Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Dec 1.
Published in final edited form as: Alcohol Clin Exp Res. 2013 Jul 29;37(12):10.1111/acer.12196. doi: 10.1111/acer.12196

Dimensions of Parental Alcohol Use/Problems and Offspring Temperament, Externalizing Behaviors and Alcohol Use/Problems

Kenneth S Kendler 1,2,3, Charles O Gardner 1,2, Alexis Edwards 1,2, Matt Hickman 4, Jon Heron 4, John Macleod 4, Glyn Lewis 4, Danielle M Dick 1,2,3
PMCID: PMC3855174  NIHMSID: NIHMS480534  PMID: 23895510

Abstract

BACKGROUND

Alcohol consumption (AC) and alcohol-related problems (AP) are complex traits. How many factors reflecting parental AC and AP are present in the large prospectively followed Avon Longitudinal Study of Parents and Children cohort? Would these factors be uniquely associated with various temperamental and alcohol related outcomes in the children?

METHODS

We factor analyzed multiple items reflecting maternal and paternal AC and AP measured over a 12 year period from before the birth of the child (n=14,093 families). We examined, by linear regression controlling for socio-economic status SES, the relationship between scales derived from these factors and offspring early childhood temperament, externalizing traits and adolescent AC and AP (n’s ranging from 9,732 to 3,454).

RESULTS

We identified 5 coherent factors: typical maternal AC, maternal AC during pregnancy, maternal AP, paternal AC, and paternal AP. In univariate analyses, maternal and paternal AC and AP were modestly and significantly associated with low shyness, sociability, hyperactivity, and conduct problems in childhood and early adolescence; delinquent behavior at age 15; and AC and AP at ages 15 and 18. AC and AP at age 18 were more strongly predicted by parental factors than at age 15. Maternal AC during pregnancy uniquely predicted externalizing traits at ages 4, 13 and 15.

CONCLUSION

Parental AC and AP are complex multidimensional traits that differ in their association with a range of relevant measures in their children. Controlling for background AC and AP, self-reported levels of maternal AC during pregnancy uniquely predicted externalizing behaviors in childhood and adolescence.

Keywords: ALSPAC, alcohol consumption, parental alcohol use, temperament, externalizing problems, fetal alcohol exposure

Alcohol consumption (AC) and the symptoms of alcohol use disorders (AUDs) (which we here term “Alcohol Problems (AP)”) are both strongly familial (Begleiter 1995;Cotton 1979). A range of twin and adoption studies in adults have suggested that genetic factors contribute substantially to resemblance in relatives for AP, although familial-environmental factors also play an etiologic role (Heath et al. 1997;Kendler et al. 1997;Cloninger et al. 1985). Studies in adolescence, by contrast, typically show that genetic effects on AC are modest and shared environmental influences considerably more important (Kendler et al. 2008;Koopmans et al. 1997;Viken et al. 1999).

A large body of work has examined children of parents with AUDs – so-called “children of alcoholics” (Sher 1991;Searles and Windle 1992;Harter 2000;Sher et al. 1991;Sher 1997;Lieberman 2000). In particular, children of alcoholics have been shown to have an elevated risk for a difficult temperament (Sher, Walitzer, Wood, & Brent 1991;Sher 1997;Harter 2000), internalizing symptoms and/or disorders (Chassin et al. 1999) and high levels of externalizing traits (Sher, Walitzer, Wood, & Brent 1991;Harter 2000;Chassin, Pitts, DeLucia, & Todd 1999;Harter 2000;Edwards et al. 2006) as toddlers, children and early adolescents, and increased rates of heavy AC and AP (Chassin, Pitts, DeLucia, & Todd 1999;Sher, Walitzer, Wood, & Brent 1991) in later adolescence and adulthood. Importantly, the transmission of APs across generations appears in nearly all studies to be partly mediated by the increased rates of externalizing traits in offspring of alcoholics with much less evidence that internalizing traits play a mediational role (Chassin, Pitts, DeLucia, & Todd 1999;Hussong et al. 1998;Lieberman 2000;Ohannessian and Hesselbrock 2008;Ohannessian and Hesselbrock 2007).

Prior studies have examined whether the negative effects of having a parent with APs are more potent for males versus females with generally mixed results (Harter 2000;Edwards, Eiden, Colder, & Leonard 2006;Sher, Walitzer, Wood, & Brent 1991;Belliveau and Stoppard 1995). As expected, in most such studies, the affected parent is the father. Limited research has been done comparing the impact of APs in the parents (e.g., (Belliveau & Stoppard 1995), with some evidence that the effects may be more severe for fathers (Stout and Mintz 1996). A limitation of this methodologic approach is that it dichotomizes complex patterns of alcohol use and misuse in parents into two categories (“non-alcoholic” and “alcoholic”) rather than utilizing potentially more informative quantitative indices of parental AP.

Intrauterine exposure to alcohol can produce a wide range of developmental anomalies making up the “fetal-alcohol syndrome.” (Riley et al. 2011) However, intrauterine exposure to doses of alcohol insufficient to produce the classic syndrome can still have appreciable behavior effects (Spear and Molina 2005) including increased risk for externalizing behaviors (Sood et al. 2001) and early AC and AP in adolescence (Baer et al. 1998;Yates et al. 1998).

In this report, we explore the extensive measures collected over multiple waves on parental AC and AP in the Avon Longitudinal Study of Parents and Children (ALSPAC) (Boyd et al. 2012). We have two major aims. The first is to clarify the structure of this risk domain. Is there only one robust dimension of parental AC and AP assessed over time or are multiple dimensions needed to capture these varied behaviors? Having clarified the structure of these measures, our second major aim is to examine their ability to predict a range of behaviors in their offspring, previously shown to be potentially abnormal in the children of alcoholics: early childhood temperament, externalizing traits from childhood to early adolescence, and AC and AP in mid to late adolescence. Our working hypothesis is that we will see multiple dimensions of alcohol-related behaviors in the parents which will differ in the observed pattern of associations with this range of traits in their offspring.

METHODS

ALSPAC is an ongoing population-based study investigating a wide range of environmental and other influences on the health and development of children (http://www.alspac.bris.ac.uk) (Boyd, Golding, Macleod, Lawlor, Fraser, Henderson, Molloy, Ness, Ring, & Davey 2012). All pregnant women resident in the Avon district of South West England with an expected date of delivery between 1 April 1991 and 31 December 1992 were invited to participate in the Avon Longitudinal Study of Parents and Children (ALSPAC). The achieved sample was 14,541 pregnant women (80% of those eligible) with 13,988 live infants at age 12 months. ALSPAC parents and children have been followed-up regularly since recruitment, with data obtained through questionnaires completed by mothers, children and teachers, and through clinical assessments. Full details of all measures, procedures, sample characteristics, and response rates are available at www.alspac.bris.ac.uk. Ethical approval for the study was obtained from the ALSPAC Ethics and Law Committee, and the Local Research Ethics Committee. To assist readers in understanding the temporal structure of the data analyzed in this report, we prepared figure 1 which shows the time of the individual assessments (in age in months of the ALSPAC proband offspring) for the four major domains considered in this report: parental AC and AP, offspring child temperament, offspring externalizing traits and offspring AC and AP.

