Social Policy Research Paper No. 36

Contents

• Executive summary
• 1. Introduction
• 2. Outcomes for children in differing circumstances
• 3. Children's use of non-parental care
• 4. Children's health
• 5. Mothers' health
• 6. Family learning environments and child outcomes
• 7. Discussion
• Appendix A: Summary of variables included in the Outcome Index
• Appendix B: LSAC Research Consortium
• Endnotes
• References
• List of tables and figures

4. Children's health

• 4.1 Introduction
• 4.2 Findings
• 4.3 Discussion

Section summary

• Although Australian infants and children are generally in very good or excellent health, many experience substantial physical health problems—for example, low birth weight (5 per cent), preterm birth (7 per cent), special health care needs (14 per cent), overweight/obesity (21 per cent), and asthma requiring medication (15 per cent).
• Although most mothers breastfeed, relatively few meet current NHMRC recommendations for exclusive breastfeeding to at least 6 months. Most children are not meeting nutritional guidelines, and many are reported to prefer sedentary to physically active pursuits.
• Child health variables strongly influence Outcome Index scores, with indicators of current health (most notably asthma and healthful nutritional behaviours) as powerful as the perinatal influences of preterm birth. Impacts of child health are often greater in the social–emotional and/or learning domain than in the physical domain, emphasising the importance of children's health to all aspects of their functioning and wellbeing.
• Impacts are less evident in the first year of life and more clearly apparent by 4 to 5 years of age. This suggests that intervention in the early years (that is, between infancy and the preschool year) may have potential to prevent the impacts of poorer health developing.

4.1 Introduction

This section addresses the following questions:

• What perinatal, postnatal, and current health issues face Australian infants and 4 to 5 year-old children, and how prevalent are these issues?
• How well are Australian 4 to 5 year-old children doing with regard to healthful levels of physical activity and nutritional habits?
• How do these issues affect children's overall health and wellbeing (measured by the Outcome Index), physical health (measured by the physical domain subscale of the Outcome Index), and other areas of functioning?
• When considered relative to each other, which of these factors make the biggest contribution to children's overall and physical health?

The health data collected in Wave 1 were mainly cross-sectional, but retrospective data regarding pregnancy and the natal and postnatal periods allow us to start to appreciate the longitudinal nature of the development of good and poor child health.

4.2 Findings

Child health variables

The child health variables (Table 16) could broadly be classified as prenatal (considered in Section 5, Maternal health), perinatal, postnatal, or current. With the exception of height, weight and head circumference (directly measured by the LSAC interviewer), all health measures were reported by the primary care giver. Television and computer use are not considered here, as they are analysed in Section 6 (Family learning environments).

The primary child outcomes relevant to this section are the overall Outcome Index and physical domain scores. Some current child health variables—such as body mass index (BMI)—contributed directly to the calculation of the physical domain (and therefore overall Outcome Index) score, but we were nonetheless interested in how these variables were related to the two other domains of child functioning. Figures 9 and 10 show the proportions of the child cohort according to each health variable scoring in the top 15 per cent (that is, above the positive cut-off ) and bottom 15 per cent (below the negative cut-off ) of overall Outcome Index scores. These analyses are bivariate (that is, not adjusted for sociodemographic or other variables). Comparable graphs for the infant cohort are not shown; relationships between health variables and high or low Outcome Index scores were generally either absent or weaker than (but in the same direction as) the child cohort.

Perinatal health variables

• Birth weight: 6 per cent of infants and 7 per cent of children were of low birth weight (‹2500g), similar to the 6.3 per cent reported for all Australian babies live-born in 2003 (AIHW 2006). Of those with low birth weight, 29 per cent of children were below the negative Outcome Index cut-off for their respective cohorts. However, the expected proportion of normal birth weight children fell above the positive cut-off, in the top 15 per cent of the Outcome Index score. (Note that no adjustment was made for gestational age in these analyses, but that future re-releases of the LSAC data may include age corrections).
• Gestational age: the great majority of children in both cohorts were born at term (37 to 41 weeks). Slightly more children (7 per cent) than infants (5 per cent) were born post-term (≥42 weeks), possibly indicating increasing obstetric reluctance to allow pregnancies to proceed past term. In both cohorts, 5 to 6 per cent were born preterm (33 to 36 weeks), and less than 2 per cent very preterm (32 weeks or less), slightly less than the 7.9 per cent of all Australian live births born ‹37 weeks in 2003 (AIHW 2006). Of the very preterm children, 44 per cent fell below the negative Outcome Index cut-off.

