Participants and procedure
The sample was drawn from the Menorcan population. All the health centers of Menorca participated: Health Centers of Mahón (Dalt Sant Joan), CS Es Castell, CS Ferreries, CS Es Banyer, CS Mercadal, CS Sant Lluís and Ciutadella (Canal Salat). Image 1 shows the proportion of participation by municipality and the number of participating subjects. The collaboration rate was higher in Canal Salat (Ciutadella), which corresponded to 75%, and CS Es Castell (70%), Dalt Sant Joan (Mahón), with 50%, and Es Banyer, with 45%. The lowest responders were those corresponding to CS Ferreries (10%), Es Mercadal (20%) and Sant Lluís (15%).
The sample size for an estimated maximum prevalence of 25% risk of having NDDs on the island of Menorca was 289 subjects to achieve a precision of +/- 5% with a 95% confidence interval and p=0.25. It was calculated according to the 2021 Registry (which refers to 01-01-21); the 5-year population (born in 2015) was 850, and the 6-year population (born in 2014) was 821. Therefore, to obtain a sample with adequate representativeness for this type of study (sampling errors between 3%-5%), 289 subjects would be necessary.
After passing the ethics committee of the Balearic Islands (CEIB) in December 2020, the sample was collected by pediatricians and nurses consecutively during the months of January, February and March 2021, the time necessary to obtain a representative sample size of 291 children. Parents of children who attended the 6-year checkup were offered to participate in the study, and subjects who agreed to participate were recruited. The investigator and collaborators evaluated the parents who agreed to collaborate in the study after signing the informed consent form. Security measures were taken to ensure the confidentiality of the data.
A total of 345 subjects were initially recruited through the pediatrician. Thirty-eight subjects were lost in this recruitment phase due to personal reasons and travel difficulties, so 307 children were finally evaluated in the first phase of the study. Of these 307 participants, the sample was finally reduced to 291, adding 16 losses that were due to incomplete assessments, lack of information and dropouts.
The study was carried out one year after the declaration of the COVID-19 pandemic situation, adopting all the required safety measures and wearing a mask during the evaluation, a fact that should be taken into account when interpreting the results.
The parents previously collected the questionnaires they had to fill out from their pediatrician's office and then handed them to the investigators on the day of the direct evaluation with the child.
Inclusion and exclusion criteria
Children who attended the consultation of the Child-Adolescent Health Program of the Primary Care Consultations in Menorca for the 6-year follow-up visit, which could be carried out from two months before turning 6 years old to one month before turning 7 years old, were included.
Children diagnosed with ND that had been detected at previous ages were not excluded, and reports were accepted if they came from accredited entities with specialized professionals. With children who had a diagnosis, the data were collected, and readministration of tests was avoided, completing the missing information and avoiding duplication.
All children under 5 years and 11 months and over 7 years at the time of the evaluation were excluded.
Study description
The study was divided into two phases. A first phase of general screening, which is the subject of this article, and a second phase of specific diagnosis in the Specialized Mental Health Unit with the altruistic collaboration of a psycho-pedagogical center and with the support of a clinical team trained in all the evaluation and research tools.
The sample was collected from routine well child consultations in primary care pediatrics according to the Child-Adolescent Health Program of the Balearic Islands (35). Subsequently, the families who decided to participate in the study were summoned by professionals trained in neurodevelopment and underwent an exhaustive general evaluation (of the child and parents separately) with different instruments that measured different areas and warning signs. Clinical data were collected through a Data Collection Notebook that included risk factors associated with NDs in the literature, such as prematurity, low birth weight, pre- and perinatal infections, medical history, parental age, and exposure to toxic substances. Instruments were included to evaluate alarm signals (shown in Fig. 1) that were used as screening to measure the risks of presenting NDs. Thus, children with or without a risk of presenting NDs were classified. During the assessment of the 291 subjects and families, 20 minutes were used for direct observation of the minor and approximately 30 minutes for the assessment of the parents.
Measures
The Minikid ADHD and TICS- Mini International Neuropsychiatric Interview for Children and Adolescents (36)
"The MINI-KID generates reliable and valid psychiatric diagnoses for children and adolescents" (36). It is a brief structured diagnostic interview for DSM-IV and ICD-10 psychiatric disorders in children and adolescents. We decided to use this interview with validity for DSM-IV to perform a quick and sensitive screening of the risk of presenting ADHD and tic disorders. As a limitation, we could say that it is based on the previous diagnostic manual, although we decided to use it because of its speed of administration and because the diagnostic criteria between the current and the previous manual do not differ significantly for initial screening. It determines the risk of presenting ADHD and tics in any of its presentations.
The Autism Spectrum Quotient (Children’s version, AQ- Child (37)
“This is an instrument that aims to quantify autistic traits in children aged 4–11 years” (37). It is a 50-question questionnaire that is answered by parents and detects autistic traits. We used the Spanish version taken from the Autism Research Center (Cambridge), which is designed to be administered to parents. A score above or equal to 75 points determines the risk of ASD.
