Publié dans Early Child Development and Care, en ligne le 23 juillet 2009 DOI: 10.1080/03004430903040417

Two generations of maternal alcohol abuse: impact on cognitive levels in mothers and their children Annick-Camille Dumareta1, Mélanie Cousinb, Maurice Titranb a

CERMES - Institut National de la Santé et de la Recherche Médicale, Unité 750, Villejuif, France b Centre d’Action Médico-Sociale Précoce, Roubaix, France

Abstract Transgenerational effects of alcohol on mothers’ and children’s intellectual functioning has been examined in 22 families from very deprived environments. Their psychosocial outcomes and IQ level were evaluated in a follow-up study on average seven years after they left the group support of a day-care center for young children; School data was collected for the offsprings. A decrease of 18 IQ points due to fetal alcohol spectrum disorders (FASD) was observed in the two generations (mothers, children). Accumulation of alcohol problems in the families still predicted school failure, even in absence of FASD. The possible gain of IQ points due to the decreasing of alcohol consumption during pregnancy could help them have better school experiences. Providing appropriate ongoing supports would be an important and practical approach to improve the conditions of life of these families.

Key words: cognitive development, fetal alcohol spectrum disorders, school achievement, lower-classes, transgenerational observation.

1

Corresponding author : A.C. Dumaret, CERMES-INSERM U750, site CNRS, 7 rue Guy Môquet, 94810 Villejuif, Cedex, France. [email protected]

1

Since the first reports on fetal alcohol syndrome FAS (Lemoine, 1968 ; Jones & Smith, 1973), numerous studies have examined the development of children with alcoholic parents (Weinberg, 1997; Johnson & Leff 1999). Given the importance of co-morbidity factors linked to alcohol, longitudinal studies are considered the most reliable. Nevertheless, none has evaluated the effects of fetal alcohol exposure on IQ variation both within the generation of the mothers and the one of their children. A global cognitive deficit is observed: Some longitudinal studies indicate a decrease in IQ from 5-7 points in children from 4-7 years old, in the United States in middle class and in France in a less privileged population (Streissguth, Barr, Sampson, 1990; Larroque et al., 1995). Effects on IQ were reported at later ages, more importantly in the case of children with facial dysmorphology (fetal alcohol syndrome, FAS) than in that of fetal alcohol effects by children (FAE), and lower verbal IQs than performance IQs were noted. New diagnostic categories were proposed for partial FAS, alcohol related neurodevelopmental disorders (ARND) for children exposed in utero to alcohol but who do not present all the criteria of FAS (Stratton, Howe & Battaglia 1996). Now, it is known that high levels of prenatal alcohol exposure are related to important IQ deficits even without severe dysmorphy (Mattson, Riley, Gramling, Delis, & Lyons 1997), and all these disorders have been termed fetal alcohol spectrum disorders (Streissguth & O’Malley 2000). Regarding school performance, reports come from the same studies. Some have demonstrated the relationship between prenatal alcohol exposure and the subsequent academic achievement of the exposed children. In Nordic countries and in the Berlin cohort, 31-55% were enrolled in special education classes and 25–30% attend medical institutions for disabled children (Aronson & Hagberg 1998; Autti-Ramö, 2000; Steinhausen & Spohr 1998). In the Seattle cohort, alcohol consumption early in pregnancy

2

was associated with increased academic performance problems in late childhood: Nearly half of the adolescents and adults with FAS/FAE failed a grade in school, 61% had disrupted school experiences and approximately 40% received special education services (Streissguth, Barr, Kogan, & Bookstein, 1996). Among studies in precarious environment, research by Bingol et al. (1987) have shown that a family history of alcoholism was almost three times more common in the lowerclasses and that mental retardation was significantly greater in children of lower-class alcoholic women. Significant deficits in academic achievement were shown; however no evidence of adverse alcohol effects on cognitive development in the absence of FAS for children aged 5-7 was noted, but the level of exposure to alcohol was moderate (Goldschmidt, Richardson, Stoffer, Geva, & Day, 1996). Similar results on IQ were found by in a sample of African-American children of 7.5 years of age (Jacobson SW, Jacobson JL, Sokol, Chiodo, & Corobana, 2004). Such studies usually report maternal education; only a few indicate the IQ of alcoholic mothers. Nevertheless maternal IQ affects the child environment with an indirect effect on academic achievement. In the low-income black population of the Atlanta cohort, mean IQ for mothers of FAS-FAE children was 78±7 and it was 91±15 for Caucasian mothers in Washington state (Coles et al., 1997; Astley, Bailey, Talbot, & Clarren, 2000). Concerning pathologies due to alcohol abuse in France, research in public health has shown marked inequalities between social classes, but impact on children from these different social classes is still not known INSERM (2001). We carried out in Northern France a transgenerational observation of families from high-risk backgrounds with a prevalent history of family alcoholism, in order to analyze their psychosocial outcome of the children, who did benefit from early intervention in a day-care treatment centre for young children in Roubaix (Centre d’action médico-sociale précoce CAMSP). In this

