Proximal Processes and Contextual Factors Associated with Early Socio-emotional Competence Development.

There is growing interest in child socio-emotional competence from parents, educators, employers and policy makers, with emphasis on developing it as early as possible. The aim of the present study was to examine contextual and proximal factors that influence socio-emotional competence development across the first five years of a child's life. We used data from 3200 mothers and their children drawn mostly from four major data collection waves (antenatal, 9 months, 2 years and 4.5 years) of the population-based longitudinal study, Growing Up in New Zealand. Regression analyses were carried out to identify the predictors of socio-emotional competence after controlling for demographics and prior score(s) of socio-emotional competence. We found that specific maternal behaviours, such as playing games and playing with toys with children, singing songs or telling stories to them, reading books with them, having rules around viewing TV, DVDs and videos, and praising children have a positive effect on socio-emotional competence. Parental relationship warmth and less family stress at 9 months also made positive contributions to socio-emotional competence at 9 months and 2 years. In contrast, attending childcare and having more siblings at home negatively predicted socio-emotional competence at 9 months. Mother's unemployment, living in neighbourhoods perceived as negative and being in contact with family and social services were negatively associated with concurrent socio-emotional competence at 2 years. Overall, more and/or stronger contemporaneous effects were found suggesting that negative effects of contextual factors may not have lasting impact on socio-emotional competence. In addition, the results showed that maternal behaviours need to be practised regularly to have positive impact on child's socio-emotional development.

Do Patterns of Levels of Socio-emotional Competence During Early Childhood Predict Executive Function at 4.5 Years?

Socio-emotional competence and executive function both work together to meet the demands of the everyday environment. While many studies have focused on how various domains of socio-emotional competence are predicted by, or associated with executive function, the predictive influence of socio-emotional competence on executive function has largely been ignored despite strong theoretical links. In addition, contradictory information exists with regard to the divergent validity of two subtypes of executive function: cool and hot. Using data from 4839 children participating in three data collection waves (9 months, 2 years and 4.5 year) in the Growing Up in New Zealand longitudinal study, we examined how different patterns of socio-emotional competence development during the early preschool years (persistent lows, recent low, improved and no lows) related to cool and hot executive function measured at aged 4.5 using a hand clap task and a gift wrap task, respectively. Findings showed that children with persistent lows (with no improvement) in the levels of early socio-emotional competence had increased odds of having below average cool and hot executive function at 4.5 years. However, no difference was found in the influence of socio-emotional competence on cool and hot executive function. Possible explanations for these associations between socio-emotional competence and executive function are discussed.

The Association Between Persistence and Change in Behavioral Difficulties During Early to Middle Childhood and Cognitive Abilities at Age 8.

We investigated the association between persistence and change in behavioral difficulties during early to middle childhood and several cognitive outcomes. We observed 3904 8-year-olds enrolled in the longitudinal study Growing Up in New Zealand (50% male/female; 23% Maori, 9% Pacific Peoples, 13% Asian, 2% Middle Eastern/Latin American/African, 9% Other, 43% European). The NIH Toolbox Cognition Battery was used to assess cognitive functioning at 8 years and the Strengths and Difficulties Questionnaire for behavioral difficulties at 4.5 and 8 years. Multivariate logistic regression analyses were conducted, controlling for well-known sociodemographic confounders. Children with persistent or later onset of behavioral difficulties were at higher risk for poorer vocabulary, reading, inhibitory control/attention, episodic memory, working memory and processing speed at age 8 compared to children with no or improved difficulties. Our study supports the importance of addressing both cognitive and behavioral aspects when planning educational programmes and interventions in early and middle childhood.

Exploring the Factor Structure of the NIH Toolbox Cognition Battery in a Large Sample of 8-Year-Old Children in Aotearoa New Zealand.

