713241

research-article2017
CCP0010.1177/1359104517713241Clinical Child Psychology and PsychiatryCrittenden

Article
Clinical Child Psychology
and Psychiatry
Formulating autism systemically: 2017, Vol. 22(3) 378­–389
© The Author(s) 2017
Part 1 – A review of the published Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
literature and case assessments DOI: 10.1177/1359104517713241
https://doi.org/10.1177/1359104517713241
journals.sagepub.com/home/ccp

Patricia M Crittenden
Family Relations Institute, USA

Abstract
Autism is a psychiatric disorder of unknown aetiology. In this article, the literature on genetic,
neurological, psychological, relational and cultural causes of autism is reviewed, beginning with the
2014 review of Crittenden, Dallos, Landini et al. (pp. 64–70) up to and including recent publications
in 2017. Some of the findings were unexpected; others led to new questions. The unexpected
findings were the minimal contribution of genes to autism, the extremely evident neurological
differences, the interpersonal quality of the psychological findings (that lacked evidence of parents’
behaviour), the relational evidence that mothers’ childhood trauma, perinatal stress and marital
stress increased the risk of autism, and the reciprocal relation between funding for treatment of
autism and diagnoses of autism. Notably, there was an abundance of genetic studies, numerous
neurological studies and only scattered psychological, relational and cultural studies, thus
rendering those findings speculative. The new questions included whether mothers used postural/
gestural signs to signal their children to maintain distance and whether mothers experienced
wariness of males as a result of childhood trauma, with their sons possibly experiencing gender
confusion. Following the literature review, a small archival set of video-recorded and transcribed
assessments of attachment of cases of autism were examined for evidence to corroborate or
refute the psychological and relational findings of the literature review. The findings were striking
in their support of mothers’ use of postural/gestural communication regarding distance, children’s
close attention to mothers’ bodily signals, without looking at mothers’ face, mothers’ greater
comfort when they approached their sons than when their sons approached them, one boy’s lack
of verbal self-representation and mothers’ childhood triangulation. These became hypotheses
regarding what to look for in Part 2 of this article, a prospective, 12-year case study.

Keywords
Autism, ASD, causes of autism, psychiatric diagnoses, triangulation

Introduction
This article updates and expands the review of the literature on the causes of autism in Crittenden,
Dallos, Landini, and Kozlowska (2014, pp. 64–70) and undergirds the prospective case study in

Corresponding author:
Patricia M Crittenden, Family Relations Institute, 9481 SW 147 St., Miami, FL 33176, USA.
Email: crittenden@patcrittenden.com
Crittenden 379

Part 2. I begin with a brief overview of the problems tied to psychiatric diagnosis, then review the
literature on autism and then briefly describe several cases of autism found in the Family Relations
Institute, Inc. archive of assessments of attachment. I conclude with a hypothesis regarding one
developmental pathway that might lead to autism. Although the review was comprehensive, stud-
ies were considered using the Dynamic-Maturational Model (DMM, Crittenden, 2015) of
Attachment and Adaptation for evidence of exposure to danger, psychological response to such
exposure and strategies to protect the self from danger. These aspects of reported findings are high-
lighted even when the researchers did not themselves highlight the finding. Because both psychi-
atric diagnoses and the causes of autism are highly controversial, I ask the reader to keep an open
and curious mind when facing unexpected information.

Psychiatric diagnoses: DSM and ICD

Psychiatric diagnoses
In the mid-20th century, the American Psychiatric Association (APA) sought a common language
to define psychiatric problems. Lacking evidence of aetiology, APA used the word ‘disorders’ (in
contrast to physical ‘diseases’ of known aetiology), with behavioural symptoms as the basis for
clustering. The Diagnostic and Statistical Manual of Psychiatric Disorders (APA, 2013) was mod-
elled on physical medicine and expected to use the emerging field of computer-based statistics to
refine its early efforts. The International Statistical Classification of Diseases and Related Health
Problems (World Health Organization (WHO), 1992) used similar diagnoses. Since then, psychi-
atric diagnoses have become so well recognized that they are understood by many professionals to
be biological conditions rather than constructs. Moreover, their application exceeded their intended
use, such that they are often required for treatment eligibility.
However, statistical support did not materialize. Symptoms overlapped among disorders, lead-
ing to changing diagnoses and co-morbidity. Lack of aetiology and pathogenesis remained prob-
lems. In the various revisions, the borders of the disorders expanded and contracted; whole
diagnoses were added, deleted or renamed, providing further evidence of the arbitrary quality of
diagnoses. Furthermore, therapists trained in psychoanalytic theory and family systems theory
tended not to use psychiatric diagnoses at all, preferring case formulations. Recently, alternative
approaches to a nosology for mental illness have been suggested (Cuthbert & Insel, 2013; Division
of Clinical Psychology, British Psychological Society (BPS), 2013).

