0
We're unable to sign you in at this time. Please try again in a few minutes.
Retry
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
Retry
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Article |

Autism Prevalence Trends Over Time in Denmark:  Changes in Prevalence and Age at Diagnosis FREE

Erik T. Parner, PhD; Diana E. Schendel, PhD; Poul Thorsen, PhD
[+] Author Affiliations

Author Affiliations: Department of Biostatistics (Dr Parner) and North Atlantic Neuro Epidemiology Alliances, Department of Epidemiology (Drs Parner and Thorsen), Institute of Public Health, University of Aarhus, Aarhus, Denmark; and National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention (Dr Schendel), and Department of Epidemiology, Rollins School of Public Health, Emory University (Dr Thorsen), Atlanta, Georgia.


Arch Pediatr Adolesc Med. 2008;162(12):1150-1156. doi:10.1001/archpedi.162.12.1150.
Text Size: A A A
Published online

Objective  To examine the effect of changing age at diagnosis on the diagnosed prevalence of autism among different birth cohorts.

Design  Population-based cohort study.

Setting  Children were identified in the Danish Medical Birth Registry and psychiatric outcomes were obtained via linkage with the Danish National Psychiatric Register.

Participants  All children born in Denmark from January 1, 1994, through December 31, 1999 (N = 407 458).

Main Outcome Measures  The age-specific prevalence, hazard ratio, and relative risk by age.

Results  Statistically significant shifts in age at diagnosis were observed for autism spectrum disorder; children diagnosed before age 9 years in the cohorts born between January 1, 1994, and December 31, 1995, between January 1, 1996, and December 31, 1997, and between January 1, 1998, and December 31, 1999, were on average diagnosed at ages 5.9 (95% confidence interval [CI], 5.8-6.0), 5.8 (95% CI, 5.7-5.9), and 5.3 (95% CI, 5.2-5.4) years, respectively. The relative risk comparing the 1996-1997 birth cohort with the 1994-1995 birth cohort at age 3 years was 1.20 (95% CI, 0.86-1.67), which decreased to 1.10 (95% CI, 1.00-1.20) at age 11 years. Similarly, the relative risk comparing the 1998-1999 birth cohort with the 1994-1995 birth cohort at age 3 years was 1.69 (95% CI, 1.24-2.31), which decreased to 1.23 (95% CI, 1.11-1.37) at age 11 years. Similar results were observed for childhood autism.

Conclusions  Shifts in age at diagnosis inflated the observed prevalence of autism in young children in the more recent cohorts compared with the oldest cohort. This study supports the argument that the apparent increase in autism in recent years is at least in part attributable to decreases in the age at diagnosis over time.

Figures in this Article

A recent review of the epidemiology of autism documented the large number of studies that have shown an increase in the age-specific prevalence of reported autism cases in the last 2 decades.1 Once considered rare, the current prevalence estimates for all autism spectrum disorders of about 6 cases per 1000 children indicate that autism spectrum disorder may be the second most common serious developmental disability after mental retardation.24 Recent studies have also shown that an upward trend in prevalence is not unique to autism but is also the case for other childhood neuropsychiatric disorders such as attention-deficit/hyperactivity disorder and Tourette syndrome.5,6 The increase in the reported prevalence of autism has raised the concerns that the incidence of autism is increasing and that the increase may be due to adverse exposures such as childhood vaccines.7

The increase in the prevalence of reported cases may be caused by many factors such as heightened public awareness, changes in referral pattern, changes in diagnostic criteria, case identification, or reporting methods. In the few studies that have investigated the effect of these potential confounders, the main focus has been on diagnostic substitution in which a child who might have received a disability label such as learning disability or mental retardation 15 to 20 years ago may now be diagnosed with autism because of shifting diagnostic practices.6,8,9 These studies suggest that diagnostic substitution may explain part of the increase in the prevalence of autism. Correct interpretation of changes in autism prevalence is important in the public health setting for services, planning, and organization of institutions and also in etiological research when searching for risk factors associated with autism. In etiological research, special attention has been on risk factors that show an increase in prevalence similar to the increase in autism prevalence.

A limited number of studies have investigated the effect of changing age at diagnosis on the prevalence of autism. Lingam et al10 presented estimates of the change in age at diagnosis and the increase in final prevalence of autism by making specific assumptions about the diagnostic rate as a function of age and changes in the age at diagnosis. These assumptions, however, cannot be validated from data. Mandell et al11 presented the average age at diagnosis in different birth cohorts, but the difference in age at diagnosis may also reflect differences in follow-up time in the birth cohorts. Wazana et al12 performed a simulation study showing that changes in the age at diagnosis may confound the observed increase in prevalence.

