Resilient Children of Parents Affected by a Dependency

Resilience: a specific concept with its own methodological implications

Resilient Children of Alcohol-dependent Parents

Children of alcohol-dependent parents are at greater risk of developing alcohol problems than are children of non-alcohol dependent parents (Goodwin, 1979; Russell, 1990). They are also at greater risk of developing cognitive or behavioural problems which can themselves lead to problems of addiction late in life (Fitzgerald et al., 1993; Sher, 1991; West and Prinz, 1987). However, not all children of alcohol-dependent parents develop such tendencies. In fact, a number of studies have shown that most children of alcohol-dependent parents do not develop alcohol or drug problems (from 60% to 75% according to Bearsdale, Son and Vaillant, 1986). Moreover, some studies report no problems which are specific to children of alcohol-dependent parents, in comparison to children of non-alcohol-dependent parents (Alterman et al., 1989, Bates and Pandina, 1992; Gilles and Hesselbrock, 1992; Pandina and Johnson, 1989). Such disparate findings suggest, at the very least, that children of alcohol-dependent parents are extremely heterogeneous.

As stated earlier, such heterogeneity may stem from methodological differences or from the fact that the degree of severity or comorbidity in parental alcoholism is highly variable. For example, Jacob and Leonard (1986) compared resilient children of alcohol-dependent parents with non-resilient children of alcohol-dependent parents, as well as with children of depressive parents. Based on their results, parents of non-resilient children present more mental health and alcohol related problems than parents of resilient children, confirming the possibility that children in the latter group are not exposed to the same level of risk as those in the former.

The heterogeneous outcomes observed among children of alcohol-dependent parents may also stem from variability in children's personal attributes and in their degree of exposure to the kinds of positive experiences (of a compensatory or protective nature) which constitute the protective or resilience and compensatory² factors which interest us here. Unfortunately, as we have already stated, there are far fewer studies on successful adaptation in children of alcohol-dependent parents (that is, studies of resilient children), than there are studies on maladaptive outcomes in such children. Moreover, few studies have employed a longitudinal-prospective design to deal specifically with children of alcohol-dependent parents (fewer still on children of drug-dependent parents or compulsive gamblers). We were able to locate approximately ten studies which looked at resilient children of alcohol-dependent parents, but the methodological qualities of these studies appear to be highly variable.

Among these studies, the oldest is that of Rydelius (1981), a Swedish researcher who tracked a group of children of alcohol-dependent fathers over a twenty-year period. Unfortunately, the absence of a control group of children of non-alcohol dependent fathers, prevents us from extracting elements of resilience or protection according to the definitions previously presented in this paper. The second and better known study, that of Werner (1986), deals with children of alcohol-dependent parents on the Island of Kauai. Werner compared groups of resilient and non-resilient children of alcohol-dependent parents at age 18 and found that resilient children scored higher in terms of easy temperament, intelligence, language skills, self-control and academic performance, beginning in infancy (at age 1 and 2), as well as during childhood (at age 10) and adolescence (at age 18). The author noted other advantages which the resilient children enjoyed, such as a positive relationship with an attachment figure, less socio-economic stress, and less turmoil within the family unit particularly during the first two years of life. The results of this study are inspiring, although hardly conclusive, owing to methodological shortcomings in at least two areas. As the Rydelius study (1981), the absence of a control group of child of non-alcohol-dependent parents makes it impossible to determine whether the presumed resilience factors constitute a more salient distinction between resilient and non-resilient children of alcohol-dependent parents than between resilient and non-resilient children of non-alcohol-dependent parents. Moreover, in the Werner study (1986), the resilient children of alcohol-dependent parents were primarily girls (21 out of 29), while the non-resilient children of alcohol-dependent parents were primarily boys (14 out of 20). Since the process of adaptation is different for boys than for girls, gender alone may explain differences in resilience among children of alcohol-dependent parents. However, Werner (1986) also took into account any prenatal and perinatal complications and problems to which the children of alcohol-dependent parents might have been exposed. Consequently, this factor can safely be eliminated as a potential explanation for the differences in resilience among children exposed to parental alcoholism.

Werner and Smith (2001), who have tracked their groups of resilient and non-resilient children into adulthood (at ages 32 and 40), point out that their resilient subjects appear to have freed themselves better emotionally from the problems which continue to plague their parents, while the non-resilient subjects remained entangled in their parents' problems. This was a surprising finding, even for the authors: one would have expected to find less detachment from parents among resilient offspring. It is important to note that the groups of resilient and non-resilient children in Werner and Smith's recent study (2001) are not limited to children of alcohol-dependent parents, as was the case in the 1986 study, but also include offspring of parents with mental health problems, long-standing marital problems, or marked socio-economic difficulties.

