External Publications Using GUI Data

Background
Height is regarded as a marker of early-life illness, adversity, nutrition and psychosocial stress, but the extent to which differences in height are determined by early-life socioeconomic circumstances, particularly in contemporary populations, is unclear. This study examined socioeconomic differences in children’s height trajectories from birth through to 21 years of age in four European countries.

Methods
Data were from six prospective cohort studies—Generation XXI, Growing Up in Ireland (infant and child cohorts), Millennium Cohort Study, EPITeen and Cardiovascular Risk in Young Finns Study—comprising a total of 49 492 children with growth measured repeatedly from 1980 to 2014. We modelled differences in children’s growth trajectories over time by maternal educational level using hierarchical models with fixed and random components for each cohort study.

Results
Across most cohorts at practically all ages, children from lower educated mothers were shorter on average. The gradient in height was consistently observed at 3 years of age with the difference in expected height between maternal education groups ranging between −0.55 and −1.53 cm for boys and −0.42 to −1.50 cm for girls across the different studies and widening across childhood. The height deficit persists into adolescence and early adulthood. By age 21, boys from primary educated maternal backgrounds lag the tertiary educated by −0.67 cm (Portugal) and −2.15 cm (Finland). The comparable figures for girls were −2.49 cm (Portugal) and −2.93 cm (Finland).

Conclusions
Significant differences in children’s height by maternal education persist in modern child populations in Europe.

Both psychological and sociological accounts have suggested that the home learning environment play an important role in children’s cognitive development and may provide insights into inequalities in cognitive outcomes. Using the infant cohort of the Growing Up in Ireland Study (GUI), this chapter investigates firstly if differences in the home learning environment at age three helps to explain the social gradient in childhood cognitive outcomes, measured as expressive vocabulary, at age five; and second, can childcare outside the home compensate for a poor home learning environment? Home learning environment is measured as the number of children’s books in the home and a combined index of parental activities with the child that includes reading, crafts and games. A rich home learning environment at age 3 is associated with higher vocabulary scores at age 5 years for all children. The children of lower educated parents tend to live in poorer home learning environments, and this partly explains their lower vocabulary scores at age 5. The chapter also provides some evidence that centre-based childcare was associated with an increase in vocabulary score for children from poor home learning environments. However, this effect is very small and only slightly reduces the gap in vocabulary scores between children from a rich and poor home learning environment.

Using the first wave of the Growing Up in Ireland Survey of nine year old children we examine whether a teacher’s assessment of their pupil’s academic ability is influenced by the weight status of the child and/or the child’s mother. Multivariate regression analyses of the teacher’s assessment, controlling for the child’s actual test performance, their BMI, their mother’s BMI, other socio-demographic and teacher characteristics were undertaken. The study highlighted that child BMI was not a significant determinant but that children whose mother was obese were more likely to be rated as below average in reading and in maths compared to those whose mother was leaner, after adjusting for their measured ability. The potential for mother’s weight status to influence teachers’ assessments of their children’s perceived ability could have long term ramifications for educational outcomes and warrants further study.

Background
The gold standard for categorisation of weight status is clinically measured body mass index (BMI), but this is often not practical in large epidemiological studies.

Objectives
To determine if a child’s weight perception or a mother’s perception of a child’s weight status is a viable alternative to measured height and weight in determining BMI classification. Secondary outcomes are to determine the influence of a mother’s BMI on her ability to categorise the child’s BMI and a child’s ability to recognise his/her own BMI.

Methods
Cross-sectional analysis of the growing up in Ireland cohort study, a nationally representative cohort of 8568 9-year-old children. The variables considered for this analysis are the child’s gender, BMI (International Obesity Taskforce grade derived from measured height and weight) and self-perceived weight status, and the mother’s weight perception of the child, BMI (derived from measured height and weight) and self-perceived weight status. Cohen’s weighted-kappa was used to evaluate the strength of the agreement between pairwise combinations of the BMI variables. Cumulative and adjacent categories logistic regression were used to predict how likely a person rates themselves as under, normal or overweight, based on explanatory variables.

Results
Mothers are more accurate at correctly classifying their child’s BMI (κ=0.5; confidence intervals (CI) 0.38–0.51) than the children themselves (κ=0.25; CI 0.23–0.26). Overweight mothers are better raters of their child’s BMI (κ=0.51; CI 0.49–0.54), compared with normal (κ=0.44; CI 0.41–0.47) or underweight mothers (κ=0.4; CI 0.22–0.58), regardless of whether the mother’s BMI is derived from measured height and weight or self-perceived. The mother’s perception of the child’s weight status is not an influencing factor on the child’s ability to correctly classify him/herself, but the child’s self-perceived weight status influences the mother’s ability to correctly classify the child.