Postnatal health variables

• Breastfeeding: the prevalence and duration of breastfeeding were comparable across the cohorts, with 13 per cent never establishing breastfeeding, and about one-quarter feeding for between one week and three months. The majority of women who breastfed for three months or more continued beyond six months; less than 10 per cent of all LSAC mothers discontinued breastfeeding between three and six months, while just over half reported feeding for more than six months (these analyses excluded the 605 infants who were aged ‹6 months at the time of interview). The median age of ceasing breastfeeding completely for the child cohort was 6 months, with 25 per cent ceasing by 8 weeks and 75 per cent by 12 months of age. 'No breastfeeding' had a much lower proportion (9 per cent) of children above the positive Outcome Index cut-off and a much higher proportion (27 per cent) below the negative cut-off. These relationships were not evident for the infant cohort.
• For the child cohort, milk or foods other than breast milk were first introduced at a median age of 4 months (25 per cent by 5 weeks, 75 per cent by 5 months). Thus, only half of all mothers were meeting the national recommendations which were in place when the child cohort was born (that is, to introduce solids after the age of 4 months). This highlights the magnitude of shift in maternal practice that would be required to meet the current NHMRC recommendations that babies should be exclusively breastfed to age 6 months (NHMRC 2003b).

Current health variables

• Health-related quality of life (child cohort only): this was measured using the PedsQL (Pediatric Quality of Life Inventory 4.0—PedsQL 4.0) (Varni et al. 2003), a 23-item validated questionnaire for children aged 2 to 18 years that generates scores between 0 and 100 for total, physical and psychosocial health. The median total (82.1), physical (84.4) and psychosocial (80.8) scores on the PedsQL for the LSAC child cohort were very similar to the published means for the normative US sample (81.2, 83.3 and 80.2, respectively).
• General health: this was rated using a widely used global health single item with response options of excellent, very good, good, fair and poor (Waters, Salmon & Wake 2000). Most parents perceived their children as being healthy, with 87 per cent in both cohorts rating their child's health as excellent/very good and only 13 per cent as good/fair/poor. This contrasts with the much lower ratings mothers applied to their own health using the same rating scale (see Section 5). The mean social–emotional domain score of children rated as having good, fair or poor health was 95.1, compared to 100.7 in those with excellent or very good health—that is, roughly half a standard deviation lower—with weaker but similar relationships seen for learning domain scores and for infants.
• Of the infants, 6 per cent were reported to have special health care needs, rising to 14 per cent for the children, commensurate with the US prevalence using the longer version of the same measure. The Children with Special Health Care Needs (CSHCN) Screener (Bethell et al. 2002) is widely used internationally, and probes for conditions expected to last for at least 12 months that need or use more care than a parent would consider usual for all children. Having special health care needs had impacts of similar magnitude on Outcome Index scores to the impact of poorer general health.
• The infants' weight-for-age standardised (z) score increased by 0.55 from birth, indicating that Australian infants gain weight more rapidly than would be expected from the normative Centers for Disease Control data (Ogden et al. 2002). This rate of growth appears plausible, as it is in line with current mean standardised (z) scores for Australian primary school children (Booth et al. 2001). This is also reflected in the rates of 'catch-up' versus 'catch-down' change in infant weight from birth, with nearly half increasing their weight by more than two-thirds of a standard deviation, that is, moving up at least one centile band on the growth charts from birth, and only 15 per cent dropping by at least one centile band from birth. Changing weight-for-age did not appear to impact on the proportions of infants above the positive or below the negative Outcome Index cut-offs.
• Of the child cohort, 15.2 per cent were classified as overweight and 5.5 per cent as obese, using the International Obesity Task Force definitions (Cole et al. 2000). This represents a very substantial increase since the 1995 National Nutrition Survey using the same categorisation (Magarey, Daniels & Boulton 2001), and is in line with recently reported prevalence for primary school children (Magarey, Daniels & Boulton 2001). Child overweight/obesity did not appear to impact on child social–emotional and learning domain scores.
• Parents of the child cohort were asked to report on their child's nutrition the previous day. While 95.5 per cent met the national recommendations of at least one serve of fruit, only 43 per cent met recommendations for vegetable intake and only 26 per cent had less than two serves of high fat/high sugar foods (NHMRC 2003b). Of the child cohort, 80 per cent had at least one soft or sugar sweetened drink (including fruit juice) the previous day, while 86 per cent had two or more drinks of water the previous day; 41 per cent had displayed two or less of the five healthful behaviours the previous day, and just 24 per cent displayed four or five of these healthful behaviours. Of those with none or one healthful behaviours, only 9 per cent fell above the positive cut-off but over 26 per cent fell below the negative cut-off of the Outcome Index; of those with four or five, 18 per cent were above the positive and only 11 per cent below the negative cut-offs.
• Nearly 80 per cent of children were reported to 'very much' enjoy physical activity, but only 29 per cent preferentially choose active over inactive pastimes. Both groups were slightly overrepresented in the lowest 15 per cent of Outcome Index scores (20 per cent of the children who were reported not to enjoy physical activity and 21 per cent of those who preferentially choose inactive pastimes).
• Asthma and wheeze: 30 per cent of infants and children were reported to have 'ever wheezed', and 15 per cent of children were reported to have doctor diagnosed asthma for which they were taking medication at the time of the LSAC survey. These figures are in keeping with other epidemiological sources for Australian children, such as the Australian arm of the International Study of Asthma and Allergies in Childhood study (Robertson et al. 1998) and the National Health Survey 2004–05 (ABS 2006). Over a quarter of the children with currently medicated asthma had overall Outcome Index scores below the negative cut-off.