Figure 1.

Figure 1

Open in a new tab

The time of the individual assessments (in age in months of the ALSPAC proband offspring) for the four major domains considered in this report: parental alcohol consumption (AC) and alcohol problems (AP), offspring child temperament, offspring externalizing traits and offspring AC and AP.

Factor Analysis

Our first goal was to clarify the underlying structure of the data collected on AC and symptoms of AUD in the mothers and their partners in the multiple assessments covering times from before the pregnancy until the ALSPAC proband was 12 years old. We selected 82 candidate variables related to alcohol consumption and abuse/dependence from interviews of mothers and their partners (hereafter fathers). Of these items, 43 were reports on alcohol-related behavior of mothers and 39 concerned alcohol-related behaviors in fathers. These items were made up of self-reports by mothers, self-reports by fathers, mother reports on fathers, and father reports on mothers. From the interview at age 8 months and beyond, the partner data was used only if the partner’s questionnaire was filled out by the biological father of the child. For the derived scores described below, we chose families where the missing data on both mother and partner were less than 50% of these candidate variables. This reduced the overall sample to 4231 mothers and 4231 partners. The items assessing alcohol abuse and dependence symptoms were binary, while the remaining variables were ordinal with a few quasi-continuous measures (e.g. counts of alcohol consumption in an interval of time). To avoid problems due to highly skewed variables and associated scaling issues, all these variables were converted to five level ordinal variables. Factor analysis was performed on these variables separately for mother and father to identify a minimal number of factors that adequately summarized the variables judged from overall fit of the factor model. Initial best estimates of the number of factors were based on scree plots. We then looked for factor models that made clinical sense and provided an adequate fit based on three fit indices which all reflected, from varying perspective, the model’s balance of explanatory power and parsimony: the Tucker-Lewis Index (TLI) (Tucker and Lewis 1973), the Comparative Fit Index (CFI) (Bentler 1990), and the root mean square error of approximation (RMSEA) (Steiger 1990). For the TLI and CFI, values between 0.90 and 0.95 are considered acceptable, and ≥0.95 as good. For the RMSEA, good models have values ≤ 0.05.

In the course of these factor analyses, we discarded variables that failed to provide substantial loadings on any of the common factors. We also found that variables derived from a tally of daily alcohol consumption for the past week (number of alcohol measures per week and maximum number of alcohol measures consumed in any day of the past week) were so highly correlated that if we included both sets of variables in the analyses these variables formed their own factor. Therefore candidate factor models were tested with weekly consumption measured at several times or maximum daily use measured at several times, but not both. We found that the weekly alcohol consumption measures worked best in factor models for mothers, but maximum daily use worked best in factor models for fathers.

Factor loadings were obtained using robust maximum likelihood estimation (with less restrictive assumptions on missing data) and a Crawford-Ferguson quartimax rotation. With this rotation we achieved very similar results with both MLR and WLSMV estimation.

Factor derived scores were created from these factors by converting each individual item score into a z-score. Then the mean of all z-scores from all non-missing items constituting a factor was used to create an overall factor mean score. For regression models the factor mean scores were normalized by the method of Blom (Blom 1958) and standardized to mean 0 and variance 1. Factor derived scores were calculated for all parents with any non-missing data as well as the proportion of missing data among the items used to construct the scores. This resulted in 14,093 families with scores on at least one of the parents.

Prediction of Traits in Offspring

Covariates for the regression models included sex of the offspring, age of mother at 12 weeks gestation, highest educational attainment of mother or partner, and a measure of social class based on occupation of mother or partner. The latter two items were measured at 32 weeks gestation. This reduced the sample used for regression models to 11,393 families. For regression all variables were normalized by the method of Blom and standardized to mean 0 and variance 1, so effect sizes are all based on a common standard metric. Regressions models were fit using weighted least squares where the weighting was based on the proportion of non-missing data from the mother, partner, or both mother and partner for models of factors derived on mother or partner or both mother and partner (for multivariate models).

To assess the validity and predictive capacity of the factors derived from parental AC and APs, we examined their association with 14 variables.

Early Childhood Temperament

As detailed elsewhere (Dick et al. 2012), we utilized information about temperament as reported by the mother at 38 months, 57 months and 69 months. Developmentally appropriate instruments were used (Carey 1977;Buss and & Plomin 1984;Elander and Rutter 1996). For this report, we factor analyzed factor derived sum scores we developed previously (Dick, Aliev, Latendresse, Hickman, Heron, Macleod, Joinson, Lewis, & Kendler 2012) to produce dimensions which contained five items each and reflected four trait domains: “activity”, “emotionality,” “shyness,” and “sociability.” Valid scores at each individual interview were obtained for those who answered at least 3 of the 5 items, and the overall score was based on standardizing all scores at individual interviews and taking the mean of available scores. The available sample size for these regression analyses was ~9732.

Conduct difficulties and hyperactivity at 42 months were assessed by the Revised Rutter Parent Scale for Preschool Children questionnaire (Elander & Rutter 1996) given to the mother using 8 and 4 items, respectively. Scores were standardized and the overall mean of the standardized scores was used. Available sample size for our regression analyses was ~9069.

Conduct symptoms at age 13 were assessed using maternal reports on the conduct problems scale of the Strengths and Difficulties Questionnaire (Goodman 1997). This scale consists of five items, such as “often lies and cheats” and “often loses temper”, and mothers are asked to report whether these statements are “not true”, “somewhat true”, or “certainly true”. This questionnaire is widely used and psychometrically valid (Smith and McVie 2003). We created a sum score across the items, with a higher score representing higher levels of conduct problems. In the case of partial data, individuals with at least 3 of the 5 items were included, and the sum of those items was scaled to be proportional to those with complete data. Available sample size for this variable was ~6236.

Antisocial Behaviors at age 15 were assessed by the administration to the adolescent during an in-person clinic a 22-item scale derived from the Edinburgh Study for Youth Transitions and Crime questionnaire (Smith & McVie 2003;Schuckit et al. 1997), aimed at measuring a variety of antisocial behaviors (e.g., breaking into a vehicle, shoplifting, etc.). Each item was scored 0-3. For those that answered at least half the items, the scores were summed and mean responses were multiplied by 22 for a potential score of 0 to 66. These scores were normalized and standardized as described above. The available sample size for this variable was ~4735.

