Multiproblem or Multirisk Families? A Broad Review of the Literature of Child & Adolescent Behavior

Numerous psychosocial factors, as they occur in multiproblem families, make psychopathology more probable in developing children. This is important because assistance for these families generally focuses on the social side only. The aim of this study is to review these psychosocial risk factors by means of a literature review and to consider whether or not they provide a more accurate definition of multiproblem families. The results contain a list of psychosocial risk factors affecting parents and families. As a result of multiple psychosocial risk factors, multiproblem families have a negative impact on development and psychopathology in young children. Conclusion: a more concrete definition of what constitutes a multiproblem family, for example based on the total number of risk factors, will make it possible to develop scientific research and therefore more effective prevention and therapy.


Introduction
In recent decades, there has been extensive research into the effects of risk factors on the development of psychopathology in children and adolescents. A risk factor is generally defined as a characteristic, an experience or an event that is associated with a higher probability of psychopathology [1][2][3][4][5] or a severe form or longer duration of a major (psychiatric) illness [6]. There is considerable overlap between the risk factors for different disorders such as depression, behavioural disorder, substance abuse or schizophrenia. The best-known risk factors are therefore social class contrasts (poor, uneducated parents without social support and many stressful life events) and mothers with a psychiatric diagnosis [7]. Psychosocial risk factors also affect the development of somatic disorders [8][9][10][11]. Multiproblem families are afflicted by a large number of social problems such as difficulties with neighbours, financial problems and domestic violence as a result of a combination of several risk factors. Furthermore, the parents in these families frequently have psychopathological disorders and the families receive psychiatric assistance (from child and adolescent services) for longer or shorter periods of time. This is often associated with stagnation in the development of the children.
It would appear to be obvious -and this is something that is also described in the literature in general -that the psychosocial risk factors in multiproblem families constitute a threat to the development of young children and that these problems may be an important factor in the development of psychopathology. Literature about the development of psychopathology in children in general terms is abundant, but studies looking specifically at the impact of multiproblem families on the development of psychopathology in young children under the age of 12 years are scarce.
In the major cities of the Netherlands, a special approach has been developed in an attempt to achieve improvements in these multiproblem families and their circumstances. However, the indications for assistance for these families are not always clearly defined and there is no concrete definition of what constitutes a multiproblem family. A commonly used definition is "a family confronted over a long period of time with a combination of socioeconomic and psychosocial problems, with important characteristics being that there are problems both in the family itself and with the services supplying assistance and that the development of individual members of the family stagnates" [12]. Verzaal [13] proceeds on the assumption of families with serious problems in more than one of the four basic areas of family life: social adaptation, health, economic behaviour and recreational needs.
In this paper, we set out a broad review of the psychosocial risk factors for the development of children (aged 0-12 years) who grow up in multiproblem families. We aimed to identify all possible risk factors and find out how multiple risk factors act in the possible development of psychopathology in those families. We then try to define multiproblem families on the basis of these risk factors with the aim of arriving at a more concrete and more operational definition of the "multiproblem family".

Method
To establish a picture of risk factors in multiproblem families that may be linked to the development of psychopathology in children (up to the age of 12 years), we used three search strategies: First of all, we used Medline/Ovid for a search with the following terms: multiproblem, multi-problem "and" children "and" risk factor(s) "and" psychopathology.
The first search generated a limited number of useful articles. It was therefore decided to adopt a second search strategy in Medline/Ovid using the following terms: child(ren) "and" risk factor(s) "and" psychosocial "and" psychopathology (from 1995 to August 2012). This search generated more than 3000 studies and so we decided to add the term "review" to the search.
In the final search, the Dutch search system Picarta was used with "multiproblem" as the keyword. Table 1 lists the numbers of articles generated using the three search strategies.  Table 1: Search strategy and articles found and used The first search strategy produced 120 articles. The second search strategy, after the addition of the term "review", produced 416 reviews. The third search identified 40 articles.
On the basis of the abstracts and the titles, we selected the studies in which the children were younger than 12 years of age from the group of 567 articles. We excluded the articles looking at adolescents, case studies, articles not written in English or Dutch and articles looking specifically at particular disorders such as ADHD, or specific situations such as sexual abuse.
We selected 2, 15 and 4 articles from the three selections of articles (N=576). The relevant articles listed in the bibliographies in this group of 21 articles were also obtained and assessed (39 articles in all).
Ultimately, we drew on 60 articles for our broad review. The final five columns of table 1 list the sections in this paper in which we quote the various articles.

