Skip to main content
PMC home page
Alcohol and Alcoholism (Oxford, Oxfordshire) logoLink to Alcohol and Alcoholism (Oxford, Oxfordshire)
. 2018 Aug 22;53(6):707–715. doi: 10.1093/alcalc/agy059

Women’s Self-Perceived Similarity to Their Mother and Associations with Patterns of Alcohol Misuse over 20 Years

Amelia E Talley 1,✉, Mackenzie L Hughes 1, Sharon C Wilsnack 2, Tonda L Hughes 3
PMCID: PMC6203126  PMID: 30137187

Women’s perceived similarity to their mother/father was examined as a factor influencing associations between perceived parental alcohol use during childhood and patterns of alcohol misuse in adulthood. Women’s self-perceived similarity to their frequent- or problem-drinking mothers increased the risk of negative consequences from drinking over time as well as potential symptoms of alcohol dependence over time.

Abstract

Aims

This study examined transgenerational transmission of risk for female alcohol misuse. Women’s perceived similarity to their mother/father in adulthood was examined in terms of its influence on the expected association between perceived maternal alcohol use and female offsprings’ trajectories of alcohol misuse. We hypothesized that a daughter’s self-perceived similarity to her mother, in instances where her mother was perceived to be a frequent- or problem-drinker, would be associated with an increase in the daughter’s count of negative consequences from alcohol use and potential symptoms of alcohol dependence across adulthood.

Short summary

Women’s perceived similarity to their mother/father was examined as a factor influencing associations between perceived parental alcohol use during childhood and patterns of alcohol misuse in adulthood. Women’s self-perceived similarity to their frequent- or problem-drinking mothers increased the risk of negative consequences from drinking over time as well as potential symptoms of alcohol dependence over time.

Methods

Analyses utilized data from a survey of women (N = 911) who were followed over a 20-year period, beginning in 1981. Women, ages 21 or older and living in households in the contiguous USA, were eligible, and women who consumed four or more alcoholic drinks per week were oversampled. Model estimates were weighted to adjust for the oversampling of heavier drinking women and to reflect national demographics. Latent growth mixture models estimated regression parameters that captured variation in participants’ alcohol misuse over time.

Results

Women who reported that their mother was a frequent- or problem-drinker and who perceived themselves to be similar to their mother, in general, showed increases in alcohol misuse. The same pattern of results was not shown for fathers.

Conclusions

Results support that interventions seeking to reduce female alcohol misuse should address the role of perceived similarity to heavy-drinking female role models or ‘female-drinker’ prototypes to reduce problem-drinking behavior among female drinkers.

INTRODUCTION

Among the almost 67% of adult women in the USA who report past-year alcohol use, 24% state that their typical consumption exceeds NIAAA’s weekly recommended limit of seven or fewer drinks, placing them at risk for a potential alcohol use disorder (AUD) and other alcohol-related consequences (NIAAA, 2008; Esser et al., 2014). Recent population-based studies indicate that women’s reported alcohol consumption and frequency of experiencing negative alcohol-related consequences have increased in recent decades (Grant et al., 2004; White et al., 2015). Thus, a greater number of mothers are potentially exposing their children to alcohol misuse and associated consequences (White et al., 2015). It is essential, given evolving cultural norms related to female alcohol use, to understand how the children of heavy-drinking women fare in terms of long-term patterns of alcohol misuse.

Previous work has shown that perceived maternal drinking behavior informs subjective norms (i.e. descriptive and injunctive norms) regarding alcohol use (Neighbors et al., 2007), with more permissive parental norms being positively related to women’s own reported alcohol use (e.g. Keefe, 1994; Maglica and Dorčić, 2016). The association between permissive alcohol norms and greater alcohol consumption may be strengthened by perceptions of similarity to heavy-drinker prototypes (Gerrard et al., 2008). To date, empirical research has not explored fully the role of perceived similarity to a heavy-drinking female role model in exacerbating risk of female alcohol misuse. We sought to examine whether perceived maternal drinking behavior positively related to participants’ prospective reports of alcohol misuse throughout adulthood, as reflected, in the current report, by negative consequences from drinking and potential symptoms of alcohol dependence. We also examined whether perceiving their mothers to be frequent- or problem-drinkers during their childhood increased risk of alcohol misuse among adult women who reported being more similar to their mothers.

Findings from previous studies support that the drinking behavior of adult offspring has been found to be reliably predicted by perceptions of parental alcohol use (Newcomb et al., 1983; Wilks et al., 1989). For example, studies have shown a positive association between mothers’ alcohol misuse and heavy alcohol consumption in their female offspring (e.g. Lieb et al., 2002). In particular, parental role modeling of alcohol use may relate more reliably to women’s alcohol use than men’s (Choquet et al., 2008). Alcohol misuse among women, in particular, is more robustly predicted by childhood environments in which mothers, more so than fathers, are frequent or problem drinkers (Dawson, 2000; Grant et al., 2001; Lieb et al., 2002). Given findings suggesting that perceptions of mothers' alcohol consumption impacts daughters’ risk of alcohol misuse, it is important to understand factors that link mothers’ and daughters’ drinking behaviors.

Women’s perceived similarity to a significant female role model who drank heavily or frequently during their early childhood may influence their subsequent alcohol-related behavior in adulthood (Perry and Bussey, 1979; Blanton et al., 1997; Gerrard et al., 2002; Rivis et al., 2006; Zimmermann and Sieverding, 2010; Lane et al., 2011). Previous work (Fromme and Ruela, 1994) suggested that only when perceived similarity to mothers or fathers was high, on five dimensions unrelated to alcohol use (i.e. personality, religiosity, physical appearance, recreational interests, political views), were associations between perceived parental drinking behavior and young adult offspring’s drinking behavior statistically significant. These authors suggested that perceived similarity to a parent who drinks could increase offsprings’ willingness to engage in alcohol use.