Alcohol Consumption and Problems

At the age 15.5 in-person clinic, ALSPAC subjects were asked a series of questions about their alcohol consumption and associated alcohol-related symptoms/problems that correspond relatively well to the DSM-IV criteria for alcohol abuse and dependence. These 16 items were extracted from the Semi-Structured Assessment of the Genetics of Alcoholism interview, developed by the Collaborative Study on the Genetics of Alcoholism (Bucholz et al. 1994;Hesselbrock et al. 1999). Quasi-continuous variables were collapsed into ordinal variables and ordinal variables were collapsed to avoid extreme categories with few responses that would lead to excessively skewed factor scores. Exploratory factor analysis in MPLUS using Crawford-Ferguson quartimax rotation yielded two correlated factors (RMSEA=0.038, CFI=0.984, TLI=0.978, r=0.67). The first factor reflected alcohol consumption and the second problems associated with alcohol use. Factor-derived scores were constructed for individuals responding to at least 8 of the 16 items by converting individual variables to z-scores, and for those that answered at least half the items, the mean z-score per item were calculated and normalized, and standardized as described above. Available sample sizes for regression models were ~3996 for the consumption factor and ~4001 for the problems factor.

We conducted a factor analysis on a total of 25 items relating to alcohol use and misuse assessed at age 18 from the Self-Rating of the Effects of alcohol scale (Schuckit, Tipp, Smith, Wiesbeck, & Kalmijn 1997), the AUDIT (Babor et al. 2001), and DSM-IV symptoms of alcohol abuse and dependence (American Psychiatric Association 1994), using a quartimin rotation. Examination of eigenvalues and interpretability of factor loadings led to the selection of a 3-factor solution. The first factor captured initial ethanol sensitivity, while the second reflected current alcohol consumption and tolerance: current SRE items, as well as 3 AUDIT items pertaining to drinking frequency and quantity, load onto this factor. The third factor captured drinking problems, with the remaining AUDIT items and all the DSM items loading onto this factor. The current consumption and current problems were only modestly correlated (r=0.32). Rotated factor scores were used as outcome variables. For these analyses, we used only the second and third factors reflecting AC and AP, as we assumed the origins of alcohol sensitivity were unlikely to be related to parental behaviors. Available sample size for these variables for our regression analyses was ~3454.

Many statistical tests are presented in this paper. We regard as potentially significant only those with a p value of under 0.01.

RESULTS

Factor Structure of Maternal Alcohol Consumption and Problems

An analysis of the scree plot based on 4231 individuals and 34 variables indicated the presence of three factors, and these were readily interpretable and provided a good fit to the data (RMSEA=0.051, CFI=0.978 and TLI=0.973). Factor loadings by Crawford-Ferguson quartimax rotation are seen in table 1. The first factor had consistently high loadings on 12 variables which reflected maternal AC and drinks per week as measured by both self and partner report from before the pregnancy through age 145 months (figure 1). We called this factor typical maternal AC. It had a Cronbach’s standardized Alpha of 0.95.

Table 1.

Crawford-Ferguson Quartimax Rotated Factor Loadings for Variables on Maternal Alcohol Consumption and Alcohol Problems - Robust Maximum Likelihood Loadings

Variable Reporter 1
Maternal
Alcohol
Consumption		 2
Maternal Problem
Drinking		 3
Maternal Alcohol
Consumption
During Pregnancy
AC - before pregnancy Self 0.70 −0.17 0.29
AC - gestation 1-3 months Self 0.10 −0.02 0.73
AC - gestation - since baby
first moved		 Self 0.19 −0.09 0.83
AC - gestation - last two
months		 Self 0.41 −0.07 0.62
AC - 8 wks Self 0.65 −0.06 0.32
AC - 21 months Self 0.74 0.02 0.17
AC - 33 months Self 0.79 0.07 0.08
AC - 61 months Self 0.84 0.08 0.01
AC - 21 months Partner 0.75 0.04 0.12
AC - 33moths Partner 0.77 0.12 0.04
AC - 47months Partner 0.85 0.07 −0.02
AC - 73 months Partner 0.85 0.17 −0.06
AC - 110 months Partner 0.77 0.12 0.04
AP (past year) - 61 months Self 0.12 0.77 −0.09
AP (past 3 years) - 110
months		 Self 0.18 0.67 −0.15
AP (past 2 years) - 145
months		 Self 0.13 0.66 −0.02
AP (past 2 years) - 110
months		 Partner 0.19 0.73 −0.18
AP (past 2 years) - 145
months		 Partner 0.30 0.56 −0.09
BD - 18 weeks gestation Self −0.29 0.47 0.68
BD - 32 weeks gestation Self −0.23 0.49 0.63
BD - 8 weeks Self −0.02 0.53 0.27
BD - 8 months Self 0.09 0.51 0.23
BD - 33 months Self 0.10 0.62 0.17
BD - 61 months Self 0.20 0.59 0.13
BD - 21 months Partner 0.13 0.54 0.18
BD - 33 months Partner 0.12 0.64 0.14
BD - 73 months Partner 0.20 0.64 0.04
BD - 110 months Partner 0.16 0.62 0.05
D/Wk - 8 wks gestation Self 0.11 0.00 0.73
D/Wk - 18 wks gestation Self 0.11 0.00 0.79
D/Wk - 32 wks gestation Self 0.32 −0.05 0.66
D/Wk - 85 months Self 0.63 0.24 0.02
D/Wk - 97 months Self 0.62 0.26 0.01
D/Wk - 145 months Self 0.54 0.30 0.00
Open in a new tab

AC – Alcohol Consumption

AP – Alcohol Problems

BD – Binge Drinking

D/Wk – Drinks per week

The second factor had consistently high loadings on 15 variables which assessed, via self and partner report, binge-drinking and alcohol problems from the time of the pregnancy through age 110 months. We termed this factor maternal AP. The third factor had high loadings on six self-report measures which reflected AC and drinks/week during gestation. We called this factor maternal AC during pregnancy. It had a Cronbach’s standardized Alpha of 0.88.

The typical maternal AC factor correlated +0.51 with maternal AP factor and +0.45 with latent factor reflecting maternal AC during pregnancy. The factors for maternal AP and AC during pregnancy were correlated +0.19. For our subsequent analyses, we constructed factor derived scales using the items with substantial loadings on each of the factors (those bolded in table 1). It had a Cronbach’s standardized Alpha of 0.85.

Factor Structure of Paternal Alcohol Consumption and Problems

An analysis of the scree plot based on 4231 individuals and 35 variables indicated the presence of two clear factors. These were readily interpretable and provided a relatively good fit to the data (RMSEA=0.066, CFI=0.980 and TLI=0.977). Factor loadings by Crawford-Ferguson quartimax rotation are seen in table 2. The first factor had consistently high loadings on 13 variables which reflected paternal AC as measured by both self and maternal report from before pregnancy through age 145 months. We called this factor paternal AC. It had a Cronbach’s standardized Alpha of 0.96. The second factor had consistently high loadings on 15 variables which assessed, via self and maternal report, binge-drinking and alcohol problems from the time of the pregnancy through age 145 months. We termed this factor paternal AP. It had a Cronbach’s standardized Alpha of 0.85. The latent paternal AC factor was correlated +0.42 with paternal AP factor. For subsequent analyses, we constructed factor derived scales using the items with substantial loadings on each of the factors (those bolded in table 2).

