The effects of low to moderate alcohol consumption and binge drinking in early pregnancy on selective and sustained attention in 5-year-old children

Transcription

1 DOI: /j x Epidemiology The effects of low to moderate alcohol consumption and binge drinking in early pregnancy on selective and sustained attention in 5-year-old children M Underbjerg, a,b US Kesmodel, a,c NI Landrø, d L Bakketeig, e J Grove, a,f T Wimberley, g TR Kilburn, a C Sværke, a P Thorsen, h EL Mortensen i a Department of Public Health, Section of Epidemiology, Aarhus University, Aarhus, Denmark b Children s Neurocenter at Vejlefjord Rehabilitation Center, Stouby, Denmark c Department of Obstetrics and Gynaecology, Aarhus University Hospital, Aarhus, Denmark d Department of Psychology, Centre for the Study of Human Cognition, University of Oslo, Oslo, Norway e National Institute of Public Health, University of Oslo, Oslo, Norway f Department of Biomedicine, Faculty of Health Sciences and Bioinformatics Research Centre (BiRC), Aarhus University, Aarhus, Denmark g Section of Biostatistics, Department of Public Health, Aarhus University, Aarhus, Denmark h Department of Obstetrics and Gynaecology, Lillebaelt Hospital, Kolding, Denmark i Institute of Public Health and Center for Healthy Aging, University of Copenhagen, Copenhagen, Denmark Correspondence: Prof EL Mortensen, Medical Psychology Unit, Institute of Public Health, University of Copenhagen, Øster Farimagsgade A, DK-1353, Copenhagen K, Denmark. elme@sund.ku.dk Accepted 10 April Published Online 20 June Objective The aim was to examine the effects of low to moderate maternal alcohol consumption and binge drinking in early pregnancy on children s attention at 5 years of age. Design Prospective follow-up study. Setting Neuropsychological testing in four Danish cities Population A cohort of 1628 women and their children sampled from the Danish National Birth Cohort. Methods Participants were sampled based on maternal alcohol consumption during pregnancy. At 5 years of age, the children were tested with the recently developed Test of Everyday Attention for Children at Five (TEACh-5). Parental education, maternal IQ, maternal smoking in pregnancy, the child s age at testing, gender, and tester were considered core confounding factors, whereas the full model also controlled the following potential confounding factors: maternal binge drinking or low to moderate alcohol consumption, age, body mass index (BMI), parity, home environment, postnatal smoking in the home, child s health status, and indicators for hearing and vision impairments. Main outcome measures TEACh-5 attention scores. Results There were no significant effects on test performance in children of mothers drinking up to 8 drinks per week compared with children of mothers who abstained, but there was a significant association between maternal consumption of 9 or more drinks per week and risk of a low overall attention score (OR 3.50, 95% CI ). No consistent or significant associations were observed between binge drinking and attention test scores. Conclusions The findings suggest an effect of maternal consumption of 9 or more drinks per week on attention functions in children, but the study detected no effects of lower levels of maternal consumption and no consistent effects of maternal binge drinking. Keywords Attention, attention deficits, binge drinking, maternal alcohol consumption, prenatal alcohol exposure, selective attention, sustained attention, TEACh-5. Please cite this paper as: Underbjerg M, Kesmodel U, Landrø N, Bakketeig L, Grove J, Wimberley T, Kilburn T, Sværke C, Thorsen P, Mortensen E. The effects of low to moderate alcohol consumption and binge drinking in early pregnancy on selective and sustained attention in 5-year-old children. BJOG 2012;119: ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG 1211

2 Underbjerg et al. Introduction Attention deficits and dysfunctions in children are among the most commonly reported effects of drinking during pregnancy, 1 5 and prenatal alcohol exposure has been suggested to constitute a key element in the development of a subtype of Attention Deficit Hyperactivity Disorder (ADHD). 6,7 However, other studies have failed to observe a relationship between prenatal alcohol exposure and attention problems Cerebral structural changes after prenatal alcohol exposure have been observed in several areas considered to be involved in attention processes, such as the parietal and orbitofrontal regions, cerebellum, and the basal ganglia. 12,13 To date, most studies have looked at the impact of relatively high doses of prenatal alcohol exposure (i.e. an intake of about one drink per day or more), but there is a growing concern that prenatal exposure to more moderate quantities of alcohol may also be damaging to cognitive and behavioural functions in the offspring. 14,15 It remains unclear whether there is a safe level of drinking during pregnancy, 16 and in spite of animal studies suggesting that the brain may be particularly susceptible to high peak blood alcohol concentrations, 17 only very few human studies have investigated the effect of exposure to binge drinking during pregnancy on neurocognitive and behavioural outcomes in offspring. 18 Results are inconsistent, and whether binge drinking, independent of the average alcohol intake, is harmful to the cognitive development of offspring remains an open question. Attention is a multidimensional construct related to memory and executive functions. 