McCord (1988) overcame the deficiencies of the earlier studies by incorporating a group of children of non-alcohol dependent parents and by focusing exclusively on sons of alcohol-dependent fathers. She also shed light on the role of the mother in these situations. She observed that the sons of mothers who show a high degree of esteem for the alcohol-dependent father (a situation which is more common in intact families), are more likely to develop drinking problems. However, when mothers show little tolerance for their partners' alcoholism and provide adequate supervision for their children, their sons appear to be as well-adjusted as those of non-alcohol dependent parents. In another set of studies, McCord (1990; 1991) observed that sons who were separated from their alcohol-dependent or criminal fathers were less likely to become delinquent than their peers who remained in intact families where the father presented similar problems.

McCord's findings are corroborated by those of Dobkin, Tremblay, Desmarais-Gervais and Dépelteau (1994) who found that the sons of alcohol-dependent parents whose families had remained intact were more likely to seek the services of a psychologist than were their counterparts from broken homes. In the case of children of non-alcohol dependent parents, the situation was reversed: those from broken homes were more likely to seek the services of a psychologist. Other studies also report that a positive relationship between mother and child of separated families is preferable for the development of the child than intact families where conflicts are numerous and other adverse conditions prevail (Rutter, 1971; West and Farrington, 1973). On the other hand, the work of Wolin and Bennet (Bennet, Wolin and Reiss, 1988b; Wolin, Bennet, Noonan and Teitlebaum, 1980), shows that children of alcohol-dependent parents whose families maintain certain rituals (birthdays and other celebrations, family vacations) are less susceptible to developing a drinking problem than those whose family rituals are disrupted by the behaviour of alcohol-dependent parents, a finding which adds nuance to the preceding results. However, the absence of a comparison group of children of non-alcohol-dependent parents limits the conclusions which can be derived from Wolin and Bennet's work.

The fourth study compared children whose alcohol-dependent parents had no psychiatric problems and children whose alcohol-dependent parents presented one or more psychiatric disorders (as defined in the DSM) (Reich et al. 1988). The authors found that the resilient children of solely alcohol-dependent parents had a better relationship with their non-drinking parent than did the non-resilient children whose parent's alcoholism was compounded by other problems; children in the first category were also less exposed to parental conflict and domestic disturbances.

The next two studies which were reviewed, both included a control group of well-adjusted children of non-alcohol dependent parents and a group of maladjusted children of non-alcohol dependent parents, established on the basis of the same criteria used to determine a group of resilient children of alcohol-dependent parents and a group of non-resilient children of alcohol-dependent parents. Moreover, the group of children of alcohol-dependent parents were matched with the group of children of non-alcohol dependent parents for several socio-demographic variables. In both studies, the presence of paternal alcoholism is based on a self-report screening tool (Short Michigan Alcohol Screening Test, Selzer et al., 1975), corroborated by a clinical diagnostic interview based on DSM criteria. Finally, both studies adopted a longitudinal perspective with respect to the presumed resilience factors, but also limited themselves to a single temporal sequence to verify the presence or absence of adjustment problems in children (boys only).

In a study conducted in Montreal by Vitaro et al. (1996) with boys from under privileged backgrounds, four groups of children of alcohol-dependent-parents and four groups of children of non-alcohol-dependent parents were established, based on whether or not the children manifested the following behaviours: psychotropic drug abuse, delinquent behaviour, and academic failure at age 14. One of the specifics of this study was that it sought to distinguish among children of alcohol-dependent parents those who presented a single problem from those who demonstrated several problems. The study also called upon teachers and peers to evaluate presumed resilience factors, thus eliminating any possible bias associated with self-evaluation or parental evaluation. Results with respect to resilience factors were collected at three different points in the children's lives (ages 6, 10 and 12). These results show that the mothers of resilient children of alcohol-dependent parents (that is, children with no problems of delinquency, substance abuse or academic failure) were better educated and exercised better supervision over their sons than did the mothers of children with adjustment problems, which corroborates the findings of McCord (1988). The results also show that resilient children of alcohol-dependent parents are perceived by their teachers and their peers as being less hyperactive or inattentive and less aggressive than maladjusted children of alcohol-dependent parents. Resilient children are also perceived as being more sociable and are more popular with their classmates; furthermore, they perform better in school at ages 10 and 12. However, no interaction between the resilience groups (no problems, a single problem, several problems) and the presence or absence of paternal alcoholism was detected, such that the differences previously cited between resilient children of alcohol-dependent parents and non-resilient children of alcohol-dependent parents can also be found among children of alcohol-dependent parents. In other words, the elements uncovered in this study are not resilience elements in the strict sense, but rather compensatory elements.

