Distribution of individuals with a 22q11.2 deletion (n = 244) (A) or duplication (n = 58) (B) as a function of age at a positive genetic test. Persons in both cohorts were primarily identified in the first 15 years of life, and those with the deletion were identified earlier than those with the duplication. The reason for referral to genetic testing is unknown, but the difference in age at identification might be due to congenital malformations, including facial features and heart defects, for individuals with the 22q11.2 deletion. Those who received a positive genetic test relatively late in life could reflect parents whose deletion or duplication was recognized secondary to a positive 22q11.2 genetic test of their child.
eTable 1. Diagnostic Classification of Psychiatric Disorders According to the ICD-10 and Corresponding ICD-8 Diagnoses
eTable 2. Incidence Rate Ratios of Psychiatric Outcomes Associated With 22q11.2 Deletion and Duplication Compared With Individuals Without the Deletion and Duplication, Respectively, Under Basic Adjustment
eTable 3. Incidence Rate Ratios of Psychiatric Outcomes for the First Individual Identified With a 22q11.2 Deletion or Duplication in Each Family
eTable 4. Incidence Rate Ratios of Psychiatric Outcomes Associated With 22q11.2 Deletion and Duplication Compared With Individuals Without the Deletion and Duplication, Respectively, When Considering 22q11.2 as a Non–Time-Dependent Variable
eFigure. Distribution of Birth Year Among Individuals With a 22q11.2 Deletion or Duplication in the Danish Cytogenetic Central Register
Customize your JAMA Network experience by selecting one or more topics from the list below.
Hoeffding LK, Trabjerg BB, Olsen L, et al. Risk of Psychiatric Disorders Among Individuals With the 22q11.2 Deletion or Duplication: A Danish Nationwide, Register-Based Study . JAMA Psychiatry. 2017;74(3):282–290. doi:10.1001/jamapsychiatry.2016.3939
What are the incidence rate ratios and absolute risks for psychiatric disorders in clinically identified individuals with 22q11.2 deletion or duplication?
In this Danish population-based study of more than 3.7 million individuals, risk estimates for psychiatric disorders were similar for clinically identified individuals with the 22q11.2 deletion as well as 22q11.2 duplication. The risk of intellectual disability was particularly high.
The study provides estimates of disease risk that are important in genetic counseling and clinical monitoring and intervention and points to individuals with the 22q11.2 duplication as being in need of the same careful and persistent clinical attention given to carriers of the 22q11.2 deletion.
Microdeletions and duplications have been described at the 22q11.2 locus. However, little is known about the clinical and epidemiologic consequences at the population level.
To identify indicators of deletions or duplications at the 22q11.2 locus and estimate the incidence rate ratios (IRRs) and absolute risk for psychiatric disorders in clinically identified individuals with 22q11.2 deletion or duplication.
Design, Setting, and Participants
A Danish nationwide register study including all individuals recorded in the Danish Cytogenetic Central Register with a 22q11.2 deletion or duplication was performed. A total of 3 768 943 individuals born in Denmark from 1955 to 2012 were followed up during the study period (total follow-up, 57.1 million person-years). Indicators of 22q11.2 deletion or duplication and cumulative incidences were estimated using a nested case-control design that included individuals from the population-based cohort. Survival analysis was used to compare risk of disease in individuals with and without the 22q11.2 deletion or duplication. The study was conducted from May 7, 2015, to August 14, 2016.
The 22q11.2 deletion or duplication.
Main Outcomes and Measures
Indicators for carrying a 22q11.2 deletion or duplication, IRR, and cumulative incidences for psychiatric diagnoses (International Statistical Classification of Diseases and Related Health Problems, 10th Revision, codes F00-F99), including schizophrenia-spectrum disorders, mood disorders, neurotic stress-related and somatoform disorders, and a range of developmental and childhood disorders.