Results
In the 60 studies we identified that looked at the risk of psychopathology in children (aged 0-12 years), we distinguished between three different levels, two of which we drew on in this paper. The three levels relate to the child, parent and family/context perspectives. Risk factors at the child level such as sex, age, temperament and chronic physical illnesses are generally linked to particular genetic constellations. From the psychosocial point of view, however, this paper focuses on describing risk factors in multiproblem families from the parent and family/context perspective [14].

Parent level
With respect to parents, most of the discussion in the literature concentrates on mothers. There has been much less research looking at the role of fathers. However, there are indications that fathers will also have an effect on the development of psychopathology [15,16]. Connel and Goodman [17] use the term "assortative mating" -the concurrence of mental health problems in both fathers and mothersin this context. This makes it hard to differentiate between the influence of fathers and mothers alone.
It emerges from the literature that different characteristics of the parents can affect the development of psychopathology, examples being the age, level of education, social skills and any psychiatric problems of the parents.
-The younger the mother when the child is born, the greater the possibility that the child will develop psychopathology in the course of time. The age threshold that is generally found in the literature is 25 years of age. Many child-care institutions therefore have special facilities for young mothers (with 21 or 18 years as the age threshold), most of which focus on good somatic care, upbringing, housing and finances [7,[18][19][20][21][22]. The study by Wille et al. [23] was, however, unable to demonstrate that younger maternal age is a risk factor.
-The level of education of the mother is a risk factor, with lower levels of education being linked to a higher probability of the development of psychopathology in these children [5,18,19,22,[24][25][26][27][28][29].
-Children of mothers with good social skills and more perceived support from the environment are less likely to develop psychiatric problems. However, more perceived stress such as recent unemployment and divorce, or other problems with relationships, makes psychiatric problems more likely [18,[27][28][29][30].
-More than a third (41-77%) of the children of parents with psychiatric problems develop chronic and severe problems themselves [16,22,23,[30][31][32][33][34][35]. Of all psychiatric disorders, personality disorders in parents probably represent the greatest risk for their children [36] but all disorders have an effect on the development of mental problems [16]. Nevertheless, research has yet to identify any difference associated with paternal and maternal disorders [17]. The impact on the child is greater if both parents have a psychiatric disorder [16]. Severity of the disorder, comorbidity and chronicity also play an important role [16,20,21,28,30,37]. Saxena et al. [6] have shown that children of depressed parents have a more than 50% probability of becoming depressed themselves. It is one of very few studies that presents a specific percentage of probability. Ezpeleta et al. [34] report that drugs and/or alcohol consumption by the mother can also be seen as a risk factor. Connell and Goodman [17] actually find that this is the case with fathers only, whereas Wille et al. [23] could not demonstrate that alcohol consumption should be seen as a risk factor. Finally, it has been found that not only psychiatric problems but also chronic physical illnesses in one of the parents involve an increased risk [23,38].

