Strand 3: The Antecedents of Physical Health in Adolescents

Leader: Debbie Lawlor Strand Publications

The main focus of this strand will be on examining obesity, metabolic, cardiovascular and respiratory health in the developing child through to adolescence.

Three periods in early life may be particularly important for the development of obesity and its associated morbidity and mortality: the perinatal period; the period of adiposity rebound; and puberty/adolescence (Dietz, 1994, 1997). In both males and females adolescence is associated with an increase in central obesity and its related metabolic and vascular abnormalities (Ferreira, Twisk, et al., 2005; Gillman, 2004) and it may also be a critical period for the development of atherosclerosis (McGill, Jr., McMahan, et al., 2000). Obesity, metabolic and cardiovascular disease in adulthood are more prevalent amongst those from lower SEP groups in childhood irrespective of their adult SEP (Parsons, Power et al., 1999), but the pathways linking childhood SEP to adult disease is unclear.

Lung function increases in healthy children reaching a maximum in late adolescence, stays at this level and then begins to decline with old age (Strachan & Sheikh, 2004). Normal lung growth and development during the intrauterine period and childhood are important for reaching maximum attainable lung function in adolescence (Dezateux & Stocks, 1997; Strachan, 1997). A number of early life factors, including maternal smoking, low birth weight and childhood infections, all of which may reflect poor early life SEP, are associated with reduced maximum lung function in 9 adolescence (Strachan & Sheikh, 2004), and recent evidence shows that low childhood SEP is related to poor lung- function into old age, independently of smoking history and other potential confounding or explanatory factors (Lawlor, Ebrahim & Davey Smith, 2004).

Using the repeated measurements of the indicators of physical health that we are interested in (see data appendix) and repeated measurements of domains of SEP our analyses will use the statistical methods described in the methodological strand, including the use of multivariate multiple imputation methods, structural equation models and multi-level modelling. We are keen to exploit the potential of genetic variation, where good candidate genes exist to allow this, to deal with problems of endogeneity and reporting bias. The project team have experience of using genetic variants in this way (Davey Smith & Ebrahim, 2003; Timpson, Lawlor et al., 2005).

We will focus on three projects:

3.1 When in the life course do socioeconomic inequalities in physical health emerge?

Researchers: Lawlor, Davey Smith, Propper, Lynch, McGill University

It is likely that socioeconomic inequalities in different aspects of physical health emerge at different stages in the life course, either because differences in health behaviours emerge at different stages of the life course or because different life course models are related with different outcomes (Ben-Shlomo & Kuh, 2002). We will use the repeated measures of SEP and health related outcomes to examine socioeconomic differentials in obesity, metabolic, vascular and respiratory outcomes at all ages from birth and infancy through to age 16. We will compare the magnitude of these differentials at different ages and thus document socioeconomic inequalities for outcomes as children age, including whether there are different patterns of socioeconomic differentials through childhood for different health outcomes.

(Back to top)

3.2 Determining the biological and behavioural pathways that mediate socioeconomic influences on physical health

Researchers: Lawlor, Davey Smith, Ness, John Lynch

There is little evidence how socioeconomic, biological and behavioural risk factors interrelate to increase the risk of poor health in adolescence. Most studies to date measure exposures and mediators at one time point, but biological, socioeconomic and behavioural exposures in the individual and the family will not be static over time. Understanding how changes in SEP, behavioural and biological mediators in early life influence physical health in adolescence will provide important insights into the mechanisms of socioeconomic inequalities and how to reduce these. We will use the rich data within ALSPAC to test different pathways between early life SEP and adolescent physical health using multilevel models and structural equation models. Our aim will be to examine the magnitudes of different direct and indirect effects along these pathways and in this way identify the main factors that mediate the relationship between early life SEP and later health.

(Back to top)

3.3 Determining the impact of peer groups, school-based factors and neighbourhoods on childhood and adolescent obesity

Researchers: Propper, Burgess, Lawlor, Lynch

10 Randomised controlled trials have examined the effectiveness of peer education and school based interventions on childhood obesity, but most studies have been of poor quality and few have been able to demonstrate long-term effects (Summerbell et al., 2005). In addition, there is relatively little on the influence of peer groups and neighbourhood effects on obesity in children and adolescents. Our aim is to examine the impact of peer groups, schools and neighbourhood on childhood and adolescent obesity. We will build on the research and approaches outlined in Strands 4 and 5, and match data on local neighbourhood characteristics (e.g. amount of green space, numbers of exercise/sports facilities and self–reported views of the neighbourhood by the mothers and children) to ALSPAC participants. We will compare the influences of peers and school based characteristics on obesity outcomes when in primary school to those when in secondary school and attempt to identify causal effects from observations of changes in peers and schools.

Finally we will combine evidence from projects 1-3 to provide a detailed picture of the social determinants of physical health in adolescence. Comparing across the other research in this programme, we will be able to highlight the similarities and the differences in the antecedent pathways leading to poor physical health, mental health, risky health behaviours, poor cognitive development and behavioural problems in childhood and adolescence.

(Back to top)