Among the 3 768 943 participants, 244 (124 [50.8%] male) and 58 (29 [50.0%] male) individuals were clinically identified with a 22q11.2 deletion or duplication, respectively. Mean (SD) age at diagnosis of any psychiatric disorder was 12.5 (8.3) years for individuals with deletions and 6.1 (0.9) years for duplication carriers. A parental diagnosis of schizophrenia—but not of other psychiatric diagnoses—was associated with a 22q11.2 deletion, and parental psychiatric diagnoses other than schizophrenia were associated with duplication carrier status. Both the 22q11.2 deletion (IRR, 4.24; 95% CI, 3.07-5.67) and duplication (IRR, 4.99; 95% CI, 1.79-10.72) was associated with increased risk of any psychiatric disorders. Furthermore, a highly increased risk of intellectual disability was found for the deletion (IRR, 34.08; 95% CI, 22.39-49.27) and duplication (IRR, 33.86; 95% CI, 8.42-87.87). Furthermore, individuals with the 22q11.2 deletion had an increased risk of several psychiatric disorders under study, for example, pervasive developmental disorders (IRR, 9.45; 95% CI, 5.64-14.69) and childhood autism (IRR, 8.94; 95% CI, 3.21-19.23).
Conclusions and Relevance
Individuals with the 22q11.2 deletion or duplication have a significantly increased risk of developing psychiatric disorders. Survival analysis of persons carrying either the 22q11.2 deletion or duplication provides estimates of direct clinical relevance useful to assist clinical ascertainment, genetic counseling, guidance of symptomatic monitoring, and early clinical intervention.
The 22q11.2 locus has been implicated in various disorders; the best characterized is the microdeletion causing 22q11.2 deletion syndrome (22q11.2DS) (also known as DiGeorge or velocardiofacial syndrome). The 22q11.2DS is the most common deletion syndrome in humans, with an estimated frequency of 1:2000 to 1:4000 live births.1,2 Most individuals with 22q11.2DS carry an approximately 3 megabase (3 Mb) deletion, whereas the remaining patients harbor smaller deletions nested within the 22q11.2 chromosomal region. These deletions typically arise de novo by nonallelic homologous recombination events between low copy repeats located in the 22q11.2 locus.3-6
The clinical presentation of 22q11.2DS has been studied intensively and shown to be highly variable. More than 180 clinical features have been described, ranging from nearly healthy individuals to those with fatal conditions, such as immune deficiencies and congenital heart defects.7-9 Psychiatric disorders have frequently been reported for the 22q11.2DS, and multiple cross-sectional studies have indicated a high prevalence of schizophrenia.10-12 In addition, reports have highlighted the high frequency of both autism spectrum disorder, attention-deficit/hyperactivity disorder (ADHD), and mood and anxiety disorders among individuals with the 22q11.2 deletion compared with the general population, but with inconsistent risk estimates between studies.12-18
Recently, individuals with the reciprocal 22q11.2 microduplication have been described,19-21 but in contrast to the 22q11.2DS, only a few studies have been published. Based on the current reports, most individuals with the 22q11.2 microduplication carry the reciprocal duplication of the approximately 3-Mb deletion that is common in the 22q11.2DS.22,23 In addition, smaller duplications within the 3 Mb region have been observed together with larger duplications of 4 to 6 Mb spanning the critical region of chromosome 22q11.2.22,24
Persons with the 22q11.2 duplication present with similar, albeit milder, phenotypic symptomology as those with the 22q11.2 deletion.19,20,23,25 The most frequently reported clinical manifestations include intellectual/learning disability, delayed psychomotor development, ADHD, autism spectrum disorder, growth retardation, muscular hypotonia, and congenital malformations.20,22,25-30 However, these manifestations are likely influenced by ascertainment bias, and 22q11.2 duplication carriers with no phenotypic appearance have been reported.20,21,27,28 To our knowledge, no longitudinal or population-based cohort studies have provided risk estimates of psychiatric disorders among individuals with the 22q11.2 duplication.
In this nationwide register-based study, we estimated incidence rate ratio (IRR) and absolute risk for psychiatric disorders among clinically identified individuals with the 22q11.2 deletion and the reciprocal duplication. In addition, we provide information about indicators for carrying a deletion or duplication at the 22q11.2 locus.