OBJECTIVE: The objective of this study was to derive a factor structure of the measures of the National Institutes of Health (NIH) Toolbox Cognition Battery (CB) that is representative of cognitive abilities in a large ethnically diverse cohort of 8-year-old children in Aotearoa New Zealand. METHODS: Our sample comprised of 4298 8-year-old children from the Growing Up in New Zealand study. We conducted exploratory and confirmatory factor analysis for the NIH Toolbox CB measures to discover the best-fitting factor structure in our sample. Measurement invariance of the identified model was tested across child's gender, socio-economic status (SES), and ethnicity. RESULTS: A three-dimensional factor structure was identified, with one factor of Crystallised Cognition (Reading and Vocabulary), and two distinguished factors of fluid cognition: Fluid Cognition I (Attention/Inhibitory Control, Processing Speed, and Cognitive Flexibility) and Fluid Cognition II (Working Memory, Episodic Memory). The results demonstrate excellent model fit, but reliability of the factors was low. Measurement invariance was confirmed for child's gender. We found configural, but neither metric nor scalar, invariance across SES and the four major ethnic groups: European, Maori, Pacific Peoples, and Asian. CONCLUSION: Our findings show that, at the age of 8 years, fluid abilities are more strongly associated with one another than with crystallised abilities and that fluid abilities need to be further differentiated. This dimensional structure allows for comparisons across child's gender, but evaluations across SES and ethnicity within the Aotearoa New Zealand context must be conducted with caution. We recommend using raw scores of the individual NIH Toolbox CB measures in future research.

Patterns of risk exposure in first 1,000 days of life and health, behavior, and education-related problems at age 4.5: evidence from Growing Up in New Zealand, a longitudinal cohort study.

BACKGROUND: Children who are high priority candidates for early intervention need to be identified to reduce their risk for experiencing problems in development. Those exposed to multiple risk factors are more likely to exhibit problems in development than those exposed to a single or no risk factor. We examined the longitudinal associations between persistence and timing of exposure to cumulative risk (CR) on three occasions by age 2 and problems in development at age 4.5 in health, behavior, and education-related domains. METHODS: Data are from Growing Up in New Zealand (NZ), a prospective longitudinal study of a birth cohort first assessed during their last trimester in 2009-10 and followed at ages 9 months and 2 and 4.5 years. All women with an expected delivery date in a 12-month period who resided within a defined region were invited to participate, with no additional eligibility criteria. Exposure was measured for 12 sociodemographic and maternal health risk factors at third trimester and ages 9 months and 2 years, from which developmental trajectories were constructed capturing persistence and timing of CR exposure. Ten developmental outcomes were measured at age 4.5 to classify problems in overall health status, obesity, and injuries; internalizing and externalizing behavior problems; and letter naming, counting forward and backward, and expectations for starting school and completing education. RESULTS: Analyses of data from 6156 children (49% female, 33% Non-European ethnicity) who participated in the 4.5-age assessment uniformly showed associations between exposure to more than consistently zero CR across early development and higher prevalence of being classified with problems for 9 of 10 outcomes. Persistent exposure to a CR ≥ 4 was generally associated with a higher prevalence of problems for 7 of 10 outcomes, whereas the timing of first exposure to CR ≥ 4 showed a less consistent association with problem outcomes. CONCLUSIONS: These findings are concerning because over 50% of NZ children are exposed to at least one of these risk factors at some point in early development. Routine screening of most of these risk factors during pregnancy is feasible and can identify priority candidates for intervention.

Early adolescent physical activity, sleep and symptoms of depression at 16 years of age.