Table 2.

Crawford-Ferguson Quartimax Rotated Factor Loadings for Variables on Paternal Alcohol Consumption and Alcohol Problems - Robust Maximum Likelihood Loadings

Variable Reporter 1
Paternal
Alcohol
Consumption		 2
Paternal Problem
Drinking
AC - 18 weeks gestation Mother 0.96 −0.15
AC - 33 weeks Mother 0.70 0.22
AC - 47 months Mother 0.65 0.27
AC - 110 months Mother 0.50 0.34
AC - 145 months Mother 0.44 0.36
AC - before pregnancy Self 0.93 −0.04
AC - 18 weeks gestation Self 0.95 −0.06
AC - last two months gestation Self 0.92 −0.04
AC - 8 weeks Self 0.93 −0.05
AC - 21 months Self 0.76 0.18
AC - 33 months Self 0.72 0.23
AC - 47 months Self 0.65 0.29
AC - 61 months Self 0.63 0.30
AP (since birth) - 8 months Mother 0.34 0.36
AP (since 8 months) - 21 months Mother 0.13 0.63
AP - (since 18 months) - 33
months		 Mother 0.18 0.50
AP (past year) - 47 months Mother 0.11 0.57
AP (past 2 years) - 110 months Mother −0.18 0.90
AP (since 10th birthday) - 145
months		 Mother −0.04 0.78
AP (past year) - 61 months Self 0.07 0.66
AP (since 5th birthday) - 73
months		 Self −0.04 0.73
AP (past 3 years) - 110 months Self −0.11 0.86
AP (past 2 years) - 145 months Self −0.13 0.91
BD - 47 months Mother 0.12 0.74
BD - 110 months Mother 0.07 0.71
BD - 145 months Mother 0.05 0.71
BD - 8 weeks gestation Self 0.38 0.44
BD - 21 months Self 0.22 0.63
BD - 33 months Self 0.20 0.68
BD - 47 months Self 0.11 0.81
BD - 61 months Self 0.16 0.72
Max/Day - 47 months Self 0.06 0.69
Max/Day - 85 moths Self 0.07 0.65
Max/Day - 97 months Self 0.06 0.67
Max/Day - 145 months Self 0.04 0.66
Open in a new tab

AC – Alcohol Consumption

AP – Alcohol Problems

BD – Binge Drinking

Max/Day Maximum drinks per day.

We then examined the observed correlations of the factor derived scales which are expected to be higher than those of the latent factors. Indeed, the factor derived AC and AP scales were correlated +0.68 and +0.75 in mothers and fathers respectively. Maternal and paternal AC and AP factor derived scales were correlated +0.62 and +0.49, respectively.

Prediction of Early Childhood Temperament

We examined 4 broad dimensions of early child temperament as assessed at three time points (figure 1) that reflected i) general activity, ii) emotionality, iii) shyness, and iv) sociability. As seen in table 3, in the univariate analyses, general activity was significantly and positively predicted by maternal and paternal AP, and by only maternal AP in the multivariate analysis. Emotionality had a modest positive association only with maternal AC during pregnancy in the multivariate analyses.

Table 3.

The Prediction of Early Childhood Temperament by Five Dimensions of Parental Alcohol Use and Problems

Outcome Univariate/
Multivariate		 Maternal
Typical AC		 Maternal
AC During
Pregnancy		 Maternal
AP		 Paternal
AC		 Paternal AP
Maternally
Reported
Active
38-69 months		 Univariate b 0.005 −0.008 0.033** 0.013 0.027**
SE 0.011 0.011 0.011 0.012 0.011
Multivariate b −0.029 −0.024 0.062*** −0.007 0.022
SE 0.019 0.014 0.017 0.018 0.016
Maternally
Reported
Emotional
38-69 months		 Univariate b 0.006 0.027 −0.005 −0.006 −0.007
SE 0.011 0.011 0.011 0.011 0.010
Multivariate b 0.005 0.038** −0.029 −0.011 0.004
SE 0.019 0.014 0.017 0.018 0.016
Maternally
Reported
Shyness
38-69 months		 Univariate b −0.073**** −0.059**** −0.068**** −0.060**** −0.061****
SE 0.011 0.011 0.011 0.011 0.010
Multivariate b −0.033 −0.026 −0.020 0.006 −0.035
SE 0.020 0.020 0.017 0.018 0.016
Maternally
Reported
Sociable
38-69 months		 Univariate b 0.058**** 0.039*** 0.033** 0.046**** 0.046****
SE 0.011 0.011 0.011 0.011 0.010
Multivariate b 0.058** 0.011 −0.028 −0.012 0.038
SE 0.019 0.014 0.017 0.018 0.016
Open in a new tab

AC – alcohol consumption; AP – alcohol problems; SE – standard error; b – regression coefficient

Covariates included in all models were: sex of the offspring, age of mother at 12 weeks gestation, highest educational attainment of mother or partner, and a measure of social class based on occupation of mother or partner.

**

p < 0.01;

***

p < 0.001;

****

p < 0.0001

In the univariate analyses, all 5 dimensions of parental alcohol use and problems significantly predicted lower levels of shyness and higher levels of sociability. In multivariate analyses which control for SES and all the other dimensions of parental alcohol use and problems, only maternal AC significantly predicted sociability.

Externalizing Traits

As seen in table 4, conduct difficulties and hyperactivity assessed at 42 months were significantly and positively predicted, in the univariate analyses, by all five dimensions of parental alcohol use and problems. For both traits, only maternal AC during pregnancy remained a significant predictor in the multivariate analyses.

Table 4.

