Pediatric schizophrenia describes schizophrenia experienced by patients with an initial presentation before adulthood and is currently diagnosed with
DSM-5 diagnostic criteria similar to those of adult schizophrenia. The description of pediatric schizophrenia has evolved over the last 50 years. In
DSM-II, what might now be identified as autism was referred to as “schizophrenia, childhood type.” It was not until
DSM-III that childhood schizophrenia became understood as a diagnosis separate from autism or pervasive developmental disorder. This separation, which occurred in 1980, was an important advance in the study of childhood psychopathology. We saw a further elaboration of the criteria for autism or pervasive developmental disorder in subsequent editions of the
DSM (i.e.,
DSM-III-R,
DSM-IV,
DSM-IV-TR, and
DSM-5), which helped to further separate autism or pervasive developmental disorder from pediatric schizophrenia (
1).
Pediatric schizophrenia is characterized as either childhood onset or early onset, depending on the age at which symptoms develop. The epidemiology, natural history, biopsychosocial underpinnings, assessment, differential diagnosis, and treatment of childhood-onset schizophrenia and early-onset schizophrenia are discussed in detail here.
Epidemiology and Natural History
Schizophrenia, a neurodevelopmental disorder manifesting usually in late adolescence or early adulthood, can cause significant personal, social, and occupational dysfunction. “Early-onset schizophrenia” refers to schizophrenia with symptom onset before the age of 18 years, whereas the rarer entity “childhood-onset schizophrenia” is defined as schizophrenia with symptom onset prior to age 13 years. Even though reported prevalence estimates of psychotic symptoms in children and adolescents reach 5% (
2,
3), the prevalence of lifetime schizophrenia is 1%, with a male-to-female ratio of about 1.4 to 1 (
4). Childhood-onset schizophrenia is far more rare. Studies on the incidence of childhood-onset schizophrenia are lacking. One estimate by the National Institute of Mental Health (NIMH) is that the incidence is less than 0.04% (
5). The NIMH childhood-onset schizophrenia study began in 1990 and has sought referrals of children experiencing psychotic symptoms before the age of 12 years. Other inclusion criteria included the absence of neurologic disorder and a premorbid IQ of 70 or higher. In their review of about 3,000 charts, the researchers found that 90% of the referrals did not meet criteria for childhood-onset schizophrenia. Three hundred children were screened in person, of whom 60% received other psychiatric diagnoses such as anxiety, affective disorders, and behavioral disorders. About 200 patients who provisionally met criteria for childhood-onset schizophrenia were admitted for observation and underwent a medication washout period followed by one to three weeks of observation off medications (
5). Of these, 20% of children no longer met criteria for childhood-onset schizophrenia, and this was confirmed by a four- to six-year follow-up study of these children who continued to not meet criteria for schizophrenia.
It is important to note that not every child who has psychotic symptoms has schizophrenia. Psychotic symptoms may be common among children with mood disorders or a history of trauma, as well as among those using illicit substances. It is also imperative to rule out medical causes of psychotic symptoms. Unlike adults, children have developmentally normal behaviors that can be misconstrued as psychotic, such as having an imaginary friend. Childhood-onset schizophrenia, like schizophrenia among adults, involves a combination of both positive and negative symptoms. Positive symptoms, such as hallucinations and delusions, are usually easy to identify. Negative symptoms may be harder to assess. Some children with schizophrenia present with cognitive deficits in the areas of memory and executive functioning. It is common to see children with schizophrenia experience academic difficulties, isolation, social withdrawal, disruptive behaviors, and problems with speech and language (
4). The onset of childhood-onset schizophrenia is often insidious, and children frequently have premorbid social and occupational dysfunction prior to developing overt psychotic symptoms.
Childhood-onset schizophrenia frequently presents with a prodrome phase prior to the onset of florid psychotic symptoms. This prodrome phase is characterized by functional decline that affects multiple domains of daily living, including academics, social functioning, and self-care. Associated comorbid psychiatric disorders may further complicate the differentiation of prodromal onset. The prodromal phase is followed by the acute phase, which is usually easier to identify because of the presentation of positive psychotic symptoms and continued decline in functioning. In the recovery phase, which usually occurs in adulthood, patients continue to be impaired, may present primarily with negative symptoms, and may continue positive symptoms. The residual phase, however, is characterized by a complete resolution of positive symptoms for an extended period. Insidious onset of symptoms, onset at a young age, lower intellectual functioning, lower premorbid functioning, and higher rates of negative symptoms are poor prognostic indicators (
6–
10). Individuals experiencing these symptoms, when followed into adulthood and compared with their peers, have been found to have higher unemployment rates and greater social deficits, and they are less likely to live independently (
10–
12). Monitoring for safety is equally important, with studies indicating that about 5% of patients with childhood-onset schizophrenia lose their life to suicide or as a direct result of dangerous behaviors when psychotic (
13).