AIM: To examine early adolescent physical activity and risk of later depressive symptoms at age 16 years in a prospective cohort study. METHODS: Participants were children and parents enrolled at birth of the child. Approximately half the children enrolled in the Auckland Birthweight Collaborative Study were small for gestational age at birth (SGA ≤10th percentile for sex and gestation) and half were appropriate for gestational age (AGA >10th percentile). Maternal demographic data were collected at birth, and children were followed through to age 16 years. Depression at 16 was assessed using the Center for Epidemiological Studies Depression Scale for Children. Accelerometer measures of physical activity and sleep were measured at 11 years of age. RESULTS: Moderate to severe depression was present in 15.6% of the 467 16-year-olds. Objectively measured physical activity and sleep at 11 years were not significantly associated with depressive symptoms at 16 years of age. CONCLUSION: Prospectively collected objective measures of physical activity levels and sleep were not predictive of depressive symptoms later in adolescence in a healthy community cohort. While interventions to promote increased physical activity and sleep in adolescents who are depressed may be effective, physical activity and sleep in the general population of adolescents does not protect against future depression.

The Association Between Persistence and Change in Early Childhood Behavioural Problems and Preschool Cognitive Outcomes.

The link between behavioural and cognitive difficulties is well established. However, research is limited on whether persistence and change in behavioural difficulties relates to cognitive outcomes, particularly during preschool. We used a large New Zealand birth cohort to investigate how persistence and change in serious behavioural problems from ages 2 to 4.5 years related to measures of cognitive delay at 4.5 years (n = 5885). Using the Strengths and Difficulties total problems score at each time point, children were categorised as showing no difficulties, improved behaviour, concurrent difficulties, and persistent difficulties. Cognitive measures assessed included receptive language, early literacy ability, and executive control. Our results showed that children with concurrent and persistent behavioural difficulties were at a greater risk of showing delays within specific cognitive domains relative to children with no difficulties and were also more likely to show comorbid delays across multiple cognitive domains.

Persistence and change in behavioural problems during early childhood.

BACKGROUND: Behavioural problems and psychopathology can present from as early as the preschool period. However there is evidence that behavioural difficulties may not be stable over this period. Therefore, the current study was interested in evaluating the persistence and change in clinically relevant behavioural problems during early childhood in a population-based New Zealand birth cohort. METHODS: Behaviour was assessed in 5896 children when they were aged 2 and 4.5 years using the Strengths and Difficulties Questionnaire (SDQ). Correlations and mean differences in subscale and total difficulties scores were examined. Scores were then dichotomised into normal/borderline and abnormal ranges to evaluate the persistence and change in significant behavioural problems. Chi-square analyses and ANOVAs were used to determine the association between sociodemographic and birth variables, and preschool behavioural stability. RESULTS: Raw scores at ages 2 and 4.5 years were moderately correlated, with most measures showing a small but significant decrease in mean scores over time. The majority of children who showed abnormal behaviour at 2 years improved at 4.5 years (57.9% for total difficulties). However, a notable proportion persisted in their difficulties from 2 to 4.5 years (42.1% for total difficulties). There was a small percentage of children who were categorised as abnormal only at 4.5 years. Children with difficulties at one or both time points had a greater proportion who were the result of an unplanned pregnancy, lived in highly deprived urban areas, and had mothers who were younger, of Maori and Pacific ethnicity and were less educated. CONCLUSIONS: Not all children who show early behavioural difficulties persist in these difficulties. Those whose difficulties persist were more likely to experience risk factors for vulnerability relative to children with no difficulties. Results suggest that repeated screening for early childhood behavioural difficulties is important.

Antenatal and Postnatal Determinants of Behavioural Difficulties in Early Childhood: Evidence from Growing Up in New Zealand.

Behavioural difficulties during early childhood have significant implications for multiple outcomes later in life. Child behavioural difficulties at 2 years of age (N = 6246) were assessed by mothers enrolled in a longitudinal, population-based New Zealand cohort study. 10.1% of children had total difficulties scores in the abnormal range on the preschool version of the Strengths and Difficulties Questionnaire. After controlling for maternal education, poverty, and child's birth age/weight, several antenatal and postnatal maternal health and family risk factors were significant for: (i) emotional problems (antenatal maternal perceived stress, lack of periconceptional folate, and moderate to severe maternal postnatal anxiety); (ii) hyperactivity-inattention (antenatal maternal perceived stress, mothers' antenatal exposure to secondhand smoke, moderate to severe maternal postnatal anxiety, and low maternal self-evaluation); (iii) conduct problems and total difficulties (antenatal maternal perceived stress, verbal inter-parental conflict and low maternal self-evaluation). The identification of risk and protective factors associated with early childhood difficulties are vital for guiding intervention and prevention efforts.

