The study showed a statistically significant effect of three preconception risk factors for ASD in offspring, i.e., mental illness in the mother/mother’s family, maternal thyroid disease and the use of oral contraception.
There are reports on the correlation between ASD and parental psychiatric disorders. A family history of psychiatric illness was associated with higher odds of ASD in the index persons. An occurrence of ASD, intellectual disability, attention deficit/hyperactivity disorder, obsessive compulsive disorder, schizophrenia and other non-affective psychotic disorders, depression, bipolar disorder and personality disorder was found. The more closely related the affected family member was, the higher the odds were of ASD for the index person. At the same time, ASD without mental retardation was evidently associated with more disorders compare with ASD with an intellectual disability [19
]. The association between maternal mental illness and ASD observed in the present study is consistent with this study from the literature.
In an Australian study, compared with mothers with no previous psychiatric contact, those with any psychiatric contact were 2.5-times as likely to have a child with ASD without an intellectual disability and more than twice as likely to have a child with ASD with an associated intellectual disability [20
]. Swedish population studies showed a 2-fold higher prevalence of ASD among children of mothers with a psychiatric illness and fathers treated for schizophrenia. Parent diagnoses were based on an inpatient hospital diagnostic evaluation and included schizophrenia, other non-affective psychoses, affective disorders, neurotic and personality disorders and other nonpsychotic disorders, alcohol and drug addiction and abuse, and autism [21
]. Similarly, Lauritsen et al. in their study on the Danish population observed that the risk of ASD was twice as high among children whose mothers were diagnosed with a psychiatric disorder compared with children of mothers with no history of psychiatric illness [22
]. In addition, the risk of ASD associated with maternal antidepressant exposure during the pre-pregnancy period vs. all unexposed women appeared statistically significantly elevated and was similar in size to that of exposure during pregnancy [23
The authors did not find data in the literature on the relationship between the occurrence of maternal thyroid disease in the preconception period and the development of ASD in children. However, a statistically significantly increased risk for ASD was observed in the offspring in mothers with hypothyroidism in pregnancy. It was found that the odds of being a probable autistic child at the age of 6-years-old increased almost 4-fold when the mother had severe hypothyroidinemia (defined as 0.03 < TSH < 2.5mIU/L and fT4
< 10.99 pmol/L) in early gestation [24
]. Maternal hypothyroidism diagnosed and treated for the first time after the birth of the child increased the risk of ASD, whereas no significant association was seen for a maternal diagnosis and treatment prior to the birth of the child [25
]. As a risk factor for ASD, autoimmune thyroiditis was also reported in pregnant women. The prevalence of maternal anti-thyroid peroxidase antibodies was significantly increased in pregnancies giving rise to autism cases compared with controls. The odds of autism were increased by nearly 80% among the offspring of mothers who had positive anti-thyroid peroxidase antibodies during pregnancy, compared with mothers negative for this autoantibody. The measures of maternal thyroid hormones did not differ between these groups [26
]. Therefore, thyroid disorders in women who plan pregnancy should be effectively treated. More research is warranted to assess the impact of the disorders of thyroid metabolism in the preconception period.
In our study, a history of epilepsy, arterial hypertension and other cardiovascular diseases, diabetes, ophthalmic diseases and other chronic diseases did not have a statistically significant effect on the manifestation of ASD. There are no data in the literature on the impact of these diseases during the preconception period. However, studies have demonstrated an increased risk of ASD in children of mothers with hypertension and/or preeclampsia during pregnancy. The adjusted pooled results of the systematic review and meta-analysis indicated that exposure to maternal gestational hypertension was associated with 35% increased odds of ASD compared with nonexposure [27
]. Similarly, in the case of diabetes, studies have confirmed a statistically significant influence of maternal diabetes on the development of ASD in the offspring, however, without distinguishing preconception glycemic disorders [30
]. Other studies reported that an increased serum glucose level in a pregnant mother was not a risk factor for ASD [29
]. Panjwani et al. reported low maternal high-density lipoprotein cholesterol (HDL-C) and above-median maternal plasma branched-chain amino acid concentrations as risk factors [31
In our study, parental epilepsy was not a risk factor for ASD. However, in the population-based cohort study of Swedish participants, having a first-degree relative with cerebral palsy or epilepsy was associated with a 2-fold increase in the odds for ASD compared with those with unaffected first-degree relatives. The differences in our results may be explained by the fact that the researchers from this publication found a correlation between neurological diseases and ASD with mental retardation, while the group we studied was entirely in the intellectual norm [19
There are interesting findings related to contraception and its duration as risk factors for ASD. On an animal model, an exposure to progesterone during pregnancy induces ASD-like behavior in the offspring. The researchers used eight kinds of clinically relevant progestins for prenatal exposure in pregnant dams, and the offspring showed autism-like behavior. Therefore, many potential clinical progestin applications (including oral contraceptive pills and preterm birth drugs), may be risk factors for ASD. The mechanism was an estrogen receptor beta (ERβ) suppression in the amygdale [33
]. In a previous study, postmortem middle frontal gyrus tissues (13 ASD and 13 control subjects) were examined with the protein levels measurement and gene expression analysis. The gene expression analysis identified a 35% decrease in the ERβ mRNA expression in the middle frontal gyrus of ASD subjects. In addition, a 38% reduction in the aromatase (CYP19A1) mRNA expression was observed in ASD subjects. Significant decreases in ERco-activators were also found. These results provided the evidence of the dysregulation of ERβ and co-factors in the brains of subjects with ASD [34
]. Similarly, prenatal levonorgestrel exposure also induced autism-like behavior in the same mechanism, which was demonstrated in the animal model [35
]. However, the Chinese population-based case–control epidemiology study revealed that the use of progesterone (to prevent miscarriage and as a contraceptive at the time of conception or prenatal consumption of progestin-contaminated seafood during the first trimester of pregnancy) was strongly associated with the prevalence of ASD. Additionally, in vivo experiments in rats were conducted to further confirm the findings. The subsequent offspring of progesterone-fed dams showed autism-like behavior, which further demonstrated that prenatal progestin exposure may induce ASD [36
]. Moreover, a statistically significantly increased risk of ASD was found in children of mothers treated with progesterone due to conception problems. Progesterone exposure during the critical period of fetal life elevated the risk of ASD, possibly reflecting an epigenetic modification [37
]. On the contrary, Lyan et al. in their population study observed no correlation between the pre-gravid use of oral contraceptives and the risk of ASD in offspring. Additionally, they presented ambiguous results about the duration (in years) of oral contraceptives: in a retrospective study, the risk of ASD was statistically significantly associated with longer use, though in the prospective sub-group, the oral contraceptives’ duration association was reversed, with a longer duration among mothers of healthy children [38
]. Due to the limitations of the available data, it is difficult to draw clear conclusions. Further research is warranted to assess a potential adverse impact of oral contraception on the development of ASD.
