Chapter 7: Discussion and Conclusions

Print page

This investigation was longitudinal in nature, and focused on a group of children and their families, at two points in time. The first phase of the study, conducted in 1990 on a sample of 185 children, concluded that 16% of children and approximately one-third of mothers showed evidence of formal psychiatric diagnosis. Childhood diagnosis was related to maternal mental health and to the economic circumstances of the family. Maternal psychological status was similarly associated with economic factors as well as dissatisfaction with marital and parenting roles and social isolation.

The second phase of the research, on which this report is based, includes ninety-seven young people who were successfully contacted and interviewed for the study. Eighty mothers were also re-interviewed. The aim of the study was to investigate a range of psycho-social outcomes for these young people and to relate those outcomes to factors identified as important in the first stage of the study. The relevant factors included respondents' diagnosis, behavioural deviancy, IQ score, maternal psychiatric diagnosis and the economic circumstances of the family. The psycho-social outcomes of interest in the young adult group included psychological adjustment, educational and employment status, involvement in crime and alcohol and drug use. Mothers were interviewed to re-assess their mental health status and to provide information on family circumstances during the participants' childhood and adolescence.

7.1 Social Development and Family Background

The mean age of the participants was twenty-one years and the majority in both phases of the study were male but gender was more evenly distributed in the present phase. They were largely homogeneous in terms of area of residence, accommodation type (over three-quarters still living in the parental home) and marital status (primarily single). Approximately one-fifth now had children. Almost all of the participants had been raised in the family home by both parents and they appeared, in general, to have good relationships with their parents. The participants were predominantly from semi-skilled manual backgrounds and only two parents had a third level qualification. A fifth of mothers reported severe financial hardship during the participant's childhood. There was wide variation in relation to educational attainment. Approximately one-third had not completed second level education. While females remained in the educational system for longer, gender differences were non- significant. Almost three-quarters of the sample were in employment, the majority being reasonably satisfied with their life and work situations. Religious observance was low yet over three-quarters of participants professed to believe in God and 40% regarded religion as important in their lives. Almost one-quarter of those interviewed, three-quarters of whom were male, had had some contact with the police but this was mainly for minor offences, and none of the respondents had been in prison. Overall, the profile of the participants' lives which emerged from the study was one of economic and social stability.

7.2 Health and Well-Being

There was no evidence of serious psychiatric disorder and only a small percentage of the group were receiving treatment for psychological difficulties. However, an assessment of psychological health status indicated a not insignificant level of symptoms, below diagnostic level. The prevalence for clinical level disorder is lower than that found in another Irish study of a similar population and may reflect differences in assessment methods (Lawlor & James, 2000). The most common symptoms were anxiety-related and the only significant gender variations related to alcohol misuse with males predominating. Alcohol consumption was high, especially amongst males respondents, and participants tended to underestimate their level of alcohol use. Respondents demonstrated adequate levels of self-esteem and locus of control and appeared to be well integrated socially as well as having sources of intimacy and affirmation. However, as in Lawlor and James's (2000) study, males were less likely to confide in others.

It is of interest that participants, including those with a psychiatric disorder at age eleven, came across as well-adjusted, psychologically healthy and competent individuals. This is in line with the majority of longitudinal studies which demonstrate, in general, a successful transition to adulthood for most individuals (Rutter, 1989). The generally positive results relating to psychological health must however be considered in terms of the possible bias that may have arisen due to sample attrition. Twenty percent of the young people could not be located and a further 21% refused to participate in the current phase of the study. As is the case with all longitudinal studies, respondents who participated in this phase may not be representative of the original sample and might be characteristic of those with the most favourable outcomes. There are some reasons for believing this might be the case.

A comparison of respondents and non-respondents revealed significantly higher levels of psychological and behavioural dysfunction among the untraced group (but not refusals) at age eleven. Almost 50% of the original sample of children who were judged vulnerable because of psychiatric disorder and/or behavioural deviance at age eleven were lost to follow-up. A small number of individuals from the original sample were not contactable because they were no longer in communication with their families. In most cases neither the young people nor their families could be traced, despite extensive inquiries, including the accession of public housing records. Thus, psychiatric morbidity and/or behavioural deviancy identifies a mobile group, either of the young people or of their families (or both). Additional analysis carried out for the study suggests that this untraced group may be more vulnerable than study participants. Non-contacts had considerably higher rates of offending than study participants (and refusals) in the intervening years.