Changes in the age at diagnosis may confound the observed increase in prevalence if there is insufficient follow-up to estimate the final prevalence. Figure 1 shows this problem for a hypothetical disorder where diagnosis of all affected persons is achieved before age 20 years. The reference cohort represents a birth cohort of the oldest children, and cohorts 1 and 2 represent more recent cohorts under 2 different scenarios. In cohort 1, the distribution of age at diagnosis is identical to the reference cohort, but there are 10% more cases at each year of age; at age 20 years, the prevalence is 10% greater than in the reference cohort. For cohort 2, there is a shorter time to diagnosis in that the diagnoses in cohort 2 are on average given 10% earlier compared with the reference cohort, but the final prevalence at age 20 years is the same for cohort 2 and the reference cohort. With complete follow-up, the prevalence curves for each cohort reach a plateau by age 20 years, indicating that no additional cases are diagnosed after that age (Figure 1A). In the case of complete follow-up, it is straightforward to estimate the increase in final prevalence between the reference cohort and cohort 1 as the relative difference between the prevalence proportions at age 20 years. If there is not sufficient follow-up to estimate the final prevalence, we can only estimate the age-specific prevalence (Figure 1B). In this example with incomplete follow-up, the curves for cohorts 1 and 2 seem to converge and are shifted away from the reference cohort—showing the potential difficulty in distinguishing the effects of an earlier age at diagnosis from the effects of an increase in prevalence.

Place holder to copy figure label and caption
Figure 1.

The prevalence of a hypothetical disorder in 3 cohorts: a reference cohort, cohort 1 with a 10% increase in prevalence compared with the reference cohort, and cohort 2 where the diagnoses are given 10% earlier than in the reference cohort. A, The prevalence curves with complete follow-up. B, The prevalence curves with incomplete follow-up to ages 9 and 11 years. C, The hazard ratio for cohorts 1 and 2 relative to the reference cohort. D, The relative risk for cohorts 1 and 2 relative to the reference cohort.

Graphic Jump Location

The aims of this study were to examine whether there is evidence for changes in the age at diagnosis and an effect on the reported prevalence of autism in Denmark and to estimate the amount of bias when using age-specific relative risk to estimate the final relative risk that is caused by changes in the age at diagnosis.

STUDY DESIGN

We designed a cohort study of all children born in Denmark from January 1, 1994, to December 31, 1999. The cohort was identified in the Danish Medical Birth Registry and consisted of 407 458 children. The Danish Medical Birth Registry comprises data on all live births and stillbirths by women with permanent residence in Denmark.13 All live-born children in Denmark are assigned a Central Population Registry number, a unique 10-digit number used for all official personal registrations in Denmark since 1968. On the basis of the Central Population Registry number, the data from the Danish Medical Birth Registry were then linked to the Danish National Psychiatric Registry (DNPR) to obtain neuropsychiatric outcomes. The DNPR includes all inpatient admissions and from January 1, 1995, also includes all outpatient contact to psychiatric hospitals, wards, and clinics in Denmark.14 For children, inpatient admissions correspond to overnight hospital stays or daily hospital visits over an extended period for diagnostic evaluation and treatment, whereas outpatient admissions correspond to clinic visits on a less regular basis. Children suspected to have autism are referred by general practitioners or school psychologists to a child psychiatric ward, where they are diagnosed and treated by a psychiatrist. In Denmark, a specialized diagnostic assessment of children suspected to have autism is generally necessary to be enrolled in special services, and the diagnostic evaluations and treatment are free of charge. The DNPR includes data on clinical diagnoses, dates of admission and discharge, and terms of admission, and the International Statistical Classification of Diseases, 10th Revision (ICD-10) diagnostic code criteria have been used since 1994.15 The age at first admission and/or contact with the health care system is used as a proxy for the age at diagnosis, and age at first contact in the diagnostic process may be a more accurate description because the timing of diagnosis once the child is admitted into the process may vary across facilities.

Follow-up for the diagnosis of autism spectrum disorder (ICD-10 codes: F84.0, childhood autism; F84.1, atypical autism; F84.5, Asperger syndrome; F84.8, other pervasive developmental disorders; and F84.9, pervasive developmental disorders unspecified) or childhood autism (ICD-10 code F84.0) began for all children from birth and continued until the diagnosis of autism, death, or the end of follow-up on December 31, 2006, whichever occurred first. All of the analyses were made separately for autism spectrum disorder and childhood autism. The children were divided into 2-year birth cohorts, born between January 1, 1994, and December 31, 1995, between January 1, 1996, and December 31, 1997, and between January 1, 1998, and December 31, 1999 (to enhance the statistical precision compared with 1-year birth cohorts), where the 1994-1995 cohort represents the reference cohort.