19 In an influential review, Posner and Petersen 20 argued, from adult neuropsychological and functional imaging evidence, that distinct attention networks are vulnerable to selective damage, and that systems involved in selective attention (e.g. finding a visual target among distractors) diverge from those involved in sustained attention or alertness (e.g. maintaining a readiness to respond over a long interval). The proposed separation does not necessarily imply that these functions share no common processes, but rather that they have distinct features and are susceptible to being differentially damaged. Several studies have investigated the effects of prenatal alcohol exposure on sustained attention (visual vigilance) using variants of the Continuous Performance Task, 21 whereas fewer studies have examined selective attention, 5,22,23 and overall it remains unclear whether some aspects of attention are more vulnerable to prenatal alcohol exposure than others. The aim of the current study was to examine the effects of low to moderate maternal alcohol consumption and binge drinking during pregnancy on attention functions at 5 years of age, using a recently developed, theoretically founded, and comprehensive test battery, the Test of Everyday Attention for Children at Five (TEACh-5), 24 designed to conduct differential assessment of selective and sustained attention. Methods Sample The design of the Lifestyle During Pregnancy Study (LDPS) has been described in detail elsewhere. 25 Briefly, the study is a prospective follow-up, based on a sample from the Danish National Birth Cohort (DNBC), 26 with oversampling of moderate to high alcohol exposure categories, including binge drinking. 25 The DNBC consists of women and their offspring, recruited from 1997 to 2003 at their first antenatal visit to a general practitioner, who routinely is the first health practitioner to see a pregnant woman. The participating women represent 60% of those invited and approximately 30% of all pregnant women in Denmark in the enrolment period. Based on their alcoholdrinking pattern before and during pregnancy, a total of 3478 women with singleton pregnancies were sampled from the DNBC and invited to participate in the LDPS between 2003 and Of the invited women, 3189 were sampled on the basis of their alcohol intake during pregnancy. Of these women, 1628 mothers and their children participated in a 5-year follow-up, and only these mother and child pairs were included in the analyses. Information on binge drinking was invalid for 11 of the 1628 mothers, and consequently the binge drinking analyses were conducted on 1617 mothers and their children. Exclusion criteria for the LDPS were an inability to speak Danish, an inability to complete the cognitive tests because of impaired hearing or vision, multiple pregnancies, and congenital diseases likely to cause mental retardation (e.g. trisomy 21). The 5-year follow-up The follow-up study was designed to assess attention functions, and also included assessment of intelligence with three verbal and three performance subtests from the Wechsler Primary and Preschool Scales of Intelligence Revised (WPPSI-R). 27 The cognitive test battery was administered to children aged from 5 years to 5 years and 4 months. Testing took place in four major cities in Denmark (Copenhagen, Odense, Aalborg, and Aarhus). Test procedures were standardised in detail and carried out by ten trained psychologists blinded to the children s exposure status. Tester s were taken into account by the inclusion of indicator variables in the statistical analyses. Exposure information Information on alcohol intake during pregnancy was derived from the first prenatal DNBC interview, 26 conducted at a 1212 ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG

3 The effects of early prenatal alcohol consumption median gestational age of 17 weeks (range: 7 39 weeks). During the interview the women were asked about the average number of beers, glasses of wine, and glasses of spirits they currently consumed during a week, and based on this information, the total number of weekly drinks was calculated. Information on binge drinking during pregnancy included data on the number of binge episodes (defined as an intake of 5 or more drinks on a single occasion) and the timing (gestational week) of these episodes. Some women reported one or more binge episodes during the early weeks of pregnancy, although their average number of drinks per week at the time of interview was zero. These women were classified accordingly as consuming 0 drinks at the time of interview, but with one or more previous binge episodes. Our definition of a drink followed that of the Danish National Board of Health, with one standard drink being equal to 12 g of pure alcohol. For this study low drinking was defined as consuming between one and 4 drinks per week, and moderate drinking was defined as consuming between five and 8 drinks per week. All mothers were sampled in strata defined by their average alcohol intake (0, 1 4, 5 8, 9 drinks/week), and by the timing and number of binge episodes (gestational weeks 1 2, 3 4, 5 8, 9, or later). 25 The higher exposure categories were oversampled in an effort to ensure that all exposure categories included enough children to attain sufficient statistical power. For each sample stratum the sampling probability was computed as the ratio between the number of sampled women and the total number of women available in the corresponding DNBC stratum. Outcome measures Attention was assessed by TEACh-5, 24 which is the most recent development in a series of comprehensive attention test batteries. 28 Briefly, TEACh-5 takes the form of a children s storybook, with puzzles and games to complete with the characters. To obtain pure measures of attention, demands on task comprehension, memory, reasoning, language, and motor skills have been minimised. For all subtests, practise trials formally checking the comprehension of instructions and basic visual and auditory skills were included. Auditory stimuli consist of animal noises that are relatively universal and usually recognisable at an earlier developmental stage than complicated language, and verbal responses are few and kept to simple yes/no answers. Several subtest scores can be derived from TEACh For the present analyses two subtests assessing selective attention and two subtests assessing sustained attention were selected. Subtests with an approximate normal distribution of test scores (or transformed test scores) and with a pattern of positive intercorrelations were selected. An overall composite mean attention score was derived from the four selected subtests, and in addition