The association between the use of assisted reproductive technology and ASD risk in offspring has been explored in several studies, but the results are still inconclusive. The meta-analysis indicated that the use of assisted reproductive technology may be associated with a higher risk of ASD in offspring. An analysis of the total 11 records (3 cohort studies and 8 case–control studies) revealed that the use of ART is associated with a higher percentage of ASD [39
]. However, some studies have not demonstrated an increased risk of developing ASD in children conceived using assisted reproductive techniques. In Spain, 231 children conceived by this technique and 208 children conceived naturally under the age of 3 were assessed. No differences were observed in the occurrence of neurodevelopmental disorders (global developmental delay, ASD or speech delay). Based on the analysis of the potential risk factors associated with assisted reproductive techniques, only a correlation between one type of technique (the transfer of frozen embryos) and speech delay was demonstrated, which had not been previously described [40
]. Similarly, the study of the Israeli population did not show the effect of in vitro fertilization on the development of ASD compared with the control group of naturally conceived children [37
]. Therefore, further prospective, large and high-quality studies are still required.
There are only a few reports on a possible relationship between miscarriage in a previous pregnancy and the manifestation of ASD. In our group of patients with ASD, no correlation was found between a history of miscarriage and the development of ASD in the subsequent children. However, the results of a German study indicated that miscarriage could be a specific risk factor for attention deficit hyperactivity disorder (ADHD) with ASD in children [41
The authors observed that parents in the reference group are more highly educated than in the study group. So far, it is hard to explain the reason for such an impact. One hypothesis suggests an influence of a healthier lifestyle, however, the authors did not ask about that. Conclusions from various studies assessing the influence of parental education on the risk of autism in offspring are inconclusive [42
]. Lee et al. presented findings similar to our results [43
]. On the contrary, some studies showed a positive correlation between higher education and ASD [44
]. Therefore, there is a need for further studies.
The strength of this work is a homogeneous study group of children with autism spectrum disorders without additional comorbidities, where a group called “pure autism” was obtained. This is the optimal group that will permit the creation of endophenotypes for further research, and at this stage allows the optimal selection of children for the control group (healthy children in the intellectual norm). Both groups include Caucasian children living in a similar environment, which is extremely important due to the participation in the formation of autism spectrum disorders of genetic and environmental factors. From the researchers’ perspective, the relatively young age of the respondents is important a sit allows the researchers to plan a prospective study in the future, which is a value in itself for population research.
The advantage is also the fact that survey was conducted personally by experienced practitioners and researchers in autistic centers known to children, in which they trust. Parents were not accompanied by children during the study, so they could freely answer the researcher.
A number of limitations should also be noted. The study group for such a common disease is small, where the study at this stage is a form of pilot study, and the conclusions drawn so far will improve the diagnostic tool which is the survey.
The questionnaire is an author’s own questionnaire—it was practically used for the first time. A detailed analysis of the data contained in the questionnaire revealed its disadvantages, including that the questions about psychiatric disorders in parents and in families of autistic children were not sufficiently developed—it would be better to enter closed questions about specific psychiatric diseases entered in the family tree.
Moreover, the data, including sensitive data regarding the family’s health status, were based on information provided by parents: no pregnancy record card or information regarding the health of the patient’s family members were analyzed. Unfortunately, bias due to the interviewers’ subjectivity cannot be ruled out on the responses to some of the questions related to pregnancy and/or family history.
Additionally, most of the children who participated in the study and control groups were unrelated, and while there were also several cases of siblings—due to the small study group—this aspect was not analyzed.
The authors in the presented work use only part of the questionnaire. There are still other questions to be analyzed, including correlations in the clinical picture of ASD or comorbidities in ASD in relation to prenatal factors. Further analysis will perhaps allow to determine endophenotypes and perform more detailed research.