The study was approved by the Danish National Board of Health and the Danish Data Protection Agency.

STATISTICAL ANALYSIS

Previous studies on autism time trends that provide data on the prevalence by age or equivalently the rate by age, including this study, do not seem to have sufficient follow-up to estimate the final prevalence of autism.5,1619 Specifically, the prevalence by age curve does not reach a plateau, or the rate by age does not converge to 0, in these studies. Because childhood mortality is low in Denmark and autism is considered a chronic disorder, the risk of being diagnosed with autism can be interpreted as the autism prevalence.

To estimate the final relative risk in 2 cohorts when follow-up is not long enough for prevalence to reach a plateau, 1 of the following 2 methods can generally be used: (1) estimation of the relative risk at a specific age, such as age 5 years; or (2) estimation of the hazard ratio, the ratio of the instantaneous rate of a new diagnosis, at a specific age. However, these methods provide valid estimates of the final relative risk only if the distribution of age at diagnosis is the same in both cohorts. On the other hand, a hazard ratio between 2 birth cohorts that is not constant as a function of age is evidence for changes in the distribution of age at diagnosis (Figure 1C in the hypothetical example).

Evidence for Changes in Age at Diagnosis

The prevalence by age for each specific cohort was estimated using the Kaplan-Meier method. The prevalence by age in different cohorts was compared in 2 ways. An overall test for equal age-specific prevalence was performed using the Breslow-Gehan test. Also, a test of constant hazard ratio was performed using flexible parametric hazard spline regression models with 3 df.20 A hazard ratio between 2 cohorts that is not constant as a function of age is evidence of changes in the distribution of age at diagnosis. Estimation of the hazard ratio by age was done using the flexible parametric hazard regression models. The hazard ratio by age compares the rate of new cases at given ages between 2 cohorts. The flexible parametric hazard spline models were applied to estimate a smooth hazard ratio by age curve.

Bias Caused by Changes in Age at Diagnosis

When the distribution of age at diagnosis is not similar between 2 cohorts, then both the hazard ratio and the age-specific relative risk are biased estimates of the final relative risk (Figure 1C and D in the hypothetical example). An initially large age-specific relative risk that decreases thereafter is evidence of an earlier age at diagnosis among the youngest ages; if a recent cohort has autism cases and the reference cohort has essentially no cases, then this pattern can only be due to an earlier average diagnosis in the recent cohort at this young age. For this study, the level of bias in reported prevalence caused by changes in age at diagnosis was evaluated by estimates of the age-specific relative risk for varying lengths of follow-up. The relative risk by age shows how the estimated increase in prevalence depends on the length of follow-up. Estimation of the relative risk by age was done using the flexible parametric hazard regression models.20 The flexible parametric hazard spline models were applied to estimate a smooth relative risk by age curve.

A total of 2649 children were diagnosed with autism. Of them, 881 were diagnosed with childhood autism (Table 1).

Table Graphic Jump LocationTable 1. ICD-10 Diagnostic Codes, Birth Cohorts, Live Births per Cohort, Autism Cases, and Prevalence in Each Cohort Reported to the Danish National Psychiatric Registry by December 31, 2006
EVIDENCE FOR CHANGES IN AGE AT DIAGNOSIS

Both autism spectrum disorder and childhood autism show statistically significant increases in the age-specific prevalence in the most recent birth cohort (1998-1999) compared with the 1994-1995 reference cohort (Figure 2) (P < .001 for both autism spectrum disorder and childhood autism); the 1996-1997 cohort also had a significant increase in autism spectrum disorder compared with the reference cohort (P = .04 for autism spectrum disorder and P = .22 for childhood autism). The tests for trend were statistically significant for both autism spectrum disorder and childhood autism (P < .001 in both cases).

Place holder to copy figure label and caption
Figure 2.

Age-specific prevalence of autism spectrum disorder (A) and childhood autism (B) for each 2-year birth cohort (1 minus the Kaplan-Meier curve).