Antibiotics in the first year of life and subsequent neurocognitive outcomes.

AIM: There may be a link between disruption to the gut microbiota in early life and later neurocognitive outcomes. We hypothesised that antibiotic use in early life is associated with a detrimental effect on later neurocognitive outcomes. METHODS: Eight hundred and seventy-one European mothers and their children enrolled in the Auckland Birthweight Collaborative Study at birth. Information on antibiotic use during the first year of life and between 12 months and three-and-a-half years of age was gathered via maternal interview. Intelligence test scores and measures of behavioural difficulties were obtained when children were three-and-a-half years, seven years and 11 years of age. RESULTS: Antibiotic use in the first year of life was reported in 70% of the 526 children with antibiotic data assessed at age three-and-a-half years. Those who had received antibiotics had more behavioural difficulties and more symptoms of depression at follow-up. Results were consistent across all standardised psychologist administered tests, as well as parent rated, teacher rated and self-report measures. CONCLUSION: This study demonstrates an association between antibiotic use in the first year of life and subsequent neurocognitive outcomes in childhood. If confirmed by further research, these findings could have implications for the use of antibiotics for minor illnesses in infancy.

A Longitudinal Study of Pre-pregnancy and Pregnancy Risk Factors Associated with Antenatal and Postnatal Symptoms of Depression: Evidence from Growing Up in New Zealand.

Objectives Antenatal and postnatal depression can lead to poor outcomes for women and their children. The aim of this study was to explore whether risk factors differ for depression symptoms that are present during pregnancy and/or after childbirth. Methods An ethnically and socioeconomically diverse sample of 5301 women completed interviews during the third trimester of pregnancy and 9 months after childbirth. Depression symptoms were measured using the Edinburgh Postnatal Depression Scale (EPDS). Depression symptoms (defined as EPDS >12) among participants and associations with pre-pregnancy and pregnancy maternal characteristics were explored using logistic regression. Results The rate of antenatal depression symptoms (ADS) only was 8.5, 5% of women had depression symptoms at 9 months postpartum (PDS) only and 3% experienced depression symptoms at both time points. Perceived stress and Pacific or Asian ethnicity were risk factors for ADS and PDS. Anxiety during and before pregnancy was a risk factor for ADS only while having a pre-pregnancy diagnosis of depression was a risk factor for PDS only. Having ADS increased the odds ratio of PDS by 1.5 (95% CI 1.01-2.30). Conclusions The results supported evidence from previous longitudinal studies that depression symptoms appear to be higher during pregnancy than in the first year following childbirth. The study found that PDS may often be a continuation or recurrence of ADS.

Associations between schizotypy and cerebral laterality.

Atypical lateralization for language has been found in schizophrenia, suggesting that language and thought disorders on the schizophrenia spectrum may be due to left hemispheric dysfunction. However, research with those with non-clinical schizotypy has been inconsistent, with some studies finding reduced or reversed language laterality (particularly with positive schizotypal traits), and others finding typical left hemispheric specialization. The aim of the current study was to use both a behavioural (dual reading-finger tapping) task and an functional magnetic resonance imaging lexical decision task to investigate language laterality in a university sample of high- and low-schizotypal adults. Findings revealed no evidence for atypical lateralization in our sample for both overall schizotypy (measured by the Oxford-Liverpool Inventory of Feelings and Experiences) and positive schizotypy (measured by the Unusual Experiences subscale) groups. Our findings provide further evidence that non-clinical schizotypy is not associated with atypical language laterality.

