The Roles of Mothers versus Fathers in Intergenerational Family Risk: A Commentary on ^{Rothenberg et al. (2022)}

Prospective intergenerational studies conducted over the past 2 decades have established that parents’ conduct problems during their own childhood are significantly related to the occurrence of these same problems in their children and have identified family risk factors implicated in this association (Conger et al., 2009). Whereas these intergenerational associations are usually found to be relatively modest in size, the findings have been replicated in numerous samples and across multiple forms of such behaviors (e.g., substance use, intimate partner violence) indicating relatively large impacts from a whole-population, public health perspective. Further, intergenerational associations are not always of modest size. In a recent study of intergenerational associations in crime with the Oregon Youth Study sample of parents and sons, we found that the odds ratio for having two or more adult arrests (vs. 0 or 1 arrest) across Generations 1 and 2 (G1 and G2; as indexed by police records) was five times greater than expected by chance (Capaldi et al., 2021). The body of basic research on intergenerational risk indicates a crucial need to intervene to reduce the transmission of risk across generations.

The Fast Track Study as a whole represents a notable achievement in many respects, including in the original secondary prevention aim of reducing future conduct problems in children already showing higher levels of conduct problems in Grade 1 by intervention throughout the school years, and in following this sample to adulthood (age 34 years) to examine whether intervention effects relate to more positive environments in the family of procreation. Conduct problems are a major domain of psychopathology symptoms that cast long shadows stretching into consequences in adulthood for the individuals themselves, for their families, and for others with whom they come into contact (Mason et al., 2004). They are relatively intractable due to a number of factors, such as family risk environments and the underlying characteristics of undercontrol and impulsivity that are key traits of individuals with higher levels of conduct problems (Gottfredson & Hirschi, 1990). Further, it is very challenging to follow such a sample well into their adult years. The value to prevention science and the wider community of such long-term data providing prospective information of intergenerational effects of a major well-designed and implemented prevention effort cannot be overstated. This work is the “gold standard.”

The major finding of the Rothenberg et al. study (2022) regarding effects of the Fast Track school years intervention on later family of procreation risk is that the intervention in childhood had significant effects on a number of aspects of adult adjustment and family environment for the women who had become mothers by age 34 years—including their depression, substance use problems, food insecurity, and use of corporal punishment with their children; no intervention effects were supported within participating fathers. As noted by the authors, this is in keeping with the findings of Rothenberg (2016) that the intergenerational process, including the emergence of the externalizing pathway, is stronger for G2 women than for men. Demonstrating intervention effects with mothers is important, given that women still carry out a larger proportion of child rearing and spend more time with their children than men.

We would also speculate that an intervention that improves mothers’ adjustment and reduces their use of corporal punishment also may benefit their partners, as we have found within at-risk families. In a study with the Oregon Youth Study Intergenerational Sample (Capaldi et al., 2008) of focal men in G2 and their women partners, we found that men’s poor and harsh discipline toward their children was influenced by the poor discipline they had received as a child but, in addition, was influenced by both their female partner’s risk behavior (antisocial behavior and substance use) and her poor and harsh discipline toward their child/children. This was the case even though the model also included additional risk factors, namely the G2 family socioeconomic status, the G2 father’s age at the birth of his first biological child, and the G2 father’s risk behavior. This indicates a second mechanism by which mothers are more influential on family environments than are fathers. It appears that at-risk men may take their parenting cues partly from their partners’ parenting. This may well be because men at least implicitly acknowledge that mothers have the primary parenting role; that they should follow their lead; and that mothers play a gatekeeping role in shaping, facilitating, or in some cases limiting fathers’ influences. Women also may have more experience dealing with young children even prior to parenthood compared with men (e.g., assisting in caring for siblings, babysitting for evening and summer jobs).

When considering the null intervention effects on physical aggression and partner violence by mothers, and on any outcome for fathers, it is important to note that power to detect effects may have been constrained by the nonparticipation of the highest risk individuals. Specifically, participating parents had lower externalizing risk scores as kindergarteners than did nonparticipants. Additionally, an inclusion criterion was that parents had to live at least 20% of the time with their child or otherwise be actively involved in parenting. This is sensible as the present focus was on family and parent functioning. However, given that an estimated five out of six single-parent American families in the U.S. are mother only, this criterion meant that fathers were less likely to be captured in this analysis than mothers (see Weinraub & Kaufman, 2019). Indeed, 58% of the age-34-years parent sample was male, compared to 69% of the original sample, and one would expect that the poorest functioning fathers (e.g., those showing this study’s violence and drug problem outcomes) would be least involved in their children’s lives and thus least represented here. In Hill and colleagues’ (2020) study of intergenerational intervention effects, they similarly found that mothers “were consistently more likely to meet eligibility criteria (regular, face-to-face contact with their G3 child) than … fathers …,” and in Riggs et al. (2009), the parent sample became disproportionately female over time, from 50% at enrollment to approximately two thirds at the time of the parent survey. Thus, understanding and accounting for the attrition of fathers, and perhaps testing noninvolvement in children’s lives as an outcome of interest, may inform our interpretation of these important studies. Struggling mothers tend to stay in their children’s lives and remain under our research lens, whereas struggling fathers may not.

Rothenberg et al. (2022) found that the unexpected association of family chaos with being in the intervention versus control group was highly related to the number of children in the family, even though this latter difference was relatively small (2.11 intervention vs. 2.01 control). A further issue to consider here is whether this finding was related to having stepchildren, which can add a number of challenges for parents, including managing communications with the ex-partner (other biological parent) regarding issues such as child support, children’s movements between families, relationships with stepsiblings, and sometimes children’s rejection of the stepparent. This may be particularly challenging for adults who had higher levels of conduct problems in childhood, who therefore may have lower levels of the social and parenting skills needed to address such complex family systems issues. Still another consideration was whether family chaos might be related to being a single parent, an already difficult job made more difficult by associated economic disadvantage. A further assumption that can be made based on American demography is that a significant portion of the participating mothers were single parents, whereas most participating fathers had a co-parent (see Weinraub & Kaufman, 2019).

Findings of this study of intergenerational effects of Fast Track have a number of prevention implications. First, such a sustained prevention effort for children with elevated levels of conduct problems across the school years can have beneficial effects not only for those children themselves in earlier adulthood (Dodge et al., 2015) but for their future family of procreation. This adds a major benefit to such interventions in terms of long-term value—both in human and economic terms. Second, the effect for the women, along with other work showing the influence of women on men’s parenting (Capaldi et al., 2008), indicates that even though the intergenerational effect was not found for men, interventions for women may have positive influence on men’s parenting, and thus additional intergenerational benefits. Third, some support beyond high school for at-risk youth, especially at the time that they are expectant or new parents (Feinberg, 2002), may provide a positive boost that can reinforce prior progress and increase beneficial intergenerational effects. Fourth, more studies need to be conducted to increase the understanding of how to improve long-term outcomes for at-risk men and functioning in the father role, including how to examine those whose involvement is inconsistent, limited, or altogether absent. For example, we know that men are likely to improve in conduct problems and associated behaviors with age but especially when they partner and at the time of first fatherhood; however, younger fathers do not experience such beneficial effects (Kerr et al., 2011). Thus, supporting delay of first biological fatherhood until men are more mature can have positive effects for the family of procreation.

The findings of Rothenberg et al. (2022) indicate that sustained school-year interventions with youth with higher levels of conduct problems can have beneficial effects on the behavior of the women who participated 18 years after the end of the intervention, and that the prevention effects will benefit their family of procreation. This is an exceptional achievement given the seriousness of this cluster of behaviors.