Autism
The term autism refers to a focus on the self, yet it is parents who usually bring children to
attention, thus suggesting an interpersonal aspect to autism. Indeed, two of the three criteria for
autism (social withdrawal, limited social communication and repetitive/rigid behaviour) are
interpersonal. Later, the symptoms were seen to reflect a gradient: autism, Asperger’s syndrome
and autistic spectrum disorder (ASD); this indicated a continuum away from normal behaviour.
Like most disorders, autism has substantial overlap in symptoms with other psychiatric disor-
ders (especially ADHD, Alexithymia, Griffin, Lombardo, & Auyeung, 2016, Rao & Landa,
2014, schizophrenia, de Lacy & King, 2013, and gender dysphoria, Jacobs, Rachlin, Erickson-
Schroth, & Janssen, 2014) and considerable co-morbidity. The outcome is both a range of ASD
and also uncertainty in some cases regarding whether or not a child is best described as having
autism (King, Navot, Bernier, & Webb, 2014). For brevity, I use the word ‘autism’ to cover all
variants.
380 Clinical Child Psychology and Psychiatry 22(3)

Figure 1.  Systemic influences on the occurrence of autism.

Autism: understanding and misunderstanding

The number of children diagnosed with autism is increasing, with a five-fold increase in preva-
lence in the United Kingdom from the 1990s to a steady 3.8/1000 boys and 0.8/1000 girls in the
early 2000s (Taylor, Jick, & MacLaughlin, 2013). A similar, but more extreme pattern exists in the
United States, with 14.7/1000 cases of autism, a 30% increase in only 2years (Centers for Disease
Control and Prevention (CDC), 2014). It is not clear, however, that the number of children with
autism is increasing; instead, the increase may reflect the confluence of teachers’ need for chil-
dren to conform in classrooms, parents’ desire for a diagnosis to ‘explain’ their children’s behav-
iour and treatment availability. Boys are diagnosed with autism four times more frequently than
girls; the reasons are not yet known. Autism runs in families, although again the reason is not
known. In addition, funding devoted to autism is substantially greater than its prevalence war-
rants, suggesting that autism has some importance beyond the disorder itself. Further, although
‘everyone’ knows that autism is a genetic disorder, the evidence is not conclusive. Moreover,
studies on environmental insults (e.g. vaccines, foods and chemical toxins) are consistently nega-
tive (e.g. DeStefano, Price, & Weintraub, 2013) and studies of family contributions are largely
lacking. The problem, in other words, is that we have a condition about which we know too little
and, possibly, believe too much.
In this brief review of the literature, I take the review by Crittenden et al. (2014) as my starting
point and discuss more recent studies. Rather than discussing every publication, I prioritize meta-
analyses, major studies and methodologically sound studies that offer new perspectives. I cluster
the studies hierarchically (see Figure 1), with the assumption of systemic interaction across differ-
ent levels of representation.

Genetic contributions to autism

Studies of the genetic contribution to autism constitute the largest subset of studies, including those
that infer genetic contributions from comparisons of monozygotic and dizygotic twins and those
that assess genes directly. The inferential studies produce substantially higher estimates of herita-
bility. Studies that offer actual genetic evidence indicate that none of the many genes or gene
clusters showing a statistically significant association with autism were either limited to autism or
Crittenden 381

universal to autism. Furthermore, none was replicated in other samples. Most of the identified
genes and gene clusters affected brain development quite generally, with most genes accounting
for less than 1% of the variance in autism. More recent studies confirm these conclusions.

Inferential genetic studies.  Using inferential techniques, augmented by new statistical methods that
amplify statistical power, a large-scale study using a network of family relationships (rather than
twins) in Sweden’s national population registry concluded that 52% of the risk for autism lay in
common genes, with only 2.6% being attributable to de novo mutations, deletions, duplications,
and so on. (Gaugler et al., 2014). The authors note, however, that this finding required a very large
sample (1.6 million families). Furthermore, each family’s genetic risk was unique and the common
genetic factors were present in the non-affected comparison group. The authors concluded that
genetic risk was likely based on a myriad of small effects but did not specify what activates those
effects, such that the behavioural symptoms of autism become manifest, nor did they explain the
41% of cases that showed no genetic contribution to autism. Because genes were not assessed, the
study cannot differentiate between genetic effects from family-wide, systemic, multi-generational
processes that could yield the symptoms of autism.

Studies of genes.  In the most comprehensive genetic study to date, 65 genes were associated with
autism (Sanders et al., 2015). They affected the development and function of synapses and chro-
matin; in other words, the 65 genes affected the basic structuring of the developing brain. These
genes affected 10% of the 2591 families with a child with autism and differed among individuals.
Most of the genes were de novo mutations. Boys were 3–4 times more likely to show these muta-
tions than girls, but the mutations were the same for boys and girls. The lead author, Stephan Sand-
ers, concluded, ‘Autism is largely a genetic disorder, so it follows that finding the genes involved
in autism is a logical first step to understanding the biology of autism’, (Brooks, 2015, p. 1). Nev-
ertheless, as in earlier research, the effects of the 65 genes were on brain development, not specifi-
cally on autism. Moreover, 90% of the sample did not have these genes, that is, the genes were
neither universal, nor even typical for autism. In addition, there were no comparison families with-
out autism. Other studies with normative comparison groups have found the focal genes to be
present in substantial proportions of non-autism families; presumably that would be true for these
65 genes as well. Finally, there were no behavioural outcome measures so it is not known whether
or not these genes affected autistic-like behaviour. The point is that this study shows an idiosyn-
cratic difference in genes in 10% of a very large sample, thus providing both evidence of some
genetic input in some families with autism, and also shows stronger evidence of unknown causa-
tion in approximately 90% of families with autism.

