Sociodemographic and Clinical Characteristics of Children who Lose the Autism Spectrum Disorder Diagnosis

Abstract

Background:

The objective of this research is to present the demographic and clinical features of children diagnosed with autism spectrum disorder (ASD) in the past, who have achieved an optimal level of improvement. One of the recurrent inquiries of the parents of children with an ASD diagnosis pertains to the prognosis of their offspring and the potential for improvement. This study aims to investigate children who lose the ASD diagnosis (LAD).

Methods:

This study retrospectively reviewed the medical records of 1465 children and adolescents aged 0-18 who were diagnosed with ASD between December 2017 and June 2021, and followed up by a child and adolescent psychiatrist. The files of a total of 50 LAD patients were analyzed. In addition to sociodemographic information, the patient's educational background, prenatal and postnatal complications, developmental milestones, other medical conditions, medications used, symptoms at the time of ASD diagnosis, and developmental/mental assessment results were also examined.

Results:

A sample of 50 cases was obtained and analyzed. The majority of the sample consisted of males, with a median age of 2.45 (min 1.17-max 7) years at the time of receiving the ASD diagnosis, and 5.5 (min 2.25-max 11.3) years at the time of losing the diagnosis. There was a positive and moderately significant relationship between the age of receiving the diagnosis and the age of losing the diagnosis (r = 0.634, P < .001). Ninety-two percent of the sample received a mean of 2.29 ± 1.77 years of special education. Additionally, 26% of the sample still had an additional psychiatric diagnosis, with attention deficit hyperactivity disorder and speech sound disorder being the most common. Eighteen percent of the sample was found to be taking medication, primarily risperidone.

Conclusion:

This study shows that a subset of monitored children may lose their diagnosis, but further research to determine the clinical characteristics, symptomatology, and biological factors of this group of children will be more informative regarding optimal outcome processes.

Main Points

• The most frequently identified psychiatric diagnoses in children who lose the autism spectrum disorder (ASD) diagnosis were attention deficit hyperactivity disorder and speech sound disorder.

• The median age at initial diagnosis was 2.45 years, whereas the median age at the last diagnosis was 5.5 years.

• There was a positive and moderately significant relationship between the age of receiving the ASD diagnosis and the age of losing the diagnosis.

Introduction

In 1943, Leo Kanner defined autism as a condition characterized by a lack of emotional contact with others, repetitive behaviors, limited interests, and problems with social communication. The severity of the disorder varies among individuals with autism and can also differ in terms of the degree of impairment experienced.¹ The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) of the American Psychiatric Association provides standard criteria for diagnosing autism spectrum disorder (ASD) and also addresses important practical and conceptual issues for the clinician making the diagnosis.² When looking at the historical process of diagnostic classification, in DSM-5, autism disorder, Asperger disorder, Rett disorder, disintegrative disorder, and Pervasive Developmental Disorder-Not Otherwise Specified (PDD-NOS) are no longer separated, unlike in DSM-IV-TR.³ Studies on the prevalence of ASD show high variability worldwide. This variability in prevalence may be due to methodological differences in the studies conducted, as well as how ASD is diagnosed, the population being studied, and how case definition is made in less populated areas.⁴ Studies on the prevalence of ASD show high variability worldwide. In 2021, the Centers for Disease Control and Prevention (CDC) reported that, based on 2020 data, approximately 1 in 36 children in the United States is diagnosed with an ASD. Specifically, 4% of boys and 1% of girls were identified with ASD.⁵