Graphic Jump Location

For both autism spectrum disorder and childhood autism there was evidence for hazard ratios that are not constant as a function of age and hence indicate changes in the age at diagnosis in the more recent cohorts (Figure 3) (autism spectrum disorder: 1994-1995 cohort vs 1996-1997 cohort, P = .04; 1994-1995 cohort vs 1998-1999 cohort, P = .01; childhood autism: 1994-1995 cohort vs 1996-1997 cohort, P = .003; 1994-1995 cohort vs 1998-1999 cohort, P = .02). The initial decrease in the hazard rate is evidence that the average age at diagnosis was younger among children in the more recent cohorts compared with the reference cohort. The earlier diagnosis is reflected in decreases in the average age at diagnosis: children diagnosed with an autism spectrum disorder before age 9 years in the 1994-1995, 1996-1997, and 1998-1999 cohorts were on average diagnosed at ages 5.9 (95% confidence interval [CI], 5.8-6.0), 5.8 (95% CI, 5.7-5.9), and 5.3 (95% CI, 5.2-5.4) years, respectively. For a diagnosis of childhood autism specifically, the corresponding average ages at diagnosis were 5.1 (95% CI, 4.9-5.4), 5.0 (95% CI, 4.8-5.2), and 4.7 (95% CI, 4.6-4.9) years, respectively. There were 69% (95% CI, 24%-131%) more cases of autism spectrum disorder and 87% (95% CI, 21%-187%) more cases of childhood autism reported at age 3 years in children in the 1998-1999 birth cohort relative to children in the 1994-1995 birth cohort (Figure 3).

Place holder to copy figure label and caption
Figure 3.

Age-specific hazard ratio of reported cases of autism spectrum disorder in the 1998-1999 birth cohort (A), autism spectrum disorder in the 1996-1997 birth cohort (B), childhood autism in the 1998-1999 birth cohort (C), and childhood autism in the 1996-1997 birth cohort (D) relative to the reference cohort of children born in 1994-1995. CI indicates confidence interval.

Graphic Jump Location
BIAS CAUSED BY CHANGES IN AGE AT DIAGNOSIS

The level of bias in reported prevalence caused by changes in age at diagnosis is shown by a graph of the age-specific relative risks for varying lengths of follow-up (Figure 4). The estimated age-specific relative risks for both autism spectrum disorder and childhood autism decreased with age as the follow-up of the children was increased.

Place holder to copy figure label and caption
Figure 4.

Age-specific relative risk of autism spectrum disorder in the 1998-1999 birth cohort (A), autism spectrum disorder in the 1996-1997 birth cohort (B), childhood autism in the 1998-1999 birth cohort (C), and childhood autism in the 1996-1997 birth cohort (D) relative to the reference cohort of children born in 1994-1995. CI indicates confidence interval.

Graphic Jump Location

Follow-up only to age 3 years resulted in an estimated relative risk of 1.20 (95% CI, 0.86-1.67) for autism spectrum disorder comparing the 1996-1997 birth cohort with the 1994-1995 birth cohort and a relative risk of 1.69 (95% CI, 1.24-2.31) comparing the 1998-1999 birth cohort with the 1994-1995 birth cohort (Figure 4). Increasing the follow-up to age 11 years resulted in an estimated relative risk of 1.10 (95% CI, 1.00-1.20) comparing the 1994-1995 and 1996-1997 birth cohorts, indicating that 52% of the estimated increase in prevalence at age 3 years could be explained by the longer follow-up to age 11 years. Similarly, increasing the follow-up from age 3 years to age 9 years resulted in an estimated relative risk of 1.23 (95% CI, 1.11-1.37) when comparing the 1994-1995 and 1998-1999 birth cohorts, indicating that at least 66% of the apparent increased prevalence at age 3 years in the 1998-1999 cohort compared with the 1994-1995 cohort could be due to shifts in age at diagnosis between the 1994-1995 and 1998-1999 cohorts (Table 2).

Table Graphic Jump LocationTable 2. Age-Specific Relative Risk in Each 2-Year Birth Cohort Relative to the Reference Cohort

The final relative risk for a given cohort compared with the reference cohort will be the limit of the age-specific relative risk as the follow-up increases, eventually reaching a plateau as the length of follow-up increases to the oldest age at diagnosis (ie, at complete follow-up). Further, when there is a decrease in age at diagnosis, the final relative risk can be no larger than the age-specific relative risk at the end of follow-up. In these data, compared with the reference cohort, the final relative risk for autism spectrum disorder at the end of follow-up was 1.10 (95% CI, 1.00-1.20) in the 1996-1997 cohort and 1.23 (95% CI, 1.11-1.37) in the 1998-1999 cohort, indicating an increase in the final prevalence for autism spectrum disorder in the youngest cohort at least. For childhood autism, the relative risk of 1.50 (95% CI, 1.25-1.75) based on follow-up to age 9 years in the 1998-1999 cohort compared with the reference cohort is also statistically significant.