Conclusions about the genetic contribution to autism.  The strongest finding in both the early and
recent literature is that autism and symptoms of autism run in families. There is some genetic
evidence, but none that can be generalized to the population of families or individuals with
autism. Instead, autism has high genotypic and phenotypic heterogeneity (Chang, Gilman,
Chiang, Sanders, & Vitkup, 2015). Moreover, the genetic differences are unique in each case
and usually involve genes that are common in the normative population. Importantly, the iden-
tified genes affect brain development and functioning (as opposed to autism itself). In addi-
tion, all the studies treat difference as detrimental, even though evolution is structured on
random beneficial variation and autism is associated with some positive characteristics (Sil-
berman, 2015). At present, the genes that differentiate autism from its absence and the
pathway(s) that transform genetic differences into neurological, psychological and interper-
sonal differences are unknown.
382 Clinical Child Psychology and Psychiatry 22(3)

Possibly, the studies have been misdirected. The question has been ‘what genes cause autism?’
Linear causation – of any sort – might miss the process. A better question might be ‘how are genes
involved in autism or autistic behaviour?’ This would include gene × environment interactions (includ-
ing epigenetic effects). Moreover, studies should recognize the equifinality of genes, that is, any
behaviour could be the outcome of many different genes and gene combinations (Chabris, Lee,
Cesarini, Benjamin, & Laibson, 2015). Instead, we should explore how nature and nurture work
together in autism (Strathearn, 2009). In conclusion, at present, the evidence does not justify consider-
ing autism a genetic disorder (although it is almost certainly a familial neurodevelopmental disorder).

Neurological aspects of autism

Neurological development reflects the interaction of genetic ‘instructions’ with experiential input;
particular ordered inputs are ‘expected’ by our evolved genome. If either genes or experience fall
outside of expectable limits, then anomalous neurological development may occur. The studies
included in the 2014 review showed clear neurological differences in the brains of children who
were and were not diagnosed with autism. The crucial neurological differences involved less prun-
ing of unused possible neural pathways (Tang et al., 2014) insufficient priming of frequently used
synaptic pathways (possibly as a function of low contingency, ambiguous stimuli or complex stim-
uli) and overproduction of inhibitory neurons (Mariani et al., 2015). The result was slower and less
efficient processing of information. Whether this is because of genetic anomalies or ‘unexpected’
experience or both cannot be determined from the structural differences themselves.

Neural structure.  Recent research confirms the brain differences, for example, some infants who
will later be diagnosed with autism have more extra-axial cerebrospinal fluid at 6 months than
other infants, and the excess is related to severity of gross motor symptomology (Shen et al., 2017).
There is continuing evidence that prenatal conditions, including maternal stress-induced antibodies
(e.g. Braunschweig et al., 2013), are associated with the symptoms of autism (albeit not necessarily
autism itself). For example, maternal depression or other psychiatric disorders accounted for most
of the risk for autism associated with prenatal use of anti-depressants (Mezzacappa et al., 2017).

Neural function. Functional neuroimaging demonstrated (1) a link between neurophysiological

inflexibility and core symptoms of autism (Uddin et al., 2015), (2) less cortical activation in indi-
viduals with autism when asked to think about faces than in normative individuals (Richey et al.,
2015), (3) less neural activation to facial images of self and others (Lu et al., 2015) and (4) predic-
tion of language competence from infancy (Lombardo et al., 2015).
In sum, ‘autism’ denotes differential brain structure and function, with specific aspects of neu-
rological difference having different impacts on development. It is not clear what causes these
differences, nor when they first appear. However, unlike studies of genes, neurological studies
consistently yield differences between individuals with and without autism.

Psychological aspects of autism

Eye contact and interpersonal signals.  Both children with autism and their siblings look at faces less
than children in families without autism. A particularly provocative finding comes from a study of
the infant siblings of older children with and without autism. At 2, 4 and 6 months, the infants were
shown videos of a research assistant. The data were whether the baby looked at the model’s face or
body. There were no differences at 2 months, slight differences at 4 months and clear differences at
6 months. Babies with older brothers with autism increasingly avoided eye contact and, instead,
Crittenden 383

looked increasingly at the model’s body. Babies with normative siblings showed the opposite trend
(Jones & Klin, 2013). Did the babies with brothers with autism discover more useful information
in bodies than in eyes? These researchers also found that 2-year-olds with autism did not avoid eye
contact, nor subtly shift their gaze to other locations; instead, they responded less to implicit social
signals for eye contact (Moriuchi, Klin, & Jones, 2017). Had 2-year-olds already learned that eyes
provided less useful information than bodies? Because these studies used videos of a research
assistant, we do not know whether or not the facial signals of mothers of children with autism
contain less information, more complex information or contradictory information as compared to
other mothers, nor whether the mothers use clearer body signals. Research that used each child’s
mother could address these questions. Others found that both poor language skills and limited
gestural/postural imitation at 2 years of age predicted language delay at 4 years, but only limited
gestural/postural imitation predicted deficits in social communication (Dohmen, Bishop, Chiat, &
Roy, 2016). Considering the findings of Jones and Klin, this highlights the importance of evaluat-
ing mothers’ gestural and postural signals to infants.