The PROLEXIA, Battery for early detection and differential diagnosis of dyslexia (38)
The PROLEXIA, is a tool for the early detection of potential cases of dyslexia. The duration of the test was 30 minutes, and the correction was performed online. Since this is a new test, there is no bibliography to date, which urges us to carry out more studies to demonstrate its replicability and validity. The scores obtained from the corrections divide children according to the risk of presenting dyslexia into very low, low, moderate, high, moderate, high, and very high risk. From a risk score (RP) of 57 onward, a moderate risk is present and therefore already considered indicative of having dyslexia according to the test scoring (38). We used it because it is a measure in the Spanish language, and there are few instruments that measure the early risk of dyslexia.
Data collection notebook
This is a questionnaire prepared by our research team in which sociodemographic variables, medical and mental health history of both mother and child, lifestyle habits and general medical information are collected. It consists of 120 questions.
Statistical analysis
The test scores are indicative and were recorded in a database in Excel format. Regarding the analysis of the information collected, SPSS (Statistical package for the Social Sciences, v. 27), (39) was used, applying a univariate analysis (percentages of the variables) and an inferential analysis of contrast between two categorical variables (Chi-square test). Finally, a predictive analysis of the variable "comorbidity" (yes or no) was performed using bivariate logistic regression.
Theory/Calculation
Risk factors found and their relationship with ND risk
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). We could say that they comprise a broad spectrum where environmental pollutants (40), (41), preperinatal risk factors (42), (43) unhealthy lifestyle habits and disadvantaged environments (social exclusion, poverty, low purchasing power) (44), (45) are collected. All these factors could act at the epigenetic level, modifying gene expression and favoring the development of a given condition.
In our work, we collected those factors considered risk factors in the literature, and we have added others for study, such as sports practice, consumption of new technologies, breastfeeding, type of childbirth, adherence to the Mediterranean diet, educational and socioeconomic level, among others. In general, there is a higher risk of suffering one or more NDs in more disadvantaged families, χ2(2)= 19,728; p< 0,05.
Risk factors and comorbidity
We have evidenced statistically significant differences between the presence of comorbid NDs and its association with low socioeconomic resources (χ2(2)= 16,901; p< 0,01) and prematurity (χ2(1)= 4,376; p< 0,05). The practice of sports was significantly associated with a lower presence of comorbidities in NDs (χ2(1)= 7,139; p< 0,01). In the Sally and Annie test, we found statistically significant differences (χ2(1)= 6,213; p< 0,05) between comorbidity and test alteration.
Risk factors and risk of ASD
The risk of ASD did not obtain statistical significance in any of the variables studied.
Risk factors and language impairment
Low socioeconomic resources obtained a statistically significant relationship with the presence of language impairment (χ2(2)= 11,616; p< 0,01), the presence of prematurity (χ2(1)= 4,239; p< 0,05) and the impairment of the Sally and Annie test (χ2(1)= 10,756; p< 0,01).
Risk factors and dyslexia risk
The risk of dyslexia did not obtain statistical significance in any of the variables studied.
Risk factors and risk of ADHD
Low economic resources were statistically significantly related to a greater presence of ADHD (χ2(2)= 11,709; p< 0,05), and no statistical significance was obtained for the other variables studied (prematurity, diet, breastfeeding, low birth weight, sports practice or parental education level).
Risk factors and risk of Tics
The presence of breastfeeding was related to a lower risk of presenting tics, and this relationship was statistically significant (χ2(1)= 3,983; p< 0,05). The rest of the variables studied did not obtain significant differences.
Comorbidity predictors among NDs
A binary logistic regression model (5-step LR method) was established for the case of the variable "comorbidity of NDs" to detect those elements that can exert a predictive effect on this variable. In the case of the predictive model on "Risk of having comorbidity", it reported a Nagelkerke corrected R-squared of 0.51 and a Cox and Snell R square of 0.312, with a prognostic probability of 85% of the sample used. The best predictor found in the analysis corresponds to the fact of presenting high and very high risk of having dyslexia according to the PROLEXIA battery, as seen in Table 2.
Table 2. Comorbidity predictors among NDs.
|
B
|
Standard Error
|
Wald
|
df
|
p.
|
Odds Ratio
|
95% C.I. for Odds Ratio
|
Lower
|
Upper
|
Step 5
|
PROLEXIA_high_R(1)
|
-3,873
|
,563
|
47,360
|
1
|
,000
|
,021
|
,007
|
,063
|
Constant
|
,210
|
,246
|
,729
|
1
|
,393
|
1,233
|
|
|
It seems that the PROLEXIA allows, in a certain way and with a certain probability, to have other types of neurodevelopmental disorders associated with at least one disorder.