Broader Autism Phenotype in Siblings of Children with ASD—A Review

Although less pronounced, social, cognitive, and personality characteristics associated with autism spectrum disorders (ASD) may be present in people who do not meet ASD diagnostic criteria, especially in first-degree relatives of individuals with ASD. Research on these characteristics, referred to as broader autism phenotype (BAP), provides valuable data on potential expressions of autism-specific deficits in the context of family relations. This paper offers a review of research on BAP in siblings of individuals with ASD, focusing on reports regarding social, communication, and cognitive deficits, published from 1993 to 2014. The studies are divided into two groups based on participants’ age: papers on preschool and older siblings of individuals with ASD; and publications on infants at risk for ASD. On the basis of this review, suggestions are offered for further research and its significance for our understanding of the genetic determinants of autism.


Introduction
Autism spectrum disorders (ASD) are neurodevelopmental disorders characterized by deficits in social communication and the presence of repetitive patterns of behavior, activity, and interests [1,2].

OPEN ACCESS
An important contributing factor in ASD etiology could be genetic liability, since the occurrence rate for ASD in siblings is on average 20-or 25-fold higher than in the general population [3][4][5].
The genetic mechanisms involved in the development of ASD are complex and heterogeneous [4,6,7]. This complexity is reflected by the variety of clinical characteristics of ASD, involving differences in the combination and expression of symptoms and severity of disorders in affected individuals. The heterogeneous, multifaceted nature of ASD is what drives research into the many possible expressions of autism that incorporate its typical deficits. The data obtained in that research may help us understand the genetic mechanisms that contribute to the development of autism-specific functional traits and atypical developmental trajectories. Moreover, this data may be useful in identifying genetic factors specific for various autism phenotypes, both in subjects with ASD and individuals from the general population.
A clear diagnostic conceptualization of autism still has yet to be established. It has been proposed that behavioral and cognitive characteristics of ASD, which include social, communication, and cognitive processes, rigid and persistent interests, and rigid and aloof personality traits, are continuously distributed throughout the general population [8]. These characteristics are likely to be more prevalent in first-degree relatives of individuals with ASD than in other groups [6,[9][10][11][12][13][14][15]. Frazier and colleagues [16] studied how caregivers reported autism symptoms among children in their charge diagnosed with ASD and their unaffected siblings. They obtained data supporting the view that ASD is best characterized as a category. However, they note that "these data do not discount the possibility of endophenotypes or subthreshold ASD symptoms in unaffected family members" [16]. Wheelwright and colleagues [17] point out that studying ASD in nonclinical samples may be valuable in the context of the variety of genetic factors that seem to be connected to ASD. The genetic research on autism may benefit from a more inclusive concept of genetic expression that comprises cognitive, social, and communication deficits, as well as personality traits [18,19]. Characteristics similar to autism but less severe are referred to as broader autism phenotype (BAP) [20]. The notion of BAP facilitated involvement in studies on ASD and related phenomena of a large number of individuals without a clinical diagnosis of ASD. Consequently, it became possible to apply methods of the quantitative genetics typically used in the studies of normally distributed characteristics [21]. Research on BAP may help identify more functionally homogeneous subgroups of affected individuals and pinpoint the genetic factors that contribute to the development of ASD symptoms or traits [22]. The identification of phenotype may, therefore, open up the possibility of hypothetically assigning responsibility to a candidate gene or chromosome region. It therefore seems that data obtained in studies on BAP may provide promising cues for more detailed hypotheses on the genetic background of ASD. These studies may also contribute to a better understanding of the lesser variants of autism [23]. Since the data collected to date are not uniform, their applicability to genetic studies remains limited, and indicating the phenotypes with a clear genetic connection is still a difficult question. This may be due to the fact that the BAP concept covers a range of cognitive and social abilities and personality traits. Precisely establishing which characteristics should be included in BAP is a somewhat controversial undertaking. The lack of standardized criteria for BAP complicates attempts to engage in a comparative review of research on the topic. The features generally recognized as the most typical are deficits in social functioning, pragmatic language difficulties, restricted, repetitive behaviors and interests, as well as cognitive deficits (with respect to theory of mind in rarely provide information on participants' age. Due to the neurodevelopmental nature of ASD and the specifics of the processes involved, the inclusion of that information would seem reasonable. In the light of increasing interest in studying infant siblings of children with ASD among researchers seeking early predictors of ASD, summarizing the data collected from siblings could provide