The Prediction of Externalizing Behaviors by Five Dimensions of Parental Alcohol Use and Problems

Outcome Univariate/
Multivariate		 Maternal
Typical AC		 Maternal
AC During
Pregnancy		 Maternal
AP		 Paternal
AC		 Paternal
AP
Maternally
Reported Rutter
Scale
42 months
Conduct difficulties		 Univariate b 0.075**** 0.106**** 0.075**** 0.034** 0.048****
SE 0.012 0.011 0.012 0.011 0.010
Multivariate b 0.017 0.092**** 0.013 −0.060** 0.007
SE 0.020 0.014 0.018 0.019 0.016
Maternally
Reported Rutter
Scale
42 months
Hyperactivity		 Univariate b 0.049**** 0.063**** 0.052**** 0.030 0.037***
SE 0.012 0.012 0.012 0.012 0.011
Multivariate b 0.001 0.052*** 0.018 −0.017 0.029
SE 0.020 0.014 0.018 0.019 0.017
Maternally
Reported Conduct
Symptoms
age13		 Univariate b 0.006 0.046** 0.034 −0.001 0.030
SE 0.014 0.014 0.016 0.014 0.012
Multivariate b −0.046 0.059*** 0.033 −0.057 0.060**
SE 0.025 0.018 0.024 0.024 0.020
Self-Reported
Antisocial Behavior
Age 15		 Univariate b 0.053** 0.071**** 0.086**** 0.041** 0.052***
SE 0.016 0.016 0.018 0.016 0.014
Multivariate b −0.030 0.051 0.067 −0.020 0.041
SE 0.029 0.020 0.027 0.028 0.023
Open in a new tab

AC – alcohol consumption; AP – alcohol problems; SE – standard error; b – regression coefficient

Covariates included in all models were: sex of the offspring, age of mother at 12 weeks gestation, highest educational attainment of mother or partner, and a measure of social class based on occupation of mother or partner.

**

p < 0.01;

***

p < 0.001;

****

p < 0.0001

In univariate analyses, conduct disorder symptoms at age 13 were significantly and positively, albeit weakly, predicted by only maternal AC during pregnancy. In the multivariate analysis, maternal AC during pregnancy and paternal AP were significant predictors. Antisocial behavior measured at age 15 was significantly predicted by all measures although, probably because of colinearity among the predictor variables, none were significant in the multivariate analysis.

Prediction of Alcohol Consumption and Alcohol Problems at ages 15 and 18

As seen in table 5, in the univariate analyses, both AC and AP at age 15 were positively and significantly predicted by all dimensions of parental AC and AP. In the multivariate analysis, age 15 AC was significantly predicted by both maternal and paternal AP, and age 15 AP predicted by only maternal AP.

Table 5.

The Prediction of Alcohol Use and Problems in Adolescence by Five Dimensions of Parental Alcohol Use and Problems

Outcome Univariate/
Multivariate		 Maternal
Typical AC		 Maternal AC
During
Pregnancy		 Maternal AP Paternal AC Paternal AP
Self-report AC
Age 15		 Univariate b 0.102**** 0.088**** 0.155**** 0.079**** 0.109****
SE 0.018 0.018 0.019 0.018 0.016
Multivariate b −0.023 0.022 0.127**** −0.029 0.086***
SE 0.031 0.022 0.030 0.030 0.026
Self-report AP
Age 15		 Univariate b 0.061*** 0.069**** 0.115**** 0.054** 0.076****
SE 0.018 0.018 0.019 0.018 0.016
Multivariate b −0.058 0.034 0.119**** −0.012 0.053
SE 0.031 0.022 0.029 0.030 0.025
Self-report AC
Age 18		 Univariate b 0.078**** 0.052** 0.121**** 0.133**** 0.153****
SE 0.019 0.018 0.021 0.019 0.017
Multivariate b −0.047 −0.012 0.085** −0.010 0.121****
SE 0.033 0.024 0.032 0.032 0.027
Self-report AP
Age 18		 Univariate b 0.133**** 0.112**** 0.168**** 0.143**** 0.168****
SE 0.019 0.019 0.021 0.019 0.017
Multivariate b −0.008 0.037 0.088** −0.010 0.131****
SE 0.033 0.024 0.031 0.032 0.027
Open in a new tab

AC – alcohol consumption; AP – alcohol problems; SE – standard error; b – regression coefficient

Covariates included in all models were: sex of the offspring, age of mother at 12 weeks gestation, highest educational attainment of mother or partner, and a measure of social class based on occupation of mother or partner.

**

p < 0.01;

***

p < 0.001;

****

p < 0.0001

In univariate analyses, both AC and AP at age 18 were significantly predicted by all five parental alcohol factors. In multivariate analyses, age 18 AC and AP were both significantly predicted by maternal and paternal AP.

Two trends in these analyses are of possible interest. First, in the univariate analyses for age 15 AC and AP, the magnitude of the associations seen for maternal factors were somewhat stronger than those seen for paternal factors. However, by age 18, paternal measures were generally more predictive than maternal measures. Second, both maternal and paternal AC and AP were considerably more predictive of AP at age 18 than at age 15.

Maternal Age and Interactions with Sex of Offspring

Maternal age was included as a covariate in all analyses. In four of the twelve dependent variables we examined, maternal age was significantly (and always negatively) related to levels of emotionality, sociability, hyperactivity and conduct disorder symptoms. We examined interactions between the sex of the offspring and our 12 dependent measures across all five parental alcohol-related factors. None of these 60 interactions were significant at the 1% level.

DISCUSSION

This report had two main goals. First, we sought to determine if the rich longitudinal data available about maternal and paternal AC and a range of alcohol-related problems in the ALSPAC cohort could be well represented by a single dimension or required multiple dimensions. The answer to this was clear. Five factors were needed to explain the diversity of items on AC and AP in the mothers and factors. These factors were coherent and easily interpretable. Three of these factors described aspects of the mother’s alcohol use: her general level of consumption, her drinking during the pregnancy of the ALSPAC proband and her level of alcohol problems. The AP factor emerged from two major sets of items – those that reflected patterns of binge drinking and those that reflected alcohol-related problems. The clarity of the factor reflecting maternal AC during pregnancy was perhaps surprising as was its low correlation with maternal AP (+0.19).

Two clear factors emerged for the fathers reflecting AC and AP. These factors look very similar, in terms of item loadings, to those seen in mothers. The inter-factor correlations between AC and APs in mothers and fathers were moderate (+0.51 and +0.42, respectively), suggesting some sharing but substantial independence.

The second goal of this paper was to illustrate potential differences in the predictive power of these five factors. We chose to examine three major domains in the offspring previously shown to frequently differ in offspring of alcoholic parents (Neale 1998;Searles & Windle 1992). Controlling for maternal age and parental social class, we examined early childhood temperament, childhood and adolescent externalizing traits, and AC and AP at ages 15 and 18.

From the many reported results, seven trends were noteworthy. First, all the relationships were statistically modest, explaining more than 1% of the variance in the dependent variable only when examining AC and AP at age 18. Second, the most consistent associations between parental AC and AP and early child temperament were positive with sociability and negative with shyness. We are not aware of prior studies reporting similar findings in children of alcoholics (Jansen et al. 1995;Carle and Chassin 2004). In his review of temperamental features of children of alcoholics, Windle noted little prior evidence that “positive affective attributes” were more prominent in COAs versus controls (Searles & Windle 1992) and Sher agreed with this conclusion with respect to the earlier literature (Sher 1997). However, two prior studies have suggested that symptoms of social phobia and/or separation anxiety can reduce risks for high levels of AC or AP in adolescence (Frojd et al. 2011;Kaplow et al. 2001). Furthermore, extraversion is associated with both alcohol consumption (Malouff et al. 2007) and heavy drinking in the normal range (Martsh and Miller 1997) in adults and adolescents and is moderately heritable (Loehlin 1992). So, our findings might arise from genetic variants in parents that decrease social anxiety and increase extraversion and thereby increase AC (and perhaps some associated problems) which are in turn transmitted to children increasing sociability and decreasing shyness.