Conclusions
A mother’s BMI classification of her child is a viable alternative to BMI measurement in large epidemiological studies.

Economic recessions have been linked to adult health, but few studies have examined how recessions influence the health of young children. This study examined the impact of life transitions linked to the recent financial crisis on the health of young children in Ireland. Data came from the Growing Up in Ireland Infant Cohort Study (n = 11,134), which assessed children before (2008), during (2011), and after (2013) the Great Recession that followed the financial crisis of 2008 and incorporated questions on the impacts of the financial crisis on families. Using fixed-effects models to control for confounding, we found that a reduction in welfare benefits during the recession was associated with a significant increase in the risks of asthma (β = 0.014, 95% confidence interval (95% CI): 0.004, 0.023) and atopy (β = 0.014, 95% CI: 0.001, 0.027). While parental job loss was not associated with child health, a reduction in working hours was associated with increased reports of child health problems (β = 0.024, 95% CI: 0.004, 0.043), as were difficulties affording basic necessities (β = 0.019, 95% CI: 0.001, 0.038). Results suggest that failing to protect vulnerable families and children during economic recessions may have long-lasting implications for child health.

This study uses Growing Up in Ireland data to examine how 9,000 children adjusted to primary school. It finds that the vast majority of five year olds are positive about school, look forward to going to school and say good things about school. The study showed that children start school with different skills and capacities and some children face greater challenges. The study suggests a number of ways to help all children experience a positive transition to primary school.

Educational expectations and the way in which they shape actual outcomes act as an important vehicle for the intergenerational reproduction of social inequality. This article draws on rich information from the Growing Up in Ireland longitudinal study to explore the factors influencing the educational expectations of 13-year-old young people and their parents. The findings point to high levels of expectations among parents but one-third of young people are found to hold lower expectations than their parents. Both sets of expectations are shaped by parental education, social class and household income, though parental expectations are more strongly structured by social background than those of young people. The article seeks to contribute to the literature on educational expectations in three interrelated ways. Firstly, it focuses on a phase of schooling, the transition to secondary education, which represents a time of significant change in young people’s learning experiences, and transition difficulties are found to dampen their expectations. Secondly, the analyses unpack the extent to which signals from the school influence parental perspectives, with academic achievement, child attitudes to maths, ability group assignment and teacher reprimands emerging as significant drivers of parental views of their children’s potential. Thirdly, the analyses take account of both the primary and secondary school attended and indicate significant between-school differences in young people’s intentions to go on to higher education.

Keywords
Educational expectations, social class differences, parent influence, school effects

Collecting accurate and detailed dietary intake data is costly at a national level. Accordingly, limited dietary assessment tools such as Short Food Questionnaires (SFQs) are increasingly used in large surveys. This paper describes a novel method linking matched datasets to improve the quality of dietary data collected. Growing Up in Ireland (GUI) is a nationally representative longitudinal study of infants in the Republic of Ireland which used a SFQ (with no portion sizes) to assess the intake of “healthy” and “unhealthy” food and drink by 3 years old preschool children. The National Preschool Nutrition Survey (NPNS) provides the most accurate estimates available for dietary intake of young children in Ireland using a detailed 4 days weighed food diary. A mapping algorithm was applied using food name, cooking method, and food description to fill all GUI food groups with information from the NPNS food datafile which included the target variables, frequency, and amount. The augmented data were analyzed to examine all food groups described in NPNS and GUI and what proportion of foods were covered, non-covered, or partially-covered by GUI food groups, as a percentage of the total number of consumptions. The term non-covered indicated a specific food consumption that could not be mapped using a GUI food group. “High sugar” food items that were non-covered included ready-to-eat breakfast cereals, fruit juice, sugars, syrups, preserves and sweeteners, and ice-cream. The average proportion of consumption frequency and amount of foods not covered by GUI was 44 and 34%, respectively. Through mapping food codes in this manner, it was possible, using density plots, to visualize the relative performance of the brief dietary instrument (SFQ) compared to the more detailed food diary (FD). The SFQ did not capture a substantial portion of habitual foods consumed by 3-year olds in Ireland. Researchers interested in focussing on specific foods, could use this approach to assess the proportion of foods covered, non-covered, or partially-covered by reference to the mapped food database. These results can be used to improve SFQs for future studies and improve the capacity to identify diet-disease relationships.

