5. Conclusions
Children exposed to multiple risks in their lives are much more likely to show psychopathology than children who are not exposed to such risks. A risk index was calculated. The risks included in this index were alcohol abuse in the mother and in the father, marital dissatisfaction, low income, the presence of depression in the PMK, large family size, teenage pregnancy, hostility in the parent-child relationship, parents having divorced and the presence of a learning disability in the child. Children not exposed to any risks showed a rate of psychiatric disorder of approximately 10 percent. Children exposed to four or more of these risks showed a rate of disorder of approximately 50 percent: a fivefold increase in the rates of disorder. This finding was evident across the PMK and the teacher reports of disorder. The association between risk and disorder was less strong when disorder was assessed through child report, although still evident. According to child reports there was over a twofold increase in the likelihood of children showing disorder when they were exposed to four or more risks in their lives compared to no risks. The less strong association between risk and disorder when using children's reports of psychopathology is likely to be because of the lower level of reliability and validity in children's own assessments of psychopathology.
Although children exposed to multiple risks in their lives are at much higher risk of disorder, there are factors occurring naturally in their environments that are associated with a lowered risk of disorder. It may be that such factors in children's lives act as "natural interventions".
There was strong evidence that the presence of good relationships in children's lives is important to their development at 6 and at 10 years of age. We also found evidence for compensatory processes in development. When children are experiencing adverse environments they may draw on other aspects of their environment to provide support and help in the face of adversity. Most of the factors that showed protective effects conformed to Model A: protective at high risk only. In the absence of good sibling relationships, friendships and good relationships with teachers, children in high risk circumstances showed very high levels of disorder. When they had such relationships their levels of disturbance were lower and more comparable to children in low risk environments. These relationships had much less of an impact on children's disturbance when the children lived in low risk circumstances. In the low risk circumstance whether a child had a good relationship with a sibling or friends showed little association with disturbance, indicating that these relationships are not essential to well-being when children are not experiencing adversity. It is clear that children in high risk circumstances draw support from their relationships with other people. Even though some children live in highly stressful environments having close relationships with others may help them to cope with the difficulties that they face.
The number of good relationships that a child has was found to be strongly related to resilience. For 6-year-olds having one good relationship was sufficient to moderate the impact of risk on the development of externalizing disorder. For 10-year-olds this was not the case as children with only one good relationship in high risk circumstances were more at risk than children with two or three good relationships. Another way of expressing this is that for 10-year-olds, having more good relationships provided more protection. For 6-year-olds, having one good relationship was as advantagous as having three. We also found that boys were less likely than girls to have good relationships. The increased vulnerability that we see in boys may in part be related to less protection from close relationships. Boys may start with more problematic behaviour than girls (showing higher levels of activity, aggressive play etc.) making it harder for people in their environment to respond affectionately, which in turn may make it harder for boys to feel confident and untroubled.
At this point when we have only cross sectional data it is difficult to know about the direction of effect of these findings. One possibility is that when children are experiencing a lot of stress in their lives they are helped by having a close relationship with a friend, sibling or teacher. Such a relationship may help them to feel that they are integrated into a community of people. A number of theorists in the area of social support and mental health have talked about this sense of integration with a community as being a central element of people's sense of well being. Their argument is that when people feel that they have an important place in the lives of others (a social role within a community) that this leads to a sense of well being and purposive activity (Durkheim, 1905; Oatley and Bolton, 1985; Scheff, 1997). It is also possible, however, that the effect goes in the opposite direction. Children who are less compromised in their development can reach out to others and pull them into their world. Children in high risk circumstances show better relationships with others because they are less disturbed. This explanation does not explain why the quality of relationships is strongly associated with disturbance in children experiencing high risk and less strongly associated with disturbance in children experiencing low risk. It will not, however, be possible to untangle issues of causality until longitudinal data are available. It will then be possible to look at how the presence of protective factors in children's lives at time 1 predict changes from time 1 to time 2 in children's externalizing and internalizing symptomatology.
There were significant differences in the buffering effects of some putative protective factors among 10-year-old boys and girls. These significant three way interactions accounted for relatively small amounts of variance, between .06 percent and 1.8 percent of the variance depending on which outcomes and which protective factors were at issue. The presence of friendships was associated with reduced levels of externalizing difficulties for both boys and girls, but boys responded even more than girls to the presence of friendships. For girls friendships were important to well-being at both low and high risk. This may relate to the importance that girls attach to social relationships throughout their lives (Golombok and Fivush, 1994), both when they are struggling in their lives and when they are not. Girls were more likely to be buffered from the deleterious effects of a high risk environment by good relationships with their teacher than boys. The quality of the relationship with the teacher was important for boys at all levels of risk, whereas for girls this only became very important to development in high risk circumstances. This may be because schools and school routines are more difficult for boys across all levels of risk. Schools require low activity levels, low distractibility, high verbal skills, aspects of behaviour and temperament that may come less naturally to boys. Both boys and girls at high risk were protected by good sibling relationships. This confirms previous work by one of the authors in which siblings found comfort from one another when the parental marriage was poor (Jenkins, 1992).
Across all analyses it is of interest that we were able to account for more variance in externalizing behaviour than internalizing behaviour. There are a number of possible explanations for such a finding. It may be due to measurement error. Internalizing disorder has been found to be harder to assess than externalizing disorder (Herjanic and Reich, 1982; Kolko and Kazdin, 1993) If there is more error in the measurement of internalizing disorder the possibilities of finding significant main effects and interactions are less likely. In other studies, externalizing disturbance has been found to be more strongly predicted by environmental risks than internalizing disturbance (Costello, 1989) It may be that internalizing disturbance is more strongly related to constitutional elements within the child, such as temperamental behavioural inhibition (Kagan, Reznick and Snidman, 1988; Rubin, 1993) with environmental factors playing less of a role in the etiology and maintenance and internalizing disturbance.
What's New