valuable information. Moreover, the methods for collecting data from children, adolescents, and adults are different, which provides even greater encouragement to have a closer look at the studies on BAP in different groups of siblings of individuals with ASD. The present paper provides a review of research on BAP traits in siblings of individuals with ASD, wherein the body of research will be divided into two groups: preschool or older siblings, and at-risk infants. The two will be treated as distinct due to the different nature of studies on infant siblings of children with ASD compared to research on older siblings. The studies on infants and toddlers are quite often prospective, enrolling both healthy children and children later diagnosed with ASD. With repeated measures it is possible to follow the siblings' developmental trajectories, unlike in most available studies on children over three or four years of age. Studies on at-risk infants have only emerged in the last decade; earlier research focused on older children, and as such these studies will be discussed first. For the reasons mentioned above, we have not attempted to propose a systematic review as defined in the PRISMA statement [38]. The present paper rather involves the subjective perspective of the authors reflected by the choice of the reviewed papers. Table 1 presents a list of studies on broader autism phenotype in preschool or older siblings of individuals with ASD, published in the years 1993-2014. The list includes mainly studies that enrolled healthy siblings not affected by autism, although in a handful of them some participants were diagnosed with ASD or another disorder (i.e., delayed language development) during the course of the study [28,39,40]. We have rejected studies in which the results of siblings were pooled in statistical analysis with the results of parents or other relatives i.e., [41,42], which was sometimes a consequence of very small sibling subgroup size (like in [43]-only four siblings). Furthermore, publications included in the list meet the following criteria: (a) Peer-reviewed articles published in English; (b) Original studies; (c) Containing a control group; and (d) The social communication, language, and cognitive characteristics of autism in siblings of individuals with autism were the objects of study.

Research on Preschool-Age or Older Siblings
As shown in Table 1, studies that involve healthy siblings of individuals with ASD vary in design. They differ in terms of compared groups, sibling ages, sample sizes, analyzed aspects of functioning, as well as instruments and methods.     (see: Table 2

Comparison Groups
All papers listed in Table 1 included one or more comparison groups. One exception is the paper by Gerdts et al. [65], which is listed despite the fact that only ASD families were included in that study. However, the inclusion of multiplex and simplex ASD families provides valuable insight into the severity of BAP traits in families with more than one child with autism.
Analysis of the groups used in comparisons with siblings of individuals with ASD reveals several strategies adopted by researchers. The most commonly-used comparison group consists of healthy siblings of typically developing individuals, or simply typically developing individuals themselves with no family history of autism [48,54,60,63,64,70]. This strategy allows for a comparison of the development of ASD siblings with their typically developing peers from families with no autism.
In many studies, one of the comparison groups was composed of individuals diagnosed with ASD, usually high-functioning autism or Asperger syndrome [22,44,54,55,58,59,62,68,72]. This way, the results of siblings could be compared with those obtained from individuals with ASD, and it was possible to determine whether the instruments used actually identified these deficits. In some of these studies, the results achieved by siblings of individuals with ASD landed in the middle, between the scores of individuals with ASD and those achieved by typically developing controls i.e., [62,71]. This could indicate qualitatively similar traits or behavioral profiles of siblings and individuals with ASD, although the results of siblings are closer to the developmental norm. Some studies evaluated affected and unaffected siblings from the same family, i.e., [71]. This empirical approach has particular merit in light of the fact that siblings share on average 50% of genes with their affected brother or sister [68]. Verifying similarities between siblings in terms of autistic traits may offer new insights into genetic susceptibility to autism.
In several studies, typically older ones, the comparison group for siblings of individuals with ASD was composed of siblings of individuals with Down syndrome, i.e., [14,[45][46][47]69]. This was to control for the effects of a family member with a hereditary disability not associated with ASD on the development and functioning of siblings. Recently, however, the focus has been on the comorbidity of Autism Spectrum Disorders and Down syndrome. Lowenthal et al. [73] showed that the frequency of Pervasive Developmental Disorders in individuals with Down syndrome was 15.6%, including 5.58% for autism. These findings must be taken into account in the selection of siblings of individuals with Down syndrome for the group compared with the siblings of ASD individuals. The presence and severity of autistic traits and symptoms in probands with Down syndrome must also be controlled for. With the high incidence of ASD (see [74]), this should be a universally-observed principle applied in the selection of comparison groups in studies on BAP, not limited to Down syndrome participants.