Biopsychosocial Underpinnings
Schizophrenia is a neurodevelopmental disorder with a multifactorial etiology. Research has shown the roles played by genetics as well as the environment in the development of childhood-onset schizophrenia and early-onset schizophrenia. Developing the disorder at an early age is a poor prognostic factor, and the phenotype may be more severe. If a first-degree relative has the disorder, the risk of developing the disorder is five to 20 times higher than the risk of the general population (
4). Concordance rates among monozygotic twins is 40%−60% and that of dizygotic twins and other siblings is 5%−15% (
4,
14).
Published studies have implicated different genomic loci and genes, including the major histocompatibility complex 6p21.1, MIR137, and ZNF804a (
15,
16). Mutations in structure at genomic “hotspots,” including chromosomes 1q21.1, 3q29, 15q11.2, 15q13.3, 16p11.2, 16p12.1, 16p13.11, 17p12, and 22q11.2, are seen in 0.5% to 1.0% of cases (
4,
17). Patients with childhood-onset schizophrenia also have higher rates of cytogenetic abnormalities, such as 22q11.2 deletion syndrome (previously known as velocardiofacial syndrome or DiGeorge syndrome). Schizophrenia is believed to be caused by multiple mutations in different genes and different genomic loci (
17). Furthermore, the same mutation or different mutations of the same gene may present in phenotypically different ways among different individuals (
17). In addition to the biologic model, the interaction of gene and environment adds another layer of complexity to the etiopathogenesis of the disorder.
Decreased brain volumes; loss of gray matter in frontal and temporal lobes; and loss of gray matter in the anterior cingulate, thalamus, hippocampus, amygdalae, and insula have been reported in studies of adults with schizophrenia (
18), although this has not been similarly verified in child and adolescent studies. Adolescence has been established as a critical period for neuronal pruning; hence, with earlier onset of symptoms, there may be disruption in the normal process of neuronal development that causes impairments with memory, abstract thinking, and emotion regulation. Loss of gray matter, abnormal cortical thinning, and reduced synaptic structures on cortical pyramidal neurons are some findings seen in the brains of those with schizophrenia. Researchers conducting a recent study hypothesized the possibility that neuron–microglia interactions via complement cascade may be the reason for the development of schizophrenia because of the stimulation of microglia and the elimination of synapses (
19).
Several environmental factors to the development of schizophrenia have been hypothesized, including marijuana exposure, in utero exposure to famine, prenatal infections, advanced paternal age, and obstetric complications. No literature supporting psychological or social factors causing schizophrenia exists at this time, but these factors may influence the onset, exacerbation, and relapse of the disorder. Other factors that may contribute include being in a racial-ethnic minority group, lacking social support, living in an urban setting, and cultural migration (
20).
Assessment and Differential Diagnoses
Many more children and adolescents present with psychotic symptoms than are eventually diagnosed as having severe and chronic childhood-onset schizophrenia. In fact, a 2011 study indicated that up to 44% of patients who initially were admitted for inpatient study with a diagnosis of schizophrenia no longer met criteria after a prolonged admission or the medication washout period (
21). It is thus vital to first thoroughly consider all other signs and symptoms and rule out differential diagnoses prior to assigning the diagnosis of childhood-onset schizophrenia. Psychotic symptoms are present as part of many nonpsychiatric medical illnesses, substance abuse, and other psychiatric disorders. Therefore, these must be considered before the determination of childhood-onset schizophrenia can be made.
Hallucinations may occur within the context of many neurologic conditions. In particular, migraines, especially migraine-preceding auras, can be confused with psychotic symptoms. These auras have a range of presentations, from visualization of bright lights, micropsia, or macropsia to fully formed visual or auditory hallucinations. Delusions have even been reported with cases describing Capgras syndrome (a delusional disorder where patients believe a family member or friend has been replaced by an imposter) or Frégoli syndrome (a delusional disorder where patients believe a person disguises himself or herself as many different people) before and after migraines (
22). Similarly, psychotic symptoms may present before, during, or after a seizure or even interictally (between seizure episodes), and there is a well-known correlation between temporal lobe epilepsy and psychosis (
23). Psychotic symptoms secondary to epilepsy frequently involve a single sensory modality correlated with the focus of the seizure activity (
24). In any complete list of neurological differential diagnoses, it is vital to include malignancies of the brain, particularly those involving the pituitary gland or the temporal, frontal, or posterior lobe, in addition to possible secondary infections of the brain, such as encephalitis (
24,
25).