means were derived for selective attention and for sustained attention. Selective attention The Great Balloon Hunt. The children were given 15 seconds to mark as many as possible of 48 balloon targets on a greyscale A4 sheet, first on a sheet with nothing but balloons, and second on a sheet with the target balloons distributed among visual distractors. The sum of the scores on the two tasks was used in the present analyses. Hide and Seek II. The task was to listen to a total of 14 sound clips of 10-seconds duration, and then to report whether the target element (a bark from the dog) was absent or present. The level of performance was measured by a compound score reflecting both accuracy and speed by weighing the mean reaction time when the bark was present based on the proportionate accuracy of the items. Only the log-transformed compound score was used in the present study. Sustained attention Barking. The children were told to listen to ten CD tracks of a single, repeated dog bark, and to keep a count of how many barks occurred on each item. The longest tone string was six, the shortest was two, and intervals between the stimuli varied between 2 and 10 seconds, with a total of four fast presented items and six slowly presented items. The number of correct slowly presented items had an approximately symmetric distribution, and was used in the present analyses. Draw-a-Line. The children were asked to use a pen to trace a line as slowly as possible without stopping and without lifting the pen from the page. To ensure that lefthanded children were not disadvantaged, the pages were presented in opposite orientations (i.e. the practise went from left to right and the test item from right to left, or vice versa). The time taken in seconds to complete the task was noted, and for data analysis this time measure was log transformed. The total number correct (Barking and Great Balloon Hunt) and the log-transformed scores (Hide and Seek II and Draw-a-Line) were first standardised to a mean of 0 and an SD of 1 (for Hide and Seek II the standardised score was reversed to ensure that higher scores reflected a better performance). Then, the mean of the four standardised scores was calculated and re-standardised to a mean of 0 and an SD of 1 for use in the statistical analyses. Similarly, mean scores were derived for selective attention (Great Balloon Hunt and Hide and Seek II), and for sustained attention (Barking and Draw-a-Line). The calculation of the mean score of overall attention was corroborated by a principal component analysis, showing that all four subtests ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG 1213

4 Underbjerg et al. loaded on the first principal component. The overall attention mean has also been analysed together with measures of intelligence and executive functioning, but this is reported in a separate article. 29 Covariates The following covariates were obtained in the prenatal interview, and were subsequently coded as follows: parity (0, 1, 2+); prenatal maternal smoking (yes/no); and maternal pre-pregnancy BMI [weight in kg/(height in m 2 )]. At the time of the 5-year follow-up, the following variables were recorded: length of parental education in years (the average educational length for the two parents or the length of maternal education if information on the father was unavailable); marital status (single either at the prenatal interview or at the follow-up/married or cohabitating at both); postnatal parental smoking (yes, if at least one of the parents smoked in the home/no if otherwise); child health status, i.e. the presence of any medical conditions (epilepsy, or syndromes such as Morpheus syndrome, von Recklinghausen s neurofibromatosis, congenital toxoplasmosis, or myxoedema) or regular medications (such as medicines for asthma and allergies, methylphenidate for ADHD, and medicines for epilepsy and respiratory conditions) that might influence test performance; dichotomised family/home environment index (presence of two or more of the following seven adverse conditions: living with only one biological parent; changes in primary caregiver; daycare for more than 8 hours a day before the age of 3 years; 14 days of separation from parents; breakfast irregularity; maternal depression; and parental alcohol use above the maximum recommended level by the Danish National Board of Health of 14 drinks/week for non-pregnant women or 21 drinks/week for men); the child s hearing ability measured on the day of the test (impaired/not impaired); and the child s vision ability measured on the day of the test (impaired/not impaired). Maternal IQ was assessed at the follow-up examination: two verbal subtests (information and vocabulary) from the Wechsler Adult Intelligence Scale 30 (WAIS) were used to assess verbal IQ, and the Raven s Standard Progressive Matrices provided non-verbal IQ. 31 Raw scores of each test were standardised based on the results from the full sample and weighted equally in a combined score that was restandardised to an IQ scale with a mean of 100 and an SD of 15. Maternal age was obtained directly from the unique Danish personal identification number, as was the child s gender and age at testing. Birthweight (g) and gestational age (days) were obtained from the Danish Medical Birth Registry. Data analysis All statistical analyses were conducted in stata 11 (Stata- Corp LP, College Station, TX, USA) and weighted by sampling probabilities with robust variance estimation. All statistical tests were two-sided and declared significant at a level of 5%. The number of missing values for the covariates ranged from two to 33, but the number of missing values on the overall attention, selective attention, and sustained attention means were 249, 136, and 162, respectively. Missing values were imputed based on two strategies: a dedicated model for imputation with missing values modelled from the other variables considered to be most predictive of each variable (specific equations are available upon request), and a blackbox strategy with all outcomes, exposures, and covariates used to predict