In some research projects, comparison groups consisted of siblings of individuals with developmental delay [48], mental retardation of unknown etiology [40,51], or psychopathology unrelated to autism, such as ADHD, affective disorders, and anxiety disorders [6]. There is also an interesting group of studies that included siblings of children with such developmental problems as dyslexia [49,50] and specific language impairment [69]. Extensive research has recently been done on potential associations between these complex developmental disorders and autism as, similarly to autism, their incidence is greater in males, they involve a number of neurophysiological and neuroanatomical abnormalities, they are likely to have strong genetic components, and they encompass language and communication deficits [75,76].
There is no question that the type of comparison groups should be taken into consideration when interpreting results and forming generalizations regarding the presence or absence of specificity in the functioning of siblings of individuals with ASD. In the case of heritable conditions that share features with ASD (e.g., SLI), it is possible that the first-degree relatives of probands will present some characteristics that overlap with the BAP [23]. This may cause additional difficulties in comparisons and the isolation of the autism phenotype.

Participants' Age and Sample Size
In the majority of studies listed in Table 1, the participants' functioning was measured only once. The exceptions are two longitudinal studies by Gamliel and colleagues [56,57] that include children aged 4 months to 4.6 and 7 years, and studies by Warren et al. [64] and Malesa et al. [66], which were a continuation of a study by Yoder et al. [67]. As these studies provide information about preschool and school-age children, we decided to include them in Table 1. There is only one other study [39] in which participants were under 4 years of age. In all other works the siblings of individuals with ASD were aged 4 or older, and in a number of them the age range was quite large. In some papers it exceeded 10 years, i.e., [44,49,55], and could be as high as 40 years, i.e., [47,69].
The age of siblings at the time of the study is not always crucial; in designs where caregivers were asked to provide information on the child's development history, that variable is less significant i.e., [46,69]. However, in those studies that measure developmental levels across various domains or mastery of specific functions, a wide age range of participants may compromise the precision of inferences. It also cannot capture any developmental delays or irregularities that may be present during a certain period. However, knowledge of the specifics of that aspect of development could help in planning support for children in ASD families. A wide sibling age range makes it difficult to determine the level of that group's functioning, especially when coupled with a small sample size, which precludes statistical analysis on subgroups more homogenous in terms of age. In some studies the ASD siblings group counted under 20 participants i.e., [14,44,49,50,55,59,62,63]. There is, however, a clear trend towards increased sample size in more recent studies, i.e., [68,69]. In turn, small groups tend to be more homogenous, including in respect of participants' age. One illustration is the study by Warren et al. [64], which enrolled 40 siblings aged 4 to 7; another comes from research on preschool and early school children aged 2 to 9 years i.e., [39,61]. Ben-Yizhak et al. [60] assessed older children in a narrow age range from 9 to 12 years, while Koh et al. [59] and Holt et al. [22] evaluated adolescents aged 12-18. Nevertheless, the majority of studies are conducted on children and adolescents with wider age ranges: from 6-8 to 16-18 years, i.e., [6,40,54,71].
As we have previously mentioned, research on siblings who are of preschool age, or on older children and adolescents, suffers from a lack of longitudinal studies. In one of a handful of reports that provide such information, Gamliel et al. [56] stated that cognitive deficits originally present in siblings of children with ASD disappeared during the preschool period (age: 54 months), but-compared to controls-some differences in terms of receptive and expressive language abilities remained. The results indicating differences in language development were confirmed in the next phase of the study, conducted when the children reached 7 years of age. Data of that sort is particularly relevant for understanding the specifics of developmental processes in children from ASD-affected families. While not meeting the diagnostic criteria for the disorder, these children may experience difficulties due to atypical development. Some of these difficulties are resolved, but there are some that may compromise their adjustment at a given moment in life. Thus, longitudinal studies that enable us to follow developmental processes, as well as to identify their determinants and effective methods of developmental support, are particularly valuable.

Functioning Characteristics of Interest
The studies listed in Table 1 cover a wide variety of characteristics. In some, non-affected siblings were assessed for autism symptomatology with instruments used in diagnosis (Autism Diagnostic Observation Schedule, ADOS [77] or Autism Diagnostic Interview-Revised, ADI-R [78]) or based on developmental history established from medical history or interviews other than ADI-R i.e., [14,28,69]. The results of these studies suggest a greater incidence of deficits in at least one domain typical for ASD. However, their results are not fully consistent in terms of the domains affected by the deficits and the depth of the deficits in question.
