univariate analysis with a later onset of illness, higher socio-economic status, less conflict, fewer stressors, less sexual abuse, fewer anxiety and ADHD comorbidities, and less medication (including stimulant use). In a multivariate analysis, this group was independently associated with less severe depression/mania, less suicidal ideation, less substance use, less sexual abuse, and less family history of mania and substance abuse.
This group had the best functioning, almost to 80 on the Children’s Global Assessment Scale (C-GAS). In comparison, despite considerable time euthymic for groups 2 and 3, these children still had considerable functional impairment, in the realm of 65 on the C-GAS scale. Even in Group 1, about half of the children had low C-GAS scores.
Birmaher suggested the importance of trying to find ways to delay the onset of the illness (to graduate more children into the good prognosis group) and allowing them time to develop socially and educationally and graduate from high school. Potential preventive strategies could include omega-3 fatty acids, more time spent exercising, good sleep hygiene, family focused therapy (FFT), dialectic behavior therapy, treating subsyndromal depression, and even treating parents with mood disorders to complete remission (which has been shown to improve behavioral health in offspring).
Editor’s Note: As this editor Post, Chang, and Frye wrote in the Journal of Clinical Psychiatry in 2013, beginning to study the effectiveness of these kinds of early primary and secondary prevention strategies in children who can now be readily identified clinically as at risk for a mood disorder, should be given the highest priority.
Children who have at least one parent with a bipolar or unipolar disorder, some further environmental risk factors (such as adversity in early childhood), and early symptoms of depression, anxiety, or prodromal bipolar disorder are at very high risk for bipolar disorder, and there is an urgent need for randomized studies (even open ones) of safe potential preventive strategies for these children.
Omega-3 fatty acids in particular have a strong record of safety, compelling rationale for use in bipolar disorder, and have already been shown to have significant preventive effects in decreasing the transition from early prodromal psychosis to full-blown schizophrenia.
Many children with bipolar disorder also present with other comorbid Axis I psychiatric illnesses. Now it seems that the worsening of these comorbidities, such as attention-deficit hyperactivity disorder (ADHD) or an anxiety disorder, can signal a more difficult course of bipolar illness itself. At a symposium on the course of bipolar disorder in children at the 2013 meeting of the American Academy of Child and Adolescent Psychiatry (AACAP), Shirley Yen from Brown University discussed findings on comorbidities of childhood onset bipolar disorder from COBY, the Collaborative Child Bipolar Network. Upon study entry, 60% of children with bipolar disorder also had ADHD, 40% had oppositional defiant disorder (ODD), 39% had an anxiety disorder, 12.5% had both oppositional defiant disorder and a conduct disorder, and 9% had a substance abuse disorder.
The prevalence of most of these comorbid illnesses increased over time (e.g. anxiety disorder rates increased from 39% to 62%). The illnesses were also related to the time it took participants to achieve recovery (eight consecutive weeks well), and the time until a recurrence of a depressive or manic episode.
Increases in anxiety were linked to longer time to achieve recovery and a shorter time to a recurrence. Increases in ADHD were linked to a more rapid onset of a depressive recurrence. Increases in oppositional defiant disorder and conduct disorder had no relationship with either remission or recurrence. Increases in substance abuse disorders were linked to a longer time to recover from a manic episode. Thus, worsening of the comorbid conditions had definite consequences for both recovery and recurrence.
At a symposium on ketamine for the treatment of depression in children at the 2013 meeting of the American Academy of Child and Adolescent Psychiatry, David Brent, a professor at the University of Pittsburg, gave the opening talk on the fact that as many as 20% of adolescents who are depressed fail to improve, develop chronic illness, and are thus in need of alternatives to traditional treatment. Predictors of non-improvement include substance use, low-level manic symptoms, poor adherence to a medication regimen, low blood levels of antidepressants, family conflict, high levels of inflammation in the body, and importantly, maternal depression. In adolescents insomnia was associated with poor response, but in younger children insomnia was associated with a better response.
Brent suggested using melatonin and sleep-focused cognitive behavioral therapy for insomnia in youth, but not using trazodone (which is commonly prescribed). Trazodone is converted to a compound called Meta-chlorophenylpiperazine or MCPP, which induces anxiety and dysphoria. MCPP is metabolized by hepatic enzymes 2D6, and fluoxetine and paroxetine inhibit 2D6, so if trazodone is combined with these antidepressants, the patient may get too much MCPP.