When the recovery periods of children diagnosed with ASD were examined, it has been shown that early developmental and behavioral interventions are effective in increasing cognitive development, reducing core ASD symptoms, and enhancing social skills and functions.⁶ In reality, due to the heterogeneous nature of the disorder, there is a need for more personalized methods rather than trying to find the best treatment for ASD. A personalized treatment plan should be created by taking into account the child’s age, level of intelligence, social communication skills, executive functions, language level, repetitive behaviors, and the presence of any accompanying psychiatric disorders.⁷ The American Academy of Child and Adolescent Psychiatry recommends the simultaneous use of multiple treatment methods, such as behavioral interventions, speech and language therapy, and school interventions, for ASD.⁸ In addition to all these intervention methods, many children who are diagnosed with ASD are prescribed psychotropic drugs. According to a study in the United Kingdom, 29% of these children are prescribed psychotropic drugs.⁹ Early diagnosis and subsequent treatment are particularly important for children who have been diagnosed with ASD, as research has found that 47% of children who receive intense behavioral therapy can achieve normal intelligence and educational functioning. The concept of recovery was first proposed by Lovaas in 1987.¹⁰ In a study conducted by Mraz et al, it was reported that out of 35 children diagnosed with ASD, none retained the diagnosis during their second evaluations.¹¹ In children with ASD, some known positive predictors of improvement include higher intelligence level or lack of accompanying intellectual disability, early diagnosis and intense intervention, higher social skills prior to treatment, particularly early language and nonverbal functional skills, and a diagnosis of higher-functioning ASDs such as PDD-NOS or Asperger syndrome (as recognized by DSM-IV).¹²

The aim of this study is to investigate the symptomatology, clinical, and sociodemographic characteristics of children who lose the ASD diagnosis (LAD). This is an area of research that has received a lot of attention in recent years due to the increasing prevalence of ASD.

Material and Methods

This study has been approved by the Local Ethics Committee of Aydın Adnan Menderes University (Approval number: 2022/26 Date: 10.02.2022).

Between December 2017 and June 2021, a total of 1465 children and adolescents between the ages of 0 and 18 who were diagnosed with ASD according to DSM-IV-TR or DSM-5 and followed up or were diagnosed by other centers and confirmed by an expert and followed up at Aydın Adnan Menderes University Faculty of Medicine, Child and Adolescent Psychiatry Department Clinic (911 cases) and Muğla Training and Research Hospital, Child and Adolescent Psychiatry Clinic (554 cases) were retrospectively reviewed. The files of a total of 50 LAD patients were examined.

The study excluded patients newly diagnosed with ASD in the last 6 months (n = 137) and those who had only been examined once and not followed up (n = 78). The data were collected from patient files and prepared by the researchers using a case report form. The form included the patient’s sociodemographic characteristics, education and special education information, prenatal, birth and postnatal complications, developmental milestones (such as walking, single-word speech, toilet training), other medical conditions (psychiatric and physical), medications used, detailed assessments of symptoms at the time of ASD diagnosis, age of ASD diagnosis and age of losing the diagnosis, and developmental/mental assessment results (including any objective evaluation criteria). The Ankara Developmental Screening Inventory (AGTE) was used for the developmental assessment of children under 6 years old,¹³ while the WISC-R intelligence test was administered for the evaluation of children aged 6 to 18 years.^14,15 Childhood Autism Rating Scale (CARS)^16,17 was applied to each case by a psychiatrist at their final examination.

The criteria for LAD were defined as not meeting the DSM-5 criteria for ASD, having CARS scores are <30, not showing ASD clinical symptoms, not requiring special education for ASD core symptoms, and demonstrating age-appropriate functional abilities clinically.

Statistical Analysis

In the study, the data obtained were recorded in the Statistical Package for the Social Sciences (SPSS) 17.0 (SPSS Inc.; Chicago, IL, USA) program. Descriptive data were presented as number, percentage, mean, standard deviation, median, min, and max. The conformity of the data to the normal distribution was evaluated with the Shapiro-Wilk test. Comparison of age at diagnosis by gender was evaluated with the Mann-Whitney U test. The relationship between the age of receiving the diagnosis and the age of losing the diagnosis was analyzed using the Spearman correlation coefficient. Statistical significance was accepted as P < .05.