Greater severity of autistic traits in the siblings of individuals with ASD compared to controls was shown in those studies that measured such traits using the Social Communication Questionnaire (SCQ [79]), Childhood Autism Rating Scale (CARS [80]), Autism Spectrum Quotient (AQ [8]) and Broader Phenotype Autism Symptom Scale (BPASS [32]) [6,62,65,79]. Another finding was that autistic traits, including less social interest, poorer conversational skills, higher rigidity, and intense interests, as well as less expressiveness in the use of nonverbal communication, are more pronounced in siblings from multiplex ASD families than in children from simplex ASD families [65]. Based on similar results obtained for social and communication domains of BAP in parents, Bernier et al. [81] concluded that different genetic transfer mechanisms may operate in families with one and families with several children with autism displaying many of these characteristics. It has been suggested that de novo mutations and non-inherited copy number variants may be particularly important risk factors in simplex ASD families, while in multiplex ASD families they are present to a much lesser degree [82]. However, this finding was not confirmed in all studies [83]. As noted by Spiker and colleagues [84], the variability of autistic phenotype expressed in multiplex families is relatively low. In their study on families with two or more children with autism, in 37 out of 44 participating families at least two children met the ADI diagnostic criteria. Out of all the children in these families, 71% met all of the criteria for autism diagnosis in ADI, 22% failed to meet any of the criteria, and 7% met one or more criteria without reaching all of the ADI cutpoints. The number of items classified as "uncertain" was small, and there was a clear-cut distinction between children affected and unaffected by autism. It should also be noted that studies involving siblings from simplex ASD families, siblings from multiplex ASD families, and controls are scarce. This reduces the possibility of identifying the scope of the differences between the three groups with respect to the BAP.
A substantial body of research into the functioning of ASD siblings concerns language and communication as part of the phenotype(s) of ASD. Findings vary, as do methods and instruments. In their grammatical analysis of spontaneous speech samples, Levy and Bar-Yuda [61] found no significant differences between ASD siblings and preschool controls. In some studies ASD siblings were found to be more likely to have language delays relative to their comparison groups [39,56,57], while in other papers such differences were not stated, i.e., [45]. A frequently encountered pattern of results is one in which the group of ASD siblings as a whole does not differ from controls, but a subgroup may be distinguished that demonstrates more pronounced BAP traits, including inferior language development, i.e., [39,60]. Interesting data comes from comparisons of ASD siblings with siblings of individuals with language disorders. Some researchers found less severe difficulties in language and communication in siblings of individuals with ASD than siblings of people with language disorders, i.e., [51], while others demonstrated that ASD siblings, similar to specific language impairment siblings, demonstrate higher levels of communication deficits in relation to siblings of individuals with Down syndrome [69].
A number of studies analyzed siblings' intellectual abilities [44,46,47,51,71]. This is an important issue since intellectual disability (ID) and autism are highly co-morbid [85]. It is estimated that ID is present in 50%-70% of all ASD cases [85]. Precise estimation of the co-morbidity of ASD and ID is complicated due to changing criteria, diagnostic procedures, and educational policy (e.g., providing more support to certain groups of pupils) as well as the methodological problems (screening tools properties) [86]. Nevertheless, it has been shown that the number of high-functioning individuals with ASD diagnosis has increased in recent years (see [74]). It is also noteworthy that IQ may be underestimated in people with high levels of autistic traits [87]. However, the interrelationships among the level of autistic traits and intellectual abilities are still far from being fully recognized, and the results of studies on IQ in siblings of individuals with ASD are equally difficult to generalize. In some of them, no differences were found between ASD siblings and controls in terms of IQ levels (e.g., [40,44,47]), while Fombonne et al. [46] and Pilowsky et al. [51] reported even higher verbal IQ in ASD siblings. Gizzonio et al. [71] found no differences between ASD siblings and controls in terms of Verbal IQ and Performance IQ, but reported a slight (non-significant) predominance of performance over verbal abilities. As the authors have suggested, this could indicate the presence of a certain cognitive profile common for individuals with ASD and their siblings, but statistically non-significant differences make this a very tentative conclusion. Fombonne et al. [46] identified a subgroup of participants who demonstrated BAP traits (referred to as BAP+) and had significantly lower IQ scores than the group of siblings non-affected with BAP. Similar results were reported by Chuthapisith et al. [39].