Using information from the Danish Civil Registration System,31 we established a population-based cohort that included all individuals born in Denmark between 1955 and 2012 who had references to both parents in the register (N = 3 768 943). In Denmark, all live-born children and new residents are assigned a unique personal identifier that can be used to link information within and across all nationwide computerized registers. The personal identifier is registered in the Danish Civil Registration system, which holds updated information on all Danish citizens who have resided at least 1 day in Denmark since April 1968, including date of birth, sex, identity of parents, place of birth and residence, and vital status (death and emigration). The study was conducted from May 7, 2015, to August 14, 2016. The study was approved by the Danish Data Protection Agency (No. 2012-41-0321). According to Danish law, informed consent is not required for register-based studies.
Information on individuals with a 22q11.2 deletion or duplication was obtained from the Danish Cytogenetic Central Registry (DCCR),32 which was established in 1968 and contains data on every karyotype obtained by clinical departments performing chromosomal analyses in Denmark. The reporting of data to the DCCR is self-imposed, and representatives from these departments administer the registry. Clinical genetic testing for 22q11.2 deletion and duplication was implemented in Denmark in 1994 and 1999 and is performed on clinical indication often owing to congenital or developmental abnormalities. Although the DCCR does not provide information about the reason for testing, it reflects usual clinical practice where parents are offered genetic testing upon identification of a 22q11.2 deletion or duplication in their child.
We identified the date of a positive genetic test result for individuals with a 22q11.2 deletion or duplication based on fluorescence in situ hybridization, multiplex ligation-dependent probe amplification, or comparative genomic hybridization arrays. Since several genetic techniques with different resolutions have been used at the clinical genetic departments over time, the 22q11.2 deletion or duplication was defined as the 3-Mb deletion or duplication, which is observed in approximately 90% of the individuals with the 22q11.2 deletion,33,34 and minor deletions or duplications nested within this region. To meet this criterion we used single-nucleotide polymorphism microarray analysis and verified the correct chromosomal location and size in nearly all of the individuals who were initially identified by fluorescence in situ hybridization.
Information on psychiatric disorders was obtained from the Danish Psychiatric Central Research Register,35 which contains information about all admissions to Danish psychiatric inpatient facilities since April 1969 and outpatients since 1995. Psychiatric diagnoses are based on the International Classification of Diseases, Eighth Revision36 between 1969 and 1993 and in the International Statistical Classification of Diseases and Related Health Problems, 10th Revision37 from 1994 onward. In Denmark, psychiatric diagnoses are given by psychiatrists. There are no private psychiatric inpatient facilities in Denmark, and all treatment is free of charge. This situation ensures that all psychiatric contacts are represented in the register.
Cohort members and their parents were classified as having a psychiatric disorder if they had been admitted to a psychiatric hospital or had received outpatient care. eTable 1 in the Supplement reports the spectrum of psychiatric disorders and the range of ages at risk considered. Date of onset was defined as the date of the first contact (inpatient/outpatient) for the diagnosis of interest. Because of comorbidity, the same individual may appear in more than 1 comparison.
To estimate indicators of 22q11.2 deletion/duplication status, we adopted a nested case-control study design, which was conducted by matching each individual with a 22q11.2 deletion (n = 244) or duplication (n = 58) to 100 randomly chosen persons from the population-based cohort with the same date of birth (±1 day) and no history of 22q11.2 deletion or duplication (time matched) to serve as controls. Incidence rate ratios were estimated using conditional logistic regression. We looked at the separate effects adjusted only for sex, an adjustment in groups, and a fully adjusted model.
To estimate IRR (relative risks) of developing a psychiatric disorder associated with 22q11.2 deletion or duplication, we used survival analysis. For each psychiatric disorder, individuals were followed up from the earliest age at which they may develop the specific disorder or January 1, 1995, whichever came later, until the date of onset of the outcome, death, emigration, or December 31, 2013, whichever came first. Because individuals were followed up from 1995, those having the diagnosis of interest before 1995 were excluded. We estimated IRRs using Poisson regressions, with the logarithm of person-years as an offset variable. This method applies identical follow-up in cases and the comparison population and is resistant to incomplete registration of cases. P values and 95% CIs were based on likelihood ratio test statistic. Age, calendar year, 22q11.2 deletion or duplication, and parental history of a psychiatric disorder were treated as time-dependent covariates, whereas all other variables were treated independent of time. We performed a sensitivity analysis that included only the first individual in each family identified with a 22q11.2 deletion or duplication to avoid bias from clinically unrecognized parents examined because of their child testing positive.