Based on data from a cohort of children born in Denmark since the mid-1990s, the study provides quantitative evidence for a change in the age at diagnosis of autism over time and the potential effect of that change on the observed risk for a reported diagnosis of autism at different ages. Statistically significant shifts in age at diagnosis were observed for both autism spectrum disorder and childhood autism. The important conclusion to be made is that the shifts in age at diagnosis—especially the earlier diagnosis at younger ages—artificially inflated the differences in the observed prevalence of autism in young children in the more recent cohorts compared with the cohort of the oldest children. With a greater length of follow-up, the differences in prevalence between the cohorts decreased. It appears that in our data, however, the length of follow-up was insufficient to observe a plateau in the autism prevalence in the more recent cohorts. With longer follow-up, the relative risks of the 1996-1997 cohort and the 1998-1999 cohort compared with the 1994-1995 cohort at the end of follow-up could potentially change. Furthermore, it remains to be determined whether the statistically significant increased prevalence at the end of follow-up could be further accounted for by other factors or is reflective of a true increase in the incidence of reported autism diagnoses.

The strengths of this study include a population-based birth cohort approach, a large cohort size, and all diagnostic data having been obtained from a nationwide register based on standardized diagnostic reporting procedures and using a common diagnostic coding system (ICD-10). The ICD-10 classification system has been used in Denmark since 1994 and both inpatient and outpatient data have been included in the DNPR since 1995. Hence, selection of the study cohort was made explicitly to ensure that administrative reporting of autism to the DNPR was stable throughout the study follow-up period between 1995 and 2006. However, the findings could also reflect better diagnostic acumen of the ICD-10 diagnostic system from the 1994-1995 cohort to the later 1996-1997 and 1998-1999 cohorts. The study is based on clinical diagnosis as a case identification strategy, and the effect of age at diagnosis may be different if other case identification strategies are used. However, clinical diagnoses are used to estimate autism prevalence, and this study highlights the difficulty in interpretation of such prevalence data.

Completeness of the DNPR for autism is believed to be high. In other studies, the prevalence of autism spectrum disorder has generally been reported to range from 33 to 106 cases per 10 000 individuals.4,16,17,21,22 Childhood autism is most commonly reported to have a prevalence of approximately 20 cases per 10 000 individuals.6,8,18,21,23 The age-specific prevalences observed in this study in the oldest cohort (Figure 2) (1994-1995 cohort, aged 13 years: 82 cases per 10 000 individuals for autism spectrum disorder and 21 cases per 10 000 individuals for childhood autism) are comparable. No validations of accuracy have been made for autism spectrum disorder diagnoses in the DNPR in general. An assessment based on expert review of diagnostic records of 499 children born in the 1990s with a reported DNPR diagnosis of childhood autism found that 94% met criteria for a correct diagnosis based on a coding scheme developed by the Centers for Disease Control and Prevention for autism surveillance in the United States (Marlene B. Lauritsen, MD, Meta Jørgensen, MD, Sanne Lemcke, RN, Kresten M. Madsen, PhD, Susanne Toft, RN, Jakob Grove, PhD, D.E.S., and P.T., unpublished data, June 2008).

The decrease in age at diagnosis may be due to better diagnostic abilities of doctors, an earlier onset of autism symptoms in new cases of autism, or an earlier onset of autism symptoms that would have otherwise appeared at an older age.

In conclusion, this study supports the argument that the apparent increase in autism in recent years is at least in part attributable to a decrease in the age at diagnosis over time. The empirical evidence for a decrease in the age at diagnosis of autism in more recent years is important as it is well known that the long-term prognosis for children with autism is enhanced by early identification and treatment.24 This study also provides an approach for quantifying the proportion of the increase in autism prevalence observed at specific ages that is attributable to changes in age-specific hazard rates of diagnosis. The results demonstrate the effect of incomplete follow-up on the potential for misinterpretation of autism trends.

Correspondence: Erik T. Parner, PhD, Department of Biostatistics, Institute of Public Health, University of Aarhus, Vennelyst Boulevard 6, DK-8000 Aarhus C, Denmark (parner@biostat.au.dk).

Accepted for Publication: May 8, 2008.

Author Contributions: Dr Parner had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Parner, Schendel, and Thorsen. Acquisition of data: Parner. Analysis and interpretation of data: Parner, Schendel, and Thorsen. Drafting of the manuscript: Parner. Critical revision of the manuscript for important intellectual content: Parner, Schendel, and Thorsen. Statistical analysis: Parner. Obtained funding: Schendel and Thorsen. Administrative, technical, and material support: Schendel and Thorsen.