This paper presents an analysis of primary schoolchildren’s attitudes to the Irish language, Gaeilge, in the context of national policy in the Republic of Ireland. In particular, the study examines the factors (social, cultural, cognitive and organisational) that may be related to a pronounced excess in disengagement with Irish over and above general engagement with schooling. The data derives from the nine-year-old child cohort of the national longitudinal study of children, the Growing Up in Ireland (GUI) survey. Unlike previous findings in relation to general engagement with school, excess disengagement with Irish is not found to be related to socio-economic or ethnic factors, factors related to teacher profile, training, self-efficacy or style or school ethos. As expected, children with less exposure to spoken Irish in school or at home are more likely to show excess disengagement with the language. Literacy activity, special education needs and school patronage are also found to be related to excess disengagement. Significant random effects indicate the importance of the individual teacher and school in the development of pupil engagement with Irish. The current study presents a timely perspective on attitudes to Irish among the younger population and possible implications for policy.

KEYWORDS: Irish language; motivation; minority languages; bilingualism; biliteracy

Objective
To longitudinally investigate body mass index (BMI) in young children in Ireland and identify factors and critical time points associated with changes in BMI.

Study design
Data on 11 134 children were collected in the nationally representative Growing Up in Ireland infant cohort study. Height and weight were measured at 9 months, 3 years, and 5 years of age. Multilevel regression was used to identify risk factors associated with changes in BMI over time (n = 10 377), combining a unique set of covariates collected from the child and the 2 main caregivers (usually the mother and father).

Results
The proportion of children ≥85th percentile of World Health Organization growth criteria was 39% at 9 months, 44% at 3 years, and 30% at 5 years. Children born large for gestational age (13%) and those with rapid infant weight gain (25%) consistently had higher BMI. Low average BMIs were consistently seen in children born small for gestational age (10%) or before 37 weeks (7%). Smaller variations in BMI existed for other factors including ethnicity, household structure, caregiver weight status, breastfeeding, sex, socioeconomic status, sleeping hours, childcare, and region.

Conclusions
In this study, differences at birth and in infancy appear to be most strongly associated with variation in BMI at all ages. Nevertheless, belonging to a number of other high-risk groups cumulatively could lead children to develop critical weight states. Policy-makers should target families with interventions before and during pregnancy when dominant risk factors are still modifiable. Longer-term follow-up of children may be needed to study associations later in childhood.

Keywords
Growing Up in Ireland, body mass index, childhood obesity, growth trajectories, multilevel modelling

The present study investigated the home learning environment of three to five-year-old children (n = 429) living in an area designated as socio-economically disadvantaged, involved in the Area Based Childhood (ABC) programme, compared to a nationally representative sample of three-year-old children (n = 9793), from the Growing Up in Ireland (GUI) Study. Statistical analysis of the frequency of engagement in home learning activities across both samples, revealed a significant difference in the environments to which children are exposed, with families from the GUI sample engaging more frequently in these activities than families from the ABC sample. Among the family demographic factors investigated, parent’s age and household type were significantly related to the frequency of engagement in home learning activities. Based on these findings, policy and practice implications are discussed.

Keywords
home learning environment, socio-economic disadvantage, family demographics, early childhood development

Caesarean section (CS) rates are increasing globally and exceed 50% in some countries. Childhood obesity has been linked to CS via lack of exposure to vaginal microflora although the literature is inconsistent. We investigated the association between CS birth and the risk of childhood obesity using the nationally representative Growing-Up-in-Ireland (GUI) cohort. The GUI study recruited randomly 11134 infants. The exposure was categorised into normal vaginal birth (VD) [reference], assisted VD, elective (planned) CS and emergency (unplanned) CS. The primary outcome measure was obesity defined according to the International Obesity Taskforce criteria. Statistical analysis included multinomial logistic regression with adjustment for potential confounders. Infants delivered by elective CS had an adjusted relative risk ratio (aRRR) = 1.32; [95% confidence interval (CI) 1.01–1.74] of being obese at age three years. This association was attenuated when macrosomic children were excluded (aRRR = 0.99; [95% CI 0.67–1.45]). Infants delivered by emergency CS had an increased risk of obesity aRRR = 1.56; [95% CI 1.20–2.03]; this association remained after excluding macrosomic children. We found insufficient evidence to support a causal relationship between elective CS and childhood obesity. An increased risk of obesity in children born by emergency CS, but not elective, suggests that there is no causal effect due to vaginal microflora.

Objectives
Migrant youths endure many challenges. Such challenges can be stressful and lead to psychological difficulties. We investigated the relationship between migration, psychopathology and stressful events in children and adolescents. We hypothesised that migrant youths would show higher levels of psychopathology and more stressful life events than non-migrant youths.

Method
Using the Child cohort (Cohort ‘98) of the ‘Growing up in Ireland’ study we investigated psychopathology, as measured by the Strengths and Difficulties questionnaire (SDQ) at age 9 and 13 and stressful life events in migrant and non-migrant youths.