Prototype willingness model

The prototype willingness model (PWM; Gerrard et al., 2008) is a parallel-process theory that has sought to disambiguate reasoned motivation from reactive motivation to improve prediction of health-related behaviors, including alcohol misuse. The PWM extended expectancy-value models (e.g. Theory of Planned Behavior, Ajzen, 1991) by explicating heuristic or socially reactive motivating factors that increase willingness to engage in a specific behavior, such as alcohol misuse, in addition to reasoned factors that increase volitional intentions to engage in that behavior (Todd et al., 2016).

In the PWM (Gibbons et al., 1998), prototypes are usually evoked to characterize a representative person in a particular setting (e.g. ‘female drinker’). Prototypes are defined as ‘images of the type of person who engages in the target behavior’ (Todd et al., 2016, p. 2). Prototypes are typically assessed with regard to both general favorability ratings and perceptions of similarity to one’s self-concept. As applied to alcohol use behavior, the PWM suggests that prototypes for alcohol use are not universally positive, but positive evaluations of and perceived similarity to a ‘drinker prototype’ may potentiate willingness to consume alcohol. The implication is that a greater willingness to engage in alcohol use, and risk of misuse, may be more likely among female participants who see themselves as similar to a female role model perceived to be a heavy drinker (Rivis et al., 2006; see also Courtenay, 2000). In the current work, we view women’s perceived similarity to a heavy-drinking female role model (i.e. their mother) as a proxy for perceived similarity to a ‘female drinker.’ Reactive pathways, delineated in the PWM, suggest that perceptions of similarity to female drinker prototypes exacerbate risk of alcohol misuse (Gerrard et al., 2008).

Among women in particular, bivariate correlations shown in Zimmermann and Sieverding (2010) suggest that perceived similarity to a drinker prototype relates to both past and subsequent drinking behavior. Unfortunately, perceptions of similarity to a ‘drinker prototype’ were unreliably associated with women’s alcohol consumption in multivariate models. The authors argue that gender differences in prototypes of the ‘typical drinker’ may have contributed to equivocal results among females. Specifically, because heavy drinking is more often regarded as a masculine behavior (Courtenay, 2000), and the prototype invoked in the study was not explicitly defined according to gender (i.e. ‘Imagine now the typical person who consumes … several (>3) glasses of alcohol …’), the imagined prototype may have been that of a typical male drinker. To the extent an adult woman reports being more similar to her mother, who functioned as a gender-specific role model and frequently engaged in alcohol consumption or had drinking problem, we would hypothesize a stronger association between perceived maternal alcohol use and risk of alcohol misuse.

Current study and hypotheses

We conducted secondary analysis of data from an existing survey of US women (N = 911) who were followed over a 20-year period (Wilsnack et al., 2006). We hypothesized that, among participants who identified their mother as a frequent- or problem-drinker during childhood, perceived similarity to their mothers, as adults, would be associated with an increase in their reported alcohol consequences over a subsequent 20-year period. By contrast, among participants who identified their mothers as abstainers or occasional drinkers during their childhood, or who did not view themselves as similar, in general, to their frequent- or heavy-drinking mother, we expected normative declines in drinking across adulthood. That is, women’s alcohol misuse typically peaks in young adulthood and steadily declines into middle and older adulthood (Wilsnack et al., 2006; White et al., 2015). Alternatively, we did not expect that similarity to fathers, even in cases where they were identified as frequent or heavy drinkers, would influence the association between participants’ perceptions of paternal drinking and changes in her alcohol misuse over time. This last prediction is based on the logic that paternal figures would not commonly serve as gender-specific role models for women’s drinking (c.f., Zimmermann and Sieverding, 2010).

METHODS

Participants

The National Study of Health and Life Experiences of Women (NSHLEW) was the first national, longitudinal survey study to focus on predictors and consequences of women’s drinking and drinking-related problems across the adult lifespan. An initial US national probability sample was drawn by the National Opinion Research Center (NORC) in 1981. Women who were age 21 or older and living in households in the contiguous USA were eligible for enrollment. All eligible women who reported consuming four or more alcoholic drinks per week (moderate-to-heavy drinkers) were selected into the sample, whereas only one of every four who reported consuming less than four drinks or who abstained from drinking was selected. Interviews were conducted by highly trained female interviewers. The first wave included 498 moderate-to-heavy-drinking women and 413 lighter-drinking and non-drinking women (see Wilsnack et al. [2006] for further details). Ten years later, in 1991, survey staff re-interviewed 695 women from the 1981 cohort (86% of all 1981 participants who were living and not incapacitated). Twenty years after the baseline survey, in 2001, survey staff re-interviewed 483 of the original 1981 cohort (66% of 1981 participants who were living and not incapacitated). Previous analyses (Wilsnack et al., 2006) of these data suggest that they can be used to inform typical patterns of alcohol use behaviors among women across the lifespan. Attrition analyses reported by the investigators showed that older women and those less educated were more likely lost in follow-up assessments. There were no differences on drinking outcomes between participants lost versus retained.

Measures

Perceptions of parental drinking

Perceptions of parents’ drinking were measured at baseline (1981; i.e. Wave 1 of the NSHLEW study) with a single, subjective question about perceptions of mothers’ and fathers’ drinking, respectively, when the participant was growing up (‘until age 14’). The questions were asked very early in the survey, assessing aspects of the participant’s family of origin (e.g. family demographics) and upbringing (e.g. religious beliefs). Response categories were 1 (Never drank), 2 (Drank occasionally), 3 (Drank frequently), 4 (Had drinking problem). Given sparsity of some cells and to improve stability of model estimates, we collapsed response category 3 (Drank frequently) and 4 (Had drinking problem) for both mothers and fathers, resulting in two trichotomized variables representing perceived maternal and perceived paternal drinking, respectively. In analyses examining perceived similarity to mothers, the father drinking variable was centered (with ‘drank occasionally,’ the most frequently endorsed category, as the zero point) and included as a covariate. In analyses examining perceived similarity to fathers, the same procedure was used but with the mother-drinking variable coded such that the zero point corresponded to mothers who ‘drank occasionally.’