Financial Disclosure: None reported.

Disclaimer: The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.

Fombonne  E Epidemiology of autistic disorder and other pervasive developmental disorders. J Clin Psychiatry 2005; (66 Suppl 10) 3- 8
PubMed
Yeargin-Allsopp  MRice  CKarapurkar  TDoernberg  NBoyle  CMurphy  C Prevalence of autism in a US metropolitan area. JAMA 2003;289 (1) 49- 55
PubMed Link to Article
Bhasin  TKBrocksen  SAvchen  RNVan Naarden Braun  K Prevalence of four developmental disabilities among children aged 8 years: Metropolitan Atlanta Developmental Disabilities Surveillance Program, 1996 and 2000. MMWR Surveill Summ 2006;55 (1) 1- 9
PubMed
Autism and Developmental Disabilities Monitoring Network Surveillance Year 2002 Principal Investigators; Centers for Disease Control and Prevention, Prevalence of autism spectrum disorders: autism and developmental disabilities monitoring network, 14 sites, United States, 2002. MMWR Surveill Summ 2007;56 (1) 12- 28
PubMed
Atladóttir  HOParner  ETSchendel  DDalsgaard  SThomsen  PHThorsen  P Time trends in the reported childhood diagnoses of neuropsychiatric disorders: a Danish cohort study. Arch Pediatr Adolesc Med 2007;161 (2) 193- 198
PubMed Link to Article
Newschaffer  CJFalb  MDGurney  JG National autism prevalence trends from United States special education data. Pediatrics 2005;115 (3) e277- e282
PubMed Link to Article
Demicheli  VJefferson  TRivetti  APrice  D Vaccines for measles, mumps and rubella in children. Cochrane Database Syst Rev 2005; (4) CD004407
PubMed
Croen  LAGrether  JKHoogstrate  JSelvin  S The changing prevalence of autism in California. J Autism Dev Disord 2002;32 (3) 207- 215
PubMed Link to Article
Shattuck  PT The contribution of diagnostic substitution to the growing administrative prevalence of autism in US special education. Pediatrics 2006;117 (4) 1028- 1037
PubMed Link to Article
Lingam  RSimmons  AAndrews  NMiller  EStowe  JTaylor  B Prevalence of autism and parentally reported triggers in a north east London population. Arch Dis Child 2003;88 (8) 666- 670
PubMed Link to Article
Mandell  DSMaytali  MNovak  MAZubritsky  CD Factors associated with age at diagnosis among children with autism spectrum disorders. Pediatrics 2005;116 (6) 1480- 1486
PubMed Link to Article
Wazana  ABresnahan  MKline  J The autism epidemic: fact or artifact? J Am Acad Child Adolesc Psychiatry 2007;46 (6) 721- 730
PubMed Link to Article
Knudsen  LBOlsen  J The Danish Medical Birth Registry. Dan Med Bull 1998;45 (3) 320- 323
PubMed
Munk-Jørgensen  PMortensen  PB The Danish Psychiatric Central Register. Dan Med Bull 1997;44 (1) 82- 84
PubMed
World Health Organization, International Statistical Classification of Diseases, 10th Revision (ICD-10).  Geneva, Switzerland World Health Organization1992;
Barbaresi  WJKatusic  SKColligan  RCWeaver  ALJacobsen  SJ The incidence of autism in Olmsted County, Minnesota, 1976-1997: results from a population-based study. Arch Pediatr Adolesc Med 2005;159 (1) 37- 44
PubMed Link to Article
Gurney  JGFritz  MSNess  KKSievers  PNewschaffer  CJShapiro  EG Analysis of prevalence trends of autism spectrum disorder in Minnesota. Arch Pediatr Adolesc Med 2003;157 (7) 622- 627
PubMed Link to Article
Lauritsen  MBPedersen  CBMortensen  PB The incidence and prevalence of pervasive developmental disorders: a Danish population-based study. Psychol Med 2004;34 (7) 1339- 1346
PubMed Link to Article
Mandell  DSThompson  WWWeintraub  ESDestefano  FBlank  MB Trends in diagnosis rates for autism and ADHD at hospital discharge in the context of other psychiatric diagnoses. Psychiatr Serv 2005;56 (1) 56- 62
PubMed Link to Article
Royston  PParmar  MK Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects. Stat Med 2002;21 (15) 2175- 2197
PubMed Link to Article
Chakrabarti  SFombonne  E Pervasive developmental disorders in preschool children: confirmation of high prevalence. Am J Psychiatry 2005;162 (6) 1133- 1141
PubMed Link to Article
Bertrand  JMars  ABoyle  CBove  FYeargin-Allsopp  MDecoufle  P Prevalence of autism in a United States population: the Brick Township, New Jersey, investigation. Pediatrics 2001;108 (5) 1155- 1161
PubMed Link to Article
Dales  LHammer  SJSmith  NJ Time trends in autism and in MMR immunization coverage in California. JAMA 2001;285 (9) 1183- 1185
PubMed Link to Article
Landa  RJ Diagnosis of autism spectrum disorders in the first 3 years of life. Nat Clin Pract Neurol 2008;4 (3) 138- 147
PubMed Link to Article