Surprisingly and contrary to some data in adults about the positive effects of therapy in those with abuse histories, in the study TORDIA (Treatment of SSRI-Resistant Depression in Adolescents), if youth with depression had experienced abuse in childhood, they did less well on the combination of cognitive behavioral therapy and selective serotonin reuptake inhibitors (SSRIs) compared to SSRIs alone.
Evidence is growing that stressful events in childhood are associated with an earlier onset of bipolar disorders and a more difficult course of illness than in those who did not experience this type of adversity. Monica Aas and colleagues in Norway have found for the first time that emotional abuse in childhood, especially before age five, also increases risk of bipolar disorder. This study indicates that while bipolar disorder has a genetic component, environmental factors also play a role.
In Norway and France, the research group surveyed patients with bipolar disorder and people in the general population about childhood trauma, including emotional abuse, sexual abuse, physical abuse, emotional neglect, and physical neglect. Among the almost 800 participants, patients with bipolar disorder were twice as likely as control participants to have experienced multiple types of trauma. However, emotional abuse was the only factor specifically linked to bipolar disorder. People who were emotionally abused in childhood were more than twice as likely to develop bipolar disorder in adulthood. Moreover, the more severe the emotional abuse, the more likely it was that a child would go on to develop bipolar disorder.
Among the adults with bipolar disorder, emotional abuse and sexual abuse in childhood predicted younger age of illness onset, more suicide attempts, more rapid cycling, and greater proneness to depression. Emotional or sexual abuse were linked to the most suicide attempts, and sexual abuse was linked to rapid cycling.
More trauma in childhood was also linked to affective instability in adults. Aas’ research was presented at the 14th International Congress on Schizophrenia Research.
At a symposium celebrating the retirement of Willem Nolen, a researcher who spent 40 years studying unipolar and bipolar disorder, from his position at Groningen Hospital in the Netherlands, his colleague Jules Angst discussed some recent findings. Angst is perhaps the world’s leading authority on the long-term course of unipolar and bipolar disorders based on his multiple prospective follow-up studies, some lasting 20-30 years.
The Sensitization-Kindling Model
Angst described evidence that supports the sensitization-kindling model of recurrent mood disorders, which this editor (Robert Post) described in 1992. Episodes tend to recur faster over time, i.e. the well interval between episodes becomes progressively shorter. While stressors often precipitate initial episodes, after multiple occurrences, episodes also begin to occur spontaneously (in the absence of apparent stressors).
This type of progressive increase in response to repetition of the same stimulus was most clearly seen in animal studies, where repeated daily electrical stimulation of the amygdala eventually produced major motor seizures (i.e. amygdala kindling). Daily electrical stimulation of rodents’ amygdala for one second initially produced no behavioral change, but eventually, minor and then full-blown seizures emerged. Once enough of the stimulated full-blown amygdala-kindled seizures had occurred, seizures began to occur spontaneously (i.e. in the absence of the triggering stimulation).
The analogy to human mood disorders is indirect, but kindling provides a model not only for how repeated triggers eventually result in full-blown depressive episodes, but also for how these triggered depressive episodes may eventually occur spontaneously as well.
Long-Term Treatment of Mood Disorders
Angst also discussed long-term treatment of mood disorders. He has found that long-term lithium treatment not only reduces suicides in patients with bipolar disorder, but also reduces the medical mortality that accompanies bipolar disorder.
Angst noted his previous surprising observations that in unipolar disorder, long-term maintenance treatment, even with low doses of tricyclic antidepressants, prevents suicide. Previously, researchers Ellen Frank and David Kupfer of Western Psychiatric Institute and Clinic at the University of Pittsburgh Medical Center had found that when patients with recurrent unipolar depression who had been stable for 5 years on the tricyclic antidepressant nortriptyline were blindly switched to half their original dose, about 90% rapidly relapsed into a new episode of depression. Their data helped establish the prevailing view that maintenance treatment with the full-dose regimen required to achieve a good initial acute response is also the optimal approach to long-term continuation and prophylactic treatment.
Angst found good results even at low doses, but his data may not be in conflict with Frank and Kupfer’s, as a person who responds well acutely to low doses may also be able to maintain good enough response to them to prevent recurrences in the long term.