3

town, it had been demonstrated that the incidence of FAS (between 1.3 and 4.8/1,000) was heavier that in the Seattle and the Cleveland Cohort (Sampson et al., 1997). This report is focused on transgenerational effects of alcohol on mothers’ and children’s intellectual functioning. The purpose is twofold: -

To describe the risk factors and psychosocial characteristics associated with FASD in these families

-

To present and compare the intellectual level (school and IQ performances) in the mothers’ and children’s groups. We observe that FASD produces the same effects in similar environments, namely poor IQs levels and school failure.

The families’ outcomes, parenthood and the impact of early intervention on child development and behaviour have been presented elsewhere (Dumaret, Constantin-Kuntz & Titran, 2009;, Dumaret, Cousin & Titran, 2008).

METHODS

Institutional context The “Tuesday group” (TG) is a friendly support group in the CAMSP, designed for mothers with infants from very deprived environments, who have mental problems or cognitive deficiencies and/or who are alcoholic. This TG was geared towards reducing social isolation and alcoholism, and enhancing parental skills (Titran, 2004). The theoretical basis for this approach lies in creating guidance and support strategies in ways that respect families and their social identity; The TG helps parents to recover self confidence, and it aims to prevent future alcohol-affected pregnancies and to protect children from the abuse and neglect that often occurs in the homes of alcoholic parents.

4

This group includes professionals (pediatrician, social worker, teachers, therapist) as well as volunteers (para-professionals and professionals). They meet once a week with warm and informal exchanges and participants share their experiences and their knowledge of activities from daily life (e.g., manual labor, cooking and so forth). Families always meet with the same staff members.

Study participants

The study focuses on alcoholism in the maternal line, as “offspring of the lower-class had a much higher incidence of chronic alcoholic maternal grandmothers” (Bingol et al, 1987, p.113). The selection criteria were the following: 1) all the families were confronted with parental or grandparental alcoholism, and had participated for one year or longer in the activities of the TG support group between 1989 and 1995; 2) all had a child, younger than four and a half years of age; and 3) all assessments were carried out more than five years after leaving the group (2000-2003). Thus, the study involves 22 families (Dumaret & Titran, 2004). Intensive research involving the participation of several medical and social services was required to access these families due to their frequent household changes, abrupt relocations, missed appointments and so forth. All the parents participating in the study signed a consent form and received a guarantee of confidentiality. All the mothers and children involved had been diagnosed several years before the beginning of the study. For the mothers, diagnosis was noted in the CAMSP records when they were examined by the pediatrician (for their own dysmorphy and malformations, their family antecedents, learning difficulties, social behavior and mental health problems). Fourteen mothers had FASD: 6 with Fetal Alcohol Syndrome [FAS] and 8 with Fetal Alcohol Effects [FAE]. Among their children, eight were FASD: 4 with FAS and 4 with

5

FAE. Five had been diagnosed at birth or before the age of three by the director of the CAMSP, specialist in disabilities and FAS diagnosis. The three others were assessed by comparing clinical data (psychomotor, language, cognitive and scholastic) with behavior problems, specific neurological anomalies and documentation of prenatal alcohol exposure obtained from the medical and social service records. Maternal alcohol consumption was documented for these eight children (on average 12-15 alcohol drinks per day). In one family, a mother and her four children were affected; in another family, the mother and two of her daughters were affected.

Measurements Interviews with families were conducted by two psychologists: a researcher from another institution interviewed the parents (first author) and a clinician, who was recently hired and did not know the families, examined the children (second author). Family assessment: Information on family antecedents abstracted from the institutional records was followed up with semi-structured interviews with parents. Interviews, all recorded, dealt with: family history, education, household, socio-professional life and financial management, parenting, health (alcohol use by the couple and in the extended family, health-care use, medication intake) and memories of the TG. Three quarters of these families lived below the level of poverty. Due to the effects of alcohol addiction on the verbal level, 4 subtests of the WAIS-III were used for the mothers (Vocabulary, Information, Similarities and Comprehension). Verbal Comprehension Index (including the first three subtests) can be considered as an estimation of IQ as it is associated with VIQ and FIQ (successively r=0.94 and r=0.85) (Wechsler, 2000).