7.3 Maternal Health & Child Well-being

Mothers' psychological health was identified in Phase One as importantly associated with childhood mental health. Then, almost a third of the mothers interviewed were diagnosed with a psychiatric disorder. Now mothers were generally healthy with no evidence of clinical-level disorder and treatment levels in the intervening years were similar for all mothers. Again, in contrast to Phase One of the study, mothers were satisfied with their marital and parenting roles. The association between maternal and child mental health is not apparent in this phase of the research which disputes a fairly consistent finding in the literature (Downey & Coyne, 1990; Weissman, Warner, Wickramaratne, Moreau & Olfson, 1997). Both adult depressive conditions and childhood psychiatric diagnosis and/or behavioural deviancy have been considered as conditions which tend to persist. The present study shows this not to be the case. Of the 50% of vulnerable children identified at Phase One and interviewed for this stage of the study the majority were doing better than might have been predicted. Mothers had similarly recovered and had not required more intensive psychiatric care in the intervening years.

This discrepancy could be due to methodological differences in that many studies concentrate on patient groups and focus exclusively on psychological outcomes (Weissman et al 1997). In this research sampling was based on random selection and other outcomes, such as educational attainment, were included. The findings reflect other longitudinal examinations of maternal/child interactions which suggest that maladaptive outcomes in children are dependent on the nature of the mother's illness, her environment and outcomes examined in the child. The life-stage which includes the care of young children is a potentially stressful phase for women (Brown & Harris, 1978; Cleary, 1997a; Christoffersen, 2000). The mothers in this study were, when assessed in Phase One, almost all full-time mothers and many were restricted (some severely) by lack of money and support. These results point to the transience of much psychiatric symptomatology amongst the general, female, population, the origins of which are often more social than biopsychiatric (Cleary, 1997b). There may be some support here for the view that the relationship between maternal and child well-being is more reciprocal than previously considered (Naerde, Tambs & Mathiesen 2002; Conrad & Hammen 1989). Mothers of children with behavioural problems experience considerable stress and at least some of the disorder seen in the mothers of ten years ago may have resulted from the strain of coping with this.

Yet, maternal psychological health did have an affect on the child's later life in that the educational attainment of children whose mother had a psychiatric disorder in Phase One were less likely to be successful educationally. And, of course the same proviso, cited above, is relevant here also in that the healthiest individuals (and their mothers) from the original sample may have been included in the present study.

7.4 Childhood Behavioural Problems

There was a good degree of overlap, as expected, between behavioural deviance and psychiatric diagnosis. None of the respondents showed evidence of current clinical level disorder and there were few differences in relation to current symptoms between the two groupings. However the deviant/diagnosed group were more likely to leave school early and less likely to go on to third level education. Approximately half of the non-disordered had gone on to third level education in contrast to less than a fifth of the categorised groups. They also differed significantly in relation to their school experiences. The diagnosed/deviant group were more likely to offend against the law and were more probably from an economically deprived background.

Early behavioural problems have been cited as one of the strongest predictors of later problems, including educational underachievement, and there is support for this in the present study. Fergusson and Horwood (1998) and Rutter (1989) have described the cycle of disadvantage set in motion by early behavioural problems. Disruptive behaviour makes it more likely that there will be an early exit from school and the individual is then less likely to attain a stable occupational status. Fergusson and Lynskey's (1998) and Kolvin et al's (1990) findings that young people with conduct disorders have higher rates of juvenile offending are reflected in this study. And, as in Olin et al's (1998) research, teacher's assessments of behaviour have good predictive value in identifying children at risk at an early stage.