Infant Behavior Questionnaire-Revised Very Short Form: A New Factor Structure's Associations With Parenting Perceptions and Child Language Outcomes.

The Infant Behavior Questionnaire Revised-Very Short Form (IBQ-R VSF; Putnam, Helbig, Gartstein, Rothbart, & Leerkes, 2014 ) is a new publicly available measure of infant temperament measuring positive affectivity/surgency (PAS), negative emotionality (NEG), and orienting and regulatory capacity (ORC). Although the initial psychometric properties of the 3-factor model appear promising, it has not been administered to a large and diverse sample and its predictive validity has not been established. This study administered the IBQ-R VSF to a diverse sample of 5,639 mothers of infants aged between 23 and 52 weeks. Confirmatory factor analysis found that the 3-factor solution did not meet the requirement for satisfactory model fit. Exploratory factor analysis found that a 5-factor solution (PAS, NEG, Orienting Capacity, Affiliation/Regulation, and Fear) was statistically and conceptually the most parsimonious. All 5 temperament dimensions were found to relate to both mother- and partner-reported infant closeness, parenting confidence, and parenting satisfaction, and four of the dimensions (PAS, Orienting Capacity, NEG, and Fear) related to the infants' communication development. Some parental differences were also found. Together these findings suggest that the 5-factor IBQ-R VSF is a promising measure of infant temperament and is related to parenting perceptions and child language development.

A Cross-Cultural Analysis of the Infant Behavior Questionnaire Very Short Form: An Item Response Theory Analysis of Infant Temperament in New Zealand.

The Infant Behavior Questionnaire-Revised Very Short Form (IBQ-R VSF; Putnam, Helbig, Gartstein, Rothbart, & Leerkes, 2014 ) is a newly published measure of infant temperament with a 3-factor structure. Recently Peterson et al. ( 2017 ) suggested that a 5-factor structure (Positive Affectivity/Surgency, Negative Emotionality, Orienting Capacity, Affiliation/Regulation, and Fear) was more parsimonious and showed promising reliability and predictive validity in a large, diverse sample. However, little is known about the 5-factor model's precision across the temperament dimensions range and whether it discriminates equally well across ethnicities. A total of 5,567 mothers responded to the IBQ-R VSF in relation to their infants (N = 5,639) between 23 and 52 weeks old. Using item response theory, we conducted a series of 2 parameter logistic item response models and found that 5 IBQ-R VSF temperament dimensions showed a good distribution of estimates across each latent trait range and these estimates centered close to the population mean. The IBQ-R VSF was also similarly precise across 4 ethnic groups (European, Maori, Pacific peoples, and Asians), suggesting that it can be used as comparable measure for infant temperament in a diversity of ethnic groups.

Dietary intakes in children born small for gestational age and appropriate for gestational age: A longitudinal study.

Children born small for gestational age (SGA) have an increased risk of cardiovascular disease (CVD) and associated risk factors in later life; however, little is known about their dietary intakes. The objective of this study was to assess dietary intakes in SGA and appropriate for gestational age (AGA) at 3.5, 7, and 11 years. The Auckland Birthweight Collaborative Study is a longitudinal case-control study of children born at term (n = 871). Children were assessed at 3.5 (n = 550), 7 (n = 591), and 11 (n = 620) years of age. Diet was assessed using a 24-hr record-assisted recall. Reported dietary intakes were analyzed and compared with the Australian and New Zealand Nutrient Reference Values. Compared with AGA, median energy intakes were significantly lower in SGA at 3.5 years (4.2 MJ [IQR, 3.0 to 5.8] vs. 5.4 MJ [IQR, 3.9 to 6.5]; p < .0001) but not at 7 and 11 years. Inadequate dietary intakes of micronutrients were more prevalent among SGA at 3.5 years and 11 years of age. A large proportion of SGA and AGA children consumed more than the recommended amounts of saturated fats, sugars, and sodium. There was no association of dietary intake and socio-demographic factors. This study reveals that dietary intake in 3.5-year-old children born SGA is lower in energy and a variety of micronutrients compared with dietary intake in AGA. These intakes may however be appropriate given their BMI z-scores. High intakes of sodium, saturated fat, and sugars are a concern for all children in this cohort.