Psychotic symptoms also occur within the context of direct head trauma with concern for coup injuries, contrecoup injuries, hemorrhages, and hematomas. It is important to recall that many metabolic changes, including hypernatremia, hyponatremia, hypocalcemia, hypomagnesemia, hypoglycemia, and hyperthyroidism, may present with psychotic symptoms. Also, autoimmune disorders such as systemic lupus erythematosus can lead to presentation with psychotic symptoms (
24). In addition, many genetic conditions are associated with psychosis. The most well-established association is the link between psychosis and 22q11.2 deletion syndrome, with approximately 14%−28% of children with the syndrome presenting with psychotic symptoms (
26). Other genetic syndromes to consider include Down and Marfan syndromes, Huntington’s disease, and acute intermittent porphyria (
24).
In addition to acute and chronic medical (nonpsychiatric) disorders, the use of certain medications and illicit substances should be excluded from the differential, as patients using them may also present with psychiatric symptoms. The medications most frequently associated with psychotic symptoms are stimulants, steroids, anticholinergics, antihistamines, antimicrobials, serotoninergic medications, analgesics (including nonsteroidal anti-inflammatory drugs), and anesthetics (
24). Psychotic symptoms can also be induced by hallucinogen intoxication with drugs such as Ecstasy, PCP, ketamine, amphetamines, opioids, synthetic opioids (“bath salts”), cannabis, and synthetic cannabis (“spice”), as well as by alcohol or sedative withdrawal (
24).
An effective assessment of psychotic symptoms of a child or an adolescent must take all of the above into account prior to a final determination of childhood-onset schizophrenia. An accurate history and physical, including detailed neurological examination, are paramount. It is important to determine whether there is history of any recent head trauma, medication use, substance abuse, other psychiatric symptoms, or associated signs or symptoms such as headaches or migraines, postictal confusion and fatigue, focal deficits, or genetic disorder stigmata. When evaluating a child or adolescent with psychotic symptoms, consider the utility of rating scales, such as the Positive and Negative Syndrome Scale, and appropriate laboratory and radiology screenings. An assessment of the complete blood count (to rule out any infection), complete metabolic panel (considering hypernatremia, hyponatremia, hypocalcemia, hypomagnesemia, and hypoglycemia and hepatic function for other metabolic disorders), thyroid function tests, and urine toxicology should be considered. Other laboratory considerations include ceruloplasmin if there is suspicion of Wilson’s disease, a heavy metal screen, or lead levels if there is the possibility of lead ingestion. An antinuclear antibody screen should be considered if lupus or other autoimmune disorders are possible. Magnetic resonance imaging should be ordered if focal deficits are observed to rule out space-occupying brain lesions or a history of head trauma. An electroencephalogram should be done if there are any indicators of seizure disorders. Family history should be considered if there are any congenital abnormalities or dysmorphologies requiring a genetics consult, karyotyping, or microarray (
24).
Treatment and Management
Although there is no cure for childhood-onset schizophrenia or even a single medication that is most efficacious, there is a consensus that early treatment is vital to improve positive and negative symptoms, as well as to minimize social, cognitive, and motor functioning deficits (
27). Multiple randomized control trials have shown that both first- and second-generation antipsychotics are superior to placebo in the treatment of psychotic symptoms (
27,
28). First-generation antipsychotics have many known adverse effects and significantly higher rates of extrapyramidal symptoms (EPS), particularly akathisia and dyskinesias; thus, second-generation antipsychotics are widely considered first-line treatment for childhood-onset schizophrenia and early-onset schizophrenia.
Further study has supported the conclusion that second-generation antipsychotics are significantly more effective than placebo in treating childhood-onset schizophrenia and early-onset schizophrenia, but the studies have also shown that second-generation antipsychotics have multiple adverse effects of their own. Children and adolescents are more likely than adults to experience sedation (although this is frequently time limited), higher levels of EPS, elevated prolactin, weight gain, and metabolic changes when taking second-generation antipsychotics (
27). Children and adolescents are less likely than adults to develop tardive dyskinesia and diabetes mellitus. Higher levels of EPS were observed with ziprasidone, followed by olanzapine and aripiprazole; the lowest levels were seen with risperidone (
27). The adverse effect of increased prolactin levels was highest with risperidone. Prolactin levels decreased with aripiprazole. Weight gain was most prominent with olanzapine, followed by clozapine, risperidone, and quetiapine (
27). Weight gain was found to occur least often with aripiprazole, although ziprasidone was not used in the comparison study and is frequently associated with lower levels of weight gain among adults (
27). Finally, clozapine has multiple specific adverse effects, including neutropenia, as well as a prohibitive monitoring regimen of laboratory tests.