Understanding of feelings and emotion.  Children with autism find it difficult to understand others’
feelings and perspectives and do not label emotions and their aged peers without autism (Crit-
tenden et al., 2014). In addition, they lack neural representation of themselves when they think
about social interactions (Just et al., 2014); this finding suggests greater psychological attention to
others than to the self. Furthermore, autism leads to a range of outcomes by early adulthood, from
resolution, to lessened intensity, to no change (Fein et al., 2013). The finding that, without chang-
ing an individual’s genetic status, the condition of autism can be changed is encouraging – and
confirms the heterogeneity of autism.

Relational aspects of autism

Earlier research indicated that families with autism had less adaptability, lower marital satisfaction
and higher divorce rates than other families. Furthermore, parent–child interaction was more nor-
mative than different, with the differences split between mothers being too under- and over-stimu-
lating (see Crittenden et al., 2014, p. 67).

Child interactive behaviour.  Despite there being very little research on families with a child with
autism, it is clear that children with autism have less eye and physical contact with their mothers
and move more slowly than those without autism (e.g. Fusaro et al., 2014). Possibly because the
researchers assumed a genetic anomaly in the child, the studies did not report mothers’ interactive
behaviour. This leaves open the contribution, if any, of mothers to children’s behaviour.

Maternal interactive behaviour.  Although mothers of children with autism were as sensitive to their
child’s signals as other mothers (van IJzendoorn, et al., 2007), their interactions were less mutual,
had more negative affect and used more maternal control (Adamson, Bakeman, Deckner, & Nel-
son, 2012; Wan et al., 2012). These studies do not indicate the direction of effects, nor differentiate
sensitivity to child’s signals from contingent responsiveness to the signals. In a thought-provoking
finding, stimulus faces that varied in the number of autistic features were rated as more masculine
when they had more ‘autistic features’ (Scott, Jones, Kramer, & Ward, 2015); the raters were
research assistants. This raises the question of how mothers of children with autism see their sons.

Attachment. Attachment reflects the protective function of relationships; pattern of attachment

describes the strategy used to elicit protection and comfort from their attachment figures when they
384 Clinical Child Psychology and Psychiatry 22(3)

feel threatened. Protective strategies are also used by parents to protect their children. Children
with autism show atypical patterns of attachment (Crittenden et al., 2014). In the Coventry Grid,
children considered to have ‘autism’ (as opposed to ‘attachment disorder’) show a dispassionate,
‘strongly cognitive approach to the world’, whereas those considered to have attachment disorder
have ‘a strongly emotional approach’ (Moran, 2010, p. 50). In either Ainsworth’s ABC model of
individual differences in attachment (Ainsworth et al, 1978) or the expanded DMM (Crittenden,
2015), children who inhibit display of feelings and seek cognitive predictability would be consid-
ered to use a Type A ‘avoidant/dismissing’ strategy, whereas children who emphasize displayed
feelings would be considered to use a Type C strategy. That is, the Coventry Grid aligns autism
with a Type A inhibitory attachment strategy. Type A is also associated with attention to others and
dismissing of the self; this fits the finding of an absence of self-representation in children with
autism. Notably, a Type A strategy is used by many people without autism; the question is whether
people with autism consistently use an extreme Type A protective strategy.

Family stress and maternal childhood trauma.  In the 2014 review, families of children with autism had
more stress, less cohesion and fewer social supports than other families, but these findings followed
the diagnosis of autism and might reflect the stress of living with a child with autism (e.g. Markowitz
et al., 2016). Mothers and fathers of children with autism who used dismissing coping strategies (i.e.
‘active avoidance’ or ‘religious/denial’) were more stressed, anxious and depressed than those who
used ‘positive coping’ (Hastings, Kovshoff, Ward, Espinosa, & Remington, 2005). In a large-scale,
prospective study, mothers of children who would later be diagnosed with autism had often experi-
enced childhood abuse, with combined physical/emotional/sexual abuse yielding a risk ratio of 3.7 for
autism (Roberts et al., 2013). Mothers of children with autism have reported troubled relationships
with their parents (Oppenheim, Koren-Karie, Dolev, & Yirmiya, 2009; Seskin et al., 2010). These find-
ings suggest both that childhood mistreatment increases risk of a woman’s sons having autism and also
that this risk may be mediated by the mothers’ coping strategies, with avoidant/denial strategies
increasing the risk of autism. In a small sample, most women with late diagnosed autism had experi-
enced child sexual abuse. It is not known whether there is a causal relation between sexual abuse and
autism and, if there is, which is the direction of effects (Bargiela, Steward, & Mandy, 2016).

The effects of treatment on autism and autistic symptoms.  Treatment studies suggest another window
on autism by identifying conditions that can reduce the symptoms of autism. A review of five meta-
analyses of treatment indicated that (1) behavioural techniques reduced the symptoms of autism,
(2) treatment was more effective for younger children, (3) treatment delivered by children’s parents
was most effective and (4) not all children responded to treatment (Reichow, 2012). Reichow con-
cluded that greater attention should be paid to child characteristics, aspects of the treatment and
therapist training. There was no mention of parental characteristics or skills, even though parental
involvement was associated with more successful treatment and more enduring effects. Although
highly contingent interactions reduce young children’s symptoms and increase their learning (e.g.
Green et al., 2015), as children with autism become older, more variability is needed or there is a
risk of increasing children’s psychological rigidity (Harris et al., 2015).

