Language, motor skills and behavior problems in preschool years

Abstract

Child language development is a complex process. This process cannot be understood without considering its relationship to other developmental domains. Language development in preschool years is associated with development of motor skills and behavior problems, and these associations are the focus of the current thesis. Despite a large number of studies examining the co-occurrence of such developmental delays and problems, few studies have examined the developmental relationship between these areas during preschool years in a population-based sample. The first aim (paper 1) is therefore to look at how variation in typical development of language skills and motor skills is related. We especially want to explore whether the developmental paths for language and motor skills are characterized by stability or change in early childhood (1.5 to 3 of age). The second aim (paper 2) is to follow up results from paper 1 later in preschool years (3 to 5 years of age). Further, we want to look at how much of the variation in language skills can be explained by motor skills and vice versa. The third aim (paper 3) is to investigate the causal direction of the co-occurrence of language delay and externalizing behavior problems. The relationship between difficulties in these two domains is well established, but few studies have tried to estimate the causal relationship between them. Our hypothesis is that there would be differences in causal directions for the relationship between language delay and two separate subdomains of externalizing problems, aggression and inattention, respectively. For the purpose of the three papers included in this thesis, questionnaire data from three waves of the population based, longitudinal Mother and Child Cohort Study (MoBa) are utilized. Mothers’ reports were collected when children were 1.5, 3 and 5 years of age. Paper 1 includes data on 62,944 children from the first two waves of data collection. Paper 2 includes data from the two last waves, and paper 3 includes data from all three waves. In paper 2 and 3, mother reports on 25,474 children are included in the analyses. In paper 1 and paper 2 we used cross lagged panel models for investigating the autoregressive and cross-lagged associations between language and motor skills. Results from paper 1 show that both communication and motor skills were quite stable over time (communication skills: .40, motor skills: .80), with motor skills being significantly more stable than communication skills. However, whereas communication skills do not positively predict motor skills, motor skills are an equally strong predictor of future communication (.38) as motor skills. We conclude that the communication skills at this age are not a reliable predictor for later motor development, whereas motor skills are. Communication and motor skills are correlated at this early age, but we argue that variation in what is considered normal language development at 1½ years is too wide to predict variation in motor skills at later stages. In paper 2 we go on to study the relations between language and a subdivision of gross and fine motor skills between the ages of 3 and 5 years, in order to understand whether one aspects of motor skills would be more predictive of language than the other, and whether language would be predictive of motors skills at this later age. The estimated models of the relationship between language and the two domains of motor skills correspond to the one presented in paper 1. Both domains are characterized by modest to high stability rather than change (language skills: .80, gross motor skills: .56, fine motor skills: .43). However, in contrast to results from paper 1, language skills at 3 years of age have significant influence on change in both gross and fine motor skills over time, whereas motor skills no longer significantly predict later language skills. We go on to calculate how much of the shared variance is explained specifically by language and gross and fine motor skills, respectively. Results from these analyses suggest that variance explained by language alone decreases, whereas variance explained by motor skills alone increases from 3 to 5 years of age. We conclude that these domains of development are best described as specific at this age. Seen together, results from paper 1 and paper 2 indicate stability in both domains, but also some variability across domains. Motor skills are highly stable from 1.5 to 3 years of age, and motor skills at 1.5 years predict later language skills. From 3 to 5 years of age language skills show higher stability than motor skills, and language skills at 3 years predict later both gross and fine motor skills. In paper 3, we change focus from variation in typical development to differences between delayed and typical development. Children with language delay are thought to be at risk for a spectrum of co-occurring difficulties, and in this paper, we investigate the causal relationship between language delay and inattention and aggression, respectively. We include data from all three waves in fixed effects models. The results show that the causal relationship between language delay and inattention is quite different from the relationship between language delay and aggression. Whereas the first is explained by common factors and a reciprocal relationship between the two, the best fitting model for the relationship between language and aggression is one where language delay predicts aggression, and not the other way around. We conclude that our results support different etiologies for the relationship between language delay and inattention and aggression, respectively. Findings from the three papers highlight the importance of knowledge about developmental change in preschool years. These findings underline the value of utilizing data from more than one measurement occasion in order to capture how language skills are related to co-occurring skills in young children. Also, estimating different outcomes simultaneously, in the same study population enable the possibility to compare parameters directly. The results also have implications for prevention and intervention. Co-occurrence of symptoms is common in preschool years and changes happen rapidly. What is considered normal at one point in time quickly changes to being abnormal at another time point. When assessing young children with language delays, it is important to be aware of the difficulties this child could have in other areas. Knowledge about how symptoms of different developmental delays influence each other over time is essential to adapt treatment strategies to each individual child. It is therefore important that clinicians follow development in more than one area closely, as both co-occurrence of symptoms, and a change in presentation of symptoms are common.