Figures

Place holder to copy figure label and caption
Figure 1.

The prevalence of a hypothetical disorder in 3 cohorts: a reference cohort, cohort 1 with a 10% increase in prevalence compared with the reference cohort, and cohort 2 where the diagnoses are given 10% earlier than in the reference cohort. A, The prevalence curves with complete follow-up. B, The prevalence curves with incomplete follow-up to ages 9 and 11 years. C, The hazard ratio for cohorts 1 and 2 relative to the reference cohort. D, The relative risk for cohorts 1 and 2 relative to the reference cohort.

Graphic Jump Location
Place holder to copy figure label and caption
Figure 2.

Age-specific prevalence of autism spectrum disorder (A) and childhood autism (B) for each 2-year birth cohort (1 minus the Kaplan-Meier curve).

Graphic Jump Location
Place holder to copy figure label and caption
Figure 3.

Age-specific hazard ratio of reported cases of autism spectrum disorder in the 1998-1999 birth cohort (A), autism spectrum disorder in the 1996-1997 birth cohort (B), childhood autism in the 1998-1999 birth cohort (C), and childhood autism in the 1996-1997 birth cohort (D) relative to the reference cohort of children born in 1994-1995. CI indicates confidence interval.

Graphic Jump Location
Place holder to copy figure label and caption
Figure 4.

Age-specific relative risk of autism spectrum disorder in the 1998-1999 birth cohort (A), autism spectrum disorder in the 1996-1997 birth cohort (B), childhood autism in the 1998-1999 birth cohort (C), and childhood autism in the 1996-1997 birth cohort (D) relative to the reference cohort of children born in 1994-1995. CI indicates confidence interval.

Graphic Jump Location

Tables

Table Graphic Jump LocationTable 1. ICD-10 Diagnostic Codes, Birth Cohorts, Live Births per Cohort, Autism Cases, and Prevalence in Each Cohort Reported to the Danish National Psychiatric Registry by December 31, 2006
Table Graphic Jump LocationTable 2. Age-Specific Relative Risk in Each 2-Year Birth Cohort Relative to the Reference Cohort