The absolute risk (cumulative incidence) of a psychiatric disorder in individuals with 22q11.2 deletion or duplication was estimated using competing risk survival analyses. The cumulative incidence measures the probability of developing a specific psychiatric disorder before a given point in time accounting for the fact that persons simultaneously are at risk of a psychiatric disorder, dying, or emigrating. Genetic testing for 22q11.2 deletion and duplication was introduced in 1994 and 1999, respectively, and the maximum observation follow-up time is therefore 14 years (until 2013) for which cumulative incidence was calculated. The cumulative incidences in controls were performed in 100 randomly chosen individuals from the population-based cohort with the same sex and same date of birth (±1 day) and no history of 22q11.2 deletion or duplication (time matched). These persons were followed up from the date that each matched 22q11.2 individual received a positive genetic test using the procedure described above. Data analysis was performed from June 29, 2015, to May 13, 2016.
Among the 3 768 943 individuals from the population-based cohort, 244 and 58 individuals were clinically identified with a 22q11.2 deletion or duplication, respectively. The Figure shows the distribution of age at a positive test for the 22q11.2 deletion and duplication. The eFigure in the Supplement illustrates the distribution of birth years for individuals with identified 22q11.2 deletion or duplication in the DCCR. Furthermore, the case-control counts and corresponding IRRs across different factors and adjustment models are reported in Table 1 and Table 2.
We found a higher risk of a positive test of 22q11.2 deletion or duplication among individuals with a history of psychiatric disorders (deletion, 13, and duplication, 10). The risk of deletion was higher among persons with a parent who had a history of schizophrenia. Individuals with a history of any psychiatric disorder in a parent faced an increased risk of 22q11.2 duplication, whereas results regarding schizophrenia were inconclusive. The differences in IRRs for schizophrenia among the matched controls (Tables 1 and 2) might be due to the different age distributions between 22q11.2 deletion and duplication carriers. There was a tendency that the lower the degree of urbanization, the higher the risk of 22q11.2 duplication; however, the opposite tendency occurred for 22q11.2 deletion (Tables 1 and 2).
A total of 3 653 517 individuals born in Denmark between 1955 and 2012 were followed up during 57.1 million person-years from 1995 to 2013; of these, 296 909 (men, 139 547 [47.0%]; women, 157 362 [53.0%]) had a psychiatric disorder. Table 3 reports the mean age at diagnosis, the IRRs, and sex-specific IRRs for each of the psychiatric outcomes. Individuals with the 22q11.2 deletion or duplication have increased risks of any psychiatric disorder and intellectual disability. The effects sizes were highly similar across models independent of adjusting for putative confounders, as reported in eTable 2 in the Supplement.
Individuals with the 22q11.2 deletion have a high risk of schizophrenia and pervasive developmental disorders; there were too few persons with the 22q11.2 duplication to study the risk of each diagnosis reliably. Restricting the analysis to the first identified individual in each family did not change the estimated IRRs (eTable 3 in the Supplement).
As reported in Table 3, we observed a tendency to sex differences in IRRs for several diagnoses but did not find an interaction between sex and 22q11.2 status. We conducted the analyses considering 22q11.2 as a non–time-dependent variable, thereby ignoring the potential impact of immortality bias,38 and found minor changes in IRR estimates. When disregarding the potential immortality bias, we obtained sufficient power to estimate a surrogate for the IRR of mood disorder, pervasive developmental disorder, and hyperkinetic disorder for the 22q11.2 duplication (eTable 4 in the Supplement).
Individuals with 22q11.2 deletion or duplication faced an increased risk for all psychiatric outcomes considered (Table 4), especially intellectual disability, with cumulative incidences of 15.08 (95% CI, 9.90-22.98) and 15.39 (95% CI, 5.63-42.11). Comparable cumulative incidences among individuals without the 22q11.2 were 0.73 (95% CI, 0.57-0.93).