Perceived similarity to mother/father

Women’s perception of similarity to their parents was assessed only once, at baseline, asking ‘Thinking about yourself as an adult, would you say you are more like your father or more like your mother?’ Response categories were ‘more like my father’ and ‘more like my mother.’ When participants volunteered alternative options, such as ‘like neither’ or ‘some ways like both,’ those responses were coded as well. This question was given toward the beginning of the survey—embedded in a series of questions about the participants’ current general self-concept and directly following a section asking them to ‘Describe Yourself.’ A total of 71 respondents indicated that their mother was a frequent or problem drinker during their childhood. Of these women, 31 reported being more like their father and 28 reported being more like their mother, with only three reporting they were ‘like neither’ and nine reporting they were, in some ways, like both. In analyses examining perceived similarity to mothers, women who reported they were ‘more like father’ or ‘like neither’ were coded as −1; ‘some ways like both’ was coded as 0; ‘more like mother’ was coded as 1. In analyses of perceived similarity to fathers, women who reported they were ‘more like mother’ or ‘like neither’ were coded as −1; ‘some ways like both’ was coded as 0; ‘more like father’ was coded as 1 (Sensitivity analyses (available upon request) that recoded ‘like neither’ to missing replicated the current pattern and significance of findings.).

Alcohol misuse

Alcohol misuse was assessed at each 10-year assessment with a series of questions about past-12-month experiences of eight problem consequences (driving a car while ‘high’ from alcohol; drinking-related accidents in the home; harmful effects of drinking on housework/chores, or on job/career opportunities; drinking-related problems with partner or with children; and starting fights with partner or with people outside the family when drinking). As in Johnson et al. (2013), any level of endorsement of a particular negative consequence was coded as present (vs. absent) in the previous 12 months; the number of negative consequences endorsed was summed across the eight items.

Serving as a second indicator of alcohol misuse and to conceptually replicate analyses examining negative consequences, five potential symptoms of alcohol dependence (AD) were assessed. A series of questions at each data collection point, asked about potential symptoms of AD in the past-12-months (blackouts, rapid drinking, morning drinking, inability to stop drinking before becoming intoxicated and inability to stop or cut down on drinking over time). As with negative consequences, a count variable summed the number of potential symptoms endorsed as having occurred, to any extent, across the five items.

Covariates

All analyses controlled for two variables: participant age (continuous) and drinking status of the parent who was not the focus of the test of moderation (ordered categorical). Preliminary models also included participant ethnicity and income but because neither was found to alter effects or conclusions, they were subsequently dropped from analyses.

Data analytic approach

Analyses were conducted in MPLUS v. 7.11 (Muthén and Muthén, 2017). We included a survey weight variable in analyses to adjust estimates for differences in selection probabilities, to take into account variations in nonresponse rates by sampling unit, and to correct for the oversampling of heavier drinking women (Wilsnack et al., 2006). Application of survey weights ensures that estimates reflect, as closely as possible, national demographics (age, education, marital status and race/ethnicity).

Latent growth model of drinking outcomes

Women who were alcohol abstainers were included in all analyses, increasing the number of zero count-values in the data. Models with count outcomes typically employ Poisson regression. Poisson regression models can include overdispersion parameters (Min and Agresti, 2005), which take into account an overabundance, or inflation, of zero values (e.g. a large proportion of alcohol-abstainers). These types of regression models have been termed Zero-inflated Poisson or ZIP models. ZIP models are two-part models, wherein a single set of predictors is used to estimate (a) the occurrence of excessive zeros in count variables (e.g. women reporting zero negative consequences due to alcohol abstention) and (b) counts of consequences/symptoms among drinkers, who are expected to report non-zero count values (Atkins et al., 2013). Because ZIP models allow for large numbers of zero count-values, they are useful in analyzing variables with highly skewed distributions, such as women’s alcohol consumption, where a large number of women typically abstain (and, thus, would have de facto zero negative alcohol-related consequences; Liu, 2007).

ZIP regression models can be extended to longitudinal designs by estimating latent growth trajectories (Liu, 2007). Latent growth models estimate intercept (i.e. baseline) and slope (i.e. change) parameters that capture variation in participants’ alcohol-misuse trajectories over time. Initially, an unconditional model, excluding covariates, is used to fit to a linear slope function. Latent growth models were used in the current study to estimate patterns of negative alcohol-related consequences among participants, from Wave 1 (1981) through Wave 3 (2001). The average level of baseline negative consequences, at Wave 1, is estimated as the intercept. The average, linear change over time in the count of negative consequences is indicated by the estimated slope parameter (Bollen and Curran, 2006). Although all model estimates are reported in tables, results in the text focus on prospective changes in negative alcohol-related consequences (i.e. slope parameters) over a 20-year period in adulthood, henceforth explicitly referred to as the ‘estimated count of negative alcohol-related consequences.’ All models were specified using full-information maximum likelihood estimation.

After latent growth parameters are fit to the data, covariates can be regressed on the latent growth factors to determine the relation between a given covariate and alcohol-misuse trajectories over time. The current models adjusted for age of the participant and perceived maternal and paternal drinking behavior during her childhood (trichotomized), both measured at baseline only, when estimating prospective trajectories of negative consequences from alcohol use. In analyses, self-perceived similarity to mothers, perceived maternal drinking behavior and the corresponding interaction between the two were entered as predictors of latent growth intercept and slope factors. In separate analyses, self-perceived similarity to fathers, perceived paternal drinking status and the corresponding interaction between these two variables were entered as predictors of intercept and slope factors, respectively, in latent growth models (a three-way interaction, among perceived maternal drinking behavior, perceived paternal drinking behavior and similarity to mother/father was tested and found to be non-significant).

