An approach for prevention planning based on the prevalence and comorbidity of neurodevelopmental disorders in 6-year-old children receiving primary care consultations on the island of Menorca

According to the latest revised version of the Statistical Manual of Mental Disorders [1], which coincides with the DSM-5 [2], neurodevelopmental disorders (NDs) are those that include a clinical manifestation in almost all developmental domains. These manifestations include intellectual disability (ID), as well as those that affect more specific domains, such as attention-deficit/hyperactivity disorder (ADHD), autism spectrum disorder (ASD), communication disorders (CDs), specific learning disorders (SLDs, including difficulties in reading, writing and mathematics), and motor skill disorders (MDs, such as Tics, Tourette’s syndrome and stereotypic disorders), among others [3].

NDs usually begin in childhood, although most of them are chronic and persist for life. A new approach is committed to the inclusion of NDs within a heterogeneous and dimensional group, leaving behind the categorical classifications of the DSM-4th edition [4] and the International Statistical Classification of Diseases and Related Health Problems (ICD) [5]. The new edition of the ICD (ICD-11) unifies its criteria with those of the DSM-5 (2013). Finally, the revised DSM-5 (i.e., DSM-5-TR) was recently published in 2022.

To our knowledge, there are only a few studies in the scientific literature that measure the prevalence of NDs in minors according to DSM-5 criteria (2013). The prevalence rates reported in 2022 were as follows: ID, 0.63%; ADHD, 5–11%; ASD, 0.70–3%; SLDs, 3–10%; CDs, 1–3.42%; and MDs, 0.76–17% [3, 6‐11].

Among the available literature, prevalence studies and meta-analyses are the most common. The prevalence rates of the most common NDs were estimated as follows: ADHD, 7.9–9.5% [12, 13]; SLDs, 0.7–2.2% [12, 14, 15]; SLDs (including developmental dyslexia [DD]), 1.2–24% [16, 17]; and MDs, 1.4–19% [18, 19]. Furthermore, reported prevalence rates for various disorders within the same study did not include comorbidity rates between disorders [10].

In the United States, according to data published by the National Center for Health Statistics (NCHS) in 2015, it is estimated that 15% of children between the ages of 3 and 17 years are affected by NDs [20].

In a previous systematic review by our research team [3], we found that the global prevalence rate of NDs fluctuates globally between 4.70% in Scotland [8], 55.5% in Norway [9], and 88.50% in Japan [10].

These variations depended on methodological aspects, such as estimation procedures and sociocontextual phenomena. The criteria used by the different studies varied, and the processes used to measure the indicators were often not explicitly stated. In addition, it should be considered that the validity and reliability of the assessment instruments used were not explicitly mentioned and, in many cases, were nonexistent. In our study, we used a diagnostic approach to consider the problems of measurement and validity with respect to the data collection techniques used not only in our study but also in general.

There was also little direct evaluation and, consequently, little diagnostic certainty regarding the clinical populations in these studies. Furthermore, the studies often did not consider the complexity and comorbidities of the disorders; instead, symptoms or risks tended to be analyzed individually. Secondary sources are important as complementary resources for diagnosis, but prevalence studies with direct sources are lacking. Among the few studies where symptoms were directly assessed, in the study of Catalonia [7] and Norway [9], clinical diagnostic measures and questionnaires were used to assess symptoms and were completed by teachers. The Japan study [10] used surveys and questionnaires completed by parents and teachers. Most prevalence studies used indirect estimates such as health database records, which are likely to be less accurate. For example, not explicitly stating the overlap of comorbidities and results could be misleading and lead to overestimating the risk. However, we believe that this is a complex aspect from an empirical point of view. In our study, we have used the added criterion of diagnostic risk in the tests administered; specifically, we report the number of diagnoses for each participant.

NDs are usually underdiagnosed [21]. Therefore, children who have not been diagnosed are more likely to suffer from emotional and behavioral problems, low self-esteem, lower-than-expected academic performance, difficulties in social relationships, unemployment, delinquent behavior and functional impairment [22, 23]. The implementation of early detection and early intervention programs is essential [24].

NDs usually present as homotypic comorbidities, and it is rare that they occur in isolation. Despite this, there is a large body of literature on specific disorders, and these disorders have rarely been evaluated as a whole.

Multimorbidity among individuals with NDs is the norm, as determined in Japan [10], low-resource countries [6], Scotland [8], Spain [11], and Norway [9]. The prevalence of NDs seems to remain stable over time in different cultures, ages, ethnic groups [25] socioeconomic strata, types of communities (rural or urban), and religions [26]. Gender differences in NDs are consistent, with males being most affected by general psychiatric psychopathology, as reflected in studies in Scotland [8] and Denmark [27].