Results

The study reviewed the files of a total of 1465 children and adolescents with ASD in the archive system, and a total of 50 LAD children were retrospectively examined. The mean age of the cases was 78.64 ± 40.85 months (min = 18 months, max = 177 months), 28% were girls (n = 14), and 72% were boys (n = 36). The mean ages of the mothers and fathers were 36.31 ± 6.13 and 39.59 ± 6.34, respectively. The mean number of siblings of the cases was 1.32 ± 1.09. The sociodemographic characteristics of the cases are presented in Table 1

Table 1.

Sociodemographic Characteristics of the Cases (n = 50)

	n	%
Gender
Female	14	28
Male	36	72
Education status
No school	7	14
Pre-school	24	48
Primary school	15	30
Middle school	4	8
Parent relationship status
Married	49	98
Divorce	1	2
Mother working status
Working	16	32
Not working	34	68
Father working status
Working	49	98
Not working	1	2
Mother education level
Primary school	12	24
Middle school	5	10
High school	12	24
Bachelor’s degree	21	42
Father education level
Primary school	9	18
Middle school	4	8
High school	11	22
Bachelor’s degree	26	52
Mother’s physical disease*
Yes	6	12
No	44	88
Father’s physical disease ***
Yes	4	8
No	46	92
Mother’s psychiatric disease**
Yes	1	2
No	49	98
Father’s psychiatric disease ****
Yes	1	2
No	49	98
Consanguineous marriage
Yes	4	8
No	46	92

*Hypothyroidism (n = 3), epilepsy (n = 1), diabetes mellitus (n = 1), thalassemia minor (n = 1), **obsessive-compulsive disorder (n = 1), ***congenital hemolytic anemia (n = 1), thalassemia minor (n = 1), asthma (n = 1), muscle disease (n = 1), ****attention deficit and hyperactivity disorder.

The analysis showed that the median age of receiving an ASD diagnosis was 2.45 (min 1.17-max 7) years, and the median age of losing the diagnosis was 5.5 (min 2.25-max 11.3) years. The correlation analysis showed a positive and moderately significant relationship between the age of receiving the diagnosis and the age of losing the diagnosis (r = 0.634, P < .001). There was no statistically significant difference in the age of losing the diagnosis between genders (z = −0.48, P = .625; Table 2). Median time until losing the diagnosis was 33.5 months (min 8-max 105.6). The cases received mean of 2.29 ± 1.77 years of special education. The developmental and clinical characteristics of the cases are presented in Table 3. All the cases had normal intelligence, and 92% received special education. Twenty-six percent (n = 13) of the cases still had an additional psychiatric diagnosis, with the most common comorbidities being attention deficit hyperactivity disorder (ADHD) and speech sound disorder. Eighteen percent (n = 9) of the cases included in the study were using medication, with the most commonly used medication being risperidone. Clinical characteristics are shown in Figure 1.

Table 2.

Comparison of the Age of Losing the Diagnosis by Gender

Gender	n	Descriptive Statistics		Mann-Whitney U		P
		Median	Min-Max	U	Z
Female	14	5.75	(3-11)	229.500	−0.489	.625
Male	36	5	(2.25-11.3)

Table 3.

Developmental and Clinical Characteristics of the Cases (n = 50)

	n	%
Birth time
Full term	43	86
Preterm	7	14
Type of delivery
Normal	16	32
Caesarean section	34	68
Birth complication
No	46	92
Meconium aspiration	3	6
Asphyxia	1	2
Neonatal incubator
Yes	7	14
No	43	86
Jaundice
Yes	18	36
No	32	64
Phototherapy
Yes	13	26
No	37	74
Physical illness
Yes*	2	4
No	48	96
Febril seizure
Yes	3	6
No	47	94
Epilepsy
No	50	100
Pregnancy complication
No	36	72
Threatened miscarriage	9	18
Bleeding	1	2
Hypertensin	1	2
Hyperemesis gravidarum	1	2
Premature birth risk	2	4
Screen exposure
Yes	40	80
Intelligence
Normal	50	100
Special education (state-supported)
Educated	46	92
No	4	8
Psychiatric diagnosis
No	37	74
Yes	13	26
ADHD	7	53.8
Speech sound disorder	5	38.4
ADHD + SLD	1	7.7
Medication use
Yes**	9	18
	Mean ± SD
Birth weight (gr)	3200.2 ± 530.14
Breast milk intake duration (month)	13.32 ± 9.46
Age of first walking (month)	14.28 ± 4.26
First single word (month)	20.91 ± 8.83
Toilet training (year)	3.18 ± 0.83
	Median (min-max)
Receiving the ASD diagnosis (year)	2.45 (min 1.17-max 7)
Losing the ASD diagnosis (year)	5.5 (min 2.25-max 11.3)