RESULTS

Descriptive statistics

Of the 911 women interviewed in the 1981 NSHLEW survey, 33 participants were excluded from current analyses because of missing data on the predictor variables. Models examining past-year alcohol consequences excluded four additional participants due to missing outcome data (n = 874); models examining past-year potential symptoms of AD excluded seven additional participants due to missing outcome data (n = 871). Therefore, a total of 874/871 participants (~96% of the 1981 Cohort) were included in the final analyses. Sample and descriptive statistics are provided in Table 1.

Table 1.

Baselinea descriptive statistics

Variable M (SD) n (%)
Age 42.69 (16.38)b
Race/ethnicity
 Non-Hispanic White 742 (85%)
 Non-Hispanic Black 92 (11%)
 Hispanic 27 (3%)
 Asian/pacific islander 4 (<1%)
 American Indian/Alaskan native 4 (<1%)
 Did not report 2 (<1%)
MO drinker status
 Abstainer 426 (49%)
 Occasional drinker 374 (43%)
 Frequent drinker 50 (6%)
 Problem drinker 21 (2%)
FA drinker status
 Abstainer 200 (23%)
 Occasional drinker 438 (50%)
 Frequent drinker 137 (16%)
 Problem drinker 96 (11%)
Similarity
 More like their mother 380 (44%)
 More like their father 311 (36%)
 Like neither 47 (5%)
 Like both 133 (15%)
Alcohol-related consequences count
 1981 0.50 (1.00) 874
 1991 0.24 (0.67) 686
 2001 0.23 (0.69) 479
Potential AD symptom count
 1981 0.31 (0.74) 871
 1991 0.14 (0.51) 683
 2001 0.15 (0.57) 478
Open in a new tab

Note. aStatistics were culled from Wave 1 data, unless otherwise noted. MO = Mother, FA = Father. bAge range: 20–86 years old. (In the 1991 interview, two respondents reported that they were enrolled into the study in 1981 at age 20, despite the fact that age-eligibility criteria were set at 21). Median income at baseline was $17,500–19,999.

Negative alcohol-related consequences

Similarity to mother

Standardized model estimates for negative alcohol-related consequences are contained in the left panel of Table 2. As predicted, perceived similarity to mother interacted with perceived maternal drinking to impact changes in estimated counts of negative alcohol-related consequences over 20 years. This interaction, depicted in Fig. 1, shows that women who perceived themselves to be more like their mother, in instances where the mother was perceived as a frequent- or problem-drinker, reported subsequent increases in expected counts of negative alcohol-related consequences over time. By contrast, all other participants reported normative reductions in the number of negative consequences into middle and older adulthood. (Sensitivity analyses (available upon request), which utilized truncated, count indicator variables in latent growth models (0, 1, 2, 3+ consequences), replicated primary findings, suggesting that outliers did not unduly influence the current results.)

Table 2.

Zero-inflated model estimates predicting negative consequences and potential AD symptoms

Negative consequences Potential symptoms of AD
Zero inflation Expected count Zero inflation Expected count
Estimate SE P Estimate SE P Estimate SE P Estimate SE P
Similarity to MO
 Age Baseline 0.001 0.072 0.000 −0.390 0.208 0.061 0.426 0.102 0.000 −0.663 0.196 0.001
Δ Across 20 years 0.802 0.731 0.273 0.280 0.880 0.750 0.888 0.192 0.000 0.965 0.068 0.000
 FA drinker status Baseline −0.250 0.075 0.001 0.204 0.163 0.211 −0.214 0.122 0.078 −0.173 0.199 0.384
Δ Across 20 years 0.465 0.549 0.397 0.473 0.246 0.055 −0.027 0.328 0.935 0.187 0.275 0.498
 MO drinker status Baseline 0.284 0.081 0.000 −0.120 0.160 0.453 −0.231 0.089 0.874 0.175 0.148 0.237
Δ Across 20 years 0.484 0.476 0.309 0.562 0.220 0.011 −0.056 0.356 0.874 0.000 0.190 0.998
 Similarity to MO Baseline 0.008 0.087 0.927 −0.098 0.128 0.441 −0.096 0.110 0.382 −0.282 0.173 0.102
Δ Across 20 years 0.147 0.507 0.772 0.136 0.305 0.655 0.617 0.421 0.142 0.394 0.236 0.094
 Similarity × MO drinker status Baseline −0.049 0.098 0.620 −0.033 0.168 0.843 −0.210 0.094 0.000 −0.030 0.179 0.867
Δ Across 20 years −0.077 0.573 0.894 0.538 0.191 0.005 0.366 0.444 0.410 0.467 0.197 0.018
Similarity to FA
 Age Baseline 0.400 0.079 0.000 −0.352 0.318 0.269 0.376 0.119 0.002 −0.619 0.199 0.002
Δ Across 20 years 0.672 0.834 0.420 0.201 0.802 0.802 0.597 0.299 0.046 0.839 0.218 0.000
 MO drinker status Baseline −0.257 0.100 0.010 −0.073 0.251 0.772 −0.154 0.089 0.084 0.253 0.096 0.008
Δ Across 20 years 0.531 0.527 0.313 0.624 0.307 0.042 −0.072 0.182 0.692 0.008 0.173 0.963
 FA drinker status Baseline −0.255 0.088 0.004 −0.211 0.163 0.194 −0.230 0.109 0.035 −0.212 0.159 0.181
Δ Across 20 years 0.434 0.538 0.432 0.274 0.371 0.461 0.017 0.288 0.953 0.195 0.290 0.501
 Similarity to FA Baseline −0.045 0.074 0.543 0.060 0.136 0.658 0.189 0.105 0.070 0.492 0.153 0.001
Δ Across 20 years 0.129 0.552 0.815 0.187 0.350 0.593 −0.805 0.220 0.000 −0.568 0.274 0.038
 Similarity × FA drinker status Baseline 0.062 0.084 0.460 0.124 0.138 0.368 0.105 0.091 0.249 0.143 0.146 0.330
Δ Across 20 years −0.202 0.507 0.691 −0.483 0.239 0.043 −0.075 0.264 0.776 −0.082 0.210 0.698
Open in a new tab

Note. Estimates with an association of P < 0.05 are in bold-face font. MO = Mother, FA = Father, Δ = Change. Standardized estimates are given as logits values. The ‘Zero Inflation’ part of the ZIP model is a logistic regression predicting log odds of a subpopulation of participants who are unable to report non-zero consequences, due to alcohol abstention. The ‘Expected Count’ part of the model is a Poisson regression, estimating the number of expected consequences (i.e. natural log of the expected count) in a subpopulation of participants who are able to report scores of zero and above (i.e. not alcohol abstainers).