Associations with specific health outcomes comprising the physical domain

To help understand where specific health issues might exert their impacts on the child, relationships between some of the key health variables and the health measures contributing to the physical domain are tabulated in Table 17. Table 17 presents results summarising the relationship between perinatal, postnatal, and current variables and the health measures contributing to the physical domain of the Outcome Index.

The physical domain score for the child cohort comprises four separate measures—health-related quality of life (PedsQL), CSHCN Screener, global health, and BMI standard (z) score, with both very low and very high BMI z-scores contributing to a lower Outcome Index score. Low birth weight and prematurity showed a markedly higher prevalence of special health care needs and poorer global health ratings, a modest reduction in mean PedsQL scores, and lower rates of overweight/obesity. On average, children who breastfed longer had higher PedsQL scores, lower special health care needs, better general health and lower rates of overweight/obesity. Mean PedsQL scores and general health both rose with increasing numbers of healthful nutritional behaviour and child preferences for active pastimes.

Impact of child health variables on outcomes: multivariable analyses

Independent associations of each variable with the overall Outcome Index score for both the infant and child cohorts were examined using multivariable regression analyses controlling for all other perinatal, postnatal and current health variables as well as for the set of socioeconomic variables referred to in Section 2.

Tables 18 (infant cohort) and 19 (child cohort) show multivariable linear regression relationships between the child health variables and the continuous Outcome Index scores for the infant and child cohorts, all adjusted for the nine key sociodemographic variables identified in Section 2. For the health variables that directly contribute to calculation of the physical domain and overall Outcome Index (BMI, general health, special health care needs, PedsQL) only relationships with social–emotional and learning domain scores are shown. For all other health variables, relationships with the overall Outcome Index score and all three domain scores are shown.

Table 18 shows that the final multivariable models for the infants accounted for only small proportions of the total variance in scores—4 per cent of the overall Outcome Index, most strongly predicting the physical domain scores (6.5 per cent); 2.2 per cent of the variance was explained by the child health variables as a group, over and above the contribution of the sociodemographic variables.