missing values. For both strategies 100 completed data sets were generated. Regardless of imputation strategy, the main conclusions were unaffected, and the point estimates of the exposure parameters never differed by more than 18% relative to their standard error. Essentially the same results were obtained when the strategy of complete case analysis was used, and only results based on the dedicated imputation strategy will be reported. All imputations were implemented with the ice add-on command and the built-in mi estimate command in stata 11. The primary analyses used the three composite attention scores as outcomes in multiple linear regression. In supplementary analyses, the three attention outcomes were dichotomised, using the sample mean minus 1 SD as the cut-off for subnormal scores on overall, selective, and sustained attention. Logistic regression was used in these analyses, and odds ratios were calculated using the category of attention scores above the cut-off as the reference group. Parental education, maternal IQ, prenatal maternal smoking, the child s age at testing, the child s gender, and tester were considered to be core confounding factors, whereas the full model also controlled for maternal age, pre-pregnancy body mass index (BMI), parity, marital status, home environment, postnatal parental smoking, child s health status, and indicators for hearing and vision impairment. In the analyses of maternal average alcohol consumption, binge drinking (coded yes/no) was included as a potential confounding factor, whereas the average number of drinks per week consumed by the mother during pregnancy was included as a potential confounding factor in the analyses of effects of binge drinking (coded as 0, 1 4, 5 8, 9 drinks/week). Birthweight, gestational age, and the child s IQ were considered potential mediators of the effects of alcohol exposure on attention. Consequently, these variables were not included in the main analyses, but were included in separate models evaluating potential mediation ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG

5 The effects of early prenatal alcohol consumption In supplementary analyses, we analysed potential interactions between average alcohol consumption and binge drinking, as well as interactions of the alcohol exposure variables with gender of the child, parental education, prenatal maternal binge drinking, and maternal smoking during pregnancy. Furthermore, raw scores of each of the four individual TEACh-5 subtests were analysed with regression models adjusting for core and all confounding factors. Results Sample characteristics are presented elsewhere. 32,33 Notably, women reporting no alcohol consumption during a typical week were significantly younger, and were also significantly more likely to be primiparous and less likely to be smokers or have suboptimal family/home conditions. Smoking and suboptimal family/home conditions were more frequent among women who reported consuming 5 or more drinks per week. 32 Women without binge drinking episodes were significantly older, less likely to be primiparous, had significantly higher BMI, were less likely to be smokers, had a lower average alcohol intake, and had significantly lower IQs, compared with bingers. 33 No substantial s were seen between participants and non-participants. Results for low to moderate alcohol intake Overall attention mean Both the crude and adjusted analyses showed the same pattern (Table 1). The overall mean scores for the children of mothers who drank 1 4 and 5 8 drinks per week drinks per week were almost identical to the mean of the offspring of the reference group of mothers who did not drink alcohol during a typical week. By contrast, the mean attention scores associated with consuming 9 drinks per week were substantially lower than that observed for the reference group, but because of the small number of children in this category the confidence intervals were wide, and the adjusted s from the reference group were not statistically significant. However, Table 2 shows that when the Table 1. Associations between maternal average alcohol intake in pregnancy and offspring mean TEACh-5 mean scores, Denmark, Average no. drinks/week in pregnancy Crude Adjusted for core confounding factors* Adjusted for potential confounding factors** Mean score Mean 95% CI Mean 95% CI Mean 95% CI Overall attention mean*** 0 )0.03 Reference Reference Reference 1 4 ) )0.14; )0.14; )0.12; ) )0.27; )0.27; )0.26; )0.56 )0.52 )1.17; 0.12 )0.56 )1.20; 0.07 )0.45 )1.08; 0.18 P**** Selective attention mean 0 )0.04 Reference Reference Reference 1 4 ) )0.14; )0.12; )0.09; )0.15; )0.15; )0.10; )0.63 )0.58 )1.11; )0.06 )0.64 )1.17; 0.10 )0.48 )1.00; 0.03 P**** Sustained attention mean 0 )0.01 Reference Reference Reference 1 4 ) )0.15; 0.16 )0.02 )0.17; 0.13 )0.02 )0.17; )0.10 )0.09 )0.38; 0.20 )0.09 )0.38; 0.19 )0.10 )0.40; )0.24 )0.24 )0.88; 0.41 )0.25 )0.89; 0.40 )0.22 )0.89; 0.45 P**** *Parental education, maternal IQ, prenatal maternal smoking, the child s gender and age, and tester. **Parental education, maternal IQ, prenatal maternal smoking and binge drinking, maternal age, parity, prenatal and postnatal marital status, postnatal parental smoking, maternal pre-pregnancy BMI, the child s gender and age, health status, hearing and vision on the day of testing, family/home environment, and tester. ***The number of children in each exposure category was 758, 675, 175, and 20 for the 0, 1 4, 5 8 and 9 drinks/week categories, respectively. ****P value for the hypothesis of no in attention scores across levels of average alcohol intake. ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG 1215