Fig. 1.

Fig. 1.

Open in a new tab

Estimated trajectories of past-year negative alcohol-related consequences based on mother’s perceived drinking status during participant’s childhood. Note. Standardized model coefficients were used to create estimated trajectories. NSHLEW, National Study of Health and Life Experiences of Women; FA, Father; MO, Mother.

Similarity to father

Unexpectedly, participants’ perceived similarity to their father also interacted with perceived paternal drinking behavior to impact changes in estimated counts of negative alcohol-related consequences over 20 years, but not in a way that was consistent with hypotheses. The form of the interaction, shown in Fig. 2, did not mirror the interaction found when women perceived greater similarity to a frequent- or problem-drinking mother, nor in a way that was predicted from theory. Instead, results suggested that participants who reported an abstaining father during childhood and greater perceived similarity toward their father as an adult, showed the fastest declines in the expected count of negative consequences over time.

Fig. 2.

Fig. 2.

Open in a new tab

Estimated trajectories of past-year negative alcohol-related consequences based on father’s perceived drinking status during participant’s childhood. Note. Standardized model coefficients were used to create estimated trajectories. NSHLEW, National Study of Health and Life Experiences of Women; FA, Father; MO, Mother.

Replication with potential AD symptoms

Standardized model estimates for potential symptoms of AD are contained in the right panel of Table 2. Participants’ perceived similarity to mothers also interacted with perceived maternal drinking behavior to predict changes in estimated counts of potential symptoms of AD across 20 years (Fig. 3). The pattern of this interaction was similar to that shown for negative consequences in Fig. 1. By contrast, the interaction between self-perceived similarity to father and perceived fathers’ drinking was unrelated to expected changes in potential symptoms of AD across adulthood.

Fig. 3.

Fig. 3.

Open in a new tab

Estimated trajectories of past-year alcohol dependence symptom count based on mother’s perceived drinking status during participant’s childhood and self-perceived resemblance to parents. Note. Standardized model coefficients were used to create estimated trajectories. AD, Alcohol Dependence; NSHLEW, National Study of Health and Life Experiences of Women; FA, Father; MO, Mother.

DISCUSSION

Results were generally supportive of hypotheses in that women who perceived that their mothers were frequent- or problem-drinkers during childhood subsequently reported increases in their own negative alcohol-related consequences across adulthood (Abrams and Niaura, 1987; Fromme and Ruela, 1994). Moreover, findings support the tenets of the PWM (Gerrard et al., 2008; Gibbons et al., 2009) and suggest that perceived patterns of maternal drinking during childhood (i.e. injunctive norms) and reported similarity to a heavy-drinking female role model (i.e. drinker prototypes), combined to influence subsequent reports of adult women’s own negative alcohol-related consequences.

Findings are consistent with previous work suggesting that parental drinking behaviors inform individuals’ injunctive norms for alcohol use (e.g. Peterson et al., 1994) and subsequent risk of alcohol misuse (White et al., 2000; Lieb et al., 2002). Notably, alcohol norms have relevance for individuals’ intentions and willingness to use alcohol (Maglica and Dorčić, 2016); among adult women, perceptions of frequent or problematic maternal alcohol use may facilitate their own intentional and reactive alcohol misuse. Perceived patterns of maternal drinking are believed to inform daughters’ alcohol-related beliefs, including perceived norms and drinker prototypes (Gerrard et al., 2008; Gibbons et al., 2009), and may ultimately influence both intentional and reactive pathways to alcohol misuse.

Mother-daughter relationships function differently from father-daughter relationships (e.g. Leaper, 2002), so perhaps it is not surprising that women’s perceptions of maternal alcohol use relate to their own alcohol misuse (Windle and Windle, 2012), particularly in instances where they perceive themselves as similar to their heavy-drinking mother. As supported in meta-analytic studies, both perceived similarity to and favorability of a drinker prototype appear to increase risk of alcohol misuse by increasing willingness to drink and subsequently affecting actual drinking behavior (Todd and Van Lettow, 2016; Todd et al., 2016). Positive feelings toward a heavy-drinking mother also might exacerbate the relation between perceived maternal alcohol use and risky drinking behavior (Gerrard et al., 2008). Because we did not measure perceived favorability of mothers or father in early childhood or adulthood, additional work is needed to explore whether, in families where mothers are frequent or heavy drinkers, perceived favorability of one’s mother also operates as an influencer of the association between perceived parental drinking and women’s risk of alcohol misuse.

Findings were interpreted in light of the PWM, which is a general model meant to disambiguate reasoned motivation from reactive motivation, as informed by prototype perception, to improve prediction of health-related behaviors. Hypotheses in the current reported were contextualized in gendered terms, acknowledging that the predictors and correlates for women's and men's drinking are distinct (Stewart et al., 2009). These findings are among the first to examine the influence of self-perceived similarity to a heavy-drinking female role model on patterns of women’s alcohol misuse in adulthood. Unfortunately, current findings were unable to address whether similar effects would be found among adult men. Nevertheless, results replicated Fromme and Ruela (1994) in that perceived similarity to a parent influenced associations between perceived parental drinking histories and filial drinking outcomes. Given that Fromme and Ruela did not formally test gender interactions, their findings provided no opportunity to clarify whether a same-sex parent was more influential in filial drinking outcomes, as has been shown in some previous work with female participants (Dawson, 2000; Grant et al., 2001; Lieb et al., 2002). For example, Windle and Windle (2012) showed stronger effects for same-gender dyads, as opposed to opposite-gender dyads, in the intergenerational transmission of alcohol use behavior, providing preliminary support for the idea that same-gender effects may be more robust. Our findings suggest that perceived similarity to a heavy-drinking, female role model in adulthood may be a unique variable that exacerbates risk of women’s alcohol misuse over time. Future work using the PWB might benefit from consideration of whether hypotheses involving prototype perceptions should be contextualized, based on the gender of the respondent and/or the gender of the drinking prototype.