Results
There was no significant difference between the proportion of migrant and non-migrant youths reporting psychopathology in childhood (p>0.05) or adolescence (p>0.05). Analysis of the SDQ subscales revealed that a significantly greater proportion of migrant youths had hyperactivity problems in childhood (p = 0.04) but a greater proportion of non-migrant youths had emotional problems in early adolescence (p = 0.04). We found that migrant youths experienced significantly more stressful life events than their non-migrant counterparts (p<0.01), however, once ‘Moving house/country‘ was removed as a stressor, there was no difference between the groups (p>0.27).

Conclusions
Contrary to our hypothesis, we observed that there were few differences between migrant and non-migrant youths in the levels of psychopathology. Migrant youths experienced a greater number of stressful life events, however, this was attributable to stressors relating to moving. An increased understanding of the factors promoting resilience, as demonstrated by the migrant youths, could aid health professionals and policy makers to effectively tailor interventions for mental health promotion.

Keywords
Early life stress, immigration, migrants, psychopathology

Purpose
This study examines the expressive language and speech of twins, relative to singletons, at 3 and 5 years, with the aim of determining if a twinning effect occurs during this developmental period. The possibility of twins outgrowing a twinning effect was investigated.

Method
A weighted population-based sample of 185 twins and 1,309 closely spaced singletons who participated in the Growing Up in Ireland study was analyzed. Their development was compared using a standardized expressive vocabulary assessment and parent interview at 3 and 5 years as well as teacher reports at 5 years. Causal steps mediation analyses were subsequently conducted to determine if gestation and birth weight mediated twin–singleton differences in expressive vocabulary development.

Results
Twins’ expressive vocabulary scores were slightly lower than those of singletons at 3 years, but there was no significant difference at 5 years. More parents of twins reported “a little” concern about their child’s expressive language and speech skills at 3 years, although there was no significant twin–singleton difference 2 years later. Teachers’ ratings of twins’ and singletons’ expressive language and speech skills at 5 years were comparable. Shorter gestations and lower birth weights mediated twins’ lower expressive vocabulary scores at 3 years.

Conclusions
A very small twinning effect on expressive vocabulary development exists at 3 years and is outgrown by 5 years. Twins born after shorter gestations or at a lower birth weight are at an increased risk of having less developed expressive vocabulary skills at 3 years.

Digital technologies have become an increasingly prominent feature of children’s lives both within and outside educational environments (McCoy, Quail, and Smyth 2012. Influences on 9-Year-Olds’ Learning: Home, School and Community. Dublin: Department of Children and Youth Affairs). Despite considerable media debate, we have little robust evidence on the impact of technology use on children’s development, both academically and socially. Much of the literature in this area relies on small-scale cross-sectional studies. Using longitudinal data on 8500 9-year-old children in Ireland, we examine the influence of early mobile phone ownership on children’s performance in reading and maths between 9 and 13 years of age. Across both reading and maths domains, children who already report owning a phone by the age of nine fare less well in terms of their academic development as they move into adolescence. The measured effects are sizeable, implying about 4 percentile lower ranking on standardised tests for an average student. Our results are consistent with the idea that there may be significant educational costs arising from early mobile phone use by children. Parents and policymakers should consider whether the benefits of phone availability for children are sufficiently large to justify such costs. We suggest a range of direct and indirect cognitive effects that could help explain these results.

KEYWORDS: Mobile phone ownership; child development; academic outcomes; early owners; longitudinal data

Background
Childhood adversity is a well-established risk factor for psychopathology; however, many who experience adversity do not go on to develop psychopathology. Poor self-concept and poor parental support are known risk factors for adolescent psychopathology, which may account for some of this mechanism.

Aims
To investigate candidate mediators in the relationship between childhood adversity and psychopathology.

Method
We used data from the age 9 and 13 waves of the child-cohort of the Growing Up in Ireland study. We undertook mediation analysis by path decomposition of the relationship between childhood adversity and psychopathology (internalising and externalising problems) at age 13 and persistent psychopathology. Candidate mediators were self-concept, parent–child relationship and hobby participation at age 9.

Results
Childhood adversity was reported by 28.2% of participants, and was significantly associated with internalising and externalising problems. Parent–child conflict mediated the relationship between childhood adversity and both age 13 and persistent psychopathology, accounting for 52.4% of the relationship between childhood adversity and persistent externalising problems (indirect odds ratio, 1.30; 95% CI 1.19–1.43) and 19.2% for persistent internalising problems (indirect odds ratio, 1.24; 95% CI 1.15–1.34). There was a small mediating effect of self-concept. Hobby participation and positive parent–child relationship did not mediate these relationships.

Conclusions
Parent–child conflict explains almost half the relationship between childhood adversity and persisting externalising problems in adolescence, and a fifth of the relationship with persisting internalising problems. This suggests parent–child conflict is a good target for interventions in childhood to prevent adolescent psychopathology.