Stimați colegi,
Avem onoarea de a vă invita să participaţi la Cel de-al XXI-lea Congres SNPCAR şi A 43-a Conferinţă Naţională de Neurologie şi Psihiatrie a Copilului şi Adolescentului şi Profesiuni Asociate, cu participare internaţională, exclusiv online, în perioada 22-25 Septembrie 2021


Autor: Cristina Gianina Anghel Liana Kobylinska Alexandru Michire Andrei Nicolae Vasilescu Cristina Maria Nedelcu Ilinca Mihailescu Victorita Tudosie Mirela Militaru Florina Rad Iuliana Dobrescu
Distribuie pe:

From epigenetic theories to assortative mating [1], a lot of attention has gone into searching for specific traits in the parents of children with autism spectrum disorders as possible predicting factors [2, 3].

In 2001 and 2006 Baron-Cohen and his team published two articles [4, 5] that showed that the mothers of children with autism spectrum disorders had low empatization quotients(EQ), whereas the fathers had high systematization quotients (SQ). This discovery later became known as the theory of assortative mating [6,7], establishing the idea that the parents of children with autism spectrum disorders present, themselves, specific behavioral phenotypes, less prone to extensive social interactions [8]. These theories have raised a very interesting point regarding the influence of nurture versus nature [9] in the development of autism spectrum disorders. In other words, if the parents exhibit non-typical attachment behaviors in the early age, it is very likely for these to have an important influence of the cortical maturation, especially if the child has a genetic predisposition towards autistic traits.

Furthermore, it has been recently reported that parents of children with autism spectrum disorder had a slower development of language related skills and an uneven evolution of their performance in different educational domains in their school years, which correlated to the severity of their children’s symptoms [10]. Also, since the description of the autistic phenotypes by Asperger and Kanner, it has been noted that the first degree relatives of children with autism spectrum disorders exhibit peculiar patterns of social interactions, with a decreased interest in social relationships and a higher degree of social withdrawal [11,12]. This has been described as a broader autistic phenotype [13]. All these observations have, in fact, led to more questions than answers: if the broader autistic phenotype existed before, why do we only now see such an increase in the incidence of the pathology? What determines it? What would happen to all the children who are now diagnosed at a very young age without therapy? Is this a consequence of current parenting patterns or are there other environmental factors responsible? All these questions have been addressed in various research papers, however, they are still under debate [14,15].

Currently, the golden standard in the treatment of autism spectrum disorders is the personalized behavioral therapy [16]. Still, just the psychotherapy is not enough, this needs to be a permanent behavioral approach for the family [16]. In general, the children with the milder forms of ASD will make good progress on the cognitive and language domains, most of them maintaining difficulties in social interactions in further development stages [17]. This is why it is important for the parents to become sensitive to the emotional cues that they need to present to their children. If they, in themselves, have difficulties in perceiving emotions, it is crucial to work on these with the parents as well, for optimal results in the child’s therapy process [18].

In this context, the main goal of our study was to verify whether the parents of children with ASD have more difficulties recognizing facial emotions and attributing appropriate valences to these.


One hundred thirty-six participants aged 19 to 56 years were included in this study, based on informed consent and with local ethical approval. Out of these, 14 (one father) had children diagnosed with autism spectrum disorders, 16 (3 fathers) had children diagnosed with other mental health pathologies, 36 (12 fathers) had children that have never been diagnosed with any mental health disorder and 70 subjects did not have children (25 males).

The children’s diagnoses were established in the Child and Adolescent Department of the “Prof. Dr. Al. Obregia” Clinical Psychiatry Hospital in Bucharest, Romania, based on ICD10 criteria and appropriate diagnosis tests.

A trained actor was photographed while expressing the following emotions: fear, anger, happiness, surprise, disgust, contempt, sadness and a neutral face, according to Paul Ekman’s Facial Action Coding System [19]. A custom program was designed in OpenSesame 3.1.2 [20], using the photos. The test consisted of two phases. In the first phase, the subjects saw the images, one emotion at a time, on a computer screen. They were asked to press one key if they found the emotion in the picture positive, and another key if they found it negative, as fast as possible. The photos changed on key press. The attributed value and the latency of response were recorded. In the second part of the experiment, the subjects were asked to pick the suitable emotion for each photograph from a list.

All the subjects that failed to complete the respective part of the test due to technical errors (they pressed the keys to fast, or had reported that they not paying sufficient attention to the images on screen) were excluded from the analysis. In the end, 123 parents completed the first part of the task (94 parents of children without pathology and adults without children (NC/CNP), 15 parents of children with non-ASD psychiatric diagnoses (PD-nASD) and 14 parents of children with ASD (ASD)). The second part of the experiment was successfully completed by 94 subjects (66 NC/CNP, 15 PD-nASD, respectively 13 ASD).