Third, consistent with a wide range of findings from children of alcoholics (Sher 1997;Searles & Windle 1992;Jansen, Fitzgerald, Ham, & Zucker 1995), we found a modest association between measures of both AC and AP in parents and symptoms of childhood hyperactivity in their children. Fourth, on average, maternal and paternal AP contributed about equally to the traits examined in childhood and early adolescence. However, for the age 18 measures of AC and AP, paternal influences became somewhat stronger. It will be interesting to see whether this trend continues as the ALSPAC cohort ages into early adulthood.

Fifth, we did not find strong evidence for specificity in the predictions of maternal AC and AP. That is, in univariate analyses, maternal AC did not predict AC in their offspring much better than it predicted offspring AP and vice-versa. Rather, there was a general trend for parental AP to be more strongly associated than parental AC with both offspring AC and AP. We speculate that this might arise because moderate to heavy drinking in adolescence might reflect risk factors more closely associated with problematic drinking rather than moderate social drinking in adulthood.

Sixth, all three of our measures of externalizing behaviors, measured at ages 3½, 13 and 15, were significantly predicted by maternal AC during pregnancy. Most of these effects persisted in multivariate analyses that included typical maternal AC and maternal AP. These results are consistent with the hypothesis that these effects are the result of fetal exposure to the intrauterine effects of alcohol but they could also result from the mother-child transmission of personality traits that increase the risk of maternal alcohol consumption during pregnancy. They are consistent with prior evidence for behavioral effects of intrauterine alcohol exposure (Spear & Molina 2005), including specifically increased risk for externalizing behaviors (Sood, Delaney-Black, Covington, Nordstrom-Klee, Ager, Templin, Janisse, Martier, & Sokol 2001). We examined whether these results could be driven by a small subgroup of mothers with quite high AC during pregnancy by looking at the relationship between AC during pregnancy – divided into 10 ordinal classes – and levels of conduct difficulties. The association was broadly linear over the range of AC, suggesting that the findings were not driven by rare outliers with very high consumption during pregnancy. Maternal AC during pregnancy was also associated with AC and AP at age 15 and AP at age 18 but in multivariate analyses, none of these effects remained significant. Overall, our data provide modest support for the hypothesis that intrauterine exposure to alcohol has an independent influence on subsequent drinking behaviors (Baer, Barr, Bookstein, Sampson, & Streissguth 1998) (Yates, Cadoret, Troughton, Stewart, & Giunta 1998).

Seventh, a number of prior studies have examined, with mixed results, whether the effects of being a child of alcohol parents varies in boys versus girls (Neale 1998;Sher, Walitzer, Wood, & Brent 1991;Belliveau & Stoppard 1995). Our results were clearly negative as we found no evidence of significant interactions between our five parental alcohol factors and any of the examined measures of early childhood temperament, childhood and adolescent externalizing traits.

In aggregate, these results suggest a considerable richness in the multiple pathways from parental alcohol consumption and problems to temperamental, behavioral and consummatory outcomes in their offspring. These findings have direct relevance to efforts to articulate developmental models for the cross-generational transmission of APs (Ohannessian & Hesselbrock 2008).

Limitations

These results should be interpreted in the context of six potentially important methodological limitations. First, the results of factor analysis are always limited to data being analyzed. Other factors of parental alcohol use and problems might have been evidence had we included other measures. Second, our analyses are vulnerable to effects of correlated errors of reporting. Mothers were the source of much of our information on parental drinking and on early childhood temperament and two of the three measures on externalizing behaviors. However, it is reassuring to see that father reported items and AC and AP in the mother clearly loaded with the self-report measures. Also, the association of our measures of parental AC and AP were actually stronger with our one externalizing trait measure obtained by self-report (at age 15) than to the two prior measures obtained using maternal report at 42 months and age 13. Furthermore, all our measures of AC and AP in the offspring were obtained by self-report. Third, working with large samples have many advantages but also the problem that statistical significance becomes a poor guide to conceptual importance. While a number of our results were statistically robust, the associated effect sizes were quite small and would typically go undetected in moderate sample sizes. Fourth, the 12 variables that we attempted to predict with our measures of parental AC and AP were to some extent arbitrary. Our goal was to illustrate rather than thoroughly characterize the associations between parental alcohol use and problems, and offspring behaviors. Fifth, we made no attempt to clarify, at a psychological or biological level, the mediating mechanisms underlying the observed associations. These are important questions that we hope to address in subsequent analyses. Finally, as the ALSPAC cohort age, sample attrition occurred which increased with the increasing age of the cohort (Boyd, Golding, Macleod, Lawlor, Fraser, Henderson, Molloy, Ness, Ring, & Davey 2012). To obtain a sense of the possible biases introduced thereby, we predicted the presence of data on all 12 of our outcome variables from a standardized score for our five parental alcohol factors. Surprisingly, these ORs ranged from 1.00 to 1.26, indicating that higher parental alcohol factors predicted greater cooperation, with the highest ORs generally seen for maternal AC (typically ~ 1.20) and the weakest (and mostly non-significant) for paternal AP with OR from the other three factors generally ranging from 1.05 to 1.15.

Conclusions

Parental alcohol consumption and problems reflect complex multi-dimensional constructs that are not well captured by a dichotomization into family history positive and negative for alcoholism. Parental AC and AP are modestly associated with both positive features of early temperament (e.g., sociability) and more negative features (e.g. hyperactivity), with measures of externalizing behavior from early childhood through adolescence, and to measures of AC and AP in middle and late adolescence. Maternal AC during pregnancy was consistently associated with childhood and adolescent externalizing behavior even when accounting for typical maternal AC and maternal AP.

Acknowledgements

We are extremely grateful to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists and nurses. The UK Medical Research Council (Grant: 74882) and the Wellcome Trust (Grant ref: 092731, 076467) and the University of Bristol provide core support for ALSPAC. This publication is the work of the authors and the corresponding author will serve as guarantor for the contents of this paper. This work was supported in part by the National Institutes of Health grants RO1 AA018333 to D.M.D. and K.S.K, P20 AA107828 and R37 AA011408, and K02 AA018755 to DMD. Additional grants that support JH include UK Medical Research Council (Grants G0800612 and G0802736).

Footnotes

The authors report no competing interests.

