|Home | About | Journals | Submit | Contact Us | Français|
This study examined parenting satisfaction among a clinical sample of 88 men with alcoholism at baseline and 6-month and 12-month follow-up. Findings indicated no overall improvements in parenting satisfaction from the outset of treatment across assessments. Alcohol consumption variables were associated with parenting satisfaction at the bivariate level, and changes in alcohol consumption from pre to post treatment emerged as a significant predictor of changes in parenting satisfaction over the course of the study. Greater decreases in alcohol consumption were associated with greater improvement in parenting satisfaction.
Parenting satisfaction is associated with several important elements of successful care-giving, including a sense of parenting efficacy, positive parent-child interactions, and healthy attachment (e.g., Owen, Thompson, & Kaslow, 2006). Low parenting satisfaction has been linked to less supportive parenting and harsh discipline techniques (Simons, Beaman, Conger & Chao, 1993). Families of fathers with alcoholism are at risk for problematic parenting and negative child outcomes (Eiden, Leonard, Hoyle, & Cavez, 2004), and children appear to benefit from reductions in fathers’ alcohol use following treatment (Burdzovic Andreas et al., 2006). Reductions in alcohol use may lead to more positive parent-child interactions, and in turn, greater parenting satisfaction and less child behavior problems.
Despite this literature indicating that problematic drinking is associated with negative parenting behaviors and child behavior problems, we are aware of no published study that has examined problem drinking and parenting satisfaction. We examined patterns and predictors of parenting satisfaction among a clinical sample of alcoholics. It was predicted that parenting satisfaction would increase upon initiation of treatment for alcoholism and across a 6-month and 12-month follow-up. Higher alcohol use was expected to be associated with poorer parenting satisfaction at the bivariate level, and reductions in alcohol use were expected to be associated with increases in parenting satisfaction across time points.
The sample consisted of 88 men who had recently begun an alcoholism treatment program. Participants were recruited from four substance abuse treatment programs in Massachusetts. At the baseline assessment, men reported living with their partners an average of 9.7 years. A mean of 2.2 children were living in the home. On average, male participants were 38.8 years old. Approximately 84% of male participants were Caucasian. Children were an average of 7.6 years old (SD = 3.5; range = 1 to 15). Participants attended substance abuse treatment sessions an average of 11.0 days (SD = 7.9 days) in the 30 days before the baseline interview, 15.9 days (SD = 32.5 days) in the 180 days before the 6-month follow-up, and 8.5 days (SD = 15.3 days) in the 180 days before the 12-month follow-up. Please see O’Farrell, Fals-Stewart and Murphy (2003) for details regarding recruitment and inclusion/exclusion criteria.
Alcohol consumption was measured with the Time Line Follow-Back interview (TLFB; Sobell & Sobell, 1996). The number of days participants used any amount of alcohol and the number of days participants engaged in heavy alcohol use were examined in this study.
We conducted a series of mixed-effects regression analyses, which offers numerous advantages to analyzing nested data structures (Singer & Willett, 2003). The current study produced a structure in which repeated measures (level-1) were nested within individuals (level-2). In addition to the unstandardized regression coefficients we report a correlation coefficient based on the amount of level-1 or within-participants variance accounted for (r) and partial regression coefficients (pr) for comparison on a common metric. First, mixed-effects regression models with time included as dummy coded variables were conducted to document change over time in all study variables (analogous to a repeated measures ANOVA within this analytic framework). Table 1 provides the estimates for the means, standard deviation, and 95% confidence intervals for each variable at each time point. Examining the 95% confidence indicated that number of days engaged in drinking and number of days engaged in heavy drinking exhibited significant decreases from baseline to the 6-month follow-up, with no further change from the 6-month follow-up period to the 12-month follow-up. Parenting satisfaction did not exhibit significant change across the assessment occasions.
The next set of analyses examined bivariate associations between the alcohol use variables and parenting satisfaction by including each predictor (separately) as a Level-1 (i.e., time varying) predictor of parenting satisfaction, which produces estimates of the predictor-parenting satisfaction association when taking into account the nested data structure. Both days of alcohol use (r = −.42, pr = −.35, p < .01) and days of heavy alcohol use (r = −.42, pr = −.35, p < .01) were negatively associated with parenting satisfaction.
To examine change in parenting satisfaction across the assessment periods, time (coded 0, 1, and 2 for the baseline, 6-month, and 12-month follow-up assessments, respectively) was entered into the level-1 regression equations predicting parenting satisfaction (i.e., standard growth curve analysis). Regressed difference scores1 were then entered into the level-2 regression equation to evaluate whether change in parenting satisfaction across the three time periods differed as a function of the change in the predictor variables from baseline to 6-month follow-up. Changes in number of days engaged in drinking (b = −.01, t = −2.86, p < .01, pr = −.29) and number of days engaged in heavy drinking (b = −.02, t = −2.41, p < .05, pr = −.25) predicted change in parenting satisfaction across the three assessment periods. As shown in Figure 1, those exhibiting larger decreases in days engaged in drinking (labeled responders) showed increases in parenting satisfaction across the assessments, while individuals who exhibited increases in days engaged in drinking (labeled non-responders) showed decreases in parenting satisfaction. Follow-up analyses revealed that changes in parenting satisfaction from baseline to the 6-month follow-up period did not predict changes in days engaged in drinking (b = −2.04, t = −1.35, ns; pr = .02) nor changes in days engaged in heavy drinking (b = −1.63, t = −1.28, ns; pr = .02).
Contrary to our hypothesis, findings suggested no overall improvements in parenting satisfaction from the outset of treatment across the 6-month and 12-month follow-ups. However, as predicted, changes in parenting satisfaction were moderated by changes in alcohol consumption. These findings suggest that reductions in alcohol consumption are associated with improved parenting satisfaction. This is consistent with prior research suggesting that alcoholism is associated with reduced quality of parent-child interactions (e.g. Eiden et al., 2004), as well as studies that have demonstrated that alcoholics’ reduction and cessation in drinking after treatment is beneficial to children (Burdzovic Andreas et al., 2006; Kelley & Fals-Stewart, 2002). Fathers with alcoholism may benefit from the incorporation of familial relationship-enhancing elements in standard interventions (e.g., O’Farrell & Fals-Stewart, 2006).
Current findings may not be generalizable to women with alcohol problems. Also, because many fathers had several children of different ages, we were unable to control for the age of the child in the analyses. Parenting satisfaction may change depending on the age of the child. This study relied on a relatively limited self-report measure of parenting satisfaction. Future investigations should incorporate a more comprehensive assessment that also examines predictors of other domains of parenting, including parenting behavior. Despite these limitations, findings provide some insight on parent-child relationships among fathers with alcoholism.
1Raw difference (change) scores are strongly influenced by initial levels of the variable in interest. Therefore, regressed change scores from pre to post treatment were computed for each of the predictor variables (Cohen, Cohen, West, & Aiken, 2003) to examine how changes in the predictor variables were related to changes in parenting satisfaction.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.