References

Fombonne  E Epidemiology of autistic disorder and other pervasive developmental disorders. J Clin Psychiatry 2005; (66 Suppl 10) 3- 8
PubMed
Yeargin-Allsopp  MRice  CKarapurkar  TDoernberg  NBoyle  CMurphy  C Prevalence of autism in a US metropolitan area. JAMA 2003;289 (1) 49- 55
PubMed Link to Article
Bhasin  TKBrocksen  SAvchen  RNVan Naarden Braun  K Prevalence of four developmental disabilities among children aged 8 years: Metropolitan Atlanta Developmental Disabilities Surveillance Program, 1996 and 2000. MMWR Surveill Summ 2006;55 (1) 1- 9
PubMed
Autism and Developmental Disabilities Monitoring Network Surveillance Year 2002 Principal Investigators; Centers for Disease Control and Prevention, Prevalence of autism spectrum disorders: autism and developmental disabilities monitoring network, 14 sites, United States, 2002. MMWR Surveill Summ 2007;56 (1) 12- 28
PubMed
Atladóttir  HOParner  ETSchendel  DDalsgaard  SThomsen  PHThorsen  P Time trends in the reported childhood diagnoses of neuropsychiatric disorders: a Danish cohort study. Arch Pediatr Adolesc Med 2007;161 (2) 193- 198
PubMed Link to Article
Newschaffer  CJFalb  MDGurney  JG National autism prevalence trends from United States special education data. Pediatrics 2005;115 (3) e277- e282
PubMed Link to Article
Demicheli  VJefferson  TRivetti  APrice  D Vaccines for measles, mumps and rubella in children. Cochrane Database Syst Rev 2005; (4) CD004407
PubMed
Croen  LAGrether  JKHoogstrate  JSelvin  S The changing prevalence of autism in California. J Autism Dev Disord 2002;32 (3) 207- 215
PubMed Link to Article
Shattuck  PT The contribution of diagnostic substitution to the growing administrative prevalence of autism in US special education. Pediatrics 2006;117 (4) 1028- 1037
PubMed Link to Article
Lingam  RSimmons  AAndrews  NMiller  EStowe  JTaylor  B Prevalence of autism and parentally reported triggers in a north east London population. Arch Dis Child 2003;88 (8) 666- 670
PubMed Link to Article
Mandell  DSMaytali  MNovak  MAZubritsky  CD Factors associated with age at diagnosis among children with autism spectrum disorders. Pediatrics 2005;116 (6) 1480- 1486
PubMed Link to Article
Wazana  ABresnahan  MKline  J The autism epidemic: fact or artifact? J Am Acad Child Adolesc Psychiatry 2007;46 (6) 721- 730
PubMed Link to Article
Knudsen  LBOlsen  J The Danish Medical Birth Registry. Dan Med Bull 1998;45 (3) 320- 323
PubMed
Munk-Jørgensen  PMortensen  PB The Danish Psychiatric Central Register. Dan Med Bull 1997;44 (1) 82- 84
PubMed
World Health Organization, International Statistical Classification of Diseases, 10th Revision (ICD-10).  Geneva, Switzerland World Health Organization1992;
Barbaresi  WJKatusic  SKColligan  RCWeaver  ALJacobsen  SJ The incidence of autism in Olmsted County, Minnesota, 1976-1997: results from a population-based study. Arch Pediatr Adolesc Med 2005;159 (1) 37- 44
PubMed Link to Article
Gurney  JGFritz  MSNess  KKSievers  PNewschaffer  CJShapiro  EG Analysis of prevalence trends of autism spectrum disorder in Minnesota. Arch Pediatr Adolesc Med 2003;157 (7) 622- 627
PubMed Link to Article
Lauritsen  MBPedersen  CBMortensen  PB The incidence and prevalence of pervasive developmental disorders: a Danish population-based study. Psychol Med 2004;34 (7) 1339- 1346
PubMed Link to Article
Mandell  DSThompson  WWWeintraub  ESDestefano  FBlank  MB Trends in diagnosis rates for autism and ADHD at hospital discharge in the context of other psychiatric diagnoses. Psychiatr Serv 2005;56 (1) 56- 62
PubMed Link to Article
Royston  PParmar  MK Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects. Stat Med 2002;21 (15) 2175- 2197
PubMed Link to Article
Chakrabarti  SFombonne  E Pervasive developmental disorders in preschool children: confirmation of high prevalence. Am J Psychiatry 2005;162 (6) 1133- 1141
PubMed Link to Article
Bertrand  JMars  ABoyle  CBove  FYeargin-Allsopp  MDecoufle  P Prevalence of autism in a United States population: the Brick Township, New Jersey, investigation. Pediatrics 2001;108 (5) 1155- 1161
PubMed Link to Article
Dales  LHammer  SJSmith  NJ Time trends in autism and in MMR immunization coverage in California. JAMA 2001;285 (9) 1183- 1185
PubMed Link to Article
Landa  RJ Diagnosis of autism spectrum disorders in the first 3 years of life. Nat Clin Pract Neurol 2008;4 (3) 138- 147
PubMed Link to Article

Correspondence

CME
Also Meets CME requirements for:
Browse CME for all U.S. States
Accreditation Information
The American Medical Association is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The AMA designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 CreditTM per course. Physicians should claim only the credit commensurate with the extent of their participation in the activity. Physicians who complete the CME course and score at least 80% correct on the quiz are eligible for AMA PRA Category 1 CreditTM.
Note: You must get at least of the answers correct to pass this quiz.
Please click the checkbox indicating that you have read the full article in order to submit your answers.
Your answers have been saved for later.
You have not filled in all the answers to complete this quiz
The following questions were not answered:
Sorry, you have unsuccessfully completed this CME quiz with a score of
The following questions were not answered correctly:
Commitment to Change (optional):
Indicate what change(s) you will implement in your practice, if any, based on this CME course.
Your quiz results:
The filled radio buttons indicate your responses. The preferred responses are highlighted
For CME Course: A Proposed Model for Initial Assessment and Management of Acute Heart Failure Syndromes
Indicate what changes(s) you will implement in your practice, if any, based on this CME course.
Submit a Comment

Multimedia

Some tools below are only available to our subscribers or users with an online account.

Web of Science® Times Cited: 47

Related Content

Customize your page view by dragging & repositioning the boxes below.

Articles Related By Topic
Related Collections
PubMed Articles