6

All the mothers agreed to participate; 21 were seen twice (at least 1 hour each time) and given the subtests of the WAIS-III. Mothers’ mean age for interview and testing was 40.6±4.6 years. Eight fathers or spouses were also interviewed. Data on several fathers who had also suffered family adversities, however, remains less complete than those on mothers. Foster families for children in out-of-home placement were given semi-structured interviews over the phone. Child assessment: In this project, the aim was to evaluate at least one child per family (the one who benefited from the longest time in the TG). Data collection included IQ tests, school curriculum, and an interview (daily family life, relationships with peers and so forth), and a behaviour questionnaire. These two last data sets are not analyzed here. IQ scores: They were assessed with the WISC-III. The composite score of the four subtests were used as a measure of IQ: Vocabulary, Similarities, Block Design and Picture Arrangement (Sattler, 1988; Grégoire, 2000). Thirty one IQ scores were obtained (9/10 of the children had participated to the group). In this sample, 20 were tested ‘blind’ by a member of the research team, eleven other IQ scores were obtained from institutions the children attended. Testing age for children was 10.8±2.8 years. Thus, for 20 families, data included an IQ score for the mother and for one or more of her children. In one family, only the mother was tested, since her children were in an institution for children with special needs (autism and psychosis). One borderline mother was interviewed once but neither she nor her child were able to be tested. School curriculum: Some families had two children when they participated in the TG, thus it was decided to enlarge the protocol to siblings. As we could not test all the mothers’offspring, the scholastic situation of all their minor children was collected by the institutions. As school performance problems are a marker of behavioral and cognitive problems, an index of school failure allows the coding of children’s difficulties in relation

7

to their age and to the ‘educational system’: (0) = no school failure, (1) or (2) years retained in grade = mild school failure, (3) = failure with special education in classes for slow learners, and (4) = very severe failure (e.g. institutions for disabled or mentally retarded children). School status was obtained for 59 children (3 had not yet been enrolled) and updated in 2004 (age during school evaluation: 13.4±2.8 years).

Analysis Data collected during the interviews from parents, children, and at times foster care givers and social workers, were compared to institutional files and relevant knowledge of the pediatrician of the child-care unit. Several maternal variables were dichotomised: maltreatment, placement, mental health problems, maternal education. IQ assessment and the school failure index were coded twice. Classical statistical tests were used (Statistica, 2004). Multivariate analyses of variance (ANOVA/MANOVA) with and without covariates were conducted to test the effects of

FASD on outcomes. Multiple regression analyses were used to test predicting factors on IQ and on school failure. For IQ and subtests, mothers and children were first considered as independent groups. Then, mothers and children were considered together: thus, four groups were taken into account, with IQ as the dependant variable and two independent ones: the mothers-children groups and the presence of FASD.

RESULTS

Family situations during the group-support and at time of the assessment All families were Caucasian. When they were admitted into the group, none of the mothers had any professional activities. None had a diploma, except two mothers without FASD:

8

one when entering the secondary school, the other when entering the second cycle of the secondary school; these two women began to drink later in life than the others. None had completed the second cycle of the secondary school (high school). Familial alcohol dependency was noted for twenty maternal grandmothers, eleven mothers and sixteen fathers (Table 1). On the maternal line, alcohol abuse involved one generation for four families (both or one alcoholic grandparent), two generations for twelve families with one alcoholic parent (both grandparents, the father or the mother and/or parental siblings), and two generations for six families with both alcoholic parents (and parental siblings). Independently of maternal FASD, alcohol dependency in the families was associated with other risk factors: severe neglect and abuse in 18 mothers during their own childhood and adolescence (82%), present intra-family violence towards the mothers and/or their children for 17 families (77%). Out-of-home placements and sexual abuse were noted respectively for 8 and 6 mothers in their past. The large majority of the 22 mothers were depressed and eleven mothers suffered from mental illness (reports of hospitalization and/or psychiatric followup of institutional dossiers). The 22 mothers had 98 children (including those of-age), with an average of 4 children per mother with FASD and 5 for the mothers without FASD. In the families, 3/4 have or have had children in out-of-home placement: 38/57 of the children of the mothers with FASD were in out-of-home care compared to 15/41 of the children of mothers without FASD (p