Cognitive schemas regarding alcohol use begin to develop in early childhood—as early as age three—and commonly pertain to levels of consumption based on gender (Zucker et al., 1995). Thus, beliefs, norms and schemas for female alcohol use behavior are likely informed by maternal alcohol consumption throughout early childhood as well as have influence on daughters’ alcohol use behavior in adulthood. Ultimately, it would be informative to examine covariations in mothers’ and daughters’ alcohol-related schemas, knowledge, and beliefs about alcohol use, particularly female alcohol use, to refine transgenerational theoretical models.

Limitations

The study had a number of limitations that should be considered when evaluating the results. Although this was a prospective, longitudinal study that spanned 20 years, perceptions of parental drinking were based on retrospective survey data. That is, a subjective, single-item indicator of perceived parental drinking during the participants’ childhood was assessed, and no objective markers of parental drinking were collected. Retrospective data have received criticism due to inherent limitations (e.g. Koriat et al., 2000). For instance, retrospective reports are dependent on memory of the event(s), which includes perception of the event, attempts at reconstructing or making-meaning of the event, and cognitive processes involved in the selection, recall and rehearsal of information (Koriat et al., 2000). Previous research, however, has demonstrated that adult participants’ retrospective reports of most childhood experiences demonstrate moderate to good reliabilities over time (Hardt et al., 2006). Further, parental substance abuse has demonstrated the highest reliability among the parental health outcomes assessed, including psychiatric diseases, prolonged parental physical illness, and chronic pain (Hardt et al., 2006). Moreover, efforts to validate retrospective reports of alcohol-related behavior suggest that dimensional judgements (i.e. of degree) may prove more reliable than dichotomized judgements of whether a particular alcohol-related behavior was observed (Hasin et al., 2015). Rather than using a formal diagnostic interview, we used a dimensional approach and calculated the number of past-year symptoms of potential AUD.

Given the wide age range of participants, it should be noted that some respondents were asked to recall their parents’ drinking from several decades or more ago. Previous criticisms of retrospective reporting (Hardt et al., 2010) have noted long intervals between an experience and the ostensible recall of that experience. Despite that some forgetting is likely to have occurred, previous research has documented similar long-term effects for adverse childhood experiences, using both prospective and retrospective reports (e.g. Hardt et al., 2010). To adjust for known bias in recall and reporting, model estimates in the current study adjusted for participant age in all analyses. Ultimately, however, conclusions are derived from adult women’s constructions of their childhood familial relationships and interpretations of past parental behaviors as they are recalled during the survey. Moreover, it should be noted that small sub-group sample sizes (e.g. those who perceived being more similar to a frequent- or heavy-drinking mother) may influence the stability of predicted trajectories over time.

Memories of historical events can also be influenced by contextual factors (e.g. mood) at the time of recall. We emphasize that the question regarding parental alcohol use appeared early in the interview, in the context of other questions pertaining to early childhood experiences and demographic features. We highlight the inherent subjective nature of these data given previous work showing that perceptions of parental drinking patterns in adulthood can be explained, in part, by individual-difference factors, such as the amount of alcohol consumed by the reporter (Fromme and Ruela, 1994). Despite the potential for recall bias on the basis of contextual factors, research suggests that the potential influence of state-based correlates may be overestimated (Brewin et al., 1993). That is, adults tend to underestimate the prevalence of negative childhood experiences; overestimation of these experiences is uncommon (Fergusson et al., 2000). In spite of limitations, retrospective reporting is quite common in the examination of associations among childhood adversities and adult outcomes (see e.g. Choi et al., 2017).

CONCLUSION

The early identification of women at risk for alcohol misuse is important for prevention and treatment programs that target women and families (e.g. Ashley et al., 2003). Findings of the present study suggest that perceived similarity to a heavy-drinking female role model (Lindgren et al., 2016) or ‘female drinker’ prototypes could be incorporated in intervention programs that seek to reduce problem-drinking behavior or improve problem recognition among female drinkers. There is promise for the prototypes of the typical ‘female drinker’ to be revised through women’s cognitive contemplation and written imagined characterizations of both the female-drinker prototype and the female-non-drinker prototype, as well as associated social comparison processes concerned with how similar the self is to each, respective prototype (Ouellette et al., 2005).

Women’s alcohol use in the USA, including prevalence, volume, frequency, and heavy episodic drinking, has increased over the last decade, and the rate of increase has been greater among women than men (Dawson et al., 2015). Current trends in women’s drinking behavior suggest that a greater number of heavy-drinking women are likely raising young girls. Self-perceptions of similarity to a heavy-drinking female role model, such as one’s mother, may potentially impact daughters’ risk of alcohol misuse. As the number of female drinkers in the USA continues to grow (Grant et al., 2001, 2004; White et al., 2015), the influence of genetic and psychosocial contributors to female alcohol misuse warrants further study. Understanding how long-held beliefs and schemas pertain to alcohol use behavior contribute to gendered drinking outcomes has promise for informing the treatment and prevention of AUDs among women.

FUNDING

We are grateful to Alethea Desrosiers and the late Susan Nolen-Hoeksema for their initial interest in this hypothesis and their early work with these data. Research activities were supported, in part, by Grant No. AA019974 (Amelia E. Talley, PI), Grant No. R01 AA004610 (Sharon C. Wilsnack, PI), and Grant No. R01 AA013328 (Tonda L. Hughes, PI) from the National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Alcohol Abuse and Alcoholism or the National Institutes of Health.

