The long-term effects of parental alcoholism in African American and white women from intact families

Maryann Amodeo
Boston University School of Social Work
264 Bay State Rd.
Boston, Mass. 02215  USA
phone:  617-353-7916
fax: 617-353-5612
e-mail: mamodeo@bu.edu
 

Mediating factors for the long-term effects of parental alcoholism in women: The contribution of other childhood stresses and resources

Margaret Griffin
Boston University School of Social Work
264 Bay State Rd.
Boston, Mass. 02215  USA
phone:  617-353-7228
fax: 617-353-5612
e-mail:  mlgrif@bu.edu
 

Testing a sexual abuse severity scale to predict adult psychosocial     outcomes in African American and white women

Irene Fassler
Boston University School of Social Work
264 Bay State Rd.
Boston, Mass. 02215  USA
phone:  617-353-3763
fax: 617-353-5612
e-mail:  irfaon@bu.edu

Maryann Amodeo
Boston University School of Social Work
264 Bay State Rd.
Boston, Mass. 02215  USA
phone:  617-353-7916
fax: 617-353-5612
e-mail:  mamodeo@bu.edu
 

Social supports among African American and white women with and without  alcoholic parents

Cassandra Clay
Boston University School of Social Work
264 Bay State Rd.
Boston, Mass. 02215  USA
phone:  617-353-3752
fax: 617-353-5612
e-mail:  cmclay@bu.edu
 

Recruiting African American women in community research: The nuts and bolts of a successful effort

Michael Ellis
Boston University School of Social Work
264 Bay State Rd.
Boston, Mass. 02215  USA
phone:  617-353-7226
fax: 617-353-5612
e-mail:  maellis@bu.edu

The heterogeneity of offspring of alcoholic parents has been established in the literature; however, predictors of the wide range of positive and negative outcomes in adulthood have not been identified.  The goal of this symposium is to describe a 3-year study that addresses this gap in the literature.  We used a model of stresses and resources in childhood to predict adult psychosocial outcomes in a community sample of 290 women.  Stresses examined were parental alcoholism, parental psychiatric disorder, domestic violence, and sexual and other physical abuse; and resources included social supports and family cohesion, conflict, expressiveness, and organization.  Adult outcomes assessed were alcohol problems and several psychosocial measures:  overall social adjustment, psychiatric disorder, depressed mood, self-esteem, life satisfaction, and satisfaction with social support.  Standardized, self-administered questionnaires and face-to-face interviews conducted by experienced interviewers were used to collect data, with siblings as collateral informants.  The balanced design required equal numbers of white and African American women and equal numbers of offspring of alcoholic and nonalcoholic parents.

The first paper assesses mediating factors in the relationship between parental alcoholism and adult psychosocial outcomes, revealing the need for social work researchers to be aware of the misleading findings about effects of parental alcoholism in bivariate analyses.  The second paper investigates the predictive validity of a severity scale for childhood sexual abuse, showing the need for social work researchers to examine the family context.  The third paper compares social supports by race and parental alcoholism, questioning the assumption by some clinicians and researchers that African American women have a broader social support system than white women.  The fourth paper describes techniques for recruiting female African American subjects from a community population in which African Americans are underrepresented, showing that researchers can increase their success in recruiting African American women by implementing recommended strategies.


Mediating factors for the long-term effects of parental alcoholism in women:  The contribution of other childhood stresses and resources

Purpose:  Predictors of the heterogeneous outcomes in adulthood among offspring of alcoholic parents are not clear.  Serious methodological flaws in the literature include samples limited to college students or clinic patients; small sample size; an absence of control groups, control variables, and multivariate analyses; and lack of attention to the range of stresses and resources in childhood.  The primary aim of this paper was to identify stresses and resources in childhood that mediate the relationship between parental alcoholism and outcomes in women.  The current study examined stresses and resources in childhood and adult outcomes among 290 community-dwelling women.

Methods:  Stresses examined were parental psychiatric disorder, domestic violence, and sexual and other physical abuse; and resources examined included social supports and family cohesion, conflict, expressiveness, and organization.  Adult outcomes examined were alcohol problems and several measures of psychosocial adjustment.  Standardized measures and a face-to-face interview were used to collect information, with siblings as collateral informants.

Results:  The initial bivariate analyses showed that, compared to women without alcoholic parents, women with alcoholic parents had more additional stresses and weaker resources in childhood, as well as worse adult outcomes.  However, multivariate regression models showed that the effect of parental alcoholism on a wide range of adult outcomes is only indirect, mediated by the other stresses and resources examined.  These findings are reported by race.

Implications:  Assumptions about family functioning have often been made by human service providers in the absence of empirical data.  Childhood resources, typically overlooked in the literature, and sexual abuse were found to have a direct effect on adult outcomes, while parental alcoholism was not statistically significant for predicting any of the adult outcomes.  Findings will assist health care providers in refining childhood risk assessments and in directing these efforts to serve those most in need.