6 Underbjerg et al. Table 2. Associations between maternal average alcohol intake in pregnancy and below-average offspring TEACh-5 scores (<1 SD below the mean), Denmark, Average no. drinks/week in pregnancy Crude Adjusted for core confounding factors* Adjusted for potential confounding factors** Odds ratio 95% CI Odds ratio 95% CI Odds ratio 95% CI Overall attention mean*** 0 Reference Reference Reference ; ; ; ; ; ; ; ; ; 9.53 P**** Selective attention mean 0 Reference Reference Reference ; ; ; ; ; ; ; ; ; P**** Sustained attention mean 0 Reference Reference Reference ; ; ; ; ; ; ; ; ; P**** *Parental education, maternal IQ, prenatal maternal smoking, the child s gender and age, and tester. **Parental education, maternal IQ, prenatal maternal smoking and binge drinking, maternal age, parity, prenatal and postnatal marital status, postnatal parental smoking, maternal pre-pregnancy BMI, the child s gender and age, health status, hearing and vision on the day of testing, family/home environment, and tester. ***The number of children in each exposure category was 758, 675, 175, and 20 for the 0, 1 4, 5 8 and 9 drinks/week categories, respectively. ****P value for the hypothesis of no in OR for attention score below the sample mean )1 SD across levels of average alcohol intake. overall attention mean was dichotomised and analysed by logistic regression models, the odds ratios associated with consuming 9 drinks per week were significant, with adjustment for both core confounding factors (OR 3.61, 95% CI ) and all potential confounding factors (OR 3.20, 95% CI ). Selective attention For the mean selective attention scores, a similar pattern was observed (Table 1). The mean scores of offspring of mothers who drank 1 4 and 5 8 drinks per week were similar to the mean scores observed for children whose mothers did not drink alcohol. However, the children of mothers who consumed 9 drinks per week performed about 0.5 SD below the reference group, and the was significant for both the crude and the core confounding factor-adjusted analyses. Table 2 shows that results were corroborated when the dichotomised selective attention score was analysed with significant odds ratios for the crude and the core confounding factor-adjusted logistic regression models (OR 3.32, 95% CI , and OR 4.42, 95% CI , respectively). However, when adjusted for all potential confounding factors, the mean and the adjusted odds ratio were insignificant (OR 3.22, 95% CI ), albeit of a comparable magnitude. Sustained attention Corresponding to the pattern for selective attention, the mean sustained attention scores of the children whose mothers drank 1 4 and 5 8 drinks per week were similar to the means of the reference group (Table 1). The mean of the children whose mothers drank 9 drinks per week was only 0.24 SD below the mean of the reference group, but analysis of the dichotomised variable with adjustment for all potential confounding factors showed a significantly increased risk of low sustained attention performance for the 9 group (OR 4.19, 95% CI ). The supplementary analyses showed no consistent effects of alcohol exposure on raw scores for individual TEACh-5 subtests. The dichotomised barking score showed a significantly increased OR associated with consuming 9 drinks per week during pregnancy, with adjustment for core 1216 ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG

7 The effects of early prenatal alcohol consumption confounding factors (OR 3.31, 95% CI ), but the increase in OR was not statistically significant when adjusted for all potential confounding factors. Tests of interactions with gender, parental education, and prenatal maternal binge drinking were not statistically significant. A test of the interaction with prenatal smoking was marginally significant for the mean selective attention score, and stratified analysis showed a trend towards stronger effects of maternal consumption of 9 drinks per week among the children of mothers who also smoked. Adjustment for gestational age and birthweight did not change the results for any of the analyses, whereas adjustment for the child s WPPSI-R full-scale IQ was associated with smaller effect estimates for the 9 drinks per week consumption category. Results for binge drinking No significant s were observed for the dichotomous binge variable, for the number of binge drinking episodes during pregnancy, or for the timing of binge drinking. For the overall attention mean, unadjusted analyses showed a small, but significant, increase in attention score associated with binge drinking during gestational weeks 5 8 (Table 3). However, when adjusting for core and potential confounding factors, no significant or clinically relevant s were observed (Table 3). These results were corroborated by linear regression analyses of the selective and sustained attention scores, which showed no significant or clinically relevant effects for the dichotomous binge variable, for the number of binge drinking episodes during pregnancy, or for the timing of binge episodes in the unadjusted or the adjusted analyses (data not shown). The number of binge episodes reported in early pregnancy did not vary with time of interview. Corroborating these findings, supplementary analyses of the dichotomised overall attention mean score, and the dichotomised selective and sustained attention scores, showed no significant associations with any binge variable (data not shown). Supplementary analyses also showed no con- Table 3. Associations between maternal binge drinking in pregnancy and TEACh-5 offspring overall attention mean, Denmark, Crude Adjusted for core confounding factors* Adjusted for potential confounding factors** Mean score Mean 95% CI Mean 95% CI Mean 95% CI Binge drinking in pregnancy*** No )0.04 Reference Reference Reference Yes )0.07; 0.17 )0.01 )0.13; 0.11 )0.02 )0.14; 0.10 P**** Number of binge drinking episodes in pregnancy*** 0 )0.04 Reference Reference Reference 1 ) )0.10; 0.16 )0.01 )0.14; 0.11 )0.02 )0.15; )0.11;.0.25 )0.02 )0.19; 0.16 )0.02 )0.20; )0.15; )0.21; )0.19; 0.27 P**** Timing of binge drinking episodes in pregnancy (gestational week)*** No binge drinking )0.04 Reference Reference Reference )0.08; )0.15; )0.16; )0.07 )0.03 )0.19; 0.13 )0.09 )0.25; 0.07 )0.11 )0.27; ; )0.02; )0.02; )0.12; )0.12; )0.13; 0.22 Multiple episodes )0.13; )0.17; )0.17; 0.21 P**** *Parental education, maternal IQ, prenatal maternal smoking, age at testing, gender of child, and tester. **Parental education, maternal IQ, prenatal maternal smoking, prenatal maternal average alcohol intake, maternal age, parity, prenatal and postnatal marital status, postnatal parental smoking, maternal pre-pregnancy BMI, gender of child, age at testing, health status, hearing and vision on the day of testing, and family/home environment and tester. ***Of the 1617 mothers, 1122 reported binge drinking on one or more occasions (range: 3 12 episodes). The number of children in each category was 495, 783, 225, and 114 for 0, 1, 2, and 3+ episodes, respectively. The number of children in each timing category was 495, 237, 261, 216, 234, and 174. ****P value for the hypothesis of no in mean attention scores across levels of average alcohol intake. ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG 1217