Conclusions and questions regarding families with a child with autism.  These findings suggest that
mother–child interaction in cases of autism is more normative than not, but the differences high-
light the developmental role of clarity and contingency, with simplicity needed early on and
increasing variability and complexity later. Furthermore, mothers’ experience of childhood or mar-
ital danger or both might contribute to the occurrence of autism through reducing maternal contin-
gency. Again, there was heterogeneity among samples.
Crittenden 385

Contextual/cultural influences on autism

After the mother-blaming work in the 1950s and 1960s, family studies almost disappeared and
research was directed towards extra-familial factors. Recent studies offer glimpses into mothers’
childhood history and current stressful relationships in some families with autism. Possibly due to
successful parental advocacy, a disproportionate amount of mental health resources have been
directed to autism, as more children are being diagnosed with autism and parents are seeking the
diagnosis more often than in the past. Possibly a circular process has been created in which psy-
chiatrists respond to parental concerns with a diagnosis that has available services.

Summative ideas
When combining findings across the levels from genes to cultural context, several ideas occur
repeatedly: (1) the role in early infancy of developmentally attuned maternal contingency and clar-
ity in promoting brain pruning and priming; (2) the association of neural, psychological and inter-
personal inhibition with autism; (3) the possibility that both children with autism and their mothers
use inhibitory Type A protective strategies focused on the well-being of the attachment figure; (4)
the lack of self-awareness of children with autism and (5) heterogeneity across all levels, such that
subgroups may exist within the diagnosis of autism. In addition, there are several new leads to
pursue: (1) children’s tracking of mothers’ bodily gestures, (2) the connection between mothers’
childhood abuse and their protective strategies in adulthood, (3) the relation of mother’s pre-exist-
ing psychiatric disorder with autism and (4) the importance of mothers in delivering treatment.
Finally, the role of fathers and families in autism is almost totally unexplored.
None of these findings suggest singular or linear causation of autism. On the contrary, they sug-
gest systemic processes in which conditions interact with other conditions yielding an array of
outcomes, sometimes including autism, from any given starting point. This review suggests that
there is unlikely to be a single, direct cause of autism. Neither does the cause appear to be multi-
factorial, that is, additive. Instead, a set of systemic processes, occurring at many levels of func-
tioning, in developmental sequence and across more than one generation, appear to be involved.
Systemic studies of autism might yield universal findings about diversity, development and risk.

Clinical observations from the Family Relations Institute archives

With these findings in mind, I gathered all the assessments of children with autism and their moth-
ers in the Family Relations Institute archive of DMM assessments of attachment. I viewed the
video-recorded procedures looking for children’s attention to mothers’ bodies and gestures, ges-
tural communication and imitation, and evidence of children’s connection (or lack thereof) to their
mothers. My observations were striking: boys with autism (1) were exquisitely attuned to their
mothers’ behaviour, (2) circled their mothers but turned away following very subtle bodily gestures
of maternal discomfort (e.g. a raised, blocking foot, a briefly turned-away face) and (3) became
immersed in rote activities before and after these aborted approaches. The rote activities may have
functioned to disguise the children’s distress, to create child-controlled contingencies or to focus
the child’s attention elsewhere – or all of these. The children’s protective strategies were classified
as inhibitory Type A, with elements of both caregiving to the mother and strong self-reliance.
The mothers (1) had closed body positions (that often protected their genitals), (2) did not mir-
ror their sons’ movements, (3) were uncomfortable with approaches from their sons, (4) were warm
when they approached their sons, (5) eluded eye contact subtly with positive facial expressions and
(6) expressed genuine pleasure in their sons and concern about their development.
386 Clinical Child Psychology and Psychiatry 22(3)

I viewed the transcribed interview assessments for evidence of Type A strategies, self-negating
children and mothers’ childhood trauma, together with use of denial to cope in adulthood. During
his School-aged Assessment of Attachment, ‘Tomm’ was too agitated to tell 7 of the 14 requested
stories; he omitted those about isolation, rejection and bullying, and told stories about family con-
flict and maternal illness. As the dangers in the requested topics increased, Tomm became more
agitated. He used a compulsively compliant (A4) self-protective strategy, with disruptive intru-
sions of forbidden negative affect and denied psychological trauma for parental conflict and ill-
ness. His mother reported a history of parental conflict, her mother’s infidelity and her grandfather’s
imprisonment for child sexual abuse, saying she had never before told this.
The Adult Attachment Interviews of mothers of children with autism indicated complex childhood
traumas that, even in adulthood, the mothers could not relate with clarity. In addition, they used either
a Type A strategy of inhibition, denial and sometimes delusional idealization of their abusive parent
or a combination of Type A/C strategy, together with triangulation by their mothers against their
fathers. These observations suggest a multi-generational, interpersonal component to autism.