Associations Between the KIAA0319 Dyslexia Susceptibility Gene Variants, Antenatal Maternal Stress, and Reading Ability in a Longitudinal Birth Cohort.

Maternal stress during pregnancy has been associated with detrimental cognitive developmental outcomes in offspring. This study investigated whether antenatal maternal perceived stress and variants of the rs12193738 and rs2179515 polymorphisms on the KIAA0319 gene interact to affect reading ability and full-scale IQ (FSIQ) in members of the longitudinal Auckland Birthweight Collaborative study. Antenatal maternal stress was measured at birth, and reading ability was assessed at ages 7 and 16. Reading data were available for 500 participants at age 7 and 479 participants at age 16. FSIQ was measured at ages 7 and 11. At age 11, DNA samples were collected. Analyses of covariance revealed that individuals with the TT genotype of the rs12193738 polymorphism exposed to high maternal stress during pregnancy possessed significantly poorer reading ability (as measured by Woodcock-Johnson Word Identification standard scores) during adolescence compared with TT carriers exposed to low maternal stress. TT carriers of the rs12193738 SNP also obtained lower IQ scores at age 7 than C allele carriers. These findings suggest that the KIAA0319 gene is associated with both reading ability and general cognition, but in different ways. The effect on IQ appears to occur earlier in development and is transient, whereas the effect of reading ability occurs later and is moderated by antenatal maternal stress. Copyright © 2016 John Wiley & Sons, Ltd.

Risk factors for migraine and tension-type headache in 11 year old children.

BACKGROUND: Though migraine and tension type headache are both commonly diagnosed in childhood, little is known about their determinants when diagnosed prior to puberty onset. Our aim was to determine psychosocial- and health-related risk factors of migraine and tension-type headache in 11 year old children. METHODS: 871 New Zealand European children were enrolled in a longitudinal study at birth and data were collected at birth, 1, 3.5, 7, and 11 years of age. Primary headache was determined at age 11 years based on the International Headache Society. Perinatal factors assessed were small for gestational age status, sex, maternal smoking during pregnancy, maternal perceived stress, and maternal school leaving age. Childhood factors assessed were sleep duration, percent body fat, television watching, parent and self-reported total problem behaviour, being bullied, and depression. RESULTS: Prevalence of migraine and tension-type headache was 10.5% and 18.6%, respectively. Both migraine and TTH were significantly associated with self-reported problem behaviour in univariable logistic regression analyses. Additionally, migraine was associated with reduced sleep duration, and both sleep and behaviour problems remained significant after multivariable analyses. TTH was also significantly associated with antenatal maternal smoking, higher body fat, and being bullied. For TTH, problem behaviour measured at ages 3.5 and 11 years both remained significant after multivariable analysis. Being born small for gestational age was not associated with either headache group. CONCLUSIONS: Although they share some commonality, migraine and tension-type headache are separate entities in childhood with different developmental characteristics. The association between primary headache and problem behaviour requires further investigation.

Prenatal determinants of anxiety symptoms in middle childhood. Evidence from Growing Up in New Zealand.