Incidence of Bipolar Disorder in Adolescents Similar to Incidence in Adults
Angst also presented data from the Adolescent Supplement to the National Comorbidity Study (NCS-A), which analyzed interviews with approximately 10,000 adolescents (aged 13-17) in the US. He found a 7.6% incidence of major depression, a 2.5% incidence of bipolar I or II disorder, and a 1.7% incidence of mania. There was an even higher incidence of sub-threshold bipolar disorder, when there are not enough symptoms or a long enough duration of symptomatology to meet diagnostic criteria for bipolar I or II disorder. These data published by Merikangas et al. in 2009 provide clear epidemiological data that there is a substantial incidence of bipolar disorder in adolescents in the US, roughly similar to that seen in adults.
At a symposium celebrating the retirement of Willem Nolen, a researcher who spent 40 years studying unipolar and bipolar disorder, from his position at Groningen Hospital in the Netherlands, this editor (Robert Post) discussed progress in the treatment of bipolar disorder over the past 40 years. Despite the availability of lithium; many new mood stabilizers (carbamazepine, valproate, lamotrigine); and many atypical antipsychotics, all of which are anti-manic and some of which are antidepressant (quetiapine and lurasidone), there is still a very high rate of continued illness and treatment resistance, especially in the US.
In fact, research from the Bipolar Collaborative Network, a treatment research network including sites around the US (one run by this editor) and in Germany and the Netherlands, shows that almost everything about bipolar disorder is worse in the US. Americans have more genetic vulnerability because more of their parents have bipolar disorder, and they are more likely to have environmental vulnerability as a result of childhood adversity. Patients in the US also reported having had more stressors at the onset of their illness and more stressors prior to the last episode they had before entering the network at an average age of 40.
Age at illness onset is much lower in the US than in the Netherlands and Germany. About two-thirds of American patients had onset in childhood or adolescence (under 19 years), while only about one-third of the European patients in this study showed these early onsets.
The course of illness is also more difficult in the US. There is more anxiety, substance abuse, and medical comorbidity, and there are more episodes and more rapid cycling. All this resulted in more US patients than European ones who did not respond to naturalistic treatment in our treatment network despite being prescribed multiple medications.
The implication of these data is that we need a new and more concerted approach to bipolar disorder in the US, beginning with early diagnosis and treatment during childhood and adolescence, instead of the 10- to 15-year average delay that was typical about twenty years ago. The duration of the delay to first treatment with a drug to treat mania or depression was an independent predictor of a worse outcome in adulthood. Early intervention should also include therapy and education.
Family-Focused Treatment (FFT), a method pioneered by researchers David Miklowitz and Kiki Chang, has been shown to be much more effective than treatment as usual in children who are at high risk for developing bipolar disorder because they have a family history of the illness and symptoms of an anxiety or depressive disorder or bipolar not otherwise specified (BP-NOS). In this way it may even be possible to head off the full-blown illness before it starts in those children at highest risk.
In research published since 2008, our Editor-in-Chief Robert M. Post and colleagues in the Bipolar Collaborative Network have compared patients with bipolar disorder in the United States to those in Germany and the Netherlands. Compared to the European sample, patients in the US have more genetic vulnerability to bipolar disorder (by having a parent with bipolar disorder), earlier onsets of their illness, more complicated courses of illness, greater treatment resistance, and more medical comorbidities. Patients in the US also have more psychosocial stress.
The researchers are now turning their attention to these psychosocial vulnerabilities, and in a new paper that will be published in Psychiatry Research (late in 2013 or early in 2014), the authors show that patients in the US had more stressors both in childhood and just prior to the onset of their illness. Childhood stressors analyzed in the study were verbal abuse, physical abuse, and sexual abuse. Stressors in adulthood included indicators of a lack of social support, troubles with finances or employment, lack of access to health care, and medical comorbidities.
The stressors patients experienced just prior to their most recent episode of bipolar illness were related to: stressors in childhood, an earlier age of illness onset, anxiety and substance abuse comorbidity, lower income, both parents having an affective illness such as depression, and feeling more stigma.
The new research suggests that for patients with bipolar disorder in the US, adverse life events in childhood and later in life are more prevalent than they are for patients in the Netherlands or Germany. Earlier and more effective approaches to these stressors, such as the Family-Focused Therapy developed by David Miklowitz and Kiki Chang, could potentially slow the onset or progression of bipolar illness in this country.