There have been numerous studies on autism-specific cognitive deficits: in theory of mind, central coherence, and executive function. Research on theory of mind, recognition of facial emotions, and face processing indicate that they are less developed in siblings of individuals with ASD compared to controls, i.e., [55,63,68,72]. By contrast, the results of a study using the "Reading the Mind in the Eyes" task were inconclusive. Dorris et al. [52] showed that siblings scored lower than controls, but Holt et al. [22] found no differences between adolescent ASD siblings and typically developing control adolescents. Similarly, no differences in theory of mind between siblings and controls were reported by Ozonoff [44]. The key in studies of this type is to take into account the age of participating siblings, which is often overlooked. It is also noteworthy that this area of study is still strongly affected by a lack of precise conceptual definitions and methodological scrutiny, which makes the discussion even more complex. It must be noted that the results of studies on deficits discussed above in individuals with ASD are mixed, and are far from conclusive. A detailed discussion of the definitional controversies relating to particular aspects of cognitive deficits typical for ASD, as well as a presentation of current opinion concerning theory of mind, emotion recognition, central coherence, and executive function in people with ASD, remain outside the scope of this paper.
A complex picture also emerges from research on executive function. Siblings of individuals with ASD were found to be no different from controls in terms of inhibitory control and processing speed [54], but performed worse in planning tasks [44,48]. Ozonoff and colleagues [44] found no differences in working memory and set-shifting, while Hughes et al. [48] concluded that a larger than expected proportion of siblings of individuals with autism demonstrated difficulties in set-shifting. In the study conducted by Pilowsky et al. [40], differences in executive function between ASD siblings and controls disappeared once two participants diagnosed with Pervasive Developmental Disorders were removed from the former group. It should be also mentioned that Happe, Briskman, and Frith [50] reported no differences in terms of weak central coherence. In recent years the field of research on siblings of individuals diagnosed with ASD has been dominated by studies focused on infants at high familial risk for ASD (see [35]). It is estimated that about 10%-20% of high-risk infants can be affected with subclinical ASD traits or other developmental problems [88]. It should be emphasized that the analyses in some of the studies on infants also included children who later received an ASD diagnosis, which is why these results should be approached with caution. Nevertheless, research on infant siblings of children with ASD can not only provide us with valuable information on early signs of autism, but also pave the way for investigation of BAP traits [35].

Research on High-Risk Infants
Most studies focus on infant siblings of older children diagnosed with ASD, aged from 4 to 24 or 36 months. Diagnosis of the ASD proband is usually confirmed with ADOS and ADI-R outcomes, and the age of the proband is not relevant. Infant siblings are usually participants in long-lasting longitudinal studies in which different aspects of infants' functioning are assessed, i.e., [56,77,89]. The control groups mainly consist of typically developing infants without familial risk for ASD. Some research projects perform comparison analyses between high-risk infants (HR, siblings of older children diagnosed with ASD) and low-risk infants (LR, infants without familial risk for ASD) [90][91][92][93][94], while others strive for interpretation of infants' functioning and test performance in the context of later ASD diagnoses or BAP characteristics [30,[95][96][97][98]. Many studies differ in the sizes of HR and LR infant samples involved in analysis (i.e., nine HR infants in [99] to 507 HR infants in [88]), making interpretation and comparison of results problematic.
A number of research projects on infant siblings are focused on early characteristics of ASD core symptoms, as well as overall risk for developing ASD, i.e., [88,94,[100][101][102]. Macari et al. [101] suggest that two thirds of infants at high risk for ASD experience some kind of developmental difficulties in the second year of life. A large study conducted by Messinger et al. [88] indicated elevated levels of autistic traits (higher mean ADOS severity scores) in HR infants compared to LR infants. Georgiades et al. [100] suggested that significantly more HR infants had exhibited higher levels of autistic-like traits than LR children, and at the age of three years these children had more social communication impairments, lower cognitive abilities, and more internalizing problems than typically developing children.