References

  1. American Psychiatric Association . Diagnostic and Statistical Manual of Mental Disorders. Fourth Edition American Psychiatric Association; Washington, DC: 1994. [Google Scholar]
  2. Babor T, Higgins-Biddle J, Saunders J, Monteiro M. The Alcohol Use Disorders Identification Test: Guidelines for Use in Primary Health Care (AUDIT) World Health Organization, Department of Mental Health and Substance Disorders; Geneva, Switzerland: 2001. [Google Scholar]
  3. Baer JS, Barr HM, Bookstein FL, Sampson PD, Streissguth AP. Prenatal alcohol exposure and family history of alcoholism in the etiology of adolescent alcohol problems. J.Stud.Alcohol. 1998;59(5):533–543. doi: 10.15288/jsa.1998.59.533. available from: PM:9718105. [DOI] [PubMed] [Google Scholar]
  4. Begleiter H. The Genetics of Alcoholism. Oxford University Press; New York: 1995. [Google Scholar]
  5. Belliveau JM, Stoppard JM. Parental alcohol abuse and gender as predictors of psychopathology in adult children of alcoholics. Addict.Behav. 1995;20(5):619–625. doi: 10.1016/0306-4603(95)00022-5. available from: PM:8712059. [DOI] [PubMed] [Google Scholar]
  6. Bentler PM. Comparative fit indexes in structural models. Psychological Bulletin. 1990;107(2):238–246. doi: 10.1037/0033-2909.107.2.238. available from: PM:2320703. [DOI] [PubMed] [Google Scholar]
  7. Blom G. Statistical Estimates and Transformed Beta Variables. John Wiley & Sons, Inc; New York: 1958. [Google Scholar]
  8. Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson J, Molloy L, Ness A, Ring S, Davey SG. Cohort Profile: The ‘Children of the 90s’--the index offspring of the Avon Longitudinal Study of Parents and Children. International Journal of Epidemiology. 2012 doi: 10.1093/ije/dys064. [epub ahead of print], available from: PM:22507743. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bucholz KK, Cadoret R, Cloninger CR, Dinwiddie SH, Hesselbrock VM, Nurnberger JI, Jr., Reich T, Schmidt I, Schuckit MA. A new, semi-structured psychiatric interview for use in genetic linkage studies: a report on the reliability of the SSAGA. Journal of Studies on Alcohol. 1994;55(2):149–158. doi: 10.15288/jsa.1994.55.149. available from: PM:8189735. [DOI] [PubMed] [Google Scholar]
  10. Buss A, Plomin R. Early Developing Personality Traits. Lawrence Erlbaum; Hillsdale, NJ: 1984. [Google Scholar]
  11. Carey WB. Infant Temperament Questionnaire. Temple University; Dept. Educational Psychology; Philadelphia: 1977. [Google Scholar]
  12. Carle AC, Chassin L. Resiillience in a communityy sample of children of alcoholics: Its prevalence and relation to internalizing symptomatology and positive affect. Applied Developmental Psychology. 2004;25:577–595. [Google Scholar]
  13. Chassin L, Pitts SC, DeLucia C, Todd M. A longitudinal study of children of alcoholics: predicting young adult substance use disorders, anxiety, and depression. J.Abnorm.Psychol. 1999;108(1):106–119. doi: 10.1037//0021-843x.108.1.106. available from: PM:10066997. [DOI] [PubMed] [Google Scholar]
  14. Cloninger CR, Bohman M, Sigvardsson S, von Knorring AL. Psychopathology in adopted-out children of alcoholics: The Stockholm adoption study. In: Galanter M, editor. Recent Developments in Alcoholism. Vol.3. Plenum Press; New York: 1985. pp. 37–51. [DOI] [PubMed] [Google Scholar]
  15. Cotton NS. The familial incidence of alcoholism: a review. J Stud.Alcohol. 1979;40(1):89–116. doi: 10.15288/jsa.1979.40.89. available from: PM:376949. [DOI] [PubMed] [Google Scholar]
  16. Dick DM, Aliev F, Latendresse SJ, Hickman M, Heron J, Macleod J, Joinson C, Lewis G, Kendler KS. Adolescent Alcohol Use is Predicted by Childhood Temperament Factors before Age 5, with Mediation through Personality and Peers. 2012. [DOI] [PMC free article] [PubMed]
  17. Edwards EP, Eiden RD, Colder C, Leonard KE. The development of aggression in 18 to 48 month old children of alcoholic parents. J Abnorm.Child Psychol. 2006;34(3):409–423. doi: 10.1007/s10802-006-9021-3. available from: PM:16649002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Elander J, Rutter M. Revised Rutter Parent Scale for Preschool Children. 1996.
  19. Frojd S, Ranta K, Kaltiala-Heino R, Marttunen M. Associations of social phobia and general anxiety with alcohol and drug use in a community sample of adolescents. Alcohol and Alcoholism. 2011;46(2):192–199. doi: 10.1093/alcalc/agq096. available from: PM:21245062. [DOI] [PubMed] [Google Scholar]
  20. Goodman R. The Strengths and Difficulties Questionnaire: a research note. Journal of Child Psychology and Psychiatry and Allied Disciplines. 1997;38(5):581–586. doi: 10.1111/j.1469-7610.1997.tb01545.x. available from: PM:9255702. [DOI] [PubMed] [Google Scholar]
  21. Harter SL. Psychosocial adjustment of adult children of alcoholics: a review of the recent empirical literature. Clin.Psychol.Rev. 2000;20(3):311–337. doi: 10.1016/s0272-7358(98)00084-1. available from: PM:10779897. [DOI] [PubMed] [Google Scholar]
  22. Heath AC, Bucholz KK, Madden PA, Dinwiddie SH, Slutske WS, Bierut LJ, Statham DJ, Dunne MP, Whitfield JB, Martin NG. Genetic and environmental contributions to alcohol dependence risk in a national twin sample: consistency of findings in women and men. Psychological Medicine. 1997;27(6):1381–1396. doi: 10.1017/s0033291797005643. available from: PM:9403910. [DOI] [PubMed] [Google Scholar]
  23. Hesselbrock M, Easton C, Bucholz KK, Schuckit M, Hesselbrock V. A validity study of the SSAGA--a comparison with the SCAN. Addiction. 1999;94(9):1361–1370. doi: 10.1046/j.1360-0443.1999.94913618.x. available from: PM:10615721. [DOI] [PubMed] [Google Scholar]
  24. Hussong AM, Curran PJ, Chassin L. Pathways of risk for accelerated heavy alcohol use among adolescent children of alcoholic parents. J Abnorm.Child Psychol. 1998;26(6):453–466. doi: 10.1023/a:1022699701996. available from: PM:9915652. [DOI] [PubMed] [Google Scholar]
  25. Jansen RE, Fitzgerald HE, Ham HP, Zucker RA. Pathways into risk: temperament and behavior problems in three- to five-year-old sons of alcoholics. Alcohol Clinical Experimental Research. 1995;19(2):501–509. doi: 10.1111/j.1530-0277.1995.tb01538.x. available from: PM:7625589. [DOI] [PubMed] [Google Scholar]