To our knowledge, this is the first large-scale longitudinal study that estimated IRR and absolute risk of psychiatric disorders among individuals with the 22q11.2 deletion or duplication found in response to clinical suspicion. Moreover, we inform about factors associated with harboring a 22q11.2 deletion or duplication. The study gains its strength from the use of the reliable, comprehensive, nationwide, and population-based registers that contain information on diagnosis and time for all inpatient and outpatient contacts and admissions to psychiatric departments, thereby providing risk estimates of direct clinical relevance.
The study shows similarly high IRRs and cumulative incidences for any psychiatric disorders and intellectual disability among clinically identified individuals with the 22q11.2 deletion or duplication (Tables 3 and 4). The risk estimates are higher than expected for the duplication,24 which may reflect ascertainment bias in non–population-based samples, and seems to argue that clinically identified duplication carriers need the same careful and persistent clinical attention as individuals with the 22q11.2 deletion. The findings of comparable effects from 22q11.2 deletions and duplications also indicate that expression of genes in the locus needs to be fine-tuned within specific upper and lower levels to sustain normal brain development and function.
Our risk estimates for individuals with the 22q11.2 deletion are lower than those in previous reports,12-18 which could reflect methodologic differences among studies. Previous studies are based mostly on locally recruited individuals with 22q11.2 deletion and apply thorough research interviews and examinations.12,39,40 In contrast, our study was population based; it included all clinically referred individuals with a 22q11.2 deletion or duplication in Denmark and analyzed data from all contacts to the free, nationwide, and uniform health care system. The study applied survival analyses to avoid bias from younger individuals who may not have realized all their disease risk.
It is consistently reported that individuals with the 22q11.2 deletion are affected by ADHD and, in particular, by the inattentive subtype with less hyperactivity and impulsivity,12,41,42 which may be overlooked in a clinical setting, especially in persons with intellectual disability and autism spectrum disorder. This lack of recognition may explain why we found only a 2-fold increase of ADHD in this clinically ascertained population sample, and it suggests that individuals with the deletion are underdiagnosed and undertreated in Denmark. In addition, anxiety and specific phobias seem to be observed less than expected,14,43 which suggests that these disorders are underreported by the psychiatric departments or that the risk might be lower in population-based studies.
A smaller than expected risk of schizophrenia was found for the 22q11.2 deletion as previously reported by Vangkilde et al.44 Unfortunately, we were unable to examine whether individuals with duplications of chromosome 22q11.2 are protected against schizophrenia as reported by Rees et al,45 since none of the persons in our sample developed this disorder.
The identified sex-specific liability for psychiatric disorders among individuals with the 22q11.2 deletion and duplication is consistent with the previously reported sex differences of these psychiatric disorders in the general population.46 We did not observe interactions between sex and 22q11.2 status, which could either result from insufficient power to detect sex differences or suggest that the chromosomal abbreviations affect both sexes similarly.
This study also informs about different factors that indicate the 22q11.2 deletion and duplication. A parental diagnosis of schizophrenia—but not of other psychiatric diagnoses—indicates a 22q11.2 deletion (Table 1). This finding might stem from the cases in which the deletion segregates from unrecognized parents with schizophrenia or point to the importance of inherited etiologic factors that, along with the deletion, underlie psychiatric disorders in this cohort.47 In contrast, parental psychiatric diagnosis (other than schizophrenia) indicates duplication carrier status, which is consistent with the notion that the duplication is frequently transmitted from affected parents.20-22 However, we did not have reliable information on the de novo rate in this cohort and therefore could not assess these hypotheses.
Parental migration patterns could in some part explain the differences in place of birth of individuals with the 22q11.2 deletion and duplication (Tables 1 and 2). It has previously been shown that patients with schizophrenia tend to migrate toward larger cities,48 which could also be true for parents with schizophrenia and the 22q11.2 deletion. Contrary to this finding, the milder parental psychiatric manifestations in individuals harboring the duplication could result in opposite migration patterns; however, this difference has never been thoroughly studied.