Males are more affected by NDs; 66.3% of the children included in a cross-sectional study in Norway [9] were male, and in a sequential cross-sectional study in Japan [28], a male:female ratio of 2.2:1 was reported. With regard to ADHD, male:female ratios of 4:1 and 2:1 were determined in a systematic review and meta-analysis in Spain [29, 30], generally corresponding with the reported ratios (3-2:1) in the systematic reviews by Sayal et al. (2018) and Faraone et al. (2021). Finally, a male:female ratio of 4.5:1 was reported in children with ASD in a retrospective analytical cohort study [31].

Considering the prevalence variations found in the different studies analyzed worldwide, we considered it necessary to carry out more studies in nonclinical samples and with direct evaluations that better reflect the reality of the population. In Catalonia [7] and the USA [32], a study was conducted with a school sample; in Galicia [11], Catalonia [31], Norway [9] and Brazil [33], a study was conducted with a clinical sample of children receiving specialized mental health services.

In countries with low socioeconomic resources [6], such as China [34] and Japan [10], a sample of the general population (rural and urban) was obtained. For this reason, we decided to conduct this study in a primary care sample, which we thought would more accurately reflect prevalence risk approximations than clinical samples. The age of 6 years was selected to insist on early detection and to demonstrate the possibility of providing an intervention through secondary prevention. These interventions can be performed at early ages when neuronal plasticity is still present, even if only the most severe forms of learning problems are usually detected at 6 years of age. Due to the heterogeneity of these disorders, the choice of 6 years of age limits our ability to diagnose the most severe cases of ASD or ID that would be detected before 3 years of age. Even so, we have observed an underdiagnosis of the more subtle forms of ASD in children with higher IQs. This is the first study in a school-age population where an exhaustive and direct assessment was carried out by professionals trained in neurodevelopment.

This is the first study of its type that has been carried out in a nonclinical population on the island of Menorca and with direct observations of the participants.

The general objective of the study was to obtain evidence on the prevalence and comorbidity of NDs to establish the fundamental elements for good health planning based on secondary prevention and early detection of subtle symptoms, which may go unnoticed if not explored in primary care services.

The specific objectives were as follows:

Finally, this study measured the risks of presenting any ND according to the DSM-5 (that is, ADHD, ASD, SLDs, MDs and CDs) and their possible comorbidities. ID, learning disabilities with difficulties in writing and mathematics and motor coordination problems were not assessed due to time and cost limitations.

There are numerous studies in the literature on the possible association between the environment and the development of NDs. Multiple and varied environmental factors have been studied (biological, social and economic factors). We could say that they comprise a broad spectrum of environmental pollutants [35, 36], pre/perinatal risk factors [37, 38] unhealthy lifestyle habits and disadvantaged environments (social exclusion, poverty, low purchasing power) [39, 40]. All these factors could act at the epigenetic level, modifying gene expression and favoring the development of a given condition.

Our work closely reflects and replicates prevalence results obtained in previous studies, and the risks of comorbidities and sex differences are consistent with previous findings. The measured risks of presenting with ADHD, ASD, dyslexia, language alterations or tics are close to those reported in the literature.

Sex (i.e., being male) and socioeconomic factors significantly influence the expression of certain conditions, such as major and minor language impairments and the presence of tics or hyperactivity, as previously demonstrated. As we have studied, there is a much higher risk of presenting with one or more NDs in disadvantaged environments, and there is a large body of literature in this regard. This was not observed for the risk of dyslexia, since it does not seem to be affected by year of schooling, socioeconomic factors, or other medical factors. This could increase the effect of genetics for future studies of this disorder; along the same lines, the genetic predisposition of ADHD [47], ASD [48] and language disorders [49] is also well known. Pennington and Lefly (2001) [50] demonstrate in their research that a total of 34% of their sample was at high risk for dyslexia. In the study, risk percentages were based on genetics and the inheritance of dyslexia in biological parents [50]. This percentage is similar to the that obtained in our research, with a total of 30.8%.

It is important to assess age and its variability in the literature, and most studies address wide age ranges. In contrast, in our study, 6-year-old children were analyzed, which could be considered a limitation. It should also be considered that in recent years, more global detection tools for NDs have been developed, and knowledge has been expanded, which may increase their prevalence, among other factors. It is challenging to determine the actual prevalence of each disorder, although with direct assessments, as in our case, we believe that some of the margin of error is reduced.