  26. Kaplow JB, Curran PJ, Angold A, Costello EJ. The prospective relation between dimensions of anxiety and the initiation of adolescent alcohol use. Journal of Clinical Child Psychiatry. 2001;30(3):316–326. doi: 10.1207/S15374424JCCP3003_4. available from: PM:11501249. [DOI] [PubMed] [Google Scholar]
  27. Kendler KS, Prescott CA, Neale MC, Pedersen NL. Temperance board registration for alcohol abuse in a national sample of Swedish male twins, born 1902 to 1949. Archives of General Psychiatry. 1997;54(2):178–184. doi: 10.1001/archpsyc.1997.01830140090015. available from: PM:9040286. [DOI] [PubMed] [Google Scholar]
  28. Kendler KS, Schmitt JE, Aggen SH, Prescott CA. Genetic and Environmental Influences on Alcohol, Caffeine, Cannabis, and Nicotine Use from adolescence to middle adulthood. Archives of General Psychiatry. 2008;65(6):674–682. doi: 10.1001/archpsyc.65.6.674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Koopmans JR, van Doornen LJ, Boomsma DI. Association between alcohol use and smoking in adolescent and young adult twins: a bivariate genetic analysis. Alcohol Clin.Exp.Res. 1997;21(3):537–546. available from: PM:9161615. [PubMed] [Google Scholar]
  30. Lieberman DZ. Children of alcoholics: an update. Curr.Opin.Pediatr. 2000;12(4):336–340. doi: 10.1097/00008480-200008000-00009. available from: PM:10943813. [DOI] [PubMed] [Google Scholar]
  31. Loehlin JC. Genes and Environment in Personality Development. Sage Publications; Newbury Park, CA: 1992. [Google Scholar]
  32. Malouff JM, Thorsteinsson EB, Rooke SE, Schutte NS. Alcohol involvement and the Five-Factor model of personality: a meta-analysis. Journal of Drug Education. 2007;37(3):277–294. doi: 10.2190/DE.37.3.d. available from: PM:18047183. [DOI] [PubMed] [Google Scholar]
  33. Martsh CT, Miller WR. Extraversion predicts heavy drinking in college students. Personality and Individual Differences. 1997;23(1):153–155. [Google Scholar]
  34. Neale MC. Modeling Interaction and Nonlinear Effects With Mx: A General Approach. In: Schumacker RE, Marcoulides GA, editors. Interaction and Nonlinear Effects in Structural Equation Modeling. Lawrence Erlbaum Associates; Mahwah, NJ: 1998. pp. 43–61. [Google Scholar]
  35. Ohannessian CM, Hesselbrock VM. Do personality characteristics and risk taking mediate the relationship between paternal substance dependence and adolescent substance use? Addict.Behav. 2007;32(9):1852–1862. doi: 10.1016/j.addbeh.2006.12.017. available from: PM:17241748. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Ohannessian CM, Hesselbrock VM. A comparison of three vulnerability models for the onset of substance use in a high-risk sample. J Stud.Alcohol Drugs. 2008;69(1):75–84. doi: 10.15288/jsad.2008.69.75. available from: PM:18080067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Riley EP, Infante MA, Warren KR. Fetal alcohol spectrum disorders: an overview. Neuropsychol.Rev. 2011;21(2):73–80. doi: 10.1007/s11065-011-9166-x. available from: PM:21499711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Schuckit MA, Tipp JE, Smith TL, Wiesbeck GA, Kalmijn J. The relationship between Self-Rating of the Effects of alcohol and alcohol challenge results in ninety-eight young men. Journal of Studies on Alcohol. 1997;58(4):397–404. doi: 10.15288/jsa.1997.58.397. [DOI] [PubMed] [Google Scholar]
  39. Searles JS, Windle M. Children of Alcoholics: Critical Perspectives. Guilford Press; 1992. [Google Scholar]
  40. Sher KJ. Children of Alcoholics: A Critical Appraisal of Theory and Research. University of Chicago Press; Chicago: 1991. [Google Scholar]
  41. Sher KJ. Psychological characteristics of children of alcoholics. Alcohol Health Res.World. 1997;21(3):247–254. available from: PM:15706777. [PMC free article] [PubMed] [Google Scholar]
  42. Sher KJ, Walitzer KS, Wood PK, Brent EE. Characteristics of children of alcoholics: putative risk factors, substance use and abuse, and psychopathology. J.Abnorm.Psychol. 1991;100(4):427–448. doi: 10.1037//0021-843x.100.4.427. available from: PM:1757657. [DOI] [PubMed] [Google Scholar]
  43. Smith DJ, McVie S. Theory and Method in the Edinburgh Study of Youth Transitions and Crime. British Journal of Criminology. 2003;43:169–195. [Google Scholar]
  44. Sood B, Delaney-Black V, Covington C, Nordstrom-Klee B, Ager J, Templin T, Janisse J, Martier S, Sokol RJ. Prenatal alcohol exposure and childhood behavior at age 6 to 7 years: I. dose-response effect. Pediatrics. 2001;108(2):E34. doi: 10.1542/peds.108.2.e34. available from: PM:11483844. [DOI] [PubMed] [Google Scholar]
  45. Spear NE, Molina JC. Fetal or infantile exposure to ethanol promotes ethanol ingestion in adolescence and adulthood: a theoretical review. Alcohol Clin.Exp.Res. 2005;29(6):909–929. doi: 10.1097/01.alc.0000171046.78556.66. available from: PM:15976517. [DOI] [PubMed] [Google Scholar]
  46. Steiger JH. Structural model evaluation and modification: an interval estimation approach. Multivariate Behavioral Research. 1990;25:173–180. doi: 10.1207/s15327906mbr2502_4. [DOI] [PubMed] [Google Scholar]
  47. Stout ML, Mintz LB. Differences among Nonclinical College Women with Alcoholic Mothers Alcoholic Fathers, and Nonalcoholic Parents. Journal of Counseling Psychology. 1996;43(4):466–472. [Google Scholar]
  48. Tucker LR, Lewis C. A reliability coefficient for maximum likelihood factor analysis. Psychometrika. 1973;38:1–10. [Google Scholar]
  49. Viken RJ, Kaprio J, Koskenvuo M, Rose RJ. Longitudinal analyses of the determinants of drinking and of drinking to intoxication in adolescent twins. Behav.Genet. 1999;29(6):455–461. doi: 10.1023/a:1021631122461. available from: PM:10857250. [DOI] [PubMed] [Google Scholar]
  50. Yates WR, Cadoret RJ, Troughton EP, Stewart M, Giunta TS. Effect of fetal alcohol exposure on adult symptoms of nicotine, alcohol, and drug dependence. Alcohol Clin.Exp.Res. 1998;22(4):914–920. available from: PM:9660322. [PubMed] [Google Scholar]

RESOURCES