It should be noted that the emergence of some of the highlighted differences are probably caused by inclusion in analyses of high-risk infant siblings who later developed ASD. For instance, Rozga et al. [97] ascertained that high-risk infant siblings who later developed ASD exhibited lower rates of joint attention and requesting behavior than typically developing children. However, HR infant siblings without ASD outcomes did not differ in these characteristics from the control infants. Similarly, Bedford et al. [111] indicated that only the high-risk infants with later emerging socio-communication difficulties (ASD and atypical development) differed significantly from the control group in the gaze following task. What is more, high-risk infants without ASD outcome performed similarly to typically developing infants. Hutman et al. [95] also suggest social impairments only in the high-risk infants with later ASD diagnosis. Hudry et al. [104] observed reduced receptive vocabulary advantage in all HR infants by 14 months, but this difference was maintained through 24 months only in children with ASD outcome, while typically developing HR infants regained a more normative profile.
Some empirical data suggest the presence of deficits in quality of mother-infant interaction and differences in responses to separation events in high-risk infant siblings. For instance, Esposito et al. [112] found differences in cry sample patterns in HR toddlers compared to LR toddlers in expression of distress during the separation phase. However, Haltigan et al. [113] concluded that HR infants are not less likely to form secure affectional bonds with their caregivers than LR infants, but also mentioned that infant siblings of children with ASD are less distressed during separation and more reserved after reunion with a caregiver compared to typically developing children. Finally, a study conducted by Wan et al. [114] indicated that infant attentiveness to parent and positive affect were lower in the high-risk group later diagnosed with ASD. These characteristics, as well as dyadic mutuality, predicted a three-year ASD outcome.
A number of studies analyzed face processing and social visual fixation patterns as predictors of autistic traits in infants. Some HR siblings demonstrated diminished gaze to the mother's eyes relative to her mouth in the Still Face episode [93]. During the face processing task, LR infants demonstrated a preference for looking at the left side of the face (characteristic left visual field bias) that emerged by 11 months of age and was absent in HR infants at any age [115]. Hutman et al. [95] observed no difference in the proportion of attention to social stimuli or attention shifting during the play condition between HR and LR infants. However, children later diagnosed with ASD tended to continue looking at a toy during the distress condition despite the salience of social information. Overall, no group differences between HR and LR infants in gaze following behavior at either age was observed in the study by Bedford et al. [111]. Nevertheless, it should be noted that HR infants with later emerging socio-communication difficulties (ASD and atypical development) allocated less attention to a congruent object compared to typically developing HR siblings and LR controls.

Matching Criteria Measures/Instruments (Samples) Main Findings
Holmboe et al.        Some authors have posited alternative temperament development trajectories in HR infants, characteristic for early ASD symptoms or broader autism phenotype condition. HR children who were diagnosed with ASD at 36 months had a temperament profile marked by lower positive affect, higher negative affect, and difficulty controlling attention and behavior. This temperamental profile also distinguished HR children without ASD diagnosis at 36 months from LR children [116]. Rosario et al. [121] compared HR infants who were or were not later diagnosed with ASD and discovered that the temperament trajectories of children with ASD reflected increases over time in activity levels and decreasing adaptability and approach behaviors relative to high-risk typically developing (TD) children.
It is also predicted that HR infants will exhibit problems with executive functions, especially attention and inhibition deficits, which are considered to be associated with frontal cortex functioning impairments [109]. HR infants had difficulty disengaging attention and showed less selective inhibition than controls (less difference between interesting and boring trials); however, they demonstrated a larger decrease in looks to the distractors in the interesting trials than in the boring trials, whereas controls showed a similar decrease in the two trial types.
Studies investigating vocalization patterns in HR infants suggest that infants later diagnosed with ASD produce low rates of canonical babbling and low volubility in comparison with typically developing infants [123]. Iverson et al. [92] also found impaired vocal-motor developmental behaviors in HR infants. Infant siblings demonstrated attenuated patterns of change in rhythmic arm activity around the time of reduplicated babble onset, and were highly likely to exhibit delayed language development at 18 months.
It should be stressed that the analyzed studies have many limitations. As previously mentioned, some authors include in the analysis high-risk infant siblings who later developed ASD, which can lead to overestimation of the differences between high-risk and low-risk children, i.e., [95,97,111]. Furthermore, some studies do not report any clear matching criteria of the control group participants during the recruitment procedure-there is only a description provided of recruited participants in terms of different parameters (such as gender, age, ethnicity, etc.), i.e., [110,115,119]. It should be noted that these features often vary across studies, which can raise questions about the comparability and recurrence of obtained results with other research projects. There are also studies that do not have any control group at all, because of the aim of the study, i.e., comparing HR siblings who later developed ASD or did not [107,121]. These research projects provide information about differences within the HR group, but lack data from comparisons of these HR infants with typically developing children that do not have an ASD family background. There are also discrepancies between infants' age ranges in different studies. Some papers focus on describing the features of children at exactly the same age (i.e., 6, 12, or 24 months [102], or other time points [30,90,96]), while others collect data covering a fixed age range (18-27 months [94], below 3 years old [105], 12-23 months [103]). These distinct approaches produce different types of data and therefore variable quality of comparisons, which should be considered during an analysis of studies.