The multivariable models for the children (Table 19) accounted for considerably larger proportions of the total variance—20 per cent for the overall Outcome Index, 10 per cent for the physical domain, 23 per cent for the social–emotional domain, and 16 per cent for the learning domain variances respectively. Again, much of this was related to the contribution of the sociodemographic circumstances, with the child health variables as a group accounting for 5.5 per cent of the total variance in this model over and above the contribution of the sociodemographic variables.

Perinatal health variables

• Birth weight and gestational age: there was strong evidence of an association between preterm birth and overall infant Outcome Index score (which fell by an average of 6 points if born ≤32 weeks gestation compared to 37 to 41 weeks, p‹0.001), but no evidence of a relationship with low birth weight (p=0.49). Although low birth weight did have a borderline impact on Outcome Index for children (p=0.04), it continued to be outweighed by gestational age (p=0.01) (which fell by an average of 4 points if born ≤32 weeks gestation compared to 37 to 41 weeks). In both cohorts, this relationship appeared to be driven by the relationship between prematurity and poorer learning domain scores; impacts on physical and social–emotional domain scores were not evident. The data suggest that child outcomes tended to improve with increasing gestational age.

Postnatal health variables

• Breastfeeding: there was strong evidence that duration of breastfeeding was associated with improved overall Outcome Index and social–emotional domain scores for the child cohort (p‹0.001). However, this effect was modest, with scores rising on average only 2 points for those breastfeeding ≥6 months compared to those who did not establish breastfeeding. There was no evidence that breastfeeding was associated with physical domain scores (p=0.11). For the infants, there was strong evidence that breastfeeding was associated with higher physical domain scores (p‹0.001), but not with a higher overall Outcome Index score (p=0.57).

Current health variables

• Regarding health-related quality of life, there was strong evidence that physically healthy children have better mental health (p‹0.001). Mean child social–emotional domain scores increased by 0.2 points with every additional PedsQL Physical Health point. General health was also strongly associated with social–emotional domain scores for both cohorts (p‹0.001), with scores on average 2 points lower for children and 3 points lower for infants whose general health was not reported to be very good or excellent. There was no evidence that having special health care needs was associated with infant outcomes (all p›0.14), but strong evidence of an association with lower social–emotional and learning domain scores for the child cohort (p‹0.001). Scores for both domains were 3 points lower on average for children with special health care needs.
• There was some or borderline evidence of an increase in infant weight-for-age from birth with slightly better overall Outcome Index (p=0.002), physical (p=0.01) and social–emotional domain scores (p=0.07). Child overweight/obesity was not related to social–emotional or learning domain scores.
• Number of healthful nutritional behaviours was the only health characteristic that indicated strong evidence of an association with higher scores on every Outcome Index domain (p‹0.005 for all three domains). The overall Outcome Index score rose by an average of 1 point for every additional healthful behaviour, so that children who demonstrated all five healthful behaviours had scores on average 6 points higher than those demonstrating none. There was also strong evidence that enjoyment of physical activity was related to better total child Outcome Index, physical domain and social–emotional domain scores (all p‹0.001), although the effects were modest.
• The highly prevalent problem of infant wheeze showed clear evidence of an association with lower overall Outcome Index (a mean decrease of 2 points, p‹0.001) and physical domain (a mean decrease of 4 points, p‹0.001) scores, as well as showing borderline evidence of an association with social–emotional domain score (a mean increase of 1 point, p=0.03). The impact of child asthma was more marked—on average a 4 point drop in overall Outcome Index score (p‹0.001), mainly due to an average 6 point drop in the physical domain score (p‹0.001).