Multiple large studies analyzing the treatment of adult schizophrenia, such as the Clinical Antipsychotic Trials of Interventional Effectiveness (CATIE) and the Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia (CutLASS study), indicate there is no significant improvement in efficacy of second-generation antipsychotics from first-generation antipsychotics. Although this holds true for adults, the Treatment of Early Onset Schizophrenia Spectrum Disorders Study (TEOSS) compared the effects of both drug types on 8- to 19-year-olds. The TEOSS compared risperidone (0.5–6 mg/day), olanzapine (2.5–20mg/day), and molindone (10–140 mg/day) and found that there was no significant difference in improvement of psychotic symptoms between the three medications—that is, across first- and second-generation antipsychotics. The study did, however, find significantly more adverse effects with olanzapine related to metabolic changes such as elevated glucose, cholesterol, and low-density lipoproteins (
29). The patients receiving olanzapine had such dramatic metabolic changes that the investigators discontinued that treatment arm prior to the completion of the study (
28).
Although a definitive algorithm for treatment of childhood-onset schizophrenia or early-onset schizophrenia does not exist, multiple algorithms have been proposed. One possible algorithm proposed by Kranzler and Cohen (
27) was to begin with a single second-generation antipsychotic; if it proves ineffective, then move to a different second-generation antipsychotic with a slow cross-taper. If psychotic symptoms persist, the algorithm proposes that the patient should be moved to clozapine for treatment-resistant schizophrenia. At present, no data support augmentation with any medications if there is a partial response to a second-generation antipsychotic. Another, more specific algorithm, proposed by Carlisle and McClellan (
28), is a six-week trial of risperidone, with a cross-taper to aripiprazole if any adverse effects are experienced. Alternately, another first- or second-generation antipsychotic (excluding clozapine and olanzapine) may be used for another six-week trial if risperidone does not control the symptoms. If both trials prove ineffective, cross-tapering to yet another antipsychotic for six weeks is indicated, with a final clozapine trial if treatment resistance continues (
28).
The recent Recover After an Initial Schizophrenia Episode (RAISE) study evaluated the benefit of a comprehensive, multidisciplinary, team-based treatment approach called NAVIGATE as compared with community care for first-episode psychosis across 34 clinics nationwide for patients ages 15–40 years old. The treatment regimen of NAVIGATE includes four main interventions: (a) medication management, (b) family psychoeducation to help the patient and family cope and support the patient’s treatment and recovery, (c) resilience-focused individual therapy to maximize strengths and increase illness management skills, and (d) supported employment and education addressing illness-related challenges to both work and school. The study found that participants receiving the NAVIGATE intervention compared with those receiving community care had significantly improved psychopathological symptoms as well as quality of life (
30).
This study highlights the importance of not just psychopharmacologic treatments but also psychotherapeutic treatments for both schizophrenic symptoms and the minimization of the adverse effects of medications. With the diagnosis of any chronic, severe psychiatric illness, psychoeducation for both the child and the family regarding the diagnosis, disease course, treatment, and prognosis is paramount, as is cognitive and social skills training. Finally, it is vital to promote a generally healthy lifestyle that includes regular exercise, a healthy diet, increased water intake, improved sleep hygiene, and the cessation or avoidance of substance abuse.
Conclusions
Pediatric schizophrenia consists of early-onset schizophrenia (onset prior to age 18 years) and childhood-onset schizophrenia (onset prior to age 13 years). The description of pediatric schizophrenia has evolved over the last 50 years, resulting in its separation from autism spectrum disorder. The lifetime prevalence of schizophrenia is 1%, and the incidence of childhood schizophrenia is rare. Therefore, it is important to evaluate the cause of psychosis of a child or an adolescent, because psychosis can occur in mood disorders, in the context of trauma, with medical conditions, or with substance abuse. Completing an accurate history and physical during the patient assessment, including a detailed neurological examination, are paramount. The antipsychotics continue to be significantly more effective than placebo in treating childhood-onset schizophrenia and early-onset schizophrenia. It is important to note the medication-specific differences in potential side effects for this population. Children and adolescents are more likely than adults to experience sedation, EPS, elevated prolactin, weight gain, and metabolic changes. Recently, the RAISE study (which included adolescents ages 15–17 years old) found significant improvement in symptoms and quality of life when medication and psychotherapeutic treatments were combined and found to be effective for adolescents with early-onset schizophrenia. Given the rarity of this disorder, further study is still needed to better understand causality, improve assessment, and develop a definitive treatment algorithm.