* n=1 bronchitis, n=1 hypoxic ischemic encephalopathy, **Risperidone, short+long acting methylphenidate, fluoxetine

ADHD, attention deficit and hyperactivity disorder; SLD, specific learning disorder.

Figure 1.

The frequency of symptoms present at the time of autism spectrum disorder diagnosis.

Discussion

This study aimed to use a retrospective file review to determine if children diagnosed with ASD no longer meet the criteria for the diagnosis. ASD, which affects millions of people worldwide, is of global importance due to its prevalence and its significant impact on families. Early identification of the clinical signs of ASD is the first step in facilitating prompt assessment and referral for diagnosis. In many cases, the symptoms of ASD and other neurodevelopmental disorders are clearly evident in the early stages of development, yet the specificity of these symptoms remains uncertain.¹⁸

Questions from parents about the recovery and disappearance of ASD diagnoses in children are common. When significant improvements in a child’s social communication and limited or repetitive behaviors are noticed by caregivers and educators, a re-evaluation should be considered. Previous studies have shown that there can be improvement over time in the symptoms of children with ASD, and a small subset of patients may be included in an “optimal outcome” group.¹⁹ This group includes children who no longer meet the criteria for ASD. ASD can have a wide age range. The majority of parents of children with ASD report recognizing delays in their child’s language and speech development, as well as deficits in common attentional abilities and sensory sensitivities, by the age of 2. However, a study has revealed that there is an approximate delay of 20-60 months between when a family first suspects ASD and when a clinician makes a diagnosis.²⁰ In one study, the mean age of initial diagnosis for ASD was reported to be 2.39 ± 0.75 years, while the mean age of exclusion from the diagnosis was 5.11 ± 1.95 years.^21,22 In this study, the median age of initial diagnosis for ASD was found to be consistent with literature. The delayed diagnosis of ASD can result in negative outcomes, as it may impede the initiation of early intervention programs and appropriate treatment methods.²³ Some researchers have demonstrated that early intervention programs, specifically those that include specialized occupational therapy and preschool educational support, can play a significant role in the identification and amelioration of autism symptoms in young children, and improvement in overall behavior and learning abilities.^21,24 Early intervention programs in the field of ASD typically focus on improving several key areas, including joint attention, functional communication, participation in communicative interactions, expressive and receptive language, social interaction, and emotional regulation. Research has shown that these targeted interventions can lead to significant improvements in the overall functioning of children with ASD.²⁵ However, it is important to note that causality cannot be definitively established based solely on the results of this study. It is possible that the provision of appropriate interventions in a timely manner following diagnosis may play a role in this phenomenon, but further research is needed to fully understand the relationship. Additionally, it is worth considering the potential impact of increased family awareness and enhanced educational programs on the transition of children out of an ASD diagnosis. Nonetheless, more extensive studies are required to establish a causal link between intervention duration and outcomes for children with ASD.