7.5 Behavioural Deviance, Economic Disadvantage and Educational Outcome

Educational attainment emerged in this study as the strongest outcome variable. Mental health status and offending against the law demonstrated weaker associations. There were few gender differences except for a relationship between parenthood in females and less favourable educational outcome which supports existing findings (NESF, 1997; McCashin, 2000). Early school leavers were more likely to have offended against the law. Socio-economic status was strongly associated with educational achievement and economic disadvantage (the most sensitive measure of which was receipt of state benefits) predicted lower educational attainment. This reflects the findings from other longitudinal studies (Kolvin et al., 1990). Mother's mental health was unrelated to educational outcome. Four variables emerged from the analysis as key to understanding outcome in terms of educational attainment. These factors, economic disadvantage, categorisation of deviancy, and another assessment of childhood functioning, IQ, all proved to have significant, independent, and direct effects on educational outcome. Thus, respondents with a categorisation of behavioural deviancy were less likely overall to be successful educationally but this likelihood increased with the accumulation of risk factors. Individuals with this classification from a disadvantaged background (i.e. family in receipt of state benefits) and a lower than average IQ were highly unlikely to complete second level schooling, in marked contrast to those participants who were not in any of the categories. IQ was important but was associated with the presence or absence of economic disadvantage. Thus respondents with an average IQ, without a deviant classification and with no evidence of economic disadvantage in their background were twice as likely to complete second level as those disadvantaged individuals with a deviant categorisation. At third level, economic factors had an even greater impact with better off individuals much more likely to reach this level than disadvantaged participants, even with similar IQ levels.

This link between economic disadvantage and educational outcome has been similarly identified in two recent studies of the Irish educational system (Hayes & Kernan, 2001; Clancy, 2001). Hayes and Kernan's (2001) work highlights the early emergence of educational inequities along socio-economic grounds and Clancy's (2001) study demonstrated the impact of these inequalities on third level access. Lower socio-economic groupings are significantly underrepresented in the third level educational sector. This study reflects Clancy's finding that middle-class students, even when educationally less able, are more likely to reach third level than individuals from other social categories.

7.6 Conclusions

In this follow-up study of young people the relatively high prevalence of psychiatric disorder and behavioural deviancy, evident ten years previously proved, in general, to be transitory and without any long-term psychological impact. Similarly, mothers previously diagnosed as suffering from psychological problems had now recovered and the absence of disorder raises the possibility that at least some of the complaints seen in the mothers of disturbed children may have resulted from the stress of coping with this disturbance. The results imply a lack of association between maternal mental disorder and long-term psychological difficulties in the child. Whether the non-contactable group were similar or dissimilar in outcome can only be speculative. There is a possibility that the untraced group might exhibit more negative outcomes than the present sample and thus would have altered the relatively healthy psychological and social profile presented by the study participants.

Yet the findings of this study clearly indicate that behavioural problems in the child do have long-term effects. Behavioural deviancy in childhood impairs individuals in terms of educational achievement and this situation is exacerbated in the context of economic disadvantage. Early behavioural difficulties are signalled here as key to understanding outcome especially educational attainment. Economic factors and IQ also have significant predictive powers although they are somewhat less important than the classification of deviancy. Economic factors are more essential than intellectual capacity in terms of educational attainment. The trajectory of the student with behavioural problems from a disadvantaged background, with possible learning difficulties, falling through the educational net is apparent here (Miech et al., 1999). This finding is reinforced by the likely possibility that had all of the original sample been included, the impact of behavioural problems would probably have been even greater.

The children interviewed for this study have benefited from the economic prosperity and widening educational opportunities available to them as they grew to adulthood. However the link between economic and educational disadvantage, especially for at risk groups, along with the shadow cast by a possibly more vulnerable group of untraced individuals, implies that inequities remain. In the light of these findings it may be opportune to re-focus on the role of economic disadvantage in the development and prolongation of behavioural difficulties in the child.


Note re Authorship
Executive Summary
Chapter 1: Introduction
Chapter 2: Methodology
Chapter 3: Risk and Protection for children
Chapter 4: Social and Psychological development
Chapter 5: Health and Behaviour in Childhoods
Chapter 6: Educational Attainment
Chapter 7: Discussion and Conclusions
Appendix 1

- Structured Clinical Interview for DSM-IV Axis I Diagnoses (SCID)
- Beck Scale for Suicide Ideation (BSSI)
- Rosenberg's Self-esteem Scale
- Arizona Social Support Interview Schedule (ASSIS)
- Locus of Control

Appendix 2

- Frequency of SCID Diagnostic Categories

Last modified:04/05/2010

 Related Topics


 Application Forms


 Intreo/Local Office Locator

Search for your local office