Socioeconomic position and trajectories of growth and adiposity across childhood

1
Socioeconomic position and trajectories of growth
and adiposity across childhood
Laura Howe, Social Medicine, University of Bristol
2
Inequalities in growth and adiposity

• Socioeconomic disadvantage is associated with
• Shorter stature
• Greater adiposity
• Little is known about when the socioeconomic
  differentials appear

3
Objectives

• Model trajectories of growth across childhood
• Height
• Adiposity
• Explore how these growth curves differ by
  socioeconomic position (SEP)

4
Research questions

• Do the overall patterns of growth and adiposity
  change differ by SEP?
• What is the magnitude of socioeconomic
  differentials at different ages?
• Are there gender differences in the patterns of
  socioeconomic differentials?

5
Methods
6
Height and weight data in ALSPAC

• Birth weights/lengths (obstetric records
  ALSPAC staff visiting hospitals)
• Routine child health records(health visitors, up
  to 4 measures per child)
• Research clinics (CiF and Focus)
• Questionnaires (mother-reports, up to 4 per
  questionnaire)

7
Data birth to ten years
Height Weight
Boys N 7194 Median 5 IQR 2-8 N 7248 Median 5 IQR 3-8
Girls N 6733 Median 5 IQR 2-8 N 6781 Median 5 IQR 3-8
8
Adiposity measures

• Ponderal index (kg/m3) for 0-2 years
• BMI (kg/m2) for 2-10 years

9
Analysis 1

1. Fractional polynomials to estimate overall shape
2. Identification of rough spline points
3. Optimisation of spline points

10
Analysis 2

• Random effects multi-level models in MLwiN
• Individual-level residuals provide an
  individuals deviation from the
  average(intercept and slope for each period)
• Interactions with SEP

11
Results

• Overall shapes of curves did not differ by SEP
• Spline points were therefore estimated for all
  SEP groups combined

12
Height

• Spline points
• Boys
• 3, 10, 29 months
• Girls
• 2, 11, 32 months

13
(No Transcript)
14
Boys mean heights by maternal education
Age (months) Less than O-Level O-Level A-Level Degree or above
12 76.18 76.41 76.43 76.48
60 108.76 109.36 109.40 109.76
120 140.56 141.40 141.36 141.97
By 10 years 1.4cm difference between highest
and lowest groups
15
(No Transcript)
16
(No Transcript)
17
Girls mean heights by maternal education
Age (months) Less than O-Level O-Level A-Level Degree or above
12 75.24 75.45 75.31 75.49
60 107.88 108.47 108.50 109.10
120 139.68 140.18 140.69 141.40
By 10 years 1.7cm difference between highest
and lowest groups
18
(No Transcript)
19
Ponderal Index

• Spline points
• Boys
• 2 months
• Girls
• 1, 4 months

20
(No Transcript)
21
Boys ponderal index by maternal education
Age (months) Less than O-Level O-Level A-Level Degree or above
6 25.33 25.49 25.48 25.36
12 23.17 23.29 23.27 23.18
18 21.01 21.10 21.06 21.00
24 18.85 18.91 18.85 18.82
22
(No Transcript)
23
Girls ponderal index by maternal education
Age (months) Less than O-Level O-Level A-Level Degree or above
6 25.68 25.74 25.73 25.48
12 23.34 23.35 23.42 23.17
18 21.00 20.95 21.12 20.86
24 18.66 18.56 18.82 18.55
24
BMI

• Spline points
• Boys
• 56, 67, 73, 79, 105 months
• Girls
• 60, 65, 75, 81, 103 months

25
(No Transcript)
26
(No Transcript)
27
Conclusions Height

• Socioeconomic differentials in height are largely
  established by birth length and growth between 1
  and 3 years
• Socioeconomic differentials in height are greater
  for girls than boys
• By 6 years-old there is a clear height gradient
  across all four categories of maternal education
  for girls. Intermediate education groups remain
  similar for boys

28
Implications Height

• Pre-natal and early life factors are the most
  important determinant of socioeconomic
  differentials in height

29
Conclusions Ponderal Index

• There is no clear socioeconomic gradient in
  ponderal index in the first 2 years of life

30
Conclusions BMI

• From 4 years onwards, boys whose mothers are
  educated to degree-level have lower BMI
• There is no BMI gradient across other categories
  of maternal education
• From 4 years onwards, girls whose mothers are
  educated to degree-level have lower BMI
• Inequalities are wider in girls, and there is a
  clearer gradient across all categories of
  maternal education

31
Implications Adiposity

• Socioeconomic inequalities in adiposity are
  established early in childhood
• Interventions could be aimed at pre-school and
  early-school children

32
Remaining questions

• WHY are inequalities wider in girls?
• Cohort effects?

33
Next steps

• Variables for ALSPAC archive will be available
  for all to use
• Associations of growth with other childhood
  outcomes/exposures
• Modelling growth later into childhood to include
  puberty

34
Acknowledgements

• Kate Tilling, Debbie Lawlor, Bruna Galobardes,
  Paul Clarke, Fiona Steele