A study conducted in Michigan by Zucker et al. (2003), compares well-adjusted children of alcohol-dependent and non-alcohol dependent parents with maladjusted children of alcohol-dependent and non-alcohol dependent parents, with no distinction as to the number or nature of the problems experienced by the latter. The authors define the absence or presence of adjustment problems by drawing a distinction between children whose score falls on either side of the 80th percentile on a composite measure of behavioural problems evaluated by the parents while their children were between the ages of 3 and 5. Resilient children are those that obtain a score lower than the 80th percentile on the scale of behavioural problems. All the children are subsequently re-evaluated at ages 6-8, 9-11 and 12-14. Interestingly, the authors take into account the degree of severity of the fathers' alcoholism and the presence or absence of anti-social tendencies in the fathers. In order to increase the validity of their study concerning internal dynamics, the authors excluded subjects affected by foetal alcohol syndrome. However, the study's external validity is compromised by the fact that it is exclusively directed at children living with both biological parents.

The results show that, at ages 3-5, resilient children of alcohol dependent parents received higher scores for intellectual functioning than their non-resilient counterparts, but did not distinguish themselves from well adjusted children of non alcohol-dependent parents. When temperament was evaluated at the same age, the resilient children of alcohol-dependent parents showed a lower level of emotional reactivity than non-resilient children of alcohol-dependent parents. Finally, at ages 3-5, resilient children of alcohol-dependent parents also presented fewer behavioural problems and a higher level of cognitive functioning than maladjusted children of alcohol-dependent parents. However, the same children of alcohol-dependent parents who were initially deemed resilient at ages 3-5 presented more externalized and internalized behavioural problems at ages 9-11 and ages 12-14 than the well-adjusted children of non-alcohol-dependent parents. These last findings demonstrate that resilience should not be viewed as a lasting characteristic since children of alcohol-dependent parents who were deemed to be resilient in early childhood nonetheless went on to develop adjustment problems later in life. In conclusion, this study is interesting for its methodological qualities which reveal the limits of an evaluation on resilience which would rest on a single measurement sequence and on a single source of information.

The penultimate study which was examined is that of Vitaro, Tremblay and Zoccolillo (1999). In this study, the criteria used to constitute a group of resilient children of alcohol-dependent parents were based on abstinence from drugs or alcohol at ages 15-16. Non-resilient children born of alcohol-dependent parents were divided into three sub-groups: subjects presenting an alcohol problem only; subjects presenting a drug problem only; and lastly, subjects presenting both forms of dependency. Only paternal alcoholism was taken into consideration, given the low number of alcohol-dependent mothers (the latter were, in fact, excluded from the study). Resilience factors were measured in childhood (ages 6, 7 and 8), and in pre-adolescences (ages 10, 11 and 12). These factors relate to three dimensions of personality, based on the Cloninger's model (1986): novelty seeking, harm avoidance and reward (or sensitivity to positive reinforcement). These factors also encompassed certain parental practices such as supervision, as well as socio-familial adversity. What is unique and interesting about this study is the very large sample of boys and girls reflective of Quebec youth. As with the preceding studies, the results of this study revealed that, in Quebec, the presence of an alcohol-dependent father substantially increases the risk of using psychotropic drugs in adolescence, even when confounding variables such as socio-familial adversity and gender are taken into account. However, low novelty seeking and the presence of harm avoidance behaviours in childhood and pre-adolescence tend to reduce this risk in mid-adolescence. Such personal elements must be viewed as beneficial or compensatory factors rather than as protection or resilience factors, since they reduce the risk of psychotropic drug use in all young people, irrespective of the presence or absence of paternal alcoholism. On the other hand, strong maternal supervision would appear to be a protective factor in that it constitutes a more salient distinction between resilient children of alcohol-dependent parents and alcohol-or drug-dependent children of alcohol-dependent parents than between well-adapted children of non-alcohol-dependent parents and alcohol-or drug-dependent children of non-alcohol-dependent parents. In statistical terms, maternal supervision produces a moderating effect on the predictive relationship between paternal alcoholism and alcohol or drug use in adolescence. This moderating effect translates into a significant interaction between maternal supervision and paternal alcoholism. It also corresponds to the definition of protective or resilience variables, as discussed earlier.

Finally, the last study deals with children prenatally exposed to alcohol (Olson, O'Connor and Fitzgerald, 2001). Although such exposure increases the prevalence of so-called internalized problems in children throughout their pre-school development (elevated levels of negative emotional feelings, few positive relationships, and significant levels of depression), it would appear that a healthy interaction between mother and child during the pre-school years has a moderating effect on these harmful consequences.

²Given the limited number of studies dealing with the factors of resilience or protection in children of alcohol-dependent parents, we have decided to include studies which examine compensatory factors since these can also lead to relevant intervention strategies or targets. We have done the same for studies which deal with children of drug-dependent parents.