BACKGROUND: Early life environments can have long-lasting impacts on future health and wellbeing. Maternal health during pregnancy, including experiencing stress or mood disorders, has been associated with psychopathology in later life. Anxiety disorders are one of the most prevalent mental health conditions, affecting approximately 7 % of children and adolescents globally, with a lifetime prevalence of 15-20 %. Identifying prenatal risk factors can support future and current public health interventions and maternity care. METHODS: Data were obtained from the Growing Up in New Zealand longitudinal study of child development. Prenatally, mothers provided sociodemographic information as well as data on their mental health, potential teratogens, and lifestyle factors such as supplement intake and exercise levels. At 8-years old, 4922 children self-completed the PROMIS-SF anxiety measure. Bivariate analyses and backward stepwise regression were used to determine the best multivariable model. RESULTS: Significant prenatal predictors of anxiety symptoms at 8-years old included elevated maternal depression symptoms, body mass index in the overweight/obese range, exercise patterns, and paracetamol, anti-inflammatory and alcohol intake. LIMITATIONS: Sample attrition from baseline to 8-year may have affected statistical power. To further untangle the effect of timing and duration of the exposures reported in this study, larger sample sizes would be required. CONCLUSIONS: Prenatal mental health and wellbeing was significantly associated with child anxiety symptoms at 8-years of age. This study highlights the importance of supporting expectant mothers' health and wellbeing during pregnancy to ensure children have the best opportunity to have good mental health.

Reduced genetic influence on childhood obesity in small for gestational age children.

BACKGROUND: Children born small-for-gestational-age (SGA) are at increased risk of developing obesity and metabolic diseases later in life, a risk which is magnified if followed by accelerated postnatal growth. We investigated whether common gene variants associated with adult obesity were associated with increased postnatal growth, as measured by BMI z-score, in children born SGA and appropriate for gestational age (AGA) in the Auckland Birthweight Collaborative. METHODS: A total of 37 candidate SNPs were genotyped on 547 European children (228 SGA and 319 AGA). Repeated measures of BMI (z-score) were used for assessing obesity status, and results were corrected for multiple testing using the false discovery rate. RESULTS: SGA children had a lower BMI z-score than non-SGA children at assessment age 3.5, 7 and 11 years. We confirmed 27 variants within 14 obesity risk genes to be individually associated with increasing early childhood BMI, predominantly in those born AGA. CONCLUSIONS: Genetic risk variants are less important in influencing early childhood BMI in those born SGA than in those born AGA, suggesting that non-genetic or environmental factors may be more important in influencing childhood BMI in those born SGA.

Age at first exposure to antibiotics and neurodevelopmental outcomes in childhood.

RATIONALE: Viral illnesses in children are common and are frequently treated with antibiotic medication. Antibiotics reduce the diversity and composition of the gut microbiota, leading to poor developmental outcomes. OBJECTIVES: To investigate the relationship between age at first exposure to antibiotics and cognitive and behavioural development at 4.5 years while controlling for multiple confounders, including otitis media. METHODS: Study participants were 5589 children enrolled in the broadly generalisable Growing Up in New Zealand cohort study, with antibiotic exposure data, maternal antenatal information, and age 4.5-year behaviour and cognitive outcome data. Children were categorised as first exposed to antibiotics according to the following mutually exclusive ages: 0-2 months; 3-5 months; 6-8 months; 9-11 months; 12-54 months or not exposed by 54 months. Developmental outcome measures included the Strengths and Difficulties Questionnaire, Luria hand clap task, and the Peabody Picture Vocabulary Test-III. RESULTS: In univariate analysis, there was an evident dose-response relationship where earlier exposure to antibiotics in the first year of life was associated with behavioural difficulties, lower executive function scores, and lower receptive language ability. After adjusting for confounders, pairwise comparisons showed that first antibiotic exposure between birth and 3 months or between 6 and 9 months was associated with lower receptive vocabulary. Antibiotic exposure at any age prior to 12 months was associated with increases in behavioural difficulties scores at 4.5 years. CONCLUSIONS: Following adjustment for socioeconomic factors and otitis media, there is evidence that antibiotic exposure during potentially sensitive windows of development is associated with receptive language and behaviour later in childhood.