Family/context level
We have classified under the family/context level all risk factors that cannot be attributed individually to the parent. This level includes the composition of the family, socio-economic status, the level of organisation of the household, ethnicity and the level of aggression in the family and the neighbourhood.
-It emerges from the literature that certain characteristics of the family, for example the family structure and the size of the family, are linked to an increased risk of psychopathology in the children. This is most clearly the case in single-parent families, but "other" family structures such as adoptive parents, homosexual couples, and foster or stepparents are also considered by some to be risk factors [5,19,20,[22][23][24]27,28,34,39]. In terms of family size, the absence of siblings (brothers or sisters) is a risk factor, as is having a large number of siblings (more than three children in the family) [18,20,25,27,28,40]. Wille et al. [23] actually conclude that the number of siblings has no effect on the development of psychopathology.
-Lower socioeconomic status (SES) is generally seen and probably the most studied risk factor for psychopathology. For this reason we discuss this risk factor more extensively, even though some studies have failed to find any link [38]. Shaw and Vondra [41] have shown that low SES in particular results in attachment insecurity, which is a predictor of behaviour problems later in life. Whether SES should be seen as a proximal risk factor, or more as a distal one, is not yet clear. For example, the risk of antisocial behaviour is determined less directly by poverty and is therefore a distal factor in the onset chain but more direct and therefore proximal in terms of the approach to upbringing. Poverty does make upbringing more difficult [36]. Lower SES is also correlated with a number of the risk factors referred to above at the parent level, such as the level of education and parental psychopathology. In addition, higher SES is linked to having paid employment, family income, family composition and family functioning [2,18,20,21,[23][24][25][26][28][29][30]34,39,[42][43][44]. Slopen et al. [45] suggest that low SES could lead to dietary deficiencies and constitute a risk factor for pathology as a result. Evans [8] has shown that parents with low SES are less responsive to their children and more authoritarian. Children from families with low SES also come into contact more with aggression, are confronted more often with the divorce of their parents and have smaller social networks. Chronic poverty has a more negative impact than temporary poverty.
-A disorganised home (an environment with high noise levels and crowding, and with little regularity or routines) also leads to unhealthy socio-emotional development. The result is that children are unstable at school and in the home situation, and this results in a negative effect upon cognitive development [8,46].
-The various studies are not unanimous with respect to ethnicity. It is probable that most disorders are not specifically linked to certain races or population groups but that the factors involved are perceived discrimination, moving to the Netherlands or another host country at a later age, or poor or inadequate mastery of Dutch or of the language of the host country [7,22,39,47]. Harland et al. [27] and Weisz et al. [48] actually failed to demonstrate that ethnicity is a risk factor. Larson et al. [10] found that, in terms of socio-emotional problems, ethnicity may have a protective effect. Gutman et al. [28] show that it is actually a risk factor if at least another two risk factors are involved. Van Os et al. [47] concluded that people from small ethnic groups are at higher risk but that there is no risk for people who are members of large ethnic groups.
-Troubled family functioning, with frequent arguments, aggression and criminality, can be seen as a risk factor for the onset of psychopathology. This also applies to neighbourhoods or areas with higher crime levels [20,23,29,30,34,39,42,[49][50][51][52]. Sameroff [7] conducted the first major longitudinal study of, among other things, environmental factors in the Rochester Longitudinal Study and found a high correlation between the mental health and intelligence of children at the ages of 4, 13 and 18 years and ten factors included in the study. These 10 factors are: (a) a history of maternal mental illness; b) high maternal anxiety; c) parental perspectives that reflected rigidity in the attitude, beliefs and values that a mother had in regard to her child's development; d) few positive maternal interactions with the child observed during infancy; (e) head of household in unskilled occupations; (f) minimal maternal education; (g) disadvantaged minority status; (h) single parenthood; (i) stressful life events, and (j) large family size. Kazdin et al. [1] found in their study that risk factors do not work solely in an additive way but also synergetically. Their study showed that four or more of the risk factors they examined resulted in a tenfold increase of the probability of later dysfunction in the children studied. Larson et al. [10] demonstrated in their study that single psychosocial risk factors in themselves have an average odds ratio of 1.5 -3.0 with respect to various health parameters. However, a clear cumulative effect was demonstrated: children with more than six of the eight risk factors studied were shown to be 17 times more at risk of poor health status than children with no risk factor at all. It is not exclusively the number of risk factors that determines the ultimate outcome, but also duration over time [14,28,53,54].