ASD was initially viewed as an intractable neurodevelopmental disorder, and effective treatment options were limited. However, in recent years, studies on interventions for children with ASD have reported that some children can show effective changes as a result of these interventions.²⁶ In most countries, the increase in awareness of ASD has not resulted in a decrease in the age of diagnosis. Despite parents expressing concerns about their child’s development between 18 and 24 months, studies indicate that the average age of diagnosis in developed countries ranges from 38 to 120 months, as per a study that looked at a number of studies.^27,28 In this study, the median age at which the cases received an ASD diagnosis was 2.45 years. Blumberg et al. found the age of diagnosis for children with ASD with a previous diagnosis to be 5.79 years, and the age of ASD with an ongoing diagnosis to be 5.68 years.²⁹ It is interesting that in this study, there were cases that were diagnosed an average of 3 years after compared to this study and recovered. In this study, there was a positive correlation between the age of diagnosis and the age of losing the diagnosis. In other words, as the initial diagnosis age increased, the age at which the diagnosis was lost also increased. Receiving an ASD diagnosis at an early age and starting treatment interventions early have been reported as positive prognostic factors.^22,26,30 On the other hand, the diagnosis age data in the study reflects the patients’ condition at the time of their initial evaluation, and not all of them were screened for ASD at an early age. Therefore, those diagnosed at an older age and later losing the diagnosis may represent a group with milder symptomatology.

According to Wiggins et al, 4% of children initially diagnosed with ASD were later reclassified as non-autistic. Factors contributing to this change included an initial diagnosis at 30 months or younger, an onset unrelated to developmental regression, the presence of specific developmental delays, and enrollment in a special needs class other than ASD by the age of 8.³⁰ Helt et al conducted a review of research studies and reported that between 3% and 25% of individuals eventually no longer meet the criteria for an ASD diagnosis. They also identified several early indicators associated with better outcomes, including higher IQ, stronger receptive language abilities, improved imitation and motor skills, earlier diagnosis and intervention, and having a diagnosis of PDD-NOS rather than autistic disorder.²⁶ Blumberg et al. reported that around 13% of children who had ever been diagnosed with ASD no longer met the diagnostic criteria.²⁹ In the longitudinal follow-up study by Elias and Lord, 19% of participants with higher IQ were reported to have “lost” their ASD diagnosis by the age of 25.³¹ Similarly, a recent study by Harstad et al. found a high diagnosis loss rate of 37% among individuals diagnosed with ASD before the age of 3.³² Various factors influencing these differing rates of losing the diagnosis across studies include ethnic background, the location where the sample was collected, early diagnosis and access to education, and the intellectual level of the sample. In this study, the analysis of all reviewed files showed that the proportion of individuals who no longer met the diagnostic criteria was 3.4%. However, this rate should be interpreted with caution. Since this study is not longitudinal, the long-term outcomes of cases that had not lost their diagnosis at the time of the study remain unknown. Future studies comparing the rates of early age of diagnosis and late diagnosis and undiagnosed rates will contribute to the literature. It is believed that early detection of ASD diagnosis can make important contributions to the healing process of children. A Türkiye model has been proposed in this field with the related studies of diagnosis and monitoring.³³ The authors believe that each country should develop an effective model for early diagnosis of ASD, taking into account its own resources. Studies in this area have shown that early diagnosis is quite effective in the healing process.²² Verbal language development in children is normal in children who show typical development. After the acquisition of first words, children typically express themselves using simple sentences around the age of 3-4 and then progress to using more complex sentences as they continue to develop.³⁴ In the follow-up and determination of improvement in ASD, the development of language is extremely important. This is one of the most common problems seen in the majority of these children, and there are difficulties in the areas of verbal and nonverbal communication.³⁵ There are significant differences in the area of language development in children with ASD, but generally, it is observed that the development of language in these children is behind that of their peers.³⁶ In addition to speech delay, some children with ASD may not have any speech at all, while others may have only echolalic speech.³⁴ Difficulties in the areas of language and communication are one of the criteria for the diagnosis of ASD and a variable that determines the severity of the syndrome in children with ASD. Additionally, the language and communication disorders seen in children with ASD appear from the early stages of development and negatively affect many developmental areas.³⁷ In line with this information, this study also found that speech began late. Because language development is an important factor in the healing process, it is a data that should be taken into account in the follow-up of these children.

Another study, which examined positive prognostic factors in children with ASD, reported that children with high IQ scores obtained optimal results faster than other children.³⁸ This is consistent with various studies and also indicates that IQ is one of the strongest prognostic factors for autism outcomes in children.^29,31 In this study, it was determined that all cases had normal development/intelligence. In light of this information, it is thought that intelligence level may be an important feature in terms of improvement criteria for children with ASD.