The data were analyzed using SPSS 22.0, respectively Excel. The distribution of the data was checked using the Shapiro-Wilk test, and parametric tests were used for p>0.05. Post-hoc comparisons were employed for multiple variables testing in order to determine the specific differences between each two groups.


In the first part of the experiment, most of the subjects in the NC/CNP group recognized the emotions in accordance to their description from FACS. There were no significant differences between the parents with neurotypical children and the adults without children, so they were considered as a single group

Table 2 – The results of Kruskall-Wallis testing of the latencies of attributing a valence to each emotion. For each test run, df=2, n=94. LpTD =latency of response in parents of children with typical development, LpOP= latency of response in parents of children with other pathologies, LpASD=latency of response in parents of children with autism spectrum disorders.

for further analysis. Table 1 presents the percentage of recognized facial emotions in each group.

Given the fact that the visual inspection of the distribution of the data within the groups suggests that there might be differences between the subjects in the different groups, paired binomial testing was applied to check the significance level of these differences.

There were significant differences in recognizing the expression only in the case of the neutral face, (χ2(123,2) =20.31, p<0.001), because the subjects in the PD-nASD group had a lower rate of emotion recognition when compared to the other two groups (p<0,01 in both cases) (figure 1)

The latency in attributing a valence to certain emotions was significantly different between the groups, as verified by the Kruskall-Wallis test, with post-hoc paired comparisons (table 2).

The within group variation between the latencies in emotion valence appropriation was significant in all three groups, as verified by Friedman’s non-para-metric ANOVA test (for NC/CNP χ2(7, 66) =397.59, p<0.001, for pASD (χ2(7, 13) =87.02, p<0.001, for pPD-nASD (χ2(7, 15) =102.95, p<0.001).

The response latencies have varied according to the valence attributed to each emotion, such as the latency for emotions that had been attributed positive valences was significantly higher for the subjects in the NC/CNP group (Z=35407.5, n1(negative valence) =355, n2(positive valence) =173, p=0.04, median latency for emotions with negative valence = 3.9s,

IQR =2.8s, median latency for emotions with positive valence =4.5s, IQR=3.7s). The subjects in the other two groups there were no identifiable variations in response latency according to valence.

Table 1 – The percentage of emotions recognized by the members of each group in accordance to thei FACS definition

Figure 1. The number of subjects that have correctly/incorrectly recognized the emotion of surprise in each group.

The response latencies have varied according to the valence attributed to each emotion, such as the latency for emotions that had been attributed positive valences was significantly higher for the subjects in the NC/CNP group (Z=35407.5, n1(negative valence) =355, n2(positive valence) =173, p=0.04, median latency for emotions with negative valence = 3.9s,

IQR =2.8s, median latency for emotions with positive valence =4.5s, IQR=3.7s). The subjects in the other two groups there were no identifiable variations in response latency according to valence.


Based on these results, it may be concluded that parents of children with psychiatric disorders might interpret facial emotions differently. This needs to be further investigated, given the importance of education in interpreting facial expressions. One remarkable aspect that has been revealed by our results is that in the group of parents with neurotypical children/ subjects with no children, the latency of response to the task was higher for the pictures perceived as positive, than for those perceived as negative. This is a studied phenomenon in human behavior, where, in order to avoid the stress induced by non-pleasant stimuli, one tends to minimize the exposure to it, whereas maximizing the exposure to pleasant factors [21]. Interestingly, the parents of children with psychiatric pathologies, be they ASD, or other pathologies, have not exhibited this discriminative response in our study group. This may be due to an altered stress response in these parents, as previously described [22]. Moreover, the task performance, as measured in the response speed was higher in for the subjects with no children/with neurotypical children than in the other two study groups. This may be, at least, partly due to the perceived difficulties in recognizing the facial emotions, as hinted by the significant differences in recognition, particularly for the ambiguous emotions.

In social situations, this may make it difficult for the parents of these children to respond accurately, or to understand subtle ques, and this is an important aspect to consider in terms of the psychotherapeutic approach and case management.

These results substantiate the idea that, for a positive outcome of behavioral therapy in children with psychiatric disorders, especially those with neurodevelopmental disorders, a systemic approach should be considered. Moreover, it seems of the utmost importance to make sure that the interpretation of the emotional ques to which the parents of these children respond is accurate and in accordance with the standard interpretation.

Another important aspect to be considered is that of the subclinical phenotypes in parents of children with psychiatric disorders, which further raises the question of the intricate issue of the nature plus nurture background in the development of various psychiatric disorders. It has already been proven that the phenotypical characteristics of parents of children with autism spectrum disorders can influence the child’s outcome in therapy.

We conclude by emphasizing the need to pursue the subject and the need for a systemic therapeutic approach for the neurodevelopmental pathologies in children and adolescents.