In summary, there are many studies suggesting the existence of a broad range of impairments in infant siblings of children with ASD. Some of the differences are probably characteristic for later ASD diagnosis, but it should be noted that infant siblings are also at a high risk of developing broader autism phenotype-like traits.

Summary and Conclusions
The review of studies on BAP traits in siblings of individuals with ASD shows that the issue has been extensively explored. This goes hand in hand with investigation of hereditary mechanisms involved in the etiology of autism spectrum disorders. Research on BAP provides important information about the varied expressions of autism-specific traits. Integration of the results of these studies presents a challenge due to differences in methods, control groups, age of participants, and other aspects of research protocols. Since the relationships between autistic traits and other individual characteristics in a general population, including first-degree relatives of individuals with ASD, have not yet been fully elaborated, this paper focuses on the phenotypic characteristics of individuals observed mainly in psychological studies.
A number of investigators reported cognitive deficits in siblings of children with ASD, which can be a part of ASD phenotype(s). These include emotion recognition tasks [52], lower levels of efficiency in planning, attention shifting, and verbal fluency [44,48]. Some studies indicated differences in social skills development [6,14,28,69]. Difficulties for siblings of children with ASD were also found with respect to communication and language. There are reports of histories of language delay and pragmatic language deficits in this group compared with the siblings of children with Down syndrome and typically developing children, i.e., [28,60,61,94,102].
Researchers have posited various BAP traits in siblings of individuals with autism as important components of the neurocognitive endophenotype for autism. For instance, Dalton et al. [55] suggested that social and emotional processing along with underlying neural circuitry constitute an important element of the endophenotype. Fiorentini and colleagues [63] point to face-coding mechanisms, emphasizing their role in the impairment of adaptive mechanisms, while Holt and colleagues [22] highlight the role of mentalizing deficits and atypical social cognition. The inclusion of language deficits in BAP remains controversial. Based on their findings from research on children aged 4-9 years, Levy and Bar-Yuda [61] question whether these deficits should indeed be included in BAP, while Gamliel et al. [56,57] consider them to be the key component in the difficulties encountered by preschool and early school individuals with ASD.
It should also be noted that some researchers found no differences between siblings of people with ASD and comparison groups with respect to BAP characteristics [40,45]. Importantly, in a number of studies subgroups demonstrating developmental deficits were distinguished from the group of healthy siblings of people with autism [39,65]. The clear distinction between ASD symptoms and BAP traits is difficult in studies involving siblings of individuals with ASD. This is especially true in studies on infant siblings, in which the risk of ASD rather than the BAP characteristics is the main concern. Prospective studies in this group of children should be continued, in order to track the developmental trajectories in these children at subsequent stages of their development.
It should be stressed that in the light of BAP research, the majority of brothers and sisters of individuals with ASD develop typically, without displaying autistic traits to a greater extent than the relevant control groups. However, traits in the siblings group are often widely dispersed, suggesting that there is much variation among these children in the course of developmental processes and their outcomes. Some data suggest that genetic susceptibility to autism may differ among families. It is likely to be higher in families with two or more children diagnosed with ASD. Other siblings in these families demonstrate more pronounced BAP traits [65,124]. Research on these families could provide valuable insights on genetic involvement in the development of autistic traits.
Longitudinal studies are especially useful, as they enable researchers to trace developmental dynamics of ASD siblings. Despite achieving the status of the gold standard in research on infants, more longitudinal studies on children over 36 months of age are needed.
The information available at present is insufficient to formulate final conclusions regarding BAP characteristics in siblings of people with ASD. There is no doubt, however, that current research is bringing us closer to an understanding of the genetic factors involved in the etiology of this group of disorders. The studies are also of fundamental importance due to the rising numbers of ASD diagnoses and the presence of autistic characteristics in the general population.