Accumulation of risk factors in multiproblem families
It is not entirely clear whether specific risk factors leads to a specific type of pathology. Fishbein [55] states, for example, that it is not so much the characteristics of the risk factors that determine the impact on the child but that this is possibly more dependent upon genetic factors such as temperament and sex. Harland et al. [27] actually showed that some risk factors were found in conjunction with more internalising disorders and others in conjunction with more externalising disorders. The Great Smokey Mountain Study [51,52] also made it clear that some disorders are more correlated with certain psychosocial risk factors than others. This study shows that parental psychopathology and crime, as well as problems in the relationships with the partner and/or child, resulted in the strongest probability of disorders. Approximately 9% of the population they studied came into this highest risk group. In particular, parental psychopathology in this group seems to us to be an important challenge facing psychiatric assistance for parents and children that should not be missed.
The most vulnerable group identified by Lanza et al. [24] was the combination of single mothers with low incomes who had their first children before the age of 19, who had little education and who were poor readers. So it is not only the total number of risk factors, but also the type and the configuration of psychosocial risk factors that appears to affect the development of psychopathology in children. The study by Wille et al. [23] showed that psychosocial risk factors do not play a gender-specific role in terms of the development of psychopathology.

Discussion
Multiproblem families are not only a social problem; they can also play a major negative role in the development of psychopathology in There have been studies of separate and isolated factors of parents and the family on the development of the child, but the effect of the accumulation of risk factors on the child has apparently not been extensively studied.
A single psychosocial risk factor results in a modest increase in the risk of pathology/psychopathology in a child. However, it is the accumulation of risk factors, as seen in multiproblem families, that results in a sharp increase in the probability of a psychiatric disorder (and probably in a somatic disorder as well). This multiplicity of risk factors is most relevant for a lack of proper adaptation by the child, resulting in an increase in the probability of psychopathology [23,34,[56][57][58]. Lyons-Ruth et al. [59] linked a multiplicity of psychosocial risk factors to disorganised attachment and postulate this as an explanation for the development of childhood psychopathology. Others have also linked risk factors to attachment: all risk factors may work to reduce caregiving quality and increase attachment risk. Attachment security can be influenced by a) the personality and psychological health of the caregiver, b) marital or adult attachment relationships and c) stress associated with poverty [14].
Multiproblem families are usually defined in pessimistic footnotes to programmes for families confronted by a mountain of psychosocial misery and care needs. Often, this begins and ends with the pinning on of the "multiproblem family" label, which is no solution. However, it should be clear from the above that this label does not do justice to the development or, more importantly for psychiatry, the psychological development of the young child. This development has proven to be dependent upon the duration, type and number of risk factors faced by the child. We believe that it would be better to define these families on the basis of the total psychosocial risk factors described in this article.
We therefore think it would be advisable, in the case of multiproblem families, to screen for risk factors at the level of the mother, at the level of the partner or former partner, and at the family/ context level. This results in a more precise picture of the problems facing these children and at least opens up avenues for assistance programmes.

Conclusion
We split up a literature review looking at psychosocial risk factors in young children up to the age of 12 years into a description of psychosocial risk factors that may result in psychopathology and a description of the accumulation of those factors as found in multiproblem families.
In 1979, Rutter was the first to demonstrate that a combination of more than two risk factors was linked to the probability of psychopathology in children. The more risk factors, the higher the probability of a negative outcome. There is remarkable little literature on risk factors in multiproblem families and further research is needed to confirm the psychosocial risk factors found in our study and the effect of accumulation on the development of psychopathology.
The idea behind a description as a first step is that this results in a better definition of these families on the basis of these factors taken as a whole. Again, further research is needed both to identify the specific factors contributing to psychopathology and the effect of duration in time of those factors.
Psychosocial problems are not the sole contributors to the development of psychopathology, social and political problems cannot be excluded, however these were not included in our study.
However, our focus on negative social circumstances would seem to be indispensable for adequate treatment not only in multiproblem families but also for children with clearly defined pathology [10]. Standard enquiries about psychosocial circumstances using a selfreport questionnaire may generate a lot of information in an efficient way that can be used to map out the problems [60]. This could generate more openings for scientific research, which will be very helpful, precisely for the people working with multiproblem families.