Figure 9: Low Outcome Index: percentage of child cohort by health characteristics

Figure 10: High Outcome Index: percentage of child cohort by health characteristics

4.3 Discussion

Although the Australian infants and children in the LSAC cohorts are generally considered by their parents to be in very good or excellent health, many children experience substantial health problems, including low birth weight (5 per cent), preterm birth (7 per cent), special health care needs (14 per cent), overweight/obesity (21 per cent), and asthma requiring medication (15 per cent). The prevalence of chronic childhood conditions such as asthma and obesity has been reported to have increased in recent decades, both in Australia and overseas (Asher et al. 2006; Booth et al. 2003; Perrin, Bloom & Gortmaker 2007; Ross Anderson et al. 2007; Shamssain 2007). The high prevalence of these problems is in line with other comparable national and international studies, supporting confidence in the generalisability of the subsequent multivariable associations.

The findings reported above suggest that many of these problems are already influencing the broad outcomes of Australian children as young as preschool age. There was strong evidence for many of the child health variables being associated with Outcome Index scores, with indicators of current health (most notably asthma and healthful nutritional behaviours) having as powerful an influence as perinatal influences (preterm birth). Given that the burden of poorer outcomes is likely to worsen over time for children with these problems, and that most of them are known to be at least partly preventable, these data lay down a very robust challenge to deliver health care and health promotion that can more effectively improve outcomes for Australian children.

Associations with health problems were often greater in the social–emotional and/or learning domain than in the physical domain, emphasising the importance of children's health to all aspects of children's functioning and wellbeing. This was further borne out by the strong relationships between the measures directly contributing to the physical domain (children's general health and the physical score of the PedsQL) and social–emotional domain scores for the child cohort.

The high prevalence of overweight and obesity (21 per cent) in the preschoolers is of particular concern, because of the rapidity with which its prevalence is rising. In the last national survey, the 1995 National Nutrition Survey involving 532 children aged 4 to 6 years (Magarey, Daniels & Boulton 2001), 13 per cent of boys and 19 per cent of girls were overweight or obese using the same cut-off points utilised in LSAC. Because BMI is incorporated into the Outcome Index, it was not possible in this report to study in depth its predictors or correlates with physical domain and overall Outcome Index score. However, at this young age, overweight and obesity were not strongly associated with poorer scores on the other two domains (social–emotional and learning). This is in keeping with limited international data regarding this age group, but contrasts with mounting evidence of poorer health and wellbeing in older children and adolescents and an unprecedented health burden in adults with obesity. LSAC should offer rich opportunities to study longitudinal relationships between physical and psychosocial health and BMI trajectories over time. Rapid early growth may be important in the early genesis of overweight and obesity (Stettler 2007), and will be an important variable to study in the infant cohort in future waves. The findings reported here (high prevalence, but as yet little morbidity) suggest that the preschool and school entry periods may be an optimal time to address excess adiposity, before health impacts develop.

However, it was possible to examine the impacts of healthful and less healthful nutritional and physical activity behaviours. Many preschoolers are not meeting nutritional guidelines for fruit and (particularly) vegetables, 75 to 80 per cent were reported to have consumed sweet drinks and foods likely to be high fat and/or high sugar the previous day, and close to 20 per cent were reported not to enjoy physical activity. All these factors were already implicated in children's health and overall outcomes, even after adjusting for sociodemographic circumstances. This was most striking for children who demonstrated no healthful nutritional behaviours the previous day, whose overall Outcome Index scores were approximately half a standard deviation lower than those with all five healthful behaviours.

With the exception of child height, weight and girth, direct child health measurements were not taken in the first wave of LSAC. Therefore, it is not possible to assess relationships between the multiple sociodemographic, non-parental care, health (both child and maternal) exposures and family learning environments on 'harder' health outcomes such as respiratory function and antecedents of diabetes and cardiovascular disease.

As seen in Sections 2, 3 and 5, associations with outcomes in the infants were less apparent than in the children, reinforcing the notion that the early years offer great potential for population interventions designed to prevent the emergence of disparities due to child health problems by the preschool year.

Return to top

Next: 5. Mothers' health

Previous: 3. Children's use of non-parental care

Copyright & Disclaimer | Privacy policy | Contact us | FaHCSIA home

© Commonwealth of Australia, 2008
Last modified 24 April, 2009