8 Underbjerg et al. sistent effects of binge drinking on raw scores from individual TEACh-5 subtests. For the overall attention mean score, the interaction between parental education and the dichotomous binge variable was significant (P = 0.05) in the model adjusting for core confounding factors, and was marginally significant in the model adjusting for all potential confounding factors (P = 0.06). Stratified analyses showed a non-significant negative effect of binge drinking among offspring of parents with high education, and a non-significant positive effect of binge drinking among offspring of parents with low education. Discussion In this study maternal binge drinking was not associated with consistent effects, but a consistent pattern was observed for the effects of maternal low to moderate consumption on the three attention composite means: the consumption of 1 4 and 5 8 drinks per week during pregnancy was not associated with adverse effects, whereas consuming 9 drinks per week during pregnancy was associated with significant or marginally significant negative effects. For the overall attention mean score no significant effects were observed in the linear models, whereas the logistic models showed significant unadjusted and adjusted odds ratios. The effects on the selective attention mean score were significant in both crude and core confounding factor-adjusted linear and logistic analyses, but when the analyses included adjustment for all potential confounding factors, the effects became smaller and non-significant. Finally, the only significant association observed for the sustained attention mean score was significantly increased odds for performing below the dichotomised cut-off level when adjusted for all potential confounding factors. Significant effects were only observed for offspring of mothers who drank 9 drinks per week, and because of the small sample size in this category, all confidence intervals were wide. This is one of very few studies to specifically address the effects of low to moderate prenatal alcohol exposure on offspring attention functions. The finding of no significant effects associated with consuming <9 drinks per week during pregnancy seems to be consistent with results from a longitudinal study reporting no association between low to moderate (or high) prenatal alcohol exposure and various attention measures, 10 with studies reporting no association between low to moderate consumption and a questionnaire on reported attention problems, 34 and with null findings in studies using computerised measures of (sustained) attention. 9,11 Furthermore, it can be argued that the apparent effects associated women consuming 9 or more drinks per week during pregnancy are in line with the findings of several studies based on the The Seattle Longitudinal Prospective Study on Alcohol and Pregnancy, reporting adverse effects of prenatal alcohol exposure on attention functions, in particular on vigilance/sustained attention. 4,5,34,35 It is well established that prenatal alcohol exposure may affect offspring intelligence, 2 and as correlations have been observed between intelligence and attention measures, 24 effects on intelligence may mediate effects on attention functions. Consequently, we conducted separate analyses including the child s IQ as a covariate to evaluate the specific effects of prenatal low to moderate alcohol exposure on attention, independent of effects on global cognitive function. Including the IQ of offspring as a covariate attenuated the effect estimates associated with consuming 9 drinks per week during pregnancy, and it is possible that the effects of prenatal alcohol exposure on general intelligence to some extent mediate effects on attention. Overall, no systematic associations between binge drinking during early pregnancy and offspring attention scores were observed in this study. The evidence of improved performance associated with binge drinking in gestational weeks 5 8 was only significant in the unadjusted analyses, and indeed the results presented in Table 3 are consistent in showing no significant effects and effect estimates that were close to zero. The analyses of the dichotomised attention composite means corroborated these findings, which were also supported when the overall attention mean score was analysed together with measures of intelligence and executive functioning. 29 Thus, this large-scale study suggests no or clinically insignificant effects of a few binge drinking episodes during early pregnancy on the attention measure used in this study. Only few studies have investigated associations between binge drinking and cognitive outcomes. Of the two studies including attention measures, one study used a vigilance paradigm and reported poorer performance in 7-year-old children of binge drinkers than in children whose mothers did not binge drink during pregnancy. 36 The other study used questionnaires and reported no statistically significant association between the attention scores of the children and the mothers binge drinking during pregnancy. 37 Only one study specifically addressed the issue of the timing of the binge episodes during pregnancy, using trimesters as a rough differentiation of time periods, but that study did not include attention measures. 38 The general inconsistency in results regarding the effects of low to moderate prenatal alcohol exposure and maternal binge drinking on attention functions may be the result of methodological s reflecting different study designs, different exposure and/or outcome measures, and different conceptualisation of attention. Our study was based on prospectively collected information on alcohol consumption during pregnancy, obtained through tele ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG

9 The effects of early prenatal alcohol consumption phone interviews asking beverage-specific questions about weekly alcohol consumption: a method that has proven reliable in a Danish pregnant population. 39 However, the distributions within the 1 4 and 5 8 drinks per week categories were skewed: on average, the observed exposure represents the lower tail of the distribution for these consumption categories, suggesting that the findings are more in line with occasional weekly drinking (1 or 2 drinks per week for the 1 4 drinks/week group) or, at most, levels of less than one drink per day. The study may not have had sufficient statistical power to detect small effects associated with such low levels of consumption, although the detected mean s were very small. On the other hand, potential problems with under-reporting and subsequent misclassification may not be negligible, and should be considered, particularly when attempting to identify a safe lower limit for alcohol consumption during pregnancy. Interpretation of results regarding binge drinking must take into account that blood alcohol concentration is not only determined by the actual dose of alcohol consumed by the women, but also by several other important factors such as the time perspective, body size, and individual metabolism. 40 Binge drinking is commonly defined as the intake of 5 or more drinks on a single occasion, but this definition does not distinguish between women who consumed 5 drinks or more within one or a few hours, versus women who spread their intake over several hours. Similarly, the definition of binge drinking does not incorporate whether the woman consumed 5 or 10 drinks on a single occasion. Our study was based on prospectively collected information obtained through telephone interviews asking the number of times the woman consumed 5 drinks or more on a single occasion, and during what week(s) of gestation these episodes took place: a method that has proven reliable in a Danish pregnant population. 39,41 The participants generally had a low to moderate average alcohol intake, and the reported number of drinks per week was included in the binge analyses, adjusting for all potential confounding factors. The availability of data on potential confounding factors is crucial, and our study took advantage of the prospectively collected and detailed information in the DNBC. 29 This information was supplemented with the assessment of other important covariates at follow-up. In particular, efforts were made to obtain reliable measures of parental education, and of maternal general intellectual abilities. Measures of maternal and paternal IQ are rarely available, and among the studies mentioned, only one seems to have assessed maternal intelligence. 10 Our study lacked information on some potentially important confounding factors such as prenatal maternal medication, and alcohol consumption in late pregnancy, and even for the included covariates there may be residual confounding as a result of gross categorisation/dichotomisation. Using the recently developed TEACh-5 battery has both strengths and weaknesses. Along with the child-friendly appearance and the relatively short time needed to administer the test battery, the most obvious advantage was the theoretical foundation of the battery that makes it possible to derive measures of both selective and sustained attention. However, although the psychometric properties of the TEACh-5 have been described in detail, 24 the test battery still needs to prove its validity as a measure of attention that can detect relatively small effects of prenatal exposures on offspring cognitive development. Finally, it is also important to consider the relatively large number of statistical tests when evaluating the significant effects of maternal average consumption. The large number of statistical tests may of course be associated with spurious significant results, but we decided against adjustment for multiple tests because of the risk of masking true associations, and because, from a public health perspective, overlooking indications of possibly harmful effects of low to moderate and binge doses of a known teratogen may be considered more serious than spurious significance. Furthermore, we believe that the observed pattern of no effects with maternal consumption below 9 drinks per week, and the indications of effects from the consumption of 9 drinks per week during pregnancy, is both biologically meaningful and consistent. In addition, we believe that the observed pattern of effects in the TEACh-5 attention composite mean scores is corroborated by other studies suggesting effects of moderate maternal alcohol consumption on offspring attention. However, the small sample of women who consumed 9 drinks per week implies a lack of statistical power, and experience with the TEACh-5 as a measure of attention is still limited. For these reasons, and because of the few and contradictory previous studies, there is an obvious need for more studies focusing on the effects of low to moderate prenatal alcohol exposure on specific attention functions in offspring. Similarly, the strengths and weaknesses of our study should be considered when interpreting the lack of consistent associations between binge drinking during early pregnancy and offspring attention, even after the adjustment for maternal intelligence and a wide set of other potential confounding factors. Disclosure of interests None to declare. Contribution to authorship MU developed the TEACh-5 attention test. MU, USK, NIL, LB, TRK, PT, and ELM contributed to the design of the Lifestyle During Pregnancy Study. MU wrote the first draft of the article, and JG, TW, and CS were responsible for the statistical analyses. All authors contributed to the interpre- ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG 1219