A summative hypothesis regarding some cases of autism

Keeping the heterogeneity of autism in mind, this review points to possible interpersonal and devel-
opmental processes that might affect some families with autism. The process begins with mothers
who had been abused in their own childhood but tried in adulthood to protect themselves and their
children by ‘forgetting’ these events. Because much abuse, especially sexual abuse, involves loved
fathers or father substitutes, some of these mothers might react to their sons, and not their daughters,
with interpersonal confusion and ambiguity. As a consequence, they might provide their infant sons
non-predictive facial information and predictive gestural/postural information; furthermore, they
might seek distance when their sons approached them but feel comfortable when they initiated con-
tact. They might both love their infant sons as unique persons and also reject them for an intrinsic
characteristic, such as male gender, associated with psychological trauma from childhood abuse. Of
course, these discrepancies would be enacted without the mothers’ awareness.
These conditions might affect the boys’ early neural development, resulting in less pruning and
priming and, therefore, less adaptive behaviour. If in addition, the boys had idiosyncratic genetic
anomalies or perinatal insults, the effect might be augmented. Later, they might experience gender
confusion. Together, these might create complex, reciprocal and multi-generational systemic pro-
cesses that children, parents and professionals could not easily identify, nor tie to autism. Of course,
given the equifinality of behaviour, this process could occur with other implicit characteristics,
including some that would involve girls or fathers. Notably, this hypothesis identifies both moth-
ers’ genuine love and concern for their sons and also the obscured process by which their own
childhood victimization, by a loved and feared family member, might result in retreat from their
sons’ approaches. Such complex, multi-generational and non-conscious processes have no begin-
ning and no culprit. This refutes the idea of ‘refrigerator mothers’ without passing the blame to the
men who harmed them as children; they too have histories that explain their behaviour. Instead of
seeking single, linear causes and assigning blame, systemic processes that include inexplicit infor-
mation should be sought. Understanding such processes could lead to interventions that would
benefit all family members.
It is important to note that such a systemic hypothesis lowers the probability of the outcome
(autism) with the inclusion of each new component. Put another way, each component is relatively
frequent in the population; for example, using a Type A strategy is frequent and normative, being
sexually abused in childhood occurs for about a fifth of women and so on. But psychological denial
of sexual abuse is infrequent. Combined in developmental order, the set of conditions described here
Crittenden 387

would be very infrequent – as is autism. The point is that these conditions are not risk markers for
autism. It is only the rare event of their systemic and developmental co-occurrence that indicates risk.
Of course, this is all speculative. It is consistent with the existing research but requires a new
generation of systemic research to accept, reject or modify the hypothesis and to develop additional
hypotheses for other pathways to the diagnosis of autism. Before a huge investment in methodo-
logically sound research is made, case studies can provide evidence regarding the probability of
significant results. Part 2 of this article offers one such study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publi-
cation of this article.

Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.

References
Adamson, L. B., Bakeman, R., Deckner, D. F., & Nelson, P. B. (2012). Rating parent-child interactions: Joint
engagement, communication dynamics, and shared topics in autism, down syndrome, and typical devel-
opment. Journal of Autism and Developmental Disorders, 42, 2622–2635.
Ainsworth, M., Blehar, M., Waters, E& Wall, S. (1978). Patterns of attachment: A psychological study of the
strange situation. Hillsdale, NJ: Erlbaum.
American Psychiatric Association. (2013). DSM-5: Diagnostic and statistical manual of mental disorders
(5th ed.). Washington, DC: Author.
Bargiela, S., Steward, R., & Mandy, W. (2016). The experiences of late-diagnosed women with autism spec-
trum conditions: An investigation of the female autism phenotype. Journal of Autism and Developmental
Disorders, 46, 3281–3294. doi:10.1007/s10803-016-2872-8
Braunschweig, D., Krakowiak, P., Duncanson, P, Boyce, R., Hansen, R. L., Ashwood, P., . . .Van De Water,
J. (2013). Autism-specific maternal autoantibodies recognize critical proteins in developing brain.
Translational Psychiatry, 3, e277.
Brooks, M. (2015, October 6). New autism genes discovered. Medscape. Retrieved from http://www.med-
scape.com/viewarticle/852193
Centers for Disease Control and Prevention. (2014). Prevalence of autism spectrum disorder among chil-
dren aged 8 years – Autism and developmental disabilities monitoring network, 11 sites, United States,
2010. MMWR Surveillance Summaries, 63, 1–21. Retrieved from http://www.cdc.gov/mmwr/preview/
mmwrhtml/ss6302a1.htm?s_cid=ss6302a1_w
Chabris, C. F., Lee, J. J., Cesarini, D., Benjamin, D. J., & Laibson, D. I. (2015). The fourth law of behavior
genetics. Current Directions in Psychological Science, 24, 304–312. doi:10.1177/0963721415580430
Chang, J., Gilman, S. R., Chiang, A. H., Sanders, S. J., & Vitkup, D. (2015). Genotype to phenotype relation-
ships in autism spectrum disorders. Nature Neuroscience, 18, 191–198. doi:10.1038/nn.3907
Crittenden, P. M. (2015). Raising parents: Attachment, representation, and treatment (2nd ed.). London,
England: Routledge.
Crittenden, P. M., Dallos, R., Landini, A., & Kozlowska, K. (2014). Attachment and family therapy. London,
England: McGraw-Hill.
Cuthbert, B. N., & Insel, T. R. (2013). Toward the future of psychiatric diagnosis: The seven pillars of RDoC.
BMC Medicine, 11, 126.
de Lacy, N., & King, B. H. (2013). Revisiting the relationship between autism and schizophrenia: Toward an
integrated neurobiology. Annual Review of Clinical Psychology, 9, 555–587.
DeStefano, F., Price, C. S., & Weintraub, E. S. (2013). Increasing exposure to antibody-stimulating proteins and
polysaccharides in vaccines is not associated with risk of autism. The Journal of Pediatrics, 163, 561–567.
Division of Clinical Psychology, British Psychological Society. (2013). Division of Clinical Psychology
position statement on the classification of behaviour and experience in relation to functional
388 Clinical Child Psychology and Psychiatry 22(3)