The strength of the study is the use of the large nationwide health registers based on contacts with inpatient and outpatient psychiatric departments in Denmark in which all treatment is provided by the free uniform and national health care system. Thus, financial factors are less likely to influence access to care in this study.49 The study population is representative of the Danish population, because all residences are included independent of demographic, social, and health-related issues.31 The large nationwide health registers allow complete lifetime follow-up, with the possibility to adjust for a wide range of confounding factors. Diagnoses, including schizophrenia, single depressive episode, affective disorders, and autism spectrum disorder, have high validity in the Danish Psychiatric Central Research Register,50-53 which reduces information bias or systematic misclassification. An additional strength is the prospective design of the study, which ensures that the exposures were recorded independently of the outcome, thereby not being subject to selection or recall bias. The study was based on clinically identified individuals with 22q11.2 deletion or duplication; thus, our risk estimates are applicable for genetic counseling of parents who have affected children who have been clinically identified. As mentioned above, this study was based on inpatient and outpatient contacts, and therefore does not include milder conditions or forms of disorders that are never seen in hospital settings or are overlooked owing to focus on severe clinical symptomatology, such as psychosis.
The 22q11.2 cohorts were extracted from the DCCR, which is surveilled by the clinical board and the Medical Society of Clinical Genetics. The DCCR is a complete registry that systematically collects all karyotypes identified in patients referred on clinical suspicion often related to congenital or developmental abnormalities.32 As such, the present study is limited by several factors. First, the register does not include 22q11.2 deletion or duplication carriers who, for some reason, are not subjected to genetic testing (eg, parents who refuse genetic testing of their children or themselves). Hence, incomplete registration may reduce the power of the study. Second, individuals diagnosed with the 22q11.2 deletion or duplication might be subject to increased medical attention, which makes them more prone to psychiatric assessment than the general population. Such information bias will inflate the estimates, but since the reported IRRs were somewhat lower than previous estimates, this bias might be less important. Third, diagnostic assessment of individuals with 22q11.2 deletion or duplication is challenging since the classification systems were not originally devised for individuals with neurodevelopmental disorders; this lack of specificity could bias the estimates.
This population-based study provides what we believe to be the first comprehensive assessment of the IRRs and absolute risks for psychiatric disorders among individuals clinically identified as having the 22q11.2 deletion or duplication. We found similar IRRs for the 22q11.2 duplication compared with the deletion, suggesting that individuals with the duplication need the same careful and persistent clinical attention as individuals with the deletion. Our results are of direct clinical relevance to assist clinical ascertainment, genetic counseling, guidance of symptomatic monitoring, and early clinical intervention.
Corresponding Author: Louise K. Hoeffding, PhD, Institute of Biological Psychiatry, Mental Health Centre Sct. Hans, Copenhagen University Hospital, Boserupvej 2, 4000 Roskilde, Denmark (email@example.com).
Accepted for Publication: November 27, 2016.
Published Online: January 18, 2017. doi:10.1001/jamapsychiatry.2016.3939
Author Contributions: Dr Hoeffding and Ms Trabjerg had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Hoeffding, Trabjerg, Mazin, Vangkilde, Pedersen, Werge.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Hoeffding, Werge.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Hoeffding, Trabjerg, Werge.
Administrative, technical, or material support: Olsen, Pedersen, Werge.
Study supervision: Trabjerg, Mazin, Mortensen, Pedersen, Werge.
Conflict of Interest Disclosures: Dr Werge has served as a paid lecturer for and consultant to H. Lundbeck A/S. No other disclosures were reported.
Funding/Support: This study was financed through grant 001-2009-2 from the Danish National Advanced Technology Foundation (Dr Werge) and by grants R219-2016-1030 (Dr Hoeffding), R155-2014-1724, R34-A3243 (Dr Werge), and R171-2014-1194 (Dr Sparsø) from the Lundbeck Foundation.
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Jan Hansen (Danish Cytogenetic Central Register, Cytogenetic Registry, University Hospital, Aarhus, Denmark), Karen Brøndum, PhD (Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark), and the scientific board of the Danish Cytogenetic Central Register provided access to the genetic data and approved the final manuscript. There was no financial compensation.