10 Underbjerg et al. tation of the results, and provided critical comments and revisions of the article. Details of ethics approval The study was approved by the DNBC Board of Directors, the DNBC Steering Committee, the Regional Ethics Committee, the Danish Data Protection Agency, and the Institutional Review Board at the Centers for Disease Control and Prevention. Signed informed consent was obtained for the LDPS. Funding This study was primarily supported by the Centers for Disease Control and Prevention (CDC), Atlanta,Georgia,USA. Acknowledgement The Danish National Research Foundation established the Danish Epidemiology Science Centre that initiated and created the Danish National Birth Cohort. The cohort was furthermore a result of a major grant from this Foundation. Additional support for the Danish National Birth Cohort was obtained from the Pharmacy Foundation, the Egmont Foundation, the March of Dimes Birth Defects Foundation, the Augustinus Foundation, and the Health Foundation. The authors would like to thank all the participants for their time, as well as Dr Tom Manly for his contribution to the development of the TEACh-5 test. j References 1 Coles CD, Platzman KA, Lynch ME, Freides D. Auditory and visual sustained attention in adolescents prenatally exposed to alcohol. Alcohol Clin Exp Res 2002;26: Kodituwakku PW. Neurocognitive profile in children with fetal alcohol spectrum disorders. Dev Disabil Res Rev 2009;15: Nanson JL, Hiscock M. Attention deficits in children exposed to alcohol prenatally. Alcohol Clin Exp Res 1990;14: Streissguth AP, Barr HM, Sampson PD, Parrish-Johnson JC, Kirchner GL, Martin DC. Attention, distraction and reaction time at age 7 years and prenatal alcohol exposure. Neurobehav Toxicol Teratol 1986;8: Streissguth AP, Sampson PD, Olson HC, Bookstein FL, Barr HM, Scott M et al. Maternal drinking during pregnancy: Attention and short-term memory in 14-year-old offspring a longitudinal prospective study. Alcohol Clin Exp R 1994;18: O Connor MJ, Paley B. Psychiatric conditions associated with prenatal alcohol exposure. Dev Disabil Res Rev 2009;15: O Malley KD, Nanson J. Clinical implications of a link between fetal alcohol spectrum disorder and attention-deficit hyperactivity disorder. Can J Psychiatry 2002;47: Boyd T, Ernhart CB, Greene T, Sokol R, Martier S. Prenatal alcohol exposure and sustained attention in the preschool years. Neurotoxicol Teratol 1991;13: Fried PA, Watkinson B, Gray R. A follow-up study of attentional behavior in 6-year-old children exposed prenatally to marihuana, cigarettes and alcohol. Neurotoxicol Teratol 1992;14: Richardson GA, Ryan C, Willford J, Day NL, Goldschmidt L. Prenatal alcohol and marijuana exposure: effects on neuropsychological outcomes at 10 years. Neurotoxicol Teratol 2002;24: Leech L, Richardson GA, Goldschmidt L, Day NL. Prenatal substance exposure: effects on attention and impulsivity of 6-year-old. Neurotoxicol Teratol 1999;21: Norman AL, Crocker N, Mattson SN, Riley EP. Neuromaging and fetal alcohol spectrum disorders. Dev Disabil Res Rev 2009;15: Spadoni AD, McGee CL, Fryer SL, Riley EP. Neuroimaging and fetal alcohol spectrum disorders. Neurosci Biobehav Rev 2007;31: Gray R, Murkherjee RAS, Rutter M. Alcohol consumption during pregnancy and its effects on neurodevelopment: what is known and what remains uncertain. Addiction 2009;104: Willford J, Leech S, Day N. Moderate prenatal alcohol exposure and cognitive status of children at age 10. Alcohol Clin Exp Res 2006;30: Henderson J, Gray R, Brocklehurst P. Systematic review of effects of low-moderate prenatal alcohol exposure on pregnancy outcome. BJOG 2007;114: Bonthius DJ, Goodlett CR, West JR. Blood alcohol concentration and severity of microencephaly in neonatal rats depend on the pattern of alcohol administration. Alcohol 1988;5: Henderson J, Kesmodel U, Gray R. Systematic review of the fetal effects of prenatal binge-drinking. J Epidemiol Community Health 2007;61: Lyon GR, Krasnegor NA. Attention, Memory and Executive Function. Baltimore, MD: Paul H. Brookes Publishing Co., Posner MI, Petersen SE. The attention system of the human brain. Annu Rev Neurosci 1990;13: Dolan PD, Stone DH, Briggs AH. A systematic review of continuous performance task research in children prentally exposed to alcohol. Alcohol Alcohol 2010;45: Mattson SN, Calarco KE, Lang AR. Focused and shifting attention in children with heavy prenatal alcohol exposure. Neuropsychology 2006;20: Burden MJ, Jacobson SW, Sokol RJ, Jacobson JL. Effects of prenatal alcohol exposure on attention and working memory at 7.5 years of age. Alcohol Clin Exp Res 2005;29: Underbjerg M. Prenatal alcohol exposure: effects on children s attention functions. PhD Thesis, Faculty of Health Sciences, University of Aarhus, Aarhus, Kesmodel US, Underbjerg M, Kilburn TR, Bakketeig LS, Mortensen EL, Landrø NI, et al. Lifestyle during pregnancy: neurodevelopmental effects at 5 years of age The design and implementation of a prospective follow-up study. Scand J Public Health 2010;38: Olsen J, Melbye M, Olsen SF, Sørensen TI, Aaby P, Andersen AM, et al. The Danish National Birth Cohort its background, structure and aim. Scand J Public Health 2001;29: Wechsler D. Manual for the Wechsler Preschool and Primary Scale of Intelligence Revised. UK edition. Sidcup, Kent: The Psychological Corporation, Manly T, Anderson V, Nimmo-Smith I, Turner A, Watson P, Robertson IH. The differential assessment of children s attention: The Test of Everyday Attention for Children (TEA-Ch), normative sample and ADHD performance. J Child Psychol Psychiatry 2001;42: Kesmodel US, Bertrand J, Støvring H, Skarpness B, Denny CH, Mortensen EL, et al. The effect of different alcohol drinking patterns in early to mid pregnancy on the child s intelligence, attention, and executive function. BJOG 2012;119; DOI: /j x ª 2012 The Authors BJOG An International Journal of Obstetrics and Gynaecology ª 2012 RCOG