It is not clear whether there are differences in the prevalences calculated among clinical, school, and population samples, since the results found in the different studies analyzed are usually similar or rather disparate and contradictory, without observed patterns to establish a clear relationship. An example includes the study of Galicia [11], which included a clinical sample, paradoxically, with lower risks of presenting dyslexia (3.26%) or ADHD (5.35%) detected than those in other studies. In contrast, in a Norwegian study [9], which also included a clinical sample and direct assessments, prevalence rates similar to those obtained in our work were detected; in our study, 55.4% of the sample suffering from one or more NDs. A study in Japan [28] obtained figures for the prevalence of ASD similar to those calculated in our work (3.22% compared to 2.8%, respectively). Age did not seem to be an influencing factor in the rest of the studies and disorders. More comparative studies between population and clinical samples should be carried out to assess whether there is variability among different populations.

As a limitation of our work, it was difficult to delve into each disorder since the literature available for each disorder is very extensive, a fact that led us to generalize and synthesize results globally. It should be emphasized that we had an age limitation for learning problems since we could not make a diagnosis at 6 years of age, but we could provide early detection and subsequent follow-up. Our study was also limited in terms of psychomotor measures since, due to time constraints, we did not use clinical assessment instruments. Highlight the importance of considering gender differences in research, where the literature is closely linked to masculinity [51], which it considers more studies of equal proportion must be carried out to better define. In any case, in our sample, the proportion of men and women was similar. Because Menorca is an island with few private services, underdetection of NDs in girls may have occurred. Another limitation that should be highlighted is the size of the sample. Since this study included a population-based sample, which was small compared to other existing prevalence studies, prevalence estimates are less reliable (made through approximations by databases or health system registries). It should be considered that in our work, each child was evaluated by a trained professional using a direct assessment and that the risks detected would have a minimal margin of error. One drawback is that it is extremely expensive to carry out population-based studies with direct assessments, which is why prevalence studies generally resort to more imprecise estimates and larger sample sizes.

In addition, of the tests used to screen for the risk of presenting an ND, we used a recently developed test, the PROLEXIA battery, concluding that it seems to be a strong predictor of comorbidity and could be useful to determine other types of disorders associated with neurodevelopment. It should be added that other types of disorders that are not directly determined by the PROLEXIA tool include those that could contribute to this test, such as communication disorders or borderline cognitive levels, among others. We should consider a global test to determine the prevalence risk of different NDs. More research is needed to refine the prediction of what disorders this tool can predict apart from dyslexia.

We consider that the island of Menorca fulfills an indispensable requirement for epidemiological studies, which is the stability of the population over time. Considering the importance of the study of prevalence in child and adolescent psychopathology carried out on the Isle of Wight in 1976 by Rutter et al. [52], we believe that we reached a real approximation of the population of the island and thus are able to compare, share and replicate the results within the scientific community.

Our work defends the importance of direct evaluation of children and their families by trained professionals to extract prevalence figures as close to reality as possible. In this sense, the sample studied was population-based, avoiding the biases that would derive from the study of a clinical sample. There are few studies that have directly evaluated individuals and their families, and this is the first study worldwide to evaluate children at an early age (6 years) from a sample recruited from primary care services. Important point about our sample are that it was a poorly selected sample, it represented the risk group for NDs in Menorca and it contrasted with the results of clinical samples.

The results obtained highlight the importance of early detection, the establishment of specific programs aimed at early detection and the training of professionals who can perform specific evaluations due to the increased demand for pediatric consultations for neurodevelopmental difficulties. Despite the increase in the number of requests for primary care pediatric consultations, the presence of an early concern of parents about their children’s language delays and other neurodevelopmental difficulties that has not been previously attended, detected, or studied by professionals is noteworthy. The study provides the following advantages: ecological validity in the area studied, emphasis on improvements in health planning, coordination between primary and specialized care services and improved early detection of subtle signs of NDs.

The existence of an evident connection between comorbidity and the involvement of these cases in populations with greater socioeconomic disadvantage and the need to provide assistance to these families is also important.

In conclusion, we recommend promoting care from the earliest stages of development, i.e., preconception, pregnancy and the first years of life, to minimize possible risks and thus model epigenetically. Moreover, the promotion of healthy lifestyle habits, such as consuming a healthy diet, practicing sports, and reducing the use of new technologies from the age of 3 years, should be emphasized. Likewise, NDs will affect the subjects and their families throughout their lives, so these disorders should be considered to reduce school failure rates and, consequently, the costs to education, society, families and children’s self-esteem.