psychiatric diagnoses: Time for a paradigm shift. Retrieved from http://dxrevisionwatch.files.word-

press.com/2013/05/position-statement-on-diagnosis-master-doc.pdf
Dohmen, A., Bishop, D. V. M., Chiat, S., & Roy, P. (2016). Body movement imitation and early language
as predictors of later social communication and language outcomes: A longitudinal study. Autism &
Developmental Language Impairments, 1, 1–15. doi:10.1177/2396941516656636
Fein, D., Barton, M., Eigsti, I. M., Kelley, E., Naigles, L., Schultz, R. T., . . .Tyson, K. (2013). Optimal out-
come in individuals with a history of autism. Journal of Child Psychology and Psychiatry, 54, 195–205.
doi:10.1111/jcpp.12037
Fusaro, V. A., Daniels, J., Duda, M., DeLuca, T. F., D’Angelo, O., Tamburello, J., . . .Wall, D. P. (2014). The
potential of accelerating early detection of autism through content analysis of YouTube videos. PLoS
ONE, 9, e93533. doi:10.1371/journal.pone.0093533
Gaugler, T., Klei, L., Sanders, S. J., Bodea, C. A., Goldberg, A. P., Lee, A. B., . . .Buxbaum, J. D. (2014). Most
genetic risk for autism resides with common variation. Nature Genetics, 46, 881–885. doi:10.1038/ng.3039
Green, J., Charman, T., Pickles, A., Wan, M. W., Elsabbagh, M., . . .Johnson, M. H. (2015). Parent-mediated
intervention versus no intervention for infants at high risk of autism: A parallel, single-blind, randomised
trial. Lancet Psychiatry, 2, 133–140. doi:10.1016/S2215-0366(14)00091-1
Griffin, C., Lombardo, M. V., & Auyeung, B. (2016). Alexithymia in children with and without autism spec-
trum disorders. Autism Research, 9, 773–780.
Harris, H., Israeli, D., Minshew, N., Bonneh, Y., Heeger, D. J., Behrmann, M., & Sagi, D. (2015). Perceptual
learning in autism: Over-specificity and possible remedies. Nature Neuroscience, 18, 1574–1576.
doi:10.1038/nn.4129
Hastings, R. P., Kovshoff, H., Ward, N. J., Espinosa, F. D., & Remington, B. (2005). Coping strategies in
mothers and fathers of preschool and school-age children with autism. Autism, 9, 377–391.
Jacobs, L. A., Rachlin, J., Erickson-Schroth, L., & Janssen, A. (2014). Gender dysphoria and co-occurring
autism spectrum disorders: Review, case examples, and treatment considerations. LGBT Health, 1, 277–
282. doi:10.1089/lgbt.2013.0045
Jones, W., & Klin, A. (2013). Attention to eyes is present but in decline in 2–6-month-old infants later diag-
nosed with autism. Nature, 504, 427–431. doi:10.1038/nature12715
Just, M. A., Cherkassky, V. L., Buchweitz, A., Keller, T. A., & Mitchell, T. M. (2014). Identifying autism
from neural representations of social interactions: Neurocognitive markers of autism. PloS ONE, 9(12),
e113879.
King, B. H., Navot, N., Bernier, R., & Webb, S. J. (2014). Update on diagnostic classification in autism.
Current Opinion in Psychiatry, 27, 105–109.
Lombardo, M. V., Pierce, K., Eyler, L. T., Barnes, C. C., Ahrens-Barbeau, C., Solso, S., . . .Courchesne, E.
(2015). Different functional neural substrates for good and poor language outcome in autism. Neuron,
86, 567–577.
Lu, J. T., Kishida, K. T., De Asis-Cruz, J., Lohrenz, T., Treadwell-Deering, D., Beauchamp, M., & Montague,
P. R. (2015). Single-stimulus functional MRI produces a neural individual difference measure for autism
spectrum disorder. Clinical Psychological Science, 3, 422–432. doi:10.1177/2167702614562042
Mariani, J., Coppola, G., Zhang, P., Abyzov, A., Provini, L., Tomasini, L., . . .Vaccarino, F. M. (2015).
FOXG1-dependent dysregulation of GABA/glutamate neuron differentiation in autism spectrum disor-
ders. Cell, 162, 375–390.
Markowitz, L. A., Reyes, C., Embacher, R. A., Speer, L. L., Roizen, N., & Frazier, T. W. (2016). Development
and psychometric evaluation of a psychosocial quality-of-life questionnaire for individuals with autism
and related developmental disorders. Autism, 20, 832–844.
Mezzacappa, A., Lasica, P. A., Gianfagna, F., Cazas, O., Hardy, P., Falissard, B., . . .Gressier, F. (2017). Risk
for autism spectrum disorders according to period of prenatal antidepressant exposure: A systematic
review and meta-analysis. JAMA Pediatrics.
Moran, H. (2010). Clinical observations of the differences between children on the autism spectrum and those
with attachment problems: The Coventry Grid. Good Autism Practice, 11, 44–57.
Moriuchi, J. M., Klin, A., & Jones, W. (2017). Mechanisms of diminished attention to eyes in autism. The
American Journal of Psychiatry, 174, 26–35. doi:10.1176/appi.ajp.2016.15091222
Crittenden 389