CONFLICT OF INTEREST STATEMENT

None declared.

REFERENCES

  1. Abrams D, Niaura R (1987) Social learning theory In Blane HT, Leonard KE (eds). Psychological Theories of Drinking and Alcoholism. New York: Guilford Press, 131–78. [Google Scholar]
  2. Ajzen I. (1991) The theory of planned behavior. Organ Behav Hum Dec 50:179–211. [Google Scholar]
  3. Ashley OS, Marsden ME, Brady TM (2003) Effectiveness of substance abuse treatment programming for women: a review. Am J Drug Alcohol Ab 29:19–53. [DOI] [PubMed] [Google Scholar]
  4. Atkins DC, Baldwin SA, Zheng C, et al. (2013) A tutorial on count regression and zero-altered count models for longitudinal substance use data. Psychol Addict Behav 27:166–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Blanton H, Gibbons FX, Gerrard M, et al. (1997) Role of family and peers in the development of prototypes associated with substance use. J Fam Psychol 11:271–88. [Google Scholar]
  6. Bollen KA, Curran PJ (2006) Latent curve models: A structural equation perspective. Hoboken: John Wiley & Sons. [Google Scholar]
  7. Brewin CR, Andrews B, Gotlib IH (1993) Psychopathology and early experience: a reappraisal of retrospective reports. Psychol Bull 113:82–98. [DOI] [PubMed] [Google Scholar]
  8. Choi NG, DiNitto DM, Marti CN, et al. (2017) Association of adverse childhood experiences with lifetime mental and substance use disorders among men and women aged 50+ years. Int Psychogeriatr 29:359–72. [DOI] [PubMed] [Google Scholar]
  9. Choquet M, Hassler C, Morin D, et al. (2008) Perceived parenting styles and tobacco, alcohol and cannabis use among French adolescents: gender and family structure differentials. Alcohol Alcohol 43:73–80. [DOI] [PubMed] [Google Scholar]
  10. Courtenay WH. (2000) Constructions of masculinity and their influence on men’s well-being: a theory of gender and health. Soc Sci Med 50:1385–401. [DOI] [PubMed] [Google Scholar]
  11. Dawson DA. (2000) The link between family history and early onset alcoholism: earlier initiation of drinking or more rapid development of dependence? J Stud Alcohol 61:637–46. [DOI] [PubMed] [Google Scholar]
  12. Dawson DA, Goldstein RB, Saha TD, et al. (2015) Changes in alcohol consumption: United States, 2001–2002 to 2012–2013. Drug Alcohol Depend 148:56–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Esser MB, Hedden SL, Kanny D, et al. (2014) Prevalence of alcohol dependence among US adult drinkers, 2009–2011. Prev Chronic Dis 11:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fergusson DM, Horwood LJ, Woodward LJ (2000) The stability of child abuse reports: a longitudinal study of the reporting behaviour of young adults. Psychol Med 30:529–44. [DOI] [PubMed] [Google Scholar]
  15. Fromme K, Ruela A (1994) Mediators and moderators of young adults’ drinking. Addiction 89:63–71. [DOI] [PubMed] [Google Scholar]
  16. Gerrard M, Gibbons FX, Houlihan AE, et al. (2008) A dual-process approach to health risk decision making: the prototype willingness model. Dev Rev 28:29–61. [Google Scholar]
  17. Gerrard M, Gibbons FX, Reis-Bergan M, et al. (2002) Inhibitory effects of drinker and nondrinker prototypes on adolescent alcohol consumption. Health Psychol 21:601–9. [DOI] [PubMed] [Google Scholar]
  18. Gibbons FX, Gerrard M, Blanton H, et al. (1998) Reasoned action and social reaction: willingness and intention as independent predictors of health risk. J Pers Soc Psychol 74:1164–80. [DOI] [PubMed] [Google Scholar]
  19. Gibbons FX, Houlihan AE, Gerrard M (2009) Reason and reaction: the utility of a dual focus, dual-processing perspective on promotion and prevention of adolescent health risk behaviour. Br J Health Psychol 14:231–48. [DOI] [PubMed] [Google Scholar]
  20. Grant BF, Dawson DA, Stinson FS, et al. (2004) The 12-month prevalence and trends in DSM-IV alcohol abuse and dependence: United States, 1991–1992 and 2001–2002. Drug Alcohol Depend 74:223–34. [DOI] [PubMed] [Google Scholar]
  21. Grant BF, Stinson FS, Harford TC (2001) Age at onset of alcohol use and DSM-IV alcohol abuse and dependence: a 12-year follow-up. J Subst Abuse 13:493–504. [DOI] [PubMed] [Google Scholar]
  22. Hardt J, Sidor A, Bracko M, et al. (2006) Reliability of retrospective assessments of childhood experiences in Germany. J Nerv Ment Dis 194:676–83. [DOI] [PubMed] [Google Scholar]
  23. Hardt J, Vellaisamy P, Schoon I (2010) Sequelae of prospective versus retrospective reports of adverse childhood experiences. Psychol Rep 107:425–40. [DOI] [PubMed] [Google Scholar]
  24. Hasin DS, Greenstein E, Aivadyan C, et al. (2015) The alcohol use disorder and associated disabilities interview schedule-5 (AUDADIS-5): procedural validity of substance use disorders modules through clinical re-appraisal in a general population sample. Drug Alcohol Depend 148:40–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Johnson TP, Hughes TL, Cho YI, et al. (2013) Hazardous drinking, depression, and anxiety among sexual-minority women: self-medication or impaired functioning? J Stud Alcohol Drugs 74:565–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Keefe K. (1994) Perceptions of normative social pressure and attitudes toward alcohol use: changes during adolescence. J Stud Alcohol 55:46–54. [DOI] [PubMed] [Google Scholar]
