Growing up in Scotland: parenting and children's health

CHAPTER 5 DO DIFFERENCES IN PARENTING CONTRIBUTE TO INEQUALITIES IN CHILD HEALTH AND HEALTH BEHAVIOURS?

5.1 Introduction

Analysis in section 2.5 illustrated how child health outcomes varied by levels of family adversity. Then, in section 3.7, the analysis demonstrated how parenting skill also varied by family adversity. Given these relationships, this chapter explores whether differences in parenting may help to explain some of the association found between family adversity and children's health.

This topic was investigated using two stages of statistical modelling. Stage 1 examined the association between family adversity and each of the health outcomes or health behaviours after controlling for a set of basic demographic factors (the child's gender, the child's age in months at sweep 5, whether the child was first born or had older siblings and the number of children in the household at sweep 5).

At stage 2, all the individual parenting measures were then added to the models. If parenting is an important explanation for inequalities in health according to level of family adversity, we would expect the strength of the association between family adversity and health (observed in stage 1) to be reduced after controlling for parenting in this way.

5.2 Key findings

• Differences in parenting accounted for some, but not all, inequalities in child health and health behaviours that are linked to family adversity.
• Parenting differences were a stronger explanation for some health inequalities than others. High parenting skill reduced the association between adversity and health by between 33% and 44% for poor general health, limiting long-term illness, social, emotional and behavioural difficulties, and poor dental health.
• Parenting skill had a lesser effect on health problems (22%) and accidents and injuries (8%).
• Parenting skill accounted for between 32% and 54% of the association between adversity and screen time, fruit and vegetable consumption and snacking on crisps, sweets and sugary drinks.

5.3 Does parenting account for inequalities in child health?

The graphs in Figure 5-A display odds ratios with 95% confidence intervals before (stage 1) and after (stage 2) parenting variables are added to the model. When the dashed line falls below the solid line this indicates a reduction in the strength of association between family adversity and child health when parenting variables are added to the model suggesting that differences in parenting across families with different levels of adversity explain some of the inequalities in that health outcome.

As the graphs show, for the most part when parenting measures were added to the models, the strength of the relationships between family adversity and poor health were reduced. For many health outcomes, the effect of the adjustment for parenting appeared greatest with higher levels of family adversity. That is, parenting skills accounted for more inequalities in health for those families experiencing the highest levels of adversity (shown by the wider gaps between the two lines towards the right-hand side of the graphs).

In order to estimate how much of the relationship between family adversity and health behaviour inequalities is explained by differences in parenting, children who had no adversity were compared with children experiencing any level of adversity (more information is provided in section 2.6 of the Technical Appendix) ¹⁰ .

• The effect of parenting was strongest in the model of limiting long-term illness, reducing the odds associated with family adversity by 44%.
• For poor general health, social, emotional and behavioural difficulties and dental health, parenting accounted for 33%, 40% and 38% of the association between adversity and health.
• Smaller reductions were achieved in the models of health problems (22%) and accidents and injuries (8%).

For health problems and accidents and injuries, the effect of parenting was very small. These findings should not be taken to imply that parents cannot do much to reduce the incidence of health problems or accidents and injuries in their children. Both outcomes may relate more to other aspects of the family environment, or to parental behaviours that have not been studied here.

The results imply that parenting may help to explain some of the inequalities in child health linked to family adversity. However, in most cases, family adversity remained significantly associated with health inequalities even after taking account of differences in parenting. This suggests that parenting is only a partial explanation for inequalities in child health. Parenting may be more important for some health outcomes than others, and overall is likely to constitute only a partial explanation for inequalities in child health.

Figure 5-A Associations between family adversity and poor health, before and after taking account of parenting

n=3343 (unweighted). The graphs display odds ratios with 95% confidence intervals before (stage 1) and after (stage 2) parenting variables are added to the model. When the dashed line falls below the solid line this indicates a reduction in the strength of association between family adversity and child health when parenting variables are added to the model suggesting that differences in parenting across families with different levels of adversity explain some of the inequalities in child health outcomes. Stage 1 adjusted for child gender, age in months at sweep 5, birth order and number of children in household at sweep 5. Stage 2 further adjusted for all parenting measures.

5.4 Does parenting account for inequalities in child health behaviours?

Figure 5-B shows associations between family adversity and child health behaviours, before (stage 1) and after (stage 2) adjusting for all parenting measures, as in the previous section. Physical activity has not been included in this table, as it did not show a clear trend according to the level of family adversity (see Chapter 2).

The graphs present a similar picture to the effect of parenting measures in the models of child health. There is a reduction in the strength of the association between family adversity and health behaviours at stage 2 when parenting is controlled for in the models. Again, the greatest reductions in the odds of poor health behaviours are seen at higher levels of family adversity. This implies that parenting may help to explain some of the inequalities in child health behaviours linked to family adversity, particularly among those families experiencing higher levels of adversity.

As in the previous section, to estimate how much of the relationship between family adversity and health behaviour inequalities is explained by differences in parenting, children with any level of family adversity greater than zero were compared with those who had no adversity. Parenting explained some, but not all of the effect of family adversity on health behaviour (see Technical Appendix for more information). The effect is estimated at:

• 33% for screen time;
• 54% for fruit and vegetable consumption; and
• 32% for unhealthy snacking.

To summarise, the findings suggest that parenting may explain some, but not all of the inequalities in child health behaviours that are linked to family adversity.

Figure 5-B Associations between family adversity and poor child health behaviours, before and after taking account of parenting

n=3343 (unweighted). Stage 1 adjusted for child gender, age in months at sweep 5, birth order and number of children in household at sweep 5. Stage 2 further adjusted for all parenting measures.