Oppenheim, D., Koren-Karie, N., Dolev, S., & Yirmiya, N. (2009). Maternal insightfulness and resolution
of the diagnosis are associated with secure attachment in preschoolers with autism spectrum disorders.
Child Development, 80, 519–527.
Rao, P., & Landa, R. J. (2014). Association between severity of behavioral phenotype and comorbid atten-
tion deficit hyperactivity disorder symptoms in children with autism spectrum disorders. Autism, 18,
272–280. doi:10.1177/1362361312470494
Reichow, B. (2012). Overview of meta-analyses on early intensive behavioral intervention for young chil-
dren with autism spectrum disorders. Journal of Autism and Developmental Disorders, 42, 512–520.
doi:10.1007/s10803-011-1218-9
Richey, J. A., Damiano, C. R., Sabatino, A., Rittenberg, A., Petty, C., Bizzell, J., . . .Dichter, G. S. (2015).
Neural mechanisms of emotion regulation in autism spectrum disorder. Journal of Autism and
Developmental Disorders, 45, 3409–3423. doi:10.1007/s10803-015-2359-z
Roberts, A. L., Lyall, K., Rich-Edwards, J. W., Ascherio, A., & Weisskopf, M. G. (2013). Association of
maternal exposure to childhood abuse with elevated risk for autism in offspring. Journal of the American
Medical Association Psychiatry, 70, 508–515.
Sanders, S. J., He, X., Willsey, A. J., Ercan-Sencicek, A. G., Samocha, K. E., Cicek, A. E., . . .State, M. W.
(2015). Insights into autism spectrum disorder genomic architecture and biology from 71 risk loci. Cell,
87, 1215–1233. doi:10.1016/j.neuron.2015.09.016
Scott, N. J., Jones, A. L., Kramer, R. S. S., & Ward, R. (2015). Facial dimorphism in autistic quotient scores.
Clinical Psychological Science, 3, 230–241. doi:10.1177/2167702614534238
Seskin, L., Feliciano, E., Tippy, G., Yedloutschnig, R., Sossin, K. M., & Yasik, A. (2010). Attachment and
autism: Parental attachment representations and relational behaviors in the parent-child dyad. Journal of
Abnormal Child Psychology, 38, 949–960.
Shen, M. D., Kim, S. H., McKinstry, R. C., Gu, H., Hazlett, H. C., Nordahl, C. W., . . . Fonov, V. S. (2017).
Increased extra-axial cerebrospinal fluid in high-risk infants who later develop autism. Biological
Psychiatry.
Silberman, S. (2015). Neuro Tribes: The legacy of autism and the future of neurodiversity. New York, NY:
Avery.
Strathearn, L. (2009). The elusive etiology of autism: Nature and nurture? Frontiers in Behavioral
Neuroscience, 3, 11. doi:10.3389/neuro.08.011.2009
Tang, G., Gudsnuk, K., Kuo, S.-H., Cotrina, M. L., Rosoklija, G., Sosunov, A., . . .Sulzer, D. (2014). Loss
of mTOR-dependent macroautophagy causes autistic-like synaptic pruning deficits. Neuron, 83, 1131–
1143. doi:10.1016/j.neuron.2014.07.040
Taylor, B., Jick, H., & MacLaughlin, D. (2013). Prevalence and incidence rates of autism in the UK: Time trend
from 2004–2010 in children aged 8 years. BMJ Open, 3, e3003219. doi:10.1136/bmjopen-2013-003219
Uddin, L. Q., Supekar, K., Lynch, C. J., Cheng, K. M., Odriozola, P., Barth, M. E., . . .Menon, V. (2015).
Brain state differentiation and behavioral inflexibility in autism. Cerebral Cortex, 25, 4740–4747.
doi:10.1093/cercor/bhu161
van Ijzendoorn, M. H., Rutgers, A. H., Bakermans-Kranenburg, M. J., Dietz, C., . . . van Engeland, H. (2007).
Parental sensitivity and attachment in children with autism spectrum disorder: Comparison with children with
mental retardation, with language delays, and with typical development. Child Development, 78, 597–608.
Wan, M. W., Green, J., Elsabbagh, M., Johnson, M., Charman, T., & Plummer, F. (2012). Parent-infant inter-
action in infant siblings at risk of autism. Research in Developmental Disabilities, 33, 924–932.
World Health Organization. (1992). International statistical classification of diseases and related health
problems. Geneva, Switzerland: WHO Press.

Author biography
Patricia M Crittenden, MEd, PhD, is a developmental psychopathologist who studied under Mary D. S.
Ainsworth. She taught special education in preschool, elementary school, and high school, for children with
intellectual and emotional disabilities, including autism. She has published more than 100 scientific papers
and chapters, plus several books on attachment, child protection, and mental illness.