2.  Testing a sexual abuse severity scale to predict adult psychosocial outcomes in African American and white women

Purpose:  Childhood sexual abuse was one of the stresses assessed in this study.  Since reports of sexual abuse varied considerably in pattern and intensity, we sought a severity measure to predict adult psychosocial outcomes.  A literature review identified single variables used to measure severity, but consensus on the most predictive variables was not found.  Further, inadequate attention was paid to the context of the abuse such as family environment.

Methods:  Special efforts were made to maximize subject reliability: a sexual abuse history was elicited via questionnaire and interview, and siblings were used as collateral informants.  28% of the subjects reported childhood sexual abuse.  We developed a 6-item severity scale, assigning subjects one point for each of the following:  the occurrence of any childhood sexual abuse, more than one episode, more than one perpetrator, incest, rape, and the perpetrator’s residence within the subject’s household.  We assessed the predictive validity of this scale on adult outcomes.

Results:  Data analysis showed that the severity scale was no better at predicting outcomes than a dichotomous variable measuring the presence or absence of sexual abuse.  We next examined the context of the abuse using 4 scales of family environment.  Two family resources were found to increase prediction of psychosocial outcomes:  conflict and expressiveness.

Implications:  Our data support an ecological perspective of childhood sexual abuse, suggesting that environmental factors may be more important in predicting adult outcomes in sexual abuse victims than the nature of the episode itself.  Recommendations include consideration of family environment in addition to details of the sexual abuse episode for researchers and clinicians to assess long-term adult outcomes.


3.  Social supports among African American and white women with and without  alcoholic parents

Purpose: Social support has been shown to be predictive of outcomes ranging from psychosocial adjustment to physical health.  Except for special populations, little empirical data has examined social supports by race or compared subjects with and without alcoholic parents.  While qualitative studies have reported that African Americans are strongly connected to family and friends, more recent empirical data calls this into question.  Our three-year study examined race differences in social supports and adult outcomes among 290 community-dwelling women.

Methods: Self-administered questionnaires were used to collect data, including several social support scales.  The balanced design required equal numbers of white and African American women and equal numbers of offspring of alcoholic and nonalcoholic parents.  Findings were adjusted for age and social class.

Results:  Women with alcoholic parents reported smaller social networks than women without alcoholic parents.  African American women reported smaller social networks, but attended religious services more often than white women.  White women with alcoholic parents scored lower on network satisfaction than the remaining subjects.  As hypothesized, social supports were significantly related to outcomes, including overall social adjustment, self-esteem, life satisfaction, depressed mood, and psychiatric problems, but not to alcohol problems.

Implications: Common beliefs that African Americans are more connected to friends and family than whites may be inaccurate.  Types of social supports and satisfaction did vary by race, as well as by parental alcoholism, suggesting that these variables need to be accounted for in research on social support.  Since social supports did predict adult outcomes regardless of race, human service providers working with women with alcoholic parents could focus on improving social supports as a way to improve other adult outcomes.


4.   Recruiting African American women in community research:  The nuts and bolts of a successful effort

Purpose:  Representation of ethnic minority groups in studies is crucial because research and evaluation provide the basis for service planning and resource allocation.  Moreover, NIH now requires the inclusion of minority subjects in all research samples.  While clinical samples present a captive audience for researchers, participation of subjects from nontreatment populations requires active outreach efforts.  This paper describes successful strategies used by social work researchers to recruit a heterogeneous, community sample with similar numbers of African American and white subjects.

Methods:  Recruitment consisted of posting flyers in dozens of community locations such as hair and nail salons, libraries, check cashing centers, ice cream shops, and health clinics; contacting African American college organizations; collaborating with members of black churches; newspaper and internet advertising; and providing incentives for subjects to recruit peers.  Recruitment required approximately 6 hours/week over the 18-month recruitment phase.  The research team monitored respondent characteristics at biweekly meetings and shifted strategies as needed.  Since the protocol required telephone screening, the subject pool had to be considerably larger than the actual participation:  1140 telephone calls yielded 290 respondents.

Results:  Our sample was 46% African American in a geographic area with 11% African Americans.  We were able to achieve socioeconomic and age diversity within each race.  White and African American groups had similar household incomes, with equal representation of subjects reporting high, medium, and low incomes.  African Americans in higher income brackets, a group especially difficult to recruit, were well represented.  Within our targeted age range of 21-60 years, all decades were represented for both races.

Implications:  Researchers can benefit from practical information about the time, funds, oversight, and creativity required for recruiting community samples.  The approaches presented for African American women may be useful for recruiting other ethnic minority groups.