According to studies in the field of ASD, generally the spectrum affects males more frequently with a ratio of 4 : 1.³⁹ In this study, the average age of the cases diagnosed and the rate of diagnosis also found that ASD was more prevalent in males than in females. There are many studies that suggest the diagnosis of ASD in girls is delayed.³⁹ Because the delay in care provided to these children can lead to more developmental difficulties, it is understood that the female phenotype of ASD needs to be studied.³⁹

Early interventions targeting the core behavioral and skill deficits associated with ASD are crucial for improvement. Currently, the American Academy of Pediatrics Council on Children with Disabilities recommends at least 25 hours of training per week for interventions related to ASD. This can include therapies such as Applied Behavioral Analysis (ABA), Speech and Language Therapy, Occupational Therapy, and Social Skills Training. These interventions can help children with ASD improve their communication, social interaction, and daily life skills.¹⁸ In this study, the authors found that the mean special education duration was 2.56 ± 1.08 years. Since the median time between diagnosis and recovery was 33.5 months, the years they received special education are not considered short for this sample group. 92% of the cases received special education. However, upon examining the 4 cases who did not receive special education, it was found that their families did not send them to state-supported special education institutions but instead arranged for individual education. Unfortunately, the authors do not have detailed data on the content of the special education programs or the duration of the lessons, which are important factors in prognosis. The effects of training and intervention programs on social communication skills in children with ASD should be further supported by more meta-analysis studies in the future.

It is also worth noting that children with ASD may have co-occurring disorders such as ADHD, depressive disorder, anxiety disorders, dysregulation/impulse-control/behavior disorders, emotional regulation disorders, self-harm, obsessions-compulsions, and sleep disorders.⁴⁰ Blumberg et al. found that the most common other current diagnoses in the previously diagnosed with ASD group between the ages of 6 and 17 were ADHD, developmental and intellectual disabilities, conduct/behavior problems, and anxiety.²⁹ Similarly, the authors found that the most common current psychiatric diagnosis was ADHD. In this study, the youngest age was 18 months and the oldest age was 177 months. Other psychiatric diagnoses may emerge in the long-term follow-up of cases with LAD.

For co-occurring conditions in individuals with ASD, psychotropic medications such as antipsychotics, antidepressants, antiepileptic drugs, and stimulants are often used for treatment. For example, in the United Kingdom, 7% of individuals with ASD have been prescribed antipsychotics.⁹ This study also found that the frequency of medication use was consistent with literature. Although some families have concerns about medication use, it can have a positive impact on improving ASD by reducing co-occurring conditions. It is important to note that medication use should always be discussed with a healthcare professional and tailored to the individual needs of the patient.⁴¹

Limitations

Having the same clinician responsible for diagnosing and monitoring all cases and using structured measurement tools could enhance the reliability of the results. In addition, the absence of using ADOS-2 or ADI-R instead of DSM-5 for the diagnosis and follow-up processes, as well as in the assessment of recovery, is considered a limitation. Additionally, the lack of comparison with the ASD group who did not lose diagnosis or typically developing children in terms of sociodemographic and clinical characteristics was insufficient to determine good prognostic factors. In this study, the authors only considered behavioral outcome measures, but the authors did not analyze biological markers, which have become a significant research topic in recent years in ASD studies.⁴² The literature on the rate of transition LAD among children and its association with sociodemographic and clinical characteristics is limited. This study aims to make a contribution to the existing body of knowledge in this area.

Conclusions

This study highlights the sociodemographic and clinical features of LAD children. The authors believe that it is important to show that at least some children with this neurodevelopmental disorder, which can lead to hopelessness and despair for families when first diagnosed, can improve. Further follow-up studies are needed to better understand the recovery process and to identify and support good prognostic factors for children with ASD.

Funding Statement

The authors declared that this study has received no financial support.

Data Availability Statement:

The data that support the findings of this study are available on request from the corresponding author.