  27. Koriat A, Goldsmith M, Pansky A (2000) Toward a psychology of memory accuracy. Annu Rev Psychol 51:481–537. [DOI] [PubMed] [Google Scholar]
  28. Lane DJ, Gibbons FX, O’Hara RE, et al. (2011) Standing out from the crowd: how comparison to prototypes can decrease health-risk behavior in young adults. Basic Appl Soc Psych 33:228–38. [Google Scholar]
  29. Leaper C. (2002) Parenting girls and boys In Bornstein MH, Bornstein MH (eds). Handbook of parenting: Children and parenting, Vol. 1, 2nd ed.Mahwah: Lawrence Erlbaum Associates Publishers, 189–225. [Google Scholar]
  30. Lieb R, Merikangas KR, Höfler M, et al. (2002) Parental alcohol use disorders and alcohol use and disorders in offspring: a community study. Psychol Med 32:63–78. [DOI] [PubMed] [Google Scholar]
  31. Lindgren KP, Gasser ML, Werntz A, et al. (2016) Moderators of implicit and explicit drinking identity in a large US adult sample. Addict Behav 60:177–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Liu H. (2007) Growth curve models for zero-inflated count data: an application to smoking behavior. Struct Equ Model 14:247–79. [Google Scholar]
  33. Maglica BK, Dorčić TM (2016) Components of the prototype/willingness model as concurrent and prospective predictors of adolescents’ alcohol consumption. J Child Adoles Subst 25:159–68. [Google Scholar]
  34. Min Y, Agresti A (2005) Random effect models for repeated measures of zero-inflated count data. Stat Model 5:1–19. [Google Scholar]
  35. Muthén LK, Muthén BO (2017) Mplus User’s Guide, 8th edn Los Angeles: Muthén & Muthén. [Google Scholar]
  36. National Institute of Alcohol Abuse and Alcoholism (NIAAA) (2008) Alcohol: A Women’s Health Issue. Department of Health and Human Services. [Google Scholar]
  37. Neighbors C, Lee CM, Lewis MA, et al. (2007) Are social norms the best predictor of outcomes among heavy-drinking college students? J Stud Alcohol Drugs 68:556–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Newcomb MD, Huba GJ, Bentler PM (1983) Mothers’ influence on the drug use of their children: confirmatory tests of direct modeling and mediational theories. Dev Psychol 19:714–26. [Google Scholar]
  39. Ouellette JA, Hessling R, Gibbons FX, et al. (2005) Using images to increase exercise behavior: prototypes versus possible selves. Pers Soc Psychol B 31:610–20. [DOI] [PubMed] [Google Scholar]
  40. Perry DG, Bussey K (1979) The social learning theory of sex differences: imitation is alive and well. J Pers Soc Psychol 37:1699–712. [Google Scholar]
  41. Peterson PL, Hawkins JD, Abbott RD, et al. (1994) Disentangling the effects of parental drinking, family management, and parental alcohol norms on current drinking by Black and White adolescents. J Res Adolesc 4:203–27. [Google Scholar]
  42. Rivis A, Sheeran P, Armitage CJ (2006) Augmenting the theory of planned behaviour with the prototype/willingness model: predictive validity of actor versus abstainer prototypes for adolescents’ health-protective and health-risk intentions. Br J Health Psychol 11:483–500. [DOI] [PubMed] [Google Scholar]
  43. Stewart SH, Gavric D, Collins P (2009) Women, girls, and alcohol In Brady KT, Back SE, Greenfield SF, Brady KT, Back SE, Greenfield SF (eds). Women and addiction: A comprehensive handbook. New York, NY, US: Guilford Press, 341–59. [Google Scholar]
  44. Todd J, Kothe E, Mullan B, et al. (2016) Reasoned versus reactive prediction of behavior: a meta-analysis of the prototype willingness model. Health Psychol Rev 10:1–24. [DOI] [PubMed] [Google Scholar]
  45. Todd J, van Lettow B (2016) A closer look at prototypes: similarity, favourability, and the prototype willingness model. A response to the commentary of Gibbons and Gerrard. Health Psychol Rev 10:47–9. [DOI] [PubMed] [Google Scholar]
  46. White A, Castle IJ, Chen CM, et al. (2015) Converging patterns of alcohol use and related outcomes among females and males in the United States, 2002 to 2012. Alcohol Clin Exp Res 39:1712–26. [DOI] [PubMed] [Google Scholar]
  47. White HR, Johnson V, Buyske S (2000) Parental modeling and parenting behavior effects on offspring alcohol and cigarette use: a growth curve analysis. J Subst Abuse 12:287–310. [DOI] [PubMed] [Google Scholar]
  48. Wilks J, Callan VJ, Austin DA (1989) Parent, peer and personal determinants of adolescent drinking. Br J Addict 84:619–30. [DOI] [PubMed] [Google Scholar]
  49. Wilsnack RW, Kristjanson AF, Wilsnack SC, et al. (2006) Are U.S. women drinking less (or more)? Historical and aging trends, 1981–2001. J Stud Alcohol 67:341–8. [DOI] [PubMed] [Google Scholar]
  50. Windle M, Windle RC (2012) Intergenerational relations for drinking motives: invariant for same- and opposite-sex parent-child dyads? J Stud Alcohol Drugs 73:63–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Zimmermann F, Sieverding M (2010) Young adults’ social drinking as explained by an augmented theory of planned behaviour: the roles of prototypes, willingness, and gender. Br J Health Psychol 15:561–81. [DOI] [PubMed] [Google Scholar]
  52. Zucker RA, Kincaid SB, Fitzgerald HE, et al. (1995) Alcohol schema acquisition in preschoolers: differences between children of alcoholics and children of nonalcoholics. Alcohol Clin Exp Res 19:1011–7. [DOI] [PubMed] [Google Scholar]

Articles from Alcohol and Alcoholism (Oxford, Oxfordshire) are provided here courtesy of Oxford University Press

RESOURCES